USF Libraries
USF Digital Collections

Characterization of cadmium zinc telluride solar cells by RF sputtering

MISSING IMAGE

Material Information

Title:
Characterization of cadmium zinc telluride solar cells by RF sputtering
Physical Description:
Book
Language:
English
Creator:
Subramanian, Senthilnathan
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
tandem solar cells
wideband gap semiconductors
thin films
CZT
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 efficiency solar cells can be attained by the development of two junctions one stacked on top of each other into tandem structures. So that, if a photon is not able to excite an electron-hole pair in the top cell can create a pair in the bottom cell, which has a smaller bandgap. For a two junction tandem device structure, the bandgap of the top cell should be 1.6-1.8eV and for the bottom cell should be 1eV to attain efficiencies in the range of 25%. Cadmium Zinc Telluride which has a tunable bandgap of 1.45- 2.2eV is a candidate for the top cell of the tandem structure. Cadmium Zinc Telluride (Cd₁-xZnxTe) films were deposited by co-sputtering of CdTe and ZnTe. Deposition of Cd₁-xZnxTe was studied in Ar and Ar/N₂ ambient. Characterization of the films was done using transmission response, X-ray diffraction (XRD), Atomic Force Microscopy (AFM), Secondary Electron Microscopy (SEM), current-voltage (I-V) and spectral response measurements. CZT deposited on CdS/SnO₂ substrates showed improved performance compared to other heterojunction partners. Doped graphite and copper were utilized as back contacts for CZT devices. Post deposition annealing treatments with ZnCl₂ on CZT films were done and their effect on the devices was also studied. The best combination of Voc and Jsc were 530mV and 3.66mA/cm² respectively.
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 Senthilnathan Subramanian.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 84 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:
oclc - 56563342
notis - AJS2475
usfldc doi - E14-SFE0000429
usfldc handle - e14.429
System ID:
SFS0025121:00001


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 2200397Ka 4500
controlfield tag 006 m d s
007 cr mn
008 040811s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000429
035
(OCoLC)56563342
9
AJS2475
b SE
040
FHM
c FHM
090
TK145 (ONLINE)
1 100
Subramanian, Senthilnathan.
0 245
Characterization of cadmium zinc telluride solar cells by RF sputtering
h [electronic resource] /
by Senthilnathan Subramanian.
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 84 pages.
520
ABSTRACT: High efficiency solar cells can be attained by the development of two junctions one stacked on top of each other into tandem structures. So that, if a photon is not able to excite an electron-hole pair in the top cell can create a pair in the bottom cell, which has a smaller bandgap. For a two junction tandem device structure, the bandgap of the top cell should be 1.6-1.8eV and for the bottom cell should be 1eV to attain efficiencies in the range of 25%. Cadmium Zinc Telluride which has a tunable bandgap of 1.45- 2.2eV is a candidate for the top cell of the tandem structure. Cadmium Zinc Telluride (Cd-xZnxTe) films were deposited by co-sputtering of CdTe and ZnTe. Deposition of Cd-xZnxTe was studied in Ar and Ar/N ambient. Characterization of the films was done using transmission response, X-ray diffraction (XRD), Atomic Force Microscopy (AFM), Secondary Electron Microscopy (SEM), current-voltage (I-V) and spectral response measurements. CZT deposited on CdS/SnO substrates showed improved performance compared to other heterojunction partners. Doped graphite and copper were utilized as back contacts for CZT devices. Post deposition annealing treatments with ZnCl on CZT films were done and their effect on the devices was also studied. The best combination of Voc and Jsc were 530mV and 3.66mA/cm respectively.
590
Adviser: Christos S. Ferekides.
653
tandem solar cells.
wideband gap semiconductors.
thin films.
CZT.
690
Dissertations, Academic
z USF
x Electrical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
949
FTS
SFERS
ETD
TK145 (ONLINE)
nkt 9/1/04
4 856
u http://digital.lib.usf.edu/?e14.429



PAGE 1

Characterization of Cadmium Zinc Telluride Solar Cells by RF Sputtering by Senthilnathan Subramanian A thesis submitted in partial fulfillment of the requirements 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. Choiu, Ph.D Date of Approval: June 29, 2004 Keywords: czt, thin films, wide bandgap semiconductors, tandem solar cells Copyright 2004 Senthilnathan Subramanian

PAGE 2

DEDICATION This thesis is dedicated to my family and friends for their love and support.

PAGE 3

ACKNOWLEDGEMENT I would like to express my gratitude to my Major Professor Dr. Chris Ferekides for his invaluable guidance and support dur ing the course of this work. He has been a great source of inspiration during my work here. I also thank Dr.Don L. Morel and Dr. Yun L. Choiu fo r their guidance in the fulfillment of this thesis and graduate program. A special thank to all my colleagues, roommates and friends for their constant support and encouragement during this work.

PAGE 4

TABLE OF CONTENTS LIST OF TABLES iv LIST OF FIGURES v LIST OF SYMBOLS vii ABSTRACT viii CHAPTER 1 INTRODUCTION 1 1.1 Introduction 1 1.2 Renewable Energy 1 1.3 Photovoltaic Effect 2 1.4 Tandem Solar Cell 4 CHAPTER 2 INTRODUCTION TO SOLAR CELLS 7 2.1 PN Junctions 7 2.2 Heterojunctions 7 2.3 Solar Cells 11 2.4 Parameters of a Solar Cell 11 2.5 Current Flow in a Solar Cell 14 CHAPTER 3 LITERATURE REVIEW 19 CHAPTER 4 PROCESSING 32 4.1 Device Structure 32 4.2 Transparent Conducting Oxides 33 i

PAGE 5

4.3 Window Layer 33 4.4 Absorber Layer (Cd 1-x Zn x Te) 34 4.5 Zinc Telluride Deposition 36 4.6 Zinc Chloride Treatment 37 4.6.1 Air Annealing 37 4.6.2 Vapor Treatment in the Oven 37 4.6.3 Oxygen Free Sublimation 38 4.7 Back Contact 39 4.8 Measurements 40 4.8.1 I-V Measurements 40 4.8.2 Spectral Response 40 4.8.3 XRD, SEM and AFM 40 CHAPTER 5 RESULTS AND DISCUSSION 41 5.1 Properties of CZT Films 41 5.2 CZT in Argon Ambient 45 5.2.1 Effects of Contact Annealing on Cell Characteristics 45 5.2.2 Heat Treatment on CZT Devices 46 5.3 CZT Films Deposited in Ar/N 2 Ambient 47 5.3.1 Effect of CdS Thickness on CZT Devices 49 5.4 CZT on Transparent Conducting Oxides 51 5.5 Transparent Back Contact on CZT Films 53 5.6 CZT Graded Devices 54 5.7 Post Deposition Treatments of CZT 57 5.8 Zinc Chloride Treatment on CZT Films 57 5.8.1 ZnCl2 Treatment on CSS-CZT Substrates 61 CHAPTER 6 CONCLUSION 64 REFERENCES 66 ii

PAGE 6

APPENDICES 69 APPENDIX A 70 APPENDIX B 71 iii

PAGE 7

LIST OF TABLES Table 1. Composition and Bandgap of Cd 1-x Zn x Te Films 20 Table 2. Condition and Results of Post Deposition Heat Treatment of Cd 1-x Zn x Te/CdS/ITO/7059 Film in Various Ambient 22 Table 3. Devices Results for Cd 1-x Zn x Te/CdS/ITO/7059 23 Table 4. Comparison of CdZnTe Cells Annealed in Different Ambient 25 Table 5. Substrate Temperature Calibrations 36 Table 6. V oc of Devices at Different Annealing Temperatures 47 Table 7. V oc and J sc of Copper Back Contacted Samples 53 Table 8. V oc of Copper Back Contacted Samples at Various Conditions 54 Table 9. V oc and J sc of Type A Graded Samples 56 Table 10. CSS ZnCl 2 Treated Oxygen Free Sublimation CZT Sample 63 Table 11. History Sheets of XRD Samples 70 Table 12. XRD Peaks for Figure 30 71 Table 13. XRD Peaks for Figure 40 72 Table 14. XRD Peaks for Figure 42 73 iv

PAGE 8

LIST OF FIGURES Figure 1. Solar Cell, Module and Solar Array 3 Figure 2. Efficiencies of PV Cells and Modules 4 Figure 3. Iso-efficieny Curves for a Two-Cell Tandem Device Versus the Energy Bandgap of the Cell 5 Figure 4. Three Layer Tandem Structure Used in Space Applications 6 Figure 5. Band Diagram of a p-N Anisotype Heterojunction Before and After Contact 9 Figure 6. Types of Heterojunctions (a) nN Isotype Heterojunctions, (b) pP Isotype Heterojunctions, (c) Np Anisotype Heterojunction, (d) nP Anisotype Heterojunction 10 Figure 7. Gradual Smoothing of the Interface Spike of an n-P Heterojunction by Compositional Grading 11 Figure 8. Equivalent Circuit of a Solar Cell 12 Figure 9. Dark-Light I-V Characteristics of a Solar Cell 13 Figure 10. Energy Band Structure of P-N Junction 15 Figure 11. Light Interaction and Current Flow in a Photovoltaic Cell 17 Figure 12. II VI Bandgaps 19 Figure 13.Variation in Bandgap and Lattice Parameter with x for Cd 1-x Zn x Te Films 21 Figure 14. X-ray Diffraction of MBE Grown CdZnTe Films 23 Figure 15. Forward Bias Dark I-V Characteristics of (a) CdTe, (b)CZT (1.7eV) and (c) CZT (1.8eV) Cell Structures 24 Figure 16. Transmission of CdZnTe Films Annealed in (a) No Anneal (b) Argon Anneal, (c) Forming Gas Anneal, (d) Air Anneal 26 Figure 17. Auger Depth Profile of CdZnTe Film Over ITO/Glass Substrate 27 Figure 18. Auger Depth Profile of CdZnTe Films Grown on CdS/ITO/Glass 27 Figure 19. Cross-section of Sample Used for Interdiffusion Studies 28 Figure 20. AES for As-deposited HT Samples 29 Figure 21. AES Data for CdCl 2 Annealed HT Samples 30 v

PAGE 9

Figure 22. AES Data of As-deposited LT Samples 30 Figure 23. AES Data of CdCl 2 Annealed LT Samples 31 Figure 24. Structure of CZT Device 32 Figure 25. CZT Sputtering Unit 35 Figure 26. Vapor Treatment in an Oven 38 Figure 27. Chamber for Oxygen Free ZnCl 2 Process 39 Figure 28. Transmission Response of CZT Film Deposited in Ar with T SUB ~ 350 o C 42 Figure 29. Bandgap of CZT Film Deposited in Ar with T SUB ~ 350 o C 42 Figure 30. XRD of CZT Film on CdS/SnO 2 Substrate 43 Figure 31. AFM Image of CZT Film on CdS Substrate 44 Figure 32. SEM of CZT Films Deposited at 300 o C 44 Figure 33. SEM of CZT Films Deposited at 400 o C 45 Figure 34. Light I-V Characteristics of As-deposited and Contact Annealed Devices 46 Figure 35. Dark and Light I-V Characteristics of CZT in Ar/N 2 and Annealed in He at 500 o C for 5 Mins 48 Figure 36. Spectral Response of CZT in Ar/N 2 and Annealed in He at 500 o C for 5 Mins 49 Figure 37. Effect of CdS Thickness on CZT 50 Figure 38. Spectral Response of CZT with Different CdS Thicknesses 51 Figure 39. XRD of SnO 2 /CZT at Different Annealing Temperatures 52 Figure 40. Spectral Response of CZT on SnO 2 Substrate 53 Figure 41. XRD of Graded Samples 55 Figure 42. Spectral Response of Graded Samples 56 Figure 43. V oc and J sc of ZnCl 2 Air Annealed Samples at Different Temperatures 58 Figure 44. Effects of ZnCl 2 Air Annealing Treatment 59 Figure 45. Effect of ZnCl 2 Annealing Temperatures 60 Figure 46. Spectral Response of CZT Device by Vapor Treatment 60 Figure 47. ZnCl 2 Thermal Annealed CSS CZT Devices 61 Figure 48. Spectral Response of Vapor Annealed CSS CZT Devices 62 Figure 49. Spectral Response of Post Deposition Hydrogen Annealed and ZnCl 2 Treated Devices 63 vi

PAGE 10

LIST OF SYMBOLS a Lattice Constant () E c Energy Of Conduction Band Edge (eV) E F Fermi Level Energy (eV) E v Energy Of Valence Band Edge (eV) FF Fill Factor h Planck Constant (J-s) h Photon Energy (eV) I Current (A) I sc Short Circuit Current (A) J sc Short Circuit Current Density (A/cm 2 ) k Boltzmann Constant (eV/k) q Electron Charge (1.6 x 10 -19 C) R s Series Resistance (/cm 2 ) R sh Shunt Resistance (/cm 2 ) V oc Open Circuit Voltage (V) Optical Absorption Coefficient (cm -1 ) E Energy Difference (eV) E c, E v Conduction and Valence Band Discontinuities (eV) Wavelength (nm) Frequency Of Light (sec -1 ) Bulk Resistivity ( cm) Work Function (eV) Electron Affinity (eV) vii

PAGE 11

CHARACTERIZATION OF CADMIUM ZINC TELLURIDE SOLAR CELLS BY RF SPUTTERING Senthilnathan Subramanian ABSTRACT High efficiency solar cells can be attained by the development of two junctions one stacked on top of each other into tandem structures. So that, if a photon is not able to excite an electron-hole pair in the top cell can create a pair in the bottom cell, which has a smaller bandgap. For a two junction tandem device structure, the bandgap of the top cell should be 1.6-1.8eV and for the bottom cell should be 1eV to attain efficiencies in the range of 25%. Cadmium Zinc Telluride which has a tunable bandgap of 1.452.2eV is a candidate for the top cell of the tandem structure. Cadmium Zinc Telluride (Cd 1-x Zn x Te) films were deposited by co-sputtering of CdTe and ZnTe. Deposition of Cd 1-x Zn x Te was studied in Ar and Ar/N 2 ambient. Characterization of the films was done using transmission response, X-ray diffraction (XRD), Atomic Force Microscopy (AFM), Secondary Electron Microscopy (SEM), current-voltage (I-V) and spectral response measurements. CZT deposited on CdS/SnO 2 substrates showed improved performance compared to other heterojunction partners. Doped graphite and copper were utilized as back contacts for CZT devices. Post deposition annealing treatments with ZnCl 2 on CZT films were done and their effect on the devices was also studied. The best combination of V oc and J sc were 530mV and 3.66mA/cm 2 respectively. viii

PAGE 12

1 CHAPTER 1 INTRODUCTION 1.1 Introduction The non-renewable energy sources such as, fossil fuels like coal, oil and natural gas were formed millions of years ago. Repeated mining of them lead to their decline. Fossil fuels are still being created even today by heat and pressure, but they are being consumed more rapidly than they are created. Moreover the non-renewable energy sources produce a large amount of bi-products which are greatly harmful to our environment and living conditions. The bi-products such as carbon-di-oxide, sulfur and nitrogen oxide may result in global warming and acid rain. So it is time that the world slowly changes from using non-renewable energy to renewable energy. 1.2 Renewable Energy Energy derived from sources that are regenerative or for all practical purposes cannot be depleted or inexhaustible are called renewable energy. Types of renewable energy resources include moving water (hydro, tidal and wave power), thermal gradients in ocean water, biomass, geothermal energy, solar energy and wind energy. Of the above renewable energy sources one of the most important renewable energy is the solar energy which is obtained from the sun. The sun which is a constant source of heat and light emits about 5,000,000 tons of energy in the form of gamma rays per second when hydrogen is converted to helium [1]. As these rays travel towards the earth’s surface, the energy is continuously absorbed and re-emitted at lower and lower temperatures, so that by the time they reach the surface it is primarily visible light. About

PAGE 13

2 98% of the total energy radiated from the sun in the outer space lies between 0.25m to 0.3m. A quantity called the solar constant is defined as the rate at which energy is received on a unit surface, perpendicular to the sun’s direction, in free space at the earth’s mean distance from the sun. And its value is given by 1.353kWm-2. The actual solar radiation in free space differs from this value by 3.35% due to the change in distance between the sun and the earth surface throughout the year [2]. The path length through the atmosphere is conveniently described in terms of an equivalent relative air mass mr. The solar spectrum in the outer space is given by AM0 and on the earth’s surface is approximately AM1.5 which has a value of 0.8318kWm-2. 1.3 Photovoltaic Effect The method of converting photon energy (packets of energy) depending on the wavelength of light into electrical energy is called the photovoltaic effect. If a photon is incident upon a semiconductor and if the photon energy is greater than the bandgap, the photon may be absorbed and an electron can be elevated from the valence band to the conduction band. This creates an electron-hole pair. Mobile charge carriers produced in this manner are called photocarriers. When these carriers combine they annihilate each other and emit a photon corresponding to the bandgap energy. The charge carriers may be collected before they combine, to form a photocurrent. A solar cell is a photovoltaic cell that converts heat from the sun (black body radiation) into work (electrical energy). The production of solar cells began long way back in the first half of the 20th century. Silicon was the first material used for solar cells. But fabricating solar cells with crystalline silicon was a costly process and since it is an indirect bandgap material, the absorption co-efficient of silicon is low. So thicker films or sophisticated light trapping techniques had to be used. So initially crystalline silicon solar cells were mostly used only in space applications and not for terrestrial applications. Due to the energy crisis and

PAGE 14

3 decrease of fossil fuels, new research began on new materials for solar cells. To decrease the cost of production of solar cells thin film materials like GaAs, CIS and CdTe are being used. The main purpose of using thin films as an alternative to single crystalline silicon are cheap substrates, manufacturing cost and the possibility of covering large areas. Technologies like screen printing [21, 23] are being tested to lower the deposition cost. CdS/CdTe solar cells with efficiencies with 16.4% [3] and CIS solar cells with efficiencies of 19.2% [4] have been obtained. A number of solar cells electrically connected to each other and mounted in a support structure or frame is called a photovoltaic module. The current produced is directly dependent on how much light strikes the module. Figure 1. Solar Cell, Module and Solar Array Multiple modules are wired together to form an array. The larger the area of the module or array, the more electricity will be produced. Photovoltaic modules and arrays produce direct current (dc) electricity. The modules can be connected both in series and parallel to produce arrays of required voltage and current. Figure 2 shows the efficiencies of all the photovoltaic cell and modules [5].

PAGE 15

4 Figure 2. Efficiencies of PV Cells and Modules 1.4 Tandem Solar Cells High efficiency solar cells can be obtained by using two or more different cells with more than one bandgap and more than one junction to generate a voltage. These kinds of cells are known as tandem cells. Tandem solar cells so formed achieve higher total conversion efficiency because they can convert more of the energy spectrum of light into electricity. Photons which pass through the first cell, because their energy was too low to create electron-hole pairs in the top cell are able to create electron-hole pairs in the next cell which has a smaller bandgap than the top cell. This procedure clearly makes use of the higher energy photons, for if they merely gave up the energy to create one electronhole pair, the remaining energy would go to heat up the lattice, instead of adding to the solar current. The bandgap of the top cell of the tandem cell should be higher than that of the bottom cell and a range of 1.6 -1.8eV is optimum for higher efficiencies [2]. The bottom cell with a bandgap value of 1-1.2eV is preferred which can be seen from figure 3.

PAGE 16

5 Figure 3. Iso-efficiency Curves for a Two-cell Tandem Device Versus the Energy Bandgaps of the Cell Tandem solar cells find their application more in space programs. Since a single layer solar cell can convert only a limited portion of the solar spectrum into efficient energy, new devices with two or three junctions are stacked over each other to convert a large portion of the solar spectrum to electricity. Figure 4 shows a three layer tandem structure of gallium indium phosphide (GaInP), gallium arsenide (GaAs) and germanium (Ge) which are grown over each other producing a simpler, cheaper and lighter energy system. This tandem structure developed by NREL has an efficiency of 30% [6, 7]. These tandem concepts are now being tried for terrestrial applications which are a clean, renewable and cost competitive way to achieve higher efficiencies in solar cells.

PAGE 17

6 Figure 4. Three Layer Tandem Structure Used in Space Applications In this thesis Cadmium Zinc Telluride (Cd1-xZnxTe) which has a variable bandgap of 1.45 to 2.2eV depending on the composition of (x) is studied for the top layer of the tandem cell with CIGS as the bottom layer. Cadmium Zinc Telluride is deposited by co-sputtering of CdTe and ZnTe targets and their characteristics are studied under various conditions.

PAGE 18

7 CHAPTER 2 SEMICONDUCTORS AND SOLAR CELLS 2.1 PN Junctions The simplest semiconductor junction is formed by bringing together a p-type and a n-type material. The three common techniques of junction formation are epitaxial growth, diffusion and ion implantation. PN junctions can be divided into: € Step junction which has a uniform p doping on one side of a sharp junction and uniform n doping on the other side and vice versa. € Graded junction in which the transition from a p-type to n-type occurs over many atomic spacing and vice versa. PN junctions can also be classified into: € Homojunction: A junction formed between n-type and p-type material of the same semiconductor € Heterojunction: A junction formed between two different semiconductor materials with different bandgap energies 2.2 Heterojunctions A heterojunction is formed when two semiconductors with different bandgap and lattice constants are brought together. Optoelectronic devices made of compound semiconductors are mostly made of heterojunctions. If the two semiconductors have the same lattice constants, then the interface between them maybe virtually free of defects. In general, the lattice constants of the two semiconductors are different. If a semiconductor is formed by epitaxy, then there will be a misfit between the two semiconductors. Lattice matching can

PAGE 19

8 be achieved by growing the appropriate semiconductor with the right composition. Example is InAlAs with a bandgap of 1.45eV is lattice matched to InP of bandgap 1.35eV. For heterojunctions, there is the inherent problem that the crystal structure changes across the junction and an interface is formed between the two semiconductors. Interfaces can be efficient recombination centers because they introduce deep trap levels in the bandgap. They can also provide sites for quantum mechanical tunneling processes, which is important for current loss mechanisms across the junction. The interface traps degrade the performance of the cells and it is necessary that heterojunctions with low density of interface traps be produced. One source of the interface traps maybe due to the lattice mismatch of the two semiconductors. Figure 5 shows the bandgap diagram of an n-type wide bandgap semiconductor and a ptype narrow bandgap semiconductor with their Fermi energies EF1 and EF2, work functions 1 and 2 and electron affinities 1 and 2. Where the work function and electron affinity are defined as the energy required to remove an electron from the Fermi level EF and from the bottom of the conduction band EC respectively, to a position just outside the material (i.e.) vacuum level [24,25]. The built in potential Vbi is simply the difference between the work functions. Vbi = 2 1 The main difference for a pn heterojunction is the discontinuity in the conduction band energies, which is equal to the difference in the electron affinities EC = ( 2 1 ) q And the band offset is, EV = ( 2 1 ) q + Eg2 – Eg1

PAGE 20

9 Figure 5. Band Diagram of p-N Anisotype Heterojunction Before and After Contact A negative EC (or EV for a p-on-n heterojunction) produces a spike in the conduction (or valence band) which is undesirable for photovoltaic applications. The spike impedes the flow of minority carriers across the junction from p-type to the n-type regions, which leads to the reduction in photocurrent. However such spikes can be avoided by a suitable combination of electron affinities and bandgap energies. Heterojunctions can be divided into two groups depending on conductivity. If the two semiconductors have the same type of conductivity the junction is called an isotype heterojunction. And when the conductivity types differ, the junctions are called anisotype heterojunctions.

PAGE 21

10 Figure 6. Types of Heterojunctions (a) nN Isotype Heterojunctions, (b) pP Isotype Heterojunctions, (c) Np Anisotype Heterojunction, (d) nP Anisotype Heterojunction The figure 6 shows nN and pP isotype heterojunctions and Np and nP anisotype heterojunctions. Here N and P stand for wide bandgap semiconductors and n and p stand for narrow bandgap semiconductors. The interfacial spikes, produced due to the band offsets of two semiconductors may sometimes be undesirable for ideal device operation. The magnitude of the spikes produced can be reduced by compositional grading at the interface, which reduces the

PAGE 22

11 electron affinity difference at the interface and therefore the field magnitude. Figure 7 shows the compositional grading of an n-P heterojunction [26]. Figure 7. Gradual Smoothing of the Interface Spike of an n-P Heterojunction by Compositional Grading 2.3 Solar Cells Solar cell is a pn junction device, which converts the energy from the sun to electric energy. Today solar cells have been an integral part of our life finding applications ranging from calculators, to watches, to solar powered irrigation systems, etc. When light or photons, falls on an unbiased junction, it creates electron-hole pairs that diffuse towards the junction and electrical power is developed across the junction which can be delivered to an external load. The primary requirement for a material to be suitable for a solar cell is a bandgap matching the solar spectrum and high mobility and lifetime of charge carriers. 2.4 Parameters of a Solar Cell Lets us consider an equivalent circuit for a solar cell as seen in figure 8. In the ideal case with the open circuit condition, the load resistance is infinite, so there is no current in the circuit and the voltage is maximum known as open circuit voltage and denoted by Voc. In

PAGE 23

12 the case of short circuit condition, the load resistance is zero, so the voltage in the circuit is zero and the current is maximum known as the short circuit current and denoted by Isc. Figure 8. Equivalent Circuit of a Solar Cell The dark and light I-V characteristics of a solar cell are given in the figure 9. Under dark conditions the solar cell behaves similar to a diode. The ideal diode equation is given by I = Io {exp [q (V IRs)/AkT] – 1} Where I is external current flow, Io is the reverse saturation current, q is the magnitude of electric charge which is 1.602 x 10 -19 Coulombs, V is the applied voltage, A is the diode factor, k is the Boltzmann constant and T is the absolute temperature. Under illumination, the above equation becomes, I = Io {exp [q (V IRs)/AkT] – 1} IL where IL is the light generated current. Due to the light generated current the currentvoltage characteristics are shifted down.

PAGE 24

13 Figure 9. Dark-Light I-V Characteristics of a Solar Cell The electrical power is the product of voltage times current and given by, I V P = And the peak power obtained in a solar cell is, I max V max P max = The Fill Factor of the solar cell is defined as the peak power to the power calculated by multiplying the Voc and Isc. FF = I sc V oc P max = I sc V oc I max V max

PAGE 25

14 Fill Factor is always less than unity, due to the presence of series resistance and shunt resistance in the solar cell as seen in figure 4, it is the reflection of how much series resistance and how little shunt resistance is present in a solar cell. A good Fill Factor for a solar cell can be obtained by infinite shunt resistance and very low series resistance. The photovoltaic conversion efficiency is another important factor to be taken into consideration. It is the measure of the amount of light energy that is converted into electrical energy and is given by = P s I sc V oc FF Where Ps is the power applied. 2.5 Current Flow in a Solar Cell When a p-type material and n-type material are figuratively joined together to form a pn junction, the electrons from the n-type region would diffuse into the electron deficient ptype region crossing the junction boundary of the p and n regions. Likewise the holes in the p-region cross to the holes deficient n-type region until a voltage equal to sum of the potential of the holes and electrons is obtained across the p-n junction. Figure 10 shows the energy band structure due to the formation of p-n junction. In figure 10 it can seen that due to the flow of electrons and holes the conduction and valence bands in the ptype material has risen relative to those of the n-type material. This is because of the Fermi level which was originally higher in the n-type material, the level is equalized within the joining crystal and across the p-n junction by virtue of the n-type material feeding electrons to the p-type material and vice versa with respect to holes.

PAGE 26

15 Figure 10. Energy Band Structure of P-N Junction The p-type region now contains a disappropriate amount of electrons and is at a more negative potential, which causes a field across the junction. Thus the electrons which are photolytically generated in the p-type region can flow towards the n-type region and the photolytically generated holes in the n-type region can flow towards the p-type region. If light of greater energy than the valence and conduction bands fall on the crystal, it may be absorbed by the electrons in the valence band, which exit to the conduction band leaving a hole in the valence band. Under the influence of the electric field the photo-excited electrons will be driven towards a lower energy state in the n-type region of the crystal while the holes moves towards a low energy state which for them is the p-type region. The photocarriers have moved into the respective regions of the crystal where like charges are the majority charge conductors, a photovoltage is created and current flows through the external circuit.

PAGE 27

16 As seen in the figure 11 [9] there are a lot of loss processes which tend to decrease the efficiency of sunlight converted into electricity. The sunlight which comprises of the spectrum of photons energies, only those with energy greater than the bandgap energy can be absorbed to produce photocarriers. Of the total light that falls on the solar cell, a part of it is reflected off the surface of the crystal due to the different refractive index of the semiconductor material. The refractivity R, for any material is given by () () 1 n 2 1 n 2 R Š + = Where n is the refractive index of the semiconductor. More than 50% of the photons in the sunlight are of proper energy and are not reflected by the surface and enters the crystal to be absorbed in them. The amount of photons absorbed in the crystal is given by 1-exp(-t) where is the absorption co-efficient and t is the thickness of the cell. The rest of the photons exp (-t) is transmitted through the cell. The optimal depth of the p-n junction beneath the face of the crystal is dictated by such factors as, where in the crystal most of the light is absorbed so as to produce electron-hole pairs, the lifetimes and mobility of the photocarriers and the resistance of the very thin side of the junction next to the surface through which the current must flow to reach the front contact electrode. This latter resistance is governed predominantly by the geometry of the cell, while other phenomenons are functions of the material characteristics of the crystal itself. A defect free p-n junction is needed to enhance the lifetimes of photocarriers to prevent their recombination and allow them to diffuse further.

PAGE 28

17 Figure 11. Light Interaction and Current Flow in a Photovoltaic Cell The collection efficiency is a measure of the proportion of the minority carriers, holes in the n-type and electrons in the p-type regions produced by the absorbed photons that reach the junction. Of the carriers generated outside the influence of the electric potential at the junction, some diffuse towards the junction while other diffuse away and recombine in the bulk or at the surface. Conduction of electricity through the very thin resistance layer between the p-n junction and the back contact of the cell is another important loss process. The geometric factors of contact location using a grid or screen instead of a spot contact and junction depth which partly governs the resistance, have to be balanced against loss of incident light through masking and collection efficiency deteriorations. The efficiency of the cell is reduced several percentage by this resistance. The voltage which is developed by a solar cell is a function of the excess minority carriers on each side of the p-n junction. The voltage so produced is always less than the bandgap

PAGE 29

18 of the cell because of the junction losses. The voltage thus developed increases with the intensity of illuminations towards a condition where the minority carrier density approached the majority carrier density and the voltage approaches the bandgap energy. The voltage cannot go above, as the junction potential will be nullified. The junction loss decreases exponentially with increasing bandgap.

PAGE 30

19 CHAPTER 3 LITERATURE REVIEW Thin film polycrystalline solar cells with high efficiencies have been obtained from low bandgap cells like CIGS with a bandgap of 1.1eV and mid bandgap cells like CdTe with bandgap of 1.45eV. However, further development of high efficiencies can be obtained by tandem cells. Efficiencies of 25% or greater can be obtained with the tandem structure with more than 16 -18% of the efficiencies contributed by the top cell and the rest by the bottom cell. To attain efficiencies in the above range, the bandgap of the top cell should be ideally 1.7eV and 1eV for the bottom cell. CIGS solar cells are well suited for bottom cell because of their 1eV bandgap and efficiencies of more than 15%. Finding a suitable material for the top cell is difficult task since the rest of the tandem structure and critical issues of fabrication of the bottom cells all depend on the top cell. Though a number of materials can be listed, materials in the II –IV groups are promising. Figure 12 shows the bandgap of cadmium and zinc compounds with their respective group IV materials [10]. Figure 12. II – VI Bandgaps From figure 12 two materials can be used as the top cell for the tandem structure CdSe and Cd1-xZnxTe. Though CdSe has a bandgap of 1.7eV and binary, Cd1-xZnxTe is more

PAGE 31

20 preferred since it is an extension of the well known CdTe and it has a tunable bandgap from 1.5eV to 2.2eV depending on the values of x, so that the bandgap can be tuned according to the bottom cell requirements. A number of fabrication techniques used for making Cd1-xZnxTe solar cells and their properties and characteristics has been discussed here. The Cd1-xZnxTe alloy film has been deposited by co-evaporation of CdTe and ZnTe powders of 6N purity, at a source substrate distance of 20cm. The alloy compositions of the film were determined by the ratio of the elemental impingement rates and their relative sticking co-efficient, which are controlled by the source temperatures of the material. The CdTe source temperature was fixed at 900oC and the ZnTe source temperature was varied to get desired composition of the alloy. Sublimation of the compounds are given by CdTe Cd + Te2 ZnTe Zn + Te2 CZT films of 3-4 microns thickness with x values from 0 to 1 were deposited over CdS/ITO/7059 glass at 325oC. Compositions were determined by Energy Dispersive Xray Spectroscopy (EDS) and lattice parameters by X-ray diffraction (XRD) using Nelson– Riley–Sinclair–Taylor analysis. Bandgap was obtained by 2 verses energy by transmission of the cells. The table 1 [11] shows the composition and bandgap of the resultant cells. Table 1. Composition and Bandgap of Cd1-xZnxTe Films

PAGE 32

21 Figure 13.Variation in Bandgap and Lattice Parameter With x for Cd1-xZnxTe Films The above figure 13 also shows the variation in bandgap and lattice parameters with a change in value of x in Cd1-xZnxTe. High efficiency CdTe/CdS thin film solar requires thermal heat treatment of the structure in CdCl2 ambient to recrystallize and partially dope the CdTe p-type. Applying the process directly to CZT films did not give favorable results due the reaction of CdCl2 with ZnTe in the formation of volatile ZnCl2 and excess CdTe. But ZnCl2 did not appreciable react with CdTe. The films were post deposition heat treated in argon, dry air, CdCl2 vapor + dry air and ZnCl2 vapor + dry air. XRD was done to study the phase changes and alloy distribution of the heat treated films. Table 2 shows the summarized results of Cd1-xZnxTe films with x ~ 0.4 [18].

PAGE 33

22 Table 2. Condition and Results of Post Deposition Heat Treatment of Cd1xZnxTe/CdS/ITO/7059 Film in Various Ambient Table can be summarized as follows: (i). Treatment in Ar at 600oC only causes primary recrystallization. (ii). Treatment in dry air preferentially removes ZnTe from the alloy as oxides. (iii). Treatment in ZnCl2:O2:Ar at low p(ZnCl2) preserves the bulk alloy composition. (iv). Treatment in ZnCl2:O2:Ar at higher p(ZnCl2) removes ZnTe from the alloy as ZnO and Te. (v). Treatment in CdCl2:O2:Ar preferentially removes ZnTe from the alloy by forming CdTe, ZnO, Te and volatile ZnCl2. Devices were fabricated after the post deposition treatments stated above. Back contact was formed by the sequential deposition of Te and Cu layers by electron beam evaporation. Graphite ink was applied and treated at 100oC for 30 minutes. The devices fabricated from Cd1-xZnxTe alloys exhibited significant dark-light crossovers and low Jsc. Poor conductivity and poor carrier transport in the bulk maybe the reason for this behavior. ZnCl2 treated devices showed low collection and long wavelength fall-off at 720nm, corresponding well to the as deposited bandgap of 1.7 eV. Devices treated with argon also showed similar behavior of low Jsc at x ~ 0.4 which can be seen from table 3.

PAGE 34

23 Table 3. Devices Results for Cd1-xZnxTe/CdS/ITO/7059 CZT films were also grown by Molecular Beam Epitaxy (MBE) using a Varian Gen II MBE system. Glass/SnO2/CdS substrates and the source materials with 5N purity were used. The substrates temperature of 275oC was used for first 30 minutes to achieve uniform nucleation and then 325oC till the end of the run. The layers were Br: Methanol etched to remove oxides formed. The cells were back contacted with ZnTe/Ni structures. Figure 14. X-ray Diffraction of MBE Grown CdZnTe Films X-ray diffraction analysis were performed on the cells and the figure 14 [12] shows the XRD image of the CZT film. From figure 14 is also seen that no other mixed phases are

PAGE 35

24 present in the CZT film. The lattice constant ‘a’ of the CZT film was determined using an extrapolation method was used to estimate the Zn content (x) in CZT according to a(x) =6.4810.381x The estimated atomic concentration of Zinc (x) was found to be ~0.40 and with a bandgap of ~1.7eV. Loss mechanisms of CZT films were studied using various measurements. C-V measurements show that the doping densities are about an order of magnitude greater than CdTe samples produced by MBE which could result in incomplete depletion of the CZT film. The spectral response measurements also showed strong wavelength dependence with reduced carrier collection at longer wavelength and are suffering from recombination in the undepleted bulk. However the large decrease in absolute spectral response cannot be explained by the undepleted bulk and the drop may also be due to contamination of CZT/CdS and back contact region. The major difference between CZT and CdTe MBE grown cells is series resistance and Jsc. Series resistance measured from a group of CZT cells with same thickness but different bandgap showed that dark series resistance increased by a factor of 4 whatever the CZT composition from figure15 [13]. Figure 15. Forward Bias Dark I-V Characteristics of (a) CdTe, (b)CZT (1.7eV) and (c) CZT (1.8eV) Cell Structures

PAGE 36

25 Since the undepleted bulk resistance cannot account for the observed series resistance, the back contact and CdS/CZT regions were investigated. A study with a thin layer of CdTe was deposited between CZT and CdS showed not a much of change in Voc and series resistance, which suggests that high series resistance is not caused by CdS/CZT interface. Measurable data were obtained only from annealed films. But air annealing decreased the bandgap of the cells from 1.7eV to 1.55eV. CZT samples showed poor performance with air annealing, so they were annealed in different ambient. The unannealed and argon annealed samples were not measurable. The samples annealed in forming gas (10% hydrogen + 90% nitrogen) and argon got highest Voc which can be seen from table 4 [13]. Cell efficiencies had no improvement due to high series resistance of the forming gas annealed samples. Forming gas anneal improves the quantum efficiencies in the film and also on the CdS/CZT junctions, which leaves only back contact region as the prime suspect. Table 4. Comparison of CdZnTe Cells Annealed in Different Ambient Series resistance of back contacted CZT cells were studied by annealing under different ambient of air, forming gas and argon for 20 minutes at 350oC. From figure 16 it is seen that the transmission of the air annealed samples showing 20% decrease in transmission from others in the long wavelength region. XPS measurements showed zinc in the oxidized form on the surface. Br: Methanol etch can make the surface Te rich, but forming

PAGE 37

26 annealed samples were found to have less Te on the surface, so various etching techniques should be used to make the surface Te rich to obtain high efficiencies on these cells. Figure 16. Transmission of CdZnTe Films Annealed in (a) No Anneal, (b) Argon Anneal, (c) Forming Gas Anneal, (d) Air Anneal A two stage technique is discussed here which is an emerging method used to deposit CuInSe2, ZnTe, CdTe and CZT layers. It is a two step process where the elements of the compound are deposited on to the substrate and then the deposited elements are reacted to form the desired compound. Annealing of the compound formed in air, is necessary to enhance the efficiency of the solar cells. Layers of Te, Cd and Zn were deposited sequentially over ITO/glass or CdS/ITO/glass by evaporation. The thicknesses of the deposited layers were adjusted to yield the compound film of derived stoichiometry and thickness after the reaction step. € A 100 thick Cd with 157 thick Te yield 298 thick CdTe € A 100 thick Zn with 223 thick Te yielded 331 thick ZnTe € In the case of Cd1-xZnxTe alloy film the composition parameter x is reacted to individual thickness of CdTe (t CdTe ) and ZnTe (t ZnTe ) tCdTe / tZnTe = 1.27(1-x)-1x The reaction step was carried out in a tube furnace, pumped down to 100mT and back filled with nitrogen or forming gas. The reaction step takes place from 15 minutes to 2

PAGE 38

27 hours at 350oC to 600oC temperature range. Gold contacts were made after Br: Methanol etch. The samples were annealed in air before the etch. Figure 17. Auger Depth Profile of CdZnTe Film Over ITO/Glass Substrate Figure 17 shows the Auger depth profile of CZT over ITO [14]. Zinc is distributed nonuniformly over the CZT layer with high concentrations on the surface along with oxygen. This is due to the zinc oxide formation on the surface due to air annealing. A zinc peak is also seen at ITO/CZT interface, which is due to the reaction of Zn with ITO at 400oC. Figure 18 shows the Auger depth profile of CdS/CZT interface which depends on the composition of x. It is observed that distribution of Zn in the CZT is non-uniform with intermixing of CdS/CZT layers and it increases with increase in concentration of zinc. The concentration of zinc is more on the surface and is in the form of oxide. Zn peak found near the ITO/CdS region is due to exchange reaction between Zn and CdS layers. The Zn which is a reducing agent reacts with CdS to form CdZnS which is thicker and more transparent. Figure 18. Auger Depth Profile of CdZnTe Films Grown on CdS/ITO/Glass

PAGE 39

28 The devices fabricated from these CZT films were poor due to the Zn reaction with CdS which gave rise to highly diffused CdS/CZT interface and poor morphology. CZT films were deposited by key approaches which have increased the performance of CdTe cells. € CSS process with which best performing CdTe solar cells are obtained. € CdCl2 treatment with which the efficiency of the CdTe cells increased. € One more key factor is oxygen ambient where the CdCl2 treatment takes place to get high performance cells. CSS process done with mixed powders of CdTe and ZnTe did not show any presence of Zinc when measured by AES. The ratio was increased from 1:1 to 1:20 without any use. This may be due to the vapor pressure of ZnTe which is lower than CdTe in the 650oC range or may be due to sticking co-efficient of zinc which is lower than cadmium at or above 400oC. Prealloyed CZT powers from CERAC, Inc were used. This too was of no use as the preformed alloys evaporated congruently. The presence of zinc was less than 5% in all cases. Figure 19. Cross-section of Sample Used for Interdiffusion Studies The above figure 19 shows the cross sectional view of the sample used for the interdiffusion studies of CdTe and ZnTe [15]. Here glass/SnO2/CdS substrates were used. CdTe was deposited by CSS at 400oC-580oC range in He/O2 ambient. ZnTe was deposited by RF sputtering at 350oC. Two types of cells were fabricated with top cell deposited at

PAGE 40

29 580oC high temperature (HT) and at 400oC low temperature (LT). The bottom layer of CdTe was deposited at 580oC. CdCl2 was done in presence of He/O2 from 400oC to 440oC temperature range. Presence of boulders at the surface of as deposited high temperature (HT) samples was found by AFM which disappeared on post deposition CdCl2 treatment. No such structures were seen in LT samples. Temperature increase of CdCl2 treatments helped in gradual grain size increase. Figure 20. AES for As-deposited HT Samples AES depth profile measurement of the as deposited HT samples is shown in figure 20. There is no CdTeZnTeCdTe structure as the cells were fabricated due to high temperature deposition of CdTe at 580oC. Large quantity of zinc is seen on the surface along with O2. They may me present in the form of oxide. The profile of Cd and Te and Zn and O2 track each other. Interdiffusion of Zn and Te into the CdS layer and S interdiffusion into CdTe layer are seen. Figure 21 shows the CdCl2 treated HT samples where Cd and Te profile are spreading towards the centre of the sample and interdiffusion of zinc to the surface. Profiles of Zn and O2 track each other. A small amount of chlorine is also seen on the top layer.

PAGE 41

30 Figure 21. AES Data for CdCl2 Annealed HT Samples Figure 22. AES Data of As-deposited LT Samples The data of the as deposited LT sample shown in figure 22 has a distinct 3-layer structure. It can be seen that Cd and Te spectra in the top and bottom layers exhibit different binding energies which implies different chemical bonding for CdTe at 580oC bottom cell and 400oC top cell. The zinc spectra of ZnTe also have different binding energies compared to interdiffusion of zinc in the top and bottom layers. The change of bonding energy of zinc with top and bottom layers of CdTe at different temperatures, both of which are deposited in O2 containing atmosphere may be due to zinc bonded to oxygen. With CdCl2 treatment as seen in figure 23 interdiffusion takes place and the distinct layers disappears. The

PAGE 42

31 amount of Zn and O2 increase and track each others profile due to the formation of ZnO. There is no interdiffusion of S from the CdS into CdTe in the LT samples. Figure 23. AES Data of CdCl2 Annealed LT Samples For samples annealed at high temperatures a gradual decrease in Zinc and Oxygen concentrations can be seen which is due to the reaction of CdCl2 with Zn forming ZnCl2 which is lost. Due to high affinity of Zinc to oxygen results in the formation of ZnO. Even a small amount of oxygen is sufficient to oxidize all zinc, which leads to the formation of p-type CdTe with n-type ZnO in the bulk. And accumulation of n-type ZnO at the surface, forming a junction at the back contact that is opposing the main junction at the CdS interface. The presence of ZnO in the bulk and the back contact junction may lead to severe performance degradation of CZT solar cells.

PAGE 43

32 CHAPTER 4 PROCESSING 4.1 Device Structure This chapter deals with the various methods used in the fabrication of the CZT solar cells. Figure 24 shows the basic cell structure used for this work. Figure 24. Structure of CZT Device Corning 7059 borosilicate glass is used as the substrate. The first layer deposited on 7059 glass is the transparent conducting oxide (TCO). For this research tin oxide (SnO2) is used as the TCO, along with indium tin oxide (ITO). n-type cadmium sulphide (CdS) is used as the window layer for the p-type CZT absorber layer. Several options are used for the formation of back contact (a) doped graphite, (b) Copper (Cu), (c) ZnTe/ doped graphite, (d) ZnTe/Cu.

PAGE 44

33 4.2 Transparent Conducting Oxides In a solar cell, current is collected through the front and back contacts and of those the front one should be transparent so that photons can pass them. These layers are generally called Transparent Conducting Oxides (TCO). Tin oxide (SnO2) is the most commonly used TCO for CdTe solar cells and it was therefore chosen as the main TCO for CZT. SnO2 used was a bilayer structure with extrinsic SnO2 doped with fluorine and an intrinsic SnO2 layer over it. SnO2 is doped with fluorine to obtain high conductivity. SnO2 is deposited by Metal Oxide Chemical Vapor Deposition (MOCVD) process in an atmospheric reactor. Tetramethyltin (TMT) is used as a source for tin and halocarbon 13B1 is used as the source for fluorine. Helium is used as the carrier gas for TMT and the process takes place in the ambient of Oxygen and Helium. The substrate temperature is maintained at 460oC. SnO2 so obtained has a thickness of 700-800. SnO2 which has a bandgap of 3.5eV absorbs all the light in the far infrared region and allows the light in the visible region into the cell. The sheet resistance of the layer is about 7-10 /. Indium tin oxide (ITO) is another TCO which was used for the experiments. ITO was deposited on substrates by RF magnetron sputtering in the ambient of Argon and Oxygen. The substrate temperature was maintained between 200300oC, and the pressure was between 2-3mT. The thickness of the film deposited varied from 2000 to 3000. The sheet resistance of the film obtained was measured to be between 7-10 /. 4.3 Window Layer Several types of window layers like cadmium sulphide (CdS) done by Chemical Bath Deposition (CBD), zinc selunide (ZnSe) and zinc oxide (ZnO) deposited by sputtering were used. Cadmium sulphide has been used as the window layer for most absorber layers like CdTe, CIGS and Cu2S etc. High efficiencies on CdTe films have been obtained with CdS as the window layer [22]. CdS has a bandgap of 2.42eV and has good photoconductivity.

PAGE 45

34 Though there are many techniques used for the deposition of CdS, CBD process is used here, which is a cheap and simple method. In CBD technique, a mixture of ammonia (NH3), ammonium acetate (CH3COONH4), cadmium acetate (Cd(CH3CO2)2) and thiorea ((NH2)2CS) are used. Here cadmium acetate is the source for cadmium and thiorea is the source for sulfur. Ammonium acetate and ammonia are used as buffer solutions. The SnO2:F coated 7059 glasses are kept in a bath of water and maintained at a temperature of 80-85oC throughout the process. The above mixture of ammonia, ammonium acetate, cadmium acetate and thiorea are added at correct proportions at regular intervals of time. The thickness of the CdS so obtained can be varied by the deposition time. Mostly a thickness of 1000 was used. The deposited CdS films are cleaned ultra sonically with water to remove any loose CdS particles. Zinc Selenide (ZnSe) is another window layer used. It is deposited by RF magnetron sputtering and the target used is 99.999% pure. The ZnSe is deposited in an ambient of Argon (Ar) at a pressure of ~5mT. The temperature was varied from 200-300oC and the thickness was varied from 200 to 1000. 4.4 Absorber Layer (Cd1-xZnxTe) Cadmium Zinc Telluride (Cd1-xZnxTe) which has a variable bandgap of 1.5 – 2.2eV is used as the p-type absorber layer. CZT was deposited by co-sputtering of Cadmium Telluride and Zinc Telluride sources which are 99.999% pure. A co-sputtering unit was constructed to sputter Cadmium Telluride (CdTe) and Zinc Telluride (ZnTe) using 3” RF magnetron sputtering guns. Figure 25 shows the sputtering unit.

PAGE 46

35 Figure 25. CZT Sputtering Unit The chamber is initially roughed by a mechanical pump, and subsequently pumped down to high vacuum of 10-7 torr range by a cryopump. The cryopump is cooled by UHP Helium from a compressor. The substrate holder is made of graphite and it is heated by a coil heater. Different substrate temperatures from 200400oC were used for deposition and the temperature of substrate was maintained within 1oC of the set point. And this was verified by several calibration runs. Table 5 gives the heater temperatures used to obtain the required substrate temperatures.

PAGE 47

36 Table 5. Substrate Temperature Calibrations Substrate Temperature (oC) Heater Temperature (oC) 400 630 350 585 300 500 250 425 200 380 The substrates were rotated at a constant speed with the help of a DC motor during the deposition in order to improve the uniformity of film deposited. The co-sputtering unit consists of two 3” RF magnetron sputtering guns which are separated by a shield to minimize over-talk. Argon (Ar) or argon/nitrogen (Ar/N2) mixture was used as the deposition ambient. The chamber pumped down to microTorr range was back filled with Ar or Ar/ N2 mixture and the pressure was maintained at ~5mT. A shutter was used to separate the guns and the substrates. Both the sputtering guns and the thickness monitors were cooled using a closed circuit cooling system. Initial set of runs were done to calibrate the process for a bandgap of 1.7eV. The required bandgap was obtained by varying the power applied to the guns, which controlled the deposition rate of the CdTe and ZnTe deposited on the substrate. A thickness of 4-5 microns was maintained throughout the experiment. 4.5 Zinc Telluride Deposition ZnTe:N2 produced by sputtering has been used as a back contact for CdTe solar cells. As it is transparent to photons below 2.2eV, it is a candidate for back contact/tunnel junction

PAGE 48

37 in the tandem cells using CdTe and CZT top cells [16]. ZnTe was deposited by RF magnetron sputtering in the ambient of Ar or Ar/ N2 mixture. The temperature of ZnTe deposition was varied between 300 -400oC. The pressure was maintained at ~5mT. The thickness of ZnTe obtained was varied from 0.51 m. 4.6 Zinc Chloride Treatment The Zinc Chloride (ZnCl2) treatment is a post deposition annealing treatment done for CZT or CZT/ZnTe layers. Zinc Chloride treatment was done by 3 methods € Air annealing € Vapor Treatment in the oven € Oxygen Free Sublimation 4.6.1 Air Annealing In this process the CZT samples are soaked in a solution of ~ 0.2 mol of ZnCl2 for a period of 5 minutes. Then they are left to dry. The samples are annealed in air on a graphite plate for 20 minutes. The temperature of the annealing process is varied between 380 – 420oC. 4.6.2 Vapor Treatment in the Oven In vapor treatment method ZnCl2 powder is used as the source which is taken in a graphite boat and the source is covered by a graphite cloth to avoid ZnCl2 powder getting deposited on the samples due to high temperatures of annealing. The sample and the source are separated by spacers. The whole setup is placed in a reactor tube and kept in an oven and annealed. The annealing time is varied from 5 – 10 minutes and the temperature is varied from 380 – 420oC. The samples are cleaned with methanol after ZnCl2 treatment to remove the excess ZnCl2 on the surface. The figure 26 shows the setup of vapor treatment in the oven.

PAGE 49

38 Figure 26. Vapor Treatment in an Oven 4.6.3 Oxygen Free Sublimation The oxygen free sublimation ZnCl2 treatment consists of a source which is made of ZnCl2 / methanol mixture. The source is made at temperature slightly higher than the room temperature, so that methanol can evaporate easily and ZnCl2 alone remains on the 7059 glass. Since ZnCl2 is anhydrous the samples are immediately loaded over the source, separated by spacers and loaded in the chamber and pumped down. The chamber is maintained at 100oC – 120oC during pump down. The sublimation process takes place at a temperature of 300oC – 400oC and in the ambient of UHP Helium. The annealing time is varied from 510 minutes. The samples are cleaned with methanol after annealing to remove excess anhydrous ZnCl2. Figure 27 shows the ZnCl2 chamber used.

PAGE 50

39 Figure 27. Chamber for Oxygen Free ZnCl2 Process 4.7 Back Contact The CZT sample is back contacted with doped graphite, Cu, ZnTe/doped graphite or ZnTe/Cu. The cell areas of the samples were defined using Kapton tape. The mixture of HgTe:Cu doped with graphite paste was applied over the exposed areas. The samples are dried for atleast six hours before the tapes are removed. Copper was deposited by RF magnetron sputtering with metal Cu target which is 99.999% pure. The cell area was defined by masks. The deposition was done at room temperature and in Ar ambient. The pressure was maintained between 3 -5mT during the deposition. The thickness of the sputtered Cu was varied from 40 100.The mask is removed and excess of CZT was removed to define cell areas and to expose TCO. Indium was melted on sides of the cell for better front contact.

PAGE 51

40 4.8 Measurements 4.8.1 I-V Measurements All the devices fabricated were subjected to both dark and light current-voltage (I-V) measurements. Dark I-V measurements were done covering the devices with a cloth and light I-V measurements were done at AM1.5 solar spectrum which was achieved using four OSRAM halogen lamps. The current was measured using a 2410 Keithley source meter. The range of the sweeping voltage across the devices is set using a labview program which is connected with the source meter. The program calculates Voc, Jsc, FF and Efficiency of the devices along with their I-V plots. 4.8.2 Spectral Response The spectral response of a solar cell is given by the variation of the short circuit current as a function of wavelength of the incident light. An Oriel Instruments spectrometer with grating is used to produce the required monochromatic beam of light to obtain the corresponding spectral response. Silicon is used as the standard source and the output value of the devices are matched with the standard value. A plot is drawn between quantum efficiency and the wavelength. 4.8.3 XRD, SEM and AFM X-ray Diffraction (XRD), Scanning Electron Microscopy (SEM), Atomic Force Microscopy (AFM) measurements were done to characterize the structural properties of the CZT films deposited .XRD measurement was done using a Philips X’pert Pro X-ray diffraction system. SEM micrographs were taken using a Hitachi S-800T SEM system. AFM images were scanned using Dimension 3100 Series Scanning Probe Microscope.

PAGE 52

41 CHAPTER 5 RESULTS AND DISCUSSION This chapter deals with the deposition and measurement of CZT solar cells deposited in Ar and Ar/N2 ambient. The effect of back contacts was also studied. The results of post deposition annealing of CZT done in ZnCl2 are also discussed. 5.1 Properties of CZT Films The first few runs of CZT were done to calibrate the process for a bandgap of 1.7eV. The deposition of CZT was done on various substrates such as 7059/SnO2/CdS, 7059/SnO2, 7059/ITO/CdS, 7059/ITO/ZnSe and 7059 glass. CZT was deposited on these substrates by co-sputtering of CdTe and ZnTe targets. The total pressure during deposition was maintained at 5mT. The deposition temperature was maintained at 350oC throughout the run. The thickness of the CZT film was 4-5 m, which was measured using a profilometer. Compositional and thickness uniformity was achieved by substrate rotation. The films were free of pinholes. Optical transmission measurements were done on all CZT films as a quality control tool. Transmission was done at different points over the deposition area, to confirm that, the bandgap and the composition at all the areas was the same. It was found that there is a small variation of 0.2-0.3 eV in bandgap. Figure 28 shows the transmission response on CZT sample deposited in Ar ambient. The transmission of the CZT film was found to be between 70-80%.

PAGE 53

42 0 10 20 30 40 50 60 70 80 90 100 40050060070080090010001100 Wavelength (nm)Transmission (%) Figure 28. Transmission Response of CZT Film Deposited in Ar with TSUB ~ 350oC 0.E+00 1.E+08 2.E+08 3.E+08 4.E+08 5.E+08 6.E+08 1.001.101.201.301.401.501.601.701.801.902.00 Photon energy (eV)2 (cm-2) Figure 29. Bandgap of CZT Film Deposited in Ar with TSUB ~ 350oC

PAGE 54

43 The bandgap of the samples was determined by plotting 2 against photon energy. Figure 29 shows the bandgap of a CZT film. The bandgap of the CZT films obtained during this work were in the range of 1.68-1.8eV. X-ray diffraction measurements were also done on CZT films to find the zinc concentration. The value of percentage of zinc in CZT (x) was calculated using Vegards law. a = 6.481 – 0.3837(x) where a is the lattice constant in , which was calculated from the XRD measurements using the equation, l 2 k 2 h 2 a d spacing + + = where dspacing is the interplanar distance in and (hkl) are the miller indices. The value of x ranged from 0.4 to 0.6. The films were found to exhibit preferential orientation along the (111) direction. Figure 30 shows the XRD of an as-deposited CZT film deposited on SnO2/CdS substrate. A mixed phase of CdTe with very low peak intensity is also found. Figure 30. XRD of CZT Film on CdS/SnO2 Substrate

PAGE 55

44 Figure 31. AFM Image of CZT Film on CdS Substrate Figure 31 shows the AFM image of CZT film deposited over a CdS substrate at 400oC. The RMS roughness for these films was approximately 440nm. The films were densely packed and pinhole free. Figure 32. SEM Image of CZT Films Deposited at 300oC

PAGE 56

45 Figure 33. SEM Image of CZT Films Deposited at 400oC SEM analysis was done on CZT deposited at different temperatures. Figure 32 and 33 show SEM images of films deposited at 300oC and 400oC temperatures. From the images the surface appears as clusters of small grains with variation in particle size between the 300oC and 400 oC deposited films. No pinholes were found in the films. Based on the images it appears that grain size decreases with increasing deposition temperature. 5.2 CZT in Argon Ambient This section describes device results obtained from CZT films sputtered in Ar. 5.2.1 Effects of Contact Anne aling on Cell Characteristics The effect of contact annealing on CZT cells was studied for different temperatures. The substrates used were SnO2/ CdS (CBD) and SnO2/CdS (CSS). Contacts were applied at room temperature. Doped graphite (see section 4.7) was used as the back contact material.

PAGE 57

46 The Voc for SnO2/ CdS (CBD)/CZT devices was in the range of 220-240mV and for SnO2/CdS (CSS)/CZT devices was in the range of 230-260mV. The contacts were subsequently annealed at 150oC for 15 minutes and the devices were measured. It was observed that there was no effect on the Voc and Jsc values of the devices. Since the contact annealing temperatures of 150oC had no effect on device performance, higher temperatures were considered. Contacts were annealed at a higher temperature of 270oC. The J-V characteristics of the devices remained the same. Performance before and after annealing is shown in figure 34. The Voc of the devices does not change with increase in annealing temperature. The Jsc values of these devices were very low and did not improve with annealing. Contact annealing was found to have no effect on the CZT devices sputtered in Ar. -1.0E-03 -5.0E-04 0.0E+00 5.0E-04 1.0E-03 1.5E-03 2.0E-03 2.5E-03 3.0E-03 -1.00-0.500.000.501.00 Voltage [Volts]Current Density [A/cm2] SnO2/CdS (CBD) As dep SnO2/CdS (CBD) Ann SnO2/CdS (CSS) As dep SnO2/CdS (CSS) Ann Figure 34. Light I-V Characteristics of Ba ck Contact As-deposited and Annealed Devices 5.2.2 Heat Treatment of CZT Devices As discussed in the previous section the as-deposited CZT devices at 350oC were poor. SnO2/CdS/CZT films were annealed in an inert ambient of He at different temperatures to investigate the effect of a simple heat treatment on solar cell performance. Different annealing temperatures from 400 -500oC and different time durations were used. It was

PAGE 58

47 observed that a slight amount of CZT evaporated during the annealing process. This was found from deposition on a glass kept at a distance of 2mm above the films during annealing. Annealing temperatures above 500oC did not yield better results because of the evaporation of CZT. Table 6 shows the Voc of the devices at different annealing temperatures and time. Table 6. Voc of Devices at Different Annealing Temperatures Annealing Temperatures (oC) Voc (mV) As-deposited* 240 400 oC / 30 mins 290 500 oC / 5 mins 430 500 oC / 30 mins 440 (* As-deposited Voc value obtained from a different substrate) The annealed devices had better characteristics than the as-deposited devices. The Voc for the annealed films at 500oC were around 440mV. But there was no significant improvement in Jsc values even after annealing and remained the same as the as deposited devices. Though the post heat treatment of the films increased the Voc values, it had no effect on the Jsc values of the devices. 5.3 CZT Deposited in Ar/N2 Ambient Since the performance of as-deposited and annealed CZT devices deposited in Ar ambient were poor, other ways to deposit CZT to modify its properties were tried. The next step was to increase the deposition temperature of CZT and to introduce nitrogen as ambient along with Ar. A series of experiments were done at different partial N2 pressures of 10%,

PAGE 59

48 25%, 50%, 75% and it was found that 50% N2 with 50% Ar produced results with improved characteristics [19]. Using Ar/N2 (50%) ambient CZT was deposited over ITO/CdS and SnO2/CdS substrates. The temperature was maintained at 400oC and the thicknesses of the films were 4-5 m. The films were uniform and pinhole free. The CZT films were fabricated to devices using doped graphite as back contact. The films had a post deposition annealing at 500oC for 5 minutes in He before the back contact was applied. The ITO/CdS/CZT samples had Voc in the range of 360-420mV and Fill Factors in the range of 33-37%. The current densities from these devices were very low. The SnO2/CdS/CZT devices had better performance than ITO/CdS/CZT devices. The Voc’s were in the range of 480-510mV and the fill factors were in the range of 22-26%. The Jsc from spectral response was approximately 2.0mA/cm2. The dark and light I-V characteristics in figure 35 show how resistance effects (shunt and series) dominate the device characteristics. Figure 35. Dark and Light I-V Characteristics of SnO2/CdS/CZT in Ar/N2 and Annealed in He at 500oC for 5 Mins -3.0E-03 -2.2E-03 -1.4E-03 -6.0E-04 2.0E-04 1.0E-03 1.8E-03 -1.00-0.500.000.501.00 Voltage [Volts]Current Density [A/cm2] -1.0E-05 -5.0E-06 0.0E+00 5.0E-06 1.0E-05 1.5E-05 2.0E-05 -1.00-0.500.000.501.00 Voltage [Volts]Current Density [A/cm2]

PAGE 60

49 0% 5% 10% 15% 20% 400500600700800900 Wavelength, (nm)Q.E. Figure 36. Spectral Response of SnO2/CdS/CZT in Ar/N2 and Annealed in He at 500oC for 5 Mins From the spectral response of SnO2/CdS/CZT devices of figure 36, shows low collection at all wavelengths and wavelength dependence. It can be seen that the collection in the CZT cells is extremely poor and suspected that CZT may be largely defective. 5.3.1 Effects of CdS Thickness on CZT Devices Though CZT was deposited on various window layers like CdS, ZnSe, and ZnO, CZT deposited on CdS had better performance than any other layers. Most of the work in this thesis was done on CdS/CZT structures. Poor spectral response on the SnO2/CdS/CZT devices also suspected the CdS/CZT interface for defects. So the thickness of the CdS was varied to examine if the thickness

PAGE 61

50 would affect the interface properties. CZT was deposited over approximated thickness of 500, 600, 700, 800 CdS. Figure 37 shows the Voc and Jsc values for various thickness of the CdS layer. 400 420 440 460 500600700800Thickness ()Voc (mV)1 1.2 1.4 1.6 1.8 2Jsc (mA) Voc (mV) Jsc (mA) Figure 37. Effect of CdS Thickness on CZT The change in Voc and Jsc values was small when the thickness of the CdS layer was varied. The Fill Factors were also in the 30-34% range with the 800 thick CdS layer having a Fill Factor of 34%. The spectral response of CZT devices on different CdS thickness are shown in figure 38.

PAGE 62

51 0% 5% 10% 15% 20% 400500600700800900 Wavelength, (nm)Q.E. 500 600 700 800 Figure 38. Spectral Response of CZT with Different CdS Thicknesses 5.4 CZT on Transparent Conducting Oxides Cadmium zinc telluride films were also deposited directly on different TCO layers without a window layer such as CdS. The films were deposited in the temperature range of 200-400oC in Ar/N2 ambient. SnO2 was deposited by MOCVD and ITO by sputtering. The thickness of CZT was 4-5 microns. Transmission response measurements were done on the films and the overall transmission was in the range of 60-70%. XRD measurements were done on SnO2/CZT films deposited at 200oC and annealed at different temperatures to study the effect of post deposition annealing. The concentration of zinc was found to be in the range of 0.45-0.6. Figure 39 shows the XRD of SnO2/CZT films in the as-deposited, annealed at 250oC for 2 hours, and annealed at 450oC for 5 minutes.

PAGE 63

52 Figure 39. XRD of SnO2/CZT at Different Annealing Temperatures Preferred orientation of (111) of the CZT was obtained with a mixed phase of CdTe and ZnTe at all conditions. Annealing at higher temperature gives rise to new peaks and improves the crystallinity of the film which can be seen from the increased peak intensities of the 450oC annealed film. The peak intensities of the film annealed at 250oC for 2 hours are low compared to the as-deposited film which shows that annealing for longer time will degrade the films. Doped graphite was applied as back contact to complete the cell structures. The ITO/CZT devices exhibited poor performance with very low current densities. The SnO2/CZT devices had a better current in the range of 2-2.5mA, with Voc ranging from 240-270mV and Fill Factor of 25-26%. The spectral response on SnO2/CZT device seen in figure 40 shows that these device too exhibit similar poor collection characteristics as cells described in the previous section.

PAGE 64

53 0% 5% 10% 15% 20% 25% 30% 400500600700800900 Wavelength, (nm)Q.E. Figure 40. Spectral Response of CZT on SnO2 Substrate 5.5 Transparent Back Contact on CZT Films In addition to graphite, copper was also used as the back contact material. Cu was deposited by RF magnetron sputtering at room temperature in Ar and at a pressure of 35mTorr. The thickness of the Cu contact was varied from 40-80. SnO2/CdS/CZT films in the as-deposited and post deposition annealed conditions were used. The Voc increased with annealing of the CZT films, but the currents were better in the as-deposited condition. Table 7 shows the values of Cu back contact devices. Table 7. Voc and Jsc of Copper Back Contact Devices Voc (mV) Jsc (mA/cm 2 ) As-Deposited 330360 0.65 Annealed 420460 0.2

PAGE 65

54 ZnTe with a bandgap of 2.2eV can be used as a transparent back contact material for CZT as the top cell of a tandem solar cell structure. A layer of ZnTe was deposited as back contact over CZT film in Ar and in Ar/N2 ambient to dope it more p-type. Cu was sputtered on ZnTe for better ohmic contact. Back contact was annealed at 150oC for 10 minutes in He. Voc for various conditions are shown in table 8. It was seen that doping of ZnTe more p-type with N2 helps in increasing current density of the cell. Better currents were obtained at as-deposited condition of CZT/ZnTe in the Ar/N2 ambient and high voltages at post deposition annealed conditions. Table 8. Voc of Copper Back Contacted Samples at Various Conditions SnO2/CdS/CZT ZnTe Voc (mV) Jsc (mA/cm2 ) As-Deposited Argon 380410 0.2 As-Deposited Argon/ Nitrogen 420 1.42 Annealed Argon 480 0.1 Annealed Argon/ Nitrogen 520570 0.25 5.6 CZT Graded Devices Grading of CZT film was done starting from CdTe to ZnTe (type A devices) and ZnTe to CdTe (type B devices). Grading was done on CdS/SnO2 and SnO2 substrates in Ar/N2 ambient. Grading of type A samples started with deposition of CdTe with deposition rate as used for CZT deposition and zero deposition rate of ZnTe. After 5 minutes, the CdTe deposition rate was reduced by 5 Watts and ZnTe deposition rate was increased by 6 Watts and this was done till CdTe deposition rate became zero and ZnTe deposition rate

PAGE 66

55 was normal deposition rate of CZT. Type B samples were deposited starting with ZnTe and ending with CdTe. The films were pinhole free and uniform. XRD measurements done on the type A and type B samples can be seen in figure 41. Figure 41. XRD of Graded Samples XRD figures show that both type A and B films have (111) orientation with a low peak intensity. As the CZT film deposited was graded, the lattice constant of the CZT changed as the ratio of CdTe and ZnTe was varied gradually. Mixed phase of ZnTe was also found with low peak intensity. Films were contacted in as-deposited and annealed conditions. Type A devices in the asdeposited condition resulted in better Voc and Jsc than annealed devices. This may be due the CdS/CdTe interface formed as in standard CdTe solar cells. The degrading on performance of devices in the annealed condition is not known. In Type B devices slight improvement were found in post deposition annealed devices. This may be due to the CdS/ZnTe interface formed as a result of annealing. However Type B devices had poor performance compared to Type A devices. Table 9 shows the Voc and Jsc of Type A devices.

PAGE 67

56 Table 9. Voc and Jsc of Type A Graded Devices It can be seen from the spectral response measurement in figure 42 that there is an increase in collection in type A devices after the CdS wavelength due to the presence of CdTe layer. But the collection is low due the very thin layer of CdTe present. There is no cut-off in collection after the CZT bandgap which also suggests that the collection is due to CdTe. Type A devices have a bandgap close to CdTe. In the case of type B devices there was poor collection and the bandgap was moved towards the blue region due to the presence of Zn. 0% 4% 8% 12% 16% 400500600700800900 Wavelength, (nm)Q.E. Type B (Post deposition annealed) Type A (As deposited) Figure 42. Spectral Respon se of Graded Samples Voc (mV) Jsc (mA/ cm 2 ) As-Deposited 450 2.83 SnO2/CdS Annealed 350 < 1 As-Deposited 250 1.95 SnO2 Annealed

PAGE 68

57 5.7 Post Deposition Treatments of CZT Post deposition treatments of CdCl2 on CdTe solar cells improved the efficiencies of the CdTe devices to a great extent. Likewise various post deposition treatments were carried on with CZT to boost the performance of the devices. As CdCl2 has enhanced the CdTe performance, CdCl2 post deposition was done on the CZT films. But the films peeled off when cleaned in methanol after the post-deposition treatment. This may be due to reaction of CdCl2 with zinc in CZT in the formation of ZnCl2, which reduces the zinc concentration in the film and may be a cause for lift off of the CZT layer [18]. A mixture of CdCl2 and ZnCl2 were taken in a 3:1 ratio in a graphite boat and post deposition treatment was done over CZT film at 400oC for 10 minutes in an oven. The devices fabricated from this method showed a small increase in Voc, but there was no increase in Jsc. The J-V measurements revealed essentially no changes in PV performance using this annealing process. Zinc and its halides were the next set of materials used for the post deposition treatment, since cadmium halides can react with zinc in CZT to form zinc halides which may lead to reduction of zinc in CZT. Zinc chloride was the material tried. 5.8 Zinc Chloride Treatment on CZT Films Zinc chloride post deposition treatment was first done on ITO/ZnSe/CZT substrates. These devices were selected for their high Voc values. The first method was solution dipping and air annealing method where the samples are dipped in a ~ 0.2 mol ZnCl2 solution for 5 minutes and left till methanol evaporates and annealed in air at 380oC for 20 minutes. The films were cleaned in methanol to remove excess ZnCl2 on the CZT and doped graphite was used as back contact. Measurements showed a slight improvement in Jsc values compared to the nonZnCl2 treated samples. However the Voc of the devices showed no improvement.

PAGE 69

58 Since SnO2/CdS/CZT substrate had better currents compared to other substrates, ZnCl2 annealing treatment was carried out on them. The post deposition annealed substrates had better performance than the as-deposited substrates. So a two stage annealing process was used where the CZT films were post deposition annealed at 500oC for 5 minutes, before dipping in ZnCl2 solution. Two different air annealing temperatures 380oC and 360oC were used to anneal the substrates. The devices annealed at 380oC gave better Jsc, more than 1mA than the devices annealed at 360oC. Voltage and current from different annealing temperatures are shown in figure 43. 60 70 80 90 100 110 120 130 140 150 160 360380 Temperature (oC)Voc (mV)1 1.5 2 2.5 3 3.5 4Jsc (mA) Voc (mV) Jsc (mA) Figure 43. Voc and Jsc of ZnCl2 Air Annealed Samples at Different Temperatures

PAGE 70

59 0% 5% 10% 15% 20% 25% 400500600700800900 Wavelength, (nm)Q.E. Figure 44. Effects of ZnCl2 Air Annealing Treatment The spectral response curve in figure 44 show that there is a shift in the bandgap of the device which may be due to the loss of zinc concentration in the CZT material. This may be due to reaction of zinc on the surface of CZT with oxygen in the air to form zinc oxide. The air annealing process was changed by vapor deposition of ZnCl2 in an oven. Here ZnCl2 powder was used as source in a graphite boat. The source and substrate were separated by spacers. Different temperatures 380oC, 400oC and 420oC were used for vapor deposition. SnO2/CdS/CZT structures used were post deposition annealed at 450oC in H2 ambient and in O2 ambient. All O2 annealed devices measured after ZnCl2 treatments were shorted, this may be due to formation of ZnO. The H2 annealed films vapor treated at 420oC had poor devices, due to over exposure to ZnCl2 at high temperature. Comparison of the Voc and Jsc by ZnCl2 treatment at different temperatures is shown in figure 45.

PAGE 71

60 430 440 450 460 470 480 490 500 As Deposited380/ 10 min400/ 10 min ZnCl2 Annealing TemperaturesVoc (mV)1 1.2 1.4 1.6 1.8 2Jsc (mA) Voc (mV) Jsc (mA) Figure 45. Effect of ZnCl2 Annealing Temperatures It is evident from the figure that both 400oC and 380oC post deposition treatment conditions have similar results, but there was no variation in the bandgap as in air annealing method which can be seen from the spectral response figure 46. Copper back contact was also deposited for the two stage hydrogen annealed and ZnCl2 annealed samples, but all the devices were shorted. 0% 5% 10% 15% 20% 400500600700800900 Wavelength, (nm)Q.E. Figure 46. Spectral Response of CZT Device by Vapor Treatment

PAGE 72

61 5.8.1 ZnCl2 Treatment of CSS-CZT Substrates ZnCl2 post deposition treatment was also carried out on CZT films prepared by the CSS process. CZT was deposited by co-sublimation of CdTe and ZnTe in 1Torr He. ZnTe was deposited over CZT at 2 Torr pressure in H2. The source temperatures were between 600oC -700oC and substrate temperature around 500oC600oC. ZnTe was deposited over the CZT films to plug the pinholes, since a significant concentration of pinholes were present in the CZT films [17]. As-deposited and post deposition hydrogen annealed samples were used for ZnCl2 treatment. ZnCl2 was vapor treated at 400oC for 10 minutes. Figure 47 shows the Voc and Jsc of the as-deposited and post deposition annealed CZT/ZnTe devices. 630 640 650 660 670 680 690 700 As Deposited Post Deposition AnnealedVoc (mV)0 0.5 1 1.5 2 2.5 3 3.5Jsc (mA) Voc (mV) Jsc (mA) Figure 47. ZnCl2 Vapor Treated CSS CZT Devices

PAGE 73

62 0% 10% 20% 30% 40% 400500600700800900 Wavelength, (nm)Q.E. Figure 48. Spectral Response of Vapor Treated CSS CZT Devices Though there was a significant difference in Voc values, the post deposition annealed device had current 2.5 times greater than the as-deposited device. The Fill Factors for both the devices were in the range of 31-32%. The spectral response curves for samples are shown in figure 48. The third process was ZnCl2 treatment by oxygen free sublimation. This was done in UHP helium ambient to avoid the reaction of oxygen with zinc resulting in ZnO. Temperature of 400oC was used.ZnCl2 post deposition was done on CZT and CZT/ZnTe. Group A samples are SnO2/CdS/CZT structures ZnCl2 post deposition treated and a layer of ZnTe deposited on it. Group B samples are SnO2/CdS/CZT/ZnTe structures ZnCl2 post deposition treated. A ZnCl2 source was made on 7059 glass from ZnCl2: methanol mixture. The films were ZnCl2 treated in as-deposited and post deposition annealed in hydrogen. Table 10 shows the Voc and Jsc values of measured devices.

PAGE 74

63 Table 10. CSS ZnCl2 Treated Oxygen Free Sublimation CZT Samples Voc (mV) Jsc (mA/cm2) (CdS/CZT)As-deposited/ ZnCl2 treated/ZnTe 180 0.37 Group A (CdS/CZT)H2 annealed/ ZnCl2 treated/ZnTe 590 1.92 (CdS/CZT/ZnTe)Asdeposited/ ZnCl2 treated 630 0.15 Group B (CdS/CZT/ZnTe)H2 annealed/ ZnCl2 treated 610 2.20 The spectral responses of the post deposition hydrogen annealed devices are shown in figure 49. Figure 49. Spectral Response of Post Deposition Hydrogen Annealed and ZnCl2 Treated Devices Poor results were obtained for samples annealed at 450oC. 0% 5% 10% 15% 20% 25% 30% 400500600700800900 Wavelength, (nm)Q.E. Group A Group B

PAGE 75

64 CHAPTER 6 CONCLUSION Successful depositions of CZT films were done in Ar and Ar/N2 mixture by co-sputtering CdTe and ZnTe targets. Uniform and pinhole free films were obtained. Films with optical transmission of 70-80% and with bandgap in the range of 1.65-1.8eV were fabricated. From XRD analysis preferential (111) orientation of CZT was obtained. Mixed phase of CdTe and ZnTe were also observed along with CZT and SnO2 peaks. Post deposition annealing increase the CZT peak intensity along with other peak intensities. SEM measurements on CZT films showed clusters of small grains at low temperature however free of pinholes. And grain size decreased with higher deposition temperature. CZT devices fabricated in Ar ambient showed poor performance. Contact annealing and post deposition heat treatments on CZT films had no effect on performance of the device. Deposition of CZT in Ar/N2 mixture of 50% improved the device performance. Post deposition annealing of films lead to loss of few of the film. Of all the window layers used cadmium sulfide was the best heterojunction partner with CZT. Thickness of CdS layer had no major effect on the performance of the devices. CdS with a thickness of 800 was mostly used. The best CdS/CZT devices had Voc in the range of 530mV and Jsc of 2.1mA. SnO2 /CZT structures were fabricated to see the performance of cells without window layer. These devices too exhibited poor collection as devices with window layer. Best Voc of 250mV and currents upto 3mA were obtained from the SnO2/CZT devices. Cu and ZnTe were used as alternative back contacts for graphite.ZnTe:N2 transparent back contact with Cu gave better Jsc in as-deposited and better Voc in the post deposition annealed conditions.

PAGE 76

65 Grading of films from CdTe to ZnTe and ZnTe to CdTe were deposited and fabricated to devices. CdTe to ZnTe devices had better performance in the as-deposited condition due to presence of a thin layer of CdTe interface with CdS. From spectral response it was seen that the bandgap was close to CdTe. ZnTe to CdTe devices showed small improvement with post deposition annealing of the films. As CdCl2 treatment of CZT films led to the loss of Zinc as ZnCl2 from CZT films, ZnCl2 post deposition treatments were done on co-sputtered CZT films and co-CSS CZT films. Air annealing treatments of CZT with ZnCl2 led to loss of Zn with oxygen by forming ZnO. Vapor treatment method had better voltages and currents. Oxygen free sublimation of ZnCl2 was done in He ambient. Efforts to implement oxygen free sublimation ZnCl2 treatment on sputtered CZT films did not succeed as the films peeled off. Though ZnCl2 post deposition treatments did not help in improving the device performance to a large extent, they helped in slight improvement on performance of the devices.

PAGE 77

66 REFERENCES [1]. “The Sun”, Detailed Information of the Sun [http://seds.lpl.arizona.edu/nineplanets/nineplanets/sol.html]. [2]. Alan L. Fahernbruch, Richard H.Bube “Fundamentals of Solar Cells”, Academic Press, p. 26, 1983. [3]. X.Wu, R.G. Dhere, D.S. Albin, T.A.Gessert, “ High-efficiency CTO/ZTO/CdS/CdTe Polycrystalline Thin-film Solar Cells”, NCPV Program Review Meeting, Lakeland, CO, Oct 14-17, 2001. [4]. K.Ramanathan, A. Contreras, C. Perkins, S. Asher, “Properties OF 19.2% Efficiency ZnO/CdS/CIGS Thin-film Solar Cells”, Prog. in Photovoltaics, Volume 11, Issue 4, pp 225-230, June 2003. [5]. Keld West, “Solar Cells Beyond Silicon”, Ris National Energy Conference, May 2003. [http://www.solarcell.dk/solarcellsbeyondsilicon.pdf]. [6]. “Tandem Cell Photovoltaic”, National Renewable Energy Laboratory [http://www.nrel.gov/technologytransfer/success_stories.html]. [7]. “Basic Physics and Design of III – V Multijunction Solar Cell”, [http://www.nrel.gov/ncpv/pdfs/11_20_dga_basics_9-13.pdf]. [8]. Matthew Buresch, “Photovoltaic Energy Systems”, McGraw-Hill Book Company, p79. [9]. Joseph A. Merrigan, “Prospects for Solar Energy Conversion by Photovoltaics”, The MIT Press, p35. [10]. 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. [11]. W.Shafarman, M.Gossla and B. McCandles, “Wide Band Gap CuInSe2 and CdTebased Thin Film for Tandem Solar Cells”, Institute of Energy Conversion, University of Delaware. [ http://www.nrel.gov/ncpv_prm/pdfs/papers/76.pdf]. [12]. A.Rohatgi, R. Sudarshanan, S.A.Ringel, P.V.Meyers and C.H.Liu, “Wide Bandgap Thin Film Solar Cells from CdTe Alloys” 20th IEEE Photovoltaic Specialists Conference, p1477-1481, 1988.

PAGE 78

67 [13]. A.Rohatgi, R. Sudarshanan, S.A.Ringel, P.V.Meyers and C.H.Liu, “Investigation of Polycrystalline CdZnTe, CdMnTe and CdTe Films for Photovoltaic Applications”, Solar Cells, vol 24 pp 285-194, 1988. [14]. Bulent M.Basol, Vijay K.Kapur, Richard L.Mitchell, “Cd1-xZnxTe Obtained by the Solid-State Reaction of Elemental Layers”, 21st IEEE Photovoltaic Specialists Conference, pp 546-549 vol.1, May 21-25, 1990. [15]. Ramesh Dhere, Tim Gessert, Jie Zhou, Sally Asher, Joel Pankow and Helio Moutinho, “Investigation of CdZnTe For Thin-Film Tandem Solar cell Applications”, Material Research Society Spring Meeting, San Francisco, CA, April 21-25, 2003. [16]. J. Drayton, C. Taylor, A. Gupta, R.G. Bohn, A.D. Compaan, B.E. McCandless and D. Ross, “Optical, Structural and transport properties of Reactively Sputtered ZnTe:N”, National Center for Photovoltaics Review Meeting, Lakewood ,CO, Oct 14-17, 2001. [17]. Gowri Sivaraman, “Characterization Cadmium Zinc Telluride Solar Cells”, Masters Thesis, University of South Florida, 2003. [18]. Brain E.McCandless, “Cadmium Zinc Telluride for Wide Band Gap Solar Cells”, 29th IEEE Photovoltaic Specialists Conference PVSC, New Orleans, 2002. [19]. Jagadish Gaduputi, “Characterization of Cadmium Zinc Telluride Films And Solar Cells on Glass and Flexible Substrates By RF Sputtering”, Master Thesis, University of South Florida, April 1, 2004. [20]. N.B.Chaure, Shwetw Chaure, R.K Pandey, “Investigation of Non Aquous Electrodeposition CdS/CZT Heterojunction Solar Cells”, Solar Energy Materials and Solar Cells 82, p 39-60, 2004. [21]. I. Clemminck, M.Burgelman, M. Cassteleyn, J. De Poorter, A. Vervaet, “Interdiffusion of CdS and CdTe in screenprinted and sintered CdS-CdTe Solar Cells”, 22nd IEEE Photovoltaic Specialist Conference, p1114-1119 vol 2, 7-17 Oct 1991. [22]. C. Ferekides, J. Britt, Y. Ma and L. Killan, “High Efficiency CdTe Solar Cells By Closed Space Sublimation”, IEEE Photovoltaic Specialist Conference, 1993. [23]. Frederik C. Kerbs, Jan Alstrup, Holger Spanggaard, Kaj Larsen, Esben Kold, “Production Of Largr-area Polymer Solar Cells By Industrial Silk Screen Printing, Lifetime Consideration And Lamination with Polyethyleneterephthalate”, Solar Energy Materials And Solar Cells.

PAGE 79

68 [24]. S.M.Sze, “Physics of Semiconductor Devices”, John Wiley and Sons, pp.122-126. [25]. Hans Joachim Moller, “Semiconductor For Solar Cells”, Artech House,Inc, pp. 58– 60, 1993. [26]. Pallab Bhattacharya, “Semiconductor Optoelectronic Devices”, Prentice-Hall Inc, pp. 194-197, 1997.

PAGE 80

APPENDICES 69

PAGE 81

APPENDIX A 70

PAGE 82

APPENDIX B 71

PAGE 83

Appendix B (continued) 72

PAGE 84

Appendix B (continued) 73