Thin Film Photovoltaics: Not Just Silicon Anymore
This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA
Overview
Download & View Thin Film Photovoltaics: Not Just Silicon Anymore as PDF for free.
More details
- Words: 1,930
- Pages: 5
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
Thin Film Photovoltaics: Not Just Silicon Anymore This article is based off the NanoMarkets report Thin Film Photovoltaics, 2008 and Beyond
Until recently, analyses of the photovoltaic industry tended to lump all thin film technologies together. And until recently, this view made sense. Wafer-based silicon technologies accounted for more than 90 percent of the total market. Thin film silicon accounted for almost all of the rest, with other technologies little more than rounding error. Times have changed. The overall market has rocketed upward. Driven in part by a shortage of silicon, and in part by the rise of building-integrated photovoltaics, thin film photovoltaics now account for about 23 percent of a much larger pie. NanoMarkets expects the thin film share to exceed 50 percent by 2015. Moreover, thin film technologies other than silicon have attracted significant attention. Nanosolar, which focuses on copper indium gallium diselenide (CIGS) cells, recently closed on a $300 million venture round, the largest sum raised by a solar startup this year. First Solar, the leading supplier of CdTe-based cells, is the fifth-largest solar cell manufacturer in the world by production, and the only thin film photovoltaic supplier in the top ten. Thin film solar cells have clearly come of age. With the emergence of technologies other than silicon, the thin film landscape has also become much more complicated. Investors and customers alike need a way to compare the various thin film technologies. This article, which is based on a report recently published by NanoMarkets, Thin Film Photovoltaics, 2008 and Beyond, aims to supply such a framework. It’s not clear what the best basis for comparison might be. A customer planning to install panels in the next six months has different requirements than an investor evaluating future market potential. The best technologies on paper have yet to prove themselves in the marketplace. Copper indium gallium diselenide: big potential, but not yet there Consider, for example, overall conversion efficiency. In the relatively mature wafer-based silicon space, commercial products have achieved more than 90 percent of the conversion efficiency of the best research-oriented “champion” cells. At this year’s IEEE Photovoltaic Specialists’ Conference in San Diego, Robert Birkmire of the Institute of Energy Conversion at the University of Delaware used this value as a rule of thumb to estimate the outlook for several thin film cell types. Table 1: Cell vs. Module e fficiency for common cell type s Cell Type c-Si mc-Si Thin film Si (single junction) CIGS CdTe
Cell Efficiency (% )
Module Efficiency (% )
Ratio
24.7 20.3 10-12 19.9 16.5
22.7 15.3 5-8 13.4 10.7
92% 75% * 67% 65%
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Page | 1
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
*
Due to the large number of proposed and commercial cell designs, a direct comparison w as not possible. S ource: Robert W. Birkmire1
Based on these results, CIGS seems to have the most commercial potential. It is the only thin film technology that even approaches the performance of crystalline silicon. Thin film silicon, in contrast, lags far behind. Indeed, the positive outlook for CIGS solar cells is reflected in the venture funding received by companies such as Nanosolar. Yet actually achieving such superior performance in commercial production has been difficult, for two reasons. First, CIGS is a complex quaternary alloy in which indium and gallium can substitute freely. The bandgap of the cell changes as the composition fluctuates. Consistent performance requires uniform deposition over the entire area of the panel. One possible solution, co-sputtering of all four elements, can fail because of target poisoning: the selenium reacts with the surfaces of the other targets, and creates an insulating layer that resists further sputtering. An alternative method, co-sputtering of copper, indium, and gallium, followed by a hydrogen selenide vapor treatment, requires substantial time at elevated temperatures. The time increases the process cost, while the elevated temperature causes phase segregation. Phase control is especially important in light of Heliovolt founder B. J. Stanbery’s identification of an intra-absorber junction (IAJ) in the CIGS layer. According to Stanbery2, the CIGS layer actually consists of a p-type Cu(In,Ga)Se2 α-phase and an In-rich Cu(In,Ga)3Se5 n-type β-phase. Carriers generated in the bulk as electron-hole pairs are swept to the appropriate terminal by the potential at the IAJ. Cell performance therefore depends not only on the overall composition uniformity, but also on nanoscale fluctuations that cause preferential formation of either αphase or β-phase material. Because of the challenges inherent in the material, the commercial production of CIGS has so far fallen short of the more exuberant projections. As well, it is difficult to quantify actual production capacity because of the uncertainty surrounding Nanosolar’s new CIGS plant. Though Nanosolar reports that it is building a massive 430 MW plant in San Jose, the first panels shipped less than a year ago and the company has yet to make any other substantial shipments. Nanosolar has published no information about the performance of its commercial cells, which is in sharp contrast with the company’s aggressive claims two years ago of solar panels for $1/W. Global Solar, which expects to reach 100 MW of capacity by 2010, appears to have the largest actual capacity in the sector. NanoMarkets expects that CIGS’ share of the thin film PV market will expand as these investments come online, but for now the technology accounts for about 10 percent of thin film PV sales. CdTe vaults to leadership Meanwhile, First Solar has exploited the much simpler CdTe technology to become one of the world’s largest thin film solar cell suppliers, with plans to reach more than a gigawatt of manufacturing capacity by the end of 2009. For CdTe, the key manufacturing challenge appears to be control of interface recombination at the electrodes, between the CdTe active layer and the CdS absorber, and elsewhere in the cell. Managing these interfaces seems to be a matter of
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Page | 2
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
careful process control. First Solar claims to have successfully replicated its process across several manufacturing facilities in Germany, Malaysia, and the U.S., while achieving a cost per watt of just $1.14, claimed to be the industry’s lowest. While CIGS cells face fundamental technical challenges, two of the biggest obstacles to CdTe’s success appear to lie in the business realm: the toxicity of cadmium, and the scarcity of tellurium. Yet neither obstacle seems to be as serious as it might first appear. Though cadmium is indeed toxic, it is also present in both coal and crude oil. At 0.02 grams per gigawatt/hour, cadmium emissions over the life of a CdTe panel are far less than would be produced by burning equivalent amounts of these fuels. According to Brookhaven National Lab scientist Vasilis Fthenakis, coal-fired generation plants emit about 3.7 grams of cadmium per gigawatt/hour, while oil-fired plants emit about 44.3 grams per Gigawatt/hour of the material.3 Since cadmium is a byproduct of zinc smelting, it can be argued that encapsulating it in solar panels is more environmentally sound than simply disposing of it as smelting waste. Tellurium, meanwhile, is produced as a byproduct of copper smelting. Fthenakis estimates that the normal growth of copper production is sufficient to support 5-10 GW per year of CdTe panels—more than five times First Solar’s planned capacity—even without improvements in tellurium extraction efficiency. Yet such improvements are likely. For example, when other “waste” metals, such as indium, have found industrial uses, extraction processes have evolved to enhance the capture of those metals. So if CIGS has the most potential but is still immature, and CdTe has overcome supposedly insurmountable obstacles, where does that leave thin film silicon? Don’t write silicon off just yet On paper, its position is fairly tenuous. After all, amorphous silicon has the lowest efficiency among major thin film technologies. On the positive side, though, silicon is an abundant, nontoxic material. Silicon deposition technology is well established in both the integrated circuit and flat panel display industries, giving solar panel makers the support of a large existing infrastructure. While CIGS and CdTe suppliers must develop their own processes, silicon solar cell manufacturers can simply buy much of their technology from companies like Applied Materials and Oerlikon Solar. This maturity gives the thin film silicon sector structural advantages that help offset its relatively inefficient cells. For instance, with the basic technology established, thin film silicon suppliers can pursue innovative cell designs, such as multijunction cells. A multijunction cell effectively stacks several (usually two or three) cells on top of each other. Each cell has a slightly different bandgap, and captures a slightly different portion of the solar spectrum. Many different combinations of amorphous, nanocrystalline, and microcrystalline silicon, are possible, with or without the addition of SiGe. It is difficult to pinpoint a single representative thin film silicon technology. Leading examples include a 13-percent efficient triple-junction cell
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Page | 3
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
from Sharp, stacking two a-Si cells on top of a μc-Si cell, and a 15-percent efficient a-Si/aSiGe/μc-Si stack from Kaneka. These designs bring silicon thin films close to the performance achieved by more exotic compositions. At the same time, variations in the band gap of thin film silicon are achieved by changing the crystallinity of the film—through either deposition temperature or the degree of hydrogen dilution—or by adding germanium. Lattice constant differences between the layers of the stack are minimal, so lattice strain is a much less serious concern than in multijunction cells in the GaAs semiconductor family. Furthermore, the similarity among layers means that they can all be deposited in a single deposition chamber, maintaining clean, low-defect interfaces. These structural advantages help explain why silicon remains the largest segment of the thin film photovoltaic market. Still, NanoMarkets expects silicon’s share to erode over the next several years as CdTe and CIGS manufacture expand. (See Table 2) Companies in the thin film silicon space, many of whom also supply wafer-based silicon cells, will need to make substantial investments in both capacity and technology in order to keep up.
Table 2: Summary of Thin-Film Photovoltaics Markets by Material ($ Million) $10,000.00 $8,000.00 $6,000.00 $4,000.00 $2,000.00 $0.00 2008
2009
2010
Amorphous Silicon
2011 CIS/CIGS
2012 CdTE
2013 GaAs
2014
2015
Other
Source: NanoMark ets report, Thin Film Photovoltaics, 2008 and Beyond .
References [1] Robert W. Birkmire, “Pathways to improved performance and processing of CdTe and CuInSe2 based modules,” 33rd IEEE PV Spec. Conf., San Diego, 2008. [2] B. J. Stanbery, “The Intra-absorber Junction (IAJ) Model for the Device Physics of Copper Indium Selenide-Based Photovoltaics,” IEEE Photovoltaic Specialists Conference, pp 355 - 358 (2005).
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Page | 4
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
[3] Vasilis Fthenakis, “Sustainability of Thin-film CIGS and CdTe Photovoltaics,” 33rd IEEE PV Spec. Conf., San Diego, 2008. For additional information about NanoMarkets and its full listing of reports and services please visit us on the web at www.nanomarkets.net or emailing us at sales ” at”nanomarkets.net. Page | 5
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
thin film | organic | printable | electronics
www.na nomarkets.net
Thin Film Photovoltaics: Not Just Silicon Anymore This article is based off the NanoMarkets report Thin Film Photovoltaics, 2008 and Beyond
Until recently, analyses of the photovoltaic industry tended to lump all thin film technologies together. And until recently, this view made sense. Wafer-based silicon technologies accounted for more than 90 percent of the total market. Thin film silicon accounted for almost all of the rest, with other technologies little more than rounding error. Times have changed. The overall market has rocketed upward. Driven in part by a shortage of silicon, and in part by the rise of building-integrated photovoltaics, thin film photovoltaics now account for about 23 percent of a much larger pie. NanoMarkets expects the thin film share to exceed 50 percent by 2015. Moreover, thin film technologies other than silicon have attracted significant attention. Nanosolar, which focuses on copper indium gallium diselenide (CIGS) cells, recently closed on a $300 million venture round, the largest sum raised by a solar startup this year. First Solar, the leading supplier of CdTe-based cells, is the fifth-largest solar cell manufacturer in the world by production, and the only thin film photovoltaic supplier in the top ten. Thin film solar cells have clearly come of age. With the emergence of technologies other than silicon, the thin film landscape has also become much more complicated. Investors and customers alike need a way to compare the various thin film technologies. This article, which is based on a report recently published by NanoMarkets, Thin Film Photovoltaics, 2008 and Beyond, aims to supply such a framework. It’s not clear what the best basis for comparison might be. A customer planning to install panels in the next six months has different requirements than an investor evaluating future market potential. The best technologies on paper have yet to prove themselves in the marketplace. Copper indium gallium diselenide: big potential, but not yet there Consider, for example, overall conversion efficiency. In the relatively mature wafer-based silicon space, commercial products have achieved more than 90 percent of the conversion efficiency of the best research-oriented “champion” cells. At this year’s IEEE Photovoltaic Specialists’ Conference in San Diego, Robert Birkmire of the Institute of Energy Conversion at the University of Delaware used this value as a rule of thumb to estimate the outlook for several thin film cell types. Table 1: Cell vs. Module e fficiency for common cell type s Cell Type c-Si mc-Si Thin film Si (single junction) CIGS CdTe
Cell Efficiency (% )
Module Efficiency (% )
Ratio
24.7 20.3 10-12 19.9 16.5
22.7 15.3 5-8 13.4 10.7
92% 75% * 67% 65%
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Page | 1
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
*
Due to the large number of proposed and commercial cell designs, a direct comparison w as not possible. S ource: Robert W. Birkmire1
Based on these results, CIGS seems to have the most commercial potential. It is the only thin film technology that even approaches the performance of crystalline silicon. Thin film silicon, in contrast, lags far behind. Indeed, the positive outlook for CIGS solar cells is reflected in the venture funding received by companies such as Nanosolar. Yet actually achieving such superior performance in commercial production has been difficult, for two reasons. First, CIGS is a complex quaternary alloy in which indium and gallium can substitute freely. The bandgap of the cell changes as the composition fluctuates. Consistent performance requires uniform deposition over the entire area of the panel. One possible solution, co-sputtering of all four elements, can fail because of target poisoning: the selenium reacts with the surfaces of the other targets, and creates an insulating layer that resists further sputtering. An alternative method, co-sputtering of copper, indium, and gallium, followed by a hydrogen selenide vapor treatment, requires substantial time at elevated temperatures. The time increases the process cost, while the elevated temperature causes phase segregation. Phase control is especially important in light of Heliovolt founder B. J. Stanbery’s identification of an intra-absorber junction (IAJ) in the CIGS layer. According to Stanbery2, the CIGS layer actually consists of a p-type Cu(In,Ga)Se2 α-phase and an In-rich Cu(In,Ga)3Se5 n-type β-phase. Carriers generated in the bulk as electron-hole pairs are swept to the appropriate terminal by the potential at the IAJ. Cell performance therefore depends not only on the overall composition uniformity, but also on nanoscale fluctuations that cause preferential formation of either αphase or β-phase material. Because of the challenges inherent in the material, the commercial production of CIGS has so far fallen short of the more exuberant projections. As well, it is difficult to quantify actual production capacity because of the uncertainty surrounding Nanosolar’s new CIGS plant. Though Nanosolar reports that it is building a massive 430 MW plant in San Jose, the first panels shipped less than a year ago and the company has yet to make any other substantial shipments. Nanosolar has published no information about the performance of its commercial cells, which is in sharp contrast with the company’s aggressive claims two years ago of solar panels for $1/W. Global Solar, which expects to reach 100 MW of capacity by 2010, appears to have the largest actual capacity in the sector. NanoMarkets expects that CIGS’ share of the thin film PV market will expand as these investments come online, but for now the technology accounts for about 10 percent of thin film PV sales. CdTe vaults to leadership Meanwhile, First Solar has exploited the much simpler CdTe technology to become one of the world’s largest thin film solar cell suppliers, with plans to reach more than a gigawatt of manufacturing capacity by the end of 2009. For CdTe, the key manufacturing challenge appears to be control of interface recombination at the electrodes, between the CdTe active layer and the CdS absorber, and elsewhere in the cell. Managing these interfaces seems to be a matter of
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Page | 2
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
careful process control. First Solar claims to have successfully replicated its process across several manufacturing facilities in Germany, Malaysia, and the U.S., while achieving a cost per watt of just $1.14, claimed to be the industry’s lowest. While CIGS cells face fundamental technical challenges, two of the biggest obstacles to CdTe’s success appear to lie in the business realm: the toxicity of cadmium, and the scarcity of tellurium. Yet neither obstacle seems to be as serious as it might first appear. Though cadmium is indeed toxic, it is also present in both coal and crude oil. At 0.02 grams per gigawatt/hour, cadmium emissions over the life of a CdTe panel are far less than would be produced by burning equivalent amounts of these fuels. According to Brookhaven National Lab scientist Vasilis Fthenakis, coal-fired generation plants emit about 3.7 grams of cadmium per gigawatt/hour, while oil-fired plants emit about 44.3 grams per Gigawatt/hour of the material.3 Since cadmium is a byproduct of zinc smelting, it can be argued that encapsulating it in solar panels is more environmentally sound than simply disposing of it as smelting waste. Tellurium, meanwhile, is produced as a byproduct of copper smelting. Fthenakis estimates that the normal growth of copper production is sufficient to support 5-10 GW per year of CdTe panels—more than five times First Solar’s planned capacity—even without improvements in tellurium extraction efficiency. Yet such improvements are likely. For example, when other “waste” metals, such as indium, have found industrial uses, extraction processes have evolved to enhance the capture of those metals. So if CIGS has the most potential but is still immature, and CdTe has overcome supposedly insurmountable obstacles, where does that leave thin film silicon? Don’t write silicon off just yet On paper, its position is fairly tenuous. After all, amorphous silicon has the lowest efficiency among major thin film technologies. On the positive side, though, silicon is an abundant, nontoxic material. Silicon deposition technology is well established in both the integrated circuit and flat panel display industries, giving solar panel makers the support of a large existing infrastructure. While CIGS and CdTe suppliers must develop their own processes, silicon solar cell manufacturers can simply buy much of their technology from companies like Applied Materials and Oerlikon Solar. This maturity gives the thin film silicon sector structural advantages that help offset its relatively inefficient cells. For instance, with the basic technology established, thin film silicon suppliers can pursue innovative cell designs, such as multijunction cells. A multijunction cell effectively stacks several (usually two or three) cells on top of each other. Each cell has a slightly different bandgap, and captures a slightly different portion of the solar spectrum. Many different combinations of amorphous, nanocrystalline, and microcrystalline silicon, are possible, with or without the addition of SiGe. It is difficult to pinpoint a single representative thin film silicon technology. Leading examples include a 13-percent efficient triple-junction cell
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Page | 3
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
from Sharp, stacking two a-Si cells on top of a μc-Si cell, and a 15-percent efficient a-Si/aSiGe/μc-Si stack from Kaneka. These designs bring silicon thin films close to the performance achieved by more exotic compositions. At the same time, variations in the band gap of thin film silicon are achieved by changing the crystallinity of the film—through either deposition temperature or the degree of hydrogen dilution—or by adding germanium. Lattice constant differences between the layers of the stack are minimal, so lattice strain is a much less serious concern than in multijunction cells in the GaAs semiconductor family. Furthermore, the similarity among layers means that they can all be deposited in a single deposition chamber, maintaining clean, low-defect interfaces. These structural advantages help explain why silicon remains the largest segment of the thin film photovoltaic market. Still, NanoMarkets expects silicon’s share to erode over the next several years as CdTe and CIGS manufacture expand. (See Table 2) Companies in the thin film silicon space, many of whom also supply wafer-based silicon cells, will need to make substantial investments in both capacity and technology in order to keep up.
Table 2: Summary of Thin-Film Photovoltaics Markets by Material ($ Million) $10,000.00 $8,000.00 $6,000.00 $4,000.00 $2,000.00 $0.00 2008
2009
2010
Amorphous Silicon
2011 CIS/CIGS
2012 CdTE
2013 GaAs
2014
2015
Other
Source: NanoMark ets report, Thin Film Photovoltaics, 2008 and Beyond .
References [1] Robert W. Birkmire, “Pathways to improved performance and processing of CdTe and CuInSe2 based modules,” 33rd IEEE PV Spec. Conf., San Diego, 2008. [2] B. J. Stanbery, “The Intra-absorber Junction (IAJ) Model for the Device Physics of Copper Indium Selenide-Based Photovoltaics,” IEEE Photovoltaic Specialists Conference, pp 355 - 358 (2005).
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Page | 4
NanoMarkets
thin film | organic | printable | electronics
www.na nomarkets.net
[3] Vasilis Fthenakis, “Sustainability of Thin-film CIGS and CdTe Photovoltaics,” 33rd IEEE PV Spec. Conf., San Diego, 2008. For additional information about NanoMarkets and its full listing of reports and services please visit us on the web at www.nanomarkets.net or emailing us at sales ” at”nanomarkets.net. Page | 5
NanoMarkets, LC | PO Box 3840 | Glen Allen, VA 23058 | TEL: 804-360-2967 | FAX: 804-360-2967
Related Documents
Thin Film Photovoltaics: Not Just Silicon Anymore
November 2019 5
Materials For Thin-film Photovoltaics
November 2019 22
Positioning Thin-film Photovoltaics For Success
November 2019 15
Amorphous Silicon And Photovoltaics
June 2020 7
Thin Film Deposition
May 2020 6
Thin Film Solar Cells
April 2020 9More Documents from ""
Amorphous Silicon And Photovoltaics
June 2020 7
Silver's Role In Solar Energy
June 2020 4
Five Opportunities For Ito Replacement
November 2019 18
Fuels Rev Up Catalyst Demand
June 2020 9
