Clemens Heske University of Nevada-Las Vegas
Miguel Contreras National Renewable Energy Laboratory
Malgorzata Igalson Warsaw University of Technology
Stuart J. C. Irvine OpTIC Technium/NEWI
Akira Yamada Tokyo Institute of Technology
Global Solar Energy Inc
National Renewable Energy Laboratory
M1: Characterization of Absorber Materials
Tuesday PM, April 14, 2009
Room 2011 (Moscone West)
9:30 AM - **M1.1
How Soft X-ray Spectroscopies Can Help to Reveal the Secrets of Chalcopyrite-based Thin Film Solar Cells.
Marcus Baer 1 2 Show Abstract
1 , University of Nevada, Las Vegas, Las Vegas, Nevada, United States, 2 , Helmholtz-Zentrum Berlin, Berlin Germany
Chalcopyrite-based solar cells, as most of the current state-of-the-art electronic devices, consist of a thin film layer stack. Since each layer has a different chemical and electronic structure, the interfaces between those films are (when not designed carefully) often the place of an increased density of defect states and/or suboptimal electronic level alignment, which both can result in an increased charge carrier recombination rate at the interface. In addition, interdiffusion processes can take place at interfaces, leading to significant intermixing, which induce changes in the optoelectronic properties of the entire device. Furthermore, the interfaces not only influence, but in most cases determine the local electric fields necessary for efficient charge carrier separation. Thus, the key issue for a deliberate improvement of chalcopyrite-based thin film solar cell devices is an exact knowledge of the chemical and electronic interface structure.
Over the last few years, a combination of electron and x-ray spectroscopies has been established to gain insight into the chemical and electronic properties of interfaces. In this presentation, we will demonstrate the potential of that approach by means of several examples from the field of chalcopyrite thin-film solar cells. By deliberately using various photon and electron spectroscopies with different information depths, we will furthermore clarify that a sample surface structure might not necessarily be the same as its bulk structure. This understanding and the new insights into interfacial properties might prove very important for future improvement of thin-film devices resulting in a reassessment of current optimization approaches.
10:00 AM - M1.2
Combined Cathodoluminescence and Electron Backscatter Diffraction Measurements on Cross-sections of CuInS2 and Cu(In,Ga)Se2 Thin Films for Solar Cells.
Daniel Abou-Ras 1 , Uwe Jahn 2 , Raquel Caballero 1 , Jo Klaer 1 , Melanie Nichterwitz 1 , Thomas Unold 1 , Hans-Werner Schock 1 Show Abstract
1 , Helmholtz Center Berlin for Materials and Energy, Berlin Germany, 2 , Paul Drude Institute for Solid State Electronics, Berlin Germany
Studies of cross-sectional samples reveal features which may affect the current transport in the solar cells. Combination of cathodoluminescence (CL) and electron backscatter diffraction (EBSD) in scanning electron microscopes allow for correlating optoelectronic and microstructural properties of CuInS2 and Cu(In,Ga)Se2 thin films on Mo-coated glass substrates and also in completed solar cells. Acquiring CL images and EBSD maps on the same sample areas, the luminescence properties can be directly related to crystal orientations and grain boundary types at high spatial resolution. Various sample preparation methods for cross-section analysis are presented. Results are discussed with respect to the CL results from CuInS2 layers. These CL images show preferentially high intensities at the top regions of the CuInS2 and Cu(In,Ga)Se2 layers, which can be related to their surface-controlled, topotaxial growth. Lines of reduced CL intensities are found, within the CL images of individual CuInS2 grains, which are attributed to planar crystal defects, containing non-radiative recombination centers. When comparing Σ3 (twin) and non-Σ3 (random) grain boundaries in CuInS2, the drop in CL intensity is much larger at the random grain boundaries, indicating a higher density of defects as compared with Σ3 grain boundaries.
10:15 AM - M1.3
Scanning Tunneling Spectroscopy on the Chalcopyrite Solar Cell Absorber Material Cu(In,Ga)Se2.
Harry Moenig 2 , Ferdinand Streicher 1 , Raquel Caballero 1 , Martha Lux-Steiner 1 2 , Sascha Sadewasser 1 Show Abstract
2 Faculty of Physics, Freie Universität Berlin, Berlin Germany, 1 Solarenergie, Helmholtz-Zentrum Berlin für Materialien und Energie, Berlin Germany
Cu(In,Ga)Se2-based thin film solar cells have reached power conversion efficiencies close to 20% . Nevertheless, little is known about electronic transport and carrier recombination in these materials on a microscopic scale. Especially grain boundaries in these polycrystalline materials are considered to play an important role in their performance [2,3]. We applied scanning tunneling microscopy and spectroscopy to gain more insight in the electronic microstructure of the material. Tunneling spectra of a KCN etched Cu(In,Ga)Se2-surface show a large density of surface states which leads to a metallic characteristic of the tunneling spectra. The occupation of the surface states can be changed by simultaneous illumination of the tunneling junction. Intensity dependent tunneling spectra suggest that occupied surface states are depopulated and unoccupied surface states are populated by laser illumination, which results in tunneling spectra showing a semiconducting characteristic. From those, a band gap can be estimated to (1.3 ± 0.1) eV, which is slightly widened with respect to the bulk band gap of 1.1 eV, in agreement with theoretical predictions of chalcopyrite semiconductor surfaces . Surface oxidation (mainly Ga and In oxides) is believed to lead to a passivation of surface states, especially at grain boundaries. This effect is expected to enhance the device performance . On an oxidized sample we find distinct differences in the tunneling spectra between different grains. Current imaging tunneling spectroscopy (CITS) images show a largely inhomogeneous passivation of the surface. Regions which show a low density of surface states show nearly no effect of laser illumination on tunneling spectra, while regions with a high density of surface states are strongly affected. At the position of grain boundaries the CITS images show a higher density of surface states in the band gap. The results will be discussed in view of defect models and the energy position of different defect complexes.  M.A. Green et al., Prog. Photovoltaics Res. Appl. 16, 61 (2008). D. Fuertes Marrón et al., Phys. Rev. B 71, 033306 (2005). C. Persson et al., Phys. Rev. Lett. 91, 266401 (2003); S. Siebentritt et al., Phys. Rev. Lett. 97, 146601 (2006). S. B. Zhang et al., Phys. Rev. B 57, 9642 (1998). L. Kronik et. al., Thin Solid Films 361-362, 353 (2000).
10:30 AM - M1.4
Preparation for and Microstructural Characterization of CdTe Thin Films using Electron Backscatter Diffraction.
Matt Nowell 1 Show Abstract
1 , EDAX-TSL, Draper, Utah, United States
Electron Backscatter Diffraction (EBSD) is a Scanning Electron Microscope (SEM) based analytical technique that measures crystallographic orientation with sub-micron spatial resolution. EBSD measurements are typically acquired by positioning the electron beam at points along a regular sampling grid, and collecting and processing the resulting diffraction pattern to determine crystallographic phase and orientation. The results can then be presented as orientation maps, where each measurement pixel is colored according to orientation. Often the distance between measurement points is smaller than the grain size of the material of interest. This sampling strategy allows for measurements of grain size, and of orientation variations within individual grains. Information on grain boundary character distributions and special grain boundaries such as twins can also be determined. However EBSD is a surface sensitive technique that requires a relatively smooth surface for consistent results. Cadmium Telluride thin films prepared by a close-spaced sublimation process often have a surface roughness that reduces the effectiveness of the EBSD data collection process. This work examines the use and effectiveness of both broad beam and focused ion beam preparation of CdTe thin films for EBSD analysis, and explores some of the resulting microstructural data that can be obtained with this technique.
10:45 AM - M1.5
In-situ Analysis of Elemental Depth Distributions by Polychromatic Synchrotron Radiation during the Growth of Cu(In,Ga)S2 Thin Films.
R. Mainz 1 , R. Klenk 1 , M. Ch. Lux-Steiner 1 2 , H. Rodriguez 1 , A. Weber 1 , C. Streeck 1 , H. Schock 1 Show Abstract
1 , Helmholtz-Zentrum Berlin, Berlin Germany, 2 Department of Physics, Freie Universität Berlin, Berlin Germany
In this contribution we present a new method for in-situ analysis of elemental depth distributions in which we combine energy-dispersive x-ray diffractometry (EDXRD) and x-ray fluorescence spectroscopy (XRF). We recorded EDXRD reflections and fluorescence signals of the elements copper, indium, gallium and molybdenum simultaneously during growth processes of Cu(In,Ga)S2 thin films out of metallic precursors. By means of the observed diffraction reflections the time evolution of phases in the thin film layers during the growth were determined. We subsequently utilized this phase information to parameterize the depth distributions of the elements in the thin films. Our parameterization model allows mixing between phases and a continuous transition between different layer sequences and assumes lateral homogeneity. The time-dependent fluorescence signals were then taken to determine the free parameters of the parameterized depth distributions. For this latter step we developed a numerical code for the calculation of fluorescence signal intensities for a given set of depth distributions . These calculations handle polychromatic excitation, arbitrary functions of depth distributions and take into account primary and secondary fluorescence. In this work we present both evolutions of phases and of depth distributions of Cu, In, Ga and S during sulphurization processes of metallic Cu-In-Ga precursors. These studies give new insight into the growth mechanism and the formation of gallium gradients during the formation of Cu(In,Ga)S2 thin films in a sequential process. R. Mainz, PhD-Thesis, FU-Berlin Germany (2008).
11:30 AM - **M1.6
Materials-intrinsic Efficiency Limitations in Cu(In,Ga)Se2 Absorbers Due to Metastable Defects.
Stephan Lany 1 , Alex Zunger 1 Show Abstract
1 , National Renewable Energy Laboratory, Golden, Colorado, United States
As the record efficiency of thin-film solar cells approaches the 20% threshold, the question arises what are the materials-intrinsic performance limiting factors? Cu(In,Ga)Se2 absorbers suffer from the fact that the open-circuit voltage stays way short of the band-gap energy, in particular when the gap is increased towards the optimum in the solar spectrum or for the purpose of a top-cell in a tandem. Our recent theoretical studies  revealed that the prominent light- and bias-induced metastability effects observed in Cu(In,Ga)Se2 absorbers can be explained by the properties of the VSe-VCu divacancy complex. Very recently, we predicted that also the intrinsic InCu and GaCu donors, and their complexes with VCu, exhibit two different stable atomic configurations , and can account for a specific type of metastability, i.e., the "red-on-bias" metastability.
Besides explaining different types of metastabilities observed experimentally, we predict that these centers act as minority-carrier traps and will limit the open-circuit voltage VOC around 1 eV. While the metastable defects are rather benign for CIGS compositions below ~30% Ga content, they are predicted to prevent the achievement of VOC significantly above 1eV when increasing the band gap by higher Ga contents. Based on the dependence of the defect densities on the growth conditions, we expect that the metastable defects are minimized under Cu-rich conditions when the CIGS films are close to stoichiometry, in contrast to the Cu-poor stoichiometry used for most high-efficiency cells. Thus, we suggest that any attempt to increase VOC by increasing the band gap need to be accompanied by a change of growth conditions, but we also highlight the need to address simultaneously other sources of VOC limitation that may exist in parallel.
 "Light- and bias-induced metastabilities in Cu(In,Ga)Se2 based solar cells caused by the (VSe-VCu) vacancy complex", S. Lany and A. Zunger, J. Appl. Phys. 100, 113725 (2006)
 S. Lany and A. Zunger, "Intrinsic DX centers in ternary chalcopyrite semiconductors", Phys. Rev. Lett. 100, 016401 (2008)
12:00 PM - M1.7
Characterizing the Effects of Silver Alloying in Chalcopyrite CIGS with Junction Capacitance Methods.
Peter Erslev 1 , J. David Cohen 1 , Gregory Hanket 2 , William Shafarman 2 Show Abstract
1 Department of Physics, University of Oregon, Eugene, Oregon, United States, 2 Institute of Energy Conversion, University of Delaware, Newark, Delaware, United States
Thin film solar cells based on the alloys of CuInSe2 (CIS) hold substantial promise for an economical source of photovoltaic energy. Because the band gap of CIS (~1 eV) does not optimally match the solar spectrum, it is normally alloyed with Ga (CIGS) to produce higher band gaps (up to 1.7 eV for CuGaSe2). Unfortunately, the predicted increase in efficiency for band gaps above roughly 1.2 eV has not yet been realized, primarily due to a deficit in the open circuit voltage. In this report we examined several sets of (AgCu)(InGa)Se2 (ACIGS) devices in hopes that such alloys will allow an increase the band gap beyond 1.2 eV in a manner that is reflected in correspondingly higher values of VOC. The endpoint AgInSe2 and AgGaSe2 compounds have band gaps 0.1 – 0.2 eV higher than the corresponding Cu compounds. In addition, because the melting point of the Ag alloys are typically 200 °C lower than the corresponding Cu alloys, one might also expect reduced structural disorder for the Ag alloys and, correspondingly, better electronic properties.ACIGS films have been deposited by elemental co-evaporation and devices were fabricated using a structure of glass/Mo/ACIGS/CdS/ZnO/ITO. We studied 2 series of ACIGS sample devices: one series of 4 samples with a constant Ag/(Ag+Cu) ratio near 0.15 with variable Ga/(In+Ga) fractions from 0.3 to 0.8, and a series of samples with Ga/(In+Ga) ≈ 0.8 and Ag/(Cu+Ag) ratios varying from 0.15 to 0.8. The resultant device efficiencies were similar to the corresponding alloys without Ag, with a highest efficiency to date of 16.1% (VOC of 0.70 V) for a sample device with Ag and Ga fractions of 0.16 and 0.45, respectively.The electronic properties of the ACIGS absorbers in these devices were characterized using the junction capacitance methods of drive level capacitance profiling (DLCP), admittance spectroscopy, and transient photocapacitance spectroscopy (TPC). One of the most significant effects of Ag alloying appears to be a decrease in the Urbach energy (the characteristic energy of the band tail) by up to 10 meV compared to corresponding CIGS samples (typically below 15 meV for ACIGS samples compared to 20-25 meV for CIGS). Indeed, one sample with Ag/(Cu+Ag) = 0.3 exhibited an Urbach energy near 10 meV, by far the smallest value we have ever recorded and close to our resolution limit. However, neither the corresponding band gaps nor the values of VOC showed any significant increases. DLCP indicated free carrier densities near 1x1014 cm-3 for most of the samples and deep defect levels which range from 1x1014 cm-3 to >1x1015 cm-3, roughly an order of magnitude lower than CIGS samples. Comparing these DLCP profiles to standard CV profiles indicates a large defect density at or near the CdS/absorber layer interface for several of these samples. We shall discuss whether this may provide a partial reason for the smaller than expected open-circuit voltages in these ACIGS sample devices.
12:15 PM - M1.8
Metastable Variations of the Fill Factor in CIGS Thin Film Solar Cells.
Aleksander Urbaniak 1 , Malgorzata Igalson 1 Show Abstract
1 Physics, Warsaw University of Technology, Warszawa Poland
A variety of metastable features depending on the history of the sample is observed in CIGS solar cells. In present work we focus on the origin of the fill factor changes induced by a reverse bias treatment. As this parameter controls solar cell efficiency its metastable behavior should be regarded as a concern for further improvement of the CIGS technology.Stressing a cell with the reverse voltage bias at temperature above 250 K leads to deterioration of its current-voltage (I-V) characteristic. Particularly, this is manifested in losses of the fill factor parameter. When the stress is removed the fill factor slowly relaxes to its steady-state value. We show that this processes are temperature dependent and have a reflection in the amount of negative charge accumulating in the near-interface region of the absorber layer.The evolution of current-voltage characteristics have been measured during stressing a cell with the reverse voltage bias and after the bias has been removed. From the characteristics, time decays of the fill factor changes induced by bias treatment have been obtained and correlated with kinetics of junction capacitance, recorded at the same conditions. Additionally capacitance-voltage characteristics have been measured to obtain net acceptor density profilesWe show that metastable variations of the fill factor have their source in the absorber layer and are related to the increased number of shallow acceptors, what can be directly observed in the shape of space charge distribution profiles. In general the kinetic of the fill factor changes consist of two steps, both having a reflection in the capacitance. When the reverse voltage bias is applied emission of holes leads to the increase of negative charge in the junction. This is manifested in the enhanced capacitance and deterioration of the fill factor. At longer times the rate of the fill factor changes is reduced and this fact is correlated with a slow decrease of the capacitance. From the point of I-V characteristics this process corresponds to enhanced double - diode behavior. We discuss our results within a model of p+ layer, showing that the change of the negative charge accumulated in the near-interface region of the absorber controls the fill factor value. We propose that the source of this negative charge is a shallow acceptor state of a VSe-VCu defect center. In the discussion we include the influence of the position of Fermi level in the absorber on the distribution of divacancies between donor and acceptor configurations.
12:30 PM - M1.9
Effects of Ga Compositional Grading on CIGS Electronic Properties Relevant to Solar Cell Performance.
JinWoo Lee 1 , Jeroen K. J. van Duren 2 , Alex Pudov 2 , Miguel Contreras 3 , J. David Cohen 1 Show Abstract
1 Department of Physics, University of Oregon, Eugene, Oregon, United States, 2 , Nanosolar, inc., San Jose, California, United States, 3 , National Renewable Energy Laborotory, Goldon, Colorado, United States
Junction capacitance methods were used to investigate the effects of Ga grading on a set of four CuIn1-xGaxSe2 (CIGS) samples grown by varying vacuum co-evaporation methods at the National Renewable Energy Laboratory (NREL). One device was deposited with a uniform Ga depth profile (x=0.30, with x=Ga/(In+Ga), η=15.1%) to serve as a reference device, one device close to NREL's optimally bandgap grading scheme (“v-shaped”, with x=0.5 near Mo, x=0.2 at roughly 0.5 μm from barrier junction, x=0.35 near barrier junction, η=16.3%), and two with the Ga fraction roughly linearly decreasing from x=0.9 at the Mo back contact to either x=0.15 (η=17.1%), or x=0.25 (η=15.3%). The compositionally graded devices all had values of Voc close to 0.66 V; Voc for the uniform control sample was slightly lower (0.64 V).Perhaps the most revealing effects of Ga grading were obtained from our transient photocapacitance (TPC) and photocurrent (TPI) spectra. These measurements provide a spectral map of the optically induced release of carriers for photon energies below 1 eV to 2 eV. By comparing the two types of spectra one can distinguish majority from minority carrier processes. Spatial information is obtained by varying the applied DC bias to weight the spectral response to different regions relative to the barrier junction. In all cases these spectra clearly identified the gap energy of the lowest Ga region for each sample (1.0 eV to 1.2 eV depending on the sample). Very narrow bandtails (varying from 13 to 23 meV) were observed indicating very low structural disorder. In the ungraded sample the optical response above 1.2 eV was nearly constant for both the TPC and TPI spectra as expected.The two samples with Ga fractions that increased monotonically with distance from the junction exhibited a decreasing TPI response but a slightly increasing TPC response for photon energies above 1.5 eV. The latter partly reflects the increasing bandgap; however, the decreasing TPC/TPI ratio clearly indicates an increased minority carrier (electron) collection in these samples. This agrees with the general understanding of how compositional grading improves overall cell performance. Evidence for increased minority carrier collection was even stronger in the sample device incorporating the v-shaped Ga-grading, particularly for photon energies 1.2 to 1.4 eV. These spectra also exhibited a quite complex temperature dependent behavior whose details are still being analyzed. However, we believe that they will provide a very complete picture of how compositional grading affects the minority collection in CIGS devices.Spatial profiles of the free hole carrier densities and deep acceptor concentrations were examined using drive-level capacitance profiling (DLCP). Results in the compositionally graded sample indicated that the free carrier density decreased and that defect density increased as the Ga fraction increased toward back contact. Possible reasons for this will be discussed.
12:45 PM - M1.10
Hole Photocarrier Drift Mobility Measurements in Polycrystalline CuIn1-xGaxSe2.
Steluta Dinca 1 , Eric Schiff 1 , Brian Egaas 2 , Rommel Noufi 2 3 , David Young 2 , William Shafarman 4 Show Abstract
1 Department of Physics, Syracuse University, Syracuse, New York, United States, 2 , National Renewable Energy Laboratory, Golden, Colorado, United States, 3 , SoloPower, Inc, San Jose, California, United States, 4 , Institute of Energy Conversion, University of Delaware, Newark, Delaware, United States
We measured temperature dependence of the hole photocarrier drift-mobility in polycrystalline CuIn1-xGaxSe2 (CIGS) thin films  incorporated into solar cell structures. We used the time-of-flight technique on twelve cells manufactured at National Renewable Energy Laboratory and at the Institute of Energy Conversion.We found drift-mobilities ranging from 0.10 to 1.0 cm2/Vs at 295 K, and a very weak temperature-dependence in the range of 100 – 300 K. These drift-mobilities are somewhat smaller than those reported from previous Hall Effect and admittance measurements. The distinction between these drift-mobility measurements and previous Hall effect work is that hole drift-mobilities are sensitive to traps (localized bandtail states) near the valence bandedge. The weak temperature-dependence indicates that the width of the valence bandtail is less than 10 meV. Previous optical absorption spectroscopy indicates that Urbach edges in similar CIGS samples are 20 meV or larger. We propose that the drift-mobilities are properties of a disorder-induced mobility-edge at the valence bandedge of CIGS. Both the magnitude and the temperature-dependence of the drift-mobility in CIGS are comparable to results in nanocrystalline and amorphous silicon that are considered to exhibit mobility-edge behavior. However, for polycrystalline CIGS the source of nanometer scale disorder is likely to be chemical composition fluctuations instead of non-crystallinity. This research was supported by the National Renewable Energy Laboratory through subcontracts NDJ-2-30630-24 (Syracuse University) and XAT-4-33624-01 (University of Delaware).. Submitted to Phy. Rev. B
Clemens Heske University of Nevada-Las Vegas
Miguel Contreras National Renewable Energy Laboratory
Malgorzata Igalson Warsaw University of Technology
Stuart J. C. Irvine OpTIC Technium/NEWI
Akira Yamada Tokyo Institute of Technology
M5: Poster Session
Wednesday PM, April 15, 2009
Salon Level (Marriott)
9:00 PM - M5.1
Influence of Se Beam Pressure on Defect Properties in CIGS Solar Cells Fabricated by Three-stage Process.
Takeaki Sakurai 1 , Monirul Islam 1 , Hideyuki Uehigashi 1 , Shogo Ishizuka 2 , Keiichiro Sakurai 2 , Akira Yamada 2 , Koji Matsubara 2 , Shigeru Niki 2 , Katsuhiro Akimoto 1 Show Abstract
1 Institute of Applied Physics, University of Tsukuba, Ibaraki Japan, 2 , Advanced Industrial Science and Technology (AIST), Ibaraki Japan
An influence of Se beam pressure on defect properties in CIGS solar cells fabricated by three-stage process was investigated by admittance spectroscopy measurements. CIGS films were prepared on Mo-coated soda-lime glass substrates by the three-stage process using a molecular beam epitaxy (MBE) system. The Se beam pressure was varied in the range from 1.8×10-3 to 4.4×10-3 Pa, at the position of the samples, by varying Se cell temperature. The cell structure was fabricated by the sequential deposition of CdS buffer layer with chemical bath deposition and ZnO window layer with rf magnetron sputtering. The admittance spectroscopy measurements were performed in the temperature range from 10 to 350 K with the frequencies between 1KHz and 1MHz. For all samples, the spectra show three distinct peaks denoted as α, β, and ζ, of which activation energies were estimated to be 10, 100, and 300 meV, respectively. We have proposed the origin of these three peaks as due to the shallow acceptor, traps at CdS (or CdS/CIGS interface), and defects in CIGS, respectively. The trap density of ζ increased from 1.9×1014 to 1.3×1015 cm-3 with decreasing Se beam pressure, although those of α and β were not varied. These results suggest the origin of the defect ζ is concerned with the Se deficiency. Since Cu/III ratio of CIGS films was found to be decreased with decreasing Se beam pressure, it may be possible to consider the defect structure of ζ as VCu-VSe complex, which will act as acceptor type defect. The density of ζ seems to correlate with the cell efficiency. Therefore, to control Se beam pressure is much important for obtaining the highly efficient solar cells.
9:00 PM - M5.10
Sulfurization Growth of SnS Films and Fabrication of SnS-related Solar Cell.
Mutsumi Sugiyama 1 , Keisuke Miyauchi 1 , Takehiro Minemura 1 , Katsuya Nakagomi 1 , Yuichiro Mori 1 , Naoki Kawada 1 , Hisayuki Nakanishi 1 Show Abstract
1 Department of Electrical Engineering, Tokyo University of Science, Noda, Chiba Japan
Binary IV-chalcogenide orthorhombic semiconductor tin sulfide (SnS) has direct bandgap energy of 1.3 eV and a high absorption coefficient of 10^5 cm-1. Therefore, SnS is a promising candidate as a light-absorbing medium for next-generation solar cells. Recently, Cu(In,Ga)(S,Se)2 (CIGS) and CdTe have attracted attention as solar cells with high conversion efficiency. However, these alloys are costly and cause environmental concerns, since In and Ga are rare metals and Se, Cd and Te are toxic. On the other hand, the elements comprising SnS are cheaper than those comprising CIGS. In fact, Sn and S are abundant in the Earth’s crust: they have Clark numbers of the order of 31 and 15, respectively. Since these elements are safe for both the environment and the human body, SnS is believed to be a safe semiconducting material.SnS films have been deposited by a variety of methods, including chemical bath deposition and sulfurization. Among these, sulfurization methods are the most desirable because they can economically deposit large-area films with well-controlled compositions. In fact, selenization and sulfurization, which are reactive solid-phase growth methods using chalcogenide (Se and/or S) vapor, are the most promising preparation methods for producing high-quality CIGS thin films. Moreover, several groups have recently fabricated SnS-based solar cells. From a technological viewpoint, sulfurization growth of SnS films has great advantages in the fabrication of SnS-based solar cells because it is simple and a dry process. However, the sulfurization conditions have not yet been clarified. For the realization of a SnS-related solar cell, the development of a conventional sulfurization technique is needed. In this presentation, we will introduce the advantages of using low-temperature sulfurization for the growth of SnS films are presented along with the characteristics of an n-CdS/p-SnS heterojunction for the SnS-related solar cell.SnS films were grown by sulfurization after tin layers were deposited on soda-lime glass (SLG) substrates using evaporation technique. The films structure was characterized by x-ray diffraction (XRD), transmittance, scanning electron microscope (SEM), and Hall-effect measurements.Polycrystalline SnS films were grown by sulfurization of a Sn precursor at low temperatures of 120–220°C. The SnS film grown at 170°C comprises densely packed 3-5-um-diameter columnar grains, which is appropriate for use in the photoabsorption layers of solar cells. The SnS film had an optical bandgap of approximately 1.3 eV and p-type conductivity. Using an appropriate SnS film, an n-CdS/p-SnS heterojunction was fabricated on Mo-coated soda-lime glass substrates. These results are the first step toward realizing an optical device as a solar cell using a SnS film grown by sulfurization.
9:00 PM - M5.11
Thermal Management of Nanostructured Semiconductor Photovoltaics.
Suchismita Ghosh 1 2 , Qinghui Shao 1 2 , Alexander Balandin 1 2 , Seth Hubbard 3 , Christopher Bailey 3 Show Abstract
1 Electrical Engineering, University of California - Riverside, Riverside, California, United States, 2 Materials Science and Engineering Program, University of California - Riverside, Riverside, California, United States, 3 Physics, Rochester Institute of Technology, Rochester, New York, United States
Nanostructured solar cells offer a number of advantages over the bulk technologies. Some of them include the ability to exceed the single-junction efficiency in multi-junction designs; tailoring the band-gap over a large range; and potentially low-cost self-assembled nanostructures. It was recently suggested that quantum dot arrays can be used for implementing the intermediate-band solar cell concept . The use of nanostructured and thin-film compound semiconductor solar cells in the concentrated solar energy system raises the issue of efficient heat removal from photovoltaic cells. The thermal management is particularly important for solar cells based on nanostructures and thin films due to the reduction of thermal conductivity. Indeed, the thermal conductivity of quantum dot superlattice can be much lower than bulk crystals due to phonon scattering from the dots, alloy disorder related point defects and interfaces . Here we report experimental investigation of the thermal conductivity of the standard GaAs-based p-i-n solar cells without quantum dots and with five layers of InAs quantum dots in the intrinsic region. We also examined QD solar cells based on structures with a strain compensation layer. The results obtained for the reference and quantum dot solar cell structures were compared with the thermal conductivity of bulk GaAs crystals. The measurements were carried out using the transient planar source “hot-disk” technique in the relevant temperature range from -20C to 100C. It was found that the thermal conductivity of solar cells with quantum dots is smaller than that of bulk GaAs by a factor of ~14. The thermal conductivity of the reference solar cells is larger than that of quantum dot solar cells by about a factor of 2. The thermal conductivity of the quantum dot solar cells slowly increases with temperature T, which is characteristic for the disordered materials. This is in contrast to the ~1/T-type dependence of the thermal conductivity of bulk crystal semiconductors. A drastic reduction of the thermal conductivity and corresponding increase in the thermal resistance of the solar cell structure may have important implication for the operation of the nanostructure-based photovoltaic cells, particularly with solar concentrators. The increase in the thermal resistance is expected to lead to larger temperature rise, which affects the photovoltaic efficiency.This work was supported in part by DOE Solar Energy Technologies Program. The work in Balandin group was also supported by the AFOSR contract FA9550-07-C-0059 and NASA contract NNC07CA20C. Q. Shao, A. A. Balandin, A. I. Fedoseyev and M. Turowski, “Intermediate-band solar cells based on quantum dot supra-crystals”, Appl. Phys. Lett. 91, 163503 (2007). M. Shamsa, W.L. Liu, A.A. Balandin and J.L. Liu “Phonon-hopping thermal conduction in quantum dot superlattices," Appl. Phys. Lett., 87, 202105 (2005).
9:00 PM - M5.12
Surface Texturing by Natural Lithography using SiO2 Nanospheres on GaAs Solar Cells.
Byung-Jae Kim 1 , Joona Bang 1 , Jihyun Kim 1 Show Abstract
1 of Chemical and Biological Engineering, Korea University, Seoul Korea (the Republic of)
Surface texturing of substrate was effective to minimize the reflectance. Surface texturing such as pyramid structures has been considered to reduce the reflection of the sun light on surface of solar cells. The efficiency of Si solar cells was 19.8% in multicrystalline Si solar cells and 24.4% in monocrystalline Si solar cells by surface texturing with lithography technique.[Zhao et al. Applied Physics Letters 73(14), 1991(1998)] In commercial Si solar cells, Si surface was textured by KOH-based wet etching. KOH-based wet etching has major drawbacks to control the etch rate and evaporation the metals. In our experiments, SiO2 nanospheres were employed to texture the GaAs surface. The surface of GaAs was coated with polymer. Then, the surface of polymer was treated with oxygen plasma to make the surface hydrophilic due to the nature of hydrophobic surface of polymer. After plasma treatment, SiO2 nanospheres were spin-cast on the surface of polymer, followed by annealing at 160 degree celcius for 5 seconds to sink SiO2 nanospheres into the polymer layer. Consequently, the surface was fabricated with nanolens arrays of SiO2 nanospheres. Our group used Benzocyclobutene(BCB) as polymer, because BCB has been widely used passivation material in semiconductor process due to compatibility with IC industry. By controlling the spin-speed and concentrations of solution, we achieved the thickness of BCB layer from 50nm to 25µm. SiO2 nanospheres were synthesized by StÖber methods. The diameters of SiO2 nanospheres were controlled by mole fraction of NH3. In addition, the dimple surface was fabricated by the same process. The hemisphere structures of surface were changed the dimple structures by HF-based wet etching because HF-based wet etching removed SiO2 nanospheres. In addition, we fabricated the needle structures on GaAs surface by ICP dry-etching. The spin-casted SiO2 nanospheres were used as the etch masks. The etching gas is a mixture of BCl3 and Cl2. The etch rate of SiO2 was slower than that of GaAs in the etching conditions. The depth of needles was controlled by the diameter of SiO2 nanospheres and etching times. The details about the texturing process will be presented at the conference.
9:00 PM - M5.13
Fabrication of Cu(In,Ga)Se2 Films by a Combination of Mechanochemical Synthesis, Wet Bead Milling, and a Screen Printing/sintering Process.
Junya Kubo 1 , Mishihiro Matuo 1 , Takahiro Wada 1 , Akira Yamada 2 3 , Makoto Konagai 2 3 Show Abstract
1 Department of materials chemistry, Ryukoku University, Otsu, shiga, Japan, 2 Department of Physical Electronics, Tokyo Institute of technology, Meguro-ku, Tokyo, Japan, 3 Photovoltaic Research Center, Tokyo Institute of technology, Meguro-ku, Tokyo, Japan
CuInSe2 (CIS) and its solid solutions with Ga and S are excellent thin-film photovoltaic materials. However, such vacuum deposition processes as “physical vapor deposition” and “sputtering and selenization,” which are typically used to fabricate CIS thin-film photovoltaic (PV) devices, are complex and expensive. Recently, some research groups have proposed various non-vacuum deposition techniques for fabricating CIS solar cells. We studied the fabrication of Cu(In,Ga)Se2 films by a combination of mechanochemical and screen printing/sintering processes [1,2]. CIGS powder suitable for screen-printing was prepared using a mechanochemical process (MCP) [2,3]. Particulate precursor ink was prepared by mixing the obtained CIGS powder with an organic solvent. The particulate precursors were deposited in a thin layer by a screen-printing technique, the remaining organic solvent was removed from the screen-printed CIGS film, and finally the porous precursor layer was sintered into a dense polycrystalline film by atmospheric-pressure firing. The CIGS solar cells with our standard Al grid/B-doped ZnO/i-ZnO/CdS/CIGS/Mo/soda-lime glass structure showed an efficiency of 2.7%, a open circuit voltage (Voc) of 0.325 V, a short circuit current density (Jsc) of 28.3 mA/cm2, and a fill factor (FF) of 0.295 . To obtain a higher energy conversion efficiency of the CIGS solar cells fabricated by a combination of mechanochemical and screen printing/sintering processes, we studied the preparation process of particulate precursors ink suitable for screen-printing and prepared the particulate precursors ink by wet bead milling, which is usually used for dispersing nanoparticles into liquids. The present wet bead milling was effective for preparing fine and homogeneous CIGS precursor powder. The grain size of the CIGS powder prepared by wet bead milling was 0.1 -0.3 μm, which is smaller than that of our conventional CIGS power. We fabricated a better CIGS absorber layer with a larger grain size of about 3 μm. The CIGS solar cell fabricated by a combination of mechanochemical synthesis, wet bead milling, and a screen printing/sintering process showed an efficiency of 3.1%, a Voc of 0.279 V, a Jsc of 28.8 mA/cm2, and a FF of 0.386. This work was supported by the Incorporated Administrative Agency New Energy and Industrial Technology Development Organization (NEDO) under the Ministry of Economy, Trade and Industry (METI).  T. Wada et al. phys. stat. sol. (a) 203, 2593-2597 (2006).  S. Nomura et al. MRS Proceedings 1012, Y03-15 (2007).  T. Wada et al. Thin Solid Films 431-432, 11-15 (2003).  T. Wada and H. Kinoshita, J. Phys. Chem. Solids 66, 1987 (2005).
9:00 PM - M5.14
Electron Beam Induced Current Studies of Grain Boundaries in Cu(In,Ga)Se2 Thin-film Solar-cells.
Melanie Nichterwitz 1 , Daniel Abou-Ras 1 , Raquel Caballero 1 , Juergen Bundesmann 1 , Thomas Unold 1 , Roland Scheer 1 , Hans-Werner Schock 1 Show Abstract
1 , Helmholtz-Zentrum Berlin für Materialien und Energie, Berlin Germany
The influence of grain boundaries on charge carrier collection in Cu(In,Ga)Se2 thin-film solar-cells was studied by means of electron-beam induced current (EBIC) and electron backscatter diffraction (EBSD) measurements in a scanning electron microscope. In order to reduce topography effects on the EBIC signal, polished instead of fractured cross sections of standard Cu(In,Ga)Se2 thin-film solar-cells with device efficiencies larger than 15% were used. It was possible to locate and classify grain boundaries of the Cu(In,Ga)Se2 absorber layer in a two dimensional EBSD map and directly study the possible modification of charge-carrier collection properties at these positions by means of EBIC. In the present work, difficulties of the used measurement setup and evaluation method are discussed, which are due to the fact that microstructure and charge-carrier collection properties are displayed in two- and not three-dimensional images. EBIC line-profiles across various grain boundaries are presented. The results indicate that grain-boundary recombination has a negative effect on charge carrier collection in some cases. The corresponding recombination velocities, which were roughly estimated, are rather low (in the range of 1×10^4 cm/s). This shows that grain boundary recombination plays a minor role in current standard devices but should be taken into account when attempting to achieve efficiencies above 20%.
9:00 PM - M5.15
Photocurrent Transport in Cu(In,Ga)S2 Based Solar Cells with High Open Circuit Voltage - Bulk vs. Interface
Saoussen Merdes 1 , Benjamin Johnson 1 , Joachim Klaer 1 , Reiner Klenk 1 , Iver Lauermann 1 , Roland Mainz 1 , Alexander Meeder 2 Show Abstract
1 SE2, Helmholtz-Zentrum-Berlin, Berlin Germany, 2 , Sulfurcell Solartechnik GmbH, Berlin Germany
Cu(In,Ga)S2 thin films prepared by rapid thermal processing of metallic precursors yielded solar cells with efficiencies reaching 12.9% . A good short circuit current density was observed together with open circuit voltages up to 850 mV. However, the fill factor was close to, but typically not exceeding, 70% . The dark jV curve shows a distinct two diode behavior which was not previously observed in Cu(In,Ga)S2-based cells. It is conceivable that this contributes to the non-optimum fill factor. Furthermore, results indicate that the photocurrent is voltage-dependent, i.e. a non-optimum transport of the photocurrent. Process control of the buffer layer preparation suggests different absorber surface properties, compared to the Ga-free reference, and the need to re-adjust cell preparation parameters in order to fully exploit the efficiency potential of the absorber layers. This implies that the interface plays a significant role for photocurrent transport. In this contribution, in addition to the straightforward approach of quantifying the effects of CdS deposition parameters, the problem is addressed using device modeling and careful surface and interface analyses. Cu(In,Ga)S2 surfaces are analyzed using near edge X-ray absorption fine structure (NEXAFS) for the conduction band edge and ultraviolet photoelectron spectroscopy (UPS) for the valence band edge. The correlation between absorber structure, buffer layer properties and cell performance is discussed.  S. Merdes, R. Kaigawa, J. Klaer, R. Klenk, R. Mainz, A. Meeder, N. Papathanasiou, D. Abou-Ras, S. Schmidt, Proc. 23rd European Photovoltaic Solar Energy Conference, Valencia (2008). R. Mainz, J. Klaer, R. Klenk, N. Papathanasiou, Proc. 22nd European Photovoltaic Solar Energy Conference, Milan (2007) 2429.Keywords: Chalcopyrite, Cu(In,Ga)S2, Rapid Thermal Processing, CdS
9:00 PM - M5.16
Characterisation of Secondary Phases in Cu Poor CuInSe2: Raman Scattering in-depth Resolved Analysis of Polycrystalline Layers.
Victor Izquierdo-Roca 1 , Xavier Fontane 1 , Lorenzo Calvo-Barrio 1 , Jacobo Alvarez-Garcia 1 2 , Alejandro Perez-Rodriguez 1 , Joan Ramon Morante 1 , Wolfram Witte 3 , Reiner Klenk 4 Show Abstract
1 EME/XaRMAE/IN<sup>2</sup>UB, Departament d'Electrònica , Universitat de Barcelona, Barcelona Spain, 2 , Centre de Recerca i Investigació de Catalunya (CRIC), Barcelona Spain, 3 , Zentrum für Sonnenenergie- und Wasserstoff-Forschung Baden-Württemberg (ZSW), Stuttgart Germany, 4 , Helmholtz-Zentrum Berlin für Materialien und Energie, Berlin Germany
Identification of secondary phases in Cu poor CuInSe2 polycrystalline layers has a strong interest because of their potential impact on the characteristics of the solar cells fabricated with these absorbers. Absorbers with Cu content below the stoichiometric composition of CuInSe2 are used in order to avoid the presence of CuySe domains that are formed when depositing the stoichiometric compound with large areas. In these layers, the formation of a surface region with a chalcopyrite ordered Cu poor ordered vacancy compound (OVC) has been reported, and the presence of this surface phase is assumed to affect the band structure at the heterojunction. However, little is known about the possible presence of such secondary phases inside the layer, as well as their possible coexistence with other phases as CuAu ordered CuInSe2.This work reports a Raman scattering in-depth resolved analysis of Cu poor CuInSe2 layers with different composition (being the Cu content in the range between 16 at. % and stoichiometric composition). Raman scattering is specially well suited for the characterisation of OVC and CuAu ordered phases in CuInSe2, being the position of the main A1 vibrational mode in the Raman spectra very sensitive to the phase structure and composition in the scattering volume [1,2]. Combined in-depth Raman/Auger Electron Spectroscopy (AES) measurements were performed by acquiring sequentially a series of Raman spectra after sputtering the layers in the AES system between every Raman measurement. The analysis of the spectra measured at different depths from the samples with lowest Cu content has allowed to confirm the formation of a surface region with a high OVC content, as well as to observe the formation of CuAu ordered domains inside the layers. In addition, a significant increase is observed in the intensity of OVC modes in the Raman spectra measured at the interface region with the back Mo layer. These data suggest a preferential formation of OVC in Cu poor layers at both the surface and back interface regions. The impact of the presence of this back region with high content of OVC domains on the formation of an interfacial MoSe2 layer is also investigated, as function of the Cu content in the layers. Presence of secondary phases at the back CuInSe2/Mo interface has a special relevance, because of their potential impact on the electrical characteristics of the back contact in the solar cells. These data will be correlated with the chemical composition of the layers, identifying the range of compositions critical for the formation of such phases in the polycrystalline layers.1.- J. Álvarez-García et al, Physical Review B 71 (2005) 054303. 2.- C.-M. Xu et al, Semicond. Sci. Technol. 19 (2004), 1201.
9:00 PM - M5.17
A Rapid Screening Method for Investigating the Effect of Processing Parameters on CdTe/CdS Solar Cell Performance.
Mohammed Al Turkestani 1 , Ken Durose 1 Show Abstract
1 , Durham University, Durham United Kingdom
A rapid screening method is reported in which material processing parameters are investigated as a function of the CdTe absorber thickness in CdTe/CdS solar cells. It has been used to investigate the effect on device performance of post-growth annealing of CdS layer with H 2, N 2, and O2. In the method the CdTe layer was grown and then prior to contacting it was chemically bevelled. The devices were then CdCl2 treated, etched and matrix of contacts then applied, each of which forms a device having a different absorber thickness. It is therefore possible to investigate ~ 25 different CdTe thickness (02, N2 and O2 at 400 °C for 4 mins, with both CSS and sputter grown material being studied. The effects on EQE, J-V performance and the shunt and series resistance were measured. Preliminary results for CSS-CdS indicated that annealing in H2 gives some benefits to VOC. However, for devices with h<3µm the shunt resistance decreased markedly for all samples. The physical integrity of the devices was studied by SEM and EBIC. The influence of systematic errors is discussed. Scattering in the data was attributed to variation in the CdS thickness. To eliminate this a second experiment series was made using sputtered CdS having a thickness of 200 ±10 nm. Results will be presented that indicate the benefits of particular annealing steps as function of CdTe thickness. Prospects for making CSS-grown cells with ultra-thin absorbers shall be discussed.
9:00 PM - M5.18
Temperature Dependence of Sodium Diffusion During the Bilayer Process of CIGS Deposition.
Ryosuke Kaneko 1 2 , Keiichiro Sakurai 2 , Shogo Ishizuka 2 , Hironori Komaki 2 , Yukiko Kamikawa-Shimizu 2 , Shuhei Shimada 1 2 , Koji Matsubara 2 , Hajime Shibata 2 , Akimasa Yamada 2 , Mutsumi Sugiyama 1 , Hisayuki Nakanishi 1 , Shigeru Niki 2 Show Abstract
1 Department of Electrical Engineering, Tokyo University of Science, Noda, Chiba, Japan, 2 Research Center of Photovoltaics, National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, Japan
CIGS-based-solar cells are among the most promising candidates for high-efficiency and low production cost solar cells.At laboratory scale, the record efficiency at ~20% have been achieved by the three-stage process, a variation of the co-evaporation method.However, the conversion efficiencies of the commercially available CIGS modules are still lower than those expected from the lab-scale records.Though higher module performances are expected by improving the deposition process, the three-stage process is relatively lower in productivity and higher in initial investment cost, than the already commercially applied processes.If the process could be simplified while maintaining the efficiency, it would further improve the cost-effectiveness of CIGS modules.One candidate is the "bilayer process", one of the co-evaporation method.Literally consisting of 2 stages, the bilayer process is a simpler process.So far the reported record efficiency is ~16% with AR coating, which is considerably lower than that of the three-stage process.Nevertheless, considering the recent progress on the co-evaporation method, there would be room to improve this already well-known technique.One point of interest is the sodium diffusion into the CIGS layer.It is well known that the sodium diffusion can improve the cell performance.However, in the case of the bilayer process, sodium diffusion have not been studied intensively as the case of the three-stage method.In this work, we have investigated the relationship between the sodium diffusion and the the deposition temperature during the bilayer process, especially focusing on the highest temperatures possible on SLG substrates.CIGS films were fabricated with the bilayer process.At the first stage, Cu, In, Ga and Se were co-evaporated on the substrate in a Cu-rich condition.Then, as the second stage, In, Ga and Se were subsequently evaporated.Consequently, a slightly Cu-poor CIGS film was fabricated.The substrate temperature (Tsub) was kept constant at 520 ~ 600 degrees C throughout the growth.The sodium composition in the surface region of the film was measured by EMPA.The film structure was observed by SEM.CIGS cells with ZnO/CdS/CIGS/Mo/SLG structure were fabricated for performance evaluation. The cell performance varied with Tsub.The FF and the VOC increased with higher Tsub.On the other hand, the JSC remained unchanged.An efficiency of 15.7% without AR coating was achieved at 600 degrees C.The grain size of the CIGS increased to about 2 micrometers at 600 degrees C from a smaller 1 micrometer at 520 degrees C.The sodium content was higher at the surface of the 600 degrees C sample, compared to the 520 degrees C sample. These preliminary results suggest that the performance of the bilayer-processed CIGS solar cells can be further improved by optimizing.
9:00 PM - M5.19
Properties of P-doped ZnO Films Deposited at Different Temperature by RF Magnetron Sputtering.
Jinzhong Wang 1 , Elamurugu Elangovan 1 , Zhiliang Pei 1 , Vincent Sallet 2 , Ana Rego 3 , Rodrigo Martins 1 , Elvira Fortunato 1 Show Abstract
1 Material Science Department, CENIMAT, Caparica, Setubal, Portugal, 2 Groupe d'Etude de la Matière Condensée, CNRS-UVSQ, Meudon France, 3 Centro de Química-Física Molecular, Institute of Nanoscience and Nanotechnology, Lisboa Portugal
As a promising candidate for blue light material, ZnO has attracted much attention due to its wide band gap and large exciton binding energy. It is well known that p type ZnO is very difficult to achieve. In order to overcome this barrier, many p type doping elements have been studied and good results have been achieved. However, most of p type ZnO films were deposited at high temperatures ( > 400 degree centigrade ). In this work, we present P-doped ZnO films sputtered at low temperature ( from RT to degree centigrade ) and their properties are studied. XRD spectra confirm that all the films were grown along the preferable <001> orientation. Further, the intensity of ZnO (002) peak decreases against the increase in the deposition temperature and the peak position shifts towards high 2θ angle. AFM images show that the surface microstructure changed with the deposition temperature. XPS spectra show a clear broad P 2p peak at about 134 eV. The films exhibit good optical properties. Their optical band gaps calculated from the transmittance spectra are about 3.2 eV and the average optical transmittance in the wavelength range from 400 to 600 nm is more than 60%. Hall measurements indicate that all films show n type conductivity. Further, the carrier concentration decreases with the increase in the deposition temperature.
9:00 PM - M5.2
Comparison of ZnS-based Buffer Layers by Chemical Bath Deposition and Atomic Layer Deposition.
Alexander Uhl 1 , Charlotte Platzer-Bjoerkman 1 Show Abstract
1 Angstrom Solar Center, Uppsala University, Uppsala Sweden
In the search for a replacement of the CdS buffer layer in Cu(In,Ga)Se2 based solar cells, buffers based on ZnS is one of the main options. For this material, several deposition techniques have yielded efficient devices, for example chemical bath deposition, CBD, and atomic layer deposition, ALD. In both cases, the buffers contain large amounts of oxygen in some form. In the case of ALD, we have previously shown that the inclusion of sulfur into ZnO or inclusion of oxygen into ZnS leads to band gap narrowing and that suitable conduction band alignment with CuIn0.7Ga0.3Se2 can be obtained for a ZnO1-xSx layer with x ~ 0.7 close to the CIGS interface. In the case of CBD-ZnS, different compositions have been reported by different groups as for example Zn(S,OH), and Zn(S,O,OH) and it is more unclear how the oxygen and hydroxide content influences the junction. In this work we compare material properties of ZnS-based buffer layers by CBD and ALD using photoelectron spectroscopy (XPS), x-ray diffraction (XRD), scanning electron microscopy and optical measurements. The CBD films are deposited using mixtures of thiourea, zincsulfate and ammonia solutions. Bath conditions such as deposition time, concentrations, mixing order and preheating of chemicals are varied. The ALD films are deposited from diethylzinc, water and hydrogen sulfide precursors. For the ALD films the oxygen to sulfur content can easily be varied by changing the pulsing sequence. For the CBD films the control of film composition is more difficult. From the XPS study it is seen that one difference between the ALD and CBD films made in this work is the broader O 1s peak in the ALD films. From grazing incidence XRD, CBD films on CIGS are x-ray amorphous for the deposition conditions used for devices. We also compare devices made from the same CIGS runs with CBD and ALD buffers. Efficiencies of up to 12% and 15 % are obtained for the CBD and ALD buffered devices respectively, as compared to 15% for the CdS reference cells. The best CBD-ZnS device showed pronounced metastable behaviour with an increase in efficiency from 1 to 12% after annealing and light soaking. Devices made from the same CIGS run with ALD-Zn(O,S) were stable. In one series of devices, the CBD-ZnS layers were covered with ALD-ZnO or ALD-(Zn,Mg)O before ZnO:Al sputtering. These devices showed large improvement in efficiency and stability as compared to devices with ZnO sputtered layers directly onto the CBD-ZnS.
9:00 PM - M5.20
Influence of the Cu Content on Structural and Vibrational Properties in Polycrystalline CuGaSe2 Thin Films
Wolfram Witte 1 , Robert Kniese 1 , Michael Powalla 1 Show Abstract
1 , Zentrum für Sonnenenergie- und Wasserstoff-Forschung Baden-Württemberg, Suttgart Germany
CuGaSe2 (CGS) thin films are a candidate for application as the top cell in tandem devices with Cu(In,Ga)Se2 (CIGS) as the bottom cell. So far, the maximum achieved conversion efficiencies of CGS solar cells are too low for such an application , motivating basic research on the properties of this material. One interesting characterization method is the non-destructive technique of Raman spectroscopy, a promising tool for the in-line process monitoring of CIGS films in solar cell production.This contribution reports on the influence of the Cu content in CGS polycrystalline thin films on the Raman spectra. X-ray diffraction (XRD) measurements reveal a change in lattice constants with varying Cu content. This change in the lattice constants is proposed to explain the shift in the Raman spectra by its direct relation to the force constants.The polycrystalline CGS thin films were co-evaporated with constant deposition rates (single-layer process) on Mo-coated soda lime glass. The Cu/Ga ratio determined with the X-ray fluorescence method varied between 0.5 and 1.2, thus including very Cu-poor samples, near-stoichiometric, and Cu-rich absorbers. XRD spectra were recorded with CuKα radiation and lattice parameters were evaluated with the Pawley method. An analysis with the sin2Ψ method on the (112) reflex revealed low tensile stress for all measured Cu contents in the region of 20-60 MPa, with no dependence on the Cu content. Therefore, an influence of stress on the analysis of the lattice parameters and Raman frequencies can be neglected. The tetragonal lattice parameters a and c increase with higher values of the Cu/Ga ratio and the tetragonal distortion c/2a decreases significantly with the Cu content.Micro-Raman spectra measured with a frequency-doubled Nd:YAG laser in backscattering geometry were analyzed with Lorentzian fits. They revealed shifts of the chalcopyrite A1- and E-Modes at 273 cm-1 to lower frequencies with increasing Cu/Ga ratio. This behavior is clearly related to the increasing values of the lattice constants, which in turn influence the force constants responsible for the observed Raman modes. Furthermore, the FWHM of the A1- and E-Modes decreases with increasing Cu contents . D.L. Young et al., Prog. Photovolt.: Res. Appl. 11, (2003), 535-541 C. Xue et al., J. Phys. D: Appl. Phys. 37, (2004), 2267-2273
9:00 PM - M5.21
DX Centers in Chalcopyrite Compounds by Junction Capacitance Methods.
Malgorzata Igalson 1 , Adam Krysztopa 1 , Aleksander Urbaniak 1 Show Abstract
1 Faculty of Physics, Warsaw University of Technology, Warszawa Poland
First principles-calculations  show that antisites InCu should behave as typical DX centers with two configurations: substitutional donor and deep interstitial DX center. Complexes of antisites with Cu vacancies should possess similar characteristics. Important consequence of these properties for solar cells might be that that these defects impose open circuit voltage limitations in high band-gap chalcopyrites. Therefore exploring properties of these defects and confronting the findings with theory is of key importance for further development of CIGS technology. One theoretically predicted property of these defects is lack of a barrier for hole capture, thus in the DX state they are rather recombination centers than electron traps. Hence it is difficult to observe a DX state using capacitance methods, because presence of holes easily destroys it. Capacitance characterization of the cell after red illumination of the reverse biased device (ROB) is an experiment in which the DX states are revealed provided the position of the Fermi-level is close enough to the shallow/deep transition level. We will compare the results for high efficiency CIGS solar cells with the ones for less efficient high band-gap chalcopyrites. Apart of capacitance-voltage characterization photocapacitance measurements in the metastable states of the junction have been performed providing an insight into optical transitions between deep defect states and the bands.In the discussion we will include the results of current-voltage characterization from which the details of transport mechanisms typical for these junctions is obtained. The data for sulphur-based chalcopyrites provide evidences that open circuit voltage limitations indeed might come from the DX-type antisites. Their presence also explains why the mechanisms of current transport are different in the dark and under illumination. The absence of ROB effect in the CuGaSe2-based devices is an indication that not DX centers, but rather VSe-related defects are responsible for photovoltaic losses in these devices.  S. Lany, A. Zunger, Phys. Rev. Lett. 100, 016401 (2008)
9:00 PM - M5.22
Effect of Shunts on the Performance of Thin-Film Solar Cells and Modules
Galymzhan Koishiyev 1 , James Sites 1 Show Abstract
1 Physics, Colorado State University, Fort Collins, Colorado, United States
Deposition of semiconductor thin films often produces a granular polycrystalline structure with the main grain axis perpendicular to the film plane. The resulting grain boundaries may allow conductivity along their surfaces, and the penetration of the junction depletion layer by such grain boundaries can lead to shunting conductance. Localized shunting may also be caused by a bridging flaw during the deposition process of the layers, or it may be the result of window layer that is too thin or not continuous. The purposes of this work are to quantitatively evaluate the performance loss of both lab-sized cells and full modules due to local shunts and to help explain the physical origins of such shunts. A distributed transparent-conducting-oxide (TCO) sheet-resistance ρS (Ω/sqr) model of a thin-film solar cell with simple rectangular geometry was used to study the effect of localized shunts on cell performance. A reliable parameterization of shunts was used to define the screening length and the performance-loss area resulting from a localized shunt. Of key importance is how the impact of the shunt is mitigated by the isolation of the shunted area due to the TCO sheet resistance. At the small-cell level, the effect of an individual shunt can be substantial, but the performance decrease is only weakly dependent on the shunt location relative to cell boundaries and grid lines. Multiple shunts generate greater loss if they are distributed throughout the cell and less if they are clustered. At the module level, however, the impact of individual shunts, even major ones, is small, but the distribution of multiple shunts is very important. In particular, the existence of multiple shunts within a single cell of a module lowers performance much more than if the same shunts were distributed over all the cells.
9:00 PM - M5.23
X-ray Absorption Studies on Cadmium Selenide Thin Films.
Selma Erat 1 2 , Artur Braun 1 , Hulya Metin 3 , Iraida N. Demchenko 4 5 , Wayne C. Stolte 4 6 , Thomas Graule 1 7 Show Abstract
1 , Empa, - Swiss Federal Laboratories for Materials Testing & Research, Dubendorf Switzerland, 2 Department for Nonmetallic Inorganic Materials, ETH Zurich, Zurich Switzerland, 3 Department of Physics, Mersin University, Mersin Turkey, 4 Department of Chemistry, University of Nevada , Nevada, California, United States, 5 Institute of Physics, Polish Academy of Sciences, Warsaw Poland, 6 , Lawrence Berkeley National Laboratory, Berkeley, California, United States, 7 , Technische Universität Freiberg, Friberg Germany
CdSe is a prospective candidate for solar cells. High temperature annealing is typically a mandatory step for the processing of the films, but the annealing temperature can have a negative effect on the performance if it is too low or too high. We present here an annealing study of CdSe films with different annealing temperature. Our films were synthesized by chemical bath deposition on glass substrates at 337 K, followed by subsequent annealing in nitrogen atmosphere from 337 K to 773 K. The films show different spectral and conductivity characteristic depending on the annealing temperature. These films were subjected to x-ray absorption spectroscopy at the Cd L2 and L3 edges (3400 eV – 3600 eV) at beamline 9.3.1 at the Advanced Light Source in Berkeley, California. Since the films present a concentrated system with a high potential for self-absorption, we recorded the spectra in the total electron yield mode. Close inspection of the Cd L3 region of the spectra suggests a small chemical shift towards lower energies with increasing annealing temperature, suggesting a chemical reduction of the cadmium. Precise quantitative analysis of the pre-edge region shows actually a chemical shift of totally ~ 0.5 eV or slightly more in the annealing range from 337 K to 773 K, confirming that the cadmium is slightly reduced with increasing annealing temperature. Interestingly, the spectra of the films annealed at 100°C and 200°C having optical band gap at around 1.8 eV are completely different from those of the others. These two films don’t have any peak splitting although others have strong peak splitting at around 3565 eV. The pre-peak growing at around 3547 eV become dominant in the spectra of the films annealed at 400°C and 500°C.
9:00 PM - M5.24
Hydrogen Diffusion in Zinc Oxide Thin Films.
Wolfhard Beyer 1 3 , Uwe Breuer 2 , Juergen Huepkes 1 , Andrea Staerk 2 , Helmut Stiebig 3 , Uwe Zastrow 1 Show Abstract
1 IEF5-Photovoltaik, Forschungszentrum Jülich GmbH, Jülich Germany, 3 , Malibu GmbH & Co.KG, Bielefeld Germany, 2 Zentralabteilung für Chemische Analysen, Forschungszentrum Jülich GmbH, Jülich Germany
Thin films of zinc oxide (ZnO) are of interest as transparent conductive oxide layers for application in thin film solar cells. One ubiquitous impurity in such layers is hydrogen which is considered a shallow donor . A strong influence of hydrogen on conductivity of single crystalline ZnO was reported as early as 1954 . Various efforts were undertaken to measure H diffusion and to determine the H diffusion coefficient, both in ZnO crystals as well as in ZnO thin films. However, most experiments confined to the indiffusion of H  and not to the diffusion of H within a given sample. Here we report on a study of H diffusion using SIMS depth profiling of deuterium implantation profiles prior to and after annealing at various annealing temperatures and annealing time. We focus on undoped and aluminium-doped polycrystalline ZnO films prepared by sputtering. The microstructure of these films was characterized by effusion of hydrogen as well as of implanted helium and neon . Annealing resulted only for a fraction of the investigated samples in an error-function like spreading of the implantation profile, as expected for random walk diffusion of H atoms. According to the effusion measurements, these films were rather dense. For films with a high degree of void-related microstructure according to the gas effusion, annealing caused a decrease of the implanted deuterium concentration either without spreading of the profile or with plateau-like spreading at low concentration, attributed to the motion of hydrogen (deuterium) molecules through interconnected voids. For the dense material, the diffusion coefficient is found to follow an Arrhenius dependence with a diffusion prefactor of 1-100 sq cm/s and a diffusion energy of 1.8- 2 eV. This diffusion energy is considerably higher than other diffusion energies reported for this material . A correspondence with a high temperature hydrogen effusion maximum near 700°C is suggested. The H diffusion coefficients are found to be time-dependent similar as observed for a-Si:H.
9:00 PM - M5.25
In-situ High Temperature Structural Investigations of SnS, ZnS, Cu2SnS3 and Cu4SnS4 by Synchrotron Radiation.
Susan Schorr 1 , Alfons Weber 2 , Nora Schulze 3 , Veijo Honkimaki 4 , Hans-Werner Schock 3 Show Abstract
1 Geosciences, Free University Berlin, Berlin Germany, 2 Solar Energy Research, Helmholtz Zentrum Berlin fuer Materialien und Energie, Berlin Germany, 3 Institute of Mineralogy, Crystallography and Materials Science, University Leipzig, Leipzig Germany, 4 , European Synchrotron Radiation Facility, Grenoble France
Highly efficient thin film solar cells are based on compound semiconductors such as Cu(In,Ga)Se2 as absorber material. Since the availability of indium is an object of concern regarding the large scale production of solar cells, its replacement with Zn and Sn is beneficial in this sense. Recently thin film solar cells using the quaternary compound Cu2ZnSnS4 (kesterite) as absorber material reaching efficiencies up to 6.7% were reported in literature . Multi-stage evaporation is a well-established method for the controlled growth of thin films. This technique can also be applied to the deposition of Cu2ZnSnS4 thin films using Cu2SnS3 as precursor to react with Zn-S or using ZnS as precursor to react with Cu-Sn-S . Other possible compounds maybe acting as precursors are Cu4SnS4 and SnS.To understand the growth process of kesterite in this multi-stage evaporation process, a detailed knowledge about the structure of the precursor compounds is indispensable. Of special interest are structural changes with temperature in the range from room temperature up to 700°C. The present work discusses the temperature dependent lattice parameter and linear thermal expansion coefficients of SnS, ZnS, Cu2SnS3 and Cu4SnS4 gained from in-situ powder diffraction experiments using synchrotron X-rays.Experiments were performed at the high energy beamline ID15B at the ESRF synchrotron radiation facility in Grenoble, France. The setup consisted of a ceramic oven with two small holes (incident and scattered beam apertures) and an on-line 2D detector, MAR 345 image plate. The samples, placed in evacuated silica tube (diameter 4 mm) to avoid sulfur losses, were heated with a rate of 180 K/h up to 700°C. Two-dimensional diffraction pattern were collected in transmission geometry in 5K steps. An aluminum reference sample was used to calibrate the beam energy (λ=1.42412 nm) and the sample-detector distance. The structural parameters of the different compounds were determined by Rietveld analysis of the data.The presentation will give an overview of the temperature dependent changes of the lattice parameter, the linear thermal expansion coefficients as well as structural phase transitions in SnS, ZnS, Cu2SnS3 and Cu4SnS4. The results will be discussed with respect to structural relations between the precursor compounds and Cu2ZnSnS4 (kesterite).  H. Katagiri et al., Appl. Phys. Express 1 (2008) 041201. A. Weber, H. Krauth, S. Perlt, B. Schubert, I. Kötschau, S. Schorr, H.W. Schock, Thin Sol. Films (2008) accepted.
9:00 PM - M5.26
Enhanced Conversion Efficiency of GaAs Photovoltaics Utilizing Anti-Reflective Indium-Tin-Oxide Nano-Columns
Chia-Hua Chang 1 , Peichen Yu 1 , Min-Hsiang Hsu 2 , Ching-Hua Chiu 1 , Hao-Chung Kuo 1 Show Abstract
1 , National Chiao-Tung University, Hsinchu Taiwan, 2 , National Tsing-Hua University , Hsinchu Taiwan
Global warming issues coupled with high oil prices have become a major driving force for the use of advanced solar power, where a key component lies in the development of high efficiency and low cost photovoltaic cells. Next-generation photovoltaics hence demand an efficiency-boosting mechanism in order to make solar energy cost-competitive with conventional sources of electricity. Fundamentally, the conversion efficiency of a photovoltaic cell depends on the photon absorption, carrier separation, and carrier collection. Therefore, an effective anti-reflection (AR) coating, minimized recombination loss, and good ohmic contacts are particularly important. Metal grids that inevitably block the transmission of solar energy also require optimization in order to reduce the series resistance. The tradeoff between the electrode and the AR coating areas is one of the efficiency-limiting factors in a conventional solar cell.The conventional AR coating is realized by a quarter-wavelength stack of dielectrics with different refractive indices. Broad angular and spectral anti-reflection is achievable at the price of multiple layers. Over the past few years, versatile sub-wavelength structures (SWS) have emerged as promising candidates for AR coatings due to the characteristics of zero-order gratings, or the so-called moth-eye effects. However, the fabrication costs that involve either electron-beam (e-beam) lithography can be significant. The resulting surface recombination loss due to dry or wet etching could further hinder the applications of SWS in commercial solar cells. Recently, multiple studies have been made on indium-tin-oxide (ITO), titanium dioxide (TiO2), and silicon dioxide (SiO2) nanostructures employing oblique-angle deposition methods, where the refractive indices of the nano-porous materials can be engineered by adjusting the air volume ratio. Still, the materials require multiple layers to effectively suppress the Fresnel reflection. In this paper, we demonstrate a practical photovoltaic application of ITO nano-columns serving as a conductive AR layer for GaAs solar cells. As in GaAs cells, using a nanostructured AR layer could be otherwise limited due to severe front surface recombination. The characteristic ITO nano-columns, prepared by glancing-angle deposition with an incident nitrogen flux, offer omnidirectional and broadband AR properties for both s- and p-polarizations, up to an incidence angle of 70° for the 300 nm-900 nm wavelength range. Calculations based on a rigorous coupled-wave analysis (RCWA) method indicate that the superior AR characteristics arise from the tapered column profiles which collectively function as a graded-refractive-index layer. The conversion efficiency of the GaAs solar cell with the nano-column AR layer increases by 28% compared to a cell without any AR treatment. Moreover, nearly 42% enhancement is achieved for photocurrents generated at wavelengths that are transparent to the window layer.
9:00 PM - M5.27
Phase Evolution and Structural Properties of CuInO2 thin Films grown by Pulsed Laser Deposition
Jong-Chul Lee 1 , Young-Woo Heo 1 , Jeong-Joo Kim 1 , Joon-Hyung Lee 1 Show Abstract
1 School of Materials Science and Engineering, Kyungpook National University, Daegu Korea (the Republic of)
CuInO2 delafossite thin films are reported to exhibit both n-type and p-type conduction by applying doping of an appropriate dopant. This bipolarity characteristic of CuInO2 is useful for homostructural p-n junction. However, the CuInO2 phase is not obtainable by using the general mixed oxide synthesis route while Cu2In2O5 phase is easily obtained. Therefore, a previous attempt reported that the Cu2In2O5 target was used in order to obtain CuInO2 thin films through a control of deposition variables such as oxygen partial pressures. In this study, we are report phase evolution and structural properties of CuInO2 thin films grown using two phase targets of Cu2O-In2O3 by PLD. The Cu2O-two phase targets of Cu2O-In2O3 which were composed of Cu2O and In2O3 with various molar ratio were fabricated by using the general mixed oxide synthesis route. Furthermore, two phase targets of Cu2O-In2O3 doped 1mol% Ca, Mg, Ti, Sn substitute for In were prepared respectively. The phase evolution of the CuInO2 thin films deposited on sapphire substrate was examined by X-ray diffraction when the oxygen partial pressure of the ambient was varied. A four-circle X-ray diffraction was also used to evaluate the crystallinity of the films. Hall-effect measurements with the van der Pauw technique were conducted to examine electrical properties and optical properties were also analyzed.
9:00 PM - M5.28
Investigations Optical and Electrical Properties of Sb and Zn co-doped In2O3 Thin Films grown by DC Magnetron Sputtering Method
Don-Hyeong Kim 1 , Young-Woo Heo 1 , Joon-Hyung Lee 1 , In-Chul Park 1 , Jeong-Joo Kim 1 Show Abstract
1 , Kyungpook National University, Daegu Korea (the Republic of)
Tin doped indium oxide(ITO), in which 6~8at% Sn forms a substityional solution in In2O3, is the most common and widely used transparent conducting oxide. However, because of an imbalance between the supply and demand of indium tin oxide, the development of indium-free or indium saving transparent conducting oxides has received more attention. In this study, we investigated optical and electrical properties of Sb and Zn co-doped In2O3 thin films grown by DC magnetron sputtering method. Sb and Zn co-doped In2O3 target was fabricated by using the general mixed oxide synthesis route. Intermediate ZnSb2O6 phase was synthesized to retard the evaporation of Sb2O5 whose melting point is very low during sintering. Furthermore, Sb and Zn co-doped In2O3 films were deposited by DC magnetron sputtering method. The phase evolution of the Sb and Zn co-doped In2O3 thin films deposited on glass(Corning Eagle2000) substrate was examined by X-ray diffraction when the oxygen and argon partial pressure of the ambient were varied. Hall-effect measurements with the van der Pauw technique were conducted to examine electrical properties and optical properties were also analyzed.
9:00 PM - M5.29
Preparation of Cu(In,Ga)Se2 Thin Films on ZnO Nanorod.
Kyun Ahn 1 , Ye Sul Jeong 1 , Je Hoon Jeon 1 , Se Young Jeung 1 , Jong Pil Kim 2 , Chae Ryong Cho 1 Show Abstract
1 nanoscience and nanotechnology, pusan national university, Busan Korea (the Republic of), 2 , Busan center, Korea Basic Science Institute, Busan Korea (the Republic of)
In this study, ZnO nanorod was used as buffer electrode to improve efficiency of Cu(In,Ga)Se2 (CIGS) thin film solar cell. We present the interface characteristics between CIGS film and ZnO nanorod during deposition at high temperature (~ 550°C). ZnO nanorod was grown on various substrates such as GaN, Al2O3 and ZnO:Al (AZO)/glass. CIGS thin film was prepared by using three-stage method of Co evaporator. As window layer, ZnO nanorod provides a direct conduction path for electrons between the CIGS light absorbing layer and transparent conducting AZO layer and may offer improved electron transport compared to ZnO of film-type. The surface morphology and structure were measured by field emission scanning electron microscope (FE-SEM) and x-ray diffraction (XRD). From the analysis results of FE-SEM and XRD, the deposited CIGS film showed a smooth surface morphology and contained a single chalcopyrite phase. Depth profiles and compositional variation of the deposited CIGS and interface properties of between CIGS and ZnO nanorod were investigated by secondary ion mass spectroscopy (SIMS) and glow discharge spectrometer (GDS). The composition ratio of Cu / (In + Ga) and Ga / (In + Ga) was 0.89 and 0.22, respectably. Using 4-prove station, electrical property was measured. The optical behaviors of CIGS and ZnO nanorod according to the deposition condition were measured by using UV-VIS-NIR spectrometer.
9:00 PM - M5.3
Solvothermal Preparation, Processing, and Characterization of Nanocrystalline CuIn1-xAlxSe2 Materials.
Christopher Exstrom 1 , Jiri Olejnicek 1 , Scott Darveau 1 , Anatole Mirasano 1 , David Paprocki 1 , Megan Schliefert 1 , Matt Ingersoll 1 , Laura Slaymaker 1 , Rodney Soukup 2 , Natale Ianno 2 , Chad Kamler 2 Show Abstract
1 Department of Chemistry, University of Nebraska at Kearney, Kearney, Nebraska, United States, 2 Department of Electrical Engineering, University of Nebraska-Lincoln, Lincoln, Nebraska, United States
Nanocrystalline chalcopyrite materials of the CuInSe2 (CIS) family are of interest for the development of solar cell absorber layers. Partial substitution of In3+ with smaller Group 13 ions has long been a standard route to increasing the CIS bandgap of 1.04 eV toward the optimum 1.37 eV. Because of its small ionic size, only 25% substitution of Al3+ for In3+ would be required to reach this bandgap. We report the preparation and characterization of the first nanocrystalline CuIn1-xAlxSe2 (CIAS) materials. The reaction of Se, CuCl, InCl3, and Al(oleate)3 in refluxing oleylamine for 30-40 minutes yields nanocrystalline CIAS. Scanning electron microscopy (SEM) images reveal morphologies consisting of hexagonal plates (100-400 nm diameter) with smaller isomorphic nodules (10-20 nm diameter). Micro-Raman spectroscopy (174-177 cm-1 A1 phonon frequencies), x-ray diffraction (XRD, d(112) 2θ = 26.93o), and optical bandgap data are consistent with Al3+ incorporation into the chalcopyrite structure. From comparison to literature XRD/composition correlations for single-crystal CIAS, the Al/(In+Al) atom ratios in our nanocrystalline CIAS are estimated at 0.12 and 0.19 from solvothermal reactions containing 0.25 and 0.50 Al/(In+Al) mole ratios, respectively. After annealing the nanocrystalline CIAS samples at 500 oC under vacuum, N2, or Ar, two significant effects are observed. SEM images reveal smoother, film-like morphologies while phase separation of Al is apparent from Raman and XRD data. Over a two-hour annealing period, up to 37% Al loss has been observed, most of this occurring within 20 minutes. Correlations of A1 phonon frequencies and XRD signal positions to Al/(In+Al) ratios in nanocrystalline CIAS samples will be compared to those in selenized CuIn1-xAlx thin-film precursors.
9:00 PM - M5.31
Fabrication of CuxSe Thin Films by Selenization of CuxSe Nanoparticles Prepared by a Low Temperature Colloidal Process.
SeJin Ahn 1 , Jae Ho Yun 1 , Kyung Hoon Yoon 1 Show Abstract
1 , Korea Institute of Energy Research, DaeJeon Korea (the Republic of)
CuSe is a well known material in growth process of CIGS thin films in that its melting point is as low as 523 oC and hence it causes a liquid assisted sintering, resulting in CIGS thin films of large grains. In this regard, we synthesized CuxSe nanoparticles to use them as flux of liquid assisted sintering and source of Cu, simultaneously, for fabrication of high quality CIGS thin films by a non-vacuum process. A simple low temperature colloidal process was involved to prepare CuxSe nanoparticles, where CuI, InI3 and GaI3 in pyridine were reacted with Na2Se in methanol at 0oC under inert atmosphere, finally resulting in formation of very fine (20nm in diameter) and uniform CuxSe nanoparticles. Further, the densification of behavior of the corresponding nanoparticle coated thin films by selenization was also investigated to confirm the liquid assisted sintering by CuSe (CuxSe(s) + Se(s) = CuSe(l) at 523 oC). For comparison, the CuxSe powders of a few μm in diameter were also selenized to investigate the effects of powder size (one is in nm scale, the other μm scale) on the densification behavior. It was found that nanoparticle-derived films were successfully densified with selenization process, while the films of large powders were not effectively densified, reflecting that the high surface energy of the nanoparticles were essential for growth of large grained thin film.
9:00 PM - M5.32
Optimization of Process Parameters for Fabrication of Flexible CIGS Thin Film Solar Cells with a Na-doped Mo Back Contact.
Jae Ho Yun 1 , SeJin Ahn 1 , Kyung Hoon Yoon 1 Show Abstract
1 , Korea Institute of Energy Research, DaeJeon Korea (the Republic of)
Flexible CIGS thin film solar cells were fabricated using a Na-doped Mo as a bottom layer of the Mo back contact on a Na-free stainless steel substrate, where Na-doped Mo layer acts as a source of Na diffusion. In this approach, Na was supplied to the CIGS bulk region from the stainless steel/SiO2/Na-doped Mo/Mo structure, and the amount of diffused Na was controlled by adjusting the thickness of Na-doped Mo layer. Process optimization results, mainly focusing on the effects of thickness of Na-doped Mo layer on the performance of solar cells, are to be reported in this presentation which includes the highest conversion efficiency of 12.38% (Jsc=36.0 mA/cm2, Voc=0.54 V and FF=64%). For a comparison, performance of solar cell without Na doping will be presented together.
9:00 PM - M5.33
Wafer Bonding and Ion-slicing as a Strategy for Breaking the Trade-off Between Quality and Cost in the Fabrication of Compound Semiconductor Multi-junction Solar Cells.
Oussama Moutanabbir 1 , Ulrich Goesele 1 Show Abstract
1 , Max-Planck Institute of Microstructure Physics, Halle (Saale) Germany
III-V semiconductors grown on silicon substrates are very attractive for lower-cost, high-efficiency multijunction solar cells, but lattice-mismatched alloys that result in high dislocation densities have been unable to achieve satisfactory performance. Alternatively, high performance can be achieved by using bulk wafers. However, the cost of these wafers presents a serious handicap for large scale production and industrial realization of III-V based triple- or four-junction solar cells. In this paper, we suggest a method capable of bridging the gap between the quality and cost of III-V multijunction photovoltaic cells. Indeed, by using direct wafer bonding in combination with hydrogen ion-cutting it is possible to integrate bulk quality thin layers onto various host materials achieving a wide variety of heterostructures frequently unattainable by epitaxy. The thickness of the layer to be transferred can be adjusted by tuning the energy of hydrogen ions. In this presentation, we will address the fundamentals of wafer bonding of dissimilar materials. The critical issue in this process has to do with the thermally induced stress which becomes very important when the bonded materials have different thermal expansion coefficients. This leads usually to debonding during thermal annealing required to strength the bonded interface and to activate the splitting process. We will describe a novel method to circumvent this interfacial stress problem. We will demonstrate how epitaxy-compatible 4-inch InP-on-Si substrates can be achieved. We will also address different issues involved in ion slicing of 2-inch high quality freestanding GaN. Besides improving the present compound semiconductor based devices, the versatility of the layer transfer process provides a playground to develop new concepts of solar cell devices.
9:00 PM - M5.35
Hydrogen and Nitrogen Incorporation in ZnO Thin Films Grown by Radio Frequency (RF) Sputtering and Chemical Vapor Deposition (CVD).
Bruno Meyer 1 , Achim Kronenberger 1 , Angelika Polity 1 , Sebastian Eisermann 1 , Stefan Lautenschlaeger 1 Show Abstract
1 1. Physics Institute, Justus Liebig University, Giessen Germany
Bipolar conduction of electron and holes is mandatory for many electronic and opto-electronic applications of ZnO and its ternary alloys. Hydrogen in ZnO plays an unusual role since it acts as a shallow donor and may control the n-type conductivity in nominally undoped material. But it can also be used as an n-type dopant - alone or in combination with the group-III-elements Al, Ga or In. We will demonstrate this behaviour in ZnO and ZnOS films prepared by RF magnetron sputtering. The incorporation of hydrogen as a function of the substrate temperature as well as its thermal stability were investigated. Among the impurity atoms suited for acceptor formation nitrogen is the prime candidate. Controlling the incorporation without the formation of deep donors and acceptors and avoiding the deterioration of the materials quality is main goal for homo- and heteroepitaxially CVD grown ZnO films (ZnO/ZnO and ZnO/GaN). We will show by SIMS and Raman spectroscopy how the surface polarity influences the nitrogen desorption. Temperature modulated growth leads to nitrogen incorporation into ZnO with substantially improved luminescence properties.
9:00 PM - M5.36
Interfacial Properties of Pulsed Laser Deposited CuIn0.7Ga0.3Se2 Thin Films on Cu Foils
Yeon Hwa Jo 1 , Bhaskar C. Mohanty 1 , Deuk Ho Yeon 1 , Ik Jin Choi 1 , Yong Soo Cho 1 Show Abstract
1 Materials science and engineering, Yonsei University , Seoul Korea (the Republic of)
One of the promising candidates for low-cost, high-efficiency photovoltaic cells is the CuIn1-xGaxSe2 (CIGS) absorber thin film. Recently, we have demonstrated that good quality thin films of CIGS using pulsed laser deposition (PLD) shows an excellent performance with respect to the compositional reproduction of compound semiconductors and high deposition rates. In this study, thin films of chalcopyrite compound CIGS have been prepared on Cu foils by PLD using a one inch CuIn0.7Ga0.3Se2 target. Besides, serving both as back contact and substrate, Cu foils can pave way for fabrication of flexible solar cells. An excimer laser (KrF; wavelength of 248 nm; pulse width 20 nm; repetition rate of 10 Hz) was focused onto the surface of the target. The temperature of the substrate, placed vertically 5 cm above the target, was varied from room temperature to 500 oC. Analysis of the X-ray diffraction patterns yielded that the films were polycrystalline with a tendency of (112) preferable orientation and exhibited enhanced crystallinity with increasing substrate temperature. Characteristic of interface were analyzed by Auger spectroscopy, which showed excellent distribution of Ga throughout the film. Se profile showed a gradual decrease at the interface leading to formation of a Cu-rich region.
9:00 PM - M5.37
Spatial Variation of Optical and Electrical Properties of ZnO:Al Thin Films Grown by RF Magnetron Sputtering.
Bhaskar Chandra Mohanty 1 , Yeon Hwa Jo 1 , Deuk Ho Yeon 1 , Ik Jin Choi 1 , Yong Soo Cho 1 Show Abstract
1 Materials science and engineering, Yonsei University , Seoul Korea (the Republic of)
ZnO:Al (AZO) thin films with high visible transmittance and good electrical conductivity have been used as window layer in CuInSe2-based solar cells with demonstrated success. A change in these properties of AZO films can affect the overall performance of the solar cells. In the present work, spatial variation of film properties, especially of the optical and electrical properties, as a result of inhomogeneous erosion of target as well as off-normal substrate placement during sputtering is studied. The thin films were grown by 30o-incident magnetron sputtering of a ZnO:Al (2 wt %) ceramic target at radio-frequency powers of 100, 150 and 200 W in pure Ar ambient with pressures in the range of 1-10 mtorr. Properties across a distance of 60 mm on the samples are investigated using optical transmission, Hall measurements and scanning electron microscopy. It was observed that the average transmittance of the central portions of the films, sputtered at the same power, decreased with increasing Ar pressure. However, it remained uniform spatially. The optical band gap of the films (central portion) sputtered at 200 W decreased monotonically from 3.42 eV to 3.29 eV for increase in Ar pressure from 1 to 10 mtorr, which can affect the amount of solar radiation reaching the absorber layer in the solar cell. For all sputtering powers, the difference in band gap of the central and edge portions of the samples was maximum for the films grown at lowest pressure. The correlation between microstructure, carrier concentration and band gap is brought out.
9:00 PM - M5.38
Texture Control of Fluorine Doped Tin Oxide Coating by Spray Pyrolysis Method
Chang-Yeoul Kim 1 , Cheol-Kyu Song 2 , Doh-Hyung Riu 1 , Seung-Hun Huh 1 , Kwang-Yeon Cho 1 Show Abstract
1 Nanomaterials Team, Korea Institute of Ceramic Eng. & Tech, Seoul Korea (the Republic of), 2 , Solar Ceramic Co., Seoul Korea (the Republic of)
Transparent conducting oxide (TCO) is widely used for the application of flat panel display like liquid crystal displays and plasma display panel. 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 (450oC) 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 ×10-4 Ω-cm, hall mobility of 4.5 cm2/Vs, and carrier oncentration of 34.5 ×1020cm-3. Microstructures, electrical resistivies and optical properties of F-doped tin oxide films were investigated by changing F/Sn ratio and thicknesses.
9:00 PM - M5.4
Comparison of Elemental Distribution Profiles in Cu(In,Ga)Se2 Acquired by Various Techniques.
Daniel Abou-Ras 1 , Christian Kaufmann 1 , Iver Lauermann 1 , Andreas Schoepke 1 , Christiane Stephan 1 , Susan Schorr 2 , Axel Eicke 3 , Max Doebeli 4 , Beate Gade 5 , Joachim Hinrichs 6 , Tim Nunney 7 , Volker Hoffmann 8 , Denis Klemm 8 , Varvara Efimova 8 , Andreas Bergmaier 9 , Guenther Dollinger 9 Show Abstract
1 , Helmholtz Center Berlin, Berlin Germany, 2 , Free University Berlin, Berlin Germany, 3 , ZSW, Stuttgart Germany, 4 , PSI/ETH, Zurich Switzerland, 5 , Thermo Fisher Scientific, Dreieich Germany, 6 , Thermo Fisher Scientific, Bremen Germany, 7 , Thermo Fisher Scientific, East Grinstead United Kingdom, 8 , IFW Dresden, Dresden Germany, 9 , Universität der Bundeswehr München, Neubiberg Germany
Thin-film solar cells with Cu(In,Ga)Se2 absorbers generally consist of a ZnO/CdS/Cu(In,Ga)Se2/Mo thin-film stack on a glass substrate. The Cu(In,Ga)Se2 absorber layers do not exhibit homogeneous compositions but gradients of the Ga and the In concentrations across the layer thickness. The analysis of these gradients is important in order to gain information on the optoelectronic properties of the Cu(In,Ga)Se2 thin films since the band-gap energy varies. between 1.0 eV (CuInSe2) and 1.7 eV (CuGaSe2)The aim of the present study is to compare various techniques which can be applied in order to determine the elemental distribution of the Cu(In,Ga)Se2 absorber layers. Measurements by energy-dispersive X-ray spectrometry (EDX) in a scanning (SEM-EDX) and in a transmission electron microscope (TEM-EDX), X-ray photoelectron (XPS), X-ray emission (XES), secondary ion-mass (SIMS) and sputtered neutral mass (SNMS), glow-discharge optical emission (GD-OES) and glow-discharge mass (GD-MS), Auger electron (AES), and also Rutherford backscattering spectrometry (RBS), as well as elastic recoil detection analysis (ERDA) and grazing-incidence X-ray diffraction (GIXRD), are performed on nominally identical Cu(In,Ga)Se2 absorber layers from the same production run. Most of these methods reproduce quantitatively similar Ga and In elemental distribution profiles across the Cu(In,Ga)Se2 thin films. The techniques are compared in terms of their availability, their spatial and depth resolutions, their measurement speed, and their accuracy of the quantitative results.
9:00 PM - M5.40
Cu-In and Cu-Zn-Sn Films as Precursors for Production of CuInSe2 and Cu2ZnSnSe4 Thin Films.
Olga Volobujeva 1 , Enn Mellikov 1 , Jaan Raudoja 1 , Maarja Grossberg 1 , Sergei Bereznev 1 , Rainer Traksmaa 2 Show Abstract
1 Department of Materials Science, Tallinn University of Technology, Tallinn Estonia, 2 Center of Materials Research, Tallinn University of Technology, Tallinn Estonia
Thin film solar cells based on different chalcopyrite and stannite absorber layers are perspective candidates for large scale solar energetic. The material quality of the polycrystalline absorber layers has critical influence to the parameters of solar cells, but is at the same time determined by the surface morphology, composition and structure of Cu-In and Cu-Zn-Sn that are used as precursors for selenization process. In this study, the evolution of surface morphology, phase composition and homogeneity of precursor Cu-In and Cu-Zn-Sn films is studied using high resolution SEM with EDS, micro-Raman and XRD techniques.Cu-In precursor layers were produced on Mo substrate at room temperature by magnetron co-sputtering of Cu-In target (Cu/In=0.8). Cu-Zn-Sn precursor layers with different ratio of Cu/(Zn-Sn) and Zn/Sn were produced on Mo substrate at temperature 150C by sequential evaporation of consistent metals.The co-sputtered Cu-In precursor layers were characterized by bi-layer surface structure in which island - type crystals were formed in a small-crystalline matrix layer. The elemental composition of the island - type crystals corresponds to the compound CuIn2 and the matrix (area) consists of copper-rich Cu11In9 phase.It was shown that surface morphology of sequentially evaporated Cu-Zn-Sn precursor layers is determined by deposition order of stacked consistent metal layers. Precursor Mo-Sn-Zn-Cu films exhibit well formed “mesa-like” structure of the surface in which larger crystals (about 1 µm) are located on a “small-crystalline” valley. The analysis of SEM and EDS data allows to conclude that these “mesa – forming” crystals originate from the deposited Sn layer. During deposition Sn forms a discontinuous layer of semispheral crystals onto the Mo substrate, leading to the “mesa-like” structure of the Sn layer. The following depositions of Zn and Cu lead to the formation layers of these metals quite uniform in thickness on the whole surface of samples that result in a “mesa-like” structure of the surface of the final Mo-Sn-Zn-Cu films. For films with other sequences of metallic layers, the mesa like structure is not so well exposed, and well formed flat precursor layers were produced replacing separate metallic Cu and Sn layers with Cu/Sn alloy layer. The analysis of XRD patterns gives evidence of the existence of separate alloy phases of Cu5Zn8 and Cu6Sn5 in the precursor films. The phase composition and structure of Cu2ZnSnSe4 films produced by the selenization of precursors in Se atmosphere in dependence of sequence of consistent layers in Cu-Zn-Sn precursor layers will discussed.
9:00 PM - M5.5
Large Area Chemical Bath Deposition of CdS on Cu(InGa)Se2.
Nirav Vora 2 1 , Ingrid Repins 1 , Steve Robbins 1 Show Abstract
2 Chemical and Biomolecular Engineering , Vanderbilt University, Nashville, Tennessee, United States, 1 , National Renewable Energy Labratory, Golden, Colorado, United States
Chemical bath deposition (CBD) is known to be a very efficient and cost-effective way of depositing uniform cadmium sulfide (CdS) films for photovoltaic application. The method is based on decomposition of a sulfur source in an alkaline solution of a cadmium source on the surface of the Cu(InGa)Se2 substrate. On the lab scale the CdS film is deposited by submerging a 1” square Cu(InGa)Se2 substrate in a heated beaker containing the chemical bath. This batch processing method provides some inherent advantages with subsequent improved device performance. The Cu(InGa)Se2 substrate is subjected to chemical etching/cleaning by ammonia in chemical bath thus resulting into for epitaxial growth of CdS on Cu(InGa)Se2. It also has been proposed that the chemical bath aids in doping of Cu(InGa)Se2 with Cd leading to formation of homojunction. There is an ongoing effort at the National Renewable Energy Laboratory to scale-up the CBD process to deposit CdS films on 6” square substrate. Apart from the above-mentioned interfacial effects, the scale-up process will have to consider factors like uniform flow and concentrations over the substrate, minimization of CdS particle precipitation in the chemical bath, and reduction of Cd containing waste. A batch reactor was designed to reproduce the deposition process in the beaker on a bigger scale with minimal chemical waste. Since a batch reactor is difficult to integrate into a manufacturing line, an attempt is made to design the second prototype as a continuous flow reactor. The films deposited using these prototype reactors will be characterized to check for uniformity. The uniform films will be made into devices to check for efficiencies. This abstract is subject to government rights.
9:00 PM - M5.6
N-Type Doping in Chalcogenides with Halogen by Co-Precipitation
Xiaofei Han 1 , Kunhee Han 2 , Meng Tao 2 Show Abstract
1 Materials Science and Engineering, University of Texas at Arlington, Arlington, Texas, United States, 2 Electrical Engineering, University of Texas at Arlington, Arlington, Texas, United States
In the microelectronics industry, doping is accomplished by either diffusion, ion implantation or co-deposition of a dopant with Si during vapor deposition. Solution-based doping techniques are desirable for low-cost, large-area and high-throughput fabrication of solar cells. In this talk, we will present an electrochemical method to dope Cu2O n-type with Cl. The principle of the doping method is believed to be co-precipitation of Cu2O with CuCl in an aqueous solution and Cl substitution of O in Cu2O to become an n-type dopant. Cu2O is naturally p-type. Photocurrent measurements reveals that Cl-doped Cu2O is n-type. Current-voltage characterization confirms that Cl doping reduces the resistivity of electrochemically-deposited Cu2O from 40 MΩ-cm to as low as 7 Ω-cm, which is close to the optimum resistivity for solar cells, ~1 Ω-cm. The significance of this work is multifold. First, Cu2O is a promising material for next-generation solar cells with its low energy input, low cost, non-toxicity and abundant source materials. The n-type doping method is expected to significantly improve the efficiency of Cu2O solar cells. More importantly, since this doping method substitutes chalcogen with halogen by co-precipitation, it is in principle universal for n-type doping in other solution-prepared chalcogenides for current and future solar cells.
9:00 PM - M5.7
Electronic Structure and Phase Stability of Cu2ZnSnSe4 by First-principles Calculations.
Satoshi Nakamura 1 , Tsuyoshi Maeda 1 , Takahiro Wada 1 Show Abstract
1 Department of Materials Chemistry, Ryukoku University, Otsu , Shiga, Japan
Chalcopyrite-type compounds such as CuInSe2 (CIS) have received attention as one of the most promising materials for thin film solar cells. Recently, the substitution of Indium has become an important issue because it is scarce and expensive. Cu2ZnSnSe4 (CZTSe), one Indium free absorber material, is a I2-II-IV-VI4 semiconductor with a stannite-type structure. The fabrication and characterization of CZTSe solar cells have been investigated . CZTSe has a direct band gap of 1.5 eV, which is suitable for a solar light absorber. The electronic structures of CIS and the related chalcopyrite compounds have been studied by first-principles calculations [2, 3]. Recently, we reported the electronic structure of stannite-type CZTSe at ICTMC-16 . The valence band maximum (VBM) consists of the antibonding orbitals of Cu 3d and Se 4p, and the conduction band minimum (CBM) consists of the antibonding orbitals of Sn 5s and Se 4p.Three kinds of crystal structures have been reported for Cu2-Zn-IV-VI4 compounds. Cu2ZnSnS4 has a kesterite-type, Cu2ZnSnSe4 has a stannite-type, and Cu2ZnGeSe4 and Cu2ZnSiSe4 have wurtz-stannite-type structures. In this research, we studied the phase stability and electronic structure of kesterite-, stannite-, and wurtz-stannite-type of CZTSe.We performed first-principles calculations within a density functional theory using a plane-wave pseudopotential method. Structural optimizations and electronic structure calculations were performed using the primitive cells of kesterite-type unit cell I-4, stannite-type unit cell I-42m, and wurtz-stannite-type unit cell Pmn21.To evaluate the phase stability of the kesterite-, stannite-, and wurtz-stannite-types of CZTSe, the formation enthalpies (ΔH) of these phases were calculated. The ΔH of the kesterite phase (-312.66 kJ/mol) is a little smaller than that of the stannite phase (-311.34 kJ/mol) and much smaller than that of the wurtz-stannite phase (-305.69 kJ/mol). This result indicates that the kesterite and the stannite phases are more stable than the wurtz-stannite phase and that the stability of the kesterite and the stannite phases is equivalent. The theoretical band gap of the stannite phase is 0.000017eV, which is a little smaller than that of the kesterite phase of 0.047eV and that of the wurtz-stannite phase of 0.062 eV. However, the band gap energies calculated with the generalized gradient approximation (GGA) functional are greatly underestimated. In the previous study, we reported that the screened exchange-LDA (sX-LDA) calculation greatly improved the band gap of CIS . Now we are calculating the band gap energies of CZTSe with the sX-LDA method. R. A. Wibowo et al, phys. stat. sol. (a) 204, 3373 (2007). J.E. Jaffe and Alex Zunger, Phys. Rev. B 28 5822 (1983). T. Maeda, T. Takeichi, and T. Wada, phys. stat. sol. (a) 203, 2634(2006). S. Nakamura, T. Maeda, and T. Wada, phys. stat. sol., submitted. T. Maeda and T. Wada, J. Appl. Phys., submitted.
9:00 PM - M5.8
Titanium Incorporation to In2S3 Thin Films for Photovoltaic Applications.
Begona Asenjo 1 , Jose Herrero 1 , Teresa Gutierrez 1 Show Abstract
1 Renewable Energy, CIEMAT, Madrid Spain
Currently diverse theoretical studies have been done focused on materials to be used for intermediate band in photovoltaic cells. The transition metal impurities have been known to introduce deep levels in the band gap of semiconductors.The transition metal impurities have been known to introduce deep levels in band gap of semiconductors. Titanium was soon known to enter the host lattice substitutionally on a cation, this element could form an intermediate band (IB) between valence band (VB) to the conduction band (CB) with the subsequent production of electric current of chemical reactions.In this work we report preliminary studies on the properties of undoped and Ti-doped In2S3 thin films grown on soda lime glass by chemical solution deposition under different conditions. Different physical, chemical and morphological properties of the films have been analysed. At the beginning of the deposition of In2S3 films, In2O3 and In(OH)3 deposited by electroless-chemical reaction are dominant. However, when adding the titanium, the formation of Titanium sulfide is predominant on short time deposition. The optical properties observed for Ti-In2S3 films show the partial contribution of the electronic transitions: two increases in absorption starting at about 0.7 eV, resulting from transitions from the VB to the IB, and at 1.4 eV resulting from transitions from the occupied IB states to the CB. There is additionally an increase at 2.2 eV resulting from the usual transition from the VB to the CB. The study is completed with SEM analysis which shows the influence of the deposition time in the morphology for incorporation of titanium at the beginning of deposition. The film is composed of nanometric round shape grains. Films deposited at longer time are composed by bigger agglomerates and X-ray diffraction of the films shows the amorphous nature of the films. The crystallinity is enhanced by using longer deposition times and/or after thermal treatment at 350 °C in N2 atmosphere.
9:00 PM - M5.9
Growth of Cu(In,Al)Se2 Thin Films by Selenization Using Diethylselenide.
Chika Fujiwara 1 , Akihisa Umezawa 1 , Toshihiro Yasuniwa 1 , Atsushi Miyama 1 , Hisayuki Nakanishi 1 , Mutsumi Sugiyama 1 , Shigefusa Chichibu 2 Show Abstract
1 Department of Electrical Engineering, Tokyo University of Science, Noda, Chiba Japan, 2 Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, Sendai Japan
Chalcopyrite-structure CuIn1−xGaxSe2 (CIGS) alloys are used as a light-absorbing medium of high conversion efficiency, low cost, lightweight, and radiation-resistant solar cells. The ideal bandgap energy for solar cells is known to be 1.4 eV, which corresponds to the bandgap of CuIn0.4Ga0.6Se2. However, growing single phase CuIn0.4Ga0.6Se2 alloys or CIGS solid solutions of high CuGaSe2 (CGS) molar faction is difficult because of unwanted compositional separation due to the difference in reaction rates of the two end-point compounds. In addition, the open-circuit voltage and conversion efficiency of CIGS solar cells do not increase proportionally with the bandgap, because of insufficient grain size and crystal quality of CIGS films.CuIn1-xAlxSe2 (CIAS) alloys are an attractive alternative of chalcopyrite semiconductor photo-absorber, since the CuAlSe2 (CAS) molar fraction, x, required for obtaining the ideal bandgap is calculated to be as low as 0.2. Therefore, phase separation is less likely than that of CIGS films. In addition, according to wide variation in the bandgap energy (1.04 - 2.67eV), multiple junction (tandem) solar cells can also be fabricated using CIAS films of separate compositions. Polycrystalline CIAS films have been deposited by a variety of methods such as evaporation, selenization, and chemical bath deposition techniques, and several research groups have recently demonstrated CIAS-based solar cells. In analogy with CIGS growth, selenization method is an attractive growth technique in fabricating large-size CIAS solar cells. However, CIAS film growth has been difficult, because it contains chemically active aluminium in the matrix.The authors have proposed the use of a less hazardous metalorganic selenide, diethylselenide [(C2H5)2Se, DESe)], for growing CIGS films by the selenization method. Since DESe decomposes into atomic Se more easily than H2Se gas or Se vapor, reaction rates capable for forming segregation-free CIGS films have been obtained. Therefore, formation of segregation-free CIAS alloy films with high x might also be able. In this presentation, advantages of using DESe for the growth of CIAS films by the selenization method will be shown.As a profit of using DESe, approximately 2.0-um-thick single-phase polycrystalline CIAS films were grown without additional thermal annealing. The films adhered well to the Mo/SLG substrate, which was confirmed by the peeling test. Photoluminescence spectra at low temperature were dominated by characteristic donor-acceptor-pair emission bands, which are peculiar to good performance CIGS solar cell material. These preliminary results suggest that the selenization growth of CIAS alloys using DESe has a potential in fabricating high efficiency solar cells.