USF Libraries
USF Digital Collections

Characterization of cadmium zinc telluride films and solar cells on glass and flexible substrates by rf sputtering

MISSING IMAGE

Material Information

Title:
Characterization of cadmium zinc telluride films and solar cells on glass and flexible substrates by rf sputtering
Physical Description:
Book
Language:
English
Creator:
Gaduputi, Jagadish
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
photovoltaics
tandem devices
CZT
polyimide
Dissertations, Academic -- Electrical Engineering -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: High performance multijunction solar cells based on polycrystalline thin films will require a wide bandgap top cell with at least 15% efficiency. With the bottom cell being CIGS which have already demonstrated the required efficiencies, this work aims to study the complete fabrication and performance of Cd1-xZnxTe solar cells with a bandgap of 1.7eV on glass and flexible polyimide substrates. Cd1-xZnxTe films were deposited by RF magnetron co-sputtering with CdTe and ZnTe sources. By varying the composition of Cd1-xZnxTe being deposited the required bandgap of 1.7eV was achieved. The optical and structural properties of the films were studied with optical transmission, SEM and XRD measurements. The films exhibited high optical transmission and pinhole free grain structure. CZT solar cells were fabricated on glass and flexible polyimide substrate and were analyzed by J-V and spectral response measurements. The effect of post deposition treatments and the effect of N2 during sputtering on CZT device performance were studied.
Thesis:
Thesis (M.S.E.E.)--University of South Florida, 2004.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Jagadish Gaduputi.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 65 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001469440
oclc - 55732477
notis - AJR1194
usfldc doi - E14-SFE0000341
usfldc handle - e14.341
System ID:
SFS0025036:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001469440
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 040524s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000341
035
(OCoLC)55732477
9
AJR1194
b SE
SFE0000341
040
FHM
c FHM
090
TK145
1 100
Gaduputi, Jagadish.
0 245
Characterization of cadmium zinc telluride films and solar cells on glass and flexible substrates by rf sputtering.
h [electronic resource] /
by Jagadish Gaduputi.
260
[Tampa, Fla.] :
University of South Florida,
2004.
502
Thesis (M.S.E.E.)--University of South Florida, 2004.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 65 pages.
520
ABSTRACT: High performance multijunction solar cells based on polycrystalline thin films will require a wide bandgap top cell with at least 15% efficiency. With the bottom cell being CIGS which have already demonstrated the required efficiencies, this work aims to study the complete fabrication and performance of Cd1-xZnxTe solar cells with a bandgap of 1.7eV on glass and flexible polyimide substrates. Cd1-xZnxTe films were deposited by RF magnetron co-sputtering with CdTe and ZnTe sources. By varying the composition of Cd1-xZnxTe being deposited the required bandgap of 1.7eV was achieved. The optical and structural properties of the films were studied with optical transmission, SEM and XRD measurements. The films exhibited high optical transmission and pinhole free grain structure. CZT solar cells were fabricated on glass and flexible polyimide substrate and were analyzed by J-V and spectral response measurements. The effect of post deposition treatments and the effect of N2 during sputtering on CZT device performance were studied.
590
Adviser: Christos S.Ferekides
653
photovoltaics.
tandem devices.
CZT.
polyimide.
690
Dissertations, Academic
z USF
x Electrical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.341



PAGE 1

Characterization Of Cadmium Zinc Telluride Films And Solar Cells On Glass And Flexible Substrates By RF Sputtering by Jagadish Gaduputi A thesis submitted in partial fulfillment of the requirement s for the degree of Master of Science in Electrical Engineering Department of Electrical Engineering College of Engineering University of South Florida Major Professor: Christos S.Ferekides, Ph.D. Don L.Morel, Ph.D. Yun.L.Chiou, Ph.D. Date of Approval: April 1, 2004 Keywords: polyimide, czt, t andem devices, photovoltaics Copyright 2004 Jagadish Gaduputi

PAGE 2

DEDICATION This work is dedicated to my family.

PAGE 3

ACKNOWLEDGMENTS I would like to thank my Major Professor, Dr.Christos S. Ferekides, for his guidance and support during the course of this work. He has been a great source of inspiration during my time at Clean Energy Research Center and will remain so in future. I would like to thank Dr.Don L. Morel and Dr.Yun L. Chiou for their guidance in the fulfillment of this thesis. I would like to thank my colleague and all my friends for their support during this work.

PAGE 4

iTABLE OF CONTENTS LIST OF TABLES iv LIST OF FIGURES v ABSTRACT viii CHAPTER 1.INTRODUCTION 1 1.1 The Path to Renewable Energy 1 1.2 Photovoltaics-An Overview of Progress 2 1.3 Beyond Single Crystal Silicon 4 CHAPTER 2.THEORY OF SOLAR CELLS 7 2.1 Solar Power 7 2.2 Solar Cell Parameters 8 CHAPTER 3.LITERATURE REVIEW 12 3.1 Why Tandem? 12 3.2 MBE Grown Cd1-xZnxTe Solar Cells 14 3.3 Effect of Pre-contact Treatments on Cd1-xZnxTe Films 16 3.4 Contact Technology 17

PAGE 5

ii 3.5 Thin Films on Flexible Polyimide Substrates 18 CHAPTER 4.PROCESSING 21 4.1 Cell Structure 21 4.2 DepositionTechnique 22 4.2.1 Transparent Conducting Oxides 22 4.2.2 Window Layer 22 4.2.3 Cadmium Zinc Telluride 23 4.2.4 Back Contact 24 4.3 Measurements 24 4.3.1 Spectral Response 24 4.3.2 I-V 25 4.3.3 SEM and XRD 25 CHAPTER 5.RESULTS AND DISCUSSION 26 5.1 CZT Material Properties 26 5.1.1 Optical Properties 26 5.1.2 Structural Properties 27 5.2 CZT Solar Cells 30 5.2.1 Effect of N2 on CZT Device Performance 30 5.2.2 ITO/ZnSe/CZT 33 5.2.3 SnO2/CZTand SnO2/CdS/CZT 35 5.2.4 ZnTe Contact 40 5.2.5 Post Deposition Treatment on CZT Films 42 5.3 CZT Thin Films on Flexible Polyimide Substrates 44

PAGE 6

iii5.3.1 Effect of Temperature on Polyimide Substrate 44 5.3.2 TCO Deposition on Polyimide 45 5.3.3 CZT Films on Polyimide 46 CHAPTER 6.CONCLUSIONS 50 REFERENCES 52

PAGE 7

ivLIST OF TABLES Table 1. Conditions and Results of 15 Minute Pre-contact Treatments on CZT 16 Table 2. Calibrated Substrate Temperatures 24 Table 3. Comparision of As Deposited and Annealed CZT Devices on SnO2 and SnO2 /CdS Substrates 36

PAGE 8

vLIST OF FIGURES Figure 1. World Energy Related Carbon Dioxide Emissions,1970-2020 2 Figure 2. PV Roadmap Goal 4 Figure 3. Best Laboratory Cell Efficiencies for Thin Films 5 Figure 4. Solar Spectrum 7 Figure 5. Equivalent Circuit of a Solar Cell 9 Figure 6. Dark and Light I-V Characterstics of a Solar Cell 10 Figure 7. Contour Map of Conversion Efficiency of a Tandem Device 13 Figure 8. Four-Terminal Tandem Solar Cell Structure 14 Figure 9. XRD of MBE Grown CZT 15 Figure 10. SR of n-p and n-i-p CZT Solar Cells 15 Figure 11. XRD(Left) and Transmission(Right) of ITO on Polyimide and Glass 19 Figure 12. SEM of As Deposited(Left) and CdCl2 Treated(Right) CdTe\CdS\TCO\Poly 20 Figure 13. I-V Characterstics of CdTe Solar Cell on Polyimide 20 Figure 14. CZT Solar Cell Structure on Glass(Top) and Flexible Polyimide Substrate(Bottom) 21 Figure 15. Co-sputtering Chamber for CZT Deposition 23 Figure 16. Optical Transmission of CZT 26 Figure 17. Measured Bandgap of CZT 27 Figure 18. XRD of As Deposited and Annealed CZT Films 28

PAGE 9

viFigure 19. SEM of As Deposited(Top) and Annealed(Bottom) CZT Films on Glass/SnO2 Substrate 29 Figure 20. Effect of N2 on Voc and Jsc for As Deposited SnO2/CZT Solar Cells 31 Figure 21. Effect of N2 on Voc and Jsc for Annealed SnO2/CZT Solar Cells 31 Figure 22. Effect of N2 on Voc and Jsc for As Deposited SnO2/CdS/CZT Solar Cells 32 Figure 23. Effect of N2 on Voc and Jsc for Annealed SnO2/CdS/CZT Solar Cells 32 Figure 24. Effect of Annealing on ITO/ZnSe/CZT Solar Cells 34 Figure 25. Spectral Response of Annealed ITO/ZnSe/CZT Devices 35 Figure 26. J-V Data for As Deposited SnO2/CZT Solar Cells 37 Figure 27. J-V Data for Annealed SnO2/CZT Solar Cells 37 Figure 28. J-V Data for As Deposited SnO2/CdS/CZT Solar Cells 38 Figure 29 J-V Data for Annealed SnO2/CdS/CZT Solar Cells 38 Figure 30. Spectral Response of As Deposited and Annealed SnO2/CZT Solar Cells 39 Figure 31. Spectral Response of As Deposited and Annealed SnO2/CdS/CZT Solar Cells 39 Figure 32. J-V Data for As Deposited SnO2/CdS/CZT/ZnTe Solar Cells 40 Figure 33. J-V Data for Annealed SnO2/CdS/CZT/ZnTe Solar Cells 41 Figure 34. Spectral Response of As Deposited SnO2/CdS/CZT/ZnTe Solar Cells 41 Figure 35. Spectral Response of Annealed SnO2/CdS/CZT/ZnTe Solar Cells 42 Figure 36. Transmission of Unprocessed and Annealed Polyimide Substrate 45 Figure 37. Transmission of ITO, ITO/ZnO and ITO/SnO2 Films on Polyimide 46 Figure 38. XRD of CZT on Polyimide 47

PAGE 10

viiFigure 39. SEM Images of CZT Deposited at 250 OC(Top) and 300 OC(Bottom) on Polyimide Substrate 48

PAGE 11

viiiCHARACTERIZATION OF CADMIUM ZINC TELLURIDE FILMS AND SOLAR CELLS ON GLASS AND FLEXIBLE SUBSTRATES BY RF SPUTTERING Jagadish Gaduputi ABSTRACT High performance multijunction solar cells based on polycrystalline thin films will require a wide bandgap top cell with at least 15% efficiency. With the bottom cell being CIGS which have already demonstrated the required efficiencies, this work aims to study the complete fabrication and performance of Cd1-xZnxTe solar cells with a bandgap of 1.7eV on glass and flexible polyimide substrates. Cd1-xZnxTe films were deposited by RF magnetron co-sputtering with CdTe and ZnTe sources. By varying the composition of Cd1-xZnxTe being deposited the required bandgap of 1.7eV was achieved. The optical and structural properties of the films were studied with optical transmission, SEM and XRD measurements. The films exhibited high optical transmission and pinhole free grain structure. CZT solar cells were fabricated on glass and flexible polyimi de substrate and were analyzed by J-V and spectral response measurements. The effect of post deposition treatments and the effect of N2 during sputtering on CZT device performance were studied.

PAGE 12

1CHAPTER 1 INTRODUCTION 1.1 The Path To Renewable Energy As the demand for more electric power increases, the need for an environmentally benign means of producing electricity becomes very important. Today fossil fuels are the primary source to heat, power homes and fuel our cars. It is convenient to use coal, oil, and natural gas for meeting our energy needs, but we have limited supply of these fuels on earth. Sooner or later these currently used power plants will be replaced with renewable energy sources. According to recent predictions the permanent decline in the global oil production rate is expected to start within the next 10-20 years. World-wide oil prices will then rise considerably, favoring the introduction of various renewable energy sources such as the direct conversion of the sunlight into electrical energy using solar cells and other renewable power sources like hydroelectric, wind or geothermal power conversion systems [2]. Even if we had unlimited supply of fossil fuels, using renewable energy is better for the environment as renewable energy technologies are clean and green. Burning fossil fuels sends greenhouse gases into the atmosphere, trapping the suns heat and contributing to global warming. About 20,000 million tons of carbon dioxide is delivered into the atmosphere every year, mainly by burning fossil fuels. Today's plants are unable to absorb this huge amount of

PAGE 13

2CO2. As a result the CO2 concentration in the atmosphere has continued to mount since the industrial revolution as shown in figure 1 [1]. This has added to the greenhouse effect considerably and will increase the global mean surface temperature by another 1 to 5.0 degrees by the year 2100 depending on future emissions and the actual climate sensitivity. This temperature change has already increased the frequency and severity of natural disasters and are likely to have more severe effects for humans and other life Figure 1. World Energy Related Carbon Dioxide Emissions By Fuel Type, 1970-2020 forms on Earth within the next few decades. Fortunately we have renewable energy sources like solar energy which neither run out nor have significant harmful effects on our environment. 1.2 Photovoltaics-An Overview of Progress The term photovoltaic is derived by combining the greek word for light, photos, with volt, the name of the unit of electromotive force. The term photovoltaic therefore signifies the generation of electricity from light. The discovery of Photovoltaic effect is generally credited to French physicist Edmond Becquerel in 1839, but it remained a curiosity of

PAGE 14

3science for the next three quarters of century. Bequeral found that certain materials would produce small amounts of electric currents when exposed to light. The effect was first studied in solids, such as selenium, by Heinrich Hertz in 1870s. Soon afterward selenium photovoltaic cells were converting light to electricity at 1% to 2% efficiency. Selenium was quickly adopted into emerging field of photography for use in light measuring devices [2]. Major steps toward commercializing PV were taken in the 1940s and early 1950s when the Czochralski process for producing highl y pure crystalline silicon was developed. In 1954, scientists at bell laboratories depended on the Czochralski process to develop the first crystalline silicon photovolatic cell, which had an efficiency of 4% [2]. Although a few attempts were made in the 1950s to use silicon solar cells in commercial products, it was the new space program that gave the technology its first major application. In 1958 the US Vanguard space satellite carried a small array of PV cells to power its radio, the cell worked so well that PV technology has been part of space program ever since. Today solar cells power virtually all satellites, including those used for communication satellites, defense and scientific research. Despite these advances, PV in 1970 was still too expensive for most terrestrial uses. In the mid 70s rising energy costs sparked by a world oil crisis renew ed interest in making PV technology more affordable. Since then government, industry and research organizations have invested hundred of millions of dollars in research, development and production. Much of this effort has gone in the development of crystalline silicon the material bell scientists used to make the first solar cells. As a result crystalline silicon have become more and more efficient, reliable and durable. Today’s commerc ial PV systems can convert from 10-15%

PAGE 15

4 Figure 2. PV Roadmap Goal of sunlight into electricity. The cost of PV has dropped down 15-20 fold and price reductions have helped to spur a growing market for photovoltaics. Shipments of PV modules have risen steadily over a past few years. With the manufacturing costs reducing steadily the average annual sales growth has also increased. To expand the markets, and to achieve the PV growth as shown in figure 2, PV costs need to be reduced further which can be achieved through research and development and technology transfer, a path that both industry and federal government are pursuing [3]. 1.3 Beyond Single Crystal Silicon Silicon is not only the material that will respond to sunlight by generating electron hole pairs, many semiconductors have similar pr operties. The silicon band gap of 1.1eV is workable but not ideal. One of the scientific discoveries of the semiconductor industry that has shown great potential for PV industry is thin film technology. Thin films can be made from variety of materials and today’s most important thin film materials are amorphous silicon ( -Si), cadmium telluride (CdTe) and copper indium diselenide (CIS).The best laboratory efficiencies for these three thin films is shown in figure 3 [4].

PAGE 16

5Amorphous silicon is a growing segment in solar market as it can be manufactured at a lower cost than crystalline silicon. Researchers at National Renewable Energy Laboratory demonstrated fabrication of CIS with added gallium devices with efficiencies upto 19% [4]. Industry is exploring a new processing system for large area CIGS deposition. Using the new system, Institute of Energy Conversion (IEC) produced Solar cell that reached 14.9% efficiency. This system was designed to use on inline process and work is going on to manufacture CIGS based solar electric modules [4]. Cadmium Telluride has several features that make it a promising material for solar electric generation. These include the bandgap, which is well matched to solar spectrum and a high absorption coefficient, meaning even one micron thickness is sufficient to produce solar cell efficiency of 10% or higher. To date 10 techniques have been developed to fabricate CdTe solar cells with efficiencies upto 15%-16% [4]. Three of these methods are currently used for module fabrication by industry and module efficiencies of 10-11% have been achieved [4]. Figure 3. Best Laboratory Cell Efficiencies for Thin Films

PAGE 17

6In order to achieve higher efficiencies greater than 25% we need to develop tandem structures with top cell efficiencies up to 16-18%. Cd1-xZnxTe with a tunable bandgap of 1.45-2.2eV is suitable as a top cell with the bottom cell having a bandgap of 1eV. The main objective of this thesis is to study and fabricate high band gap Cd1-xZnxTe solar cells on glass and polyimide substrates by s puttering that could well serve in next generation solar cell technologies.

PAGE 18

7CHAPTER 2 THEORY OF SOLAR CELLS 2.1 Solar Power The energy of the sun, according to current knowledge, is created by the nuclear fusion reaction of hydrogen to helium. This reaction takes place inside the star at several million degrees. The mass reduction occurring in this process is converted to energy. Since all elements are ionized to some degree at this temperature and their spectral lines are strongly broadened, the radiation consists of a multitude of spectral lines, so that the gaseous surface of the sun radiates like a black-body at 6000 K. In space solar radiation is obviously unaffected by earth’s atmosphere and has power density of approximately 1365 W/m2.The characterstic spectral power distribution of solar radiation Figure 4. Solar Spectrum

PAGE 19

8as measured in space is described as Air Mass Zero(AM 0) distribution. At the earth’s surface solar radiation is attenuated selectively at different wavelengths because of absorption by the atmosphere’s constituent gases (oxygen, ozone, water vapor, carbon dioxide, etc). Because of constantly changing position of the sun, the light path through the atmosphere also changes. The spectral power distribution observed when sun’s radiation is coming from an angle to overhead of about 48o corresponds to Air Mass 1.5 distribution which is the standard used for terrestrial applications and its total power density is 844 W/m2 [5]. 2.2 Solar Cell Parameters Photovoltaic energy conversion requires the separation of electrons and holes by an internal electric field. When light made up of photons with energy larger than the bandgap is incident on the junction, it will be absorbed by the semiconductor and electron–hole pairs are created. These electron hole pairs are separated by an internal electric field and this gives rise to phot ocurrent. These charges lower the built-in potential of the junction, Vbi to V bi Voc where Voc is open circuit voltage [5].The generation rate of EHP in the bulk region is determined by the illumination flux and optical absorption constant of the semiconductor. The current voltage (I-V) characterstics of an ideal diode under dark are given by, I=I0 { exp qV/KT-1 } Where I is the external current flow, Io is the reverse saturation current, q is the fundamental electronic charge of 1.602 x 10-19 Couloumb, V is the applied voltage, K is the Boltzmann constant and T is the absolute tempertaure.

PAGE 20

9A non-ideal solar cell can be described with an equivalent-circuit model. The most simple equivalent circuit used consists of a non-ideal diode, representing the p-n junction, coupled with a current source to account for the collection of light-generated carriers. A shunt resistance Rsh and a series resistance Rs are also included in the circuit The complete equivalent-circuit diagram is shown in figure 5. Figure 5. Equivalent Circuit of a Solar Cell The series resistance Rs is the resistance the carriers find on their way and there are many effects that give rise to this resistance: • Low conductivity of the window layer or of the absorber. • Non-ohmic contacts. • Recombination of carriers. The shunt resistance is included to model is due to the effect of low resistivity of some parts of the material, such as grain boundaries or secondary phases and the effect of pinholes. The series resitance can be minimized by reducing the contact resistance and the shunt resistance can be maximized by reducing shunt paths in the junction [6]. Under AM 1.5 illumination the I-V characterstics are simply shifted down due to light generated current IL as shown in figure 6,

PAGE 21

10 Figure 6. Dark and Light I-V Characterstics of a Solar Cell Under illumination the theoretical behaviour is represented by, I=I0 { exp qV/KT-1 } IL The curve describes three important parameters that give complete description of solar cell, open circuit voltage (VOC), short circuit current (ISC) and fill factor (FF). under open circuit conditions, no current flows through the device and VOC is observed at terminals of the junction. This quantity is unaffected by series resistance losses in the cell but is sensitive to shunt losses. Under short circuit conditions the entire light generated current, ISC is seen with near zero voltage. This value is primarily affected by series resistance losses in device. The Parameters described above represent the maximum values that it is possible to measure and their product is ideal power. The fill factor is then defined as the the area of maximum power rectangle to the product of short circuit current and open circuit voltage,

PAGE 22

11Where Im and Vm are the current and voltage for maximum power output. The most important parameter of solar cell is photovolta ic conversion efficiency defined as ratio of output power to input power calculated as, Where Pin is the incident light power.

PAGE 23

12CHAPTER 3 LITERATURE REVIEW 3.1 Why Tandem? The Bandgap of 1.45eV of CdTe films is nearly optimum for a single junction high efiiciency solar cell. Polycrystalline thin film heterojunction solar cells fabricated from CdTe films have achieved efficiencies upto 16% [7]. However we need efficiencies >25%, so that the module efficiencies of 15% or higher can be achieved by which low cost electricity can be provided. This can be realized with a tandem structure which has a bandgap of 1.7eV and 1.0 eV for the upper and lower cells respectively. Tandem structures can be two terminal or four terminal. The disadvantage of two terminal tandem structure is current matching between series connected top and bottom cell. The bandgaps are dictated by the requirement that the photocurrents from each cell must be matched to achieve maximum efficiency. If not, the cascade current will be limited by the cell generating the lowest current, and the efficiency will fall [8]. Temperature variations also play a important role in limiting the efficiency and the choice of optimum energy gap combination of four terminal or two terminal structures [9]. An alternative better design compared to two terminal tandem structure is a multijunction device with independent terminals for each cell. For this structure the ranges of acceptable bandgaps that produce the higher efficiency are considerably broader. The contour map of conversion efficiency of a-Si /c-Si four terminal tandem solar cell is

PAGE 24

13shown in figure 7 [10]. As shown in figure, the maximum efficiency exceeding 30% appears for a bandgap combination of 2.1/1.5 eV and 2.3/1.4eV. For a bandgap combination of 1.75/1.2ev, which corresponds to a-Si/c-Si tandem device, more than 28% efficiency can be seen. Figure 7.Contour Map of Conversion Efficiency of a Tandem Device Cd1-xZnxTe is the one choice for the top cell in four terminal tandem structures as its bandgap can be varied between 1.45-2.2eV by varying the Composition value. While the terenaries are complex they offer flexibility of tuning the bandgap to bottom cell.. Using Penn State EPRI AMPS code for tandem simulations an efficiency more than 25% was predicted for Cd1-xZnxTe/CIGS tandem structure [11]. A four terminal tandem structure with with CIGS as bottom cell and Cd1-xZnxTe as top cell is shown in the figure 8.

PAGE 25

14 Figure 8. Four Terminal Tandem Solar Cell Structure 3.2 MBE Grown Cd1-xZnxTe Solar Cells The terenary films of Cd1-xZnxTe were grown by MBE (Molecular beam epitaxy) on Glass/SnO2/CdS Substrates with a target bandgap of 1.7-1.8eV. n-i-p Cd1-xZnxTe solar cells had better efficiencies compared to n-p cells. The optimum temperature was maintained at less than 400 OC. The substrates were baked out 250 OC for 3-4 hours with growth rate of 1 m/hr. Before film growth the substrate temperature was kept at 275OC for 30minutes to commence film growth and increased to 300 OC for remaining run. Bandgap, compositional uniformity were determined by XRD technique. From the x-ray plot shown in figure 9, it is clear that there are no other mixed phase present in the Cd1-xZnxTe films since only peaks corresponding to Cd1-xZnxTe are observed. The lattice constant of Cd1-xZnxTe film was determined by extrapolation method used to estimate zinc content in films according to a(x)=6.481-0.381x [7] The zinc concentration was 0.4 corresponding to a bandgap of 1.7eV.

PAGE 26

15 Figure 9. XRD of MBE Grown CZT Both n-p and n-i-p Cd1-xZnxTe devices were fabricated, for n-i-p p-type ZnTe was deposited before contact metallization. The n-i-p spectral response was higher over most of the spectral range as shown in figure 11. Figure 10. Spectral Response of n-p and n-i-p CZT Solar Cells From spectral response we can see that bandgap was reduced from 1.7eV to 1.55eV after processing. This bandgap shift can be attributed to loss of zinc from its substituitional sites in the film. Auger depth profile analysis on this film after processing indicated a non uniform distribution of Te throughout the film with some zinc segregation to the ZnTe/Cd1-xZnxTe interface within the bulk film. This possibly indicates that Zn has

PAGE 27

16left its lattice sites in Cd1-xZnxTe and may be forming other defect complexes which result in lowering of bandgap. 3.3 Effect of Pre-c ontact Treatments on Cd1-xZnxTe Films Effect of CdCl2 and ZnCl2 treatment were observed in single phase Cd1-xZnxTe alloy films deposited by a co-evaporation from bi nary CdTe and ZnTe sources of thickness 34 m with bandgap varying between 1.5-2.25eV [12]. Efforts to directly implement CdCl2 treatment on Cd1-xZnxTe films similar to CdTe films did not work out due to high reactivity of ZnTe with CdCl2. Cd1-xZnxTe was fabricated on 7059/ITO/CdS substrates and XRD was used to determine phase changes and alloy distribution in the films after treatments in argon, dry air, CdCl2 vapor + dry air and ZnCl2 vapor +dry air which are summarized below. Table 1. Conditions and Results of 15minute Pre-Contact Treatments on CZT

PAGE 28

17From above results table we can infer that Treatment in argon at 600 OC only causes primary recrystallization. Treatment in dry air preferentially removes ZnTe from the alloy as oxides. Treatment in ZnCl2:O2:Ar at low pressure, preserves the bulk alloy composition. Treatment in ZnCl2:O2:Ar at high pressure, removes ZnTe from alloy as ZnO and Te. Treatment in CdCl2:O2:Ar removes ZnTe from the alloy forming CdTe ,ZnO, Te and volatile ZnCl2. Devices were fabricated similar to CdTe/CdS but utilizing selected post deposition treatments. Back contact was formed by s equential deposition of Te and cu layers by electron beam evaporation, application of graphite ink and heat treatment at 100 OC for 30min. However these devices exhibited significant dark-light crossover and low Jsc. These properties suggest poor conductivity and poor carrier transport in the absorber layer. Spectral response measurements of ZnCl2 treated devices show low collection and long wavelength fall off at 720nm corresponding to as-deposited bandgap of 1.7eV. 3.4 Contact Technology The formation of a stable, reproducible, low resistance contact to pCd1-xZnxTe is one of the major problems in fabrication of efficient solar cells. The reason for this is high work function of Cadmium zinc Telluride. CZT with a bandgap of 1.7eV has a work function of ~5.2eV [13] No metals present such a high value and thus most of the contacts are rectifying. Metals like platinum satisfy the condition for ohmic contacts on p-type CZT, however the stability of the contact for solar cell purposes is questionable. Electrical properties of contacts on p-type Cd0.8Zn0.2Te were studied. Au, Al, In and electroless Au

PAGE 29

18were used for this purpose. Results showed that electroless Au films deposited by chemical method can form a heavily doped p+ layer on Cd0.8Zn0.2Te smooth surface which is nearly ohmic on p-type material. The CZT films were precontact treated with bromine methanol (BM etchant) followed by lactic in ethylene glycol(LB etchant). Post annealing treatments were also found to improve ohmic quality and adhesion of contact layers to Cd0.8Zn0.2Te surface [14]. ZnTe:Cu contact is one of the contact that has been successfully used in CdTe/CdS solar cells [15]. Research has shown that reproducible p-type doping of ZnTe can be achieved by reactive RF magnetron sputtering with Ar/N2 gas mixtures eliminating the need for Cu doping [16]. In our case ZnTe was deposited in RF magnetron sputtering chamber with Ar/N2 gas mixture and doped graphite was used as metal contact. 3.5 Thin Films on Flexi ble Polyimide Substrates In recent years there has been an increased interest in photovoltaic structures on light weight flexible substrates. Photovoltaic structures on lightweight substrates have several advantages over the heavy glass based structures in both terrestrial and space applications [17]. They are important for novel added products such as portable and lightweight sources of power for emergencies and recreational use, PV integrated buildings, Solar boats and cars, smart cards, data and telecommunication products [18]. CdTe Solar cells with a efficiency of 8.6% [18] and CIGS solar cells with efficiency of 11.3% [20] were fabricated on polyimide substrates for both terrestrial and space applications. A lot of work has been done on TCO’s such as ITO and ZnO:Al deposited on polyimide which have good optoelectronic properties [19]. Reported difficulties while processing with polyimide substrates in clude low processing temperatures and poor adhesion of films.

PAGE 30

19ITO films were deposited on Kapton polyimide by RF sputtering at room temperature. Before the deposition process plasma etching on the surface of the substrates was done with low N2 flows and low power of 1W/cm2 to improve adhesion of ITO on polyimide [19].The thickness on the glass reference sample was around 500nm. From the XRD data as shown in figure 11, we can see that ITO polycrystalline films are grown with preferred (222) orientation. The average transmission of ITO on polyimide was found to be 72.5% as seen in figure 11 with sheet resistance of about 14 / Figure 11. XRD(Left) and Transmission(Right) of ITO on Polyimide and Glass CdTe solar cells were fabricated on polyimide substrates, but instead of commercially available polyimide, a specific type of pol yimide was prepared in house [18]. One of the advantages of this is that the thickness of polyimide film can be minimized thereby reducing the absorption loss in the substrate. A thin buffer layer was evaporated onto the glass substrate, then a polyimide layer was spin coated and cured at 430 OC. CdTe/CdS solar cells were fabricated on this spin coated polyimide substrate with a process in which all layers were grown by evaporation method. After complete processing the entire stack was separated from glass by dissolving NaCl buffer layer in water. The structural

PAGE 31

20properties of CdTe on polyimide were similar to that of glass substrates. As shown in figure 12 as deposited CdTe layer is compact with grain size upto ~1m. Figure 12. SEM of As deposited (Left) and CdCl2 Treated (Right) CdTe/CdS/O/Poly The CdCl2 treated layers are crack free with grain size upto ~5 m .The as deposited layers had (111) orientation, but loss in orientation was observed in CdCl2 annealed samples [21]. The I-V Characterstics of CdS/CdTe solar cell on polyimide is shown in figure 13. Efficiencies upto 8% were obtained on polyimide substrate. The lower efficiencies are attributed to high resistivity of TCO ZnO:Al films which do not yield efficient CdTe/CdS solar cells. Higher efficiencies can be expected if the same process is done on ITO or SnO2 front contacts[18]. Figure 13. I-V Characterstics of CdTe Solar Cell on Polyimide

PAGE 32

21CHAPTER 4 PROCESSING 4.1 Cell Structure High performance multijunction tandem solar cells based on polycrystalline thin films will require a wide band gap top cell with atleast 15% efficiency. This work aims to study the feasibility of Cd1-xZnxTe solar cell as the top cell with a band gap of 1.7eV on glass and polyimide substrates. The basic cell structure of CZT solar cell on glass and polyimide substrate is shown in figure 14. The processing of each layer is described in the later part of the chapter. Figure 14. CZT Solar Cell Structure on Glass (top) and Flexible Polyimide Substrate (bottom) 7059 Glass TCO Window Layer CZT Back contact In Pol y imide TCO Window La y e r CZT Back contact

PAGE 33

224.2 Deposition Technique 4.2.1 Transparent Conducting Oxides Transparent conducting oxide (TCO) acts as a Front Contact for a solar cell. Ideally for a solar cell the TCO should have low resistivity and optical transmission greater than 90% [22]. Tin Oxide(SnO2 ) or Indium Tin Oxide (ITO) were used as transparent conducting oxides for our experiments. SnO2 was deposited by MOCVD technique with O2 and Tetramethyltin (Tin source). High purity helium is used as a carrier gas. The thickness of this film is ~ 700-800 Indium tin oxide(ITO) is deposited by a 3” RF magnetron sputtering with Ar as plasma source. The total pressure was maintained at 2-3 x 10-3 torr at 200-300 OC substrate temperature.The thickness of these films is about 1500-2000 Both the films had resistivity less than 2 x10-4 -cm and transmission greater than 90%. 4.2.2 Window Layer Cadmium sulphide (CdS) or Zinc selenide (ZnSe) were used as a window layers to form the junction with p-type absorber layer. CdS has a proven record of highest efficiencies with CdTe solar cells. CdS in this case was deposited by chemical bath deposition with a cadmium salt ,ammonium salt, ammonia and thiourea as the reactants used in aqueous solution. The amount of ammonia and ammonium salt can be used to control CdS deposition rate [23].The solution was heated to 80-85 OC during the process and final thickness obtained is < 1200 ZnSe is deposited in a 3” RF magnetron sputtering chamber with Ar as plasma source. The substrate is heated upto 250-300 OC and pressure is kept at 4-5 x 10-3 torr. The thickness is varied between 500-1000 in different cases.

PAGE 34

234.2.3 Cadmium Zinc Telluride Cadmium Zinc Telluride (CZT) is the absorber layer. CZT was deposited in a RF magnetron co-sputtering chamber with 99.999% purity CdTe and ZnTe sources. For high efficiency tandem solar cell we need a top cell with a bandgap of 1.7eV.This can be achieved by varying the composition of CdTe and ZnTe, so that bandgap can be varied from 1.45ev-2.2eV. The set up for depositing CZT is shown in figure 15. A Varian CRYOSTACK-8 Cryo pump is used to evacuate the chamber to a base pressure in the range of 10-7 torr and sputtering pressure is maintained between 4-5 x 10-3 torr with Ar/N2 mixture as the plasma source. The thickness on the substrate was monitored using crystal monitor. Figure 15. Co-sputtering Chamber for CZT Deposition Thickness calibration and temperature calibration were done before starting actual experiments. A THERMCOAX heater is used as substrate heating element. The

PAGE 35

24calibrated temperatures for substrate are shown in table 2. Substrate holder is rotated during deposition to get uniform heating on the substrate and uniform deposition. The thickness of the films is maintained between 4-5 m. Table 2. Calibrated Substrate Temperatures Heater Temperature (OC) Substrate temperature (OC) 600 400 500 300 420 200 4.2.4 Back Contact Graphite paste doped with HgTe:Cu is one of the contact used for CZT solar cells. ZnTe is the other back contact that was used. For our experiments ZnTe is deposited in RF magnetron sputtering chamber with Ar/N2 gas mixture and doped graphite is used as metal contact. 4.3 Measurements 4.3.1 Spectral Response The quantum efficiency is the measure of number of electron-hole pairs generated and collected for every incident photon absorbed. The external quantum efficiency of the device is determined and plotted as function of wavelength. A spectrometer is used which has a grating. By varying the angle of grating to incident light, monochromatic light of different wavelengths can be produced. The source is calibrated from silicon standard cell obtained from the National Renewable Energy Laboratory (NREL). The output Jsc, of

PAGE 36

25the cell at each wavelength is normalized against the same for the Si standard cell. The plot of external quantum efficiency versus wavelength is plotted. 4.3.2 I-V All the devices were characterized by light and dark current-voltage(I-V) measurements and the main device parameters like Voc, Jsc, FF and efficiency were determined. Light I-V measurements were taken with a co mputer controlled Keithley source meter. Light from four OSRAM halogen lamps simulated the AM !.5 solar spectrum and provided uniform distribution over an area large enough to place a sample (4x4) cm. A labview program runs in the computer which is used to control sweep voltage across the device. Once the currents are logged the program calculates main device parameters and characterstic dark and light I-V curves are plotted. 4.3.3 SEM and XRD The structural properties of CZT films were characterized by X-ray diffraction (XRD) and Scanning electron microscopy (SEM). XRD measurements were done using Philips X’pert PRO X-ray diffraction system and SEM micrographs were taken on a Hitachi 800 SEM system.

PAGE 37

26CHAPTER 5 RESULTS AND DISCUSSION 5.1 CZT Material Properties 5.1.1 Optical Properties The CZT films optical properties were charac terized by transmission measurements. As discussed in previous chapter CZT was deposited in a co-sputtering chamber with Ar/N2 mixture as the plasma source. The deposition was carried out with a total pressure of 4-5 x 10-3 torr and substrate temperature in the 300-400 OC range. Initial experimentswere carried out to obtain a bandgap of 1.7eV. Transmission and bandgap 0 10 20 30 40 50 60 70 80 90 100 500550600650700750800850900950100010501100 Wavelength(nm)Tr(% ) Figure 16. Optical Transmission of CZT

PAGE 38

27measurements were done for each substrate. The transmission of CZT on Glass/SnO2 substrate was around 70% as shown in figure 16. The optical bandgap of a material can be found from transmission spectra if the film thickness is known. Since = 0(E Eg)1/2 where Eg is the bandgap, E is the energy of incident photons, 0 is the constant and and is the absorption co-efficient [25]. Then a plot of 2 Vs photon energy(E) should yield a line with an intercept equal to Eg which shown in figure 17. The bandgap deduced from the plot is 1.7eV which is the desired bandgap for our experiments. 0.E+00 9.E+07 2.E+08 3.E+08 4.E+08 5.E+08 11.11.21.31.41.51.61.71.81.922.12.22.32.4 Photon Energy(eV)( )2 cm-2 Figure 17. Measured Bandgap of CZT 5.1.2 Structural Properties The structural properties of CZT films were characterized by using X-ray diffraction and scanning electron microscopy. For this purpose Glass/SnO2/CZT substrates were used. CZT was deposited with Ar/N2 mixture as plasma source at 400 OC and one set of samples were annealed in helium ambient so that measurements could be carried on

PAGE 39

28as-deposited and annealed samples. XRD measurements were done on the asdeposited and annealed SnO2/CZT samples as shown in figure 18 to determine film Figure 18. XRD of As Deposited and Annealed CZT Films composition, orientation and possible formation of mixed phase. The as-deposited CZT films exhibit strong (111) texture. The zinc concentration in Cd1-xZnxTe was determined using the lattice constant relation a(x)=6.481-0.381x .The value of x is ~0.4 which is consistant with measured optical bandgap of 1.7eV After annealing the intensity of (111) reflection increased and became sharper than as deposited films, all other reflections also sharpened with increased peak intensity which indicates annealing improved crystallinity of CZT films. It can be seen that both as-deposited and annealed films had CdTe peaks which gives hint of mixed phase.

PAGE 40

29 Figure 19. SEM of As Deposited(Top) and Annealed (Bottom) CZT Films on Glass/SnO2 Substrate

PAGE 41

30SEM images for as deposited and annealed CZT f ilms are shown in figure 19. From the results it can be seen that both as deposited and annealed films had closely packed dense structure with large number of small grains and SEM analysis did not reveal significant grain growth after anneal, which indicates anneal induced primary recrystallization without changing the alloy composition of the films. 5.2 CZT Solar Cells 5.2.1 Effect of N2 on CZT Device Performance Nitrogen doping of II-VI compounds has been successfully reported. High doping levels can been achieved with ZnTe and slightly lower for CdTe, but very low doping levels have been achieved with terenary compounds such as CdZnTe and CdMgTe, which might be due to formation of nitrides which act as a limiting factor [24]. After initial work on CZT films deposited with 50:50 Ar/N2 mixture, our further work was concentrated on the varying N2 partial pressure during CZT deposition and its effect on the CZT device performance. A series of experiments were conducted with varying N2 partial pressure. These experiments were conducted on SnO2/CdS and SnO2 substrates. The total pressure was maintained at 4-5 x 10-3 torr with Ar/N2 mixture. The deposition was done at 300-400 OC and the N2 partial pressure varied by 10%, 25%, 50%, and 75% of the total pressure. The devices were grouped into as-deposited and annealed. Annealing was carried out at 500 OC in Helium ambient and doped graphite was used as back contact for all these devices. The results are summarized below.

PAGE 42

31 0 50 100 150 200 250 300 01020304050607080 N2(%)Voc(mV)0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2Jsc(mA/cm2) Voc Jsc Figure 20. Effect of N2 on Voc and J sc for As Deposited SnO2/CZT Solar Cells 0 50 100 150 200 250 300 01020304050607080 N2(%)Voc(mV)0 0.5 1 1.5 2 2.5Jsc(mA/cm2) Voc Jsc Figure 21. Effect of N2 on Voc and J sc for Annealed SnO2/CZT Solar Cells

PAGE 43

32 0 50 100 150 200 250 300 350 400 450 01020304050607080 N2(%)Voc(mV)0 0.5 1 1.5 2 2.5Jsc(mA/cm2) Voc Jsc Figure 22. Effect of N2 on Voc and Jsc for As Deposited SnO2/CdS/CZT Solar Cells 0 50 100 150 200 250 300 350 400 450 500 550 01020304050607080 N2(%)Voc(mV)0 0.5 1 1.5 2 2.5Jsc(mA/cm2) Voc Jsc Figure 23. Effect of N2 on Voc and Jsc for Annealed SnO2/CdS/CZT Solar Cells

PAGE 44

33From the above results it can be seen that N2 concentration increases both Voc and Jsc values in SnO2/CdS/CZT and SnO2/CZT devices up to 50% N2 partial pressure beyond which the CZT device performance degrades. This effect of N2 is not exactly known but it can be speculated that N2 improves electronic properties of CZT films eventhough the mechanism of which is not yet known. It can also be seen that annealing of CZT films helps to improve the performance of device irrespective of N2 concentration. 5.2.2 ITO/ZnSe/CZT After establishing deposition conditions for CZT with bandgap of 1.7eV, CZT solar cells were fabricated with ITO/ZnSe substrates. Both ITO and ZnSe were deposited by RF magnetron sputtering with Ar ambient at 200 -250 OC. The thickness of ITO was 10002000 and ZnSe was 500-1000 CZT was deposited with Ar/N2 mixture as plasma source. The total pressure was 4-5 x 10-3 torr with N2 partial pressure of 2.5 x 10-3 torr. The processing was categorized into two groups, (a) as-deposited. (b) annealed. The samples were contacted with doped graphite. The films were annealed at different temperatures in Helium ambient to study their performance. The Voc and Jsc values of as deposited and cells annealed in Helium at 450 OC and 500 OC are shown in figure 24. The fill factors of all the devices were less than 30%. Efforts to anneal at temperatures greater than 500 OC did not yield better results as the CZT films were evaporating.

PAGE 45

34 0 100 200 300 400 500 600 700 0450500 Annealing Temperature(oC)Voc(mV)0 0.02 0.04 0.06 0.08 0.1 0.12Jsc(mA/cm2) Voc Jsc Figure 24. Effect of Annealing on ITO/ZnSe/CZT Solar Cells The cells annealed at 500 OC exhibited better characterstics compared to other annealing temperatures. High Voc‘s were obtained for the films annealed at 500 OC but currents were less than 1mA/cm2. From the high Voc values it can be speculated that annealing improves junction properties to some extent. The spectral response for the devices annealed at 500 OC is shown in figure 25, from which it can be seen that low collection in the 550-700nm region is the main cause for currents to be low The Jsc of the devices calculated from the SR was less than 1mA/cm2 which is well below the ideal value based on CZT bandgap which is around 22 mA/cm2.

PAGE 46

35 0% 2% 4% 6% 8% 10% 12% 14% 400500600700800900 Wavelength, (nm)Q.E. Figure 25. Spectral Response of Annealed ITO/ZnSe/CZT Devices 5.2.3 SnO2/CZT and SnO2/CdS/CZT After studying ZnSe/CZT devices, CZT solar cells were fabricated with SnO2 and SnO2/CdS substrates. CZT was deposited with Ar/N2 mixture as plasma source. The total pressure is 4-5 x 10-3 torr with N2 partial pressure of 2.5 x 10-3 torr. SnO2 was deposited by MOCVD and CdS was deposited by chemical bath deposition. The devices were grouped into as-deposited and annealed. A temperature of 500 OC was chosen as annealing temperature for the devices as films above that temperature were evaporating. The samples were contacted with doped graphite. The results of asdeposited and samples annealed at 500 OC in He ambient are summarized in table 3.

PAGE 47

36Table 3. Comparision of As Deposited and Annealed CZT Devices on SnO2 and SnO2/CdS Substrates Device structure Pre contact condition Voc (mV) Jsc(mA/cm2) SnO2/CZT As deposited 240 1.84 SnO2/CZT Annealed 280 2.36 SnO2/CdS/CZT As deposited 410 1.87 SnO2/CdS/CZT Annealed 510 2.07 From the results it can be seen that currents were comparable in SnO2/CZT and SnO2/CdS/CZT devices, however Voc’s were high in SnO2/CdS/CZT devices. It can also be seen that annealed devices had better Voc’s and Jsc’s compared to as deposited devices in both the device structures. This shows that even though annealing did not increase the grain size of CZT films as shown in SEM micrographs in figure 19, it might have helped to improve some of the interface/junction properties and crystallinity of the films which resulted in high Voc. The J-V and spectral response measurements were done on these devices and are summarized below. From the spectral response results it can be seen that low collection in 550nm-700nm region leads to low currents. The cause for this low collection is not exactly known but it may be due to carrier recombination within the bulk before they reach the junction and proceed to external circuit or may be due to low electric field in the space charge region. Therefore it can be speculated that sputtered CZT films have poor electronic properties. The fill factors of all the devices were less than 30%. From t he J-V curves it can be seen that devices have high series resistance which contributes to low fill factors.

PAGE 48

37 -0.003 -0.002 -0.001 0.000 0.001 0.002 0.003 -1.00-0.75-0.50-0.250.000.250.500.751.00 Voltage [Volts]Current Density [A/cm2] Figure 26. J-V Data for As Deposited SnO2/CZT Solar Cells -0.002 -0.001 0.000 0.001 0.002 -1.00-0.75-0.50-0.250.000.250.500.751.00 Voltage [Volts]Current Density [A/cm2] Figure 27. J-V Data for Annealed SnO2/CZT Solar Cells

PAGE 49

38 -0.003 -0.002 -0.001 0.000 0.001 0.002 0.003 -1.00-0.75-0.50-0.250.000.250.500.751.00 Voltage [Volts]Current Density [A/cm2] Figure 28. J-V Data for As Deposited SnO2/CdS/CZT Solar Cells -0.0009 -0.0006 -0.0003 0.0000 0.0003 0.0006 0.0009 -1.00-0.75-0.50-0.250.000.250.500.751.00 Voltage [Volts]Current Density [A/cm2] Figure 29. J-V Data for Annealed SnO2/CdS/CZT Solar Cells

PAGE 50

39 0% 4% 8% 12% 16% 20% 24% 28% 400500600700800900 Wavelength, (nm)Q.E. As Deposited Annealed at 500c in He Figure 30. Spectral Response of As Deposited and Annealed SnO2/CZT Solar Cells 0% 3% 6% 9% 12% 15%400500600700800900 Wavelength, (nm)Q.E. As Deposited Annealed at 500C in He Figure 31. Spectral Response of As Deposited and Annealed SnO2/CdS/CZT Solar Cells

PAGE 51

405.2.4 ZnTe Contact ZnTe doped with N2 has been successfully fabricated to obtain good ohmic contacts in CdTe/CdS solar cells [15]. For the CZT devices, ZnTe:N2 was deposited on CZT before contacting with doped graphite to see if it can improve contact properties and reduce the series resistance of the device. ZnTe wa s deposited By RF magnetron sputtering in Ar/N2 plasma at 200-250 OC to a thickness of 1000-2000 The J-V curves and spectral response are shown below, from which it can be seen that the performance of the devices is similar to non ZnTe devices and t here was no significant change in either Voc or Jsc from non ZnTe devices. -0.0008 -0.0006 -0.0004 -0.0002 0.0000 0.0002 0.0004 0.0006 0.0008 -1.00-0.75-0.50-0.250.000.250.500.751.00 Voltage [Volts]Current Density [A/cm2] Figure 32. J-V Data for As Deposited SnO2/CdS/CZT/ZnTe Solar Cells

PAGE 52

41 -0.0008 -0.0006 -0.0004 -0.0002 0.0000 0.0002 0.0004 0.0006 0.0008 -1.00-0.75-0.50-0.250.000.250.500.751.00 Voltage [Volts]Current Density [A/cm2] Figure 33. J-V Data for Annealed SnO2/CdS/CZT/ZnTe Solar Cells 0% 2% 4% 6% 8% 10% 12% 14% 400500600700800900 Wavelength, (nm)Q.E. Figure 34. Spectral Response of As Deposited SnO2/CdS/CZT/ZnTe Solar Cells

PAGE 53

42 0% 3% 6% 9% 12% 15% 400500600700800900 Wavelength, (nm)Q.E. Figure 35. Spectral Response of Annealed SnO2/CdS/CZT/ZnTe Solar Cells 5.2.5 Post Deposition Treatment on CZT Films Efforts to utilize chloride treatments on Cd1-xZnxTe similar to CdCl2 treatment on CdTe have not been successful. This is due to chemical reactivity of ZnTe with CdCl2 resulting in a reaction yielding volatile ZnCl2 and excess CdTe. However ZnCl2 will not react appreciably with CdTe which is of great interest [12]. ZnCl2 treatment was done prior to contacting of the Cd1-xZnxTe devices to see if it enhances some of the electronic properties of the Cd1-xZnxTe material. Initial experiments were done using ~ 0.2 mol ZnCl2 solution. CZT deposited on SnO2/CdS substrates were dipped in the solution for 4-5minutes and these when taken out of the solution were air annealed at 350-400 OC for 10min.This experiment resulted in better currents upto 4.5 mA/cm2 This might have been due to removal of ZnTe from the alloy as ZnO and Te [12].

PAGE 54

43 This solution treatment was later replaced by ZnCl2 vapor treatment. The samples were kept above ZnCl2 powder placed in a graphite boat and the whole set up was heated in a oven at different temperature so that CZT was exposed to ZnCl2 vapors. This experiment resulted in better results without causing the loss of Zinc from the CZT bulk as compared to solution treatment. However further studies are needed in order to completely understand the effect of ZnCl2 on CdS/Cd1-xZnxTe junction and also on electronic properties of Cd1-xZnxTe films [26].

PAGE 55

445.3 CZT Thin Films on Flexi ble Polyimide Substrates After partially optimizing the deposition conditions for CZT on glass substrate, further work was concentrated on fabricating CZT solar cells on flexible substrates. Lightweight flexible polyimide substrates were used for this purpose. Flexible substrates have several advantages over the heavy glass based structures in particular for space applications. Although flexible substrates have advantages over glass substrate, they present certain process challenges as they cannot be exposed to high temperatures and and also adhesion of films is one of the issue, which in turn reduce the device performance. 5.3.1 Effect of Temperat ure on Polyimide Substrate For this work commercially available KAPTON polyimide was used as a substrate which could withstand upto 400 OC. initial experiments were conducted on polyimide to find the effect of temperature on this substrate. Three substrates were heated at three different temperatures 375 OC, 400 OC, 425 OC and transmission measurements were done on these samples. The only noticeable change was observed for substrate heated at 425 OC which had some physical damage visible to naked eye. The transmission measurements are shown in the figure 37, from which we can see that cutoff is around 450nm which is due to absorption in the polyimide substrate and it can be seen that the transmission of the annealed substrates remains same in most of the region except in the infrared region it increases as the temperature increases, which may just be due to transmission error.

PAGE 56

45 0 10 20 30 40 50 60 70 80 90 100 40050060070080090010001100 Wavelength(nm)Tr(% ) Polyimide 375 C 400 C 425 C Figure 36. Transmission of Unprocessed and Annealed Polyimide Substrate 5.3.2 TCO Deposition on Polyimide Indium tin oxide (ITO) films were deposited on polyimide substrates by RF magnetron sputtering at 200-250 OC with Ar and 5% O2 as plasma source. The total pressure was in the range of 2-3 x10-3 torr. The thickness of ITO was ~3000 The sheet resistance of this film was 35 / SnO2 or ZnO were also deposited as bi-layers in different cases on polyimide/ITO substrates in the same chamber without breaking the vaccum to find the adhesion and transmission properties of these films on polyimide. ITO in this case was deposited with Ar only. The thickness of SnO2 and ZnO was at ~1000 and total thickness was maintained at ~3000 The sheet resistance of these films was 7-8 / From the transmission response shown in figure 38, it can be seen that ITO deposited with Ar only results in low optical transmission in infrared region which may

PAGE 57

46be due to free carrier absorption because of highly conductive films. The transmission of ITO deposited with Ar/O2 plasma source on polyimide is greater than 90% in all the regions which is due to presence of oxygen during deposition which increases the resistivity of films. The adhesion of film s on polyimide was good and they did not show any physical damage at 200-250 OC but substrate would curl after heat treatment which was natural for plastics. 0 10 20 30 40 50 60 70 80 90 100 400500600700800900100011001200 Wavelength(nm)Transmission(% ) ITO ITO/SnO2 ITO/ZnO Figure 37. Transmission of ITO, ITO/ZnO and ITO/SnO2 Films on Polyimide 5.3.3 CZT Films on Polyimide Cadmium Zinc Telluride (Cd1-xZnxTe) was deposited on ITO/CdS, ITO/ZnO/CdS and ITO/SnO2/CdS substrates. CdS was grown by chemical bath depositon. With CZT problems with adhesion existed which at this time is believed to be due to stress in the CZT films. Further studies are needed to overcome these CZT adhesion problems. However we were able to get better films on some of the substrates. Bandgap,

PAGE 58

47compositional uniformity and grain structure of CZT on polyimide were studied using XRD and SEM images. With XRD data the zinc concentration in Cd1-xZnxTe was determined using the lattice constant relation a(x)=6.481-0.381x .The value of x was ~0.4 corresponding to a bandgap of 1.7eV. 0 50 100 150 200 250 300 350 400 450 500 20304050607080 2 thetaCounts CdTe ZnTe CdZnTe Figure 38. XRD of CZT on Polyimide From the XRD data we can see that CZT films have preferential (111) orientation. It can also be seen that only peaks corresponding to CZT were observed and no secondary phase can be seen. This may be due to low processing temperature of CZT on polyimide which does not exceed 300 OC. Previously CdTe peaks were seen in XRD of CZT on glass substrate which had deposition temperatures up to 400 OC. This suggests lower substrate temperatures are favorable for single phase CZT films.

PAGE 59

48 Figure 39. SEM Images of CZT Deposited at 250 OC (Top) and 300 OC (Bottom) on Polyimide Substrate

PAGE 60

49From the SEM data shown in figure 40, it can be seen that CZT deposited on polyimide at 250 OC is less dense and grain structure looks li ke clusters of small particles. As the deposition temperature is increased to 300 OC, it can be seen that more closely packed dense films are obtained similar to that on glass substrates which were deposited at 400 OC as shown in figure 19. CZT solar cells were fabricated on these substrates and were contacted with doped graphite. J-V characterstics for these devices were measured. The cells exhibited Voc’s up to 100mV and currents were less than 1mA/cm2 due to excess series resistance. Nevertheless further studies are needed on flexible substrates which provide an interesting alternative to glass substrate in both terrestrial and space applications.

PAGE 61

50CHAPTER 6 CONCLUSIONS Cadmium Zinc Telluride films have been succe ssfully deposited by co-sputtering under low temperature conditions on glass and Polyimide substrates. The effects of processing and post deposition heat treatments on CZT fi lms structural and optical properties have been studied and several observations were made. 1) Heat treatments improved crystallinity of the CZT films as indicated by the increased intensity of the XRD peaks. 2) Higher deposition temperature leads to secondary phase in CZT films on glass substrates, however single phase CZT films can be fabricated through lower deposition temperatures (Polyimide). 3) Due to the low processing temperatures used with polyimide substrates CZT films on this type of substrate were less dense with clusters of small particles. 4) Heat treatments did not induce significant grain growth but improved the device performance in terms of Voc and Jsc when compared to as-deposited devices. The influence of N2 during the deposition of Cd1-xZnxTe on the device performance was studied. Films deposited with Ar and 50% N2 as partial pressure exhibited better performance. Voc‘s up to 670 mV and Jsc‘s up to 2.5 mA/cm2 were obtained with these deposition conditions. The use of ZnTe:N2 as the back contact did not improve the performance of the devices to large extent.

PAGE 62

51Post deposition ZnCl2 treatment was done on Cd1-xZnxTe films similar to CdCl2 treatment in high efficiency CdTe solar cells. ZnCl2 Vapor treatment did improve the device performance to certain extent with out change in bulk composition, but further studies are needed in order to completely understand the effect of ZnCl2 on Cd1-xZnxTe films. Cd1-xZnxTe solar cells were fabricated for the first time on flexible polyimide substrates. Commercially available KAPTON polyimide was used for our experiments which could withstand up to 400 OC. Devices made out of these films had low Voc‘s and Jsc‘s due to high series resistance. Further studies are needed to completely study the device performance on these flexible polyimide substrates. The low currents in the CZT devices can be attributed to low collection in long wavelength region. This may be due to poor interface/junction properties or poor electronic properties of CZT.

PAGE 63

52REFERENCES [1] “International Energy Outlook 2003”, Energy Information Administration, US Department of Energy, 2003. [http://www.eia.doe.gov/oiaf/ieo/index.html] [2] C.Gary, B.Lynn, A.Rick, “Photovolataic Fundamentals”, Solar Energy Research Institute, Springfield, VA : Available from National Technical Information Service, 1991. [3] “Photovoltaic Energy Program Overview”, US Department of Energy, 2000. [http://www.nrel.gov/docs/fy01osti/28976.pdf]. [4] Photovoltaics Overview 2001, Office of Solar Energy technologies, US department of Energy, 2001.[http://www.nrel.gov/docs/fy02osti/30729.pdf]. [5] P.Selvaraj, “Effect of Cadmium Chloride Treatment on CdS-CdTe Solar Cells”, Master’s Thesis, University of South Florida, Tampa,1999. [6] C.E. Backus “Solar cells”, IEEE press 1976, p.24. [7] A. Rohatgi, R. Sudarshanan, S.A.Ringel, “Wide Bandgap Thin Film Solar Cells from CdTe Alloys”, 20th IEEE Photovoltaic Specialists Conference, pp.1477-1481, 1988. [8] J.S.Britt, “II-VI Compound Thin Films by MOCVD for Tandem Solar Cells”, Ph.D Dissertation, University of South Florida, Tampa, 1992. [9] L.Zhihong, H.Xiaokang and W.B.Berry, “ Temperature dependence on the Energy Gap Selection in Two Cell, Four Terminal Tandem Solar Cell design” 22nd IEEE Photovoltaic Specialists Conference, pp.323-328. [10] W.Ma, T.Horiuchi, C.C.Lim, K.Goda, H.Okamoto, Y.Hamakawa, “ An Optimum Design of a a-Si/poly-Si Tandem Solar Cell”, 23rd IEEE Photovoltaic Specialists Conference, pp.833-838. [11] C.S.Ferekides and D.L.Morel, “ Development of II-VI-based High Performance High Bandgap Device for Thin Film Tandem Solar Cells”, National Center for Photovoltaics Program Review Meeting, Lakewood, CO, Oct. 14-17, 2001. [12] B.E.Mccandless, “ Cadmium Zinc Telluride Films for Wide Bandgap Solar Cells”, 29th IEEE Photovoltaic Specialists Conference, pp.488-491, 2002.

PAGE 64

53[13] A.Burger, H.Chen, K.Chattopadhayay, D.Shi, S.H.Morgan, W.E.Collins, R.B.Jones, “Characterization of Metal Contacts on and Surfaces of Cadmium Zinc Telluride”, Nuclear Instruments and methods in Physics Research, pp.8-13,1999. [14] Linjun wang, wenbin sang, Weimin Shi, Yongbiao Quin, jiahua Min, Donghua Liu, Yiben Xia, “Electrical Properties of Contacts on P-type Cd 0.8Zn0.2Te Crystal Surfaces”, Nuclear Instruments and methods in Physics Research, pp.581-585, 2000. [15] T.A.Gessert, P.Sheldon, X.Li, D.Dunlavy, D.Niles, “Studies of ZnTe Back Contacts to CdS/CdTe Solar Cells”, IEEE 26th Photovoltaic Specialists Conference, pp.419422,1997. [16] Makhratchev, K.J.Price, D.A.simmons, J.Drayton, A.Gupta, R.G.Bohnand, A.D.Compaan, “ZnTe:N Back Contacts to CdS/CdTe Solar Cells”, IEEE 28th Photovoltaic Specialists Conference, pp.475-478, 2000. [17] M.Xavier, T.Gerald., V.P.Singh, J.C.McClure, S.Velumani, R.Mathew, P.J.Sebastian, “Development of CdTe Thin Films on Flexible Substrates”, Solar Energy Materials and Solar Cells, Vol 76, pp.293-303, 2002. [18] A.N.Tiwari, A. Romeo, D.Baetzner and H.Zogg, “Flexible CdTe Solar Cells on Polymer Films”, Progress in Photovoltaics: Research and Applications pp.211215, 2001. [19] J.Herrero,C.Guillen, “Transparent Films on Polymers for Photovoltaic Applications”, Surface Engineering ,Surface Instrumentation and Vaccum Technology, Vol 67, pp.611-616, 2002. [20] S.Wiedman, R. Butcher, J.Fogleboch, J.Muha, R.G.Wendt, M.E.Beck, J.S.Britt, “CIGS Processing on Flexible Polyim ide Substrates”, National Center for Photovoltaics Program Review Meeting, Colarado, Oct 15,2001. [21] A.N.Tiwari, A.Romeo, D.Baetzner and H.Zogg, “Influence of Proton Irradiation and Development of Flexible CdTe Solar Cells on Polyimide”, Proceedings of 2001 Material Research Society Spring Meeting, San Francisco CA, USA, 16-20 April 2001. [22] S.N.Alamri and A.W.Brinkman, “The Effect of Transparent Conductive Oxide on the Performance of Thin Film CdS/CdTe Solar Cells”, J. Phys. D : Appl Phys.33, L1L4, 2000. [23] T.L.Chu, S.S.Chu, C.S.Ferekides, C.Q.Wu, J.Britt, C.Wang,“High Efficiency CdS/CdTe Solar Cells from Solution Grown CdS Films”, IEEE 22nd Photovoltaic Specialists Conference, Vol 2, pp.952-956, 1991. [24] S.Tatarenko, T.Baron, A.Arnoult, J.Cibert, M.Grun, A.Haury, Y.Merle, A.Wasiela, K.Saminadayar, “Nitrogen Doping of Te Based II-VI Compounds”, Journal of Crystal Growth, pp.682-687,1997.

PAGE 65

54[25] D.N.Marinskiy, “Cadmium Sulphide Thin Films Deposited by Close Spaced Sublimation and CdS/CdTe Solar Cells”, Ph.D. Dissertation, University of South Florida,Tampa,1998. [ 26 ] These results are from the work of my colleague S.Subramanian and will be included in his thesis to be submitted in summer 2004 to Electrical Engineering Department of University of south Florida.