SYMPOSIUM G

11:00 AM *G1-S2.1 (invited)
Photovoltaics from "Dirty" Silicon. (#1440) Eicke Weber, Fraunhofer-Institut für Solare Energiesysteme ISE, Germany.

Abstract unavailable as yet.

11:30 AM *G1-S2.2 (invited)
Growth of High-Quality Si Multicrystals with Same Grain Orientation and Large Grains Using Dendritic Casting Method and Formation Mechanism of Dendrite Crystals with Parallel Twins. (#220) Kazuo Nakajima¹, Kozo Fujiwara¹, Noritaka Usami¹, Satoshi Okamoto²; ¹Institute for Materials Research, Tohoku University, Japan ; ²SHARP Corporation, Japan.

Si multicrystals have many grains with different orientations and different sizes. If we can obtain Si multicrystals that have grains with a same orientation and proper sizes and grain boundaries are electrically inactive, such as Σ3, the conversion efficiency of the Si multicrystals should be very close to that of Si single crystals. This concept has not been tried yet for crystal growth of semiconductor multicrystals because there are no breakthrough ideas to obtain such high-quality Si multicrystals by a practical method. We have recently developed a reliable method named as "the dendritic casting method", for both purification and structure control to obtain high-quality Si multicrystals with the same grain orientation and large grain sizes. This method is on the basis of using dendrite crystals which grow along the bottom of the crucible at the initial stage of the growth and which can be used to control grain orientation and grain size. Using this method, we have obtained Si multicrystals with same grain orientation and large grain sizes, whose quality is very close to that of Si single crystals. In this paper, first we found a new formation mechanism of dendrite crystals with parallel twins using our newly developed "in situ observation system" for the observation of growing interface of Si multicrystals in Si melts at a temperature higher than 1400?C. We found the new formation mechanism of parallel twins that are prerequisite for the growth of Si dendrite crystals. In this mechanism, at first a twin boundary at an {111} facet plane of a step on the growth interface is formed, and then another parallel twin to the first one is formed. Subsequently, a faceted dendrite crystal grows parallel to the {111} facet plane of the step on the growth interface. Then, we found that dendrite crystals grow along the bottom of the crucible at the initial stage of the growth. By application of this mechanism, we have developed the dendritic casting method to control grain orientation and grain size to constantly obtain textured high-quality Si multicrystals. At first, we grow dendrite crystals with the <112> or <110> direction along the bottom of the crucible. Then, we grow Si multicrystals on the upper surface of the dendrite crystals. Textured high-quality Si multicrystal ingots were fabricated, which were grown on the (112) upper surface of dendrite crystals. The diameters of the ingots are 15 cm? and 30 cm?. The grain size is quite large and larger than 3 cm. The Si multicrystal ingots have the longer lifetime more than 35 ?s without any hydrogen passivation and gettering. The conversion efficiency of solar cells prepared by Si multicrystals with the same orientation is found to be as high (>17%) as that of solar cells prepared by Si single crystals using the same processes. The wafer size used to obtain the best conversion efficiency of 17.1% was 10 cm x 10 cm. The new formation mechanism of dendrite crystals with parallel twins was found using our newly developed in situ observation system. By application of this mechanism, the dendritic casting method was developed to control grain orientation and grain size to constantly obtain textured high-quality Si. The conversion efficiency (17.1%) of the multi-crystalline solar cells with the same grain orientation is found to be as high as that (17.9%) of the single-crystalline solar cells.

12:00 PM *G1-S2.3 (invited)
Impurities in Solar-Grade Silicon. (#152) Daniel Harold Macdonald, Department of Engineering, Australian National University, Australia.

As the photovoltaic industry grows over the coming years, new forms of low-cost, 'solar-grade' silicon feedstock will necessarily be introduced. These new feedstock sources will almost certainly contain greater quantities of impurities than the semiconductor-grade off-cuts and scraps that have mostly been used for solar cells until now. Of course, using such 'dirty' material carries some risk, in that it may have a detrimental impact on cell efficiency through increased recombination. It will be very important for the photovoltaic industry to minimise, or preferably eliminate, any such performance reductions. The purpose of this paper, therefore, is to review our current understanding of the impact of specific impurities in solar-grade silicon, and to build towards specifying tolerance levels for the most important impurities. The impurities to be discussed fall into three categories. The first and probably most important type is the metallic impurities, which can exist in point-like form or as precipitates. Point-like metals tend to have the greatest recombination activity, whereas metal precipitates, although also strongly recombining, yield a lower recombination activity 'per atom'. Some fast-diffusing point-like metals (such as Fe, Ni and Cu) are susceptible to impurity gettering during cell fabrication, which can greatly reduce their impact. The amount of a certain metal which is precipitated depends on the crystal growth technique, the diffusivity and solubility of the metal, the presence of precipitation sites such as grain boundaries, and subsequent wafer processing. This raises the prospect of 'defect engineering', whereby the most favourable chemical form of a metal impurity is deliberately induced through careful processing. The second type of unwanted impurities are excess dopants, such as B, P, Al and Ga. These are especially difficult to remove during feedstock purification, and may cause the resulting wafers to be quite strongly compensated. The potential impacts of compensation on device performance will be discussed. This is an issue which has been largely overlooked until recently. The final group of impurities to consider are the relatively benign, ubiquitous, non-metallic, non-dopants such as C, O and N. These are highly soluble in silicon, and tend to only create problems when supersaturation during crystal growth causes them to precipitate. These precipitates can act as recombination centres, and sometimes as conductive shunting paths across a p-n junction. Point-like oxygen also forms a well-known recombination centre with B. Considering these aspects allows us to begin to specify approximate tolerance levels for some of the most important impurities in solar-grade silicon.

LUNCH 12:30 PM - 2:00 PM

2:00 PM *G1-S3.1 (invited)
Compatibility of the RGS Silicon Wafer with Solar Cell Processing. (#759) Axel Schonecker¹, Astrid Gutjahr², Antonius R. Burgers², Sven Seren³, Giso Hahn³; ¹RGS Development B.V., Oudkarspel, Netherlands ; ²ECN Solar, Netherlands ; ³University of Konstanz, Germany.

The ribbon-growth-on substrate (RGS) silicon wafer technology is a unique casting technology for the next generation of silicon wafer manufacturing for photovoltaic application. Compared to today's cut wafer technology, the silicon yield in the process is increased from about 40% to more than 90% due to the direct casting of the wafers. Production speed per machine is in the order of 50 MWp per year, which is an order of magnitude higher than cut multi-crystalline silicon wafer manufacturing lines or other silicon ribbon technologies such as EFG or String Ribbon. This will result in a strongly reduced investment for the wafer manufacturing equipment. RGS is thus a key technology for enabling the high growth rates of the PV sector in the future. However, due to the unique wafer manufacturing process, the characteristics of RGS wafers differ from the known solar silicon wafers such as cast multi-crystalline silicon wafers. The crystallisation from the substrate with a freely crystallising top surface, results in a wafer with a perfectly defined bottom surface and a rough top surface. Further investigations are needed in order to use thick film metal printing on RGS wafers. From an electronic property point of view, the main limiting factor was the high oxygen concentration in the wafer. This caused long wafer processing times, both for hydrogen passivation and to prevent the formation of oxygen related recombination centres. By reducing the oxygen concentration in the wafer, the compatibility of the RGS wafer with an industrial solar cell process as also the overall wafer quality could be enhanced. As result the efficiency of RGS wafer based solar cells could be increased from less than 10% to 13% using a basic screen printing process.

2:30 PM G1-S3.2
Effect of Intentionally Fe-doped Feedstock on p- and n-type Multicrystalline Silicon Solar Cell Efficiency. (#647) Gianluca Coletti¹, Rannveig Kvande², Valentin Mihailetchi¹, Bart Geerligs¹, Lars Arnberg², Eivind Ovrelid³; ¹ECN Solar Energy, Petten, Netherlands ; ²Norwegian University of Science and Technology (NTNU), Norway ; ³SINTEF Materials and Chemistry, Norway.

In this work we study the impact of Fe on p-type and n-type multicrystalline silicon grown by directional solidification, with particular emphasis on the properties of the solar cells. The Fe has been added to the silicon feedstock before melting and the subsequent solidification of the ingot. This allows us to study the impact of Fe at different stages of the manufacturing chain of the solar cells: from the crystallisation of the ingot to the completion of the solar cell process. The amount of Fe introduced in the melt (53 ppm wt) has been chosen targeting typical concentration of total Fe in commercial multicrystalline silicon ingots (reported to be 1E13-1E15 cm-3 ) assuming an equilibrium segregation coefficient (6E-6). A p-type and n-type solar cell process was applied to both the Fe contaminated wafers and reference wafers produced from non-contaminated silicon feedstock. The p-type solar cell process is the state of the art industrial Al-BSF SiNx firing through. The n-type cell process comprises of a boron front emitter and phosphorous BSF, both passivated by SiNx and firing through metallisation with open rear side grid. Neighbouring wafers were used to characterise the minority carrier lifetime after P-gettering and hydrogenation steps. The initial lifetime in p- and n-type Fe contaminated wafers is about 1-2 &mus and 6-20 &mus, respectively. The differences in lifetime between the p- and n-type Fe ingots are explained with SRH theory for interstitial Fe defects. The improvement of lifetime after gettering and hydrogenation of the p-type Fe ingot is up to 50 times and of the n-type Fe ingot 5-6 times. After P-gettering and hydrogenation the lifetimes in the p- and n-type Fe-doped ingots approach each other. The most harmful impurities in n-type wafers are mainly substitutional. The reduced gettering effectiveness in n-type compared to p-type might be explained by the presence of a 'ceiling' due to non-getterable defects, such as substitutional impurities. In order to quantify the gettering effectiveness, the interstitial Fe (Fei) concentration was derived from lifetime measurements before and after FeB pair dissociation for the p-type ingots. The initial [Fei] in the p-type Fe ingot is about 1E13 cm-3 in the middle of the ingot and decreases to about 1E11 cm-3 after gettering. The addition of Fe in the p- and n-type ingots causes a decrease of the solar cell diffusion length in the top of the ingots due to segregation. Segregation is, however, not the only effect. An increased crystal defect concentration in the top and bottom of the ingots contributes considerably to solar cell performance degradation. The high defect concentration in the top of the ingots may be related to the higher Fe concentration that affects the crystallisation. The high defect concentration in the bottom region may be due to transient nucleation and growth phenomena, possibly also due to Fe, during the early solidification phase.

2:45 PM G1-S3.3
Photo- and Cathodoluminescence Characterisation of Multicrystalline Silicon for Solar Cells. (#420) Martin C. Schubert¹, Manuel Romero², Sergei Ostapenko³, Paul Gundel¹, Wilhelm Warta¹, Stefan W. Glunz¹; ¹Fraunhofer-Institut für Solare Energiesysteme ISE, Germany ; ²National Renewable Energy Laboratory, USA ; ³University of South Florida, USA.

The goal of this work is to provide an in-depth analysis of the material quality of solar grade multicrystalline silicon based on photo- and cathodoluminescence measurements and to examine the potential of these measurement techniques in the material analysis. In recent publications the band-to-band-luminescence was used to determine the minority carrier lifetime with high spatial resolution. But also defect photoluminescence at longer wavelengths is promising to yield significant information, since wavelength bands in the range 1.2-1.7 um have been related to crystal defects, stress, oxygen precipitates and metallic impurities. In particular four prominent lines, entitled D1-D4 bands, have been revealed in this spectral range at low temperatures. Furthermore Tajima, Kittler and Ostapenko observed a defect band centred at 800 meV in the photoluminescence spectrum of multicrystalline silicon, which is detectable at room temperature and whose energy position matches with the D1-band. For this article the spatially and spectrally resolved photoluminescence spectra of several multicrystalline samples were measured. In order to gain more information about the physical nature of the involved defects temperature and injection dependent photoluminescence measurement were conducted. In addition the spatial correlation between the intensity of defect bands and the density of recombination inactive trapping centres was examined. The trap density was measured with Carrier Density Imaging (CDI) / Infrared Lifetime Mapping (ILM). These traps have been attributed to crystal defects and impurities. Recent results give rise to the assumption that oxygen constitutes a dominant contribution to the trap density. Our photoluminescence examinations of multicrystalline silicon for solar cells can be summarised as follows: A defect band at 880 meV was observed. Temperature dependent measurements of this band revealed an energy difference to the band edge of 50 meV respectively 79 meV depending on the theoretical model for this defect. The energy of this band allows the assumption that it is connected to the D2 band. High trap densities can be found in regions with high intensity of the 838 meV band at room temperature. From the spatially resolved cathodoluminescence measurements the following information was extracted: Regions with high trap densities show complex cathodoluminescence spectra at 20 K, which can be clearly distinguished from regions with low trap densities. The trap density shows a high spatial correlation with the intensity of a defect band at 910-930 meV at 80 K. In the conclusion the nature of the defect which is responsible for the 838 meV photoluminescence band is illustrated in detail and the importance of oxygen in the formation of trap centres is discussed. The results of this article highlight the potential of luminescence spectroscopy in the material analysis of multicrystalline silicon.

3:00 PM G1-S3.4
Crystallisation of Multi-Crystalline Silicon Ingots Using Purified and Compensated Metallurgical Grade Silicon. (#902) Roland Einhaus¹, Jed Kraiem¹, Francois Lissalde²; ¹ApollonSolar, France ; ²Cyberstar, France.

Introduction: The presented work is part of the French PHOTOSIL project, which deals with the purification and crystallisation of metallurgical grade Silicon, in order to arrive at multi-crystalline Silicon ingots of solar grade quality. A new, innovative crystallisation process and equipment have been developed and realised by CYBERSTAR and APOLLONSOLAR, which both are specially adapted to Silicon of metallurgical origin. Important key features, such as two independent inductive heating elements and the use of a thermally anisotropic crucible allow for a very uniform heat distribution and excellent temperature control during crystallization. With adapted processes for the growth of ingots and the realization of solar cells metallurgical Silicon represents an alternative, widely available and low cost Silicon feedstock source for photovoltaics. Experimental work: The thermal configuration of the new furnaces allows for a perfect lateral and vertical temperature control during crystallization, so that high temperature gradients can be obtained, which are especially beneficial for the segregation of remaining impurities in lower quality silicon. An important point related to the use of purified metallurgical grade Silicon is the presence of both, n-type and p-type dopant atoms (Phosphorous and Boron), even after the purification that often does not arrive at their complete removal. The remaining dopant atoms lead to compensation effects while arriving at acceptable resistivity levels of around 0.5 Ohmcm and p-type. However, due to the more effective segregation of Phosphorous often the final upper part of the ingots show n-type polarity, whereas the lower part of the ingots remain of the desired p-type polarity. Both, compensation effects and change of polarity are investigated in detail. On the other hand, on recently grown multi-crystalline Silicon ingots from purified metallurgical grade Silicon regions with a high compensation ratio could be identified. These regions are characterized by a relatively high minority carrier diffusion length compared to non-compensated regions or regions with a low compensation ratio. This indicates that compensated doping might be a way to improve the electrical performance of multi-crystalline Silicon from metallurgical origin. Results: A temperature gradient in the order of 20 K/cm is regularly obtained during the crystallisation process, assuring a stable, planar interface, which is important for an efficient segregation. An ingot from purified metallurgical Silicon has allowed to obtain top efficiencies of 14% with a resistivity of 0.5Ohmcm. Detailed characterisation of this ingot has shown that dopant compensation has increased the minority carrier diffusion length by 50% and thus the solar cell efficiency. However this ingot turned n-type at around 75% of its height due to the different segregation coefficient of the major dopants, Boron and Phosphorous.

3:15 PM G1-S3.5
Modelling of Compensated Silicon Solar Cells. (#478) Andres Cuevas¹, Daniel Harold Macdonald¹, Bart Geerligs²; ¹Department of Engineering, Australian National University, Australia ; ²Energy Research Centre of the Netherlands: ECN, Netherlands.

The introduction of lower grade silicon materials produced via metallurgical methods to the field of solar cells is creating new scientific challenges. Silicon wafers are commonly selected by the manufacturers based on their resistivity. Today the norm is to use p-type wafers with a resistivity in the proximity of 1?cm, but UMG-Si usually has a much lower resistivity, down to about 0.2?cm, due to a high concentration of both p-type and n-type dopants. Such a resistivity implies 4x1017cm-3 boron and 2.4x1017cm-3 phosphorus atoms, compared to just 1017cm-3 boron atoms in non-compensated material. The higher concentration of foreign impurities can have a negative impact on the carrier lifetime, and indeed our preliminary experiments on compensated silicon have shown that carrier lifetimes are lower, approximately a factor of two, compared to normal silicon. Dopant compensation can also be expected to affect the mobility of electrons and holes. Most empirical models used to calculate the mobility as a function of doping predict that the increased number of scattering centres present in compensated silicon (that is, the ionised boron and phosphorus atoms) will produce a reduction of the mobility. A lower carrier mobility can have significant impacts on solar cell performance. In this paper we present modelling implications of compensation on the performance of silicon solar cells. We have used a simple, but comprehensive modelling tool, QSS-Model, to study the impact of compensation on effective lifetimes, device current, voltage and efficiency. An obvious effect of a lower mobility is a shorter carrier diffusion length, with adverse consequences on the carrier collection at the pn junction, resulting in a lower output current. There can be other, completely unexpected, effects of a lower mobility. Our modelling indicates that it can result in an increase of the output voltage, in cases where the device is dominated by recombination in the bulk of the wafer, which is the most likely scenario for "dirty" silicon materials. This increase in Voc (up to 20mV in some cases) offsets to a large extent the reduction in Jsc. Another interesting aspect is that surface recombination can have a greater impact on a compensated wafer than on a non-compensated wafer of the same resistivity. Because of this, the interpretation of effective lifetime measurements need to take into account whether the material is compensated or not. Although it is highly likely that our effective lifetime experiments are affected by surface recombination, our calculations indicate that the different impact of surface recombination is not sufficient to explain the observed differences between compensated and non-compensated materials and that this difference is largely due to a reduction of the bulk carrier lifetime. These modeling results are, nevertheless, critically dependent on the assumed mobility models and further testing of alternative models will be performed to clarify the matter.

AFTERNOON BREAK 3:30 PM - 4:00 PM

4:00 PM *G1-S4.1 (invited)
Selective Emitters Made by LCP for High-Efficiency Silicon Solar Cells. (#1135) Daniel Kray, Sybille Hopman, Andreas Fell, Christoph Fleischmann, Kuno Mayer, Stefan W. Glunz; Micro-patterning and Laser-chemical technologies, Department Silicon Solar Cells - Development and Characterization, Fraunhofer-Institut für Solare Energiesysteme ISE, Germany.

Selective emitters allow for the decoupling of the technological requirements of contacted and uncontacted emitter regions. While doping concentrations above 10^(19) cm^(-3) are necessary to allow for low contact resistance [1], shallower emitters are beneficial for the illuminated areas of solar cell. This is due to the strong dependence of the emitter saturation current on the surface dopant concentration and the sheet resistance [2]. Therefore, the most efficient silicon solar cells use selective emitters which are fabricated using photolithography and two subsequent diffusions [3]. Laser Chemical Processing (LCP) [4] is investigated as a very elegant and simple method to generate selective emitters without the need of photolithography or a second diffusion. After the shallow diffusion and passivation LCP can be used to open locally the passivation layer, melt or partially evaporate the surface and diffuse e.g. phosphorous in the molten surface layer. These steps can be performed in parallel due to the guidance of the laser light inside a dopant containing liquid jet. Since no second diffusion step is needed, this method is especially interesting for multicrystalline silicon. We will present investigations of LCP-doped grooves in p- and n-type silicon with QSSPC and SunsVoc measurements for determination of implied Voc depending on laser and jet parameters. These are supported by defect etch experiments on cleaved samples as well as numerical simulations to determine the thermal load of the different LCP processes. We could already show [5] that the developed Matlab(R) code is performing well by comparison with experimental groove shapes for dry laser processes. The necessary model extensions for the liquid-jet-guided laser are discussed. The main focus of the paper is the generation of process-induced defects and their characterization. We present SEM investigations of cross sections of LCP grooves after different thermal treatments and defect etches to show the influence of the etchant choice on the defect visualization. The defect etch results are correlated to the SunsVoc and QSSPC measurements which quantify the effects on the electrical performance of the device. 1. Schroder, D.K., Semiconductor Material and Device Characterization. 1990, New York: John Wiley & Sons. 599. 2. King, R.R., R.A. Sinton, and R.M. Swanson, Studies of diffused phosphorus emitters: saturation current, surface recombination velocity, and quantum efficiency. IEEE Transactions on Electron Devices, 1990. 37(2): p. 365-71. 3. Zhao, J., A. Wang, and M.A. Green, 24.5% Efficiency silicon PERT cells on MCZ substrates and 24.7% efficiency PERL cells on FZ substrates. Progress in Photovoltaics: Research and Applications, 1999. 7(6): p. 471-4. 4. Kray, D., et al. Progress in laser chemical processing (LCP) for innovative solar cell microstructuring and wafering applications. in Proceedings of the 22nd European Photovoltaic Solar Energy Conference 2007. Milan, Italy. 5. Fell, A., et al. Transient 3d-simulation of laser-induced ablation of silicon. in Proceedings of the 22nd European Photovoltaic Solar Energy Conference 2007. Milan, Italy.

4:30 PM G1-S4.2
Fabrication of Textured Silicon Solar Cell Using Microlens as Anti-Reflection Coating Layer. (#256) Hye Jin Nam, Soon Cheol Jeong, Juyeon Chang, Jae-Eok Han, Hyeon Ju Ahn, Duk-Young Jung, Young Kuk Kim, Kyu Min Han, Jun Sin Yi; Department of Chemistry, Sungkyunkwan University, Republic of Korea.

This paper presents facile, reproducible soft-lithographic technique for fabricating textured crystalline silicon solar cell with hexagonally close-packed hemispherical arrays using ordered polymeric microspheres. Close-packed monolayers of polymer microspheres were deposited on flat cleaned glass substrate using spin-casting technique. The relief structure of highly ordered microspheres successfully generated their negative replica of elastomers. Hemispherical microlens arrays were molded by dispensing and curing liquid ultraviolet-curable photopolymers into the negative replica on crystalline silicon solar cell module. Sub-micrometer scaled microlens with uniform, cleaned surface morphology were easily reproduced by using the prepared molds, without multi-step engineering processes. The microlens-coated crystalline silicon solar cell showed additional increment of 1% in solar cell efficiency, compared to that of SiNx ARC-coated one, as well as remarkable decrease of reflectance. This technique opens the way to provide reliable route to fabricate embossed thin layers from nanometer to micrometer diameters by controlling particle sizes of polymer microspheres on the substrates up to centimeter-scale, which may improve the efficiency to receive a light irradiated on the solar cell module.

4:45 PM G1-S4.3
Fixed Negative Charge in Nanoscale Al₂O₃ and its Role in the Surface Passivation of c-Si. (#1441) Bram Hoex, Eindhoven University of Technology, Netherlands.

Fixed negative charge in nanoscale Al2O3 and its role in the surface passivation of c-Si

5:00 PM G1-S4.4
Back Enhanced Heterostructure with Interdigitated Contact - BEHIND - Solar Cell. (#1186) Mario Tucci¹, Luca Serenelli¹, Pierino Martufi¹, Enrico Salza¹, Giampiero de Cesare², Domenico Caputo², Matteo Ceccarelli², Simona De Iuliis³, Lambert J Geerligs³; ¹ENEA Research Center Casaccia, Roma, Italy ; ²Electronic Engineering Department, University of Rome 'Sapienza', Italy ; ³ECN, Netherlands.

The tendency towards shrinking wafer thickness is driving solar cell process temperature down. This is especially the case for multi-crystalline silicon; where wafer bowing and thermal stresses in the bulk material can be problematic. On the other hand, an effective back surface field, usually provided by high temperature processes, is an increasing necessity with decreasing substrate thickness. Both of these issues can be faced with amorphous/crystalline (a-Si/c-Si) silicon heterostructure, which can be deposited by Plasma Enhanced Chemical Vapour Deposition (PECVD) at temperature below 400?C. Such technology can increase the open-circuit voltage, even though the n-type a-Si window layer can still affect the short-circuit current of p-type c-Si based cell due to the amorphous silicon absorption in the blue region of the solar spectrum. To overcome this and to remove any shadowing effect from a top metal grid, we have developed a new cell design transferring the emitter and both contacts to the rear side of the device: the BEHIND (Back Enhanced Heterostructure with InterDigitated contact) solar cell. The purpose of this work is to develop further our low temperature very high efficiency BEHIND cell process, by improving front surface passivation, reducing absorption in front coating, and by improving contact and series resistance. Our device has been fabricated starting from a mono-crystalline p-type silicon substrate. After front side alkaline texturing and industrial cleaning, we have deposited a double layer stack of SiNx on the sunward side, acting as passivation and anti-reflection layer. Then, on the whole polished backside of the wafer, after an HF-dip, an intrinsic a-Si:H buffer layer has been deposited followed by a n-type doped a-Si:H one. Over these films a chromium silicide (CrSi) layer has been formed [1] and an ITO layer has been sputtered. By using an opportunely designed metallic mask, we have dry etched the ITO/CrSi/n-type a-Si:H to form the trench for the base contact. Subsequently, keeping the mask in the same position, a p-type a-Si:H film has been deposited and 2 mm of Al has been evaporated inside the trench. Finally, the mask has been rotated 180 degrees and 2 mm of Ag has been evaporated to contact the ITO region, creating the interdigitated shape with respect to the p-type contact. For n-type a-Si/p-type c-Si heterojunction surface passivation is still a challenge in order to obtain higher open-circuit voltage (>700 mV) [2]. Different approaches have been investigated, to improve the interface with the c-Si substrate [1]. Efficiency and carrier transport properties of a-Si/c-Si heterojunction solar cells are significantly affected by processing and plasma deposition conditions. While rear-junction, back contact solar cells are not new, employing the silicon heterostructure technology enables the use of very thin silicon wafers since low thermal budget is needed for cell fabrication. The metallic mask forms the interdigitated pattern on the rear side of the device. Several aspects of the proposed new structure were investigated. The dimensions of both p- and n- type amorphous contacts were carefully calculated to ensure complete collection of the diffusion current. The geometrical constraints for both contacts, the quality of the crystalline wafer needed to completely collect the diffusion current, and the effectiveness of the front surface passivation were considered and will be discussed in detail. We found that the mask alignment is not critical since the emitter is much wider than the base region. Preliminary results indicate that the intrinsic buffer layer plays a fundamental role in surface passivation and isolation between the two contacts. High open circuit voltage (695 mV) has been reached. [1] M. Tucci, G. de Cesare, J. Non-Cryst. Solids 338 (2004) 663. [2] H. Angermann, J. Rappich, K. v. Maydell, E. Conrad, I. Sieber, D. Schaffarzik and M. Schmidt, Proceeding of the 21th EPVSEC, Dresden (2006) 895.

5:15 PM G1-S4.5
Comparison of Photoconductance and Phototoluminescence Based Lifetime Measurement Techniques. (#623) Thomas Roth, Philipp Rosenits, Marc Rüdiger, Wilhelm Warta, Meinrad Spitz, Stefan Rein, Stefan W Glunz; Department Silicon Solar Cells - Development and Characterization, Fraunhofer-Institut für Solare Energiesysteme ISE, Germany.

At Fraunhofer ISE different techniques for measuring the area-integrated effective lifetime of excess carriers in silicon samples are used. Photoconductance based measurement techniques such as microwave-detected photoconductance decay (MW-PCD), quasi-steady-state photoconductance (QSS-PC) and transient photoconductance (TR-PC) are used as well as the quasi-steady-state photoluminescence (QSS-PL), where the emitted light of the radiative recombination of electron-hole-pairs is evaluated. The purpose of this paper is a comparison of the different techniques in order to achieve maximum accuracy for the determination of this important material parameter. For MW-PCD measurements, the decaying excess carrier density after a laser pulse is measured using microwave reflectance and fitted using an exponential function, yielding the effective lifetime. Such an evaluation of the decaying excess carrier density is also used for TR-PC measurements, where the excess carrier density is measured after a short flash pulse using an inductively coupled rf circuit. These transient techniques have in common that the optical parameters of the samples do not influence the measured effective lifetimes. However, for quasi-steady-state measurement techniques (like QSS-PC and QSS-PL), where the illumination intensity is varying very slowly compared to the lifetime of the excess carriers, the actual generation rate within the sample has to be known. In the case of a monochromatic light source, the generation rate can be calculated quite easily using a calibrated reference cell and the reflections of the sample. If the illumination of the sample is carried out using a flash lamp, optical simulations have to be performed in order to access the generation rate. Since all techniques measure the depth averaged excess carrier density care has to be taken when measuring samples with very low effective lifetimes, since the carrier distribution may not be homogeneous if the excess carriers are generated near the front surface of the sample. This carrier profile inhomogeneity can be minimized using an appropriate light source, which generates excess carriers throughout the whole bulk of the sample. In this oncoming contribution, these different aspects of lifetime measurements will be investigated in detail. Different sets of samples with varying lifetime, thickness and surface passivation will be analysed using the different measurement techniques. In addition, different kinds of illumination sources will be used, including light emitting diodes and flash lamps with several spectral filters. Guidelines will be presented how to handle different kinds of samples in order to measure the excess carrier lifetime with maximum reliability.

G1-S5.1
Microstructure of Cu(In,Ga)Se2 Thin Film Prepared by Selenization of Metallic and Selenide Precursors and its Application to Solar Cells. (#470) Min Sik Kim, Jong Chul Lee, Rawee Balaji Venkata Chalapathy, Jae Ho Yun, Byung Tae Ahn; Korea Advanced Institute of Science and Technology (KAIST), Republic of Korea.

Thin film solar cells based on Cu(In,Ga)Se₂ (CIGS) continue to be a leading candidate for thin film photovoltaic devices due to its bandgap, high absorption coefficient, long-term stable performance and potential for low-cost production. Many groups have reported within the past years on a variety of processes for preparing CIGS films. Up to date, the most successful technique for deposition of CIGS absorber layer for the highest efficiency solar cells has been based on Cu, In, Ga and Se co-evaporated process, so-called 3-stage process, achieving an efficiency of greater than 19 %. However 3-stage evaporation process is difficult to scale-up for large area, economic manufacture. The selenization process has been a promising method for a low cost and large scale production of high quality CIGS film. In conventional selenization process, the sputtered precursor with alloyed or stacked metallic Cu-In-Ga layers is deposited on the substrate and it is followed by selenization process in toxic H₂Se ambient gas. However The CIGS films had the poor microstructure when the metal layers were annealed in a Se vapor atmosphere. In this study, we prepared a precursor which is a mixture of metals and Cu_2-xSe selenide compound on Mo-coated soda-lime glass by sputtering. The precursors were selenized to form CIGS film using Se vapor in a vacuum evaporation system without H₂Se ambient gas. The microstructure of the CIGS films was investigated with various Cu contents and deposition temperatures. And the CIGS films were applied to the fabrication of solar cell with an Al/n-ZnO/i-ZnO/CdS/CIGS/Mo/glass structure and its photovoltaic properties were characterized. The details will be reported in the meeting.

G1-S5.2
Fabrication of Organic Semiconductor Thin Film of Melanin Biopolymer by Electrochemical Method. (#506) Hyeon Ju Ahn, Hye Jin Nam, Mi Jin Park, Juyeon Chang, Soon Cheol Jeong, Jea Eok Han, Duk-Young Jung; Department of Chemistry, Sungkyunkwan University, Republic of Korea.

This paper describes a reliable synthetic route to fabricate thin films of melanin, which is amorphous p-type semiconducting natural pigment, by two electrochemical methods of constant potential (CP) and cyclic voltammetry (CV), respectively. Thin film of melanoid pigments were formed on ITO glass by using 5,6-dihydroxyindole (DHI) precursor solution in both tris- and phosphate-buffer solutions. Thin film deposition using CP method showed faster deposition rate than those using CV. The weak interaction between melanin thin film and substrate in phosphate buffer solution allowed to produce a freestanding melanin thin film. As the deposition time increased from 5 minutes to 2 hours, the film thickness linearly increased from 100 nm to 1500 nm, proving effective control of film thickness depending on deposition time. Raman spectra showed broad two peaks at 1350 and 1550 cm-1, corresponding to combinational vibration modes by several functional groups of melanin. Thin film deposition using flexible PDMS mask generated patterned melanin thin films. The electrical resistivity of melanin films was 15 ?ocm, close to that of silicon, which provides potential applications in photovoltaics, optoelectronic devices and sensors.

G1-S5.3
Effect of P3HT:PCBM Concentration in Solvent on the Performance of Photovoltaic Cell. (#919) Woon-Hyuk Baek, Hyun Yang, Jung-Min Kim, Tae-Sik Yoon, C. J. Kang, Hyun Ho Lee, Yong-Sang Kim; Myongji University, Republic of Korea.

The bulk Heterojunction (BHJ) solar cells have become a very attractive research topic because of their low cost, low weight, simple fabrication process and flexibility. Although their photovoltaic (PV) performances were very poor in the early stage, recent studies are reporting remarkably improved efficiency of the BHJ solar cell. These researches include the study of an annealing time and temperature, molecular weight of electron donor material (P3HT) and ratio of electron donor (P3HT) and acceptor (PCBM), etc. Especially, as annealing has become a fixed procedure for BHJ solar cells, the evolution of the film morphology and device efficiency has been investigated after annealing at elevated temperatures. Through annealing, an optimal PV performance can be obtained determined by de-mix the P3HT and PCBM constitution and the self-assemble of the BJH materials. However, the influences of annealing time and temperature on PV parameters are not still clearly elucidated. In this study, it is found that the concentration of P3HT:PCBM (1:1 by weight) in solvent will affect the performance of the cell, even though the devices have same active layer thickness. To analyze the effect of the concentration, we investigated the effect of annealing time on the PV performance and different concentration (1, 2, 3 wt%) of P3HT:PCBM in chlorobenzene with device structure of ITO / PEDOT / P3HT:PCBM / Al. In addition, PV characteristics(Voc, Isc, FF, efficiency), P3HT:PCBM phase separation and morphological study of the active layer by Transmission Electron Microscope (TEM) are discussed.

G1-S5.5
Thermally Stimulated Current Measurements Applied to the Study of Proton-Irradiation Induced Defects in CdS Thin Films. (#53) Veta Ghenescu¹, Lucian Ion², Marian Ghenescu¹, Cezar Tazlaoanu², Rosemary Bazavan², Dan Bazavan², Stefan Antohe²; ¹Institute for Space Sciences, Romania ; ²University of Bucharest, Romania.

Thin films of AII-BVI compounds are potential candidates for the manufacturing of electronic and optoelectronic devices, especially solar cells. In this paper the effects of irradiation with protons on electrical properties of polycrystalline CdS thin films have been investigated. The samples were irradiated with 3 MeV protons, up to fluency of 10^13 protons/cm2. The defects induced by ionizing radiations have been studied by thermally stimulated current spectroscopy (TSC). Thermally stimulated current method is largely used for that purpose and different procedures were proposed. The electrical conduction of the samples, both before and after irradiation, is controlled by deep donor centers. Several electrically active defects were detected and their parameters (activation energy, density) were determined. The possible origin of those defects is discussed.

G1-S5.7
Efficient Hole-extraction in Bulk-Heterojunction Organic Photovoltaics Using Chemically Derivatized Carbon Nanotubes. (#899) Ross Hatton, Li Wei Tan, S. Ravi P. Silva; Department of Chemistry, University of Warwick, Coventry, United Kingdom.

For efficient hole-extraction in solution processed organic photovoltaics the transparent indium-tin oxide electrode is invariably pre-coated with a thin layer of the high work function conducting polymer blend poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT:PSS). We show that chemically derivatized single-wall and multi-wall carbon nanotube thin films are viable alternatives to PEDOT:PSS in efficient bulk heterojunction organic photovoltaics based on poly(3-hexylthiophene) (P3HT): [6,6]-phenyl-C61 butyric acid methyl ester (PCBM). Crucially, these thin films are deposited in the same way as PEDOT:PSS, with the important advantage of being processed from stable dispersions of low acidity. The functionality of these nano-structured films is demonstrated in relatively large area devices [~ 0.35 cm-2] tested under standard test conditions of 100 mWcm-2 AM1.5G. The effectiveness and hole-selectivity of these three dimensional electrode materials can be rationalised based on measurements of their electronic structure and morphology.

G1-S5.8
Synthesis and Characterization of CuInGa(S,Se)2 Thin Films for Photovoltaic Applications. (#916) Ahmed Ihlal¹, Khalid Bouabid¹, Mbarek Nya¹, Abdeslam Elfanaoui¹, Lhoussaine Elhamri¹, Xavier Portier², Gerard Nouet²; ¹LMER, University Ibn Zohr, Agadir, Morocco ; ²SIFCOM, France.

The efficiency of solar cells using the hetero-structure ZnO/CdS/CIGS/Mo/Glass has recently reached a high level of performance (20%), approaching those of Si based solar cells. In such structures, the absorber CIGS films were prepared using a high vacuum technique (evaporation. of elements). However, such techniques induce a high cost of processing. In this contribution we present our recent results on the synthesis conditions of CuInGaSe2 on different substrates: Mo, ITO and FTO by means of a non vacuum cost effective method: one step electrodeposition. The CuInGaSe2 films were prepared using an acidic solution formed by: CuSO4, In2(SO4)3, Se02 and GaCl3. The as-deposited films were then annealed, at 420?C, in 5% H2S/Ar atmosphere to prepare CuInGa(SSe)2 films. The XRD spectra of heat treated films under sulphur atmosphere show the well suitable chalcopyrite structure with displacement of the position of the principal peaks. Optical measurements show that our films exhibit a band gap value of about 1.3 eV approaching the ideal value for the conversion of sun light. SEM and EDX analyses were then performed to study the morphological and chemical composition out the prepared films. Finally, Solar cells using the hetero-structure ZnO/CdS/CuIn(SSe)2/ITO/glass were processed and investigated by J(V) measurements in dark and under light.

G1-S5.10
Low Resistivity Fluorine-Doped Tin Oxide Coating for Photovoltaic Solar Cell Substrate. (#1025) Chang-Yeoul Kim, Doh-Hyung Riu, Seung-Heon Huh; Nanomaterials, Korea Institute of Ceramic Engineering and Technology, Republic of Korea.

Transparent conducting oxide (TCO) is widely used for the application of flat panel display like liquid crystal displays and plasma display panel.[1] It is also applied in the field of touch panel, solar cell electrode, low-emisstivity glass, defrost window, anti-static material. In the field of display, tin-doped indium oxide (ITO) is used for its low electrical resitivity and patternability, but ITO thin film could not used for the application of high temperature stability and chemical durability. For example, ITO thin film is degraded in the environment of high temperature and chemicals. So, in the application for dye sensitized solar cell (DSC) and silicon thin film solar cell, fluorine-doped tin oxide (FTO) thin film are being used. Texture control or microstructure control of FTO thin film is very important for the application of solar cell, because there is a need to hold light wave by textured FTO thin film as much as possible. ?F-doped tin oxide (FTO) films were prepared on alumino-borosilicate glass (450?) by a sol-gel spray method. FTO films show good qualities: high optical transmittance of 80% in the visible range, low electric resistivity of 3.6 x10-4 ?-cm, hall mobility of 4.5 cm2/Vs, and carrier concentration of 34.5 x1020cm-3. Microstructures, electrical resistivies and optical properties of F-doped tin oxide films were investigated by changing F/Sn ratio and thicknesses.

G1-S5.11
Characteristics of Flexible Indium Tin Oxide Films Grown on PET Substrate Using Roll-To-Roll Sputtering for Flexible Organic Solar Cells. (#606) Han-Ki Kim, Kwang-Hyuk Choi, Sung-Woo Cho, Jae-Wook Kang, Dong Yu Kim, Seok Soon Kim; Department of Information and Nano Materials Engineering, Kumoh National Institute of Technology, Republic of Korea.

We reported on the electrical, optical, structural, mechanical and surface properties of flexible indium tin oxide (ITO) film grown on PET substrate by using a specially designed roll-to-roll (R2R) sputtering system as a function of DC power, Ar/O2 flow ratio, and rolling speed. It was observed that both the electrical and optical properties of ITO film on PET substrate were critically dependent on the Ar/O2 flow ratio. In addition, x-ray diffraction examination results showed that the structure of the ITO film on PET substrate was amorphous structure regardless of DC power, and Ar/O2 flow ratio due to a low substrate temperature which was maintained constant by main cooling drum. At optimized conditions, we could obtain ITO film with sheet resistance of 30~40-cm/sqare and transparency of 80~86 % even though it was prepared at room temperature. The bending test result of flexible ITO electrode exhibits R2R-grown ITO film had good flexibility due to ion-gun treated effect duing the sputtering. Furthermore, the flexible organic solar cell fabricated on R2R sputter grown ITO anode shows the power conversin efficiency of 2% under AM 1.5 illumination. This indicates that the R2R sputtering technology is a promising thin film technologies for flexible electrode in flexible organic solar cells.

G1-S5.12
Low Temperature Growth of Flexible Al Cathode on PET Subsrate Using Linear Facing Target Sputtering for Flexible Organic Solar Cells. (#609) Han-Ki Kim, Jin-A Jeong, Ho-Kyun Park; Department of Information and Nano Materials Engineering, Kumoh National Institute of Technology, Republic of Korea.

Linear Facing target sputtering (LFTS) system with a configuration of vertically parallel facing Al targets and a substrate holder perpendicular to the Al target plane has been designed for direct sputtering of Al cathode on organic layers at a low substrate temperature. The linear FTS system has a linear twin target gun with ladder-type magnet arrays for more effective and uniform confinement of high-density plasma. Due to more effective confinement of high-density plasma and perpendicularly placed substrate position, we were able to deposit Al cathode layer directly on an organic layer at the low temperature of ~50 degree without an additional substrate cooling system. In this work, we report on the characteristics of flexible Al cathode layer grown on PET substrate using linear FTS as a function of DC power, working pressure and substrate-target distance, respectively. Using XRD, FESEM, AFM, UV-visible spectrometer and XPS, the structural, optical, and surface properties of LFTS grown Al cathode were investigated. In addition, the bending test result obtained from optimized Al cathode exhibits the good flexibility which would be applicable to fleixlbe organic solar cells. Furthermore, the flexible organic solar cell with Al cathod fabricated by LFTS and conventional thermal evporator shows identical electrical properties and power conversion efficiency to those of flexible organic solar cell fabricated by thermal evporator. This indicates that the LFTS system with ladder type magnet array could be effective in direct Al cathode sputtering on organic photovoltaics.

G1-S5.13
Preparation and Characterization of Sputtered CuInSe₂ Thin Films Using a Single Target Composed of a Mixture CuSe and InSe Binary Selenides Powders. (#1093) Kyoo Ho Kim, Rachmat Adhi Wibowo; Department of Materials Science and Engineering, Yeungnam University, Republic of Korea.

In this study, the viability of growing a CuInSe2 thin film solar absorber by sputtering using a single target composed of CuSe and InSe powders was investigated. To accomplish this, qualitative and quantitative analyses of the structural and optoelectronic features of the sputtered films were carried out. It was found that the ability to obtain a sputtered film with a stoichiometric composition was greatly dependent on the substrate temperature and that the optimum conditions could be obtained by adjusting the sputtering radio frequency power. A single phase chalcopyrite CuInSe2 film with a nearly stoichiometric composition and strong (112), (220/204) and (312/116) reflections was successfully grown under the optimized growth condition. The CuInSe2 films exhibit an absorption coefficient of 104 cm-1 and an optical band gap of 1.0 eV. According to the Hall measurements, the sputtered CuInSe2 film showed p-type semiconductor characteristics, which is in good agreement with the non-stoichiometry (Δy>0) and non-molecularity (Δx<0) model.

G1-S5.14
Investigation of Cu₂ZnSnSe₄ Thin Films for Indium-Free Thin Film Photovoltaic Application. (#1097) Rachmat Adhi Wibowo, Kyoo Ho Kim; Department of Materials Science and Engineering, Yeungnam University, Republic of Korea.

In this study, sputtered Cu2ZnSnSe4 thin films were grown on Corning 1737 glass substrate by radio frequency magnetron sputtering from powder targets composed of CuSe, Cu2Se, ZnSe and SnSe binary chalcogenide powders. Various target compositions and substrate temperature were employed in order to attain a stoichiometric composition of sputtered films. From the films bulk composition analysis results, an appropriate Cu concentration in the sputtered films could be attained by adjusting the Cu concentration up to 4 moles. All sputtered films possess a slight excessive Cu and Se deficiency. The films grew as a stannite single phase with (112), (220/204) and (116/312) reflections. Under an optimum condition, Cu2ZnSnSe4 films demonstrated a p-type semiconductor characteristic with an absorption coefficient of 104 cm-1 and a band gap of 1.5 eV. These results show that Cu2ZnSnSe4 thin films could potentially be applied as an Indium-free thin film solar absorber for photovoltaic application.

G1-S5.15
Enhancement of Surface Morphology of CdTe Thin Film by Chemical Mechanical Polishing and its Influence on Characteristics and Performance of Solar Cells. (#515) Nam-Hoon Kim¹, Pil-Ju Ko², Woo-Sun Lee², Hyun-Yong Lee³; ¹The Research Institute for Catalysis, Chonnam National University, Gwangju, Republic of Korea ; ²Department of Electrical Engineering, Chosun University, Gwangju, Republic of Korea ; ³Faculty of Applied Chemical Engineering, Chonnam National University, Gwangju, Republic of Korea.

Cadmium telluride (CdTe) is one of the most attractive photovoltaic materials for thin film solar cells due to its low cost, high efficiency and stable performance in physical, optical and electronic properties. It has a direct band gap of about 1.45 eV and high optical absorption coefficients [1,2]. It is generally known that the surface morphology influences the performances of solar cells such as the gain in photocurrent by increase of light scattering [3]. CdTe thin film was prepared by sputtering method in this study, which showed the rough surface morphology. Some researches were reported for enhancement of the surface morphology by a plasma treatment. Chemical mechanical polishing (CMP) processing was firstly proposed for improving the surface morphology of CdTe thin film in this study. Surface roughness and within-wafer non-uniformity (WIWNU%) of the sputtered CdTe thin film was examined with a change of CMP process parameters. The optimized process condition was selected considering to both the surface roughness and the hillock-free surface with the good uniformity. The effects of the enhanced surface morphology on the characteristics and performance of CdTe thin film were also confirmed. X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) were employed to examine the characteristics of CdTe thin film before and after CMP process. The optical properties of CdTe thin film including the photoluminescence and optical absorptance as well as some electrical properties such as the electrical conductivity, Hall coefficient and thermostimulated current (TSC) were also investigated with the application of CMP process. Consequently, the improved performance of CdTe thin film for solar cells was obtained after CMP process by the enahced surface morphology. Acknowledgement: This work was supported by Korea Research Foundation Grant (KRF-2007-412-J02003). [1] G. Pürçek et al., J. Mater. Process. Tech. 2008: [2] S. Krishnan et al., Sol. Energy Mater. Sol. Cells 2008: [3] M. Phyton et al., J. Non-Cryst. Solids 2008.

G1-S5.16
Characteristic Observation and Analysis of TiO₂ Doped with Carbon Nanotubes in the Dye-Sensitized Solar Cells. (#1142) Mi Chen, Horng Show Koo, Yi Ru Chen, Cheng Wei Yan; Department of Materials Science and Engineering, Minghsin University of Science and Technology, Hsinchu, Taiwan.

Following the mature development and silicon materials shortage in market of Si-based solar energy materials and solar cells, novel energy-related materials have been expanded in research and commercial production, especially the dye-sensitized solar cells. This emerging solar energy materials and solar cells have some merits of low power consumption, low environmental loading, low cost manufacturing in commercial production and comparable to conversion efficiency. This paper demonstrates the physical characterization and analysis of TiO2 doped with carbon nanotubes in the dye-sensitized solar cells. The nanostructured and porous TiO2 particles were synthesized by a nanotechnolgy. Besides, The self-synthesized fullerene have been purified by microwave digestion method. The purity of carbon nanotubes greatly affect the cell characteristics of the dye-sensitized solar cells. The paste of the micro-sized and nano-sized TiO2 particles doped with various weight ratio of carbon nanotubes were coated on the surface of ITO-coated glass substrate, which is the working electrode in the dye-sensitized solar cells. The various weight ratio of carbon nanotubes in porous TiO2 particles will lead to the variation of the cell characterization, energy conversion efficiency and Ish, of the dye-sensitized solar cells. These improvements maybe due to the enhanced adsorption of micro-sized and nano-sized TiO2 particles and carbon nanotubes for dye and electrolytic solution in the cells.

G1-S5.17
Diffusion and Chemical Reactions at AgIn₅S₈/CuInS₂ Interface. (#1350) Liudmila Makhova¹, Liudmila Roussak², Gerald Wagner², Igor Konovalov¹; ¹Wilhelm-Ostwald Institute for Physical and Theoretical Chemistry, Leipzig University, Germany ; ²Institute for Mineralogy, Crystallography and Materials Science, Leipzig University, Germany.

The n-AgIn5S8/i-CuInS2 (spinel/chalcopyrite) structure could be regarded as heterojunction for perspective solid-state solar cells. Possibilities for application of these ternary sulphide semiconductors in photovoltaic devices in combination with suitable p-type buffer layers are considered. We present deposition and detailed investigation of CuInS2 chalcopyrite layer prepared on AgIn5S8 spinel single crystal by evaporation from three sources. The AgIn5S8 spinel single crystal has been grown using the vertical gradient freezing method from binary sulphides. The clean spinel surface (111) was obtained by the annealing in sulphur vapour in vacuum chamber at 523 K. Low energy electron diffraction study shows the presence of clear (1x1) patterns. It was found that surface roughening of spinel single crystal takes place after substrate cleaning. The deposition of chalcopyrite layer was carried out by co-evaporation of copper and indium from effusion cells and sulphur from a two-stage evaporation cell. Variation of the spinel substrate temperature in the range 473-723 K during the deposition of the chalcopyrite layers leads to the formation of surface agglomerates of different form (islands, pillars), as demonstrated by scanning electron microscopy micrographs. X-ray photoelectron spectroiscopy and energy dispersive x-ray analysis data have shown, that Ag diffuses from the substrate through CuInS2 film and forms Ag-rich agglomerates at the surface of the structure. Copper interdiffusion from chalcopyrite layer to spinel crystal and copper-for-silver substitution effect in these structure are discussed. Thermochemical possibility of reactions regarding decomposition, formation and cation substitution are considered.

G1-S5.18
TCO Layer with an Effective Light Harvest for Dye-Sensitized Solar Cells. (#1011) Masayuki Okuya, Kenji Ohashi, Takafumi Yamamoto; Shizuoka University, Japan.

A dye-sensitized solar cell has been a hot topic during the decade due to its scientific and technological importance. In order to fabricate the solar cell in a practical use, a novel industrial electrode production technique should be developed. Among many thin film processing techniques, a spray pyrolysis deposition (SPD) technique is one of the most promising ones, since the film formation is carried out in air by a simple apparatus. We report here an optimization of surface morphology of TCO layer for dye-sensitized solar cells by SPD technique. Fluorine doped tin oxide (SnO₂:F) with various surface morphologies were employed as TCO layer. TCO layers were prepared from di-n-butyltin(IV) diacetate or tin(IV) chloride ethanol solution with ammonium fluoride as an additive. The source solution was sprayed by a compressed air onto the heated glass substrate (Corning 1737; 25x25x1 mm³ in size). The mist was pyrolized on the substrate to form SnO₂:F film in air. Since the mist cooled down the substrate, the spraying was not carried out continuously but intermittently. Porous TiO₂ layer was deposited on TCO layer in the same way that mentioned above from the mixture of TiO₂ powder and sols (P-25; Degussa, and TKC-302; TAYCA, respectively). Ruthenium(II) cis-di(thiocyano)bis(2,2'-bipyridyl-4,4'-dicarboxylic acid) dye was adsorbed on the surface of TiO₂ films by refluxing electrodes in the ethanol solution. An anhydrous electrolyte containing I^-/I₃^- was sandwiched between the dye-adsorbed TiO₂ electrode and the platinum coated glass to construct a dye-sensitized solar cell. The current-voltage characteristic measurements were done under the quasi light of AM-1.5 and 100 mW/cm² illumination. Surface morphology of TCO layer was strongly influenced by a source solution. The flat TCO layer prepared from di-n-butyltin(IV) diacetate consists of small particles approximately 30-50 nm in size. On the other hand, a pyramidal texture structure with a particle size of 100-200 nm was produced from tin(IV) chloride.Dye-sensitized solar cells were fabricated with the TCO layers. The short circuit current was enhanced by the texture structure of TCO layer, while the open circuit voltage and fill factor did not depend on the surface morphology of TCO. A rough surface and large particle grain size with a texture structure in TCO layer increased the short circuit current to enhance the overall conversion efficiency up to 8.1%. We found that the conversion efficiency was strongly influenced by Haze rate of TCO layer. The higher Haze rate corresponds to increasing a light harvesting within a working electrode due to the light confinement within the cell to enhance the short circuit current. In fact, higher quantum efficiency was detected for TCO layer with the texture structure.

G1-S5.19
Effects of Anion of Imidazolium Salt Derivatives on Photovoltaic Performance of Dye-Sensitized Solar Cells. (#593) Chaiyuth Sae-Kung¹, Arkom Heawchin¹, Saroj Rattanachot¹, Sanya Keawket¹, Patcharin Wechkama², Passawadee Vijitjunya¹; ¹SOLARTEC dept, National Science and Technology Development Agency, Thailand ; ²National Science and Development Agency, Thailand.

The Effect of Imidazolium in electrolytes on photovoltaic performance of dye-sensitized solar cells was investigated. Electrolytes were mixed with lithium iodide, iodine, tert-butyl-pyridine and imidazolium salt derivatives in acetronitrile. The effect of anion in imidazolium derivatives was also studied. It is found that imidazolium iodide improves short-circuit current density of the cell whereas the open-circuit voltage is not significantly increased. Nonetheless, the imidazolium bromide can increase the open-circuit voltage.

G1-S5.21
SEM Analysis and Selenization of Cu-Zn-Sn Sequential Films Produced by Evaporation of Metals. (#1125) Olga Volobujeva, Tallinn University of Technology, Estonia.

O. Volobujeva, M. Altosaar, J. Raudoja, E. Mellikov, M.Grossberg, S.Bereznev, R.Traksmaa Tallinn University of Technology, Ehitajate tee 5, Tallinn 19086, ESTONIA Cost reductions in PV will be mainly achieved by the introduction of new technologies and materials such as Cu2ZnSnSe4 (hereafter denoted as CZTS, including also the sulfur-based compounds), which can be regarded as an alternative to CuInSe2, and Cu(InxGa1-x)Se2 materials, with the extremely expensive and resource limited indium being replaced by cheap and abundant zinc (Zn) and tin (Sn). Recent attempts to produce quaternary CZTS thin films include pulsed laser ablation, RF magnetron sputtering, synthesis from the melt, and thermal evaporation of the elements [1-4]. The implementation of these processes to produce materials with suitable characteristics for PV application requires a coordinated effort to avoid the formation of secondary phases (CuxSe, ZnSe, SnSex) and to obtain the required stoichiometry of materials. All it lead to the necessity to control mechanism of formation CZTS from metallic precursors. In spite of a lot of investigations of formation of CuInSe2 and CuInS2 in chalcogenisation of metallic films in elemental Se (S) [5,6] there are no results of such investigations till the time published in the literature for CZTS. The SEM SE surface studies and results of EDS analyze of Cu-Zn-Sn sequential precursor films reveal the homogeneity of their composition in macroscopic level. At the same time tin locates as globules on the surface of Mo leading to the inhomogeneity in level of hundreds of nanometers. The selenized films were characterized using SEM SE images to observe the surface morphology, SEM BSE images for compositional analyze and XRD and micro-Raman spectroscopy to identify the phases. It was shown that the parameters of selenization, such as heating temperature strongly influence the morphology of surface and the composition of obtained films. Very intensive out-diffusion of Cu was detected at lower temperatures and the formation on the surface of selenized films copper selenides with composition that was determined by the temperature of selenization . The formation of CZTS began at the temperature around 300C and was nearly completed after annealing during 1hour at 420C. The possible phases in selenized films at different temperatures and selenization mechanism of Cu-Zn-Sn sequential precursor films up to CZTS formation are discussed in detail. Literature 1. R.A. Wibowo et al. J. Phys. and Chem. of Solids, 68 (2007) 1908 2. H. Matsushita et al. Journal of Crystal Growth, 208 (2000) 416. 3. K. Jimbo et al., Thin Solid Films, 515 (2007) 5997 4 M. Altosaar, et al, physica status solidi (a) 205 (2008) 167-170 5. Ch. von Klopmann, et al., Journal of Crystal Growth 289 (2006) 121-133 6. R. Caballero, et al., Thin Solid Films 474 (2005) 70- 76

G1-S5.22
Studies on SnS Thin Films for Solar Energy Application. (#1285) Guang-Pu Wei¹, Wei-Ming Shi¹, Dingyun Gao², Jianping Fang²; ¹School of Materials Science and Engineering, Shanghai University, China ; ²Shanghai Aerospace Energy CO. LTD., China.

Tin sulfide (SnS) is a promising material for low cost solar cell and solar water heater due to its high optical absorption, suitable bandgap, non-toxic and cheaply available. In these papers we report some research results about the preparation and characterization of SnS thin films. We have fabricated SnS thin films by sputtering method,chemical bath deposition (CBD) method and vacuum-evaporation method. The absorption coefficients ? of these films were calculated out and found to be in the order of 10^4 ~ 10^5 cm^-1 . The optical bandgap of SnS thin film was determined by extrapolating from (?^2 ~h?) plots and ranged from 1.3 eV to 1.5eV for different film fabricated by different method. It has been demonstrated that the optical bandgap of SnS film changes with the film composition (Sn/S atom ratio). The Sn/S atom ratio can be controlled by changing the preparation condition, such as the sputtering conditions. SnS films can be used for making solar water heater as a high absorption layer. Key Words: Tin sulfide (SnS), Coefficient, Bandgap, Solar cell, Solar water heater, Photoconductivity.

G1-S5.23
Study of the Properties of Cd_1-xZn_xTe Thin Films Deposited by Vacuum Co-Evaporation Technique. (#215) Lili Wu, Lianghuan Feng, Wei Cai, Wei Li, Jingquan Zhang, Bing Li, Zhi Lei, Jiagui Zheng, Yaping Cai; College of Materials Science and Engineering, Sichuan University, China.

Cd_1-xZn_xTe(CZT) thin films with different zinc content (0<x<1) were fabricated by vacuum co-evaporation technique at different substrate temperatures. The structure, composition, optical and electrical properties of CZT films have been characterized. X-ray diffraction (XRD) spectra showed that the CZT films with different x were cubic phase with highly preferred orientation in (111) direction. There was no obvious relationship between the film structure and the substrate temperature. Energy dispersive X ray diffraction(EDX) results suggested that the as-grown CZT films were rich in Te element and the tendency of Te rich was more obvious with the increasing zinc content and higher substrate temperatures. The bandgap of CZT films varied linearly with x value. The electrical resistivity and conduction active energy of CZT films decreased with increasing x value. The influence of annealing on the structure and morphology of CZT films were also investigated.

G1-S5.24
Fabrication of Cu(In,Ga)Se₂ Solar Cell with ZnS/CdS Double Layer as an Alternative Buffer. (#907) Donghyeop Shin, Liudmila Larina, Kyung Hoon Yoon, Byung Tae Ahn; Korea Advanced Institute of Science and Technology (KAIST), Republic of Korea.

In this paper, we report on the fabrication and characterization of ZnS/CdS double layer being applied as a counterpart in heterojunction Cu(In,Ga)Se₂ (CIGS) solar cells in order to increase light transmittance and to fabricate favorable junction between buffer layer and CIGS absorber layer. As a result, the CIGS solar cell with ZnS/CdS double layer demonstrated the improved energy conversion performance and an increased value of open circuit voltage, compared to CIGS solar cell configuration with buffer that consists only of a single ZnS layer. Normally, a conventional device configuration of CIGS solar cell containing chemical bath deposited CdS buffer layer between the CIGS absorber layer and ZnO window layer gives rise to the cell efficiency is high as 19.5 %. However, the short wavelength response of CIGS solar cell is limited by low CdS band gap of around 2.4 eV. Taking into consideration the environmental aspect, the Cd being toxic element should be replaced by a different material. It is why during last decades many efforts have been provided to construct high efficiency Cd-free CIGS solar cells. Nevertheless, due to unfavorable conduction band alignments at the heterojunction, the cell efficiency of Cd-free CIGS solar cells is still lower than that observed in the conventional CIGS solar cells. Additionally, it is very difficult to form uniform Cd-free buffer with desirable thickness and chemical composition. As a result of optimization deposition process, we provide the uniform and hole-free ZnS/CdS double layer that consists of around 25 nm thick CdS layer and ZnS layer with desirable thickness. Both layers were grown by using chemical bath deposition technique. We have found that CdS layer was covered uniformly forming favorable junction and minimizing the blocking effect in short wavelength region. No cracks have been observed in ZnS/CdS double layer structure. In short wavelength range, the optical transmittance of ZnS/CdS double layer was high and nearly the same as a transmittance of ZnS layer only. As a result, we have demonstrated that ZnS/CdS double layer being used in CIGS solar cell increases the number of high energy photons reaching the junction. Due to this, we succeeded in increasing the value of short-circuit current compared to the value reached in cell configuration using CdS buffer layer. This difference can be clearly seen when comparing the external quantum efficiencies of ZnS/CdS/CIGS and CdS/CIGS solar cell made with identical absorbers. The I-V curve of ZnS/CdS/CIGS solar cell demonstrates the higher open circuit voltage than ZnS/CIGS solar cell. The observed effect can be explained by diminishing the value of series resistance in ZnS/CdS/CIGS solar cell and forming the favorable conduction band alignments at the heterojunction. The cell efficiency was found to be 10.08 % on the active area of 0.44 cm2.

G1-S5.25
The Impedance Spectroscopy Study of CdTe Thin Film Solar Cells. (#736) Kai Yang, Jingquan Zhang, Lianghuan Feng, Zhi Lei, Yaping Cai, Lili Wu, Wei Li, Bing Li; Institute of Solar Energy Materials and Device, College of Materials Science and Engineering, Sichuan University, China.

The effects of the back contact structure and CdCl2 heat treatment of CdTe layer on the CdTe thin film solar cells characteristics have been studied with impedance spectroscopy measurement and current - voltages, capacitance - volatge measurements. The device structure is glass/SnO2:F/CdS/CdTe/back contact, where the back contacts in this work are Cu-doped graphite/Ni and Au respectively . The CdCl2 heat treat temperatures of the CdTe layer are 370?,400? and 430?.A reasonable equivalent circuit model has been used to simulate the admittance spectroscopy measurement results.The results have been analyzed combined with photo- and dark- current-voltage measurements, capacitance-voltage measurements. The effects of CdS/CdTe main electronic junction, defect trap level in energy gap of CdTe absorber layer and back contact structure on device characteristics are analyzied. The results show that (1)There are a comparatively wide defect band in the gap of CdTe absorber layer;(2) The device whose CdTe absorber layer was annealed in higher temperature has a lower back contact barrier;(3)The devices with Au back contact has a lower back contact barrier than them with graphite/Ni back contact.

G1-S5.26
The Preparation and Characterization of Rare-Earth Doped TiO₂ Film. (#638) Xuefeng Zhang,

Abstract: Rare-earth doped nanocrystalline TiO2 film depositing onto a glass substrate were prepared via sol-gel method. The effect of rare-earth doping on the properties of the resulting film, such as surface topography and photocatalysis properties, was analyzed with the aid of X-ray diffraction(XRD), Scanning electron microscope(SEM) and ultraviolet-visible(UV-Vis) spectrophotometer. Key words: TiO2 film rare-earth doping sol-gel method

11:00 AM *G2-S2.1 (invited)
Highly Ordered Nano Structure Hybrid Material for High Efficient Photovoltaic Device. (#662) Wei-Fang Su, Chun-Wei Chen, Chi-An Dai, Lee-Yi Wang; Department of Materials Science and Engineering, National Taiwan University, Taipei, Taiwan.

Hybrid materials made from conducting polymer-nanoparticle are attractive for photovoltaic devices because of the prospect of light weight, low cost, high throughput, high energy density using reel-to-reel or spray deposition on flexible substrate. The hybrid materials serve as the photo-electrical active layer in the photovoltaic device. Upon the absorption of solar energy by hybrid material, excitons are formed and diffused into the interfaces between polymer and nanoparticle. Then the excitons are separated into electrons and holes. The electrons are transported through the pathways built by nanoparticles and the holes are transported through the polymer. Thus the electricity is generated. In this research, we are able to greatly improve the efficiency of the hybrid solar cell by fabricating highly ordered nano structure hybrids, studying the interlayer characteristics of hybrid and device physics, and modifying the surface of nanoparticles. The device usually has the construction of ITO/PEDOT/hybrid/Al four layers. We have established time resolved photoluminescence (TRPL) and time of flight (TOF) to measure exciton life and charge carrier mobility respectively. The inclusion of CdSe nanoparticles into MEHPPV conducting polymer in the hybrid solar cell increases the ordering in polymer and its absorption spectrum was red shifted. The hybrid also exhibits an order increase in power efficiency (2006 Nanotechnology 17, 1260-1263). The exciton life time of hybrid can be decreased to less than half of the neat polymer by blending low cost and nontoxic TiO2 nanorod into MEHPPV (2006 Nanotechnology, 17, 5781-5785). The efficiency of MEHPPV-TiO2 solar cell can be increased by 2.5 times by inserting a TiO2 nanorod layer between the hybrid active layer and Al electrode due to the enlargement of the interconnecting network between the hybrid and electrode (2006 Nanotechnology 17, 5387-5392). The carrier mobility can be increased by 9 times using column structured ZnO electron transport layer infiltrated with the P3HT-TiO2 hybrid due to efficient charge transport (2007 J. Mater. Chem., 17, 4571-4576). We have used conducting molecules to modify the surface of CdSe nanoparticles such as thiophenol or p-hydroxy thiophenol to facilitate the charge transport, 10 times increase in output current is observed (2008, Journal of Materials Chemistry, 18, 675-682). The solar cell fabricated from surface modified TiO2 nanoparticles with carboxylate linker and P3HT hybrid has achieved the relatively high power conversion efficiency of 2.2% under simulated AM 1.5 illumination (100 mW/cm2) because of the reduced recombination of charge carriers.(2008, J. Mater. Chem. in press and 2008 Appl. Phys. Lett. 92, 053312). The efficiency of the device is expected to be further improved by using newly developed self assembled highly ordered nano structure copolymers of P3HT-P2VP (2007 J. Am. Chem. Soc., 129(36), 11036-11038) and low bandgap polyfluorene-co-polythiophene (2007 Macromolecules 40, 8189-8194). This project is supported by the National Science Council of Taiwan (95-3114-P-002-003-MY3) and the AOARD of US Air Force (AOARD-07-4014). Key words: block copolymer, self-assemble, CdSe, TiO2, hybrid, conducting polymer, solar cell

11:30 AM *G2-S2.2 (invited)
Charge Transport in Self-Assembled Materials for Organic Photovoltaics. (#1136) Tobin J. Marks, Michael R Wasielewski; Department of Chemistry, Northwestern University, Evanston, Illinois, USA.

An important goal of our research is to explore the fundamental structural and electronic requirements for efficient energy and electron transfer in extended arrays of donor-acceptor molecules. In this presentation we will describe new molecules designed to provide building blocks for the self-assembly of photoactive materials for solar energy conversion using organic photovoltaics. We are currently exploring molecular building blocks in which photoinduced multi-step electron transfer leads to long-lived radical ion pairs. These structures are comprised of a donor-acceptor system in which two-step charge separation occurs following photoexcitation. This building block self-assembles into a stacked, helical supramolecular structure in which the electron donors and acceptors are segregated. The long lifetime of the photogenerated radical ion pairs within the covalent building blocks allows competitive charge hopping to occur between non-covalent donors and acceptors in the segregated charge pathways within the supramolecular structure. The structure of the supramolecular assembly was determined in solution using X-ray scattering techniques at the Advanced Photon Source synchrotron at Argonne National Laboratory. The structures are determined at concentrations comparable to those at which spectroscopic measurements of their energy and electron transfer dynamics are made; thus providing a direct measure of how structure controls photodriven charge separation within the assembly. Implementation of segregated pathways for electron and hole transport using self-assembly is important for developing new materials for organic photovoltaics.

12:00 PM G2-S2.3
Origin of Open-Circuit Voltage of Hybrid Metal Oxide / Polymer Solar Cells. (#1147) Punniamoorthy Ravirajan¹, Thilini Ishwara², Donal Bradley², Jenny Nelson²; ¹Department of Physics, University of Jaffna, Sri Lanka ; ²Imperial College London, United Kingdom.

The origin of open-circuit voltage (VOC) in organic donor-acceptor heterojunction devices has been the subject of much interest in recent years. Whilst arguments based on the semiconductor metal-insulator-metal picture indicate that VOC should be limited by the difference in work functions between the two electrodes, experimental studies of polymer/fullerene devices showed a dependence of VOC on the fullerene acceptor strength and negligible dependence on the cathode work function. Experimental and modeling studies confirmed that in bilayer devices VOC can be controlled by the difference between the highest occupied molecular orbital (HOMO) of the donor and the lowest unoccupied molecular orbital (LUMO) of the acceptor, and is only partly dependent on the difference in work functions of the electrodes. In bulk heterojunctions where both donor and acceptor phases contact both electrodes, the electrode work functions are also expected to influence VOC. Nevertheless, the most detailed study of polymer/fullerene devices confirmed that VOC varies linearly with the energy difference between the LUMO of the acceptor and the HOMO level of the donor, using a total of different donor materials. In this work, we use polymers with different ionization potential (IP) as the donor and metal oxide with different quasi Fermi-level as acceptor in hybrid polymer/metal oxide solar cell devices. We show that in order to obtain the correct relationship between VOC and the ionization potential of the donor, variations in the photogenerated carrier density between different polymer materials must be taken into account. Once corrected for differences in optical absorption, the measured eVOC values vary linearly with energy separation between HOMO and quasi Fermi-level of metal oxide, with a slope of 0.8 +/- 0.1, in good agreement with a recent study of VOC in polymer:fullerene devices. This indicates that VOC is limited by the difference between the energy separation and the quasi Fermi level of the metal oxide. VOC can be tuned further by modifying the metal oxide Fermi level using an adsorbed monolayer of permanent dipole moment molecules. We demonstrate that treatment of metal oxide with such molecular monolayers results in changes in charge injection and VOC that are consistent with the sign of the dipole moment. The implications for optimization of hybrid polymer - metal oxide device efficiency will also be discussed.

12:15 PM G2-S2.4
Nanoscaled Morphology and Performance of Molecular-Weight-Dependent poly(3-hexylthiophene)/TiO₂ Nanorod Hybrid Solar Cells. (#932) Ming-Chung Wu, Hsi-Hsing Lo, Hsueh-Chung Liao, Wei-Che Yen, Yun-Yue Lin, Chun-Wei Chen, Yang-Fang Chen, Wei-Fang Su; National Taiwan University, Taipei, Taiwan.

Polymer solar cells have attracted a great amount of interest recently due to their potential applications in developing low-cost, solution processable, large-area, mechanically flexible photovoltaic devices. The environmentally friendly and low-cost TiO2 nanocrystal is a promising material in hybrid organic/inorganic photovoltaic device applications. It has been known that the device performance of polymer-based solar cells strongly depends on the morphology, efficiency of charge separation and transport and interfacial contact. McGehee et al. have used atomic force microscopy to study the morphology of P3HT thin films made from different molecular weights (MWs). They have found that the field effect mobility of P3HT increases as the MW increases. Similar observation of the molecular-weight-dependent mobilities was also found in a diode geometry using the space charge limited current measurement. Heeger et al. have also demonstrated the dependence of MWs on the morphology and photovoltaic performance based on P3HT/PCBM. They concluded that the best performance can be achieved by using an optimum ratio between high MW and low MW components. Recently, scanning near-field optical microscopy (SNOM) becomes one of the high-resolution microscopic techniques that have been used to identify the relative distribution of polymers within a blend. The SNOM is a particularly valuable analytical tool for the study of conjugated polymer since many of the important processes pertinent to their applications involve the emission or absorption of photons. In this article, we would like to report the nanoscaled morphology dependence of P3HT/TiO2 nanorod bulk heterojunctions with different polymer MWs by employing the scanning near field microscopy (SNOM), atomic force microscopy (AFM), and confocal Raman microscopy. In addition, the correlation between the morphology and transport behavior for P3HT/TiO2 nanorod hybrids with different MWs are also explored by the time of flight (TOF) mobility measurement. Finally, the photovoltaic performances of the devices based on the hybrids with different MWs are also presented. In this study, we have investigated the effect of polymer molecular weight on the morphology and performance of poly (3-hexylthiophene)/TiO2 nanorod hybrid photovoltaic devices by using scanning near field optical microscopy (SNOM), atomic force microscopy (AFM) and confocal Raman microscopy. From the topography and absorption mapping images, it has been found that the rod-like structure of the low molecular weight polymer hybrid film generates a great amount of grain boundaries and forms a less continuous absorption mapping image. In contrast, the larger domain structure of the high molecular weight polymer hybrid film exhibits continuous absorption mapping, as a result of enhanced polymer stacking and electronic delocalization. The nanoscaled morphology of the hybrid samples with different molecular weights also reveals the nature of photovoltaic performance and carrier transport behavior investigated by the time-of-flight technique.

LUNCH 12:30 PM - 2:00 PM

2:00 PM *G2-S3.1 (invited)
Iodine-Free Solid-State Dye-Sensitized Solar Cells: Effect of Ionic Liquid on Hole conducting PEDOT. (#259) Shozo Yanagida, Center for Advanced Science and Innovation, Osaka University, Japan.

Research and development on dye-sensitized nano-crystalline (nc) TiO2 solar cells (DSC) has been successively progressed, and they are now being recognized as an energy solution photovoltaic in the light of silicone crisis situation. DSC photovoltaics must achieve high stability and reliability under outdoor use at high (<85?C) and low (>-40?C) temperature, in other words, DSC manufacturing must pass standardized packaging durability tests of solar cells under such severe temperature conditions. The conventional DSC contains corrosive and volatile iodine in the hole- conducting phase. One of the solutions for stability issues is the introduction of organic hole-conductors as a substitute of the iodine/iodide electrolyte phase of DSC. Our research has clarified that poly-3,4-ethylenedioxythiophene (PEDOT) works not only as an excellent hole-conductor as a substitute of the iodide/iodine electrolytes [1] but also as excellent electron transfer catalysts on the cathode of fluorine-doped SnO2 (FTO) as well as gold works effectively so far [2]. In addition, the molecular-level integration at the interfaces plays a decisive role in optimization of the dyed-nc-TiO2/PEDOT-DSC, The in-situ photo-electrochemical polymerization of bis(ethylenedioxythiophene) (bis-EDOT) is one effective way to fill the nano-size pore spaces of the nano-crystalline TiO2 electrodes with hole conducting PEDOT. Further, the newly designed sensitizing dye molecules, cis-Ru[4,4'-di(hexythinelvinyl)-2,2'-bipyridyl](4,4'-dicarboxylic acid-2 ,2'-bipyridyl)(NCS)2 (HRS-1), which is quite comparable with N719 in efficiency, was introduced to the PEDOT-DSC devices, clarifying the important role of the molecular-level integration of PEDOT with 2-thiophen-2-yl-vinyl-conjugated pyridine ligand on the dyed-TiO2/PEDOT interfaces [3]. Taking into account the importance of the interface molecular integration and molecular self-organization of DSC, we will discuss about optimization of photo-electrochemical polymerization of bis-EDOT using commercially available Z907 ruthenium complex as sensitizing dye molecules and PEDOT/FTO electrodes as counter cathodes of DSC [4]. References 1) N. Fukuri, N. Masaki, T. Kitamura, Y. Wada, and S.Yanagida, J. Phys. Chem. B 110, 25251-25258, 2006. 2) Y. Saito, T. Kitamura, Y. Wada and S. Yanagida, Chem. Lett., 1060-1061 (2002). 3) A. J. Mozera, Y. Wada, K.-J. Jiang, N. Masaki, S. Yanagida, and S. N. Mori, Appl Phys. Lett. 89, 043509, 2006. 4) Y. Kim, Y. Sung, J. Xia, M. Lira-Cantu, N. Masaki, S. Yanagida, J. Photo. Photobiol. A: Chemistry, 193 77 (2008)

2:30 PM G2-S3.2
Three Dimensional TiO₂ Electrodes for Dye Sensitized Solar Cells Covering Wide Range of Wavelength and All-Metal-TCO-Less Dye Sensitized Solar Cells. (#267) Shuzi Hayase, Yuhei Kashiwa, Yoshitaka Hara, Shinichi Kojima, Tatsuya Yokoyama, Yoshitaka Beppu, Yuhei Ogomi, Mitsuru Kono, Yoshihiro Yamaguchi; Kyushu Institute of Technology, Fukuoka Prefecture, Japan.

In order to increase solar cell efficiency, acquisition of lights up to near IR region is needed. However, it is difficult for single dye to cover the wide range of the wavelength. We propose TiO2 electrodes consisting of two-layer-hybrid TiO2 layer stained by two dyes selectively (two-layer-cell) and all-metal-TCO-less dye sensitized solar cells (TCO-less cell). 1. Two-layer-cell: The two-layer-cell consists of an upper layer stained with a dye absorbing longer wavelength and a bottom layer stained with a dye absorbing shorter wavelength. It was difficult to stain the TiO2 layer selectively with two dyes respectively by using a mere dipping process. We found that two-layer-cell can be prepared by dye-staining under pressurized CO2 atmosphere. It was proved that the two-layer-cell has advantage over cells consisting of randomly-stained-TiO2 layer (Cocktail type cell) in terms of high efficiency. We concluded that the advantage of the two-layer-cell was brought about by inhibiting inconvenient aggregations between the two dyes. The reason why the pressurized CO2 atmosphere makes the selective staining possible was explained by the low concentration of dye molecules in the pressurized CO2 atmosphere and high reactivity between the dye molecule and TiO2 surface. 2. All-metal-TCO-less cells: Since TCO glasses are expensive, TCO-less DSC has been desired. The TCO-less cell structure consisting of all methal electrodes is reported. In order to realize the cell structure, both of high conductivity of Ti electrode and high diffusion of ions through the Ti electrode, which are conflicting items, are needed. When the Ti electrode thickness became thicker than 150nm, the electrolyte diffusion through the Ti electrode was disturbed at the sacrifice of the increase in the conductivity of the Ti layer. We solved the problem by preparing thick Ti thick electrodes with straight nano-pores. All-metal-TCO-less dye sensitized solar cells with 7.4% efficiency is reported.

2:45 PM G2-S3.3
Dye-sensitized Solar Cells with Organic/Inorganic Composite Electrolytes. (#896) Qingbo Meng, Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Zhongguancun, Beijing, China.

Dye-sensitized Solar Cells with Organic/Inorganic Composite Electrolytes Qingbo Meng Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100080, China qbmeng@aphy.iphy.ac.cn Dye-sensitized solar cells (DSCs) have attracted great attention over the past decade owing to high-energy conversion efficiency and low production cost. Above 11% light to electricity conversion efficiency has been achieved in a DSC with an organic liquid based electrolyte containing I3-/I- as the redox couple.(Refs.1,2) However, several practical issues related to the presence of a liquid electrolyte, such as leakage and evaporation of the electrolyte, dye desorption and the corrosive action of the iodide/iodine redox couple, have limited the large-scale technological development of such devices. Approaches towards various electrolytes have facilitated the development of quasi-solid-state DSCs, such as polymeric ionic gels, room temperature molten salts. (Refs.3-5) Recently, several kinds of quasi-solid and solid state electrolytes, composed of organic-inorganic nanocomposites, have been investigated in our group. These solid electrolytes were based on lithium iodide (LiI) and organic ligand. LiI and organic ligand can form addition compound, which has the feature of monoionic transport of I- along three-dimensional (3D) diffusion paths in the crystal. To restrain the formation of large crystallites and improve the contact of TiO2 electrode with electrolyte, inorganic nanoparticles are dispersed into these systems. These electrolytes were successfully used to fabricate DSCs. The energy conversion efficiencies of these DSCs reached to 5.48% with electrolyte of LiI/hydroxypropionitrile/SiO2 and 6.1% with electrolyte of LiI/ethanol/SiO2. (Refs. 6-11) More recently, we also synthesized a new kind organic composite electrolyte by the in-situ reaction of acetylacetone, pyridine and iodine in 3-methoxypropionitrile. The efficiency of DSCs using this electrolyte reaches 6.72%.(Ref.12 ) Acknowledgement: We gratefully acknowledge the support of the National Natural Science Foundation of China under Grant No 20725311, 20673141, 20703063 and 20721140647, the National Basic Research Program of China ("973") under Grant No 2006CB202606, the National High Technology Research and Development Program ("863") under Grant No 2006AA03Z341. References 1 O'Regan B. and Gr?tzel M., Nature, 1991, 353, 737. 2 Gr?tzel, M. Inorg. Chem. 2005, 44, 6841. 3 Tennakone, K., Fernando, C. A. N., Dewasurendra, M. J., J. Photochem. 1987, 38, 75. 4 Meng Q.B.,Takahashi K., Zhang, X.T., Sutanto I., Rao T. N., Sato O., Fujishima A. Langmuir 2003, 19, 3572. 5 Wang, P., Zakeeruddin, S. M., Moser, J.-E., Gr?tzel, M. J. Phys. Chem. B 2003, 107, 13280. 6 Xue B.F., Wang H.X., Hu Y.S., Li H., Wang Z.X., Meng Q.B.,et al Photochem. Photobio. Sci. 2004, 3, 918. 7 Wang H.X., Wang Z.X., Xue B.F., Meng Q.B., et al, Chem. Comm. 2004, 2186 8 Wang H.X., Li H., Xue B.F., Wang Z.X., Meng Q.B., Chen L.Q., J. Am. Chem. Soc. 2005, 127, 6394. 9 Wang H.X., Liu X.Z., Wang Z.X., Li H., Li D.M., Meng Q.B., Chen L.Q., J. Phys. Chem. B 2006, 110, 5970. 10 An H.L., Xue B.F., Li DM, Li H., Meng Q.B., Guo L., Chen L.Q., Electrochem. Comm. 2006, 8, 170. 11 Xue B.F., Wang H.X., Hu Y.S., Li H., Wang Z.X., Meng Q.B., et al, Comp. Rendus Chim., 2006, 9, 627. 12 Liu X. Z., Qin D., Fan Y. Z., Li K. X., Li D. M., Meng Q. B. Electrochem. Comm. 2007, 9, 1735.

3:00 PM G2-S3.4
ZnO Nanorods via Hydrothermal Processes. (#455) Serene Ng, John Wang; Department of Materials Science and Engineering, Faculty of Engineering, National University of Singapore, Singapore.

With a wide direct band gap of 3.37eV and large exciton binding energy of 60meV, ZnO exhibits the desired electronic and photonic properties, making it the material-of-choice for applications such as dye-sensitised solar cells. For this purpose, we have successfully synthesized ZnO nanorods assembled on ITO-glass substrates via a hydrothermal process, where the crystallisation and growth of ZnO nanorods are delicately controlled at low temperatures. To optimize the growth, alignment, packing density and uniformity of the ZnO nanorods, ITO-glass substrates were either spin-coated or RF-sputtered with a layer of ZnO thin film. It was observed that RF-sputtering gave an improved size distribution and surface roughness. The nanorods also had a wurtzite structure and were up to 3um in length and 0.1-0.2um in diameter. Their packing density is approximately 6 x 10^7 rods/cm^2, with a strong (002) preferential orientation, confirming the vertical alignment of nanorods. Several synthesis parameters are shown to strongly affect the growth, morphology and orientation of the nanorods, including the hydrothermal duration, temperature and pH of the solution. The nanorods have been investigated for their electrical transport and photoluminescence behaviour, which are discussed in relation to their structural parameters. In addition, the behaviour of these nanorods in dye-sensitised solar cells will be presented and addressed.

3:15 PM G2-S3.5
Influence of Molecular Volumes of Ruthenium Sensitizers on the Conversion Efficiency of Dye-Sensitized Solar Cells (DSSCs). (#1283) Byeong-Kwan An, Rhiannon Mulherin, Benjamin Langley, Paul Meredith, Paul Burn; Centre for Organic Photonics and Electronics, The University of Queensland, Brisbane, Australia.

Nanostructured TiO2 dye-sensitized solar cells (DSSCs) have been intensively investigated in the past decade as a promising renewable energy source. In such DSSCs, the dye sensitizer is one of the key components for high solar-to-electrical energy conversion efficiency. Thus, considerable effort has been expended in developing new photovoltaic sensitizers with high power conversion efficiencies. So far, heteroleptic ruthenium(II) complexes with anchoring groups have been considered as one of the best candidates for DSSC sensitizers and they have achieved power conversion efficiencies over 11% in standard AM 1.5 sunlight. In this presentation, we report a study on the development of structural guidelines for heteroleptic Ru sensitizers aimed at improving the solar-to-electrical energy conversion efficiency. In particular, we will illustrate how the sensitizer molecular volume effects dye adsorption onto the TiO2 substrates and relate these to the photovoltaic properties of the cells.

AFTERNOON BREAK 3:30 PM - 4:00 PM

4:00 PM *G2-S4.1 (invited)
Developments Towards Efficient, Low-Cost Dye Solar Modules. (#1435) Andreas Hinsch, Fraunhofer-Institut für Solare Energiesysteme ISE, Germany.

Dye solar cells (DSC) are under intensive investigation as a new type of solar cell technology since 15 years. Although still not commercially available, prototyping of DSC modules for certain applications like flexible solar cells and integration into decorative glass facades^I is underway. Recently, progress has been reported on the efficiency^I and on the long-term stability^II of so-called monolithic dye solar cells^III, a concept which in particular is attractive in terms of low material costs. Last year, we reported on results gained in a German network project on material development for monolithic dye solar cells^V. In this paper, the potential of the monolithic dye solar cell concept in terms of efficiency is evaluated. It turned out, that the monolithic concept is well suited for the further efficiency enhancement of dye solar cells beyond 10 %. In particular, the optimisation of the diffusion limited charge transport in the electrolyte and the implementation of new efficient dyes and photoelectrode materials are addressed. I. A. Hinsch, H. Brandt, W. Veurman, S. Hemminga, M. Nittel, U. Würfel, P.Putyra, C. Lang-Koetz, M. Stabe, S. Beucker, K. Fichter, ,,Dye solar modules for facade applications:recent results from project Colorsol″, Proc. 17th International Photovoltaic Science and Engineering Conference, 3 - 7 December 2007, Fukuoka, Japan II. H. Pettersson, T. Gruszecki, R. Bernhard, L. Haggman, M. Gorlov, G. Boschloo, T. Edvinsson, L. Kloo and A. Hagfeldt, ″ The Monolithic Multicell: A Tool for Testing Material Components in Dye-Sensitized Solar Cells″, Prog. Photovolt: Res. Appl. 2007; 15: 113-121. III. N. Kato, K. Higuchi, Y. Takeda, A. Takeichi, T. Motohiro, T. Sano, and T. Toyoda, ″Long-term stability of the DSC module under outdoor working conditions″, Proc. Renewable Energy 2006, October 9-13, 2006, Makuhari, Japan IV. A. Hinsch, R. Kinderman, M. Späth, E. Rijnberg, J.A.M. van Roosmalen, ″The perfomance of dye-sensitised solar cells with a one-facial, monolithic layer built-up prepared by screen printing″, Proc. 2^nd World Conference and Exhibition on Photovoltaic Solar Energy Conversion, Vienna (1998) V. A. Hinsch, P. Putyra, U. Würfel, H. Brandt, K. Skupien, A. Drewitz, F. Einsele, D. Gerhard, H. Gores, S. Hemming, S. Himmler, G. Khelashvili, G. Nasmudinova, U. Opara-Krasovec, U. Rau, S. Sensfuß , J. Walter, P. Wasserscheid, ″Developments towards low-cost dye solar modules″, Proc. 22nd European Photovoltaic Solar Energy Conference and Exhibition, 3 - 7 September 2007, Milan, Italy

4:30 PM *G2-S4.2 (invited)
Catalyst-Free Growth of Vertically Aligned Zinc Oxide Nanowires for Photovoltaic Application. (#965) Shu-Te Ho, Heh-Nan Lin; Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu, Taiwan.

Zinc oxide nanostructures have been the focus of extensive research activities in recent years due to their widespread applications in electrical, optical, and sensing devices. We report large-scale and selective-area growths of vertically aligned zinc oxide nanowires by thermal evaporation via a catalyst-free surface-roughness-assisted mechanism on various substrates including sapphire, glass, indium-tin-oxide (ITO) coated glass, and fluorine-doped tin oxide (FTO) coated glass. The nanowires are single-crystalline and have diameters ranging between 20 and 100 nm and lengths ranging from a few hundred nanometers to a few microns depending on the growth periods. Dye-sensitized solar cells (DSSC) are constructed on ITO and FTO coated glass substrates. The relationships between the solar cell efficiencies and the nanowire spatial properties such as diameter, length, spacing, etc. are explored. Additionally, single nanowires are placed between gold electrodes by a dielectrophoresis process and the electrical transport properties are studied. The correlation between the electrical properties of the nanowires and the solar cell efficiencies will be discussed.

5:00 PM G2-S4.3
Enhancing the Efficiencies of Bulk Heterojunction Solar Cells. (#1281) Karsten Krueger, Paul Schwenn, Paul Burn, Paul Meredith; Department of Physics, The University of Queensland, Brisbane, Australia.

With serious concerns regarding climate change as a result of burning fossil fuels for energy it is important that efficient, cost effective and environmentally friendly forms of energy are developed. Organic thin film bulk heterojunction solar cells are one promising candidate for energy harvesting because of their low cost, mechanical flexibility and simple, scalable large surface area production. So far, verified efficiencies of more than 5.4% have been reported for these types of devices. The optimal choice of the donor and acceptor materials is critical for the development of highly efficient devices. In addition, the precise control of the morphology of the donor-acceptor blend is of the utmost importance to achieve high performance. In this presentation, the effects of several performance enhancing methods, like slow solvent evaporation ("solvent annealing") and thermal annealing are investigated to optimize the morphology of the active layer. Furthermore, other device modifications such as the inclusion of cathode interfacial layers are investigated to improve the charge injection/extraction properties of the device. We have obtained conversion efficiencies of more than 4% with devices based on poly(3-hexylthiophene) as donor and phenyl-C61-butyric acid methyl ester as the acceptor material. During this process, we identified several pit falls regarding processing that can limit the performance of bulk heterojunction solar cells.

5:15 PM G2-S4.4
Organic Photovoltaics at CSIRO. (#758) Lynn Rozanski, Scott Watkins, Chris Fell; Australian Commonwealth Scientific and Research Organization (CSIRO) Energy Technology, Australia.

Organic photovoltaics (OPVs) are emerging as an inexpensive alternative to traditional silicon-based photovoltaics. The active materials used in OPVs can be solution processed under ambient conditions, allowing them to be applied to flexible substrates and potentially utilise existing roll-to-roll printing technology. At present, OPV technology is limited in terms of both performance and durability. The world's best devices have efficiencies around 5%, and even with encapsulation, lifetimes are currently insufficient for large-scale commercialisation. However, with careful design of new molecular materials, great potential exists to overcome these performance barriers. Device efficiency has suffered as a result of limited absorption across the solar spectrum as well as due to poor charge separation and mobility, when compared to inorganic alternatives. CSIRO Energy Technology has partnered with CSIRO Molecular and Health Technologies to design, synthesise and test novel materials for OPV devices. Novel block co-polymers containing pendant electron and hole transport groups have the potential to be a single-component active material, improving both charge separation and transport. In addition, the solar spectrum absorption can be enhanced through the use of polymer-functionalised nanoparticles. At our National Solar Energy Centre in Newcastle we are also examining numerous other challenges associated with scaling OPV devices to commercial volumes, including module design, printing and encapsulation.

G2-S5.1
Au/Pt/InGaN/GaN Heterostructure Schottky Prototype Solar Cells. (#61) Dunjun Chen, Zili Xie, Rong Zhang, Youdou Zheng; Department of Physics, Nanjing University, China.

It has been demonstrated that the band gap of In1-xGaxN alloys can be varied continuously from 0.7 to 3.4 eV or wavelengths from 1771 to 365 nm, a near-perfect match to the solar energy spectrum.This creates an unprecedented opportunity to design and fabricate new solar cells with greatly improved efficiencies. In particular, InGaN alloys show a much higher resistance against proton irradiation damage as compared to conventional GaAs and GaInP solar cell materials, and hence, present a great potential for applications in space solar cells. However, work on InGaN solar cells has not yet progressed due to the big challenge of p-type doping of InGaN alloys. It is well known that InGaN alloys have an extremely propensity of being doped n-type by native point defects and present a high n-type background carrier concentration. Therefore, it is difficult to study p-n junction InGaN solar cells before resolving the problem of p-type InGaN materials. Although multi-junction solar cells show more promising due to their higher power conversion efficiency, Schottky barrier solar cells are convenient for investigation of the influence of materials and electrodes properties on the solar cells performance. It is worth mentioning that it is also a rather difficult work for n-InGaN to obtain high-quality Schottky contacts due to a large amount of surface defects in InGaN layers. Therefore, there are only limited works on Schottky barriers of metal/n-InGaN. In this work, a patterned Au/Pt/In0.2Ga0.8N/GaN heterostructure Schottky prototype solar cell has been fabricated. The forward current-voltage characteristics indicate that the thermionic emission is a dominating current transport mechanism at the Pt/InGaN interface in our fabricated Schottky solar cell. The spectral responsivity of the solar cell measured at zero-bias exhibits two response steps. The higher response step with the wavelength cutoff at around 360 nm is assigned to the GaN layer and the lower one with longer wavelength cutoff at around 437 nm is attributed to the In0.2Ga0.8N layer. The Au/Pt/In0.2Ga0.8N/GaN heterostructure Schottky solar cell has a very low dark current density, and open-circult voltage of 0.91 V and short-circuit current density of 7 mA/cm2 when illuminated by a Xe lamp with the light intensity of 30 mW/cm2.

G2-S5.2
Optical Bandgap of Si Quantum Dot Superlattice. (#935) Eun-Chel Cho, Xiaojing Hao, Sangwook Park, Gavin Conibeer; The University of New South Wales, Sydney, Australia.

All-silicon tandem solar cells using quantum-confinement in silicon quantum dots (QDs) have been proposed to enhance efficiency through the increased spectral collection efficiency. The optimal bandgap of the top cell to maximize conversion efficiency is 1.7 ~ 1.8 eV for a 2-cell tandem with a Si bottom cell [1], where there is a problem to determine the optical bandgap of Si QDs. In recent years it has been shown that, accompanying the reduction in size of silicon nanocrystals, i.e., quantum dots, zero-phonon optical transitions are partially allowed and the oscillator strength of zero-phonon transitions is significantly enhanced. This increases the radiative recombination rate via a direct band-to-band recombination process. The quantum confinement effect is suggested as the origin of the strong luminescence band, which causes an enlargement of the band gap. Photoluminescence was used to determine the its effective bandgap as well as quantum confinement effect.[1] The optical bandgap Eopt of Si QDs is determined by measuring transmission and reflection spectra of the films deposited on quartz substrate. The Eopt was extracted from Tauc function [ 2,3 ]. Due to the indirect nature of silicon, square root is used to determine the optical bandgap. However it is hard to get a linear relation in Taus's plot due to a quantum confinement of Si QDs. In addition, there is a difference in PL peak energy and absorption edge, which can be explained by Stoke's shift.[4] Si QD superlattice fabricated by alternate deposition of silicon-rich oxide (SRO) and SiO2 by a radio-frequency magnetron sputtering technique and subsequently high temperature annealing. SRO was realized by a co-sputtering of Si and SiO2 targets at a room temperature. Similarity and difference in optical bandgap by Taus's plot and effective bandgap by photoluminescence measurement will be discussed and compared. 1. E. C. Cho, M. A. Green, G. Conibeer, D. Y. Song, Y-H, Cho, G. Scardera, S. J. Huang, S. Park, X. J. Hao, Y. D. Huang, L. V. Dao, Advances on Optoelectronics, 69578 (2007). 2. J. Tauc, R. Grigorovici, and A.Vancu, Phys. Status Solidi 15, 627 (1966). 3. A. Matsuda, T. Yamaoka, S. Wolff, M. Koyama, Y. Imanishi, H. Kataoka, H. Matsuura, and K. Tanaka, J. Appl. Phys. 60, 4025 (1986). 4. Z. Ma, X. Liao, G. Kong, J. Chu, Appl. Phys. Lett. 75, 1857 (1999).

G2-S5.4
Improvement of Figh Rate Growth μc-Si:H Solar Cell by Insertion of Buffer Layer to p/i Interface. (#464) Xiaoyan Han, Nankai University, Tianjin, China.

In the process of high rate growth ?c-Si:H film by very high frequency plasma enhanced chemical vapor deposition (VHF-PECVD), the high energy ion species impinging on the growing surface could deteriorate the device performance. Incorporation of a low rate growth intrinsic ?c-Si:H to p/i interface was advanced in this paper. The results shown that the introduced low rate growth buffer layer could improve the characteristics of p/i interface and the vertical uniformity of intrinsic layer. It was found that the defects in intrinsic layer were first decreased and then increased with increasing the thickness of the buffer layers. These results led to an optimal thickness for the buffer layers. The efficiency of solar cells was increased about 1% when the thickness was optimized. As a result, the efficiency of 8.11% has been achieved at an i-layer deposition rate of 8.5?/s.

G2-S5.7
Spiral Heterosturucture-Based New Solar Cells with Constant Output Voltage. (#1076) Nobuyoshi Kawaguchi, Akira Ishibashi; Laboratory of Quantum Electronics, Research Institute for Electronic Science, Hokkaido University, Sapporo, Japan.

We propose a next generation solar cell, based upon a spiral heterostructure, with a constant output voltage. The spiral heterostructure-based solar cell (SHSC) has a disk shape, being composed of a rolled-up pn junction layer on a flexible substrate sandwiched between an anode electrode and cathode electrodes. Photons impinge on the disk surface and photogenerated electrons and holes drift in the radial direction. The photon propagation direction being parallel to the vertical axis of the spiral heterostructure is perpendicular to the photogenerated carrier drift direction that is the radial direction of SHSC. Since these two directions are orthogonal, our new solar cell, being different from conventional solar cells, can optimize the absorption of light and the photogenerated carrier collection independently and simultaneously. We prepare 2n (n is a natural number) semiconductor stripes with different band gaps Eg^(k) (k = 1 to 2n) on the flexible substrate to be rolled up into a disk. With respect to the sunlight, we set the disk of SHSC so that the incoming photons first encounter the widest gap semiconductor with energy gap Eg = Eg^(1) (≡ Eg_max), then narrower gap semiconductors with Eg = Eg^(k) (k = 2 to 2n − 1), and the narrowest with Eg = Eg^(2n) (≡ Eg_min), to convert the whole spectrum of black body radiation into electricity. By choosing appropriate energy gaps, we set Eg^(k) = Eg_max − (k − 1)ΔE. Each of the semiconductor stripes is sandwiched between the anode electrode and the cathode electrode to form a unit cell. The anode electrode is in common and 2n cathode electrodes are connected to 2n semiconductor materials independently. As a consequence, in the multistriped SHSC, each unit of the solar cell outputs a different voltage depending on the bandgaps of the semiconductor materials used for the stripes. Multi voltage output with small current amount from a single cell is not convenient to use. To solve this problem, we prepare two 2n striped SHSCs, i.e., SHSC_1 and SHSC_2, and connect serially the unit cell having Eg = Eg^(k) in SHSC_1 with that having Eg = Eg^(2n − k + 1) in SHSC_2. Then the output voltage is {Eg^(k) + Eg^(2n − k + 1)}/e = {Eg_max − (k − 1)ΔE + Eg_max − (2n − k)ΔE }/e = {Eg_max + Eg_max − (2n − 1)ΔE }/e ={ Eg_max + Eg_min }/e = const., i.e., each of 2n serially connected unit cells outputs a single constant voltage, where e is the electronic charge. In contrast to a hot carrier solar cell that needs to utilize large excess kinetic energy of the photogenerated carriers before they relax, our new solar cell does not, thanks to multi striped SHSC structure. The connected multistriped spiral heterostructure-based solar cells can convert virtually the whole spectrum of black body radiation into the electricity with a single output voltage, being a candidate for next generation solar cells with high energy conversion efficiency.

G2-S5.8
Effect of Surface Textured AZO:H Films on Light Scattering in Thin Film Silicon Photovoltaics. (#1139) Sung Ju Tark, Min Gu Kang, Ji Hoon Jang, Jeong Chul Lee, Joon Sung Lee, Hee-jin Lim, Donghwan Kim; Department of Materials Science and Engineering, Korea University, Seoul, Republic of Korea.

The requirements for applications of transparent conducting oxide (TCO) films in thin film silicon solar cells are reviewed with a focus on sputtered Al doped zinc oxide and fluorine doped tin oxide films. TCO films are employed as a front contact and as part of a highly reflective back contact so that the silicon absorber layer is embedded by TCO films. The optoelectronic properties of TCO layers and their influence on the solar cell performance are discussed. ZnO can be used as a TCO films for the front electrode of silicon thin film solar cells due to its low resistivity and high optical transmittance. Hydrogenated Al-doped zinc oxide (AZO:H) thin films were grown on the glass (corning eagle 2000) by rf magnetron sputtering using a ceramic target (98 wt. % ZnO, 2 wt.% Al2O3) in the H2/Ar ambient. We changed the growth condition of AZO films such as the ratio of H2/(Ar+H2) and the substrate temperature from RT to 250?. As results, in the 2% H2 addition with the substrate temperature of 150?, it showed the excellent electrical property of the AZO:H films with the resistivity of 3.21x10-4 ?ocm. UV-measurement showed that the optical transmission of AZO:H films is above 85% in the visible range with widen optical band gap. After deposition, the smooth surface of AZO:H films were etched in diluted HCl (0.5%) to investigate the variation of optical properties and surface morphology due to a textured surface. The prepared etched thin films had excellent scattering properties (spectral haze: 52%) and low reflectance. We will show promising results that demonstrates an improved light trapping due to the textured surface formed by an etching technique. We will discuss the effect of hydrogen on the electrical and optical properties of AZO:H films and apply AZO:H films to ?c-Si solar cell.

G2-S5.9
Preparations of TiO₂ Nanoparticles and its Application to Light Scattering Layer for Dye-Sensitized Solar Cells. (#982) Jin-Kook Lee, Bo-Hwa Jeong, Sung-Il Jang, Young-Guen Kim, Ji-Un Kim, Young-Wook Jang, Mi-Ra Kim; Department Polymer Science and Engineering, Pusan National University, Republic of Korea.

Dye-sensitized solar cells (DSSCs) constructed using dye molecules, nanocrystalline metal oxides and organic liquid electrolytes, have attractive features of high energy conversion efficiency and low production cost [1]. Since Gr?tzel's group reported an overall efficiency of DSSC exceeding 11%, many research groups have been focused on the improving the photocurrent and photovoltage by developing new dye-sensitizer, suppressing the charge recombination, improving the interfacial interaction and/or modifying the electrolyte components [2]. Recently, improvement of light harvest efficiency in the dye-adsorbed TiO2 electrode by light scattering has been reported [3]. Light scattering effect can be achieved by additional layers in the TiO2 layers. Addition of the scattering layers with the large particles ensures adequate light trapping in the device [4], due to the increase of absorption path length of photons and optical confinement. Ferber et al. [5] and Rothenberger el al. [6] confirmed the light scattering effect with the transport theory and the many-flux model, respectively. With scattering abilities in the TiO2 films, it is also important that TiO2 electrode has the higher surface area, which is responsible for the optimal dye loading and effective photocurrent generation. In this work, four kinds of TiO2 paste solutions for the electrode of the DSSC device were composed of TiO2 particles of 9 nm, 20 nm, 123 nm and 300 nm in the average diameter, respectively. 9 nm- and 20 nm-TiO2 particle solutions were used for the dense layer. 123 nm- and 300 nm-TiO2 particle solutions were used for light scattering layer from the incident light. Especially, 20 nm- and 123 nm-TiO2 paste solutions were prepared by using sol-gel method. Here, to confirm a significant role of a light scattering layer, we determined the overall photovoltaic performances of DSSC device with the light scattering layer at the up 123nm- or 300 nm-TiO2 layer. Five types of the TiO2 electrodes were prepared by layer-by-layer deposition of mono- or multi-layer. The best efficiency of 6.03 % under illumination of simulated AM 1.5 solar light (100 mW/cm2) was attained with a multilayer structure using 123nm-TiO2 layer for the light scattering layer and 9 nm-TiO2 layer for the dense layer. 1. B.O'Regan, M. Gr?tzel, Nature 353 (1991) 737. 2. F.C. Krebs, M. Biancardo, Sol. Energ. Mat. & Sol. Cells 90 (2006) 142. 3. A. Usami, Chem. Phys. Lett. 277 (1997) 105. 4. S. Hore, C. Vetter, R. Kern. H. Smit, A. Hinsch, Sol. Energ. Mat. & Sol. Cells 90 (2006) 1176. 5. J. Ferber, J. Luther, Sol. Energ. Mat. & Sol. Cells 54 (1998) 265. 6. A. Usami, Sol. Energ. Mat. & Sol. Cells 64 (2000) 73.

G2-S5.10
Effect of Oxygen Rich-Seed Layer on Material Properties of ZnO:Al Films. (#1039) Young-Jin Kim, Jun-Sik Cho, Jeong-Chul Lee, Jin-Suk Wang, Jinsoo Song, Kyung-Hoon Yoon; Korea Institute of Energy Research, Republic of Korea.

Young-Jin Kim1,2, Jun-Sik Cho1, Jeong Chul Lee1, Jin Suk Wang2, Jin Soo Song1, and Kyung Hoon Yoon1 Solar Cells Research Center, Korea Institute of Energy Research, 71-2 Jang-dong, Yuseong-gu, Daejeon, 305-343, Korea1 Chung-nam National University, Korea2 High quality ZnO:Al films for silicon thin film solar cells were prepared on glass substrates by rf magnetron sputtering using oxygen rich-seed layers which affected the microstructure and crystallinity of the final films. The ZnO:Al films were deposited by two-step method in which the seed layer was deposited in oxygen and argon gas mixture and the bulk layer in only argon gas. Effects of oxygen rich-seed layer on the structural and electrical properties of final films were investigated using various characterization methods such as XRD, SEM, AFM and Hall measurement. The as-deposited ZnO:Al films were etched chemically in diluted HCl and changes in the surface morphology and light scattering of the etched ZnO:Al films were examined. The influence of surface texturing of ZnO:Al films on the performance of silicon thin film solar cells will be explained in detail in terms of the changes in structural, electrical and optical properties of the ZnO:Al films.

G2-S5.11
Technological Infrastructure and Business Model of Solar Cell Industry. (#454) Horng-Show Koo, Department of Optoelectronic System Engineering, Ming-Hsin University of Science and Technology, Hsinchu, Taiwan.

Horng-Show Koo1,* and Mi Chen2 1 Department of Optoelectronic System Engineering, Ming-Hsin University of Science and Technology, Hsinfong, Hsinchu 304 Taiwan, R.O.C. 2 Department of Materials Science and Engineering, Ming-Hsin University of Science and Technology, Hsinfong, Hsinchu 304 Taiwan, R.O.C. * Corresponding author : frankkoo@must.edu.tw Abstract This paper demonstrates present status and future development of solar cell technology and industrial infrastructure in Taiwan. The industrial and technological development of Si-based microelectronics semiconductors and GaAs-based optoelectronic devices are the main driving force in evaluating and investing the next high-tech industry such as solar cell. The basic research on flexible energy-related electronics is merely planning and executing in national institute organizations and academic universities, the enterprises in industry are under the investigation and observation. The integrated infrastructure and industrial supply chain of IC, LED, OLED and LCD industry in Taiwan are gradually formed and constructed. The emerging foundry business model in semiconductor industry is also introduced and feasible in the industrial field of solar cell technologies. A series of the conventional vertical integration manufacturing model is gradually replaced by the innovative horizontal disintegration production mode - foundry business model. The solar cell industry mode have been divided three sections of upper- stream industry in raw materials, medium-stream one in professional manufacturing and lower-stream industry in module system assembly. The complete supply chain have constructed in medium-stream and lower-stream industrial enterprises in Taiwan, but the upper-stream industry in raw materials have not completely infrastructures. In the seeable future, the establishment and production of the raw materials for solar cell application should be achieved, because some global cooperation and alliance enterprises have being proceeded. The technical development, supply chain and innovative business model of the upper-stream, medium-stream and lower-stream industries in Taiwan will be specified in this article.

G2-S5.12
To a New Concept of High-Effective Solar Cells. (#884) Viktor Laptev, Halyna Khlyap; Institute of Physical Electronics, Universität Stuttgart, Germany.

Abstract presents the very first results obtained from investigations of formation and operation of a solar cell with metallic clusters in front contacts. All different structures operating as solar cells are produced with metallic contacts (fingers) making a very sufficient contribution in effectively functioning solar cell. Technology and functionality of fingers are seemed to be of special importance for manufacturing high-effective solar elements. We report some quiet new ideas suitable for solving the problem of good contacts without dependence on the solar cell type (organic, inorganic, etc.). Effectivity of solar energy conversion is in a very close connection with a structure of the solar cell and, more specially, with a structure of metallic parts of the device (in particular, fingers). The traditional way of fingers' formation does not imply effects of finger structure on performance of the device in general. There no principal difficulties in replacing Cu-clusters in pores of metallic contacts by dye molecules. Clusters-like structure of metallic stripes as well as clustering solar cells (for example, dye-based and other types of molecular-based active elements) are seemed to be of special interest. The sequence of ground and excited states of atoms is presented as a thermodynamic process. The requirements to processes, under which the electron energy transitions with maximal work production are realized, are described. The number of such transitions increases from diamond to graphite and soot. On the example of silicon it is suggested that clusters with small number of atoms can be comparable with ideal absorber. Electric properties of Cu/Ag system contacts obtained for different types of solar cells are investigated. The crystalline Cu-clusters are placed in Ag-matrix pores and on the surface of silver contact stripes. Alternation illumination-darkness in the same experiment led to the unique results obtained under standard measurements of specific resistance of Cu/Ag-fingers.

G2-S5.13
Effect of p µc-Si:H Window Layers on Electrical Properties of Intrinsic µc-Si:H Layer and Superstrate p-i-n Solar Cells. (#1196) Ji Hoon Jang, Jeong Chul Lee, Jun-Sik Cho, Hae Seok Lee, Kun Ho Ahn, Heon-Min Lee, Don-Hee Lee, Young Kuk Kim, Jun Sin Yi, Jinsoo Song, Kyung Hoon Yoon; Korea Institute of Energy Research, Republic of Korea.

This paper investigates the effect of p-type microcrystalline silicon (p µc-Si:H) window layers on electrical properties of intrinsic µc-Si:H light absorbing layer and superstrate p-i-n solar cells. The crystalline properties of p µc-Si:H seed layers give significant effects on the grain growth of i-type µc-Si:H films in the initial stage of nucleation. The i µc-Si:H deposited on p µc-Si:H window layers with high crystalline volume fraction shows high dark conductivity due to the high defect density from grain boundaries. The performances of superstrate p-i-n µc-Si:H solar cells are also significantly affected by the crystalline properties of p layers. The open circuit voltage and fill factor linearly decreases with increasing crystallinity of p layers although i µc-Si:H layers are deposited at the same deposition conditions.

G2-S5.14
Purification of MG-Silicon by Gas Blowing during the Fractional Melting Process. (#239) Jaewoo Lee, Wooyoung Yoon; Department of Materials Science and Engineering, Korea University, Seoul, Republic of Korea.

Fractional melting process involves heating an alloy within its liquid-solid region while simultaneously ejecting liquid from the solid-liquid mixture. A previous study has demonstrated that a purifying process utilizing the principle of the fractional melting is more effective than zone refining or fractional solidification process. To obtain theoretically high refining ratio, however, two conditions are required. First, liquid must be completely removed from the sample as soon as it is formed. Liquid that remains in the sample acts as a source of solute on heating and lowers the refining ratio obtainable. Second, heating must be sufficiently slow that complete diffusion takes place in the solid. For these reasons, FM process is difficult to apply for commercial processes. In this research, the new fractional melting process in which the gas blowing force is used for separating the liquid from the cake has been developed and applied to the purification of the metallurgical grade Si (MG-Si). The major impurities in MG-Si such as Fe, Ti, Al, and Cu can significantly degrade the efficiency of the solar cell, so it is important to remove these metal elements from MG-Si to obtain high-quality silicon. Since these elements have low segregation coefficients in silicon, a high purification is possible through the fractional melting process.

G2-S5.15
Determination of Surface Compositions on c-Si Solar Cell by AES/ XPS. (#533) Z.Q Ma, F Li, B He; Department of Physics, Shanghai University, China.

Abstract: In this paper, the specific surface and interface characterization of phosphorous-included silicon oxide / Si (111) materials structure for mono-crystal Si solar cells have been examined by using Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS), respectively. This methodology has been well established by our laboratory in the previous works [1]. It is found that the various chemical compositions such as C, O, SiC, (CH3)n, (C6H5O)3PO, and the atomic bonds such as Si-C, , C-O, P-O, C6H5-O, C-O-C or Si-O-Si bridges, and Si-O-C linkages were co-present at top multi-layers of the as-doped Si wafer. These parasitical micro-phases on the surface or subsurface of the impurity doped silicon directly influence the short-circuit current, open-circuit voltage, series and shunt resistance, filled factor, the minority carrier lifetime and the conversion efficiency of the c-Si based solar cell. The photoelectric conversion property is strongly dependent on the electronic states and the depth profile of the chemical compositions. Keywords: XPS/AES, Si-based solar cells, Surface and interface characterization, Microstructure Reference: 1. Z.Q.Ma, Q.Guo and Tao.Jin, "Energy spectroscopy studies of radiation-induced damaged surfaces and interfaces in SiO2/Si by light charged particles" Nucl. Instr. & Meth. B71, (1992) 278-290.

G2-S5.16
A Diagram for Estimating Cost-Effectiveness of Solar Modules. (#409) Yasuto Maruoka, Doitsu Giken Ltd., Tokyo, Japan.

Determination of cost-effectiveness of solar modules is cumbersome. A diagram has been de-veloped for estimating maximal power of solar modules and front surface area as well as list price on the photovoltaic market. The diagram is titled "Cost-Effectiveness Diagram". This paper sets out a method for the use of the diagram with supporting arguments. The "Cost-Effectiveness Diagram" has been derived from a previous market analysis by this au-thor. For the analysis, information from independent public sources was collected. Seven solar cell materials: poly-Si, c-Si, a-Si, a-Si/c-Si, a-Si/μ-Si, CIS, and CIGS were compared using a sample base in excess of 120 modules produced by 26 different manufacturers in the German, US and Japa-nese markets. On the "Cost-Effectiveness Diagram", relative maximal power Pmax, mod/Pmax, sys is presented as a function of the product of relative area Amod/Asys and relative cost Cmod/Csys with the cost-effectiveness factor G as a parameter. The indexes 'mod' and 'sys' denote a module and an as-sumed reference photovoltaic system. The dimensionless values incorporated in the equation upon which this diagram is based, have been numerically evaluated using an empirical equation devel-oped in the aforementioned analysis. The relative maximal power increases with increasing product of the relative area and cost as well as the cost-effectiveness factor. The cost-effectiveness of a module can be estimated by providing two input values: the relative maximal power of the module and the product of the relative area and cost. The confluence point of both values on the diagram gives us the cost-effectiveness of the module. This point can be easily interpreted as the cost-effectiveness of the module by referring to the G-line Scale clearly marked on the diagram. If the cost-effectiveness factor of the module is at G ≥ 3.3, its cost-effectiveness is above the average on the market. However, if G < 3.3, it means that, either the maximal power of the module is lower, or its list price is higher than market average. In this case, the surface area may be larger than average. Using this graphic representation, it is clearly evident how far the cost-effectiveness of the module is from average. If the cost-effectiveness factor is assumed G = 3.3, an additional ad-vantage of the diagram is the ability to calculate, in reverse, an estimation of maximal power, sur-face area required, as well as a desirable list price reflecting market reality. The conclusion is that the "Cost-Effectiveness Diagram" is a useful tool for the assessment of the performance, cost as well as list price of solar modules. The diagram imparts information about the value of modules quickly and easily for experts and end-users alike. ***

G2-S5.17
Photovoltaic Properties of N-type β-FeSi₂/Intrinsic Si/P-type Si Heterojunctions. (#743) Keita Nomoto, Mahmoud Abdel-Rady Ahmed Shaban, Tsuyoshi Yoshitake; Department of Applied Science for Electronics and Materials, Kyushu University, Fukuoka, Japan.

The semiconducting iron disilicid (β-FeSi₂) features a direct optical band gap of 0.85 eV and a large absorption coefficient, which is 10⁵ cm‾¹ at 1.5 eV. It can be epitaxially grown on Si due to the small lattice mismatch. Moreover, β-FeSi₂ is an ecologically friendly material due to its nontoxicity and its elements, Fe and Si, are abundant in nature. In a β-FeSi₂/Si solar cell, the front-surface Si layer absorbs the ultraviolet light and visible light and the near-infrared light transmitted the Si layer is absorbed by the back-surface β-FeSi₂ layer, therefore a high photovoltaic performance is expected. Previously, we succeeded in epitaxially growing n-type β-FeSi₂ thin films on p-type Si (111) substrates at a substrate temperature of 600°C by the facing-target DC sputtering. The fabricated p-n heterojunctions exhibited rectifying and photovoltaic properties accompanied with a large leakage current, which is mainly owing to the inter-diffusion of Fe into Si. Fe atoms easily diffuse into the Si substrate at 600°C. The diffused Fe into Si generates deep trap levels and heterojunction degrades the properties. In this study, we inserted the very thin intrinsic-Si layer (i-layer) less than 100 nm to suppress the Fe atoms inter-diffusion. The photovoltaic properties of the p-i-n structure were evaluated based on the current-voltage characterization both in the dark and under illumination. The rectifying and photovoltaic properties of the p-i-n structure were improved as compared with the p-n structure. The inserted i-layer is apparently effective in the leakage current suppression.

G2-S5.18
Photovoltaic Properties of n-Type Nanocrystalline-FeSi2/Intrinsic-Si/p-Type Si Heterojunctions Fabricated by Facing-Target DC Sputtering. (#803) Mahmoud Shaban, Haruhiko Kondo, Tsuyoshi Yoshitake; Department of Applied Science for Electronics and Materials, Kyushu University, Fukuoka, Japan.

Nanocrystalline iron disilicide (NC-FeSi₂) is a novel semiconducting material comprising crystallites with diameters ranging from 3 to 5 nm [1]. It has a direct optical bandgap close to that of crystalline NC-FeSi₂ and an optical absorption coefficient larger than 10⁵cm^-1 at 1.2 eV [2]. Moreover, NC-FeSi₂ is an eco-friendly material and can be grown at room temperature. Previously, we experimentally demonstrated the photovoltaic effect of n-type NC-FeSi₂/p-type Si heterojunction photovoltaics wherein all the fabrication processes were carried out at room temperature. These heterojunction exhibited rectifying behavior similar to that of conventional p-n heterojunctions. On the other hand, their conversion efficiency was low due to their large leakage currents. The main source of this leakage is the dangling bonds and the high residual carrier density of the NC-FeSi₂ film, which exceeds 10¹⁹cm^-3. In this work, a thin intrinsic-Si layer was grown on the Si substrate as a buffer layer to reduce the heterojunction leakage current. The leakage current of the p-i-n heterostructure was markedly reduced and therefore the photovoltaic properties were improved. Thin intrinsic-Si layers (i-Si) with a thickness of 40 nm was homoepitaxially grown on p-type Cz-Si(111) substrates at a substrate temperature of 600°C. After reducing the samples temperature to room temperature, n-type NC-FeSi₂ films with thickness of 320 nm were deposited, in hydrogen ambient, to form p-i-n heterostructure. The i-Si layers and the NC-FeSi₂ thin films were deposited by facing-target DC sputtering (FTDCS) using Si and FeSi2 alloy targets, respectively. The base pressure in the chamber was less than 5.0 x 10^-5 Pa, and the discharge gas pressure was maintained at 1.33 x 10^-1 Pa by introducing Ar gas. After the deposition of the NC-FeSi₂ films, the samples were transferred into a radio-frequency magnetron sputtering apparatus to deposit the ohmic contacts. A finger-shaped Pd electrode was deposited on the front of the Si surface and an Al electrode was deposited on the back of the NC-FeSi₂ surface at room temperature. The film morphology was observed by scanning electron microscopy (SEM). The device photovoltaic properties were evaluated on the basis of the current-voltage characterization and the photoresponse spectrum measurement. The SEM observation results showed that the NC-FeSi₂ film was continuous and uniform in thickness. The interface between the NC-FeSi₂ film and the Si substrate was extremely sharp owing to the room temperature growth and the plasma-free deposition that is specific to FTDCS. An extremely smooth film surface without pinholes was also confirmed from the bird-view and plane-view SEM images. The leakage current of the p-i-n heterojunction was reduced by at least an order of magnitude as compare to that of the p-n heterojunction. While the short-circuit current density was slightly enhanced from 4.5 to 5.5 mA/cm², the open-circuit voltage showed a remarkable increase from 150 to 320 mV. The inserted i-Si layer reduced the leakage current and improved the device photovoltaic properties. REFERENCES [1] T. Yoshitake, M. Yatabe, M. Itakura, N. Kuwano, Y. Tomokiyo, and K. Nagayama, Appl. Phys. Lett. 83, 3057 (2003). [2] K. Takarabe, H. Doi, Y. Mori, K. Fukui, Y. Shim, N. Yamamoto, T. Yoshitake, and K. Nagayama, Appl. Phys. Lett. 88, 061911 (2006).

G2-S5.19
Growth and Low-Temperature Annealing Effects of n-Type β-FeSi2/p-Type Si Heterojunctions on the Rectification and Photovoltaic Properties. (#801) Mahmoud Shaban, Keita Nomoto, Tsuyoshi Yoshitake; Department of Applied Science for Electronics and Materials, Kyushu University, Fukuoka, Japan.

The orthorhombic iron disilicide (β-FeSi₂) is a semiconducting material having several interesting electrical and optical properties [1, 2]. Crystalline β-FeSi₂ has an optical absorption coefficient larger by at least two orders of magnitude than that of crystalline Si at 1.5 eV. Moreover, β-FeSi₂ is an eco-friendly material due to its nontoxicity and its elements (Fe and Si) are abundance in earth's crust. n-Type β-FeSi₂ thin films were as-grown on p-type Si(111) substrates at different substrate temperatures by the facing-target DC sputtering method. It was confirmed that 600°C [3, 4] is the minimum growth temperature that required for the epitaxial growth of single-crystal β-FeSi₂ thin films. The β-FeSi₂/Si heterojunctions exhibited rectifying behaviour similar to that of conventional p-n heterojunctions. On the other hand, their conversion efficiency was low due to their large leakage currents. The main source of this leakage is the Fe atoms that diffuse from the β-FeSi₂ to the Si substrate during the film preparation at 600°C. In order to reduce the Fe atom concentration that diffused into the Si substrate, a low-temperature annealing was conducted at different temperatures from 200-400°C. The annealed heterojunctions showed markedly reduction in the leakage current and improvement in the photovoltaic properties at annealing temperature of 300°C. REFERENCES [1] M. C. Bost, J. E. Mahan, J. Appl. Phys. 64 (1988) 2034. [2] D. N. Leong, M. A. Harry, K. J. Reeson, K. P. Homewood, Appl. Phys. Lett. 68 (1996) 1649. [3] T. Yoshitake, M. Yatabe, M. Itakura, N. Kuwano, Y. Tomokiyo, and K. Nagayama, Appl. Phys. Lett. 83, 3057 (2003). [4] M. Shaban, K. Nakashima, W. Yokyama, T. Yoshitake, Jpn J. Appl. Phys. 46 (2007) L667.

G2-S5.20
Elimination of Hydrogenation-induced Contact Degradation of Evaporated Poly-Si Thin-Film Solar Cells on Glass . (#397) Lei Shi, Armin Gerhard Aberle; ARC Photovoltaics Centre of Excellence, The University of New South Wales, Sydney, Australia.

PURPOSE OF THE WORK The purpose of this paper is to present a detailed contact resistance study on poly-Si thin-film solar cells on glass, whereby the silicon film was deposited by PECVD or electron-beam evaporation. We believe that this work is the first systematic study of the contact resistance of aluminium on evaporated poly-Si thin-film solar cells. APPROACH Three types of thin-film solar cells under development in our group - EVA, ALICIA and PLASMA - were investigated. First, a doped a-Si film was deposited onto the glass (or onto a seed layer on the glass) using PECVD or e beam evaporation. The a-Si film was then crystallised by either solid phase crystallisation (SPC) or solid phase epitaxy (SPE). Then followed two post-crystallisation steps (defect anneal, hydrogenation) to boost the efficiency of the solar cells. The Transmission Line Model (TLM), including in-line TLM and circular TLM, was then employed to measure Rc. We found that hydrogenation alters the surface properties of the different types of solar cells to different extents, which causes non-ohmic and non-uniform contacts. Two approaches, thermal annealing and etching with coloured HF, were employed to solve this problem. In order to reveal the general features of Al contacts on evaporated Si PV materials more accurately, uniformly doped poly-Si films on glass (i.e., no p-n junctions) were investigated. No hydrogenation step was applied to these samples to avoid the hydrogenation induced damage of the surface region. SCIENTIFIC INNOVATION AND RELEVANCE This is the first systematic contact resistance study on evaporated poly-Si thin-film solar cells on glass. Surface degradation was observed due to the hydrogenation step, which was found to depend strongly on the type of Si material investigated. Two methods were employed and optimised to satisfactorily solve this problem. The work provides significant insight into the general features of Al contacts on evaporated Si material for PV applications. RESULTS AND CONCLUSIONS Good ohmic contacts were achieved, with Rc of below 10-3 &Omegacm2 for both emitter and BSF layers of all types of UNSW thin-film solar cells investigated in this work. Rc was found to depend strongly on the resistivity of the poly-Si film, with excellent Rc values of less than 10-5 &Omegacm2 realised for p-Si films with resistivities below about 10 m&Omegacm and for n-Si films with resistivities below about 1 m&Omegacm. Rc values of annealed samples were found to be stable during storage in air ambient for 15 months. In summary, low contact resistance values have been achieved for UNSW poly-Si thin-film solar cells on glass, which are believed to offer sufficient room for optimising the other parameters of our metallisation scheme. Al is demonstrated to be well suited for evaporated poly-Si materials for PV applications.

G2-S5.21
High-Efficeincy, 0.8 M2 Thin-Film SI Modules Fabricated by a Batch. (#522) Porponth Sichanugrist, Nopphadol Sitthiphol; Institute of Solar Energy Technology Development, NSTDA, Thailand.

A large batch PECVD system has been developed in order to deposit a-Si/a-SiGe and a-Si/?c-Si solar cell on 20 pieces of SnO2-coated glass with the area of 0.8 m2. Lower manufacturing cost can be expected with this one large batch system due to lower equipment cost, higher throughput and better gas utilization. Newly designed electrode has been used in order to achieve uniform ?c-Si film on 0.8 m2. Beside ?c-Si n-layer, p(?c-SiO) has been used in the p-layer of the bottom cell in order to increase the module performance. Alcohol flush has been used in order to minimize the boron contamination. Furthermore, ZnO/Ag back electrode has been used instead of ZnO/Al electrode. Up to now, initial module efficiency of more than 6.5% and 7% have been achieved with a-SiO/a-Si/a-SiGe and a-Si/?c-Si cell structure. The update work of higher performance development will be reported in the conference.

G2-S5.22
The a-Si PV/T Collectors for Solar Air-Conditioning System in Thailand. (#700) Suriya Sukarin, Porponth Sichanugrist, Kamonpan Chumpolrat, Thipjak Nualboonrueng; National Science and Technology Development Agency, Thailand.

This work is concerned about heat source of adsorption chiller machine . This machine is installed at Ministry of Science and Technology, Thailand.The PV/T (Photovoltaic/Thermal) solar system, the main source , produces 17.4 kW of heat supplied to adsorption chiller machine. The a-Si PV collector area of 93 m2 is installed on roof of building with 6.3 kW of electricity capacity. Electricity generated is connected to grid by DC to AC inverter. This system produces 12,000 liter per day at 60? C. The auxilliary heat is the heat recovery system. This system has two of 30 kW heat exchanger connected to 190 ton of conventional air-conditioning system (Compressed refrigerant). In summary the system can produce 77.4 kW of heat supplied for 52.3 kW adsorption chiller machines at COP 0.45. The adsorption chillers produce cold water 12?C for a surface of 230 m2 8 hour per day. This system can save cost for operate conventional air-conditioning system for this room around 15,120 Thai baht per month.

G2-S5.23
Texturing Crystalline Silicon Wafers using Hydrogen Bubbles as Pseudo Etch Mask in Alkaline Etchants without Surfactant. (#517) Ann-Kuo Chu, Jung-Shin Wang, Zhao-Yuan Tsai, Chao-Kuei Lee; Institute of Electro-optical Engineering and Semiconductor Technology Research and Development Center National Sun Yat-sen University, Kaohsiung, Taiwan.

Surface texturization of (100)-oriented crystalline silicon wafers is a frequently used technique in modern solar cell processing to reduce optical reflections. Alkaline etchants such as sodium hydroxide (NaOH) or potassium hydroxide (KOH) at low concentration in water expose Si {111} faces resulting in square-based pyramids randomly distributed over the cell surface. In general, isopropyl alcohol (IPA) is added to the alkaline etchants to improve the uniformity of the random pyramid texture. Though IPA is expensive, it is essential for the process because it removes hydrogen bubbles (with 2-3 mm diameter) sticking on the silicon wafer by improving the wettability of the wafer surface. In a typical texturing condition, high concentration of IPA is required to obtain uniform pyramid textures. However, the etching rate of silicon decreases drastically with the IPA concentration. In this work, a simple and cost effective approach for texturing crystalline silicon wafers is proposed. The advantage over conventional texturization processes is the fact that no surfactant is added in alkaline etchants. Hydrogen bubbles which are regarded as residuals in conventional texturization processes are utilized in this approach as etch masks on silicon surface. This is accomplished by placing a metal grid with suitable openings on silicon wafers to trap the hydrogen bubbles created during the etching. Pyramids with a size ranging from 6 to 9 &mum are uniformly fabricated using this method. The etching experiments were carried out using 3", p-type, <100> oriented, crystalline silicon wafers with resistivity of 1-3 &Omegacm. The wafers were etched in KOH (1 wt%) solutions at different temperatures for 10, 15, and 20 minutes. No surfactant was added in the etchants during the processes. The hydrogen bubbles produced during the etching were trapped on the wafer surfaces utilizing the stainless steel metal grids with different square openings. To optimize the size and uniformity of the pyramid, we used four different metal grids with 1, 1.5, 2, and 3 mm square opening for texturing. The separation between the metal grids and the silicon wafers was kept at 1 mm. After the etching process, the surface morphology and the optical reflection of the silicon wafers were characterized by a surface scanning electron microscope and an integrating sphere/spectroradiometer in the wavelength range from 400 nm to 1100 nm. The pyramids fabricated using the proposed approach is dependent not only on the conditions of the KOH etchants but also on the structures of the metal grids to silicon wafers. All of these factors must be considered to obtain pyramid structures with satisfactory optical performance. It was observed from the experiments that the size and surface coverage of pyramid structures increase with increasing etchant temperature and the dimensions of the grid openings up to 2 mm. Using the metal grid with 2 mm square opening and 1 mm separation between the grid and the wafers, our optimized process lasted for 20 minutes at 90°C, 1% KOH solution without adding any IPA. After texturing, we obtained reliable and uniform pyramid structures with an average weighted reflectance of 15.1% and the surface overall coverage of better than 95%.

G2-S5.24
Plasma Dry Etching for Selective Emitter Formation in Multi-Crystalline Silicon Based Solar Cell. (#1187) Mario Tucci, Luca Serenelli, Pierino Martufi, Massimo Izzi, Luisa Pirozzi; ENEA Research Center Casaccia, Roma, Italy.

Selective emitter formation on top of silicon based solar cell has recently becoming a common technique to enhance the blue response of solar cell. This concept is based on differential doping of the solar cell front emitter allowing higher doping region under metal grid contact to reduce series resistance, and lower doping region off grid to promote a better surface passivation. To develop this concept several processes have been proposed and investigated [1, 2], but their complexity often due to realignment necessity between front grid and high doped emitter regions, still limits a strong industrial interest. To overcome the realignment problem in this work we suggest a novel procedure based on a self alignment thought. The idea is based on a plasma dry etching procedure of the emitter region using the metal grid of the cell as mask. In particular the plasma treatment can reduce the thickness of an homogeneous highly doped emitter reducing its thickness as well as its Ro-sheet. Starting from a RCA cleaned 1 ohm.cm 12.5x12.5 cm2 textured mc-Si n-type wafer, we have diffused a deep p+ region as for top emitter and a n+ region as for back surface field. After a wet chemical cleaning to remove the post diffusion oxide formation of wafer both sides, Aluminum front grid and Silver back contact have been deposited by screen printing procedure and thermally annealed to ensure both front and rear ohmic contact. Then the cell front side has been treated by NF3 in a 13.56 MHz RF Plasma Enhanced Chemical Vapor Deposition chamber to dry etch and reduce the emitter thickness following a particular recipe already developed by the same group [3]. Argon dilution in the gas mixture during the dry etching plasma process is the key point to reach a controlled etching rate useful to reduce the emitter thickness. Since the Al screen printed grid is a mask for the plasma process, only the area outside the Al grid has been etched resulting in a selective distribution of the emitter region. In the same plasma system an a-Si:H/SiNx antireflection-passivation double coating has been deposited on the base of previous experience [4] on top of the solar cell. Then the cell has been annealed for 10 min at 350 ?C to enhance the passivation performances [5]. Finally the edge isolation has been ensured cutting by dicing machine. This concept can be adopted on both type of doped silicon wafers and in this work we show a comparison of solar cell performances fabricated by this proposed process starting on both p-type and n-type multi-crystalline silicon wafers. Since there is no necessity of realignment step in the proposed selective emitter fabrication process it can be useful and appealing for industrial PV manufacturing. [1] L. Pirozzi, S. De Iuliis, M. Tucci, L. Serenelli, F. Ferrazza, M. Zarcone, B. Terheiden, A. Zerga, N. Eisenberg, R. Kopecek, S. Dewallef, Y. Veschetti. Proceeding of XXI European Photovoltaic Solar Energy Conference Dresden (2006) 1254. [2] T. Pletzer, E. Stegemann, L. Janssen, H. Windgassen, D. L. B?tzner and H. Kurz, Proceeding of XXII European Photovoltaic Solar Energy Conference Milan (2007) 1604. [3] M. Tucci, L. Serenelli, S. De Iuliis, E.Salza, L. Pirozzi. Proceeding of XXI European Photovoltaic Solar Energy Conference (2006) 1250. [4] M. Tucci, L. Serenelli, S. De Iuliis, M. Izzi, Thin solid Films 515 (2007) 7625. [5] M. Tucci and L. Serenelli; accepted for publication in Thin Solid Films, available on website since 10-1-2008.

G2-S5.25
Dielectric Bragg Back Reflecting Mirror in a-Si:H / c-Si Heterostructure Solar Cell. (#1185) Mario Tucci¹, Luca Serenelli¹, Pierino Martufi¹, Enrico Salza¹, Giampiero de Cesare², Domenico Caputo², Matteo Ceccarelli²; ¹ENEA Research Center Casaccia, Roma, Italy ; ²Electronic Engineering Department, University of Rome 'Sapienza', Italy.

The amorphous/crystalline silicon (a-Si/c-Si) heterojunction have recently attracted new interest due to get higher open circuit voltage with respect to homojunction and the low temperature fabrication processes. Reducing the wafer thickness these characteristics become a necessity, together with the requirement of an excellent passivation of both front and back crystalline wafer surfaces as well as an optical confinement of light traveling inside the solar cell. To this aim a back dielectric reflecting mirrors can be adopted in the rear side of the solar cells, together with a local process of laser fired back Al contact. Taking advantage of a-Si/SiNx passivation properties of c-Si surface, a Bragg reflector configuration can be formed on the rear side of the c-Si wafer by plasma enhanced chemical vapor deposition tuning and alternating several couples of a-Si/SiNx and choosing their thicknesses to maximize the reflectance inward the c-Si wafer in the NIR spectrum. We have found out that 4 couples of a-Si/SiNx (66 nm and 130 nm respectively thick) followed by a 2 mm Al layer are able to produce a reflectance of 99% inward the c-Si wafer at 1050 nm wavelength of light. So in this work we have introduced this dielectric mirror on the rear side of an ITO/n-a-Si/i-a-Si/p-c-Si heterostructure solar cell. In particular it has been dimensioned by an optical model of the Bragg structure and its reflectance characteristics have been experimentally evaluated with the aid of XOP software [1]. The cell back contact has been ensured by a Nd-YAG pulsed laser treatment of a boron rich film and Aluminum layer, deposited on the rear side of the Bragg mirror, by spin-on coating and e-beam evaporation respectively. Optimizing the laser treatment a back contact has been obtained by B/A localized diffusion into the silicon wafer through the dielectric mirror. The cell front contact has been enhanced by chromium silicide CrSi formation on top of the n-a-Si layer and ITO deposition followed by an Ag grid [2]. Finally the heterostructure solar cells have been characterized by current voltage (I-V) measurements both in AM1.5G and in dark conditions, and by quantum efficiency (QE) in the range of 350-1200 nm. A Voc of 690mV and 90% of IQE at 1050 nm have been achieved. [1] www.esrf.eu/computing/scientific/xop2.1/ [2] M. Tucci, G. de Cesare, J. Non-Cryst. Solids 338 (2004) 663

G2-S5.28
Multiple Exciton Generation Quantum Dots Intermediate-Band Solar Cells Modified by MeV Si Ion Bombardment. (#1166) Bangke Zheng, Center for Irradiation of Materials (CIM), Department of Physics, Alabama A&M University, USA.

In multilayer of GaAs p- i-n type photovoltaic device, self assembled CdSe quantum dots are grown by molecular beam epitaxy in the Stranski Krastanov growth mode, and the miniband formed in quantum dots provide intermediate band and make absorption of sub- bandgap photon possible, yielding higher solar conversion efficiency in wide spectrum, also result in multiple exciton generation. The nanoscale clusters formed by MeV Si ion beam bombardment in p type absorber layer further increase the absorption of photon, increasing the creation of exciton, increasing photocurrent.

G2-S5.29
Cu-Chalcopyrite Thin Films Prepared by Chalcogenization of Mechanically Milled Nanoparticles. (#706) Ki Hyun Kim, H.-C. Michael Jin, Rajalakshmi Sundaramoorthy, Alex Alphonse; Materials Science and Engineering of University of Texas, Arlington, USA.

CuInGaSe₂ (CIGS) has emerged as a promising material for high efficiency and low cost thin film solar cells. Power conversion efficiencies over 19 % on small scale laboratory cells and 13 % on large area modules were achieved. Various methods such as co-evaporation, selenization and other low cost non-vacuum processes were used to deposit CIGS absorbers. However, low cost non-vacuum methods such as paste coating and inkjet printing of CIGS nanoparticles and subsequent chalcogenization offer further cost reduction due to higher material utilization. Many groups reported the fabrication of CIGS absorbers and devices using CIGS nanoparticles derived from oxides and nitrate based starting materials. Recently in December, 2007, the author has presented a paper about the preparation of CuInSe₂ (CIS) absorber layer by non-vacuum process at 17-PVSEC in Fukuoka, Japan. In the paper, CIS absorber layer was prepared by paste coating followed by the selenization of Cu₂Se and In₂Se₂ nanoparticles. The mixture of the powder was ball-milled with propylene glycol binder for 18 hours. The mixed paste was coated manually onto a Mo-coated glass substrate by a doctor-blade method. While Cu₂Se films with large grains were grown densely when the substrate temperature was 550°C and the Se source temperature was 450°C, the CIS film made under similar condition was relatively porous and small-grained. The thickness of CIS films was in the range of 3 - 6 µm and a MoSe₂ interlayer was observed between CIS and Mo. Our result suggested that the appropriate understanding of Cu₂Se film growth will enhance the grain growth of CIS and potentially CIGS film. In this study, further experiments will be preformed to improve the morphology of CIS thin films by varying Cu/In ratio and commercially available CIS powder will be also tried to fabricate CIS thin films. Detailed characterization of the films deposited using SEM/EDS, XRD, and XPS together with solar cell performance will be included in the presentation.

11:00 AM *G3-S2.1 (invited)
Preparation of CIGS Solar Cells by Non-vacuum Particles-based Techniques. (#353) Takahiro Wada, Department of Materials Chemistry, Ryukoku University, Kyoto, Japan.

CuInSe2 (CIS) and its solid solutions with Ga and S are excellent thin-film photovoltaic materials. However, the vacuum deposition processes, such as physical vapor deposition and sputtering, typically used to fabricate CIS thin-film photovoltaic (PV) devices are complex and expensive. Recently, some research groups proposed various kinds of non-vacuum deposition techniques used to fabricate CIS solar cells [1-4]. I will review non-vacuum deposition techniques used to fabricate CIS solar cells. One class of non-vacuum deposition technique is a non-vacuum processing of particulate materials. This non-vacuum processing of particulate materials offers a simple, inexpensive alternative to fabricating high-efficiency CIS solar cells. Suitable particulate materials can be prepared by a variety of techniques, including precipitation, pulverization, aerosol pyrolysis, vapor condensation, etc. Particulate precursors can be deposited in thin layers by simple techniques such as screen-printing, spraying or dip coating. Porous precursor layers can be sintered into dense polycrystalline films by atmospheric-pressure heat-treatment [2]. Recently some topics concerning non-vacuum particles-based techniques for the fabrication of CIGS solar cells were reported [5-7]. And then, I will introduce our research. Recently, Ryukoku University group successfully synthesized CIS and its solid solution, Cu(In,Ga)Se2, CuIn(S,Se)2 and (Cu,Ag)InSe2 from elemental powders by means of a mechano-chemical process (MCP) without any additional heating [8-10]. The MCP is a process that induces physical and/or chemical change by mechanical energy, such as pulverization, friction or compression. This method has some advantages for the mass production of CIS solar cells, such as high energy efficiency, high productivity and short processing cycle time. We prepared CIGS powders suitable for the screen-printing ink by the MCP. Particulate precursors were deposited in thin layers by screen-printing technique. Then, the porous precursor layers were sintered into dense polycrystalline films by atmospheric-pressure heat-treatment [11]. References [1] Vijay K. Kapur et al., Thin Solid Films 431-432, 53-57 (2003). [2] Chris Eberspacher et al., Thin Solid Films 387, 18-22 (2001). [3] M. Kaelin et al., Thin Solid Films 480-481, 486-490 (2005). [4] S. Taunier et al., Thin Solid Films 480-481, 526-531 (2005). [5] J. K. J. van Duren et al., MRS Proceedings 1012, Y05-03 (2007). [6] J. K. J. van Duren et al., Technical Digest of PVSEC-17, PL5-3 (2007). [7] V. Kapur et al., Technical Digest of PVSEC-17, 6O-A11-05 (2007). [8] T. Wada et al., Thin Solid Films 431-432, 11-15 (2003). [9] T. Wada and H. Kinoshita, J. Phys. Chem. Solids 66, 1987 (2005). [10] S. Nomura et al., MRS Proceedings 1012, Y03-15 (2007). [11] T. Wada et al., phys. stat. sol. (a) 203, 2593-2597 (2006).

11:30 AM G3-S2.2
Sonochemical Synthesis of CuInSe₂ (CIS) Nanoparticles and Preparation of CIS Thin Films for Solar Cells. (#257) Juyeon Chang, Jae-Eok Han, Soon Cheol Jeong, Hyeon Ju Ahn, Hye Jin Nam, Duk-Young Jung; Department of Chemistry, Sungkyunkwan University, Republic of Korea.

Sonochemical synthesis of CuInSe₂ (CIS) nanoparticles and subsequent fabrication of CIS absorber layer from the nanoparticles were investigated. According to this approach, nanosized CIS powder was synthesized by ultrasound irradiation under ambient pressure at 100 ^oC and characterized by powder X-ray diffraction (XRD), scanning electron microscopy (SEM), optical absorption spectroscopy and energy-dispersive X-ray (EDX) analyses. The XRD spectra of the prepared samples were indexed to reference patterns for the chalcopyrite crystal structure. The CIS samples have particle sizes less than 50 nm, which were confirmed by SEM and XRD peak widths. Optical measurement and elemental analysis by EDX also support that the product has a single phase chalcopyrite structure. Subsequently, the nanoparticulate CIS layer was deposited by non-vacuum techniques such as spin coating and bar coating. After thermal processing of the precursor films under a selenium ambient, the conversion of nanosized CIS particles into microstructured layer and the prepared CIS layer have been investigated by XRD, SEM and other measurements.

11:45 AM G3-S2.3
Selenization of Copper-Indium-Boron Thin Film Photovoltaic Absrober Materials. (#274) Ned Ianno¹, Rodney Soukup¹, Chad Kamler¹, Shuchi Sharma¹, James Huguenin-Love¹, Jiri Olejnicek¹, Christopher Exstrom², Scott Darveau²; ¹Department of Electrical Enginering, University of Nebraska-Lincoln, USA ; ²University of Nebraska at Kearney, USA.

It has recently been established that the ideal bandgap for terrestrial photovoltaics is 1.37 eV and the bandgap for CuInSe₂ is only 1.04 eV. Various elements have been used for substitutions to increase the bandgaps, such as Ga and Al for In and S for Se. To date, the substitution of Ga for In has been the most successful. In order to achieve the ideal bandgap, a substitution level of 67% Ga for In would be needed. At that level, the efficiency for CIGS solar cells is reduced to approximately 14%. Boron, an even smaller atom, should require less atomic substitution than either Ga or Al in order to achieve a wider bandgap. We have previously reported a simple approach used in determining the bandgap of CuBSe₂, and using this and equation 1, it is estimated that the amount of B substitution needed to reach a bandgap of 1.37 eV is only 16.9%. E_g(x) = (1-x) E_g(A) + x E_g(B) -bx(1-x) (1) In this equation A is In and B Boron. In order to fabricate a thin film of CuIn_xB_1-xSe₂, Cu, In and B were deposited from a variety of sputtering targets in a variety of manners. The targets used were pure Cu, In, and B; a Cu_0.45In_0.55; and a Cu₃B₂ target. Films were deposited simultaneously and sequentially. After deposition these films were post selenized in a separate vacuum chamber employing solid selenium in close proximity to the substrate where the selenezation was driven by a quartz halogen based rapid thermal processing system. Analysis of these films was accomplished using Raman spectroscopy, X-ray diffraction, spectroscopic ellipsometry, and Auger electron spectroscopy. The results presented illustrate the difficulties encountered in making a CIBS material using the post selenization technique, namely the migration of the copper, indium and selenium material to the front of the film during the selenization process, leaving the boron to accumulate at the substrate surface. This migration was discovered by a set of experiments using Auger analysis which will be illustrated in this presentation. With the difficulties encountered in the post selenization approach, the materials were also deposited in a selenium atmosphere. This was accomplished by installing a small effusion cell, which was operated at 350 ^oC, in the sputtering chamber. The sputter rates of the Cu, In and B are adjusted to yield films of appropriate composition. Analysis of these films was accomplished using Raman spectroscopy, X-ray diffraction, spectroscopic ellipsometry, and Auger electron spectroscopy and the results are presented.

12:00 PM G3-S2.4
Control of the Na Density in Cu(In,Ga)Se₂ Thin Films and Application in High-Efficiency Flexible Solar Cells. (#479) Shogo Ishizuka, Akimasa Yamada, Paul Fons, Hajime Shibata, Keiichiro Sakurai, Koji Matsubara, Shigeru Niki; Research Center for Photovoltaics (RCPV), National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, Japan.

Sodium (Na) has been widely known as a beneficial dopant for Cu(In1-xGax)Se2 (CIGS) absorber layers that serves to improve p-type conductivity and concomitant cell efficiencies. We have developed a novel technique to control the Na doping level in Cu(In1-xGax)Se2 (CIGS) thin films using a stable Na-compound as a dopant material. Using this technique, a 17.6%-efficiency flexible CIGS solar cell with an open circuit voltage of 0.642 V, a short circuit current density of 36.2 mA/cm2, and a fill factor of 0.759 (AM 1.5G, active area 0.475 cm2) has been demonstrated. This efficiency is the highest value reported to date for flexible CIGS solar cells. In this study, CIGS films were grown by the three-stage process using a molecular beam epitaxy (MBE) apparatus on flexible substrates such as metal foils for device fabrication and rigid alkali-free glass substrates for analysis. The Na-compound layers were deposited on the substrates before CIGS growth. Detailed experimental procedures will be presented at the conference. Variations in CIGS film properties with doping levels and related effects on solar cell performance have been investigated. Secondary ion mass spectroscopy (SIMS) measurements showed that the amount of Na incorporated into CIGS films could be controlled by adjusting the Na-compound layer thickness. The photovoltaic performance of CIGS cells dramatically improved with use of the Na-compound layers and the extent of improvement was found to depend on the Na-compound layer thickness. In addition to this, the quantum efficiency curves of solar cells fabricated using CIGS films grown with these Na-compound layers showed enhanced absorption in the long wavelength region in comparison with cells fabricated using non-doped CIGS films or conventional CIGS films grown on soda-lime glass substrates. The Na doping technique developed in this study can be applied for a variety of substrates such as flexible metal foils, polyimide films, and other rigid substrates as well, and will lead to a wide variety of types of high-efficiency flexible CIGS cells and modules

12:15 PM G3-S2.5
ZnO Nanowire-CuInS₂ Nanocomposites as Light-Harvesting Assemblies: Photoinduced Charge-Transfer Interactions. (#498) Kuan-Ting Kuo¹, San-Yuan Chen¹, Bing-Ming Cheng², Chin-Ching Lin³; ¹National Chiao Tung University, Hsinchu, Taiwan ; ²National Synchrotron Radiation Research Center, Taiwan ; ³Industrial Technology Research Institute, Taiwan.

A new-type quantum-dot-sensitized solar cell consisting of CuInS2 semiconductor nanocrystals and ZnO nanowires was developed and studied under different atmospheres. After An n-type layer of tangled ZnO nanowires was grown from a aluminum-doped tin oxide cooducting substrate, CIS quantum dot was atteched to the surface of ZnO nanowires by modifying surface function with mercaptoproponic acid. The results show that depostion of CIS nanocrystals can remarkably increase the steady state photocurrent but without increasing dark current, revealing that CIS apparently influences the photoconductivity of ZnO nanowires by minimizing its interaction with oxygen in air as well as decreasing the possibility of photogeneration-recombination of electron-hole pairs. Similar result was also obtained under nitrogen atmosphere that evidanced additional trap decreasing photoconductivity of ZnO from oxygen. Furthermore, solar power conversion efficiency of ZnO nanowires substrate was enhanced by a factor of 7 times after deposition of CIS under AM1.5 illumination.

LUNCH 12:30 PM - 2:00 PM

2:00 PM *G3-S3.1 (invited)
Fabrication of Cu(InGa)Se₂ Thin-Film Solar Cells Grown with Ionized Ga Source. (#656) Akira Yamada, Tetsuya Nakashiba, Li Zhang, Makoto Konagai; Qunatum Nanoelectronics Research Center, Tokyo Institute of Technology, Japan.

Cu(InGa)Se2 (CIGS) is a promising compound semiconductor as an absorber layer for solar cells, and efficiencies of CIGS thin-film solar cells exceeded 19% with a band gap energy of approximately 1.15 eV. For further improvement of the efficiency, the band gap energy should be increased up to 1.4 eV, however efficiencies of solar cells monotonously decrease with increasing band gap energy by the Ga addition. This is the major challenging issue for researchers. In the paper, we have proposed a new growth concept for CIGS films in which ionized Ga source is used as an elemental source. The newly developed K-cell which has an ionization system was used for growth of CIGS films. The K-cell was designed to ionize Ga by imposing ionize-voltage and to accelerate Ga by an acceleration voltage. The high energy Ga releases their energy on the CIGS surface, resulting in the local heating of the growing surface. Thus, the migration enhancement is expected by using ionized Ga, and the film quality is improved. The CIGS films were grown by the co-evaporation method, and the typical growth temperature was 520 C. The films were characterized by SEM, XRD, and Raman measurement, and the defect density was evaluated by the admittance spectroscopy (AS). Finally CIGS solar cells with a CBD-CdS buffer layer and a MOCVD-ZnO transparent electrode were fabricated. CIGS films deposited with ionized Ga showed equivalent quality to CIGS films deposited with normal Ga by XRD and Raman measurements. The peaks originated from Cu-Se and In-Ga-Se were not detected for both CIGS films, and the films were preferentially oriented toward (112)-plane. The FWHM of (112) peak of the film deposited with ionized Ga was 0.2 degree, and this was almost equivalent value of the normally deposited CIGS films. The Raman spectra of both samples showed strong A1 mode and weak B2 and E modes, showing the high quality CIGS films. From the SEM measurement, we observed slight improvement of grain growth of the films deposited with ionized Ga. The quality differences between the films were observed in the electrical characterization. The samples showed the distribution of density of states centered at about 220 meV above the valence band maximum measured by the AS. The defect densities were 1.1e14 cm-3 and 0.6e14 cm-3 for the samples grown with normal Ga and ionized Ga, respectively. Furthermore, the reduction of the reverse saturation current was observed in the I-V measurements for CIGS solar cells grown with ionized Ga. These results demonstrated that utilization of ionized Ga is promising in order to improve the film quality of CIGS thin films. Finally we fabricated CIGS thin-film solar cells with and without the Ga ionization technique. By using ionized Ga the improvement of Voc and FF was observed, and the efficiency of 15.1 % has been achieved with a Ga content of 25% under AM 1.5 irradiation. These results clearly showed the potential of newly proposed Ga ionization method for growth of high quality CIGS films.

2:30 PM G3-S3.2
Effect of Mo Back Contact Thickness on the Properties of CIGS Solar Cells. (#594) Yukiko Shimizu Kamikawa¹, Shuuhei Shimada², Manabu Watanabe², Akimasa Yamada¹, Keiichiro Sakurai¹, Shogo Ishizuka¹, Hironori Komaki¹, Koji Matsubara¹, Hajime Shibata¹, Hitoshi Tampo¹, Keigou Maejima¹, Shigeru Niki¹; ¹Thin Film Compound Semiconductor, Research Center for Photovoltaics (RCPV), National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, Japan ; ²Tokyo University of Science, Japan.

Thin-film Cu(In,Ga)(Se,S)² (CIGS) is a promising material for the production of highly efficient and low cost PV modules. Conversion efficiencies up to 19.52 % for small area laboratory cells have been demonstrated and some companies have already launched into mass productions [1-4]. The sputter-deposited Mo back contact has been widely used for CIGS solar cells taking advantages of its ideal ohmic contact. However, rising demand as a feedstock of steel in China inflate the price of Mo rapidly in these 5 years. Since, the purpose of this work is to reduce the thickness of Mo back contact to dispel the rising price. We reduced the thickness of Mo and investigated the effect on the photovoltaic properties of CIGS solar cells. High quality CIGS thin films were deposited by the three-stage process using a molecular beam epitaxy (MBE) apparatus. The highest growth temperature was set to 520 ^oC measured by thermocouple. Mo back contacts were deposited by RF magnetron sputtering at 80 ^oC on the soda lime glass (SLG). Ar pressure and RF power were set to 1.5 Pa and 100 W, respectively. The target to substrate distance was about 7 cm. Six holders were attachable and rotating to get good uniformity during sputtering. The deposition rate was 270 nm / hour. The surface morphology of Mo back contact showed scale like shape and grain size at the surface (longest diameter) gradually increased with the thickness of Mo increased. The grain size changed from 20 nm to 180 nm as thickness of Mo was increased from 70 nm to 600 nm. A reduction of Na concentrations in CIGS absorbers was observed when the grain size was very small as ~10 nm or thickness of Mo was thicker than 200 nm. The structure of Mo could affect the diffusion of Na to CIGS absorber from SLG. In this study, Na concentration became maximum around the Mo thickness with 200 nm. The electrical resistivity showed constant value of 5x10^-5 ohm-cm independent of thickness of Mo. Since the sheet resistance changed inversely proportional to the thickness, too thin Mo became high series resistance which mainly deteriorate the fill factors. Also, nonnegligible reduction in V_OC was observed in low Na samples with thick Mo and some of very thin Mo. As a result of effects of both Na concentration and series resistance, highest conversion efficiency of as high as 15.8 % was obtained when we reduced the thickness of Mo toward 200 nm which is 1/4 times thick of our default value. [1] R. N. Bhattacharya et. al., Appl. Phys. Lett. 89 (2006) 253503. [2] K. Kushiya, Technical Digest of the International PVSEC-17, Fukuoka, Japan, 2007. [3] K. Matsunaga, Technical Digest of the International PVSEC-17, Fukuoka, Japan, 2007. [4] B. Dimmler and R. Wachter, Thin Solid Films 515 (2007) 5973.

2:45 PM G3-S3.3
High Electron Injection Structure in Hybrid Solar Cells. (#826) Hung-Chou Liao¹, Chin-Ching Lin², San-Yuan Chen¹; ¹National Chiao Tung University, Hsinchu, Taiwan ; ²Materials and Chemical Engineering Research Laboratories, Taiwan.

An ordered organic-inorganic solar cell architecture based on CdS QD-decorated ZnO nanorod arrays encased in the hole-conducting polymer Poly(3-hexylthiophene-2,5-diyl) (P3HT) was investigated. In this work, CdS quantum dot with different sizes, capped with mercaptopropionic acid, were attached to the surface of the nanorods. TEM analysis shows that the CdS QDs are uniformly attached on the surface of ZnO nanorods. After the structure was devloped, the PHT was dipped into the nanostructure. The results show that the CdS QDs on the ZnO nanorods significantly increase the voltage and fill factor relative to devices compared to that without CdS QDs. The photovoltaic device based on the nanorod structure with CdS QDs decorated yields power conversion efficiency over 4 times greater than that for a similar device without CdS QDs. The best device yields a short circuit current density of 1 mAcm-2 under AM1.5 illumination (100 mW cm-2), resulting in a power conversion efficiency of 0.21% The increase demonstrates that the modification with CdS QDs can serve not only as a photosensitizer but mainly as an energy funnel and/or an electronic mediator to significantly improve the electron injection efficiency from P3HT to ZnO nanorods.

AFTERNOON BREAK 3:30 PM - 4:00 PM

4:00 PM *G3-S4.1 (invited)
Materials and Devices for Photovoltaic Electricity Generation and Potential for Climate Change Impact. (#1223) Martin Andrew Green, School of Photovoltaic and Renewable Energy Engineering, The University of New South Wales, Sydney, Australia.

Most solar cells to date have been based on crystalline silicon wafers, similar to those used in microelectronics. These cells have performed exceptionally well and still show considerable potential for further cost reduction. However, solar cells based on thin-films of photoactive material show more potential to reach to costs required to make a large scale energy impact in the future. Progress in material and devices for such large scale application will be described. Work within the author's group on silicon thin-film photovoltaics and on the use of silicon quantum dots for 'all-silicon tandem' and 'hot carrier' cells will also be described.

4:30 PM *G3-S4.2 (invited)
Photocatalytic Oxide Materials. (#1452) John Stride, The University of New South Wales, Sydney, Australia.

TBA

5:00 PM G3-S4.3
Titanium Dioxide-Based Oxide Materials for Solar Eenergy Conversion. (#1222) Leigh R. Sheppard, Tadeusz Bak, Janusz Nowotny; Centre for Materials Research in Energy Conversion, The University of New South Wales, Sydney, Australia.

The commonly available TiO₂ may be used as a raw material for the processing of a wide range of TiO₂-based materials for environmentally friendly applications. These materials may be applied to harness solar energy for the production of solar-hydrogen by water splitting and water purification as well as for the production of cheap n- and p-type semiconductors. Their processing is based on defect engineering, which allows to impose controlled properties that are desired for specific applications. It is shown that TiO₂ involves a number of lattice species involving both ionic and electronic species in their normal lattice sites and interstitial sites. The properties of TiO₂ may be imposed in a controlled manner by specific combinations of these species. This may be achieved by doping TiO₂ with oxygen and aliovalent ions. A general chemical formula, representing the composition of TiO₂ and its solid solutions, is derived. This formula is reflective of the effect of both thermodynamically reversible defects and extrinsic defects on properties. It is shown that defect chemistry may be used as a framework for the imposition of controlled properties that are desired for specific applications.

5:15 PM G3-S4.4
Titanium Dioxide Modification for Application as the Electrode in the Dye-Sensitised Solar Cell. (#1005) Rachel Anne Caruso, Fuzhi Huang, Yi-Bing Cheng; The University of Melbourne, Parkville, Victoria, Australia.

The efficiency of the dye sensitised solar cell can be enhanced by tailoring the composition, crystal phase and morphology of the titanium dioxide-based electrode. In this study both templating and doping techniques have been combined during the preparation of the titanium dioxide material. The precursor solutions were infiltrated within a porous template, an agarose gel, followed by sol-gel chemistry and a heating step which resulted in the formation of porous inorganic structures. A second metal oxide precursor could be added at various stages during the synthesis. The materials were characterised by electron microscopy to gain an understanding of the morphology, and X-ray diffraction to obtain crystallisation characteristics as a function of the second metal oxide wt. % and heating temperature. It was found that the presence of the second metal oxide could result in higher surface areas with a decrease in the anatase crystal size. The surface properties of the samples were also studied by both nitrogen sorption and monitoring the adsorption of the sensitizing dye, N719. The time at which the second metal oxide was added during synthesis, as well as the type of metal being added, influenced the overall efficiency of the electrode, with some metals improving the open circuit voltage, while in other cases the short circuit current increased.

2:00 PM *G4-S3.1 (invited)
Roll to Roll Producible Si Thin film Single, Tandem and Triple Cells on Stainless Steel Foils and on Plastic. (#1288) R. E. I. Schropp, Debeye Institute for Nanomaterials Science, Utrecht University, Netherlands.

This presentation discusses recent progress and insight obtained in the field of thin film silicon solar cells by VHFCVD and Hot Wire CVD. The technology of Hot Wire Chemical Vapor Deposition (HWCVD) or Catalytic CVD (Cat-CVD) has made great progress during the last couple of years. The quality of thin film materials can be accurately controlled and there is increasing evidence that large area continuous coating is feasible. The cell efficiency reached for (nanocrystalline) nc Si:H n-i-p solar cells on textured Ag/ZnO presently is 8.6 %, in line with the state-of-the-art level for such cells for any method of deposition. Such cells, used in triple junction cells together with hot-wire deposited protocrystalline Si:H and plasma-deposited proto-SiGe:H, have approached 11 % efficiency. The single junction ? = μc-Si:H n i p cell is entirely stable under prolonged light soaking. The triple junction cell, that is only 2.5 ? = μm thick, including protocrystalline i layers, is within ? = Δ? = η/? = η =3.5 % stable. We also report on our research on translating this technology to inexpensive polymer foil substrates, such as PET. To this end, the deposition temperature has to be reduced from the optimal temperature of 200-250 ? = οC to 100 ? = οC. It is a challenge to preserve good optoelectronic properties for the active materials and to deposit efficiently doped n-type and p-type layers at this temperature. Amorphous silicon deposited at 100 ? = οC by VHF plasma deposition has been most successful in this respect and could successfully be used in p-i-n solar cells, which yielded 7.3% efficiency under AM1.5 illumination. In the n i p configuration, which can be used on opaque polymer foils, we reached an efficiency of 5.3%. In case of Hot Wire CVD we have reached an efficiency of 3.4%. Another approach towards roll-to-roll deposition is to use a temporary substrate that is resistant to high temperature as well as to reactive gas species and to transfer the complete cell to a permanent flexible polymer plastic substrate. Recently, experimental laboratory scale tandem modules with HWCVD active layers have reached initial efficiencies of 7.4% at an aperture area of 25 cm2. This Helianthos concept of manufacturing of flexible thin film silicon modules was originally invented by researchers at Akzo Nobel and Utrecht University, and was first published in 1998. Recently, 6-meter long modules have been made in the roll-to-roll pilot line system at Nuon Helianthos b.v. with efficiencies of around 6%.

2:30 PM G4-S3.2
Device Fabrication Scheme for Evaporated SPC poly-Si Thin-Film Solar Cells on Glass (EVA). (#282) Oliver Kunz, Zi Ouyang, Johnson Wong, Armin Gerhard Aberle; The University of New South Wales, Sydney, Australia.

Polycrystalline silicon thin-film solar cells on glass obtained by SPC of PECVD-deposited a-Si precursor diodes are capable of producing large-area devices with respectable photovoltaic efficiency. This has not yet been shown for equivalent devices made from evaporated Si precursor diodes ("EVA" solar cells). In this work we show that there are two main problems for the metallization of EVA solar cells: (i) Shunting of the p-n junction when the air-side metal contact is deposited; (ii) Formation of the glass-side contact with sufficiently low contact resistance and high shunt resistance. The first problem can be overcome by decreasing the metallisation fraction on the BSF layer (line contacts) and by subsequently thinning of the metal on the steep sidewalls of the pinholes using a wet-chemical etching step. The second problem can be removed by plasma etching down to the highly doped glass-side layer and subsequent alignment of the emitter contact fingers to entirely avoid contact to the base material. The best laboratory cells produced using these procedures have so far achieved fill factors in excess of 65%, series resistance values of about 5 Ohm-cm2, and short-circuit currents of up to 14 mA/cm2 on planar glass superstates using white paint as back surface reflector. Computer modelling of finished EVA solar cells with PC1D reveals that the diffusion length in the absorber layer is larger than the thickness of this layer (~2 microns), giving significant room for further device optimisation. It can be concluded that the device fabrication scheme presented in this paper is well suited for the metallization of EVA solar cells and that the electronic properties of evaporated SPC poly-Si materials are sufficient for PV applications. Optimisation of the presented metallisation scheme is expected to lead to efficient EVA solar cells.

2:45 PM G4-S3.3
Microcrystalline Silicon, Grain Boundaries and Role of Oxygen. (#250) Jan Kocka, Ha Stuchlikova, Martin Ledinsky, Jiri Stuchlik, Tomas Mates, Antonin Fejfar; Department of Thin Films, Institute of Physics, Academy of Sciences of the Czech Republic, Praha, Czech Republic.

While the stability of hydrogenated microcrystalline silicon (μc-Si:H) against the light soaking is its great advantage, there are many problems, related to its very complex microstructure, often changing with the thickness. By AFM topography combined with the local current map [1] it was shown, that there are at least 2 sizes of crystalline Si grains, the small ones (10-30nm) and the large ones (100-500nm), often called aggregates or columns. The characteristic features of a-Si:H/μc-Si:H transition [1], like abrupt change of dark conductivity and crystallinity have been verified on many series of samples. In contrary, the nature and the role of grain boundaries (GB) as concerns the transport in μc-Si:H, are still far from full understanding. Recent attempts to use combined AFM techniques for GB characterization [2,3], renewed the discussion whether GBs represent conductive channel or potential barrier. We have illustrated, that the properties of the GBs can be modified by H [4]. The purpose of this work is to correlate the content and bonding of oxygen with the transport properties and GBs in μc-Si:H. We have prepared by 54 MHz PECVD with the multiholes cathode [5] the 3 dilution series of samples with dilution rH = H2/SiH4 = 10 - 32: • Series "I", TS = 250°C, (4 W/cm2), • Series "II", TS = 250°C, (8 W/cm2), • Series "III", TS = 80°C, (4 W/cm2). For sample characterization we have used the room temperature DC dark conductivity (σD), its activation energy (Ea) and prefactor (σ0) from its temperature dependence, the crystallinity, derived from the Raman spectra and H and O content and their bonding, evaluated from IR spectra. We have found the following: • For series "I" the "large GBs" (columns, detrimental for transport properties and ageing) are formed at rH= 13, above which O content increases up to 2 at%. • For series "II", prepared with the increased power, the activation energy (Ea) and the prefactor (σ0) are practically constant, no columns are formed, the O content is surprisingly high, up to 12 at%, but samples are highly crystalline and stable . • For low TS series "III" the dramatic changes of the activation energy (Ea) and the prefactor (σ0) , correlated with the O content, are observed already before a-Si:H/μc-Si:H transition (at rH= 17). These results will be discussed within our model of transport [1], based on the idea of the GB potential barriers, related to the O and H rich a-Si:H based GB tissue. [1] J. Kocka et al, physica status solidi (c), 1, No.5, 1097-1114 (2004). [2] Z. Shen et al, Jap. J. of Appl. Phys. 46, 2858-2864 (2007). [3] D. Domine, et al, preprint from PVSEC, Milano (2007) [4] J. Kocka, et al, Thin Solid Films, 501, 107 (2006) [5] Ch. Niikura, et al, J. Non-Cryst.Sol. 338-340, 42 (2004)

3:00 PM G4-S3.4
Recent Progress and Advancement of Poly-Si Film Solar Cell Technology. (#174) Horng-Show Koo, Department of Optoelectronic System Engineering, Ming-Hsin University of Science and Technology, Hsinchu, Taiwan.

Yoshihiro Morimoto and Horng-Show Koo* Department of Optoelectronic System Engineering, Ming-Hsin University of Science and Technology, Hsinfong, Hsinchu 304 Taiwan, R.O.C. * Corresponding author : frankkoo@must.edu.tw Abstract a-Si thin film solar cell has been developed to realize high efficiency and large size solar cell with low cost, and already achieved relatively high efficiency soar cell and widely used. However light induced degradation is a crucial problem that need to be solved. Since micro crystal (?-C)Si:H Si deposited by RF PECVD was reported, considerable attention has been focused on this ?-C Si:H as a promising material that can realize high efficiency, large size solar cell without light induce degradation by low cost. Large grain size poly Si has also attracted as promising technology to realize high efficiency low-cost solar cell. This paper reviews the poly crystalline Si formation technologies and discusses in detail and clarifies the present problems of these technologies. For ?-C Si:H film, development history of ?-C Si:H film is briefly reviewed first. Then we discuss the importance of radical hydrogen role during deposition of ?-C:H Si and also discuss its mechanism. After that, deposition technologies including RF PECVD, VHF CVD, Hot Wire-CVD and High Density Plasma CVD are surveyed in detail and clarify the present problems of these technologies to achieve high quality ?-C:H Si film with high deposition rate. Tandem structure of a-Si:H/?-C Si:H that has great deal attention as promising structure being able to realize high performance solar cell is also reviewed and its issues are addressed. After review of ?-C:H Si technology, large grain size poly Si technologies are reviewed. Aluminum Induced Crystallization (AIC) and Zone Melting Crystallization (ZMC) technologies are surveyed in detail. For the AIC case, influence of intra-grain defect on the efficiency is discussed in detail and clarify the issue of this technology.

3:15 PM G4-S3.5
Doping Dependence of the Optical Properties of Solid Phase Crystallized Evaporated Poly-Si Thin-Films on Glass. (#203) Song He, Alistair Bruce Sproul, Armin Aberle; The University of New South Wales, Sydney, Australia.

Doping Dependence of the Optical Properties of Solid Phase Crystallized Evaporated Poly-Si Thin-Films on Glass Song He, Alistair B. Sproul, Armin G. Aberle ARC Photovoltaics Center of Excellence, The University of New South Wales, UNSW Sydney NSW 2052, Australia Poly-Si thin-film solar cells have been receiving significant interest in recent years due to the potential of significant reduced manufacturing cost. Poly-Si thin-film solar cells are being developed at The University of New South Wales (UNSW) by solid phase crystallization (SPC) of e-beam evaporated a-Si films. In this work, the boron and phosphorous doped a-Si films (~1e16cm-3, ~5e17cm-3 and ~1e19cm-3) with thickness of 1.5 um were evaporated onto SiN coated glass at 200°C with a base pressure of 3e-8 Torr. For comparison, intrinsic a-Si films were also prepared on SiN coated glass. The a-Si samples were then transferred to a tube furnace and crystallized by thermal annealing at 600°C. Rapid thermal annealing (RTA) (4 min at 900°C) was then applied to the poly-Si films to activate the dopants. Both the reflection and transmission spectra were measured across the wavelength range of 250-2000 nm with a spectrophotometer. The spectral data were then fitted simultaneously with commercial software (WVASE) to extract the refractive index and absorption coefficient. A Tauc-Lorentz model with one oscillator was used for the fitting of intrinsic poly-Si films (for wavelengths in the range of 450 nm to 2000 nm), while a Tauc-Lorentz model with two oscillators was used for the fitting of doped poly-Si films (due to the free carrier absorption at the infrared (IR) wavelengths). The fitting had very low mean-squared error (MSE) of less than 1, indicating that the fittings are good and the applied Tauc-Lorentz models are well suited for modeling the optical properties of the evaporated SPC poly-Si films on glass. UV reflectance measurements show that the crystal quality of the poly-Si films improves with the increasing of phosphorous doping concentration. In contrast, boron has little or even a negative impact on the crystal quality of poly-Si films. The refractive indexes of intrinsic and moderately doped (~1e16cm-3, ~5e17cm-3) poly-Si thin films are very similar to that of Si wafer. The refractive indices of the heavily doped poly-Si films (both boron and phosphorous, ~1e19cm-3) are lower than c-Si values, owing to the strong interaction of the radiation with free charge carriers. The intrinsic poly-Si films have a larger absorption coefficient than Si wafer in the visible wavelength range due to the defects such as grain boundaries. The boron doped poly-Si films show a larger absorption coefficient than intrinsic poly-Si film, whereas that of phosphorous doped poly-Si films is lower. This is attributed to the poorer crystal quality of the boron-doped films and superior crystal quality of the phosphorous- doped films. At IR wavelengths, the absorption coefficients increase linearly with the doping concentrations, for both boron and phosphorous-doped poly-Si films. In summary, optical characterization allows the material quality of evaporated SPC poly-Si films to be assessed without damaging the samples. Variation in absorption for wavelengths in the range of 450 nm - 900 nm is attributed to the differences in the defect density in the films. IR absorption in doped poly-Si films is a useful measure for the density of electrically active dopants.

AFTERNOON BREAK 3:30 PM - 4:00 PM

4:00 PM *G4-S4.1 (invited)
High Efficiency Thin Film Silicon Solar Cell and Module with Newly Developed Interlayer. (#1314) Kenji Yamamoto, Fontier Materials Research Laboratories, Kaneka Corporation, Osaka, Japan.

The paper discusses the development of a novel enhancement of light trapping scheme for a thin film Si based stacked cell and module, where a newly developed interlayer by plasma CVD is inserted between amorphous and microcrystalline Si thin film. This new interlayer leads both intentional controlling the refractive index of it and nothing shunting issues for fabricating series interconnection of large area module. An initial aperture efficiency of 13.4% has been achieved for 910x455mm2 thin film Si tandem module fabricated in a single chamber process of p-i-n microcrystalline cell, which was independently confirmed by AIST. We also proved by numerical simulation that the effect of the interlayer on light trapping towards the top cell is more enhanced with textured substrate. Namely, the optimum combination between the interlayer with lower refractive index and the optimized feature size of the texture of front TCO enhances the internal light trapping of stacked cell.

4:30 PM G4-S4.2
Wide Optical Bandgap Boron-Doped µc-Si:O_x:H using Tantalum HW-CVD. (#752) Yasuhrio Matsumoto¹, Mauricio Ortega¹, Zhenrui Yu²; ¹Electrical Engineering Department, CINVESTAV-IPN, Mexico ; ²INAOE-Puebla, Mexico.

Oxygen-rich boron-doped hydrogenated microcrystalline silicon (p-µc-Si:Ox:H) films have been prepared using Hot-Wire Chemical Vapor Deposition (HW-CVD) system. Pure silane (SiH4), hydrogen (H2), oxygen (O2), and diluted diborane (B2H6) gases were used. The tantalum (Ta) catalyst temperature (Tfil) was varied from 1800 to 1950 ?C and films were deposited at substrate temperatures (Tsub) of 200 and 300 ?C. In addition, same p-µc-Si:Ox:H deposition were made using tungsten (W) catalyst in order to compare its optical and electrical properties. HW-CVD system was described elsewhere [1]. Film deposition was carried out on Corning 2947 glass substrates. The gas sources were directed toward 0.5-mm-diameter, with 15-cm-long, double-coiled Ta catalyst. All of the p-µc-Si:Ox:H films were prepared at the deposition chamber pressure (Ps) of 0.1 Torr during 20 minutes and the source gases SiH4:H2:O2 were fixed all to 5 SCCM, and hydrogen diluted B2H6 to 10 SCCM. The samples were deposited with fixed filament to substrate distances (dfs) of 5 cm. Optical-bandgap (Egopt) of the deposited films was estimated from transmittance data using Swanepoel`s procedure [2]. The thickness was measured by means of Tencor P-15 profile meter. Egopt of the prepared samples has the tendency to increase as a function of Tfil. The obtained Egopt changes from 1.78 eV to 1.91 eV as Tfil increases from 1800 to 1950 ?C. The electrical conductivity (&sigma???varies in 0.3 ~ 7.0x10-5 &Omega -1cm-1 and the deposition rate (dr) varies from 0.40 to 0.54 nm/s. In comparison, the deposited p-µc-Si:Ox:H films using W catalyst revealed bit a lower Egopt than that of deposited using Ta-catalyst, within 1.73 to 1.85eV. Even though, the corresponding &sigma was better achieving 5.0x10-3 &Omega -1cm-1 with similar dr as Ta catalyst deposited samples. From the optical properties of the samples prepared using W- and Ta-catalysts, it could conclude that in both type of catalysts, its temperature promotes oxygen incorporation to obtain wider- bandgap materials. On the other hand, &sigma also has the same tendency to increases with catalyst temperature as occurred with Egopt, but at higher temperatures it tends to saturate. The conductivity increment due to higher Tfil, it could be explained because of the formation of larger microcrystalline grains which was demonstrated by XRD spectra. The best wide-gap material using Ta-catalyst was obtained at Tfil = 1950?C and Tsub = 200?C, having Egopt = 1.91 eV with &sigma??= 2.0x10-5 &Omega -1cm-1. The tantalum catalyst has no evident degradation after film deposition process using pure oxygen. References: [1] Y. Matsumoto, M. A. Reyes, and A. Escobosa, J. Appl. Phys. 98 (2005) 014909. [2] R. Swanepoel, J. Phys. E: Sci. instrum. 16 (1983) 1214.

4:45 PM G4-S4.3
Heterogeneous Integration using III-V Compound Semiconductors for Solar and Electronic Applications. (#1179) Mark S Goorsky, Sumiko L Hayashi, Monali B Joshi; Department of Materials Science and Engineering, University of California, Los Angeles, USA.

Heterogeneous integration of III-V compounds represents a means by which to achieve engineered multilayers with a combination of thermal and mechanical properties that are not available in a single material, and greatly reduce the need for lattice-matched epitaxial processes. There are several key technologies involved in this process. First, the principle of semiconductor layer transfer using hydrogen implantation - while well known for silicon - has not been widely considered for other semiconducting materials because techniques to investigate the exfoliation process have not been well understood. Understanding exfoliation through a nucleation, growth, and diffusion model has allowed us to successfully apply this technique from materials used for multi-junction solar cells, particularly GaAs and InP. Second, we have developed novel transfer techniques that can be used to transfer active solar cell structures to a separate substrate without the limitations usually imposed by techniques such as epitaxial layer liftoff. This process involves the use of porous silicon and III-V exfoliation. Third, controlled chemical mechanical polishing provides a means to produce epi-ready surfaces upon which devise structures can be grown. Finally, understanding the role of thermal expansion of the structure can greatly reduce plastic deformation. In each of these cases, techniques involving x-ray scattering have proved to be highly beneficial to the overall study of heterogeneous materials integration.

5:00 PM *G4-S4.4 (invited)
Advances in Defect Control for High Performance Metamorphic III-V on Si Photovoltaics. (#1369) Steven Ringel, Tyler Grassman, Mark Brenner; Department of Electrical and Computer Engineering, The Ohio State University, Columbus, USA.

Metamorphic epitaxy, by which lattice constants different from an original host substrate can be achieved through careful grading of material compositions and equilibrium lattice constants during epitaxy, has been transforming advanced semiconductor technologies for the past 20 years from SiGe electronics to metamorphic high electron mobility transistors. The more recent impact of metamorphic III-V epitaxy in photovoltaics is particularly profound, since by tuning a material lattice constant, one can obtain optimum bandgap profiles leading toward ideal solar cell designs with respect to use of the solar spectrum. Moreover, metamorphic growth is the most promising pathway to integrate III-V multijunction cell technologies with Si wafer technologies, which can ultimately lead to creating a high performance III-V multijunction cell manufacturing base that is simultaneously high throughput, high capacity and high yield, while leveraging existing global infrastructure. The challenge for metamorphic III-V photovoltaic devices however, is in defect control. It is necessary to maintain a very high level of electronic quality, e.g. long carrier diffusion lengths, so that the designed optical partitioning advantages of metamorphic multijunctions, which portend concentrator efficiencies in excess of 50% and/or integration with Si technology, are not overwhelmed by crystalline defects that result from grading the lattice constant, which can impede carrier collection and thus electrical performance. This talk focuses on advances in two areas of metamorphic III-V photovoltaic materials leading to integrated III-V/Si photovoltaics, the use of low defect density, graded SiGe alloys as intermediary steps that enable scalable integration of III-V multijunction solar cells with Si in which a virtual (metamorphic) Ge substrate supports III-V cell growth, and our recent work based on virtual GaAsyP1-y on Si substrates. For the former, single junction GaAs and dual junction GaInP/GaAs cells on Si with high performance (18.5% AM1.5 for single junction on Si and open circuit voltages in excess of 2.2 V for the dual junction on Si) has been achieved, with a clear path defined toward higher performance. The latter is based on our development of low defect density, anti-phase domain-free, atomically smooth (0.38 nm RMS roughness) and highly uniform GaP layers on Si, upon which we have fabricated very promising 1.7 eV and 1.8 eV bandgap GaAsP solar cells on Si via anion-based metamorphic GaAsyP1-y layers. The presentation will review material properties, growth dependencies and cell results from both approaches and will focus on fundamental similarities and differences of both GaP/Si and GaAs/Ge as analogous III-V/IV heterovalent interfaces that are closely lattice-matched.

9:00 AM G5-S1.1
The Development of a High Efficiency Advanced Quad-Junction Solar Cell. (#825) Sherif Michael, Electrical and Computer Engineering Department, Naval Postgraduate School, Monterey, California, USA.

In this paper the design and optimization of advanced multi-junction photovoltaic devices, utilizing a newly introduced modeling technique[1], is demonstrated. In our opinion, the introduction of this modeling technique to the Photovoltaic community will prove to be of great importance in aiding the design and optimization of advanced solar cells. A new four-junction solar cell was simulated and optimized to maximize efficiency and power output of this prospective device. To accomplish this, exotic materials such as Indium Gallium Phosphide (InGaP) and Indium Gallium Nitride Arsenide (InGaNAs), were physically modeled. The methodology and design considerations for this process are explained. The flexibility of the proposed methodology is illustrated and example results are shown throughout the whole process. The Silvaco Software used here, is a simulation software tool targeting the area of electronic design. One of its major products is the Virtual Wafer Fabrication package (VWF). This is a large suite of highly sophisticated tools - among which is ATLAS - that aid in the design and development of all types of semiconductor and VLSI devices. The phenomena modeled start from simple electrical conductivity and extend to thermal analysis, radiation, laser effects etc. A big variety of detailed layer growth processes and material properties (mobilities, recombination parameters, ionization coefficients, optical parameters, etc) add to the accuracy of the simulation. However, we have not seen any effort, from researchers or from solar cell manufacturers, to utilize this powerful tool for the modeling of advanced solar cells and designing future state-of-the art solar cells. The multi-junction cell is constructed of three stacked Ge, GaAs, and InGaP cells. To verify this novel modeling technique as presented in [1], these cells were first developed and tested individually. That way, their efficiency could be optimized and their performance characteristics could be adjusted. I-V curves and frequency responses were also plotted. Producing snapshots of the structure and how various parameters are distributed throughout its area provides a very valuable insight. Such parameters include potential, electron and hole concentrations, photogeneration rates and electrostatic fields. Using these parameters from the triple cell, the InGaNAs fourth junction was created between the GaAs and the Ge cells to form the quad structure. In order to optimize the cell, it must be noted that each layer current produced is a factor of the amount of light the cell receives and how thick the cell's base is. With the addition of the fourth cell, it cannot be taken for granted that the bottom Ge cell will be able to produce as much current as the other cells. Therefore, the optimization becomes a variation of the thickness of three cells to match four currents. The solution to optimizing any multijunction design involves both the design of individual junction layers which produce an optimum output power and the design of a series-stacked configuration of these junction layers which yields the highest possible overall output current. This paper proposes a two-part genetic algorithm process to refine a given multijunction solar cell design for near-optimal output power for a desired light spectrum. [1] Michael, S."A Novel Approach for the Modeling of Advanced Photovoltaic Devices Using the SILVACO- ATLAS virtual wafer fabrication tools," Journal of Solar Energy Materials & Solar Cells, Vol. 87, 2005.

9:15 AM G5-S1.2
Study on Boron Doped Silicon Quantum Dot Superlattices for All-Silicon Tandem Solar Cells. (#349) Xiaojing Hao, Eun-Chel Cho, Sangwook Park, Gavin Conibeer, Martin Andrew Green; School of Photovoltaic and Renewable Energy Engineering, The University of New South Wales, Sydney, Australia.

Doping bulk semiconductors has made the new era of modern electronic devices [1]. Conventional semiconductor devices, such as the transistor, would not operate without such impurities. Similarly, as the electronic device uses nanometer-scale particles, selected impurities incorporation into these particles have been investigating. Thus, doping is a critical step for tailoring their properties for specific applications in solar cells [2], optoelectronics [3] and other electronic devices. However, doping a small amount of foreign atoms in nanometer-scale semiconductors has been under a debate for a few years [4,5]. Question arises whether the regularities accepted for doping in bulk semiconductors can be employed to low-dimensional case. Thus, it is of the ultra importance to understand how doping operates at this nanometer-scale. This paper is focused on p-type Si QD superlattices, for the sake of realizing quantum dot (QD) P-N junction for its possible application in all-Si tandem solar cells. For photovoltaic application, optical, structural and electrical properties of Si QD superlattice with various boron (B) concentrations were investigated. P-type Silicon QD superlattices was realized with boron as a doping source. Boron-doped Si QD superlattices with various boron concentrations were obtained by alternate deposition of boron-doped silicon-rich oxide (SRO) and SiO2 by a co-sputtering technique and subsequently high temperature annealing. Boron-doped SRO was realized by a co-sputtering of Si, SiO2 and B targets at a room temperature. B concentrations in the SRO were varied by control of deposition rates of three targets. Si wafer and quartz substrate were used for various characterization purposes. The chemical composition and coordination information of as-deposited boron-doped SRO layer were studied by X-ray photoelectron spectroscopy (XPS). The XPS results indicate existence of boron oxide (193 eV),boron (187 eV) and/or silicon boride (188 eV) in the as-deposited boron-doped film. The boron bonding environments vary with high temperature annealing. Moreover, the boron bonding environments are dependent on the O/Si ratio. Formation of Si nanocrystals was studied by transmission electron microscopy (TEM) and glancing incidence X-ray diffraction (GIXRD). Optical absorption properties were studied by a UV-Visable-Near-infrared spectroscopy. Lateral conductivity was also investigated. Reference 1. Ben G. Streetman and Sanjay Banerjee (2000) Solid State Electronic Devices. 5th ed. New Jersey: Prentice Hall. 2. E.C. Cho, M.A. Green, G. Conibeer, D.Y. Song, Y-H, Cho, G. Scardera, S.J. Huang, S. Park, X.J. Hao, Y.D. Huang, L.V. Dao, Journal of Advances on Optoelectronics, 69578 (2007). 3. L. Pavesi, L. D. Negro, C. Mazzoleni, G. Franz? and F. Priolo, Nature 408, 440 (2000). 4. S.C. Erwin, L.J. Zu, M.I. Haftel, A.L. Efros, T.A. Kennedy, D.J. Norris, Nature 436 (2005) 91. 5. G.M. Dalpian, J.R. Chelikowsky, Phys. Rev. Lett., 96 (2006) 226802.

9:30 AM G5-S1.3
Miniature Silicon Solar Cells for High Efficiency Tandem Cells. (#888) Ngwe Soe Zin, Andrew Blakers, Evan Franklin, Vernie Everett; The Australian National University, Australian Capital Territory, Australia.

A major objective for photovoltaic conversion is to develop high-efficiency and low-cost solar cells. Many approaches are under investigation - Multiple Junction Solar Cell, Multiple Spectrum Solar Cell, Multiple Absorption Path Solar Cell, Multiple Energy Solar Cell, and Multiple Temperature Solar Cells [1]. The Multiple Junction Solar Cell approach based on six-junction tandem solar cell has been adopted to achieve conversion efficiency of greater than 50% in the VHESC program sponsored by DARPA [2]. A consortium with around 20 contributors, led by DuPont and the University of Delaware with substantial funding from DARPA, aims to create devices that operate at 50% efficiency in production for $1,000 m2. With the recent rapid advances in materials processing, non-imaging optics, and solar cell architectures, this aggressive goal is achievable. In six-junction tandem solar cell approach, individual solar cells are stacked on one another and each solar cell absorbs the respective solar spectrum closest to one another. Silicon is one of the cells in the tandem stacks and absorbs energy of 1.42 - 1.1 eV. The role of the silicon cell is to convert 7% of the light incident on the tandem stack into electricity. Other cells in the stack contribute the balance of the electricity. Key design parameters for the silicon cells are that it should have dimensions of 2.5mm x 8mm and it needs to transfer light energy of less than 1.1ev to the underlying solar cells. CSES (Centre of Sustainable Energy Systems) group in ANU has joined the consortium to work on silicon cells for tandem stacks. In this paper, a discussion is made of the design of the silicon cell. Minority carrier recombination at surfaces and in the volume, internal quantum efficiency, resistance losses, free carrier parasitic absorption, optical reflection, light trapping, and light absorption must be traded off against each other. Modelling was used to analyse the various parameters and produce estimates of short-circuit current, fill factor and open-circuit voltage of the cell. In addition, characterization of solar cell by photoconductance measurement to analyse carrier recombination and emitter saturation current (Joe) as well as to predict the open-circuit voltage of solar cell [3, 4] is presented. Based on characterisation results completed wafers were maintained with decent carrier life-time and low Joe value even after a series of high temperature processes. For metallisation of such small solar cells, alternate methods of making proper contact such as light-induced plating and electrolyte plating in addition to evaporating metal on the contacts were explored and employed. Numerical resistive loss modelling was made to calculate the optimum metal thickness achieved by light-induced and electroplating to minimise resistive losses generated by metal contacts. Experiments were conducted to determine the proper plating rate by light-induced and electrolyte platings. Cells were fabricated by standard silicon processing techniques followed by testing of IV curves using current-voltage flash-tester [5] to achieve the target efficiency.

9:45 AM G5-S1.4
Advanced Upconverter Systems with Spectral and Geometric Concentration for High Upconversion Efficiencies. (#750) Jan Christoph Goldschmidt¹, Philipp Löper¹, Stefan Fischer¹, Stefan Janz¹, Marius Peters¹, Stefan Glunz¹, Gerhard Willeke¹, Efrat Lifshitz², Karl Krämer³, Daniel Biner³; ¹Department Silicon Solar Cells - Development and Characterization, Fraunhofer-Institut für Solare Energiesysteme ISE, Germany ; ²TECHNION Haifa, Israel ; ³Universität Bern, Switzerland.

Motivation Silicon solar cells loose at least 20% of the incident power because photons with energies below the bandgap are transmitted. Upconversion of low energy photons is an approach to overcome this principal problem. Er doped NaYF4 is up to now the most promising material for upconversion. However, upconversion efficiency is still quite low and only a small fraction of the spectrum is upconverted. Approach The combination of an upconverter with a fluorescent material helps to overcome this limitation. The fluorescent material should absorb all photons with wavelengths between the bandgap of the solar cell and the absorption range of the upconverter and emit in the narrow absorption range of the upconverter. Because the solar photon density of a broad spectral range is funnelled to a smaller range, we call this process spectral concentration. This approach increases the upconversion efficiency by two mechanisms: Firstly, more light is absorbed that is potentially upconvertible. Secondly, the photon density in the absorption range of the upconverter is increased. As upconversion is a nonlinear process, the quantum efficiency of the upconverter increases with increasing impinging intensity. The incorporation of the fluorescent material into a fluorescent concentrator yields additional geometric concentration. However, applicable fluorescent materials, such as PbSe quantum dots, absorb the upconverted radiation. Thus, upconverter and fluorescent material have to be separated from each other and the upconverted radiation should be prevented to enter the fluorescent concentrator. This can be done with selectively reflective photonic structures. Preliminary results First calculations show that the number of upconverted photons can be increased by a factor of 30 by spectral concentration for an Er-based system. Geometric concentration additionally increases efficiency by approximately the concentration ratio. We set up a detailed simulation of the upconversion system based on rate equations and the Einstein relations. It will be matched to experimental photoluminescence and absorption data to perform realistic calculations. We synthesized PbSe quantum dots as fluorescent material and prepared photonic structures based on SiC with promising characteristics. The spectral response of first upconverting systems without spectral concentration was measured. Realisation of a complete system will start soon, so results can be presented at the conference.

10:00 AM G5-S1.5
Towards an Ab-intio Characterization of Novel Intermediate Band Photovoltaic Materials. (#425) Perla Wahnon, Pablo Palacios, Irene Aguilera, Kefren Sanchez; Instituto de Energía Solar, ETSI Telecomunicacion, Universidad Politécnica de Madrid, Spain.

An ab-initio study of novel photovoltaic materials with enhanced optoelectronic properties is presented in this contribution. Predictions of absorption coefficients agree completely with the characterization of the first experimental samples grown in the laboratory. Compounds selected for the study are derived from chalcogenide semiconductors in which several atoms are substituted by transition elements. These substitutions modify the electronic band structure in such a way that a new narrow band appears inside the parent semiconductor band-gap. Partial occupation of this band enables that additional carriers could be obtained from absorption of photons with energy lower than that of the band-gap, thus enhancing the photovoltaic conversion properties of the material. It was estimated than a photovoltaic solar cell designed from this novel concept could reach a thermodynamic efficiency of 63.2% compared to 43.1% corresponding to the thermodynamic efficiency limit of conventional semiconductor based solar cells. Chalcogenide semiconductors are well known and spreadly used in the photovoltaic field, often as thin film solar cell materials or window layers. In the present study several transition metals M in a +3 valence state substitute some selected group III elements of CuGaS2 chalcopyrite, or In2S3, MgIn2S4 thiospinels. Thus in the chalcopyrite based compounds MCu4Ga3S8 M turn out to be in tetrahedric coordination while in thiospinels derived materials M2In14S24 and MMg2In3S8 it has an octahedric coordination. The intermediate band in all of these materials is formed by electrons of the 3d shell of M. In the case of chalcopyrite based structure substituted by Ti the intermediate band corresponds to the low-energy eg doublet split from the 3d shell in a tetrahedral environment, while for the spinel based compounds taking Ti or V as substitute the intermediate band comes from the low-energy t2g triplet originated from the splitting of the 3d shell in an octahedral crystalline field. A broad study of material properties has been carried out from state of the art calculations made from first principles following the Density Functional Theory (DFT) and beyond DFT. The study includes relaxation of forces and stresses to predict the structure of the novel materials. The electronic band structures show in some of the compounds an intermediate band with the desired features. Dielectric function was also calculated from converged wave functions, allowing obtaining the optical absorption spectra. Results show a significant enhancement of the absorption coefficient respecting to the corresponding parent semiconductors in the main emission region of the solar spectrum. For some of the theoretically proposed compounds, optoelectronic properties of recently synthesized samples have been obtained experimentally. Comparison of experimental absorption measurements with results of the calculations presented here shows a very good agreement.

10:15 AM G5-S1.6
Study of Silicon Quantum Dot P-I-N Junction Device. (#1106) Sangwook Park, Eun-Chel Cho, Xiaojing Hao, Gavin Conibeer, Martin A Green; The University of New South Wales, Sydney, Australia.

The tandem stack of cells is one of good approaches for using a full solar spectrum and improving solar cell performance [1]. By stacking solar cells of different bandgap material on top of one another, with the highest bandgap cell uppermost, light is automatically filtered as it passes through the stack. By restricting the dimensions of silicon to less than Bohr radius of bulk crystalline silicon (~ 5 nm), quantum confinement cause its effective bandgap to increase. Therefore silicon quantum dot superlattice can be a good candidate for realizing all silicon tandem solar cells. In this work, silicon quantum dot and crystalline silicon heterostructure device and p-i-n homojunction device will be fabricated to understand the electrical properties of these junctions. We are experimentally investigating on material properties of silicon (Si) quantum dot (Si QD) superlattices and fabricating the device as a first step towards silicon based tandem cells. Photovoltaic properties and electrical properties of n-type silicon quantum dot and p-type crystalline Si (c-Si) heterostructure device were fabricated and characterized [2]. p-type Si QD and n-type c-Si heterostructure device will be fabricated for comparison and p-i-n homojunction device also will be fabricated and investigated. The phosphorous doping in n-type or Boron doping in p-type Si QDs superlattices were realized by P2O5 or Boron co-sputtering during the deposition of silicon-rich oxide (SRO, Si and SiO2 co-sputtering), which forms Si QDs upon high temperature post-annealing. The n-type or p-type region typically includes 15 bi-layers formed by alternating deposition of SRO layer and SiO2. The structural characterisations will be accomplished by transmission electron microscopy (TEM) and Raman spectroscopy. The optical bandgap energy (Eopt) of films will be determined by the Tauc's equation. The electronic and photovoltaic properties of the heterostructure and p-i-n homojunction devices also will be characterised by illuminated and dark current-density-voltage characterisation, spectral response, and capacitance-voltage (C-V) measurements. Reference 1. MA Green et al, 22nd European Photovoltaic Solar Energy Conf., Milan, 2007, pp. 1-4. 2. S.W. Park et al, 17th International Photovoltaic Science and Engineering Conference, Fukuoka, 2007