当前位置:首页 >> 能源/化工 >>

CIS-based Thin-Film Photovoltaic Modules Potential and Prospects


PROGRESS IN PHOTOVOLTAICS: RESEARCH AND APPLICATIONS Prog. Photovolt: Res. Appl. 2002; 10:149–157 (DOI: 10.1002/pip.413)

Special Issue

CIS-based Thin-Film Photovoltaic Modules: Potential and Prospects
B. Dimmler1,*,y M. Powalla2 and H. W. Schock3
¨ Wurth Solar GmbH & Co. KG, Ludwigsburger Strasse 100, D-71672 Marbach am Neckar, Germany ¨ ¨ ¨ Zentrum fur Sonnenenergie- und Wasserstoff-Forschung (ZSW), Baden-Wurttemberg, Hessbruhlstrasse 21c, D-70565 Stuttgart, Germany 3 ¨ ¨ Institut fur Physikalische Elektronik (IPE), Universitat Stuttgart, Pfaffenwaldring 47, D-70569 Stuttgart, Germany
2 1

Cu(In,Ga)Se2 (CIGS), one of the main candidates of thin ?lm absorber materials, is considered to have the highest potential with respect to module quality. To achieve very large capacities further developments of the manufacturing technologies are still necessary. On the basis of the fundamental work of IPE the ZSW has developed the technologies for all process steps on a module size of 30 ? 30 cm2 on glass substrates. Average module ef?ciencies are well above 10%, with a tendency to improve ef?ciency to a maximum of 12?7% on a 30 ? 30 cm2 module area. The next ¨ step of scaling up has started as a joint venture of ZSW, the Wurth group and the ¨ EnBW AG, forming Wurth Solar GmbH & Co. KG. The objective of the company is to commercialise CuIn Se2 (CIS) technology and PV products. Operation of a pilot line with ?nal nominal capacity of about 1 MWp /yr was began in spring 2000. After transfer of ZSW baseline technology and successful optimisation of the pilot line, the ?rst modules of 60 ? 120 cm2 area, with ef?ciencies exceeding 8%, have been fabricated. Copyright # 2002 John Wiley & Sons, Ltd.

HISTORY: FROM BASIC RESEARCH TO COMMERCIAL FABRICATION (STUTTGART)
ince the early 1980s the fundamentals of material and processing for CIGS-based solar cells have been developed at the Institute of Physical Electronics (IPE) at the University of Stuttgart. About 10 years later the technology was transferred and scaled up in size, together with module technology applying ¨ production-relevant technologies, at the ZSW. On this basis Wurth Solar constructed and operates a pilot line to establish manufacturing technology for CIS modules at a size of up to 60 ? 120 cm2. After proof of the concept of high productivity under industrial conditions, mass production for CuInSe2 (CIS) modules will begin. The historical development is illustrated in Figure 1. This shows the improvement of the cell ef?ciency and the increase of module size over the past 20 years within the CIS development in Stuttgart. Very important milestones in this development were the quality of the solar cells and modules. The scalingup activities of ZSW, ?rst to 100 cm2, was not started until a small-area ef?ciency above 15% for the 1 cm2 size was reached at IPE. After reaching 10% for 100 cm2, ZSW further scaled up the size of the modules to about
¨ * Correspondence to: B. Dimmler, Wurth Solar GmbH & Co. KG, Ludwigsburger Strase 100, D-71672 Marbach am Neckar, Germany. y E-mail: wuerth.solar@we-online.de ¨ ¨ Contract/grant sponsors: BMBF; BMWi; Wirtschaftsministerium Baden-Wurttemberg; Stiftung Energieforschung Baden-Wurttemberg; EC. Published online 28 January 2002 Copyright # 2002 John Wiley & Sons, Ltd. Received 2 April 2001 Revised 2 August 2001

S

150

B. DIMMLER, M. POWALLA AND H. W. SCHOCK

Figure 1. The route from laboratory scale, with intermediate steps towards manufacturing

1000 cm2. This step began the development of industrially compatible process technology including process control, because most of the speci?c equipment was not available from industrial sources. On the basis of this technology, the ?nal step in scaling up the size to 120 ? 60 cm2, which will be the standard product size, has started. At this stage the development is more and more concentrated on all productivity parameters, mainly throughput, yield and materials cost. So the next step, to the 10 MWP/yr capacity range, depends on the productivity parameter of the 1 MWP/yr pilot production.

BASIC PROPERTIES AND POTENTIAL OF CIS
CuInSe2 has been known to be an ef?cient photovoltaic material for more than 25 years. However, production at pilot line level has started only recently. The reason for this slow development lies in the complexity of the CuInSe2-based solar cell and the processes involved. The dif?culty lies either in the complex process technology, such as co-evaporation of the elements, or in controlling such simple processes as selenisation of metal ?lms. Speci?c equipment for the deposition has to be developed. At ?rst, solar cells were based upon pure CuInSe2 single crystals with a n-type window layer deposited by the evaporation of CdS.1 Thin-?lm solar cells fabricated by co-evaporation of the elements soon reached an ef?ciency of 10%.2 Control of the deposition process and scaling up turned out to be dif?cult. Therefore, the ?rst approaches to large-area production used the so-called two-stage process where the deposition of metal ?lms by sputtering is followed by a selenisation step in H2Se.3 The small bandgap of about 1 eV of CuInSe2 can be easily increased by adding gallium or/and sulphur in order to form the ?ve-component alloy Cu(In,Ga)(S,Se)2. This alloy system opens up many possibilities for device optimisation. The addition of new components to the absorber layer and the device structure contributed to the advance of high-ef?ciency solar cell development:4
*

*

* *

adding Ga to the absorber layer increased process tolerance and made it possible to increase the open-circuit voltage; replacement of the evaporated CdS layer by a combination of a thin CdS layer with ZnO as a transparent conductor improved junction performance and photocurrent; incorporation of Sulphur at the absorber surface helped to increase the open-circuit voltage; understanding the role of sodium made it possible to control carrier concentration in the CIGS absorber layers.5

Together with an improved understanding of the material and the development of innovative technologies, co-evaporation of the elements became a viable and ?exible method for large-area deposition of the CIGS ?lms,6 with the following features:
Copyright # 2002 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2002; 10:149–157

CIS-BASED MODULES

151

* * * *

automated control and monitoring of complex processes; rate control by atomic absorption spectroscopy; control of composition by X-ray ?uorescence; development of suitable evaporation sources.

The potential of the multi-component material Cu(In,Ga)(S,Se)2 as a photovoltaic material is not yet fully exploited, even though the ef?ciency of laboratory-scale devices exceeds 18%. Features of high-ef?ciency crystalline cells, such as back-surface ?eld, passivation of front surface and contacts are not yet deliberately implemented. However, the potential for optimisation needs careful evaluation because some features are already intrinsic to the material and device design. The wide bandgap window of the heterojunction does not absorb a signi?cant amount of light and thus does not contribute to photocurrent generation. It could have the function of a front-surface passivation. Furthermore, the surface of the absorber layer is type-inverted, so that an intrinsic pn junction is formed. Copper depletion towards the surface lowers the valence band.7 Thus, an additional barrier for holes is formed, so that interface recombination is suppressed. The Mo back contact forms an interfacial MoSe2 layer which passivates the back contact. In addition, the tendency of Ga to accumulate at the back surface provides an additional back-surface ?eld, due to the increase of the bandgap towards the back contact. Owing to the complexity of the material, the possibilities for modi?cations of the device depend on to the interaction between the single layers and the stability of the interfaces. The superior stability of CIGS solar cells is based on the favourable defect structure of the absorber layer and the thermodynamic stability of the interfaces.8 The aspects of future developments are manifold and a summary is given elsewhere.9

SCALING UP
Worldwide state of the art Several research institutes and companies are working in the ?eld of CIS thin-?lm solar modules, mainly in USA, Japan and Europe. The best ef?ciencies of laboratory cells, mini-modules and modules are compiled in Figure 2 as a function of cell or module area. Ef?ciencies in the range 16–19% for small cells (<1 cm2) are achieved by several groups (UU/IPE,10,11 NREL12 and Matsushita13,14). These cells are all prepared by thermal co-evaporation of CIGS, and demonstrate the high potential of the CIS technology. In 1999 UU15 and Siemens Solar3 demonstrated ef?ciencies of mini-modules close to 15%. Also, 16?6% ef?ciency for a small module has been demonstrated by UU.16 The results marked with circles in Figure 2 were made with sophisticated laboratory processes. There is no apparent obstacle to transferring these results to large areas. However, modules in the square meter range have not yet been demonstrated. All larger modules have ef?ciencies below

Figure 2. Ef?ciencies of CIGS cells and modules from different laboratories and companies as a function of size
Copyright # 2002 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2002; 10:149–157

152

B. DIMMLER, M. POWALLA AND H. W. SCHOCK

¨ 13% (ZSW, 17 Siemens Solar, 18 Showa Shell 19). Wurth Solar expects to market 120 ? 60 cm2 modules with ef?ciencies exceeding 8%. Losses in performance of the ?rst-generation modules are mainly caused by the non-active area of the patterning and as yet unsatisfactory homogeneity. It can be expected that 14–15% module ef?ciencies might be attained after a thorough optimisation of the process technology. Baseline processes at the 30 ? 30 cm2 CIS line at ZSW The standard fabrication sequence for 30 ? 30 cm2 CIGS modules at ZSW is the following:20
* *

* *

*

*

* *

soda-lime sheet glass with a thickness of 1–3 mm is used as substrate for the thin ?lms; a molybdenum back contact is deposited by d.c. magnetron sputtering. The preparation conditions for Mo in combination with the Na containing glass have signi?cant in?uence on cell performance; for monolithic integration the back contact is patterned by laser scribing; simultaneous evaporation of the elements on continuously moved substrates (in-line) is applied for the deposition of the absorber layer; for the formation of the heterojunction, a thin buffer layer of CdS is prepared by chemical bath deposition (CBD). Alternatively other compounds, such as In(OH, S) and others, are applied to produce a completely Cd-free device; ZnO has been found to be the most suitable material for the n-type transparent electrode. ZnO is fabricated in a two-layer process (undoped i-ZnO and Al-doped ZnO) by DC magnetron sputtering from ceramic targets; the semiconductor patterning (second and third patterning step) is done by mechanical scribing; the modules are ?nished by bonding electrical contacts and standard glass–glass lamination to encapsulate the CIS module for long-term stable operation in any outdoor environment.

CIS fabrication by co-evaporation Most of the deposition and patterning techniques used for CIS module fabrication are new and not directly compatible with well-known industrial processing. One of the most critical issues is how to make CIS ?lms with suf?cient homogeneity and quality at high throughput and yield on large areas. We use the co-evaporation

Figure 3. The development of the thermal co-evaporation method of CIGS from the laboratory to the industrial scale ¨ Arrangements (a), (b) are realised at IPE, (c) and (d) at ZSW, (e) is the production version at Wurth Solar
Copyright # 2002 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2002; 10:149–157

CIS-BASED MODULES

153

method because this technique has the highest quality potential up to now. The steps from the deposition arrangement from the square centimeter batch process to the square meter in-line process is shown in Figure 3. Arrangements (a), (b) and (c) are batch processes, but offer the possibility to optimise the main process parameters, such as evaporation pro?les for the different elements, substrate temperature pro?les, process control of evaporation rates and composition. The in-line approach (c) was simulated by rotating the substrate holder. Con?guration (d) at ZSW, and (e) at ¨ Wurth Solar are real in-line coaters. Key issues are the evaporation sources, which have a linear evaporation pro?le, perpendicular to the moving direction and the temperature budget of the whole system. The transport of the glass substrates at the high process temperatures near the softening point enhances process complexity. A key issue to achieve homogeneous large-area ?lms, beside the evaporation source itself, is the rate control of evaporation sources. High reproducibility and homogeneity of the deposition process is realised with atomic absorption spectroscopy of the ?ux and X-ray ?uorescence measurements of the ?lms, together with an automation system. Process statistics The maturity of process technology is assessed by product quality statistics. Accordingly, more than 200 CIGS modules were produced in the ZSW line in different batches with minor process variation within a year. Figure 4 gives the distribution of the ef?ciency. The overall yield of the single process steps (deposition and scribing) are well above 90%, and average ef?ciency is 11?1%. A few modules have been excluded because of problems in manual handling. Figure 4 contains all the modules, which pass through the line. With more practical experience and additional automation the variation of the distribution could be further limited. The challenge is now to ¨ transfer this process technology to the Wurth Solar pilot line with an even narrower ef?ciency distribution, which should be possible with increased automation. Cost estimation Cost estimates are very important in this state of development, because the pro?tability of thin-?lm solar cell manufacturing has not been demonstrated up to now. At the moment, the investment costs to get into the technology are rather high. This is the main reason that the capacity of a pro?table manufacturing plant has to be at least about 10 MWP/yr. A European study on ‘multi-megawatt upscaling of PV technologies’ (APAS/MUSIC?FM21 compared PV technologies. It was clear, that for thin-?lm solar modules, it should be possible, for the three most important materials: a-Si (Phototronics/ASE), CIGS (ZSW) and CdTe (BP Solar), to fabricate modules at costs far below

Figure 4. Ef?ciency distribution of 218 CIGS modules (30 ? 30 cm processed in several batches at the ZSW line, not encapsulated)
Copyright # 2002 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2002; 10:149–157

154

B. DIMMLER, M. POWALLA AND H. W. SCHOCK

Figure 5. Estimation of production costs with increasing production capacity

1 s/Wp at a production capacity of 60 MWp/yr. In comparison, costs for crystalline silicon are not expected to fall below 1 s/Wp, even at production levels of 500 MWp/yr, except with the EFG process used by ASE Americas. Figure 5 shows the estimated data of the reduction of the production costs for CIS modules with increasing production capacity. The calculation of very large production volumes, gives a rough indication of the learning curve for CIS technology. The cost categories for equipment, plant area, energy, labour, materials and miscellaneous items are taken into account. The main presumed effects are the reduction of equipment costs, due to a decreased share of development costs for the supplier, as well as reduced energy and labour cost, due to optimised and highly automated systems. Additionally it is assumed that device quality is increased to 12% on average at a process yield of 90%. It is very important to note that these estimates imply combined efforts of R&D and production experience. Reliability Within the framework of a joint European research project (ANTHEM), several representative CIS modules of ZSW were tested at the Joint Research Centre (Ispra) under the test procedures according to the EN 61646 standard. The test sequence comprised the most demanding test levels, such as damp heat (1000 h at 85 C, 85% relative humidity), thermal cycling (50 cycles ? 40 to 85 C), humidity freeze (10 cycles 85 C, 85% relative humidity to ? 40 C), UV exposure (UV-A and UV-B, total 15 kWh), insulation tests, and mechanical tests, including preconditioning steps, such as controlled light soaking and annealing, as speci?ed in the standard. The prototype CIS modules fabricated at ZSW have passed all relevant tests, except for minor problems which can easily be solved by slight changes in module technology. Figure 6 gives an example of the most critical test for CIS modules in damp heat conditions (tested and measured at ZSW).

Figure 6. CIGS module ef?ciencies (encapsulated) under damp heat test conditions (according to IEC 1646 for 1000 h exposure, module size 30 ? 30 cm)
Copyright # 2002 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2002; 10:149–157

CIS-BASED MODULES

155

The modules withstand these very severe test conditions, which allow only 5% degradation after 1000 h at 85% relative humidity and 85 C. The initial ef?ciency of the modules is in the range of 11%. Additionally, outdoor measurements of CIS modules are done at ZSW Widderstall test ?eld near Stuttgart. The investigations indicate that CIS modules are in principle as stable as crystalline Si modules, with comparable properties with regard to temperature and light intensity effects.

LARGE-SCALE MANUFACTURING
¨ With the foundation of Wurth Solar, the next step to increase module size to 60 ? 120 cm2 and to prove the concept for high yield and cycle time, while maintaining high quality, has already started. A new company was founded with the purpose of proving material and fabrication technology to be applicable to a highly ¨ productive manufacturing line. The company was founded in spring 1999 as a joint venture of the Wurth Group, the EnBW and the ZSW. ¨ Adolf Wurth GmbH & Co. KG, a company mainly specialised in supplying all kinds of accessories and tools to craftworkers and workshops worldwide, holds 79?5% of the shares, EnBW AG, the most important energy ¨ ¨ provider in Baden-Wurttemberg, Germany holds 20% and the ZSW 0?5%. Wurth Solar is associated with ¨ ¨ Wurth SOLERGY as part of the Wurth-Elektronik group, which is a manufacturer of printed circuits, and an international distributor of photovoltaic systems. The objective of the company is to manufacture CIS-based thin-?lm modules on a low-cost basis. In the ?rst step, a pilot line was constructed and brought into operation to fabricate modules on glass plates with an area of 60 ? 120 cm2. The aim of the pilot line is to develop the industrial processing under real manufacturing conditions in continuous operation. All productivity parameters, such as throughput, process yield, availability of equipment, are being optimised to be assessed with the aim of constructing a multi-megawatt line with at least 10 MW/yr capacity to minimise the investment risk. The main tasks for the development and implementing of production are as follows:
* * * * * * *

optimisation the prototype equipment together with the supplier; reoptimisation of process parameters with respect to homogeneity and quality; assessment of productivity of the pilot line under industrial conditions; training of employees; increase of the degree of automation of the single process steps as well as the linking of steps; recalculation of the manufacturing costs given in Figure 5 with proven data; de?nition of standard products and interfaces to the PV system.

The commercialisation of the CIS technology is supported by the close collaboration with the IPE and ZSW with respect to further developments and innovations of materials science and process technology. With crosschecks to the ZSW line, single process steps are optimised. The ZSW crew supports the start-up of the pilot line. Direct long-term support of the commercial activity by ZSW’s advanced CIS module technology is foreseen to make the commercialisation of the CIS technology successful. After the planning phase, the building and infrastructure for the pilot line were constructed within an old electricity plant in Marbach, near Stuttgart. The installation of the equipment was completed in May 2000. Since then standard processes developed at the ZSW line have been transferred to the pilot line and optimised within the rest of the year 2000. Since January 2001 optimisation of all process steps led to the ?rst 60 ? 120 cm2 CIS modules with ef?ciencies exceeding 8%. It is planned to offer a wide product diversity to the PV market. Beside standard modules, customer-designed modules for building integration and consumer or industrial product integration are planned. Figure 7 shows a standard 60 ? 120 cm2 CIGS module. The average ef?ciency of the increasing number of modules is expected to reach 10% this year and 12% next year. To reach the cost goals depicted in Figure 5, a bigger line with capacity higher than 10 MWp /yr will follow as soon as proof of the concept of high productivity is shown. In a realistic scenario this could be the case within the next 2–3 years.
Copyright # 2002 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2002; 10:149–157

156

B. DIMMLER, M. POWALLA AND H. W. SCHOCK

¨ Figure 7. CIGS module (120 ? 60 cm2) fabricated in the Wurth Solar pilot line

CONCLUSION
CIGS, and especially ?lms fabricated by the co-evaporation method, has shown the highest quality potential of all thin-?lm materials and is comparable to the well-established crystalline silicon technology. The potential of the multi-component compound Cu(In,Ga)(S,Se)2 as a photovoltaic material is not yet fully exploited. Nevertheless, the CIGS solar cell exhibits the best economic prospects among all PV technologies. However, for a ?nal optimisation, important material properties are still lacking and dif?cult to access. Fabrication technologies are being developed with prototype equipment to start manufacturing. However, industrial fabrication at ¨ high productivity is still to be achieved. Under these circumstances Wurth Solar GmbH & Co. KG started operation of a pilot line in Marbach early in 2000. It is clear that the 1 MWP pilot line is not the optimal size to reach the cost goals, but is an important intermediate step towards economically reasonable production of CIS modules. After successful demonstration of the technology in this pilot line, the annual capacity of around 1 MWp will be raised by at least one order of magnitude within a short time, with the prospect of fabrication costs much lower than all other PV technologies. Acknowledgements ¨ The authors thank the CIS teams at IPE, ZSW and Wurth Solar for their successful work in scaling up CIS technology. The authors thank J. Werner for the continuous support of CIS research. The work has been ¨ supported by the Bundesministerium fur Bildung, Wissenschaft, Forschung und Technologie (BMBF), the ¨ ¨ Bundesministerium fur Wirtschaft und Technologie (BMWi), the Wirtschaftsministerium Baden-Wurttemberg, ¨ the Stiftung Energieforschung Baden-Wurttemberg and by the European Commission under various contracts during the last 25 years.

REFERENCES
1. Wagner S, Shay JL, Migliorato P, Kasper HM. CuInSe2/CdS heterojunction photovoltaic detectors. Applied Physics Letters 1974; 25: 434–435. 2. Mickelsen RA, Chen WS. Polycrystalline thin-?lm CuInSe2 solar cells. Conference, Record 16th IEEE Photovoltaic Specialists Conference, Piscataway: IEEE Press, San Diego, 1982; 781–785.
Copyright # 2002 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2002; 10:149–157

CIS-BASED MODULES

157

3. Karg FH. Development and manufacturing of CIS thin ?lm solar modules. Solar Energy Materials and Solar Cells 2001; 66: 645–653. 4. Rau U, Schock HW. Electronic properties of Cu(In,Ga)Se2 solar cells—recent achievements, current understanding, and future challenges. Applied Physics A 1999; 69: 131. 5. Braunger D, Hariskos D, Bilger G, Rau U, Schock HW. In?uence of sodium on the growth of polycrystalline Cu(In,Ga)Se2 thin ?lms. Thin Solid Films 2000; 361–362: 161–166. ¨ 6. Dimmler B, Gross E, Menner R, Powalla M, Hariskos D, Ruckh M, Ruhle U, Schock HW. Thin ?lm solar modules based on CIS prepared by the co-evaporation method. Proceedings of the 25th Photovoltaic Specialists Conference. IEEE Press: New York, 1996; 757–762. 7. Schmid D, Ruckh M, Grunwald F, Schock HW. Chalcopyrite/defect chalcopyrite heterojunctions on basis of CuInSe2. Journal of Applied Physics 1993; 73: 2902–2909. 8. Guillemoles JF, Kronik L, Cahen D, Rau U, Jasenek A, Schock HW. Stability issues of Cu(In,Ga)Se2-based solar cells. Journal of Physics Chemistry 2000; B104: 4849–4862. 9. Schock HW, Nou? R. CIGS-based solar cells for the next millennium. Progress in Photovoltaics: Research and Applications 2000; 8(1): 151–160. 10. Dimmler B, Menner R, Powalla M, Ruehle R, Schock HW. Thin ?lm solar modules based on Cu(In,Ga)Se2 prepared by the coevaporation method. Proceeding of the 13th Photovoltaic Solar Energy Conference. H. S. Stephens & Associates: Bedford, UK, 1995; 1451. ¨ 11. Schock HW, Rau U, Dullweber T, Hanna G, Balboul M, Margorian-Friedlmeier T, Jasenek A, Kotschau I, Kerber H, Wiesner H. High ef?ciency, high voltage solar cells by bandgap and defect engineering in Cu(In,Ga)(S,Se)2 chalcopyrite semiconductors. Proceedings of the 16th European Photovoltaic Solar Energy Conference. James & James: Glasgow, UK, 2000; 304–308. 12. Contreras MA, Egaas B, Ramanathan K, Hiltner J, Swartzlander A, Hasoon F, Nou? R. Progress toward 20% ef?ciency in Cu(In,Ga)Se2 polycrystalline thin-?lm solar cells. Progress in Photovoltaics: Research and Applications 1999; 7: 311–316. 13. Negami T, Hashimoto Y, Nishiwaki S. Large-area CIGS absorbers prepared by physical vapor deposition. Solar Energy Materials and Solar Cells 2001; 67: 1–9. 14. Kirtagawa M, Negami T. R&D of high ef?cient CIGS thin ?lm solar cells for production. Technical Digest of the 14th Sunshine Workshop on Thin Film Solar Cells. M. Konagai, Meguro: Tokyo, Japan, 2001; 86–94. ? ¨ 15. Kessler J, Bodegard M, Hedstrom J, Stolt L. Baseline Cu(In Ga)Se2 device production: control and statistical signi?cance. Technical Digest of the 11th International Photovoltaic Science and Engineering Conference. T. Saitoh: Tokyo, Japan, 1999; 631–632. ? ¨ 16. Kessler J, Bodegard M, Hedstrom J, Stolt L. New world record Cu(In Ga)Se2 based mini-module: 16?6%. Proceeding of the 16th European Photovoltaic Solar Energy Conference. James & James: Glasgow, UK, 2000; 2057–2060. 17. Powalla M, Dimmler B. Scaling up issues of CIGS solar cells. Thin Solid Films 2000; 361–362: 540. 18. Karg FH. Development and manufacturing of CIS thin ?lm solar modules. Solar Energy Materials and Solar Cells 2001; 66: 645–653. 19. Kushiya K, Tachiyuki M, Nagoya Y, Fujimaki A, Sang B, Okumura D, Satoh M, Yamase O. Progress in large-area Cu(In Ga)Se2-based thin-?lm modules with a Zn(O,S,OH)x buffer layer. Solar Energy Materials and Solar Cells 2001; 67: 11–20. 20. Dimmler B, Gross E, Kessler F, Lotter E, Powalla M, Springer J, Stein U, Voorwinden G, Gaeng M, Schleicher S. thin ?lm module technology: towards commercialization. Proceedings of the 2nd World Conference on Photovolatiac Solar Energy Conversion 1998; 419–423. 21. Woodcock JM, Schade H, Maurus H, Dimmler B, Springer J, Ricaud A. A study of the upscaling of thin ?lm solar cell manufacture towards 500 MWp per annum. Proceedings of the 14th European Photovoltaic Solar Energy Conference. Barcelona. H. S. Stephens & Associates: Bedford, UK, 1997; 857–860.

Copyright # 2002 John Wiley & Sons, Ltd.

Prog. Photovolt: Res. Appl. 2002; 10:149–157


相关文章:
更多相关标签: