i

Design and Analysis of a High Efficiency and Ultra

Thin ZnxCd1-xS/ZnCdTe Solar Cell

Md. Muin Uddin

Dhaka University of Engineering and Technology, Bangladesh ii

Design and Analysis of a High Efficiency and Ultra

Thin ZnxCd1-xS/ZnCdTe Solar Cell

THESIS SUBMITTED IN FULFILMENT FOR THE DEGREE OF

MASTER OF SCIENCE IN ELECTRICAL AND ELECTRONIC ENGINEERING

By Md. Muin Uddin

Department of Electrical and Electronic Engineering Dhaka University of Engineering and Technology, Bangladesh 2015 iii

CERTIFICATION

The thesis titled “DESIGN AND ANALYSIS OF A HIGH EFFICIENCY AND ULTRA THIN

ZnxCd1-xS/ZnCdTe SOLAR CELL” submitted by Md. Muin Uddin, Roll No.: 112221P, Session: 2011-2012 has been accepted as satisfactory in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE IN ELECTRICAL & ELECTRONIC ENGINEERING on 02 April, 2015.

BOARD OF EXAMINERS

1.______Dr. Md. Anwarul Abedin Chairman Professor & Head Department of Electrical and Electronic Engineering Dhaka University of Engineering & Technology Gazipur-1700, Bangladesh

2.______Dr. Md. Bashir Uddin Member Professor Department of Electrical and Electronic Engineering Dhaka University of Engineering & Technology Gazipur-1700, Bangladesh

3.______Dr. Md. Shaheen Hasan Chowdhury Member Professor Department of Electrical and Electronic Engineering Dhaka University of Engineering & Technology Gazipur-1700, Bangladesh

4.______Dr. Md. Sharafat Hossain Member Associate Professor (Supervisor) Department of Electrical and Electronic Engineering Dhaka University of Engineering & Technology Gazipur-1700, Bangladesh

5. ______Dr. Kazi Khairul Islam Member Professor (External) Department of Electrical and Electronic Engineering Islamic University of Technology, Gazipur-1704, Bangladesh iv

DECLARATION

I hereby declared that this thesis or any part of it has not been submitted elsewhere for the award of any degree or diploma.

02 April, 2015 Md. Muin Uddin 112221P

v

ACKNOWLEDGMENT

First of all, I would like to express my deepest gratitude to Almighty ALLAH (S.T.) who helped, supported and guided me by every means throughout this work. The author would like to take this opportunity to thank everyone who has given significant help and support for this work.

I would like to convey my sincere gratitude and profound respect to my thesis supervisor Dr. Md. Sharafat Hossain, Associate Professor, Department of Electrical and Electronic Engineering (EEE), Dhaka University of Engineering and Technology (DUET), for his continuous guidance, suggestion and whole hearted supervision throughout this research work. I am grateful to him for his encouragement and patience. I would like to thank him for giving me extra sessions from his valuable time, even in holidays. I am deeply indebted to him in fulfilling this research work which would be impossible to materialize without his support. I also express my appreciation to Dr. Md. Saifuddin Faruk, Professor, Department of Electrical and Electronic Engineering, Dhaka University of Engineering and Technology, Bangladesh for giving me encouragement, suggestions and invaluable source of materials to complete this work.

I would like to thank the Department of Electrical & Electronic Engineering of Dhaka University of Engineering & Technology for providing me continues support and facilities to carry on my research. I would like to thank my parents and wife for their enormous sacrifices and supports during my study. Last but not least, I would like to express my most sincere gratefulness to all those who have helped me in many different ways to complete this work but their names are not appeared here.

vi

ABSTRACT

The polycrystalline CdTe based solar cells have shown good prospects for low cost and large scale solar PV energy source. This work investigates the feasibility of the ultra thin ZnCdS/ZnCdTe solar cell with back surface field (BSF) for higher efficiency solar cells utilizing AMPS numerical simulation as the investigation tool. Among thin films, CdTe is the favorable choice due to its numerous material advantages for solar cell applications, such as optimum direct energy gap, higher absorption coefficient and its ability to keep good electronic properties at various low cost fabrication technologies. This work also investigates the prospects of zinc (Zn) incorporation in CdS/CdTe solar cell both in window layer (CdS) and absorber layer (CdTe). The main goal of this work is to enhance the performance of thin film ZnCdS/ZnCdTe solar cells with submicron absorbers, which could lead to significant reduction of production cost and wider commercial usage. Analysis shows that 0.5-0.6 μm ZnCdTe absorber layer gives good performance with suitable back surface field (BSF). The specific BSF materials selected to be investigated are ZnTe, As2Te3, Cu2Te, PbTe. The best calculated ultra thin ZnCdS/ZnCdTe solar cell with a 100 nm of ZnTe BSF layer should have an absorber layer thickness of 0.6 μm (efficiency 21.17%), higher stability with a linear temperature coefficient of -0.18%/°C and a competitive ZnCdS/ZnCdTe cell performance at a absorber thickness of 0.5 μm should be possible. Based on this research, good quality ultra thin ZnCdS/ZnCdTe solar cells are feasible with higher conversion efficiency.

vii

Contents

Page

DICLARATION iv

ACKNOWLEDGEMNT v

ABSTRACT vi

CONTENTS vii

LIST OF FIGURE xi

LIST OF TBLE xiv

LIST OF SYMBOLS xv

LIST OF ABBREVIATIONS xix

CHAPTER I INTRODUCTION 1

1.1 Motivation 1

1.2 Photovoltaic Technology 2

1.3 Problem Statement 7

1.4 Objectives of the Study 10

1.5 Dissertation Organization 12

CHAPTER II LITERATURE REVIEW 14

2.1 Introduction 14

2.2 Solar Cell Basics 14

2.2.1 Solar Cells and Solar Energy Materials 14

2.2.2 Solar Cell Operation 15

2.2.3 Solar Cell Output Parameters 16

2.2.3.1 Short Circuit Current (Isc) 17 viii

2.2.3.2 Open Circuit Voltage (Voc) 17

2.2.3.3 Fill Factor (FF) 18

2.2.3.4 Efficiency (η) 18

2.2.4 Series and Shunt Resistance (Rs and Rsh) 18

2.2.4.1 Series Resistance 19

2.2.4.2 Shunt Resistance 20

2.2.5 Diode Factor 21

2.2.6 J-V Characteristics 21

2.2.7 Quantum Efficiency 23

2.3 A Brief History on Thin Film CdTe Solar Cells 25

2.4 CdS/CdTe Solar Cell 27

2.4.1 Substrate 28

2.4.2 Front Electrical Contact 28

2.4.3 CdS Window Layer 29

2.4.4 CdTe Absorber Layer 29

2.4.5 Back Contact 30

2.5 Research Target Area– ZnxCd1-xS Window Layer and ZnCdTe 31 Absorber Layer for CdTe Solar Cell

2.6 Chapter Summary 34

CHAPTER III METHODOLOGY 35

3.1 Introduction 35

3.2 Solar Cell Modeling 36

3.3 Equations Involved for Simulation 37

3.3.1 Boundary Conditions 40

3.3.2 Solution Technique 41

3.3.3 Cell Performance Parameters Determination 42

3.4 Simulation Software and AMPS 43 ix

3.5 Initial Setting of AMPS Simulation Parameters 44

3.5.1 Material Parameters 45

3.5.2 General Device Parameters 46

3.5.3 Layer Information 47

3.6 Proposed Cell Structure and Parameter Set 48

3.7 Current-voltage (I-V) characteristics of Solar Cells 49

3.8 Chapter Summary 51

CHAPTER IV OPTIMISATION OF CELL STRUCTURE FROM 53 NUMERICAL ANALYSIS

4.1 Introduction 53

4.2 Proposed Solar Cell Structure 54

4.3 Conventional and Modified Cell Structure Analysis 56

4.4 Selection of Front Contact 57

4.5 Selection of Buffer Layer 58

4.6 Selection of Window Layer 60

4.7 Selection of Absorber Layer 61

4.8 Back Contact and Barrier Height Analysis 62

4.9 Back Surface Recombination Rate Analysis 65

4.10 Addition of BSF Layer 67

4.11 Chapter Summary 68

CHAPTER V RESULTS & DISCUSSION 69

5.1 Modeling & Simulation 69

5.2 Optimization of ZnxCd1-XS Window Layer 71

5.3 Optimization of ZnCdTe Absorber Layer 74

5.4 Analysis of ZnCdTe Cell with BSF 77

5.4.1 ZnTe Back Surface Field 79

5.4.2 Cu2Te Back Surface Field 83 x

5.4.3 As2Te3 Back Surface Field 88

5.4.4 PbTe Back Surface Field 93

5.5 Comparison among Proposed Cells 96

5.6 Comparison between Recent Published Work and Proposed Work 100

CHAPTER VI CONCLUSIONS 102

6.1 General Summary 102

6.2 Future Scopes 103

6.3 General Conclusions 104

REFERENCES 106

xi

LIST OF FIGURE

Figure No. Title Page No.

Figure 1.1 Overview of the dissertation organization 13

Figure 2.1 Equivalent circuit of idealized solar cell under illumination (In the 15 diagram, IL is the photocurrent generated by light hγ, and Idiode is the current of an ideal diode under voltage of V across it, and we assume the positive current direction is along Idiode)

Figure 2.2 Energy band diagram of a p-n junction in thermal equilibrium 16

Figure 2.3 Dark and light I-V characteristic of solar cell 17

Figure 2.4 Equivalent circuits for solar cell including series and shunt 19 resistance

Figure 2.5 Effect of Series Resistance on the I-V Curve of a Solar Cell 20

Figure 2.6 Effect of Shunt Resistance on the I-V Curve of a Solar Cell 20

Figure 2.7 (a) Light and dark J-V curves for an ideal solar cell; standard J-V 22 parameters that describe performance are indicated. (b) Non-ideal behavior includes parasitic resistances and a diode quality factor greater unity.

Figure 2.8 Example of a QE curve with various loss mechanisms in CdS/CdTe 24 solar cells

Figure 2.9 CdS/CdTe Photovoltaic Structure 28

Figure 2.10 Efficiency of ideal solar cells at 300 K for 1 sun under 1.5 AM 29

Figure 2.11 Proposed ZnCdS/ZnCdTe solar cell structure 33

Figure 3.1 Proposed solar cell structure used for the simulation. 38

Figure 3.2 Intervals and grids used in numerical method. There are N intervals 41 (dashed lines) and N + 1 major grid points (solid lines)

Figure 3.3 The conventional baseline case cell structure 45

Figure 3.4 AMPS device definition window with layers parameters, front and 47 back contact parameters

Figure 3.5 Modified cell structure for higher performance 48

Figure 3.6 J-V curves of solar cells 50 xii

Figure 4.1 Structure of the (a) conventional CdTe thin film solar cell and (b) 55 modified solar cell structure for higher performance

Figure 4.2 Structure of the proposed ultra thin film ZnCdTe solar cell structure 57 for higher performance.

Figure 4.3 Effect of ZnO and Zn2SnO4 on the ZnCdTe solar cell I-V 59 characteristics

Figure 4.4 Comparison of JV characteristics curve of CdTe and ZnCdTe cell. 61

Figure 4.5 Ohmic and rectifying metal/p-semiconductor contacts 63

Figure 4.6 J-V curve with rollover 64

Figure 4.7 Effect of back surface recombination velocity on ZnCdTe cell 66

Figure 5.1 Structure of ZnCdTe solar cell: (a) Conventional baseline case 70 structure and (b) modified structure for higher performance.

Figure 5.2 Effect of Zn content (x) on ZnxCd1-xS/ZnCdTe cell performance 72 using theparameters of ZnxCd1-xS from Table 5.2.

Figure 5.3 Effect of Zn0.08Cd0.92S thickness on ZnCdS/ZnCdTe cell 73 parameters.

Figure 5.4 Effect of Zn0.08Cd0.92S thickness on ZnCdS/ZnCdTe cell quantum 74 efficiency (QE)

Figure 5.5 J-V characteristics curve of the 1µm proposed ZnCdTe solar cell. 75

Figure 5.6 The effect of ZnCdTe thickness variation on the output parameters 76 of ZnCdS/ZnCdTe solar cell

Figure 5.7 The effect of ZnCdTe thickness on SR of proposed ZnCdS/ZnCdTe 77 solar cell.

Figure 5.8 CdTe/BSF/Metal contact 78

Figure 5.9 CdTe/ZnTe/Metal junctions 79

Figure 5.10 The proposed ultra thin ZnCdTe cell with ZnTe BSF 80

Figure 5.11 J-V characteristics of the proposed ZnCdTe cell with and without 81 ZnTe BSF.

Figure 5.12 Effect of ZnCdTe thickness on the cell with and without ZnTe 82 BSF.

Figure 5.13 Effect of operating temperature on the proposed ultra thin ZnCdTe 83 solar cell. xiii

Figure 5.14 CdTe/CuxTe/Metal junctions 84

Figure 5.15 Proposed ultra thin ZnCdTe cell with Cu2Te BSF layer. 85

Figure 5.16 J-V characteristics of the ZnCdTe cell with and without Cu2Te BSF 86

Figure 5.17 Effect of ZnCdTe thickness on the proposed ultra thin solar cell 87 with and without Cu2Te BSF.

Figure 5.18 Effect of operating temperature on the proposed cell. 88

Figure 5.19 CdTe/As2Te3/Metal junctions 89

Figure 5.20 The proposed ultrathin solar cell with As2Te3 BSF. 90

Figure 5.21 J-V characteristics of the ZnCdTe cell with and without As2Te3 91 BSF

Figure 5.22 Effect of ZnCdTe thickness for the cell with and without As2Te3 92 BSF

Figure 5.23 Effect of operating temperature on the proposed ultra thin ZnCdTe 93 cell

Figure 5.24 Proposed ultra thinZnCdTe solar cell with PbTe BSF. 94

Figure 5.25 J-V characteristics of the proposed cell with and without PbTe 95 BSF.

Figure 5.26 Effect of ZnCdTe thickness for the cell with and without PbTe 96 BSF.

Figure 5.27 Effect of temperature on the proposed cell. 97

Figure 5.28 J-V characteristics of the proposed cell with BSF 99

Figure 5.29 Effect of operating temperature on the proposed cells 99

xiv

LIST OF TABLE

Table No. Title Page No.

Table 1.1 Advantages and Disadvantages of Photovoltaic Technology 3

Table 2.1 Basic parameters for J-V characteristics 22

Table 2.2 Related Published Works on CdTe PV cell 26

Table 3.1 SnO2/CdS/CdTe solar cell baseline parameters 44

Table 4.1 Properties of some common TCO and buffer layers 58

Table 4.2 Metal work function Фm and resulting hole barrier Фb in ZnCdTe 64

Table 5.1 Material parameters used in AMPS simulation 71

Table 5.2 Parameters of ZnxCd1-xS used in simulation to optimize value of ‘x’ 71

Table 5.3 Output parameters of the modified cells without and with ZnTe BSF 81

Table 5.4 Output parameters of the modified cells without and with Cu2Te BSF 85

Table 5.5 Output parameters of the modified cells without and with As2Te3 90 BSF

Table 5.6 Performance parameter comparison of proposed ZnCdTe cell 94 without and with PbTe BSF layer

Table 5.7 The best output parameters of different cells with and without BSF 98

Table 5.8 Comparison between recent published work and the proposed work 101

xv

LIST OF SYMBOLS

α Absorption coefficient

φe Barrier height Al Aluminum a-Si Amorphous silicon & And

Jp And hole current density Å Angstrom Sb Antimony Ar Argon As Arsenide

Φbc Back contact barrier

RB Back surface reflectivity Br Bromine

Vbi Built in potential

фb Built potential Cd Cadmium

CdCl2 Cadmium chloride

Cd2SnO4, Cadmium stannate CdS Cadmium sulphide

CdSxTe1-x Cadmium sulphide telluride CdTe Cadmium telluride CdZnTe, Cadmium zinc telluride

CO2 Carbon dioxide ρ Charge density Cl, Chlorine

ΔEC Conduction band offset η Conversion efficiency Cu Copper

Cu2Se Copper selenide

CuxTe Copper telluride xvi

c-Si Crystalline silicon °C Degrees celsius d Depletion-region width A Diode quality factor

J0 Diode saturation current E Electric field χ Electron affinity

Jn Electron current density

De Electron diffusitivity

Gn Electron generation rates

Gn(x) Electron generation rates

τe Electron lifetime

µc Electron mobility

EFn Electron quasi-Fermi level Ψ Electrostatic potential

EF Fermi level F Fluorine

SnO2:F/Al Fluorine/aluminum doped tin oxide n Free electron p Free hole n(x) Free-electron density p(x) Free-hole density

RF Front surface reflectivity Ga Gallium GaAs Gallium arsenide Au Gold

Dp Hole diffusion coefficient

Gp Hole generation rates

Gp(x) Hole generation rates as function of distance

µp Hole mobility

EFp Hole quasi-Fermi level xvii

Pin Incident light power InSb Indium antimonide InP Indium phosphide I(x) Intensity at a distance x

I(0) Intensity on the surface I Iodine − NA , Ionized donor doping + ND Ionized doping acceptor W Layer thickness Le Diffusion length for electron

JL Light generated current Li, Lithium

Ln Diffusion length Jmax Maximum available photon current density Jmp Maximum power current Vmp Maximum power voltage HgSe Mercury selenide HgTe, Mercury telluride

φm Metal work function μm Micrometer mbar Milli bar mTorr Milli torr Mo Molybdenum Jo Multimode interferometer n N type Doping concentration nm Nanometer ns Nanosecond Ni Nickel N Nitrogen Ω/Sq Ohm per square Voc Open circuit voltage α Optical absorption coefficient xviii

O2 Oxygen CdS:O Oxygenated cds np P type Doping concentration Pd Palladium Pa Pascal Wp Peak watt ε Permittivity P Phosphor hν Photon energy Φ Photon flux Pt Platinum $/Watt Price figure Ω Resistance Rs Series resistance Jsc Short circuit current density Jsc Short circuit current density Rsh Shunt resistance Si Silicon a-Si:H Silicon: hydrogen alloys Ag Silver AgI Silver iodide Na Sodium Te Tellurium

SnO2 Tin oxide nt Trapped electron pt Trapped hole ZnO Zinc oxide

Zn3P2 Zinc phosphide

ZnSiAs2 Zinc silicon arsenide

Zn2SnO4, Zinc stannate ZnTe Zinc telluride

xix

LIST OF ABBREVIATIONS

AM Air Mass AM0 Air Mass 0 AM 1.5 Air Mass 1.5 AMPS Analysis of Microelectronic and Photonic Structures AFM Atomic Force Microscope ALD Atomic Layer Deposition ALE Atomic Layer Epitaxy BSF Back Surface Field BSRR Back Surface Recombination Rate BSRV Back-Surface Recombination Velocity CO2 Carbon Dioxide CBD Chemical Bath Deposition CVD Chemical Vapor Deposition CSS Close-Spaced Sublimation CIGS Copper Indium Gallium Diselenide J-V Current Density–Voltage I-V Current- Voltage DOS Density of States DC Direct Current EHP Electron- Hole Pair FF Fill Factor GW Giga Watt ITO Indium Tin Oxide IR Infrared NREL National Renewable Energy Laboratory PV Photovoltaic PVD Physical Vapor Deposition PSI Pounds Per Square Inch QE Quantum Efficiency RF Radio Frequency SEM Scanning Electron Microscope xx

SLG Soda Lime Glass SCAPS Solar Cell Capacitor Simulator SCR Space Charge Region SR Spectral Response SEL Stacked Elemental Layers TC Temperature Coefficient TCO Transparent Conducting Oxide TW Tetra Watt UV Ultra Violet USD United States Dollar XRD X-Ray-Diffraction

CHAPTER I

INTRODUCTION

1.1 MOTIVATION

One of the most important ingredients for continued civilization development is an abundant supply of energy. Energy is a key to the advancement and prosperity of humans. With rapid technological development the energy consumption has increased steadily. To sustain human development, more electricity consumption is expected in future decades. Our primary solution to this increasing electricity consumption has been to burn more fossil fuels (coal, oil, or natural gas) or build more nuclear plants. Such traditional sources of energy coal, liquid fossil fuels, and natural gas will become scarce or run out as the present rates of use in the near future. Moreover the scientific community is largely in agreement that green house gas emissions by burning fossil fuels are the most probable cause of the global warming [1], that the damage to the environment is irreversible on a timescale of hundreds of years [2], and thus reducing CO2 emissions should be a priority [3]. Furthermore, safe disposal of high-level radioactive waste from nuclear plants raises several issues. These impacts are irreversible.

The fossil fuels pollute the atmosphere and by product gasses like sulfur dioxide and nitrous oxides are the main causes of acid rains and smog. At the current estimates, our present life style results in a daily 16 million tons of CO2 into the atmosphere. This large amount of CO2 emission is one of the primary causes for global warming and climate changes. The impact on eco systems and environmental awareness from the increasing energy demands is a vital problem. Unsolved problem 2

with oil is that it is not a renewable resource. Its supply will eventually be depleted, which too several million years to produce will finish in a single life time. There are different estimates for predicting how fast we will finish the oil, ranging between 10 to 100 years. Oil is spent million times faster than it is formed and it is agreable that complete depletionis unavoidable. The political and economical tensions are generating due to the heterogeneous distribution of oil reserves, which might be increased in the future with oil reserves goes lower. The damages that are brought by fossil fuel to the environment are unjustifiable. These impacts are permanent and therefore, alternative energy resources are really needed. The substantial changes in climate are definitely a good reason to look for alternative energy sources. Eliminating the use of fossil fuels will require a replacement by an alternative source of energy. Clean and renewable sources of energy are the sound alternative of the energy crisis.

Besides nuclear power, there is an excess of renewable energy supplies, including hydroelectric, biomass, wind, geothermal, and solar power [4]. However, most of these resources are effectively limited in supply. The world already uses most of the economically feasible hydroelectric power. Though enough biomass fuel could be theoretically grown, such crops would need to occupy an unrealistic 31% of the land on earth [4]. Wind power is irregular and relatively scarce, with only 2 TW available globally, in practical terms. Geothermal power is also in short supply as there are few ideal sites, and drilling technology is not sufficiently advanced to economically access most of the remainder of the subsurface, solar energy resource [4]. On the other side, the sun constantly showers the earth with enormous amount of photons. This resource must be harvested; it is the most promising sources of alternative energy. Solar energy has received particular interest due to its plentiful availability and its potential for significantly reducing non renewable fossil fuel consumption.

1.2 PHOTOVOLTAIC TECHNOLOGY

The power of sunlight incident on the earth is about 125,000 TW [5]. That is about 10,000 times more than the world current energy demand. The sun will be available for a very long time for all of us as energy from the sun makes the life possible in the earth. The best continent way to utilize the energy of the sun is by 3

converting it into electricity. Solar cells perform this conversion based on the photovoltaic (PV) effect. The advantages and disadvantages of photovoltaic are mentioned in the following Table 1.1.

Table 1.1 Advantages and Disadvantages of Photovoltaic Technology

Advantages Disadvantages a. Source of energy is vast and a. Source energy is diffuse (Sun light essentially infinite is a relatively low density of energy) b. No emission, no combustion or radioactive fuel for disposal (does not contribute perceptibly to global climate change or pollution) c. Low operating cost (no fuel) b. High installation cost d. No moving parts (no wear) e. Ambient temperature operation c. Poorer reliability of auxiliary (no high temperature corrosion or (balance of system) elements safety issues ) including storage f. High reliability in modules (more than 20 years) g. Can be integrated into new or existing building structures h. Quick installation i. Can be installed at nearly any d. Lack of widespread commercially point of use available system integration and installation so far j. Daily output peak may match local demand k. High public acceptance e. Lack of economical efficient energy storage l. Excellent safety record

Source: Ref. [6]

Solar PV technology is a promising alternative energy source to meet the world energy demand because it harvests energy from the Sun free. Only a fraction of the total amount of solar energy is needed to be harvested to fulfill the entire demand of mankind. However, solar power depends upon location and climate, but can be used almost anywhere globally to at least some extent. Solar cells are devices that 4

produce electric energy when exposed to sunlight, which is free and available any where. PV cells can operate continuously without any maintenance, are totally silent, non-polluting and have a long lifetime of about 20-25 years. The main limit is their cost to the initial investment needed for purchase and installation. As a commercial product they produced in modules, and manufacturers typically guarantee these to last 20 years or longer. Compared to other renewable counterpart such as wind and water, they are available every where, less noisy, less likely to fail and has higher possibility for low cost production.

Eventually, as fossil fuel reserves become scarcer, their cost will rise until PV cells become a cost effective source of energy. PV cells, commonly known as solar cells, are devices made of semiconductor materials which convert sunlight into direct electricity. Solar cells were first designed for very specialized uses such in spacecrafts and satellites, but they can now be found on every day used devices from wrist watches to central power stations. The idea of converting light energy into useful electricity dates back to the 19th century, when A. E. Becquerel invented solar cell and the era of the modern solar cells starts in the 1950’s at Bell labs wherethe first 6%- efficient silicon solar cell was produced by Chapin et al.in 1954 [7]. The early efforts for development of PV cells were directed towards space applications and PV cells are still today the main source of power in space. PV use was expanded to terrestrial applications from 1970 and it keeps on growing rapidly. That growth is inspired by many advantages of solar cells, the primary one being its low impact on the environment. Another very important advantage is the more or less homogenous distribution of solar energy on the earth surface, making it accessible anywhere and to everyone. The portability of solar modules makes them commercially attractive especially for remote areas. The number of villages in the world which are not connected to an electrical grid is estimated to be two million. One third of the world’s population lives in these villages. Extending the electricity connections to remote areas, which are scattered and usually have lower energy demands compared to the cities, is often considered financially unattractive for the electricity providing companies. Basic electricity needs for agriculture, education, heating, and light in these areas are usually satisfied with electrical sources such as rechargeable batteries and small fossil fuel generators [8]. These sources are very often expensive and 5

unreliable. It is easy to predict the impact this lack of power on the development and standard of living in these areas. Providing solar powered electricity to remote areas may be one of very few ecologically and commercially acceptable solutions. Even though solar PV usage has clear advantages over conventional energy sources there are obstacles to its wider use. The main one is the higher production cost of the solar cells and modules. In order to lower the price and expand its usage, steps need to be taken for low cost production such as to reduce the quantity and cost of material process and the energy used to produce them.

To increase the power output of PV cells, individual cells are connected in series and parallel according to the voltage and current demand to form module or a photovoltaic system. Solar cells are good source of energy that is environmentally friendly with long term economic advantages. The scientific and technical challenges addressed by the universities, research groups and leading PV industries are the key to reaching this potential. The biggest hurdle now facing solar power is cost reduction. For a large scale solar system, the simple combination of cost per square meter divided by output power per square meter yields the key parameter ($/Wp), which is commonly used as the key PV metric. A peak watt (Wp), is the maximum power generated by a solar cell in the course of an ideal day. There are two ways to lower the cost of $/Wp: one is reducing the manufacturing cost $/m2 [5, 8] and the other isincreasing the output power Wp/m2 of the cells by increasing the cell’s conversion efficiency. Reduction of manufacturing cost can be achieved by using lower amounts of material, shorter processing time, lower energy input and cheaper processing technology. By increasing the cell conversion efficiencythe cell output power can be increased.

PV electrical power generation is gaining acceptance as an alternative for electrical power production. Because of its importance in new power generation, promising solar cell materials have been intensively researched targeting mainly on an acceptable tradeoff between conversion efficiency and low-cost manufacturing. It is anticipated that the price of PV will be lower than the commercial electricity production cost in near future with thin film technology. At a current cost of 0.50 USD per kilowatt hour, solar power is in general more than twice as expensive as conventional electrical power supply for residences. This high cost is mainly due to 6

the current expensive solar cell fabrication from Silicon (Si). Until now Si based solar cells are the dominant technology in the commercial production of solar cells. The transition to a cheaper thin film solar cell is essential for low costs in the long term, since thin films are much cheaper for a given output. Thin film PV modules production is rapidly increasing but, most of the PV cell until today is from the crystalline silicon. A transition to thin film technology for cost effective PV cells is essential in near future as the thin films offer significant reductions in material usage and cost. The thin film polycrystalline solar cells such as CdTe, copper indium selenide (CIS) and copper indium gallium diselenide (CIGS) have a great advantage of significantly lower production cost over conventional silicon based cell technology. Lower cost comes not only from lesser cell materials usage, but also from lower deposition temperatures (<1000°C) and from higher tolerance towards non uniformities, which allows lower purity usage of cell materials. However, several groups of researchers have investigated different solar cell materials targeting mainly on an acceptable trade off between conversion efficiency and low cost manufacturing. Cost effectiveness motivated the creation of the second generation of solar cells which is called thin film technology. Examples of these thin film materials are gallium arsenide, CIS, amorphous silicon (a-Si), and CdTe.

Thin-film CdTe based solar cell is one of the most promising candidates for PV energy conversion due to greater potential in low cost, high efficiency and stable solar cell fabrication. CdTe has near ideal band gap of 1.45 eV and high optical absorption coefficient of over 5×105/cm. It is a potential material for ultra thin film solar cell with low cost solar cell fabrication. The polycrystalline layers of a CdTe solar cell can be deposited using a variety of low cost techniques, such as close-spaced sublimation (CSS), physical vapor deposition (PVD), and chemical bath deposition (CBD) and sputtering [9]. CdTe thin film solar cells have shown long-term stable performance [10] and high efficiency [11] under AM1.5 illumination for terrestrial usage. It is note worthy that the often raised concerns on the toxicity of CdTe are perhaps based on the toxicity of elemental Cd but in the CdTe based solar cell, the Cd is converted to the stable compounds which are not toxic. The facts on the environmental and health issues of CdTe based on several extensive investigations prove that the environmental and health issues of CdTe based photovoltaic are 7

insignificant. The production of CdTe PV is a better way to restrain the risks created by mining of zinc and several other metals that are routinely used worldwide. Currently the main goal of PV cell research and development is the commercially viable cells that have mainly few mutually related features such as low cost, higher conversion efficiency and stable operating lifetime. It is anticipated that the CdTe PV cells are the potential candidate to the shared dream of humanity of clean and affordable energy from the sun, for all in everywhere.

1.3 PROBLEM STATEMENT

A solar cell is a p-n junction device which converts solar energy directly into electricity by the photovoltaic effect. The development of terrestrial solar PV accelerated in response to the oil crises of the 1970s. Over the past 40 years, solar cell and module conversion efficiencies and reliabilities have been increasing, manufacturing costs and prices of PV modules have been decreasing and markets have been growing at increasing rates. Today, most of the solar cells are made of crystalline silicon. Silicon’s availability is huge, but crystalline silicon (c-Si) cells require an expensive production process and it is doubtful that the single crystal silicon technology canreach module cost below $1/Wp. Silicon require relatively thick cells as it an indirect band gap material and some more it is fragile.

In the last decade, 24.7% efficiency was achieved in C-Si solar cells [12-13], 16.5% efficiency was achieved in thin film CdTe solar cells [11] and 19.5% efficiencywas achieved in thin film CIGS solar cells [8]. Polycrystalline CdTe is the leading thin film material for it numerous beneficial properties to realize low cost and high efficiency solar cells. Thin film CdTe based PV cells are one of the most promising candidates for low cost PV energy conversion because of the possibility of higher cell efficiency with reduced materials, reliable and stable cell operation. First of all the CdTe based cells are produced from polycrystalline materials and glass. Secondly, the layers of a CdTe solar cell can be deposited using different low cost techniques. Thirdly, CdTe has direct optimum band gap with a high absorption coefficient over 5×105/cm, which means that all the potential photons with energy greater than the band gap can be absorbed within a few µm of CdTe absorber layer. Hence, the small thickness required for an absorbing layer makes the cost of material 8

for CdTe based solar cells relatively very low. Certainly the main objective of PV cell research is using less material by making the cells thinner [8, 14]. Thinning will save material, lower production time and energy, all these factors will lead to decrease the cell production cost.

This work has focused on CdTe-based solar cells due to its good electronic properties and low cost possibility. Thin film polycrystalline CdTe based solar cells are one of the leading candidates for thin film PV cells mainly due to the optimal direct band gap of 1.45 eV which is closer match to the solar spectrum. Thus, only few micron CdTe absorber thicknesses is required to absorb all the potential photons which interns require less material and low cost fabrication. The reduction in price in turns fueled further global demand for PVs. The challenge is to develop high efficiency CdTe cells that can be produced at lower cost and it is the main driving force of this work. The efficiency of CdTe based solar cell can be increased using ternary CdZnTe material as absorber layer [15]. Cd1-xZnxTe has tunable band gap depending on the composition. In this work the band gap of CdZnTe layer (1.53 eV) which is in the optimum range, can be achieved with Zn composition of x=0.09. Furthermore, the uses of poly CdS as a window layer in CdTe solar cells have several drawbacks [16] although it has been found as a best suited heterojunction partner of CdTe. At first, poly-CdS film has a band gap of 2.42 eV, which causes considerable absorption in the short-wave length region. Theoretically, it is transparent for the light of wave length above 510 nm. The coefficients of absorption in II-VI compounds in the spectrum window are high [17], which is about 104 to 105 cm-1. As a result, the electron hole pair (EHP) generated by higher energy photons in the n-CdS layer takes place at the surface far away from the depletion region where the generated EHP can be collected. Thus, the EHPs are often lost to surface recombination current. Consequently, the dark current is increased which reduces the useful current delivered by the cell to a load. The CdS film with a thickness of 100 nm can absorb about 63% of the incident radiation with energy greater than its band gap [18], which cause for reduced Jsc and lower quantum efficiency (QE) at the blue region. To decrease the absorption and minimize the surface recombination current in CdS/CdTe cells, reducing the thickness of CdS layer [19] to less than 100 nm is routinely done. Although thinner CdS films are desirable, they are impractical due to excessive 9

pinhole formation and possible electrical shunting across the heterojunction. Below 70 nm, there is a general degradation [19] in the cell performance owing to considerable decrease in shunt resistance and due to excessive pinhole formation across the heterojunction which can adversely affect the device open-circuit voltage (Voc) and fill factor (FF). Then, a lower band gap (1.3 eV) CdS1-yTey can be formed due to the inter diffusion between CdTe and CdS layer at the p-n junction which also affects the performance. Finally, there exists nearly 10% lattice mismatch between poly CdTe and poly CdS, which causes high defect density around the junction region. Lattice mismatch can be reduced by device fabrication in high temperature but it enhances the inter-diffusion again and forms intermixed layers (CdS1-yTey and CdTexS1-x) causing pinholes and limits the improvement of Voc and FF.

It should have, therefore, been more appropriate to use a thick layer like ZnS or ZnTe films with higher band gap as a window ZnS/CdTe, ZnTe/CdTe [20-21]. However, there are several negative aspects related with this scheme. First, ZnS is exceedingly resistive and very hard to dope [22] and could considerably increase the cell's series resistance. Second, the lattice mismatch between ZnS and CdTe is about 16%, making it an inferior heterojunction partner to CdTe compare to CdS. Therefore, a new window material that has higher optical band gap, better lattice match with CdTe absorber, and minimum Te diffusion is sought for.

Considering the above discussion, ZnxCd1-xS can be chosen as an alternative window structure in CdTe solar cells. Its band gap can be tailored from 2.42 eV (CdS) to 3.7 eV (ZnS). In heterojunction solar cells, the use of ZnxCd1-xS instead of CdS can lead to an increase in photocurrent by providing a match in the electron affinities of the window and absorber material [23-28]. ZnxCd1-xS are known to have properties in between those of CdS and ZnS. In CdS/CdTe solar cells, the replacement of CdS with the higher band gap ternary ZnxCd1-xS film can lead to a decrease in window absorption losses and has resulted in an increase in the short circuit current [29]. Several novel device structure concerning ultra thin ZnCdTe/ZnCdS solar cells with back surface field (BSF) have been proposed with all possibilities being explored to increase the underlying efficiency with developed cell technologies. Investigations that are yet to be done may lead this vision to reality for higher ZnCdTe cell 10

conversion efficiency. Thus, the main problem is to develop high efficiency ZnCdS/ZnCdTe solar cell with least material input for low cost fabrication.

1.4 OBJECTIVES OF THE STUDY

To attain the expected breakthrough of photovoltaic technology as a competitive energy source against fossil fuels both the cell conversion efficiency and reduction of production cost are necessary. The ultimate goal of this study is to develop a high efficiency ultra thin ZnCdS/ZnCdTe solar cell. The BSF and ultra thin ZnCdTe absorber would increase cell efficiency and reduced fabrication material and finally cost of production. The following efforts were explored to develop ultra thin ZnCdS/ZnCdTe solar cell with suitable BSF.

(I) To design a high efficiency ZnCdS/ZnCdTe thin film solar cell

The most determinant of the performance of the solar cell, and necessary prerequisites for its design, are the choice of the material, structure and technology. The ZnCdTe absorber material and thin film were chosen to obtain high conversion efficiency design (thickness, doping concentration, contact material and structure). The baseline CdTe cell was modified with suitable front and back contact for higher performance. By incorporating an extra Zn2SnO4 buffer layer in conventional superstrate structure, the ZnCdS window layer thickness was reduced to improve the cell performance. The ultra thin ZnCdTe absorber layer was incorporated with BSF layer for higher efficiency.

(II) To optimize the ZnCdTe absorber and ZnCdS window layer by simulation

The challenging issue of ZnCdTe cells is related to the lesser material usage. The polycrystalline CdTe has a high absorption coefficient of over 5×105/cm, which means that all the potential photons of sunlight with energy greater than the band gap can be absorbed within a thin CdTe absorber layer. Moreover, CdTe has a direct optical bandgap of 1.45 eV which is close to the optimum bandgap for PV cells. If we incorporate small amount of Zn in CdTe its band gap can be increased and ultimately 11

increases the absorption of photon. The lesser thickness required for ZnCdTe absorber layer can lead to reduced cell material usage and lower cost of fabrication. Currently the main goal of PV research is towards less semiconductor material usage. Thus, reduction of ZnCdTe absorber and ZnCdS window layers thickness to the extreme limit will be explored in this work.

(III) To investigate the effect of BSF layer at the back contact of ultra thin ZnCdS/ZnCdTe cell

The major differences of ultra thin cells compared to the thicker ones is the absorber/back contact interface is located closer to the main junction (p-ZnCdTe/n- ZnCdS) of a ultra thin cell and the back contact material therefore has a high impact on the ultra thin cell overall performance. A stable back contact which is not rectifying and has a low resistance is required for good performance and long term stability of ZnCdTe based PV cells. As ZnCdTe is a p-type semiconductor material with a high electron affinity (χ=4.4 eV) and band gap (1.53 eV), usually a high work function metal is necessary to make good ohmic back contact to p-ZnCdTe. Most of the potential metals do not have sufficiently high work functions and eventually most of the metals form non ohomic contacts to p-ZnCdTe absorber layers. The presence of a high (<0.5 eV) back contact barrier would affect the current-voltage characteristics of the ZnCdTe cells. A possible method to overcome this problem is to reduce the barrier height by inserting an extra layer in between the CdTe and the metal back contact to form back surface field (BSF) at the back of the cell. The main purposes of BSF are to reduce the back surface recombination losses and another purpose is to reduce the barrier height at back contact to improve voltage and efficiency of the ZnCdTe cells. In case of ultra thin (<1 µm) ZnCdTe absorption layer, the loss at the back contact due to the carrier recombination is significant. The justification of such back surface field layer proved from the theoretical analysis of ultra thin ZnCdTe solar cells. Therefore, the insertion of suitable materials such as ZnTe, As2Te3, PbTe and Cu2Te were proposed at the back contact of ultra thin ZnCdTe cells those could slow down the recombination of generated carriers by bouncing them back to the ZnCdS/ZnCdTe junction for effective collection. The stability of cell with BSF anticipated being better than that of without BSF. 12

1.5 DISSERTATION ORGANIZATION

This dissertation is organized into six chapters as illustrated in Figure 1.1. Chapter 1 exposes a brief motivation, problem formulation and objectives of this study. First of all, energy demand and potential source of energy including alternative source of energy are explored finally solar energy as the ultimate solution is indicated. Then, in spite of a wide use of Si based solar cell, the increasing demand to develop thin-film solar cells as a suitable energy source is discussed. Again, among various kinds of thin-film solar cells, numerous advantages of low cost, high efficiency solar cells based on a ZnCdTe thin film absorber are selected for this work. Chapter 2 reviews the present status of the most common cell technologies and the fundamental principles of the solar energy conversion process. The most widely used PV cells principle is the operation of a cell, photovoltaic effect, fundamental semiconductor concepts, thin film CdTe solar cell fundamental aspects and key features, CdTe solar cells structure, conventional and baseline case CdTe cells are briefly discussed. Brief reviews of research works on CdTe and ZnCdTe thin films and solar cells are also summarized. Chapter 3 presents the theoretical modeling of ZnCdTe thin film solar cells based on available numerical tools as it has been recognized as an indispensable tool for the design of efficient cells. This particular portion of the study describes a one dimensional simulation program Analysis of Microelectronic and Photonic Structures (AMPS-1D) and its application to ZnCdTe solar cells from different point of views. It would mainly emphasize the posibility of thickness reduction of ZnCdTe absorber from theoretical analysis and furthermore would include the justification of a BSF layer. Description of the CdTe base case, base case parameters and parameters used for ultra thin film ZnCdTe solar cell simulation are discussed. Chapter 4 focuses on the design and analysis of ZnCdS and ZnCdTe thickness reduction with minimum material uses without much compromise of conversion efficiency. Proposed cell structure, front contact, buffer layer and other layer selection are performed. To make the ZnCdTe PV cells further attractive by reducing the amount of rear earth material used like Tellurium, the absorption layer of ZnCdTe could be of a minimum thickness without compromising conversion efficiency are analyzed. To improve the open circuit voltage (Voc) and short circuit current density (Jsc) through controlling the 13

recombination loss at ZnCdTe bulk with suitable BSF layer at the back contact were proposed. Chapter 5 narrates the BSF mechanism, strategy, approach and related issues regarding the back contact of ultra thin ZnCdTe solar cells, which includes the insertion of suitable BSF layer at the back of the cells. These BSF materials are expected to have an effect of reflecting back the generated carriers (hole) from recombination loss at the back contact. The last part in this chapter is intended for the subjects to solve in this study concerning the ultra thin ZnCdTe absorber layer with suitable BSF including ZnTe, As2Te3, PbTe and Cu2Te are explored individually and finally these BSF insertions are compared. Chapter 6 summarizes the works performed and results obtained in this study. Areas for further improvement of efficiency and optimization of cell fabrication steps with ZnCdTe ultra thin-film are demonstrated with final remarks.

Introduction [Chapter 1]

Literature Review [Chapter 2]

Methodology [Chapter 3]

Optimisation of Cell Structure from Numerical Analysis [Chapter 4]

Result & Discussion [Chapter 5]

Conclusions [Chapter 6]

Figure 1.1 Overview of the dissertation organization

CHAPTER II

LITERATURE REVIEW

2.1 INTRODUCTION

Important principles and properties of semiconductors and solar cells are briefly reviewed as these are helpful for the understanding of this work. This section includes a brief description of state-of-the-art CdTe solar cells. All the solar cell parameters and CdTe cell structure are extensively reviewed in this chapter.

2.2 SOLAR CELL BASICS

2.2.1 Solar Cells and Solar Energy Materials

Solar energy is one of the profuse, most consistent renewable energy in our planet which has no adverse effect on our environment. During the last three decades considerable progress has been achieved in developing technologies to harness electricity from solar radiation. Silicon is the most common material using in solar cell which average efficiency is 24% [30]. Although silicon single-crystal cells have enormous advantages in terms of high efficiency and durability, but these devices are not best suited for terrestrial applications because they are very expensive and it is doubtful that the single-crystal silicon technology can reach module cost as expected below $1/Wp. In the last two decades considerable work has been done in thin film solar cells to replace silicon to reduce solar cell cost. CdTe and Cu (In, Ga) Se2 are the leading thin film photovoltaic materials with efficiencies exceeding 16.4% and 19.2% [31-32], respectively. But the theoretically predicted efficiencies are higher than the reported values; therefore much work is needed to optimize the material properties of the films, the junctions and also the fabrication techniques. 15

2.2.2 Solar Cell Operation

Solar cell directly converts solar energy into electrical energy. It basically is a p-n junction diode, which is optimized in order to get maximum electrical power when exposed to sunlight. The equivalent circuit of an ideal solar cell is shown in Figure 2.1.

Figure 2.1: Equivalent circuit of idealized solar cell under illumination (In the diagram, IL is the photocurrent generated by light hγ, and Idiode is the current of an ideal diode under voltage of V across it, and we assume the positive current direction is along Idiode)

Source: Ref. [33]

When a photon strikes the cell it is absorbed by an electron in the valence band and therefore provides energy to the electron elevating it to the electronic state in the conduction band leaving a hole in the valence band. These electrons and holes act as free charge carriers and contribute to the current. In the absence of the electric field both electron and hole randomly move until they recombine. If during this random movement they get into the space charge region of the p-n junction they are separated by the electric field and collected as a part of photocurrent. Figure 2.2 shows the energy band diagram for p-n junction under equilibrium.

16

Figure 2.2: Energy band diagram of a p-n junction in thermal equilibrium

Source: Ref. [34]

To maximize the efficiency of the solar cells the following conditions have to be met: 1. A major portion of the incident light should be absorbed by the cell to generate electron-hole pairs. 2. The cell structure should provide a mechanism for separating the generated charge carriers. 3. The lifetimes of the minority charge carriers are high to ensure good collection efficiency.

2.2.3 Solar Cell Output Parameters

Figure 2.1 shows the equivalent circuit diagram of an ideal solar cell, which includes a current generator representing photo-excitation. The current flowing through the circuit in the presence of light is

푞푉 ( ) 퐼 = 퐼0 [푒 퐴푘푇 − 1] − 퐼퐿 2.1 Where,

Io is the reverse saturation current, A is ideality factor (which determines how the J-V curve is bent and usually is achieved from experiments),

IL is the light generated current. 17

q is the charge of an electron, k is the Boltzmann constant and T is the temperature in kelvin.

We shall now look at each of the parameters, quantifying the performance of a solar cell:

2.2.3.1 Short Circuit Current (Isc)

Short circuit current is the current when no bias voltage is applied to the cell. Isc can be obtained by substituting V=0 in the equation 2.1.

퐼푠푐 = −퐼퐿 2.2

2.2.3.2 Open Circuit Voltage (Voc)

The voltage developed when terminals are open (Infinite load resistance) is called open circuit voltage, Voc is given by

푘푇 퐼퐿 푉표푐 = 퐴 ln [ + 1] 2.3 푞 퐼0

Figure 2.3: Dark and light I-V characteristic of solar cell

18

2.2.3.3 Fill Factor (FF)

From Figure 2.3, Pmax is maximum power that can be obtained at the point where voltage and current are maximum, i.e at point Vmax and Imax. So, Pmax is area of the rectangle with sides Vmax and Imax.

푃푚푎푥 = 퐼푚푎푥푉푚푎푥 2.4

FF is defined as the ratio of the maximum power to the product of Voc and Isc.

푃 퐹퐹 = 푚푎푥 2.5 퐼푆퐶푉푂퐶

It is a measure of the ‘squareness’ of the light I-V curve.

2.2.3.4 Efficiency (η)

The efficiency of the solar cell is defined as the ratio of Pout, electrically generated power to the input power on the cell.

푉 퐼 휂 = 푂퐶 푆퐶 퐹퐹 2.6 푃𝑖푛

2.2.4 Series and Shunt Resistance (Rs and Rsh)

The equation 2.1 is for an ideal case of solar cell but departure from ideal case usually happens in real world solar cells. The difference than idealized solar cells is that they have extra resistance associated with diodes - series and shunt resistance. Then, the diode equation of real world solar cells is modified to be like:

푞(푉−푅 .퐼) ( 푆 )−1 푉−푅 .퐼 퐴푘푇 푆 퐼 = 퐼0 [푒 ] + − 퐼퐿 2.7 푅푆ℎ

Where, RS and RSh are series and shunt resistance respectively. And the equivalent circuit of a solar cell in this case is shown in Figure 2.4.

19

Figure 2.4: Equivalent circuits for solar cell including series and shunt resistance

Source: Ref. [35]

2.2.4.1 Series Resistance

The series resistance (Rs) is the resistance the carriers find on their way. The series resistance of a solar cell mainly arises from the ohmic resistance in semiconductor substrate, ohmic resistance in metal contacts, and metal-semiconductor contact resistance. The low conductivity of the window layer and absorber layer and recombination of carriers into the bulk materials are the core fact for Rs. The Fill Factor (FF) is directly affected by series resistance as it softens the I–V characteristics of a solar cell in the fourth quadrant, eventually the efficiency decreases of the solar cell. It is found that the fill factor of a solar cell decreases by about 2.5% for each 0.1 increase in series resistance [36].

In open-circuit condition the current is zero and the voltage drop across the Rs is also zero, so the Voc is not affected by the Rs. But in short-circuit condition, current flows through the series resistance and Rs affect the short circuit current and degrade Isc. Variation in I-V characteristic due to change in Rs is shown in Figure 2.5

20

Figure 2.5: Effect of Series Resistance on the I-V Curve of a Solar Cell

Source: Ref. [35]

2.2.4.2 Shunt Resistance

The shunt resistance arises from leakage currents through the cell, pinholes and voids. Shunt resistance degrades the fill factor and Voc. As the shunt resistance decreases, the current flow through it increases thereby affecting the Voc. Variation in

I-V characteristics with RSh is shown in Figure 2.6

Figure 2.6: Effect of Shunt Resistance on the I-V Curve of a Solar Cell

Source: Ref. [35] 21

2.2.5 Diode Factor

The diode factor is an important parameter for solar cell that characterizing the ‘perfectness’ of the junction. The value of diode factor is dependent on the mechanism of junction transport, e.g. if the transport is dominated by diffusion then the value of A is closer to 1; if recombination is the primary transport mechanism then A is closer to 2. This factor is dependent on the process and material. The value of A should be close to 1 to get optimum output for solar cells. The values of ‘A’ increase due to the defect and impurity precipitation. ‘A’ can be determined by the slope of the ln (J)-V curves.

2.2.6 J-V Characteristics

In the wide range of characterization methods available for thin-film solar cells measurement, the current-density vs. voltage (J-V) curves is of great importance, since it determines the efficiency at which the solar cell converts the Sun’s power into electricity.

The standard J-V measurement is performed under a standardized “one-sun” illumination at room temperature [37]. Experimentally, it is necessary to cool the solar cell during illumination, as otherwise the intense light will increase the cell’s temperature. In short, the solar cell is subjected to a calibrated light source, two contacts are used to apply a voltage bias, and two additional contacts are used to determine the resulting cell current. An example of a J-V curve is shown in Figure 2.7(a) and from such a curve, the basic performance parameters can be extracted as listed in Table 2.1. Assuming that the curve follows an exponential behavior, there are redundancy in these parameters and only three (e.g., Voc, Jsc, and η) are necessary to specify them all. Ideally, a solar-cell J-V curve equals that of a p-n junction diode shifted by the light generated current JL≌Jsc as written in equation 2.8.

However, with including parasitic losses of thin film solar cell to the equation (2.9), the diode equation becomes [38]: 22

Where, series resistance RS, shunt-resistance rsh, and diode quality factor A describe these nonidealities. The effect of each of these parameters on the J-V curve is indicated in Figure 2.7(b).

Figure 2.7: (a) Light and dark J-V curves for an ideal solar cell; standard J-V parameters that describe performance are indicated. (b) Non-ideal behavior includes parasitic resistances and a diode quality factor greater unity. Source: Author’s estimation

Table 2.1: Basic parameters for J-V characteristics

Parameter Symbol Unit Determined by

Open circuit voltage Voc V J = 0 2 Short circuit current Jsc mA/cm V = 0 Maximum power voltage Vmax V V at (JV)max 2 Maximum power current Jmax mA/cm J at (JV)max Fill factor FF % (JV)max/ (Voc. Jsc) Efficiency η % (JV)max/ (Pincident)

Source: Author’s estimation

23

2.2.7 Quantum Efficiency

The J-V curve tells us a little about the basic characteristic of conversion efficiency on about the actual device. One extension of a regular J-V measurement is the measurement of the wavelength-dependent current response, generally known as the “quantum efficiency” (QE). The QE represents the fraction of photons of wavelength that are converted into electron-hole pairs and collected as current. Typically QE performed at zero bias, its measurements under light or voltage bias, can also be misleading and careful analysis is necessary [39]-[40].

Measurement of the current response at zero voltage and normalization of this current to the flux density of the incoming light allows calculation of the quantum efficiency:

QE (λ) = (횫J/q)/φ = (collected electron-hole pairs / incident photons) 2.10

The optical beam of total flux density φ (units of [1/cm2s]) is monochromatic, but its peak wavelength varies over the spectral range that is relevant to the solar cell application.

Eq. 2.10 can be integrated to express the relation between the short-circuit current density and QE,

Where, F is the flux density per unit wavelength. A QE curve with different energy loss mechanism is shown in Figure 2.8.

24

Figure 2.8: Example of a QE curve with various loss mechanisms in CdS/CdTe solar cells

Source: Author’s estimation

The losses in quantum efficiency can be described as follows:

 ”Reflection losses” is introduced by the Non transparent contacts (i.e., metal contact fingers). Some reflection also happens at the material interfaces. Experimentally, these losses can be minimized by the applying anti-reflective coatings.  “Window” absorption in the short-wavelength region is negligible due to the high band-gap energy of this material. Free electron absorption in the ITO layer can lower the quantum efficiency in the high wavelength region, but this effect is typically small.  “Buffer” absorption represents one of the major losses in today’s for CdTe and CIGS thin-film solar cells. Thinning of the CdS or replacing it with a higher band-gap materials are possible alternatives. Here is one of the main objectives of our work. We introduce Zn on to the CdS layer to increase its band gap and minimize other defect of it.  “Recombination” losses are introduced by less-than-ideal collection efficiencies of photo-generated carriers. The longer wavelengths are causing 25

for the generation of carriers at the deeper region, and that increase the likelihood of recombination.  “Deep penetration” losses are inherent to every semiconductor as light with

photon energy of hν < Eg is not absorbed.

2.3 A BRIEF HISTORY ON THIN FILM CdTe SOLAR CELLS

The first silicon solar cell discovered by Russel Ohl accidentally in 1940 [41]. He was surprised to measure a large electrical voltage from what he thought was a pure rod of silicon when he shone a flashlight on it. Closer investigation showed that small concentrations of impurities were giving portions of the silicon properties dubbed as ‘negative’ (n-type). These properties are now known to be due to a surplus of mobile electrons with their negative charge. Other regions had ‘positive’ (p-type) properties, now known to be due to a deficiency of electrons, causing an effect similar to a surplus of positive charge (something close to a physical demonstration of the mathematical adage that two negatives make a positive).

William Shockley worked out the theory of the devices formed from junctions between `positive’ and `negative’ regions (pn junctions) in 1949 and soon used this theory to design the first practical transistors. The semiconductor revolution of the 1950s followed, which also resulted in the first efficient solar cells by Bell Laboratories in 1954 in the form of a 6%-efficient crystalline-silicon (c-Si) device [7]. Loferski, (1956) showed that the optimum band gap of the semiconductor materials for photovoltaic solar energy conversion is 1.5 eV. He suggests that CdTe, which has a direct energy band gap near to ideal value, and a high optical absorption coefficient for photons with energies greater than the energy band gap (α >105 cm-1) is a promising PV material [42-43]. It was realized that a very thin layer (few microns) of CdTe absorber is enough to absorb most of the incident light which will minimize material costs. A long minority carrier diffusion length would not be required (as most of the light would be absorbed in the depletion region) for CdTe polycrystalline layer and it might be deposited using low cost processing methods for producing efficient cells. The earliest CdTe cell was developed using the structure of n-CdTe/p-Cu2Te

[44]. The n-CdTe/p-Cu2Te cell showed a promising conversion efficiency of 6% [44].

However, this device suffered from instabilities due to Cu diffusion from the p-Cu2Te 26

layer to the front contact, thus an appropriate p-type window material for n-type CdTe absorber layer was essential. To escape from this problem, focus was given to device structure that would employ p-type CdTe and n-CdS. The n-CdS was found to be appropriate heterojunction partner to p-CdTe.

Table 2.2: Related Published Works on CdTe PV cell

Year Eff. Technology Structure T Comments ( %) (0C) 1991 13.4 CSS SnO2/CdS/CdTe/Metal NM p-CdTe by CSS

1992 14.6 CSS SnO2/CdS/CdTe/ Metal NM Voc improved

1993 15.8 CSS SnO2/CdS/CdTe/ Metal NM FF and Jsc high

1997 16 CSS SnO2/CdS/CdTe/C:Cu NM C:Cu & high Jsc

2000 13.5 CSS SnO2/CdS/CdTe/Ag:C 650 2 μmCdTe 2001 16.5 Magn. CTO/ZTO/CdS/CdTe/C 625 Champion Cell, Sputt. u:HgTe:CuxTe CdTe 10 μm +CBD+CSS 2002 13.5 NM ITO/CdS/CdTe/ Metal 334 CdTe 2 μm 2003 16.4 NM ITO/CdS/CdTe/Cu:Au NM Base case 2004 14 Sputtering ITO/ZnO/CdS/CdTe/M 387 All Sputtering 2005 11.4 CVD ITO/CdS/CdTe/Poly 450 Flexible Sub. 2006 12.5 HVE ITO/FTO/CdS/CdTe/Pol 450 Flexible, 4 µm 2006 13 RF Sputt. SnO2/CdS/CdTe/Cu+Au 390 CdTe is 2.3 µm 2007 10 PVD ITO/CdS/CdTe /FTO 350 Bifacial, 2.5 µm 2007 21 NM ITO/CdS/CdTe /BSF NA Numerical Ans. 2007 >20 NM ITO/CdS/CdTe /ZnTe NA Numerical Ans. 2008 11.5 CSS+Sputt. ITO/CdS/CdTe/ZnTe NM CdTe is 1.2 µm 2009 10.2 CSS+PVD FTO/CdS/CdTe/Cu+Au 660 CdTe 3.7 μm 2010 15.8 DC and RF ITO/ZnO/CdS/CdTe/As2 500 CdTe> 5 Sputt.+ CSS Te3/Cu+Au+Mo μmAs2Te3 BSF 2011 17.3 VTD CdTe First Solar 2013 18.7 VTD CdTe First Solar

Source: Ref: [11, 18, 19, 45 & 46]

A thin film p-CdTe/n-CdS heterojunction was first demonstrated in 1969 [47]. Bonnet & Rabenhorst (1972) demonstrated the excellent potential of the n-CdS/p- CdTe cell by producing devices with efficiencies of 5-6%. For this cell, the layers were deposited in the ‘‘superstrate configuration’’ as light pass through the glass substrate first and then CdS and CdTe layers, respectively. Afterwards, Bosol, (1990) reported 10% efficiency CdS/CdTe solar cell. Further research by Bonnet and co-workers and complementary work led to the 27

development of CdS/CdTe devices with efficiencies over 10% and Mr. Bonnet set up ANTEC GmbH (in Germany) to manufacture commercial modules [18 & 45]. In 1991, a thin film CdTe solar cell with 13.4% efficiency was reported [48]. Record cells were then reported as having an efficiency of 14.6% in 1992 [18], 15.8% in 1993 [19 & 49], 16% in 1997 [46], 16.5% in 2001 [11], 17.3% in 2011 by First Solar [50] and 18.7% in 2013 again by First Solar [51]. The efficiency 16.5% CdTe-based solar cell produced was certified by NREL in 2004 made by Wu, (2004). This device was fabricated by depositing layers in the sequence of Cd2SnO4/Zn2SnO4/CdS/CdTe/Cu.

The improvement is due to the Cd2SnO4 layer being both more transparent to light in visual region and more conductive than usually used SnO2. The structures of 18.7% and 17.3% CdTe solar cells made by First Solar are still undefined. Table 2.2 above shows some reported works on CdTe PV.

2.4 CdS/CdTe SOLAR CELL

Cadmium (Cd) based II-VI semiconductors gained importance in making photovoltaic devices and CdTe is the one that receives more interest because of its low cost and high efficiency in making thin film solar cells. First, the cell is produced from polycrystalline materials and glass, which are potentially much cheaper than bulk silicon. Second, the polycrystalline layers of a CdTe solar cell can be deposited using a variety of different techniques [51]-[52], such as close-spaced sublimation (CSS), which has been used to produce the highest efficiency cells so far, chemical vapor deposition (CVD), chemical bath deposition (CBD), physical vapor deposition (PVD i.e. RF sputtering), molecular beam epitaxy (MBE) and vapor transpot deposition (VTD). Third, CdTe has a high absorption coefficient because it has a direct band gap which is very close to the optimum band gap to absorb solar spectrum, so that approximately 99% of the incident light is absorbed by a layer thickness of about 1μm. CIS and CIGS materials also show almost same ability as mentioned in above but CdTe has a better advantage over them regarding the electronic properties. The electron properties are one of the most important properties to make high efficiency solar cell and CdTe is unique in this regards. The CdS/CdTe solar cell structure can be divided into five major components: (1) Substrate, (2) Front contact, (3) Window 28

layer, (4) Absorber layer and (5) Back contact. Figure 2.9 shows the schematic of the different components of a CdS/CdTe thin film solar cell.

Figure 2.9: CdS/CdTe Photovoltaic Structure

Source: Ref. [53]

2.4.1 Substrate

Substrate is very important component of the solar cell; it should withstand the cell fabrication process temperatures and must not contaminate the layers that are grown later. CdS/CdTe solar cells have a superstrate structure which implies that light enters the substrate first; hence CdTe solar cells require a transparent substrate so that reflection and absorption do not occur in the substrate, which would be detrimental to the current generation of the cell. The best choice for a transparent substrate is glass because it is cheap and withstands high temperatures. Common types used are soda- lime glass, which is inexpensive, and borosilicate glass [45]. The outer face of the glass often has an anti-reflective coat to enhance its optical properties.

2.4.2 Front Electrical Contact

In general, transparent conducting oxides (TCO) are used as front contact [45]. It is an n+ transparent conductive oxide with very high band gap. It should have high optical transmittance and low resistivity. The most widely used materials is tin oxide

(SnO2), it is deposited onto the glass either by sputtering or atmospheric pressure 29

chemical vapor deposition. As tin oxide has too low conductivity for making a good contact, it is often doped either with indium, forming indium tin oxide (ITO) or with fluorine giving a compound fluorine tin oxide (FTO). For low-temperature CdS and CdTe deposition processes, ITO/FTO is the material of choice, and for CdS or CdTe deposition requiring high temperature, SnO2 is the best material, since it is very stable under high temperature.

2.4.3 CdS Window Layer

The polycrystalline CdS layer acts as the n part of the p-n junction. It can be deposited by different deposition techniques; such as RF sputtering, chemical bath deposition (CBD), thermal evaporation chemical vapour deposition (CVD), close spaced sublimation (CSS), molecular beam epitaxy (MBE), spray pyrolysis and hot wall epitaxy and so on [54]-[55]. The structural and optical properties of the layer are strongly depending on the deposition technique [56].

2.4.4 CdTe Absorber Layer

The polycrystalline CdTe layer should be electrically p-type to form the p-n junction with n-CdS. It has an energy gap of 1.45 eV, which gives the highest theoretical efficiency because the band gap is very close to the optimum band gap for solar cells.

Figure 2.10: Efficiency of ideal solar cells at 300 K for 1 sun under 1.5 AM

Source: Ref. [57] 30

Figure 2.10 shows the ideal solar-cell efficiency at 300K under one-sun illumination and AM 1.5G as a function of energy band gap. Here the maximum value of ideal efficiency occurs near a band gap of 1.5 eV, which is approximately the band gap of CdTe. The corresponding ideal efficiency for a CdTe solar cell is about 29%. Many factors will degrade the ideal efficiency, so that efficiencies actually achieved should be lower. The record laboratory efficiency for CdTe thin-film solar cell has reached 18.7% 18.7% in 2013 by First Solar [51]. The CdTe module performance is over 10% [52] and 14.4% by First Solar which is world record for module efficiency till now [58].

2.4.5 Back Contact

One of the major challenges associated with the fabrication of efficient CdTe based solar cells is that the formation of stable, low resistance back contact to p-CdTe. It is difficult due to large electron affinity and low native carrier concentration of CdTe. An ohmic contact to CdTe requires either a metal with a work function greater than 5.7 eV or a sufficiently narrow Schottky barrier to enable tunneling. No metal have been found which have such large work function (>5.7eV) to properly match with CdTe and so, there naturally exist a wide Schottky barrier between the CdTe and metal back contact, resulting in a significant limitation to hole transport from the p- CdTe. In a circuit model, this barrier will form a diode of opposite polarity to the primary junction [59], decreases device efficiency by reducing the fill factor and by limiting the Voc. Generally a Cu containing back contact is used improve device performance of thin-film n-CdS/p-CdTe solar cells [60], which creates a quasi ohmic, non-rectifying contact and additionally dopes the CdTe layer. These types of cells exhibit good efficiencies in the beginning, however the efficiency degrade with time due to Cu diffusion to the front contact which causes shunting effect [10]. Another strategy to overcome the naturally existing Schottky barrier is to create a heavily p- doped CdTe surface by chemical etching and applying a buffer layer or back surface field (BSF) layer of high carrier concentration and low resistive between CdTe and the metal [61]-[62]. This process decreases the barrier width at the back contact interface. The tunneling barrier formed in this way is quasi ohmic and also acts as a minority carrier mirror.

31

2.5 RESEARCH TARGET AREA– ZnxCd1-xS WINDOW LAYER AND ZnCdTe ABSORBER LAYER FOR CdTe SOLAR CELL

In traditional CdTe solar cells CdS is popularly used as a window material and as an n-type heterojunction partner to p-CdTe. The highest efficiency of the laboratory-device in 2004 has reached 16.5% [63] and in 2013 by First Solar [51]. However, in conventional CdS/CdTe cell has the three main issues that limit device performance. At first, a lower band gap (1.3 eV) CdS1-yTey can be formed due to the inter diffusion between CdTe and CdS layer at the p-n junction which affects device performance. Secondly, poly-CdS film has a band gap of 2.42 eV, which causes considerable absorption in the short-wave length region. Theoretically, it is transparent for the light of wave length above 510 nm. The coefficients of absorption in II-VI compounds in the spectrum window are high [17]), which is about 104 to 105 cm-1. As a result, the EHP generated by higher energy photons in the n-CdS layer takes place at the surface far away from the depletion region where the generated EHP can be collected. Thus, the EHPs are often lost to surface recombination current. Consequently, the dark current is increased which reduces the useful current delivered by the cell to a load. The CdS film with a thickness of 100 nm can absorb about 63% of the incident radiation with energy greater than its band gap [18], which cause for reducing Jsc and lower quantum efficiency (QE) at the blue region. To decrease the absorption and minimize the surface recombination current in CdS/CdTe cells, reducing the thickness of CdS layer [19] to less than 100 nm is routinely done. Below 70 nm, there is a general degradation [19] in the cell performance owing to considerable decrease in shunt resistance and due to excessive pinhole formation across the heterojunction which can adversely affect the device open-circuit voltage (Voc) and fill factor (FF). Thirdly, there is nearly 10% lattice mismatch between poly CdTe and poly CdS, which causes high defect density at the junction region. Lattice mismatch can be reduced by device fabrication in high temperature but it enhances the inter-diffusion again and form intermixed layers (CdS1-yTey and CdTezS1-z) cause of pinhole effect and limits the improvement of Voc and FF.

It should have, therefore, been more appropriate to use a thick layer like ZnS film with 3.7 eV band gap as a window, but there are several negative aspects related with this scheme. First, ZnS is exceedingly resistive and very hard to dope [22] and 32

could considerably increase the cell's series resistance. Second, the lattice mismatch between ZnS and CdTe is about 16%, making it an inferior heterojunction partner to CdTe compare to CdS.

Considering the above discussion, we have chosen ZnxCd1-xS as an alternative window structure in CdTe solar cells. Its bandgap can be tailored from 2.42 eV (CdS) to 3.7 eV (ZnS). In heterojunction solar cells, the use of ZnxCd1-xS instead of CdS can lead to an increase in photocurrent by providing a match in the electron affinities of the window and absorber material [24 & 64]. ZnxCd1-xS are known to have properties in between those of CdS and ZnS. In CdS/CdTe solar cells, the replacement of CdS with the higher band gap ternary ZnxCd1-xS film can lead to a decrease in window absorption losses and has resulted in an increase in the short circuit current [29]. It has been used as a potential window layer to form heterojunction solar cells with Si [65],

CuxS [66], CuInSe2 [67], CuGaSe2 [68], Cu(In,Ga)Se2 [69] and CdTe [70-71] for its large and variable band gap between 2.42 and 3.7 eV. The band gap energy of the window layer can be increased significantly and the difference in thermal expansion coefficient is reduced [23].

The addition of Zn to CdS improves the Voc and Jsc in heterojunction devices as a result of the decrease of absorption losses in the window layer [29 & 72]. Under

AM1.5 illumination, ZnxCd1-xS is transparent to this spectrum [73], the EHPs generated by the higher energies photons now well removed from the cell surface, are no longer lost to surface recombination. Instead, they are separated and collected by the assistance of the field present in this region. It enhances the spectral response [23 & 74] in the short wavelength region without compromising the layer thickness. Therefore, the outcome is a strong spectral response of the solar cell in almost all AM1.5 spectrums resulting in higher efficiency solar cell.

33

Figure 2.11: Proposed ZnCdS/ZnCdTe solar cell structure

However, there are still major works for improvement in balancing the effects of ZnxCd1-xS on cell output parameters Voc, Jsc and FF for optimum composition of Zn and Cadmium (Cd) with suitable back contact buffer layers which are termed as

BSF (As2Te3, PbTe, ZnTe, Cu2Te) by modified design using standard numerical technique and fabrication method. The proposed ZnCdS/ZnCdTe solar cell structure is shown in Figure 2.11 above.

The optimization of deposition parameters of existing solar cells and fabrication of new solar cells and new optoelectronic devices repeatedly demand preparing ZnxCd1-xS window layer with a controllable distribution of Zn content. Keeping these aspects in view, much attention is being given in developing good quality ZnxCd1-xS thin films for comprehensive optical studies and their various applications. However, problems, for example, resistivity increasing from 1 Ω-cm to 1010 Ω-cm with increasing zinc content in the compound cannot be overlooked [75]. Therefore, in this study, an attempt was made to determine the optimum value of zinc content in ZnxCd1-xS for CdTe solar cells by standard numerical simulation technique (AMPS 1D). However, for these cells to compete adequately with conventional energy sources there is need to improve its efficiency. It is envisaged that to achieve 34

low cost PV energy supplies, module efficiencies of 15% or higher cell efficiencies greater than 25% are required [76]. To obtain efficiencies greater than 25 %, the band gap of the absorber layer (CdTe) was increased by incorporating Zn in CdTe [76-79]. A limitation of conventional CdTe cell structure is that the absorber layer can absorb only photons with energy equal to or greater than its band gap. Moreover photons with energy near or just below the band gap may pass through to the bottom of the cell, where most of their energy is lost by thermalization, resulting in loss of efficiency. In an attempt to overcome the limited efficiency of the CdTe solar cell, this study proposes a different cell absorber, whereby zinc cadmium telluride (ZnCdTe), as an absorber is used, instead of using CdTe as absorber. ZnxCd1-xTe has a uniformly varying band gap with variation of x. This way, all the photons whose energy lies within the band gap range is supposedly absorbed, leading to higher electron hole-pair generation and thus higher efficiency. This is possible because the compound (ZnxCd1- xTe) has a tuneable band gap of 1.45-2.2eV, depending on the value of x. [76, 77, and

80]. However, for increasing the value of x in ZnxCd1-xTe, Voc is in decreasing nature.

Hence, optimum value of ‘x’ was found in this work through simulation of ZnxCd1- xS/ZnxCd1-xTe solar cell. Till now, no reports (except present work’s reports) are available to simulate ZnxCd1-xS/ZnxCd1-xS through numerical modeling solar cells by AMPS-1D.

2.6 CHAPTER SUMMARY

Solar cell does not achieve its goal-low cost and high efficiency till now. Thin film solar cell give a dream to achieve it and CdTe thin film solar cell came on leading in this regards because CdTe has an optimum band gap, high absorption coefficients, good electronic properties and easy to get large scale production. Although CdTe has important advantages as mentioned above but CdTe based solar cell suffers from lack of high efficiency due to not absorbing all high energetic photons in the absorber layer (CdTe) and because of the drawback of its window layer (CdS) and back contact difficulties. Incorporating Zn both in absorber layer (CdTe) and window layer (CdS) could be one of the best choices to overcome the problem. The use of BSF layer may help to overcome the back contact problem. 35

CHAPTER III

METHODOLOGY

3.1 INTRODUCTION

The methodology of the research has been inotroduced in this chapter. One- dimensional AMPS numerical simulation of CdTe thin-film solar cells is the investigative approach used in this study. It is a powerful tool to build a reasonable physical model to test the viability and numerical simulations can help to predict any change in cell performance resulting from the modified resonable parameters. Moreover, difficult experimental tasks can sometimes be by-passed with numerical modeling and simulations. The operation of semiconductor devices can be described by a set of basic equations. These equations are coupled partial differential equations, for which it is often not possible to find general analytical solutions. These equations can be transformed into a set of nonlinear algebraic equations, which can be solved numerically with a computer. In this work, computer based program AMPS are utilized. This chapter discusses the set-up of the conventional CdS/CdT ecell model with the AMPS software. Input parameters are explained in the context of the base case, which was the baseline for subsequent variation investigations. A three-layer device model of a SnO2/CdS/CdTe solar cell is the starting point for the calculations in this work and it was modified to achieve better performance with extra Back Surface Field (BSF) layer for ultra thin cells. The material parameters used in this modeling, which were selected based on literature values, theory, or in some case reasonable estimations. CdTe is a tolerant material, good quality CdTe cells have been fabricated by a multitude of deposition techniques and in this chapter a summary of different deposition techniques are explored. Two characterization techniques for the 36

complete CdTe cells and ultra thin layer characterizations are also discussed here. The complete methodology has main three parts: numerical modeling and simulation, cell fabrication, measurements and characterization.

3.2 SOLAR CELL MODELING

In solar cell, a cell model is a theoretical struture created to represent real processes and parameters that might influence cell performance. Models are widely used in science and technology to simplify complicated systems, by neglecting the noncrucial features. Consideration of important and nonimportant parametersis the most important task for a successful cell model. In a good model all the important parameters are taken into account and the non important ones are omitted. Modeling is widely used in analysis of silicon solar cells. For thin film polycrystalline solar cells the need for numerical modeling methods is higher due to the complex nature [81]. Numerical modeling would help to understand the solar cells behaviour and should give the additional ideas to control the fabrication parameters to improve the cell performance. Once the cell is produced, it needs to be characterized to determine the performance and losses. Analysis of characterization curves indicates the nature and source of losses. Modeling techniques can provide physical explanation of the mechanism behind the cell performance. Numerical modeling may predict the changes in material properties on the solar cell performances and could suggest ways to change the fabrication process to improve the cell performances. Modeling would be a valuable tool to improve the performance, interpretation of modeling results requires great carefulness [82].

Modeling can be used to provide insight, to interpret measurements and to asses the potential merits of a cell structure. The number of variable parameters thouse could be varied in solar cell model would be larger than 50 [81]. A problem with about 50 variables is difficult to solve reliably. Thus, it is necessary to minimize the number of parameters by fixing many of them. The large number of parameters (<50) needed for numerical simulationis, however, a real concern. Therefore, numerical modeling needs input from measurements of layers to fix variable parameters as many as possible. Choosing a suitable parameter set to simulate a solar cell is not an 37

unsighted job which might be left to software, but it requires physical insight and experimental knowledge. There is a considerable scattering in the literature data on almost all electrical material properties and the parameters used for CdTe solar cell modeling [82]. It is a real challenge to choose the appropriate parameters of cell structure and different layers at the beginning of a cell model. Many of them depend on fabrication techniques and deposition methods and can thus vary among devices of the same batch of fabrication. Due to the large number of variable parameters, different sets of parameters can lead to similar output performance curves. Therefore, drawing conclusions through a fit of a single set of experimental data could be misleading. To use simulations for explaining the experimentally observation complete sets of experimental data and knowledge are necessary.

To simulate the CdTe based thin film solar cells, any numerical program which can solve the basic semiconductor equations could be used. These basic semiconductor equations are: Poisson’s equation, the continuity equation for electrons and the continuity equation for holes [82]. The physics of device transport can be achieved by solving these three governing equations along with necessary boundary conditions.

3.3 EQUATIONS INVOLVED FOR SIMULATION

To model the charge transport processes in the present structure shown in the Figure 3.1, the drift-diffusion approach is used as a function of device length (x). The three main equations are the Poisson’s equation, continuality equation for free electrons and continuity equation of free holes. Generally the Poisson’s equation is [83]:

푑 푑휓 (−휀(푥) ) = 푞. [푝(푥) − 푛(푥) + 푁+(푥) − 푁−(푥) + 푝 (푥) − 푛 (푥)] (3.1) 푑푥 푑푥 퐷 퐴 푡 푡

Where, ψ is the electrostatic potential, n, p are the concentration of free electrons and holes, nt, pt are the trapped electron and holes, ND⁺, ND⁻ are the concentrations of ionized donors and acceptors, ε is the dielectric permittivity of semiconductors and q is the electron charge. 38

X=0

X=L

Figure 3.1: Proposed solar cell structure used for the simulation.

The transport characteristics of an electronic device may be derived by the continuity equation for the holes and electrons. The continuity equations in steady state conditions are [83]:

1 푑퐽 푛 = 푅 (푥) − 퐺(푥) (3.2) 푞 푑푥 푛

1 푑퐽 푝 = 퐺(푥) − 푅 (푥) (3.3) 푞 푑푥 푝

Where, Jn, Jp are electron and hole current density, Rn, Rp are electron and holes recombination velocities for direct band-to-band and indirect transitions, and G is the optical generation rate which is expressed as a function of x is [83],

푑 푑 퐺(푥) = − ∑ ф퐹푂푅 (휆 ) + ∑ ф푅퐸푉( 휆 ) (3.4) 푑푥 푖 푖 푖 푑푥 푖 푖 푖

FOR Where φi (λi) represent the photon flux of frequency λi at some point x which is REV moving forward direction and φi (λi) represent the photon flux of frequency λi at some point x which is moving to the reverse direction, depending on the light absorption coefficient and the light reflectance in the forward and reverse direction and can be expressed as [83]:

2 퐹푂푅 [−훼(휆푖)푥] (−훼(휆푖)퐿) [−훼(휆푖)푥] ф푖 (휆푖) = ф0푖(휆푖). {푒 + 푅퐹푅퐵[푒 ] . [푒 + ⋯ }(3.5) 39

3 푅퐸푉 [−훼(휆푖)퐿] [−훼(휆푖)(퐿−푥)] (−훼(휆푖)퐿) [−훼(휆푖)(퐿−푥)] ф푖 (휆푖) = 푅퐵ф0푖(휆푖). {푒 . 푒 + 푅퐹푅퐵[푒 ] . 푒 + ⋯ } (3.6)

In these expression RF is the reflection coefficient for the internal surface at x=0 and

RB is the reflection coefficient for the internal surface at x=L.

Typically the current density for holes and electrons are represented by [83]

푑퐸푓 퐽 = −푞µ 푝 ( 푝) (3.7) 푝 푝 푑푥

푑퐸 퐽 = 푞µ 푛 ( 푓푛) (3.8) 푛 푛 푑푥

Here n is the free electron concentration and expressed as:

퐸퐹 −ф −휒푒(퐿)+휒푒−휓(푥) 푛 = 푁 퐹 [ 푝 푏퐿 ] (3.9) 푐 1⁄2 푘푇

Where,

푛 퐸퐹 푛 = 퐸푐 + 푘푏푇 푙푛 (3.10) 푁푐

퐸푐 = ф푏퐿 + 휒푒(퐿) − 휒푒 + 휓(푥) (3.11)

And p is the free hole concentration and express as:

ф +휒푒(퐿)−휒푒−퐸푔+휓(푥)+퐸퐹 푝 = 푁 퐹 [ 푏퐿 푝] (3.12) 푣 1⁄2 푘푇

Where,

푝 퐸퐹 푝 = 퐸푣 − 푘푏푇 푙푛 (3.13) 푁푣

퐸푣 = ф푏퐿 + 휒푒(퐿) − 휒푒 − 퐸푔 + 휓(푥) (3.14)

From equations 3.7-3.14 we have seen that current density of the solar cell is depends on the electron mobility (µn), hole mobility (µp), electron and hole concentration (n and p), band gap (Eg), Electron affinity (χ), effective density of state in the conduction band (NC) and effective density of state in the valance band (NV), barrier height at back contact (фbL), as well as device length (x).

40

3.3.1 Boundary Conditions

The three governing equations (3.1), (3.2), and (3.3) must hold at every position in a device and the solution to these equations involves determining the state variables 휓(x), EFn(x), and EFp(x) or, equivalently, 휓(x), n(x), and p(x) which completely defines the system at every point x. Because the governing equations for

휓(x), EFn(x), and EFp(x) (or, equivalently, 휓(x), n(x), and p(x)) are non-linear and coupled, they cannot be solved analytically. Hence, numerical methods must be utilized. Newton-Raphson technique, which is used in AMPS and AMPS to numerically solve the resulting algebraic equations. Like any other mathematical analysis, there must be boundary conditions imposed on the set of equations. These are expressed in terms of conditions on the local vacuum level and the currents at the contacts. To be specific the solutions to equations (3.1), (3.2), and (3.3) must satisfy the following boundary conditions [83]:

휓(0) = 휓0 − 푉 (3.15)

휓(퐿) = 0 (3.16)

퐽푝(0) = −푞푆푝표(푃표(0) − 푃(0)) (3.17)

퐽푝(퐿) = 푞푆푝퐿(푃(퐿) − 푃표(퐿)) (3.18)

퐽푛(0) = 푞푆푛표(푛(0) − 푛표(0)) (3.19)

퐽푛(퐿) = −푞푆푛퐿(푛표(퐿) − 푛(퐿)) (3.20)

Where x=0 refers to the front side and x=L to the back side of any general device structure under consideration. In boundary conditions 3.15 and 3.16 the quantities 휓(0) and 휓(L) are the function 휓 in equation 1 evaluated at x=0 and x=L in. In boundary condition statements 3.17-3.20 P0(0) and P0(L) are the valence band hole populations at x=0 and x=L, respectively. The quantity p(0) and p(L) are the corresponding hole populations, under operating conditions at x=0 and x=L, respectively. The quantity n(0) and n(L) are the corresponding electron populations, under operating conditions at x=0 and x=L, respectively. The quantities Sp0, SpL, Sn0, and SnL, appearing in conditions 3.17-3.20 are effective surface recombination speeds for holes at x=0 and x=L, respectively, and for electrons at x=0 and x=L, respectively. 41

Conditions 3.17-3.20 must be matched by equations 3.2 and 3.3 at x=0 and x=L under operating conditions.

3.3.2 Solution Technique

To numerically solve the three non-linear, coupled governing equations, the one-dimension device with length L is divided into N intervals and N +1 major grid points, shown as in Figure 3.2. The grid spacing need not be uniform. The Poisson’s equation and the continuity equations will be solved for each interval along with the appropriate boundary conditionsand the set of three variables Ψ, EFn and EFp are then solved at each particular grid point 1 to N + 1, represented by solid lines. In a computer algorithm, the solar cell structure is first divided into slices, typically some tens to more than a hundred per semiconductor layer [83]. For a five layer system, the number of slices can be e.g. N = 500. Each differential equation is approximated by a system of N nonlinear algebraic equations. Once the variables Ψ, EFn and EFp are determined under the light, voltage and temperature conditions and other variables such as electric field, carrier concentration, or trapped charges are defined, the recombination profiles, electron and hole current densitiesand other transport information may be obtained. Then the total current–voltage (J-V) characteristic can be obtained from J(x) = Jn(x) + Jp(x) [83]).

Intervals 1 2 3 N-2 N-1 N

x = 0 x = L

Grids 1 2 3 4 N-2 N-1 N N+1

Figure 3.2: Intervals and grids used in numerical method. There are N intervals (dashed lines) and N + 1 major grid points (solid lines) Source: Ref. [83]

42

3.3.3 Cell Performance Parameters Determination

Once the state variables 휓, EFn, and EFp are determined for a given set of biasing conditions (voltage, light, or both) and temperatures, the current density- voltage (J-V) characteristics for these conditions can be generated. The J-V characteristic for some temperature T, with or without the presence of light, is obtained from the fact that J=Jp(x) + Jn(x) where x is any plane in the device and Jn and Jp are obtained from Equations 3.7 and 3.8. Similarly the electrostatic field (ξ) throughout the device can be generated for the various conditions using ξ=d휓/dx and recombination can be generated using equation R=βnp, where β is a proportionality constant which depends on the energy-band structure of the material under analysis, and n and p are the band carrier concentrations present when devices are subjected to a voltage bias, light bias or both. Jsc=Jp(x)+Jn(x) (3.21)

The short circuit current density, Jsc = Isc/A, is often used to compare solar cells [84].

푞푉 ( ) 푘 푇 퐽 = 퐽0 (푒 퐵 − 1) − 퐽푠푐, 퐽푠푐 ≈ 퐽퐿 (3.22)

To find an expression for the open circuit voltage, Voc, we use (22) setting J = 0. The resulting expression is [84]:

푘퐵푇 퐽푠푐 푘퐵푇 퐼푠푐 푉표푐 = ln ( + 1) ≈ ln ( + 1) (3.23) 푞 퐽0 푞 퐼0

푑푃 When, = 0 then 퐷푉

푘퐵푇 1+(퐼푠푐⁄퐼0) 퐾퐵푇 푞푉푚 푉푚 = ln [ ] ≈ 푉표푐 − ln (1 + ) (3.24) 푞 1+(푞푉푚⁄푘퐵푇) 푞 푘퐵푇 and

푞푉푚 푞푉푚 푘 푇 1 퐼푚 = 퐼0 ( ) 푒 퐵 ≈ 퐼퐿 (1 − ) (3.25) 푘퐵푇 푞푉푚⁄푘퐵푇

The maximum power that can be delivering to the load is the combination of Vm and

Im:

푘퐵푇 푞푉푚 푘퐵푇 푃푚 = 푉푚. 퐼푚 = [푉표푐 − ln (1 + − )] . 퐼푠푐 (3.26) 푞 푘퐵푇 푞 The efficiency of the solar cell η can be drive [84]:

푘퐵푇 푞푉푚 푘퐵푇 푃 푉 퐼 퐹퐹푉 퐼 [푉표푐− 푙푛(1+ − )].퐼푠푐 휂 = 푚 = 푚 푚 = 표푐 푠푐 = 푞 푘퐵푇 푞 (3.27) 푃푖푛 푃푖푛 푃푖푛 푃푖푛 43

Pin is the incident power and FF is the fill factor which defines how the solar cell gets to the theoretical maximum value of power. The fill factor can be drive as [84]:

푃 푉 퐼 푘 푇 푞푉 푘 푇 퐹퐹 = 푚 = 푚 푚 = 1 − 퐵 ln (1 + 푚) − 퐵 (3.28) 푉표푐퐼푠푐 푉표푐퐼푠푐 푞푉표푐 푘퐵푇 푞푉표푐

3.4 SIMULATION SOFTWARE AND AMPS

A numerical solar cell simulator is a computer programme that numerically solves the ‘semiconductor equations’ for a given solar cell structure and parameters. Numerical simulation of a solar cell is a very important technique to predict the effect of physical changes on solar cell performance and to test the viability of the proposed cell structures. Several numerical programs have been developed and widely used. An overview of numerical simulators used for thin film solar cells is given in [81]. In this work, AMPS-1D software has been utilized to simulate the thin film CdS/CdTe solar cells with BSF.

The AMPS-1D is a one-dimension device simulation program was developed by Prof. S. J. Fonashet al. [83] in the Electronic Materials and Processing Research Laboratory at Pennsylvania State University. BETA version 1.0, which was revised in 2007, is used here. It can analyze the transport in a variety of crystalline, polycrystalline, or amorphous solar-cell materials and device structures including homojunction, heterojunction, or multijunction solar cells and detectors [40, 85]. The AMPS-1D program asks the user to input the specific parameters of the different layers to build the structure of the device need to be tested. When running the AMPS simulation, the program expects a set of default parameters. The user can save the default case and reset the parameters to be varied for a particular configuration. The advantages of AMPS include its user friendliness and the stability in general. It also has a very flexible plotting program, in which the user can generate output plots such as J-V curves, spectral response, band diagrams, carrier concentrations, currents, electrical field distribution and recombination profiles. However, AMPS has some disadvantages, such as the need to input all information including spectrum parameters by hand and the lack of interface treatment so that an interface in some cases need to be approximated by thin layers. In this work, the AMPS-1D has used to 44

analyze the proposed structures of ultra thin CdTe cells with and without back surface field.

3.5 INITIAL SETTING OF AMPS SIMULATION PARAMETERS

A starting parameter set for simulating the CdS/CdTe thin film solar cells is the ‘baseline set’ given by Gloeckler and Sites [86]. For analytical numerical simulation, it is a prerequisite to find a consistent parameter set. This means a single parameter set for one cell, which describes all measured effects. Three layers: n-SnO2 as transparent conducting oxide and cell front contact, n-CdS window and p-CdTe absorber from the baseline case structure was selected as the starting platform as shown in Figure 3.3. Baseline case parameters of CdTe cell are listed in Table 3.1, where S is the surface recombination velocity, W is the layer thickness, ΔEC is the conduction band offset. The other symbols in Table 3.1 have their usual definitions.

Table 3.1: SnO2/CdS/CdTe solar cell baseline parameters General Device Properties Front Back Фb [eV] Фbn = 0.1 Фbp= 0.4 7 7 Se [cm/s] 10 10 7 7 Sh [cm/s] 10 10 Reflectivity Rf 0.1 0.8 Layer Properties SnO2 CdS CdTe W [nm] 500 25 4000 ε/ε0 9 10 9.4 2 µe [cm /Vs] 100 100 320 2 µb [cm /Ns] 25 25 25 n, p [cm"3] n:l017 n: 1017 p: 2×1014 Eg [eV] 3.6 2.4 1.5 3 18 I8 17 Nc [cm 1 2.2×10 2.2×10 8×10 19 19 19 Nv [cm ] l.8×10 1.8×10 1.8×10

ΔEc [eV] 0.0 -0.1

Source: Ref. [86]

There are three types of parameters that are necessary to enter in the device simulation window before any AMPS simulation starts:

A) The material properties for each layer with front and back contacts. 45

B) Environmental conditions such as operating temperature. C) Modeling settings: model type, grid spacing for the numerical calculations, bias voltages.

Figure 3.3: The conventional baseline case cell structure

3.5.1 Material Parameters

The material parameters that define the base case are presented in Table 3.1. A brief discussion of these numbers is given in following sections. The number of incident photons was entered for wavelengths between 0.38 mm and 0.86 nm, with a step size of 0.02 nm. It is not necessary to specify the illumination spectrum ahead of these limits since, as CdTe (1.45 eV) solar cells have only measurable quantum efficiency in the range of 0.35 – 0.86 nm. The temperature for the base case and for all other cells was set to 298 K (25 °C) if not specify. The effect Operating temperature on selected cell’s performance is addressed in Chapter 5.

The AMPS software can operate in two distinct modes: the density of state (DOS) mode or the lifetime mode. A details description of both modes can be found in the AMPS manual [84]; a comparative discussion can be found in A.L. Fahrenbruch’s 46

work on CdTe solar cells [87]. However, the lifetime mode accepts inputs in the form of carrier lifetimes, which are assumed constant, independent of light and voltage bias. The DOS mode allows the definition of multiple defect states, using densities, energy distributions and capture cross-sections. All modeling for this work the lifetime mode isused as lifetime is more important than defect states in CdTe cells.Denser grid spacing was selected for the thinner layers of the cell, where more rapid changes are expected.

3.5.2 General Device Parameters

The front and back contacts are exclusively defined by their work function, possible recombination velocities and the reflectivity of the contact semiconductor interface: фb0 = -0.1 eV fit the conduction band at 0.1 eV belowthe Fermi level (EF) at x = 0 nm (front contact) with a small cliff (-0.1 eV) and calculated фbL = 1.1 eV fit the conduction band at 1.1 eV above EF at x = L (back contact). These numbers create an ohmic contact at the front and a negligible barrier at the back contact. Strong back barriers are frequently observed in a typical thin film CdTe solar cell with different back contact formation and will be discussed in Chapter 4. Interfaces between polycrystalline layers are rich in defect states, generated by lattice mismatch and impurities that cause recombination losses. The parameter used to describe this recombination current is given in terms of a surface recombination velocity [83]. Allnumerical calculation for this work unless specified used a surface recombination velocity of 107 cm/s, which corresponds approximately to the thermal velocity of the electrons, meaning that the entire carrier will recombine if they can reach surface.The front surface reflectivity limits quantum efficiency and therefore Jsc of the cells. This parameter is set to RF = 0.1 (10%) in order to reflect the experimental spectral response data of CdS/CdTe solar cells with typical window layer. The back surface reflection has negligible influence on the thicker cell performance since a few absorbable photons (close to the CdTe band gap) transverse the device and therefore gets a chance of being reflected but it gains more importat for the thinner absorber layer.Thus, in the case of thicker cells the reflection photon has a little chance but for ultra thin cells it might be an important factor. This parameter is set to RB = 0.9 (90%) in order to get reflected back energetic photons from the back surface (if any). The 47

voltage biasing window contains information about the voltage which needs to be applied for proper simulation.

3.5.3 Layer Information

The thicknesses of the different layers (TCO, ZnCdS, ZnCdTe and BSF) are chosen as they are found in typical experimental devices and literatures. Dielectric constants, band gaps, mobility and effective density of states values are common knowledge and can be found in literatures [82]. About carrier concentrations ZnO and ZnCdS layers are designated as n-type materials with high carrier concentrations. CdTe is p-type with a 2×1014 commonly measured carrier concentration. All these layer parameters have been inserted in the simulation window of AMPS simulator as shown in Figure 3.4. The band offsets in the model are defined via the electron affinities of the different layers. The entered absorption coefficient α(λ) of each layer reflects how strongly light of a specific wavelength (λ) is absorbed in the layer or pass through to the next layer.

Figure 3.4: AMPS device definition window with layers parameters, front and back contact parameters

48

3.6 PROPOSED CELL STRUCTURE AND PARAMETER SET

A starting parameter set for simulating the CdS/CdTethin film solar cells is the ‘baseline set’ given by Gloeckler and Sites [86]. In this stage of work the main objective is to construct a valued numerical simulation, which is a prerequisite to find a consistent cell parameter set. Thismeans a single parameter set for one cell, which describes all measured effects in the CdS/CdTe cell. The baseline case was utilized to approximate the reasonable cell efficiency and it willmodify for higher performance of CdTe cells. Many researchers have carried out simulations with CdTe based solar cells and a list can be found in the references given in the review paper by Burgelman et al. (2004) [81]. The modified structure of the proposed solar cell

(Glass/SnO2/ZnO/ZnCdS/ZnCdTe/BSF/Metal) from the base line case cell is shown in Figure 3.5.

Figure 3.5: Modified cell structure for higher performance

It is clear from figure 3.5 that the new structure has an extra layer (Zn2SnO4) in between ITO and ZnCdS layer to achieve thinner ZnCdS window. Thus, the front contact consists of ITO plus a buffer layer. Suitable BSF were inserted to reduce 49

minority carrier recombination loss at the back contact in ultra thin cellsand to formation of flat barrier height at back contact. Cheaper metals (Cu/Te/Al/Mo/Ag/Ni) were used as final back contact material in the proposed cells for the low cost aspect of the cell. The number of parameters that can be varied in a particular solar cell model is larger than 50 [81]. A problem with 50 variables is difficult to solve reliably. It is therefore necessary to minimize the number of variable parameters by fixing many of them at reasonable values. It was a real challenge to choose the appropriate parameter set of the cells. Many of them depend on fabrication techniques and deposition methods and can thus vary even between devices fabricated in the same batch.

Four layers that were emphasized in the modified cells are the buffer layer

(ZnO/Zn2SnO4), n-ZnCdS layer, p-ZnCdTe layer and a BSF (ZnTe/ PbTe/ As2Te3/

Cu2Te) layer along with suitable front and back contact. In the modeling, all these layer thickness were varied (one at a time by keeping all other parameter at the fixed values): the ZnCdTe layer from 100 nm up to 4 µm, ZnCdS window layer from 25 nm to 200 nm, ZnO or Zn2SnO4 buffer layer from 25 nm to 200 nm and BSF layer from 10 nm to 500 nm while keeping all other cell parameters at fixed value as described in the next chapter. This modeling has investigated the possibility of achieving efficient ultra thin ZnCdS/ZnCdTe solar cells, the details is presented in Chapter 4 and 5.

3.7 CURRENT-VOLTAGE (I-V) CHARACTERISTICS OF SOLAR CELLS

The most common way to describe the performance of a solar cell is with its current- voltage characteristics. The current-voltage curve of an ideal solar cell in the dark follows the exponential diode law (long-dash line in Figure 3.6). When the cell is illuminated, the curve is shifted downwards by an amount referred to as the light generated current JL, in ideal case it is independent of voltage and the same as the short-circuit current Jsc [88].

푞푉 퐽 = 퐽 퐽 = 퐽 푒푥푝 [ − 1] − 퐽 (3.29) 퐷− 퐿 0 퐴푘푇 퐿

Where, J0 is the diode saturation current, A is diode quality factor (equal to 1 in the ideal case), q is elemental charge, k is Boltzman constant and T is temperature. 50

Figure 3.6: J-V curves of solar cells

Conversion efficiency is generally the parameter of most interest for solar cell applications. It is often broken down into three different parameters: Jsc, Voc and FF. The Jsc, i.e. the current at V = 0, depends on the number of photo- generated carriers and there collection efficiency. The number of generated carriers can be maximized by minimizing the area taken by contact grids (if any on the front contact) and by sufficiently thick absorbers, which allow all of the photons with sufficient energy to be absorbed. Collection efficiency depends on the recombination mechanisms as well. The losses in Jsc can be analyzed from the quantum efficiency curves (Chapter 4). The Voc is the voltage at zero current, when the forward current balances the photogenerated current. The Voc is equal to the difference in quasi-Fermi levels for holes and electrons between the two sides of the cell. From the diode equation, Voc is equal to [88]:

퐴푘푇 퐽퐿 푉표푐 = 푙푛 ( + 1) (3.30) 푞 퐽표

The saturation current J0 depends on the material properties, the cell structure and is limited by recombination coming from several different recombination mechanisms: recombination in the bulk CdTe, in the SCR and at the CdS/CdTe interface. 51

The power produces from a solar cell is a product of current and voltage. At some point on the current-voltage curve (for a conostant load) this product has a maximum 1 (100%). The current density will be referred to as current throughout the content value. That point is maximum power pointand the corresponding current and voltage are referred to as the maximum power current Jmp and the maximum power voltage Vmp. In addition to Voc and Jsc, the maximum power depends on how “square” the I-V curve is. The “squareness” of the I-V curve is defined by fill factor (FF) and can be defined [88]:

푉 퐽 퐹퐹 = 푚푝 푚푝 (3.31) 푉표푐퐽푠푐

Finally, the conversion efficiency involves all these three parameters and can be calculated as [88]:

푃 푉 휂 = 표푢푡 = 표푐퐽푠푐 퐹퐹 (3.32) 푃푖푛 푃푖푛

Where, Pin is the incident light power on the cell. It is commonly taken to be 100 mW/cm2 for standard solar illumination. This illumination is referred to as AM1.5 illuminationand it is equivalent to sunlight passing through 1.5 times the air mass of vertical illumination. Equation (3.29) is the ideal diode equation, but real solar cells have additional losses, such as series resistance and leakage or shunt conductance. The series resistance Rs comes from the resistance in the bulk of the semiconductor and the contact resistances. Shunt resistance Rsh representsthe parallel paths for any current flow.

3.8 CHAPTER SUMMARY

Modeling is widely used in analysis of silicon solar cells. For thin film ZnCdTe solar cells the need of numerical modeling methods is higher due to the complex nature of the polycrystalline materials.The insight operation of the ZnCdTe solar cells can be achieved by solving mentioned three governing equations along with the appropriate boundary conditions and layer parameters.In this work, AMPS-1D was utilized to simulate the thin film ZnCdS/ZnCdTe solar cells.The baseline case of CdTe solar cells was was the starting point and it was modified to achieve higher performance of the ultra thin cell with BSF.The material parameters used in the modeling, which were 52

selected based on literature values, theory or in some case, reasonable estimations.CdTe is the most promising thin film photovoltaic materials as low cost variety of deposition methods are available.For the complete solar cell the I- Vcharacterization techniques isused to evaluate the cell. Thus, the entire research methodology has three main parts: numerically analysis, fabrication and evaluation. The next chapters will discuss the details of numerical analysis of the proposed solar cells.

53

CHAPTER IV

OPTIMISATION OF CELL STRUCTURE FROM NUMERICAL ANALYSIS

4.1 INTRODUCTION

In this work the thin film CdS/CdTe solar cell with conventional structure were chosen for modifying its structure to achieve an ultra thin high efficient solar cell. At the beginning of this work, the base line CdTe solar cell was chosen and it was modified by changing different layer structure, varying layer thickness and selecting suitable front and back contact for getting higher performance. Commercially available ITO coated glass was selected as substrate and low resistive conducting ITO as the front contact for all solar cells.

By adding an extra Zn2SnO4 buffer layer in the conventional cell structure and incorporating Zn with the both CdS window layer and CdTe absorber layer, the window layer thickness was reduce to 80 nm as well as it provide more transparent window layer in the blue region. And one of the main goals of today’s solar cell research is using less semiconductor material by making the cell thinner. Thinning will not only save materials but also reduce cell production energy and time without compromising its conversion efficiency. Therefore in an attempt towards a higher performance thin film solar cell structure, this study proposed the use of a ternary semiconductor Zink Cadmium Telluride (ZnCdTe) as an absorber layer to form ZnCdS/ZnCdTe solar cell. The formation of stable and non rectifying back contact to ZnCdTe is one of the major and critical problems for ZnCdTe based solar cells. Typically, metals with a high work function (φm≥ 5.9 eV) are required to make an ohmic contact to p- ZnCdTe but most metals do not possess such high work functions. Potential metals are tried as the final metal back contact of the proposed cells, but only a few of them which are rare- earth and expensive like Te, Au, Pd and Pt show acceptable performance but others are have shown inferior cell performance. To overcome this obstacle a usual approach is to either 54 reduce the barrier height or moderate its width by inserting a highly doping extra layer to produce back surface field (BSF) with appropriate material in between the ZnCdTe and final metal back contact. This BSF has been found to reduce the back surface recombination velocity (BSRV) and can solve back contact problem of proposed ZnCdS/ZnCdTesolar cells.

4.2 PROPOSED SOLAR CELL STRUCTURE

The typical solar cell structure of a CdTe/CdS solar cell is mainly composed of four layers which are shown in the figure 4.1 (a):

1. A transparent & conducting oxide (TCO) about 500 nm which acts as a front contact generally deposited on high quality glass substrate. 2. A thin CdS film which is the so-called window layer about 50nm thickness on top of the front contact (TCO). 3. Few µm thick CdTe layer, which is the absorber layer and deposited on top of the CdS to complete the pn junction. 4. Finally the back contact to be formed on the top of CdTe layer to complete the cell and contacts.

The light is incident from the glass side through the front contact into the CdS layer then finally to the CdTe absorber layer where the cell is designed to absorb most of the incident photon in the CdTe layer only.

The baseline case of CdTe cell [86] was utilized to approximate the highest efficiency CdS/CdTe solar cells at that time from numerical modelling, and it was modified in this work to explore the possibility of efficient ultra thin ZnxCd1-xS/ZnCdTe cells with suitable BSF. The modified solar cell structures are shown in Figure 4.1 (b), which is composed of five main layers:

1. A transparent and conducting oxide (TCO) about few hundred nm which acts as a front contact generally deposited on high quality glass substrate.

2. A thin Zn2SnO4 film which is called buffer layer about 100nm thickness top of the front contact TCO.

3. A thin ZnxCd1-xS film which is the so-called window layer about 80 nm thicknesses

on top of the Zn2SnO4 buffer layer. 55

4. Few μm thick ZnCdTe layers, which is the absorber layer and deposited on top of

ZnxCd1-xS to complete the p-n junction. 5. Finally, the back contact with BSF to be formed on top of the ZnCdTe layer to complete the cell. The light is incident from the glass side through the front contact

into the ZnxCd1-xS layer then finally to the ZnCdTe absorber layer where the cell is designed to absorb most of the incident photon in the ZnCdTe layer only.

Figure 4.1: Structure of the (a) conventional CdTe thin film solar cell and (b) modified solar cell structure for higher performance.

It is clear from the figure 4.1 (b) that the proposed structure has an extra layer

(Zn2SnO4) in between front contact and ZnCdS layer to achieve thinner ZnCdS window layer and ITO material, thus, the front contact consist of ITO plus a buffer layer of Zn2SnO4 and this buffer layer is absent in the base line case cell, rather it has 500 nm SnO2 as the front contact. The doping concentration (~1014 cm-3) used in the base line case was changed to (~1015 cm-3) for today’s achievable values for ZnCdTe material. To form ohomic contact on ZnCdTe, there is no low cost metals with appropriate high work function (5.9 eV) thus, Au:Cu has been used in most of the cells although some other contact material like Ni, Mo or Al based contacts have shown promising results. Among other approach surface etching or a 56 pseudo-ohmic contact was developed by Bosio et al. (2006) [89] for CdTe solar cells. In the pseudo-ohmic contact approach, a highly doped semiconductor is first deposited on the treated ZnCdTe surface as a BSF layer followed by the application of a thin metal contact.

Semiconductors including HgTe, ZnTe, Sb2Te3, As2Te3, and CuxTe (x=2), might be deposited between ZnCdTe and the contact metal to achieve a good pseudo-ohmic contact. ZnTe,

Cu2Te, PbTe and As2Te3 were used in this work for the same purpose with expected BSF effects on the ZnCdTe cell. Etching of the ZnCdTe back surface with Br– methanol, NP solution, etc., might creates a Te-rich p+ region, which could be useful for producing good contact, it might be attributed to the elemental of Cu which might leads to p++ Cu2Te layer for the CdCl2 treated ZnCdTe based cells. All of the above ideas have been simulated utilizing AMPS simulator to obtain high efficiency ultra thin ZnCdTe solar cells with improved performance. These BSF material insertions might reduce minority carrier recombination loss at the back contact in ultra thin ZnCdS/ZnCdTe cells and help to improve the cell stability at higher operating temperature. To investigate the back contact problems and stability issues different potential back contact material (Te/Cu/Ni/Ag/Mo/Al) and different BSF material were inserted in the modified cells.

4.3 CONVENTIONAL AND MODIFIED CELL STRUCTURE ANALYSIS

Before modelling the modified ZnCdTe cell structures, the following issues of the widely used conventional CdTe cells of Figure 4.1 are addressed. High-efficiency ZnCdTe devices are generally fabricated with a buffer layer to prevent forward current leakage. The conventional CdS/CdTe cells usually have lower open circuit voltage (Voc) than their counterparts like CIGS or CIS cells. The Voc of the conventional cell can be improved by higher carrier density of ZnCdTe (~1015 cm-3) and higher absorber lifetime (>1 ns) and by reducing the back-contact barrier height with proper back contact material system. The fill factor (FF) might be improved by formation of good ohmic back contact and by reducing the thickness of ZnCdTe absorber material whose usual value is 5-10 μm. The short circuit current density (Jsc) of the cell can be improved by selecting proper window layer with higher transmittance, lower reflection and with reduced CdS window layer, in this view. The stability of the cell might be improved by inserting stable BSF with low band gap material like As2Te3 and PbTe or with higher band gap material like Cu2T or ZnTe. All of the above ideas were examined in this work through the simulation of different cell structures with and without BSF layer. The starting parameter set for simulation of the CdS/CdTe thin film solar 57 cells is the ‘baseline case set’ given by Gloeckler et al. (2003) [86]. For a valued numerical simulation, it is a very important to find a consistent materials parameter set. This means a single acceptable parameter set for the cell, which describes all experimental effects of the cell. Thus, the starting point of this work was the base line case and the structure of this baseline case was modified with different front contact and different back contact with BSF layer. Figure 4.1 illustrates the ZnCdTe baseline case structure and the modified structures investigated in this work.

Glass

Front contact: SnO2 ⁻ Buffer layer: Zn2SnO4

Window layer: ZnCdS Absorber layer: ZnCdTe

BSF layer: As2Te3, ZnTe, ,PbTe, Cu2Te

Back contact: Ag, Al, Cu, Te ⁺

Figure 4.2: Structure of the proposed ultra thin film ZnCdTe solar cell structure for higher performance.

As can be seen in Figure 4.2, the modified structures have an extra layer (Zn2SnO4) between ITO and ZnCdS window layer to support ultra-thin ZnCdS window layer for better performance in the blue region. Thus, the front contact consists of ITO along with a buffer layer. The doping concentration used in the baseline case (~1014 cm-3) has been changed to ~1015cm-3 for achieving higher efficiency as discussed before. To investigate the back contact issue and stability problem different back contact materials (Te/Cu/Ni/Ag/Mo/Al) and BSF are inserted in the modified cells. In some of the modified cell structures, Te/Cu/Ni/Ag/Mo/Al is used as back contact material with reasonable back contact barrier height in place of the Au:Cu (Фb= 0.4 eV) as back contact material of base line case considering the low cost and availability of these materials for industrial mass production.

4.4 SELECTION OF FRONT CONTACT

In general, any suitable member of the TCO family with higher transmittance can be used as the front contact of superstrate configured ZnCdTe based thin film solar cells. For a front contact of the ZnCdS/ZnCdTe based solar cells, the most important characteristics that 58 is required such as high transparency greater than 85% in the effective wavelength region, low resistivity in the order of 2×10-4 Ωcm or a sheet resistance less than 10 Ω/ square, good stability at higher operating temperature. The later one means that no diffusion from the front contact TCO into the other layers which need to be deposited subsequently to complete cell structure. The properties of some of the most important materials are used as TCO in ZnCdS/ZnCdTe solar cells [90-91] that are given in Table 4.1.

Table 4.1 Properties of some common TCO and buffer layers

Material Resistivity (Ω cm) Transparency (%) -4 SnO2 8×10 80 -4 In2O3:Sn (ITO) 2×10 >80 -4 In2O3:Ga (IGO) 2×10 85 -2 In2O3:F 10 85 -4 Cd2SnO4 (CTO) 2×10 85 -2 Zn2SnO4 (ZTO) 10 90 ZnO:In 8×10-4 85

Source: Ref. [90]

It is clear from the table that IGO, CTO and ITO are suitable for front contact of the cells. The IGO has Ga content, which is avoided considering its cost of mass production. The -4 Cd2SnO4 has a transparency about 85% and resistivity is low about 2×10 Ωcm, which seem the best material as front contact, champion cell also used it but in order to reduce the Cd use in solar cell the ITO which has a transparency above 80% and resistivity around 2×10-4 Ωcm was selected to investigate the performance. Moreover, ITO coated glass substrate are commercially available but Cd2SnO4 need to be sputtered on the glass substrate. Complete solar cells with CTO and ITO were simulated to compare the performance of CTO and ITO; it was found that cells with ITO have almost similar performance to those with CTO, details about CTO and ITO as a front contact material can be found [91]. In this analysis Zn2SnO4 is used as buffer layer.

4.5 SELECTION OF BUFFER LAYER

In order to improve the performance of thin film ZnCdS/ZnCdTe solar cell the ZnCdS window layer needs to be reduced; but reduced ZnCdS layer has been attributed to the 59 forward leakage current to front contact through pinholes in the reduced ZnCdS layer. This approach consist of a thinner high resistive buffer layer of a suitable material, to minimize the forward leakage current through pinholes of the ZnCdS layer and a highly conducting layer for front contact and current collection. In this work, ITO was selected as front contact layer, reasons explained earlier but it might require a buffer layer. The presence of the smoother high resistive suitable buffer layer, improves the ZnCdS film morphology by producing large grains during film growth. Among the several potential buffer layer materials, ZnO and

Zn2SnO4 were investigated in this work. Doped ZnO showed promising results to be implemented in thin film solar cells [93] and Zn2SnO4 was used as a buffer layer in the highest efficiency CdS/CdTe solar cell [11]. Doped and undoped ZnO and Zn2SnO4 as a buffer layers were inserted in the designed ZnCdTe cell to compare their performance numerically. A numerical analysis was done using AMPS-1D to investigate the cells performances with proposed buffer layers (ZnO or Zn2SnO4). The ZnO insertion shows almost the same performance as with Zn2SnO4 insertion. The buffer layers were used with 1 µm ZnCdTe, 80 nm ZnCdS, 100 nm TCO layers and all other cell parameter set as in Table 5.1. The I-V characteristics of the proposed cells from AMPS-1D simulation with buffer layers are shown in Figure 4.3.

Figure 4.3: Effect of ZnO and Zn2SnO4 on the ZnCdTe solar cell I-V characteristics

It is evident from the Figure 4.3 that with the insertion of ZnO or Zn2SnO4 buffer layer the I-V curve shows acceptable performance with Jsc over 30mA/cm2. The insertion of

ZnO has shown almost the same result with Zn2SnO4 insertion. It might be attributed to the 60 material properties which are very close to each other. Further numerical calculation was done to observe the effect of buffer layer thickness on cell performance. It has been found that cell characteristics are independent of the ZnO/Zn2SnO4 thickness from 50 nm up to 500 nm. Moreover, it was observed that the spectral response (SR) of the cells were unaffected by the ZnO/Zn2SnO4 thickness variation from 50 nm to 500 nm, meaning that a very thin buffer layer is adequate for the purpose. Considering the reality in fabrication, we have selected the

Zn2SnO4 buffer layer thickness of 100 nm with acceptable conversion efficiency for all the cells in this work

4.6 SELECTION OF WINDOW LAYER

One of the main goals of today’s solar cell research is using less semiconductor material by making the cells thinner. Thinning will not only save material, but will also reduce production time, and the energy requires producing them. All of these aspects will decrease the manufacturing cost of cells. The CdTe thin film solar cells have shown promising efficiency [11] and long-term stable performance [10] under AM1.5 illumination for global usage. However, conversion efficiencies of CdTe solar cells with homojunction have not shown encouraging results. Thus, in CdS/CdTe solar cells incorporate of zinc (Zn) in the window layer (CdS) is very promising in order to achieve high efficiency, reliable and stable solar cells than the other counterparts [15, 94]. ZnxCd1-xS is gaining prominence as good candidate for wide band-gap material in the field of photovoltaic solar cells.

The CdS window layer has a lower band gap, which causes significant absorption in the short-wavelength region which is below 500 nm. Substituting ZnxCd1-xS as an alternative window layer with a higher band-gap than CdS is a promising approach. In this study,

ZnxCd1-xS has been substituted for CdS as it can provide a more transparent window in the blue region (<500nm). As demonstrated by Oladeji et al. [70] and several other researchers

[23, 74], the spectral response in the blue region can be significantly enhanced using ZnxCd1- xS as a window layer in CdTe solar cells. Moreover, ZnxCd1-xS films can be deposited in a variety of ways: Vacuum evaporation [75], metal organic chemical vapor deposition (MOCVD) [95], spray pyrolysis [96], successive ionic absorption and reaction (SILAR) [97], photochemical deposition [98] and chemical bath deposition [99]. The publication of Yin et al. [96] reported on CdZnS/CdTe junctions in which the Zinc concentration was chosen to be 61 round 8%. Oladeiji’s et al. [70] showed improved Quantum Efficiency (QE) in comparison to CdS device.

4.7 SELECTION OF ABSORBER LAYER

CdTe semiconducting material has shown great potentials as an absorber material for thin film solar cells with solar cells based on this material attaining a maximum efficiency of 16.5%. However, for this thin film technology to compete with other conventional energy sources, its efficiency needs to be improved, where tandem cell structure is considered as possible means of obtaining up to 30% efficiency. In an attempt towards a higher efficiency thin film solar cell structure; this study proposes the use of a ternary semiconductor cadmium zinc telluride (CdZnTe).

On the other hand to get optimum performance from CdS/CdTe cell we have seen that the thickness of the cell have to be few micrometers which increase the thickness of the cell as well as increase the production cost and time of the layer production process. So that it is a great challenge to reduce the absorber layer to increase the cell performance as well as reduce the production cost. In this connection we are introducing Zn with the CdTe layer to reduce production cost as well as production time.

Figure 4.4: Comparison of JV characteristics curve of CdTe and ZnCdTe cell.

As discussed in the previous section, in this proposed cell added layer, Zn2SnO4 between TCO and window layer which reduce the window layer thickness as well as 62 increases the photon collection capacity. The J-V characteristic curve comparison between conventional cell and our proposed cell was shown in the figure 4.4. In both cell structure the absorber layer is kept constant at 1 µm and temperature was set at 25o C.

4.8 BACK CONTACT AND BARRIER HEIGHT ANALYSIS

The main target of this work is the development of an ultra thin solar cell without much compromise on cell performance. The ZnCdTe absorber and back contact interface are located closer to the main p-n junction in the case of thinner cells. Thus, the choice of the back contact material in ultra thin cells has high impact on the cell performance quite unlike of thicker cells. The formation of a low resistance as well as low barrier back-contact is one of the most challenging aspects in the fabrication of high performance ZnCdTe based solar cells.

In general, metal-to-semiconductor contacts can be having either as a rectifying (Schottky) or as an ohomic contact depending on the characteristics of the interface. For a p- type semiconductor with band gap Eg and electron affinity χ, and a metal with work function

Фm, an ohomic metal/semiconductor contact is formed when:

Фm>Eg + χ (4.1) and rectifying contact is formed when:

Фm

Both ohmic and Schottky metal/p-semiconductor interfaces are shown in Figure 4.5.

At the Schottky-contact interface, majority carriers (holes) see a barrier Фm as they travel from the semiconductor towards the metal, but such a barrier is absent in the case of ohmic contact interface. 63

Figure 4.5: Ohmic and rectifying metal/p-semiconductor contacts.

A tunneling contact can in some cases be formed by heavily doping the semiconductor surface. Tunneling contacts, however, require high doping (1018 cm −3) and a thin layer (<3 nm). Since ZnCdTe is heavily compensated, it is difficult to achieve a hole density near 1018cm −3, and hence tunneling mechanism is unlikely to happen. ZnCdTe is a p- type semiconductor with a high electron affinity (χ=4.4eV) and a high band gap (Eg=1.53eV), and thus a metal with a high work function (Фm ≥ 5.9ev) is required to make an ohmic contact to ZnCdTe. Most metals, however, do not have sufficiently high work functions and therefore form Schottky-barrier contact to ZnCdTe absorber layer. Table 4.2 tabulates the work functions of typical metals that can be making contact to p-ZnCdTe.

If any of the metals in Table 4.2 is deposited directly on a p-ZnCdTe surface, then a

Schottky contact is formed at the junction, and the contact barrier height, Фb, for holes at the interface (figure 4.5), in the absence of interface states, is given by the difference between the valence band edge and the Fermi energy in the metal.

Eg Фb = + (χ − Фm) (4.3) q where q is the elementary charge. The barrier height at the metal/p-ZnCdTe interface is calculated using equation (4.3) which is also shown in Table 4.2. A zero or negative value would be preferred, though in practice thermal effects make 0.3 eV or less sufficient. 64

Table 4.2: Metal work function Фm and resulting hole barrier Фb in ZnCdTe

Metal Фm (eV) Фb (eV) Ag 4.26 1.69 Al 4.28 1.67 Au 5.10 .85 Cu 4.65 1.30 Cr 4.50 1.45 In 4.12 1.83 Mo 4.60 1.35 Ni 5.15 0.80 Pd 5.12 0.83 Pt 5.65 0.3 Sb 4.55 1.40 Te 4.95 1.00 Ti 4.33 1.62 V 4.30 1.65

The presence of a back contact barrier can significantly affect the current-voltage characteristics of a p-ZnCdTe based solar cell, primarily by impeding hole transport. This current limiting effect is commonly referred to as “rollover”. Figure 4.6, shows a typical J–V curve from a ZnCdTe device with a rollover caused by a 0.52 eV barrier at the back contact.

Figure 4.6: J-V curve with rollover 65

Since no metal has a sufficiently high work–function to make an ohmic contact to p- ZnCdTe numerous approaches to form low resistance ohmic contacts have been attempted. Most have been directed towards special etching treatments of the ZnCdTe surface prior to metal deposition. The metal/semiconductor contact depends strongly on the surface condition of the semiconductor, and some surface treatments have achieved improved contacts.

Often the ZnCdTe surface is chemically etched to create a Te-rich surface, and then

Cu is incorporated as a key element on the surface followed by anneals to form a CuxTe layer that improves the back-contact property. Cu can be incorporated as elemental Cu, followed by a metal or carbon paint to form the electrode [100–102]. Alternatively, CuxTe and HgTe doped graphite paste may be used as a Cu source [103]. Other compounds such as HgTe

[104], ZnTe:Cu [105] and As2Te3 [106] have also been used as contacts to p-ZnCdTe. High performance ZnCdS/ZnCdTe solar cells often utilize CuxTe and HgTe doped graphite paste, and this is the back-contact process for standard devices at the National Renewable Energy Laboratory (NREL). The use of a Cu-doped graphite paste offers a quick and simple process. It is difficult, however, to reproduce or control this process. Moreover, it may not be scalable to large volume production. Hence, an alternative approach that incorporates evaporated Cu as the primary route to a CuxTe contact layer and evaporated metals for a secondary contact is utilized on NREL devices in this study. Though Cu inclusion at the back contact layer insures a non–rectifying contact, its use has been also linked to stability problems in ZnCdTe solar cells.

4.9 BACK SURFACE RECOMBINATION RATE ANALYSIS

An ideal back contact for thick cells which has low resistivity, allows for diffusion of Na from the soda lime glass (SLG) substrate into the absorber, and blocks the diffusion of other impurities. The optical properties of the back contact material for thick cells are not important, since for thicknesses above 1.5 µm of absorber layer, none of the incident photons can reach the back contact [107]. Current losses due to absorption at the back contact can increase when the absorber layer is ultra thin.

66

Figure 4.7: Effect of back surface recombination velocity on ZnCdTe cell

The main target of this work is to form a ultra-thin ZnCdTe cell using different technique, so that back surface recombination rate (BSRR) is a important issue which will be discussed here. In this work, keeping the ZnCdTe thickness constant to 1 µm, the back surface recombination rate is changed to investigate the effect in the cell performance. In this simulation, 1 µm ZnCdTe along with other selected parameters are kept at constant value, the back surface recombination rate was changed from 103 cm/s to 109 cm/s. In thinner cells, however, as the generation profile shifts closer to the back contact, the properties of the back contact material start to affect the device performance significantly. Once the thickness of the absorber (W) is lower than d+Ln, where d is the depletion-region width and Ln is the diffusion length for electrons in the absorber, the recombination for minority electrons at the interface between the ZnCdTe and the back-contact material becomes a significant loss. From this 67 analysis, it was found that if the BSRR decreases all the cell output characteristics increase drastically. This clearly demonstrates the effect that occurs in the case of ultra thin ZnCdTe cells as results of carrier recombination near the back contact. To reduce the BSRR of this ultra thin cell some new configure ration with BSF need to be introduced. The BSF will reflect back the minority carrier (electron) towards front contact before possible recombination at the back.

4.10 ADDITION OF BSF LAYER

As found in the previous analysis, to form an ultra thin solar cell Zn is incorporated in the conventional CdS/CdTe cell. Now suitable BSF layer will be added in between p-ZnCdTe and back contact of the ZnCds/ZnCdTe cell to restrain the possible recombination losses and reflect back the minority carrier (electron) towards front contact before possible recombination in such ultra thin (<1 μm) ZnCdTe cell. The specific BSF materials chosen to investigate in this work are Zinc Telluride (ZnTe), Cupric Telluride (Cu2Te), Arsenic

Telluride (As2Te3) and Lead Telluride (PbTe). The ZnTe, a p-type semiconductor with a direct band gap of 2.25 eV and formation of ohmic contacts to ZnTe is easier due to its lower electron affinity (3.65 eV) and the ability to dope it highly p-type. The Cu2Te is a p-type semiconductor with a wide optical band gap of 1.08 eV is used as BSF of ZnCdS/ZnCdTe solar cells. Another potential As2Te3 material as BSF has already introduced in the CdTe thin film solar cell. As2Te3 is a p-type semiconductor which has a forbidden gap of about 0.6 eV and exhibits resistivity of 10-3 Ω-cm at room temperature. It melts at 360°C and can evaporate at temperatures higher than 250°C in vacuum; they have reported a CdS/CdTe cell with efficiency of 15.8% using As2Te3 material. In this work we introduce a new BSF material, Lead Telluride (PbTe). PbTe is a p-type semiconductor with a narrow band gap of 0.32 eV and used as a BSF in ZnCdS/ZnCdTe cell. All these BSF layer are used in this analysis to investigate the improvement of proposed cell performance which will be discussed in the next chapter.

68

4.11 CHAPTER SUMMARY

The specific structure of ultra thin ZnCdS/ZnCdTe solar cell is numerically optimized in this chapter. First of all, the entire window layer (TCO/Zn2SnO4/ZnCdS) is designed and optimized for higher Jsc. The conventional baseline cell was chosen as reference and from there the modified cell structure was proposed for higher performance. In the first step of the numerical calculation, the ZnCdS layer was reduced to the extreme limit for better response in the blue region. The main focus was to reduce the ZnCdTe absorber layer thickness, which is normally more than 4 μm in the typical CdS/CdTe cells. Less than 1 μm ZnCdTe layer showed some good performance in the modified cell and further reduction of ZnCdTe layer outlined from the electric field distribution and carrier generation rate. Different suitable back contact material has been tried with success and finally BSF insertion was proposed from the BSRR of the ultra thin cells. The specific BSF materials selected to investigate are Zinc

Telluride (ZnTe), Lead Telluride (PbTe), Cupric Telluride (Cu2Te) and Arsenic Telluride

(As2Te3), the implementation of these material for BSF effect will be discussed in the next chapter. 69

CHAPTER V

RESULTS & DISCUSSION

5.1 MODELING & SIMULATION

The purpose of numerical modelling and simulation in photovoltaic cell analysis is to check the validity of proposed physical structure maintaining cell geometry and cell performance. The AMPS-1D program has been developed to visualize the details of the physical operation of solar cell. In this work modelling and simulation were done utilizing

AMPS-1D simulator to explore the possibilities of ultra thin ZnCdTe absorber and ZnxCd1-xS window layer with improved cell output like open circuit voltage (Voc), the short circuit current density (Jsc), fill factor (FF) and ultimately the conversion efficiency (η). In AMPS- 1D, the optical model was set to 85% reflection without BSF and 95% reflection with BSF at the back contact and 5% reflection for the front contact, which mean that 95% of the incoming light will be transmitted to the absorber layer. The temperature was set to 298K (25°C). The base line case of CdTe cell was utilized to approximate the highest efficiency of CdTe solar cell at that time, and it was modified in this work to analyze the possibility of efficient ultra thin cells with proper BSF. The first modification was to decrease ZnxCd1-xS window layer to 80 nm with Zn2SnO4 buffer layer. The ZnxCd1-xS layer was included as an alternative of CdS layer to improve the absorption of photon energy in the short wavelength region (blue region). The front contact of the modified cells consist of a highly conducting layer of SnO2 as transparent conducting oxide (TCO) for low resistance due to contact and lateral current collection and a much thinner high resistivity buffer layer of Zn2SnO4 to prevent small amount of forward leakage current through ZnxCd1-xS layer which is significantly less than the leakage current in CdS layer in comparison to CdS/CdTe solar cell

[108]. This CTO/Zn2SnO4 has replaced the SnO2 as front contact layer of the conventional cell. The next modification was to reduce the ZnCdTe absorber thickness to the extreme limit for achieving ultra thin ZnxCd1-xS/ZnCdTe cell and inserting different BSF layer to reduce the barrier height and the minority carrier recombination loss at the back contact of the ultra 70

thin ZnCdTe cell. Figure 5.1 shows the CdTe baseline case structure and (Glass/ SnO2/

Zn2SnO4/ ZnxCd1-xS/ ZnCdTe/ BSF/ MBC) that is the modified structure in which CdS and

CdTe from baseline case are replaced by ZnxCd1-xS and ZnCdTe respectively for higher performance.

Figure 5.1: Structure of ZnCdTe solar cell: (a) Conventional baseline case structure and (b) modified structure for higher performance.

Four layers that were highlighted in this analysis are the n-Zn2SnO4 buffer layer, n-

ZnxCd1-xS window layer, p-ZnCdTe absorber layer and p-Cu2Te; ZnTe, As2Te3 as well as PbTe are BSF layers. The ZnCdTe absorber layer thickness was varied from 0.1 µm to 4 µm and the other layers thickness were fixed to the optimum values found in literature [108] and is shown in Table 5.1 and also the material parameters used in this modelling which were chosen based on literature values. Hence, it is crucial to minimize the number of variable parameters by fixing many of them at reasonable values. It was a tough challenge to select the appropriate parameters to be used for the individual layers of the cells. Many of them depend on fabrication processes and deposition techniques and can thus vary even between devices fabricated at the same chamber. The dependability of this analysis, of course relies on the selection of the material parameters that are going to be used in the simulation. Table 5.1 shows the material parameters used in this modelling, which were chosen based on experimental data, literature values, and theoretical study. 71

Table 5.1: Material parameters used in AMPS simulation

General Device Properties Parameter Front Back Фb [eV] Фbn = 0.1 Фbp= 0.3 7 7 Se [cm/s] 10 10 7 7 Sh [cm/s] 10 10 Reflectivity Rf 0.05 0.85 Layer Properties

Parameter n-SnO2 n-Zn2SnO4 n-ZnxCd1-xS p-ZnCdTe p-As2Te3/PbTe/ZnTe/Cu2Te W [nm] 100 100 80 1000 100 ε/ε0 9 9 10 10.2 20/20/14/10 2 µe [cm /Vs] 100 32 100 250 500/6000/70/500 2 µb [cm /Ns] 25 3 25 70 210/4000/50/100 n, p [cm"3] n: l017 n: 1019 n: 2.5×1016 p: 5.0×1014 p:6.8x1019/5x1021/7.5x1019/1020 Eg [eV] 3.6 3.35 2.5 1.53 0.6/.32/2.25/1.18 3 18 18 I8 17 18 17 17 Nc [cm 1 2.2×10 2.2×10 2.2×10 1.5×10 5x10 /7.5x10 /7.8x10 19 19 19 18 19 19 19 Nv [cm ] l.8×10 1.8×10 1.8×10 1.8×10 1.8x10 /1.5x10 /1.6x10 Χ (eV) 4.5 4.5 4.46 4.4 4.0/4.6/3.65/4.2

5.2 OPTIMIZATION OF ZNXCD1-XS WINDOW LAYER

In this work the conventional structure CdS/CdTe has modified and here CdS is replaced by ZnCdS and CdTe is replaced by ZnCdTe. To optimize the composition of x of

window layer (ZnxCd1-xS) in the modified structure the numerical simulation has done using AMPS-1D. Numerical simulation has been done to see the effect of Zn content on conversion efficiency from x=0 to x=1 using the parameters from Table 5.2 which was adopted from [23, 109-113].

Table 5.2: Parameters of ZnxCd1-xS used in simulation to optimize value of ‘x’

Parameter X=0.05 X=0.08 X=0.1 X=0.2 X=0.3 X=0.5 X=0.6 X=0.8 W (µm) 0.1 0.02-0.3 0.1 0.1 0.1 0.1 0.1 0.1 ɛ/ ɛo 10 10 10 10 10 10 10 10 2 µc (cm /Vs) 100 100 95 85 75 70 65 60 2 µp (cm /Vs) 40 40 35 30 25 20 15 10 -3 n, p (cm ) 3.0×1016 2.5×1016 2.5×1016 1.7×1016 1.6×1016 4.1×1015 2.5×1015 1.7×1015 Eg (eV) 2.48 2.50 2.55 2.58 2.64 2.70 3.07 3.33 -3 Nc (cm ) 2.2×1018 2.2×1018 2.2×1018 2.2×1018 2.2×1018 2.2×1018 2.2×1018 2.2×1018 -3 Nv (cm ) 1.8×1019 1.8×1019 1.8×1019 1.8×1019 1.8×1019 1.8×1019 1.8×1019 1.8×1019 Χ (eV) 4.47 4.46 4.44 4.38 4.32 4.26 4.14 4.02

72

Figure 5.2: Effect of Zn content (x) on ZnxCd1-xS/ZnCdTe cell performance using the

parameters of ZnxCd1-xS from Table 5.2.

It is clear from Figure 5.2, for low content of Zn (≤ 10%), the conversion efficiency (Eff), Voc, and FF are higher than that for high content of Zn (> 10%) and Jsc decreased continuously from x=0.08 to x=0.3. It is also clear that the optimum efficiency 20.62% (Voc 2 = 0.99 V, Jsc = 29.40 mA/cm and FF = 0.779) has found for x=0.08. The electrical resistivity 10 of ZnxCd1-xS layer increases from 1 Ω-cm (x=0.0) to 10 Ω-cm (x=1.0) [111] although band gap increases with increasing ‘x’. The band gap of ZnxCd1-xS is not significantly higher than that of CdS for x ≤ 0.1. The high band gap is the key requirement to replace ZnxCd1-xS instead of CdS as it increases transmission in the blue region and consequently quantum efficiency of ZnCdTe solar cells. In consideration to fabrication complexity, band gap and simulation results, x=0.08 was selected for this work. In the following section, ZnxCd1-xS has been replaced to Zn0.08Cd0.92S for the proposed cell. 73

In this part of work, the 80 nm Zn0.08Cd0.92S layer will be optimized for higher cell performance. Figure 5.3 shows the effects in details of Zn0.08Cd0.92S layer thickness variation from 25 nm to 200 nm on the cell output parameters such as Jsc, Voc, and FF and conversion efficiency from AMPS calculation.

Figure 5.3: Effect of Zn0.08Cd0.92S thickness on ZnCdS/ZnCdTe cell parameters.

It has been observed that the thicknesses of Zn0.08Cd0.92S window layer directly affect the cell performance. If the thickness is less than 50 nm, Voc as well as FF remain almost same. It is well known that very thin window layer cause for pin-hole effect and it may be happened due to this effect. To date, it is almost impractical to fabricate window layer thickness below 50 nm for high quality ZnCdTe solar cells. On the other hand, as the thickness is more than 50 nm, the efficiency and Jsc decrease. It happened due to the light absorption by the window layer. As the thickness increases, the light absorption also 74

increases, and as a result less efficiency decreases. Due to fabrication limitation, Zn0.08Cd0.92S layer thickness was selected as 80 nm with conversion efficiency 20.62% (Voc = 0.99 V, Jsc = 29.40 mA/cm2 and FF = 0.779) outcomes from the simulation results as shown in Figure 5.3 above.

Figure 5.4: Effect of Zn0.08Cd0.92S thickness on ZnCdS/ZnCdTe cell quantum efficiency (QE)

The spectral response (SR) of Zn0.08Cd0.92S thickness variation from 25 nm to 200 nm is shown in Figure 5.4. It can be seen from Figure 5.4 that, when the wavelength is below 510 nm the quantum efficiency (QE) is slight affected with the increasing Zn0.08Cd0.92S layer thickness which influences the cell Jsc and finally conversion efficiency.

5.3 OPTIMIZATION OF ZnCdTe ABSORBER LAYER

Theoretically, the minimum thickness required to absorb 90% of the incident photons with energy greater than the band gap is nearly 1 µm which was in literature for CdTe cell. But, it is remarkable that in most high efficiency CdTe solar cells, the CdTe absorber layer is purposely kept at 5 µm and above. The key idea of this analysis is to obtain the acceptable cell output parameters using ZnxCd1-xS (x=0.08) window at reduced ZnCdTe absorber layer thickness. This will reduce the cost of cell deposition and material usage of ZnCdTe cells. Numerical analysis was done to lessen the thickness of ZnCdTe absorber layer to the extreme limit aiming to preserve the absorber ZnCdTe materials use. And it was also found that all the 75 cell output parameters were constant from 4 µm to 10 µm of ZnCdTe layer. Thus, the cell output characteristics below 4 µm ZnCdTe layer were explored and are shown in Fig. 5.5.

35.00

30.00

25.00

20.00

15.00 1µm ZnCdTe Jsc Jsc mA/cm2 10.00

5.00

0.00 0 0.2 0.4 0.6 0.8 1 V (Volt)

Figure 5.5: J-V characteristics curve of the proposed 1µm ZnCdTe solar cell.

Further numerical analysis was done with As2Te3/PbTe/ZnTe/Cu2Te BSF to investigate the effects of BSF in the ultra thin cell. It was found that the proposed cell with 1 µm ZnCdTe without BSF gives conversion efficiency of 18.09% but with BSF 1 μm ZnCdTe gives 20.35% and with 0.6 μm ZnCdTe gives the highest conversion efficiency of around 21.1%. Simulation results are shown in Figure 5.6 with variable thickness of ZnCdTe absorber layer from 0.1 μm to 4 μm with 100 nm different BSF layer and along with the

ZnxCd1-xS (x=0.08). The enhancement in efficiency with BSF resulted from the improvement of all the cell output parameters like Jsc, Voc and FF. 76

Figure 5.6: The effect of ZnCdTe thickness variation on the output parameters of ZnCdS/ZnCdTe solar cell

It is evident from the Figure 5.7 that the long wavelength spectral response (SR) shows a decrease with the decrease of ZnCdTe thickness. The 1 µm and 10 µm CdTe absorber thickness shows the more or less same SR. Thus, there was no change in cell performance with thicker ZnCdTe layers (>1 µm) as found from the same results in the previous calculation. The reduced ZnCdTe thickness has shown lower SR in the long wavelength region, which is the main cause of reduced Jsc and cell performance of the thinner cells. Therefore, insufficient absorption of photon is occurred in the case of thinner

ZnCdTe absorber which reduces the cell Jsc and Voc and finally the conversion efficiency. The importance of thickness reduction of ZnCdTe to 0.6 µm as found in this analysis indicates that the carriers after this thickness have a little effect on the cell current. The 77 decreased SR for ultra thin (<1 µm) ZnCdTe absorber might be attributed to poor absorption and the recombination losses at the back contact or from both.

1

0.9

0.8 0.2µm 0.6µm 0.7 0.8µm 0.6 1µm QuantumEfficiency 4µm 0.5 10µm 0.4 0.4 0.5 0.6 0.7 0.8 0.9 Wavelength (nm)

Figure 5.7: The effect of ZnCdTe thickness on spectral response of proposed ZnCdS/ZnCdTe solar cell.

5.4 ANALYSIS OF ZnCdTe CELL WITH BSF

A back surface field (BSF) material is a highly p-doped semiconductor material which can act as a conduction-band energy barrier at the back surface of the solar cell. Sometimes it is called as an electron reflector because it reflects the electron at the conduction band. An efficient back surface field is an indispensable structural element to attain high efficiency in a solar cell. The importance of BSF layers in the field solar cells attracted some attention in the 1980s [114-116]. The main role of this layer is to provide confinement for the photo- generated minority carriers and keep them within the reach of the p-n junction to be efficiently collected. This has to be accomplished without increasing the series resistance of the device. Additionally, photon confinement capabilities are an interesting ancillary property for properly matched BSF layer. Highly doped BSF materials can reduce the recombination by reducing barrier height and junction resistance at the back surface.

Introducing BSF materials such as CuxTe, ZnTe, As2Te3, PbTe, etc adds additional flexibility for Eg and χ [117], but adds another junction and another barrier height as shown in Figure 5.8. Dipole layers at both the CdTe/BSF and BSF/metal interfaces are likely to be 78 present. A highly doped BSF allows tunneling to the metal at the BSF/Metal junction and reduces Φbc at the CdTe/BSF junction. The required doping density for tunneling contact at the BSF/Metal junction is 1019 cm-3 or higher.

Figure 5.8: CdTe/BSF/Metal contact

One of the major differences of thin cells compared to the thicker ones is that the absorber/back contact interface is located closer to the p-ZnCdTe and n-ZnCdS junction and the choice of the back contact material therefore has a high impact on the cell performance. A stable back contact that is not significantly rectifying is essential for good performance and long term stability of ZnCdS/ZnCdTe cells. The formation of a low resistance, low barrier back contact is one of the most challenging aspects for high performance ZnCdTe solar cells.

The specific BSF materials like Zinc Telluride (ZnTe), Cupric Telluride (Cu2Te),

Arsenic Telluride (As2Te3) and Lead Telluride (PbTe) will be examined in this part of work to restrain the possible recombination losses and reflect back the minority carrier (electron) towards front contact before possible recombination in ultra thin ZnCdTe cell. The ZnTe, a p- type semiconductor with a direct band gap of 2.26 eV and formation of ohmic contacts to ZnTe is easier due to its lower electron affinity (3.65 eV) and the ability to dope it highly p- type [118]. The Cu2Te is a p-type semiconductor with a wide optical band gap of 1.08 eV is used as BSF of ZnCdS/ZnCdTe solar cells. As2Te3 is a p-type semiconductor which has a band gap of about 0.6 eV and exhibits resistivity of 10-3 Ωcm at room temperature. In this work we introduce another new BSF material like PbTe with a narrow band gap 0.32eV. All these BSF materials are investigated in the following subsections.

79

5.4.1 ZnTe Back Surface Field

Another most common method is the insertion of p-ZnTe in between CdTe and Metal which have favorable valence band discontinuity ΔEvb ≈ 0 to 0.1 eV [119-120]. This valance band discontinuity increases series resistance of solar cell which has been discussed in chapter 3. To overcome this problem, ZnTe is used with N or Cu doping and it also increases the carrier density higher than 5.0 ×1018 cm-3 [121], which is enough to allow tunneling to an outer metal contact. Figure 5.9 has shown the band diagram for CdTe/ZnTe/Metal junction. In addition, ZnTe appears to have a lower Φb to the metal than does CdTe.

Figure 5.9: CdTe/ZnTe/Metal junctions

Extensive research on ZnTe doped with Cu has been done by Gessert et al. [122] and achieved efficiency of 12.1% and Tang et al. achieved efficiency of 12.9% [123]. ZnTe:Cu does make a low resistance contact, although the control of the Cu diffusing into the bulk CdTe and CdS is critical and lowering the cell stability. From the Hall Effect measurement, Tang et.al. [123] have found the hole densities in the range of 1.0 ×1019 to 1.0 ×1020 cm3 for films with Cu.

Although the cell with ZnTe:Cu BSF shows a higher efficiency than ZnTe:N BSF at preliminary time but Makhratchev et al. [121-122] have found that the efficiency degradation of cell with ZnTe:N is smaller and final efficiencies are higher.

In order to reduce possible recombination loss and the barrier height at the back contact of the ultra thin ZnCdTe cell a high band gap material ZnTe (Eg=2.25) was inserted 80 at the back contact. This high band gap material requires final metal back contact with higher work function (Фm> 5.9 eV) for ФbL=1.6 eV or higher to form ohomic contact without ZnTe to explore the effect of BSF with the modelled ultra thin ZnCdTe cell. In this part of analysis, all the layers of the proposed cell are similar except one extra BSF layer (100 nm of ZnTe) at the back contact of the cell with Ni as a final back contact as shown in figure 5.10.

Figure 5.10: The proposed ultra thin ZnCdTe cell with ZnTe BSF

It is evident from Table 5.3 that the proposed cell with 1 µm ZnCdTe layer with and without BSF show conversion efficiencies of 20.525% and 18.094%, respectively. And the maximum cell efficiency was found at 0.6µm absorber layer thickness which was 21.172%. The improvement in efficiency with BSF came from the improvement of the cell output parameters like Jsc and Voc due to the BSF effect as can be seen from the J-V curve of the cells in Figure 5.11. Further calculations have been carried out for the

SnO2/Zn2SnO4/ZnCdS/ZnCdTe/ZnTe/Ni configuration to find the effect of ZnCdTe thickness reduction with ZnTe BSF. The variation of the thickness of ZnCdTe absorber layer from 0.1 μm to 4 μm with 100 nm ZnTe BSF layer and together with the proposed cell results are shown in Figure 5.12.

81

Table 5.3: Output parameters of the modified cells without and with ZnTe BSF

ZnCdTe Eff Jsc Voc Cell structure/ output parameters FF thickness (%) (mA/cm2) (V)

1µm Glass/SnO2/Zn2SnO4/ ZnCdS/ZnCdTe/Te 18.094 0.702 30.159 0.93

1µm Glass/SnO2/Zn2SnO4/ ZnCdS/ZnCdTe/ZnTe/Ni 20.525 0.758 30.396 0.98

0.6µm Glass/SnO2/Zn2SnO4/ ZnCdS/ZnCdTe/ZnTe/Ni 21.172 0.781 30.037 0.99

Figure 5.11: J-V characteristics of the proposed ZnCdTe cell with and without ZnTe BSF.

The calculated results of proposed cell without BSF layer are shown in this Figure

5.12 for comparison. The Voc and FF have shown increasing trend as expected with decrease of the ZnCdTe layer thickness in the presence of highly doped ZnTe BSF whereas without

ZnTe BSF the Voc and FF have shown decreasing trend. The rise of Voc has the largest contribution for the efficiency improvement in the case of ultra thin ZnCdTe solar cell.

The Jsc with ZnTe BSF has shown higher value than without BSF layer but follow the previous trend of decrease drastically below 0.5 μm of ZnCdTe layer as without BSF. The increase in Voc, FF and Jsc are due to the minority carriers (electron) that are reflected from the back surface of the SnO2/Zn2SnO4/ ZnCdS/ZnCdTe/ZnTe/Ni cell and the smoother flow of hole at the back contact. These results of BSF layer are agreeable with related published works [20]. Thus, the highest cell conversion efficiency showed the highest value of 21.172% (Voc= 0.99 V, Jsc= 30.037 mA/cm2, FF = 0.781) at 0.6μm of ZnCdTe layer with BSF. 82

Therefore, it is a clear indication of further reduction of ZnCdTe absorber layer thickness from 1 μm to 0.6μm with ZnTe BSF as found in this analysis.

Figure: 5.12: Effect of ZnCdTe thickness on the cell with and without ZnTe BSF.

Before final conclusion on the ZnTe BSF performance, it is important to investigate the stability of the proposed cells at higher operating temperatures. The operating temperature plays a very important role for cell performances. At higher operating temperature parameters such as the electron and hole mobility, carrier concentrations, density of states and band gaps of the materials are affected. In order to investigate the effects of higher operating temperature on cells performances with and without ZnTe BSF, simulation were carried out with cell operating temperature ranged from 25°C to 100°C and the calculated results are shown in Figure 5.13. 83

Figure 5.13: Effect of operating temperature on the proposed ultra thin ZnCdTe solar cell.

From the Figure 5.13, it is evident that without ZnTe BSF layer the cells normalized efficiency linearly decreased with the increase of operating temperature at a temperature coefficient (TC) of -0.26%/°C. This TC indicates stability of the cells at higher operating temperature, which are in good agreement with related works [71]. The cell with ZnTe BSF showed better stability, the cell conversion efficiency decreased with increasing operating temperature range from 25° C to 40°C with a TC of - 0.18%/°C. But in the temperature range from 40°C to 100°C it gives better stability as its efficiency remains almost unchanged. However, the performances of the cells with ZnTe BSF have shown better overall stability throughout the range than without ZnTe BSF layer. Therefore, the stability of cell with ZnTe BSF has favourable effects on the cell stability at higher operating temperature. Thus, this cell can be used where operating temperature are higher.

5.4.2 Cu2Te Back Surface Field

CuxTe(x=2) as a BSF material is widely used in CdTe solar cell. In this technique, at first a thin Te layer is made by etching the CdTe or by Te deposition, is then converted to

CuxTe by inter-diffusion of Cu, either from a Cu paste or from deposition of Cu [124]. Figure

5.14 has shown the band diagram for CdTe/CuxTe/Metal junction. The highest efficiencies for CdTe cells have been achieved with CuxTe layers by Wu [125] where, Jsc = 25.9 2 mA/cm , VOC = 0.845 V, FF = 0.755, Eff = 16.5%) and Britt and Ferekides [49]. The high p doped CuxTe allows tunneling at the CuxTe/Metal junction, so many different metal layers 84 can be used such as Al, Ni, Cr, ITO, graphite, etc [126]. The band gap and doping concentration are changed with the change of ratio between Cu and Te. As the barrier height between Cu2Te and metal depends on carrier concentration, it should maintain a certain value to get the optimum back contact. Farag found a direct band gap = 1.18 eV for CuxTe with x ≈ 1.98, and p ≈ 1.0×1021 cm-3 due to Cu vacancies [127]. Späth et al. have found Eg = 1.04 eV and Φbc = 0.7 to 0.8 eV (depending on p) by using UPS measurements during growth of

CuxTe contacts for 1.5 < x < 1.9[120]. The Φbc found in this case is much larger than indicated by cell J-V measurements. They concluded that contact transport was by tunneling through the CdTe/CuxTe barrier, and involved Cu states in the CdTe.

Figure 5.14: CdTe/CuxTe/Metal junctions Source: Ref. [119] However, Cu-containing contacts are potentially problematic, as Cu is a fast diffuser in CdTe, which degrades the PV performance of the cell [60 and128]. This degradation is thought to occur via: a) segregation at grain boundaries in CdTe, which creates shunting paths, and b) diffusion to the CdTe/CdS interface, therefore increasing the resistivity of CdS by creating deep levels capturing electrons. However, in a recent work Romeo et al. (2009) [129] reported that if a very thin layer of Cu (~ 2 nm) is evaporated on an etched CdTe back surface, devices do not suffer from degradation.

To see the effect in ultra thin ZnCdTe/Zn0.08Cd0.92S solar cell, in this study, wide band gap material Cu2Te (Eg = 1.08 eV) was inserted at the back contact as BSF layer. A numerical analysis was done with Cu2Te BSF layer to explore the effect of BSF in the proposed ultra thin solar cell ZnCdTe. In this analysis, all the layers of the proposed cell are similar except one extra layer (100 nm of Cu2Te) at the back contact of the cell as shown in Figure 5.15. 85

Figure 5.15: Proposed ultra thin ZnCdTe cell with Cu2Te BSF layer.

It is evident from Table 5.4 that the proposed cell with 1 µm ZnCdTe layer with and without BSF show conversion efficiencies of 20.388% and 18.094%, respectively. And the maximum cell efficiency was found at 0.5µm absorber thickness which was 20.877%. The improvement in efficiency with BSF came from the improvement of the cell output parameters like Jsc and Voc due to the BSF effect as can be seen from the J-V curve of the cells in Figure 5.16. Further calculations have been carried out for the SnO2/Zn2SnO4/

ZnCdS/ZnCdTe/Cu2Te/Mo configuration to find the effect of ZnCdTe thickness reduction with Cu2Te BSF. The variation of the thickness of ZnCdTe absorber layer from 0.1 μm to 4

μm with 100 nm Cu2Te BSF layer and together with the proposed cell results are shown in Figure 5.17.

Table: 5.4: Output parameters of the modified cells without and with Cu2Te BSF

ZnCdTe Eff Jsc Voc Cell structure/ output parameters FF thickness (%) (mA/cm2) (V)

1µm Glass/SnO2/Zn2SnO4/ ZnCdS/ZnCdTe/Te 18.094 0.702 30.159 0.93

1µm Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/Cu2Te/Mo 20.388 0.756 30.332 0.97

0.5µm Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/Cu2Te/Mo 20.877 0.782 29.613 0.98 86

The calculated results of proposed cell without BSF layer are shown in this Figure

5.17 for comparison. The Voc and FF have shown increasing trend as expected with decrease of the ZnCdTe layer thickness in the presence of highly doped Cu2Te BSF whereas without

Cu2Te BSF the Voc and FF have shown decreasing trend. The rise of Voc has the largest contribution for the efficiency improvement in the case of ultra thin ZnCdTe absorber.

Figure 5.16: J-V characteristics of the ZnCdTe cell with and without Cu2Te BSF.

The Jsc with Cu2Te BSF has shown higher value than without BSF layer but follow the previous trend of decrease drastically below 0.5 μm of ZnCdTe layer as without BSF. The increase in Voc, FF and Jsc are due to the minority carriers (electron) that are reflected from the back surface of the SnO2/Zn2SnO4/ ZnCdS/ZnCdTe/ZnTe/Ni cell and the smoother flow of hole at the back contact. These results of BSF layer are agreeable with related published works [20]. Thus, the highest cell conversion efficiency showed the highest value of 20.877% (Voc = 0.98 V, Jsc = 29.612 mA/cm2, FF = 0.782) at 0.5μm of ZnCdTe layer with BSF. Therefore, it is a clear indication of further reduction of ZnCdTe absorber layer thickness from 1 μm to 0.5 μm with Cu2Te BSF as found in this analysis.

87

Figure 5.17: Effect of ZnCdTe thickness on the proposed ultra thin solar cell with and

without Cu2Te BSF.

Finally, the operating temperature plays a very important role on cell performances. At higher operating temperature parameters such as the electron and hole mobility, carrier concentrations, density of states and band gaps of the materials are affected. In order to investigate the effects of higher operating temperature on cells performances with and without Cu2Te BSF, simulation were carried out with cell operating temperature ranged from 25°C to 100°C and the calculated results are shown in Figure 5.18.

88

Figure 5.18: Effect of operating temperature on the proposed cell.

From the Figure 5.18, it is evident that with Cu2Te BSF layer the cells normalized efficiency linearly decreased with the increase of operating temperature from 25°C to 40°C at a temperature coefficient (TC) of -0.15%/°C. After then it’s remain almost constant. This TC indicates stability of the cells at higher operating temperature, which are in good agreement with related works [71]. The cell with Cu2Te BSF showed better stability, the cell conversion efficiency decreased with increasing operating temperature with an average TC of

– 0.04%/°C. However, the performances of the cells with Cu2Te BSF have shown better overall stability throughout the range than without Cu2Te BSF layer. Thus, this cell can be used where operating temperature are higher.

5.4.3 As2Te3 Back Surface Field

Romeo et al. (2009) [129] first reported the use of As2Te3 as a BSF layer on to the as grown CdTe Layer. The back contact barrier height φb for the As2Te3 contact was 0.42 eV for non-etched CdTe. This process is unlike to others. They reported a stable and low resistance back contact by sputtering onto as-grown CdTe (i.e. no etching was applied prior metallization) in a sequence As2Te3, Cu (≤ 20nm) and Mo, respectively which is shown in

Figure 5.19. In order to form a CuxTe (x < 1.4) phase onto CdTe by replacing As by Cu, Cu was deposited at a temperature 200°C and they reported the cell efficiency 15.8 % (Romeo et al. 2009) [129]. 89

Figure 5.19: CdTe/As2Te3/Metal junctions

Source: Ref. [118]

A numerical analysis was done with As2Te3 BSF layer to explore the effect of BSF in the proposed ultra thin solar cell ZnCdTe. In this analysis, all the layers of the proposed cell are similar except one extra layer (100 nm of As2Te3) at the back contact of the cell as shown in figure 5.20.

To reduce the possible recombination loss and barrier height at the back contact of the ultra thin ZnCdTe solar cell a narrow band gap material As2Te3 (Eg = 0.6 eV) was inserted at the back contact as BSF layer. This low band gap material would produce almost flat barrier height with Ag as final back contact materials and act as a BSF to bounce back the carrier

(electrons) from the ZnCdTe/ As2Te3 junction and thus would contribute in the enhancement of carrier. 90

Figure 5.20: The proposed ultrathin solar cell with As2Te3 BSF.

The modified solar cell along with all the parameters of the proposed cell with As2Te3 BSF layer was simulated and the results are caparisoned in Table 5.5.

Table 5.5: Output parameters of the modified cells without and with As2Te3 BSF

ZnCdTe Eff Jsc Voc Cell structure/ output parameters FF thickness (%) (mA/cm2) (V)

1µm Glass/SnO2/Zn2SnO4/ ZnCdS/ZnCdTe/Te 18.094 0.702 30.159 0.93

1µm Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/As2Te3/Ag 20.357 0.754 30.333 0.97

0.5µm Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/As2Te3/Ag 21.019 0.784 29.532 0.99

It is evident from Table 5.5 that the proposed solar cell with 1 µm ZnCdTe layer with and without BSF shows conversion efficiency of 20.357% and 18.094%, respectively. The improvement in efficiency with BSF came from the improvement of all the cell output parameters like Jsc, Voc and FF which will be much clearer from the J-V characteristics of the cells. The simulated J-V characteristics of the cells are shown in Figure 5.21. 91

Figure 5.21: J-V characteristics of the ZnCdTe cell with and without As2Te3 BSF.

From Figure 5.22, the structure with As2Te3 BSF showed higher Voc, Jsc and FF than the cell without BSF due to reduced back surface recombination and improved back contact formation with p-ZnCdTe. Thus, the conversion efficiency of the cell with As2Te3 BSF is higher than without BSF layer.

Calculations have been carried out for the ZnCdS/ZnCdTe/As2Te3 configuration to find the effect of ZnCdTe thickness reduction with BSF. The variation of the thickness of

ZnCdTe absorber layer from 0.1 μm to 4 μm with 100 nm As2Te3 BSF layer and together with the ZnCdS/ZnCdTe simulation results are shown in Figure 5.22.

The calculated results of cell without BSF layer are also shown in this figure for comparison. The Voc and FF show increasing trend as expected with decrease of the ZnCdTe layer thickness in the presence of highly doped As2Te3 BSF, where without BSF both Voc and FF show decreasing trend.

The increase of Voc has the largest contribution to the efficiency improvement in the case of ultra thin ZnCdTe absorber. The Jsc with As2Te3 BSF has shown higher value than without BSF layer but follows the same trend of decrease drastically below 0.5μm of ZnCdTe layer as without BSF. The increase in Voc, FF and Jsc are due to the minority carriers

(electron) that are reflected from the back surface of the ZnCdS/ZnCdTe/As2Te3 cell and the smother flow of hole at the back contact. These results of BSF layer are agreeable to the related published works [130]. Thus, the cell conversion efficiency showed highest value of 92

21.019% (Voc= 0.99 V, Jsc= 29.532 mA/cm2, FF = 0.784) at 0.5μm of ZnCdTe layer with BSF. Therefore, it is a clear indication of further reduction of ZnCdTe absorber layer thickness from 1 μm to 0.5μm with As2Te3 BSF as proved in this analysis.

Figure 5.22: Effect of ZnCdTe thickness for the cell with and without As2Te3 BSF.

Before as for the As2Te3 performance, it is also important to investigate the stability of the proposed cells at higher operating temperatures. In order to investigate the effects of higher operating temperature on cells performances with and without BSF simulation were carried out with cell operating temperature ranged from 25°C to 100°C and the calculated results are shown in Figure 5.23. 93

Figure 5.23: Effect of operating temperature on the proposed ultra thin ZnCdTe cell.

From the figure 5.23, it is evident that without BSF layer and with As2Te3 BSF layer the cells normalized efficiency linearly decreased with the increase of operating temperature at a temperature coefficient (TC) of -0.25%/°C. This TC indicates the stability of the cells at higher operating temperature, which are in good agreement with related work [130]. The stability of cell with As2Te3 BSF showed the almost same stability as the cell without BSF.

Thus, the As2Te3 BSF has no unfavourable effects on the cell stability at higher operating temperature.

5.4.4 PbTe Back Surface Field

Another low band gap materials PbTe (Eg=0.32) has been inserted as a back contact of the proposed cell structure to reduce the barrier height and possible recombination loss at the back contact of the ultra thin ZnCdTe solar cell. This low band gap material would produce almost flat barrier height with Al as a final back contact material and expected to act as a BSF to bounce back the minority carriers from the ZnCdTe/PbTe junction and thus would contribute in the enhancement of carrier. A numerical analysis has done with PbTe to explore the effect of BSF with the ultra thin ZnCdTe solar cell. In this analysis, all the layers of the proposed cell are similar except one extra BSF layer (100 nm of PbTe) at the back contact of the cell as shown in Figure 5.24. 94

Figure 5.24: Proposed ultra thin ZnCdTe solar cell with PbTe BSF.

Then modified cell along with all the parameters of the proposed cell with PbTe BSF layer were simulated and the calculated result are compared in Table 5.6.

Table 5.6: Performance parameter comparison of proposed ZnCdTe cell without and with PbTe BSF layer

ZnCdTe Eff Jsc Voc Cell structure/ output parameters FF thickness (%) (mA/cm2) (V)

1µm Glass/SnO2/Zn2SnO4/ ZnCdS/ZnCdTe/Te 18.094 0.702 30.159 0.93

1µm Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/PbTe/Ag 20.345 0.757 30.187 0.97

0.6µm Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/PbTe/Ag 20.626 0.779 29.397 0.99

It is evident from the Table 5.6 that the proposed cell ZnCdS/ZnCdTe has higher conversion efficiency (18.094%) than Cds/CdTe cell (17.884%). After insertion of PbTe BSF layer it is clear that the improvement occurred in the cell efficiency. The improvement in efficiency with BSF came from the improvement of Voc and FF but Jsc shows little decrease 95 value due to the thinner absorber layer as can be seen from the J-V curve of the cell from Figure 5.25.

Figure 5.25: J-V characteristics of the proposed cell with and without PbTe BSF.

Further simulations have been carried out for the ZnCdS/ZnCdTe/PbTe/Al configuration to find the effect of ZnCdTe thickness reduction with BSF. The variation of the thickness of ZnCdTe absorber layer from 0.1 μm to 4 μm with 100 nm PbTe BSF layer and together with the ZnCdS/ZnCdTe simulation results are shown in Figure 5.26. The calculated results of cell without BSF layer are also shown for comparison.

The Voc and FF show increasing trend as expected with the decrease of the ZnCdTe layer thickness in the presence of highly doped PbTe BSF, but without BSF Voc and FF show decreasing trend. The increase of Voc has the largest contribution for the efficiency improvement in the case of ultra thin ZnCdTe absorber. The Jsc with PbTe BSF has shown higher value than without BSF layer throughout but follow the same trend of decrease drastically below 0.6μm of ZnCdTe layer with BSF. The increase in Voc, FF and Jsc are due to the minority carriers (electron) that are reflected from the back surface of the ZnCdS/ZnCdTe/PbTe/Al cell and the smoother flow of hole at the back contact. Thus, the cell conversion efficiency has the highest value of 20.626% (Voc= 0.99 V, Jsc= 29.397 mA/cm2, FF = 0.779) at 0.6μm of ZnCdTe layer with PbTe. 96

Figure 5.26: Effect of ZnCdTe thickness for the cell with and without PbTe BSF.

Again as for the PbTe performance, there is need to investigate the stability of the proposed cells at higher operating temperatures. In order to investigate the effects of higher operating temperature on cells performances with and without this BSF, AMPS simulations were carried out with cell operating temperature ranged from 25°C to 100°C and the simulated results are shown in Figure 5.27.

From Figure 5.27 it is clear that without BSF layer and with PbTe BSF layer, the cell’s normalized efficiency linearly decreased with the increase of operating temperature at a temperature coefficient (TC) of -0.26%/°C. The stability of cell with PbTe BSF showed the almost same stability as the cell without BSF. Thus, the PbTe BSF has no unfavourable effects on the cell stability at higher operating temperature.

97

Figure 5.27: Effect of temperature on the proposed cell.

5.5 COMPARISON AMONG PROPOSED CELLS

In this part of work, the proposed three BSF material for ultra thin CdTe solar cell are compared from the cells conversion efficiency, CdTe absorber layer thickness and stability at higher operating temperature. A comparison has been done with the best ultra thin cell

(Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/Te) with thinner ZnCdTe layer aiming to explore the performances with four different BSF materials. The best cell output parameters for each structure with ultra thin ZnCdTe absorber layer are summarized in Table 5.7.

It is evident from Table 5.7 that the designed best cell without any BSF layer shows conversion efficiency of 18.094%, which higher than CdS/CdTe cells. When the wide band gap material like ZnTe is inserted as BSF at the back contact, the conversion efficiency increased up to 21.172% with only 0.6 μm ZnCdTe absorber layer as Voc and FF increased due to expected BSF effect on the ultra thin cell. At the same way when Cu2Te is inserted as BSF at the back contact, the conversion efficiency increased up to 20.877% with only 0.5 μm ZnCdTe absorber layer. As well as Voc and FF also increased due to expected BSF effect on the ultra thin cell.

When the lower band gap material like PbTe was inserted as BSF the conversion efficiency increased up to 20.626 % with 0.6 μm ZnCdTe absorber layer as Voc and FF increased due to BSF effect but Jsc showed lower value, it might be attributed to the poor back contact with 98

this material. Moreover, another low band gap material, As2Te3, as BSF produced cell performances of 21.019% with 0.6 μm ZnCdTe absorber. The improvement in efficiency of this cell came from the improvement of all the cell output parameters like FF and Voc.

Table 5.7: The best output parameters of different cells with and without BSF

Cell structure/ output parameters Eff Jsc FF Voc (V) (%) (mA/cm2)

Glass/ZnO/CdS/CdTe/Cu 17.884 0.817 26.579 0.90

Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/Te 18.094 0.702 30.159 0.93

Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/PbTe/Al 20.626 0.779 29.945 0.98

Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/As2Te3/Ag 21.019 0.784 29.893 0.99

Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/Cu2Te/Mo 20.877 0.782 29.945 0.98

Glass/ZnO/Zn2SnO4/ ZnCdS/ZnCdTe/ZnTe/Ni 21.172 0.781 30.037 0.99

In general, the structure with BSF showed higher Voc, FF and Jsc than the cell without BSF. All the cell with BSF showed higher Voc and FF but poor Jsc due to supper thinning (≤.6µm) of ZnCdTe material and from poor back contact formation. The As2Te3/Ag has shown moderate cell output parameters. In this analysis as well as in general, the ZnTe/Ni layer has shown higher performance than As2Te3, PbTe and Cu2Te BSF with 1μm ZnCdTe layer. But in particular with ZnTe BSF the cell with 0.6 μm ZnCdTe layer showed highest conversion efficiency of 21.172%. The J-V characteristics of the cells with BSF and without BSF, all with 0.6 μm ZnCdTe, 80 nm ZnCdS and same front contact are shown in Figure 5.28 to compare their performance.

99

Figure 5.28: J-V characteristics of the proposed cell with BSF

Before drawing the final conclusion on BSF choice there is a need to compare the stability of the cells at higher operating temperature with BSF. As it is well known that operating temperature plays a very important role on the cell output performance in practical fields where the operating temperature is higher than 25°C. At higher operating temperature, cell properties are affected as also found in this analysis. A comparison was done at operating temperature ranged from 25°C to 100°C for all the cells with 0.6 μm ZnCdTe, 80 nm ZnCdS layer and with same front contact (ITO/ZnO). The results obtained from AMPS calculation are shown in Figure 5.29.

Figure 5.29: Effect of operating temperature on the proposed cells 100

From the figure 5.29 it is clear that cell conversion efficiency without BSF layer linearly decreased with the increase of operating temperature with a temperature coefficient (TC) of -0.26%/°C, which is in good agreement with related works [71]. The cell with ZnTe BSF showed better stability with TC of - 0.18%/°C in the operating temperature range from 25°C to 40°C. After 40°C temperature the cell with ZnTe/Ni gives better stability as its efficiency remains almost unchanged in the operating temperature up to 100°C with a TC

0.04%/°C. Cells with Cu2Te/Mo gives better stability as its efficiency remains almost unchanged in the operating temperature range from 40°C to 100°C. But in the operating temperature range from 25°C to 40°C its efficiency decreases with a TC of -0.15%/°C. The stability of cell with As2Te3/Al is almost similar to that without BSF and which is -0.25/°C.

Another cell with PbTe/Ag shows similar temperature coefficient of cell with As2Te3/Al.

However, the performances of the cells with Cu2Te BSF have shown better stability than all other cells. There are good indications of Cu2Te and ZnTe BSF layer for higher ultra thin cell stability.

5.6 COMPARISON BETWEEN RECENT PUBLISHED WORK AND PROPOSED WORK

From the comparison Table 5.8, it is seen that the proposed work is much better than the recent published works in comparison open circuit voltage (Voc), shot circuit current (Jsc), conversion efficiency (Eff%) and stability (TC%) of the cell structure. The superior performance of the proposed cell structure Glass/ITO/Zn2SnO4/ZnCdS/ZnCdTe/BSF/Back contact is due to high Jsc and Voc which collectively contributes to the higher conversion efficiency. The low value of FF might be due defect states in any of the layer in the full device. Nevertheless, the 21.17% efficiency of Zn0.2Cd0.8S/CdTe solar cells outcomes in this work is much higher than 20.78% of CdS/CdTe cell structure [131], 18.6% of CdS/CdTe cell structure [108] and 20.4% of CdS/CdZnTe cell structure [15]. In comparison to stability with respect to temperature coefficient (TC%) the proposed solar cell structure with As2Te3, ZnTe,

PbTe and Cu2Te BSF show better stability than the solar cell structure in recent published work shown in the table below. This design approach then represents the better approach if the proposed cell structure can be fabricated successfully.

101

Table 5.8: Comparison between recent published work and the proposed work

This work and BSF 2 Voc (V) Jsc (mA/cm ) FF Eff (%) TC (%) published work Material Matin et al ZnTe 0.960 25.63 74.4 18.37 -0.3/°C [131], 2013 Matin et al As2Te3 0.990 24.73 84.5 20.78 -0.4/°C [131], 2013

Matin et al. Cu2Te 0.990 24.73 79.8 19.54 -0.35/°C [131], 2013

Nowshad et al. ZnTe 0.90 24.92 0.70 15.8 -0.3/°C [108], 2010

Nowshad et al. As2Te3 0.92 24.97 0.81 18.6 -0.4/°C [108], 2010

Aliyu et al. - 0.96 24.90 0.86 20.4 - [15], 2010

Proposed work As2Te3 0.99 29.893 0.776 21.019 -0.25/°C -0.15/°C, (25°-40°C) Proposed work Cu2Te 0.98 29.945 0.773 20.877 -0. 01/°C, (40°-100°C) -0.18/°C, (25°-40°C) Proposed work ZnTe 0.99 30.037 0.781 21.172 -0. 005/°C, (40°-100°C)

Proposed work PbTe 0.98 29.945 0.773 20.626 -0.26/°C

102

CHAPTER VI

CONCLUSIONS

6.1 GENERAL SUMMARY

The global population exploration coupled with the technological advancement as a result the energy consumption has increased abruptly. PV cells are the best way to supply the energy demand for all anywhere. Among thin film PV cells the CdTe, is the favorable choice because it has many favrable advantages as a material for solar cells. The cost-driven trend of CdTe solar cells is toward thinner CdTe layers. The targeted areas of CdTe solar cells for improving the performance with reduced cost have investigated. These are better window layer materials, ZnCdS layer with suitable

Zn2SnO4 buffer layer, incorporating Zn in CdTe absorber layer for higher short circuit current, thinning the ZnCdTe absorber layer with higher film quality, improvement of doping. Insertion of suitable BSF for the higher performance of ZnCdS/ZnCdTe solar cells will overcome the problems associated with ultra thin ZnCdTe absorber and suitable back contact formation. In order to achieve these goals innovative design and detailed analysis were explored. The main focus is to develop high efficiency ZnCdS/ZnCdTe solar cell using appropriate BSF with least material input for low cost fabrication.

In order to achieve these goals a modified cell structure was proposed and modeled utilizing the widely used AMPS software. The baseline case of CdTe solar cells was the strating point and it was modified to achieve higher performance with BSF. The specific structure of an ultra thin ITO/CdS/CdTe solar cell is numerically optimized in this work. In the first step of the numerical calculation the ZnCdS layer was reduced for better response in the blue region absorption benefits. Another 103

objective was to reduce the ZnCdTe absorber layer thickness, which is normally more than 5 μm in the typical CdS/CdTe cells. The higher efficiency of CdS/CdTe solar cells have been reported by Britt and Ferekides (1993) [49] and Wu et al. (2001) [11] using much thicker CdTe layers. At ~3 μm of CdTe absorber, the availability of elemental Te may be a concern if production levels increase above 20GW per year (Zweibel 2011) [132]. The development of CdS/CdTe cells with thinner CdTe layer could help the lagging Te supplies and might have additional advantages in fabrication. The 0.6μm ZnCdTe layer showed some good performance in the modified cell. The insertion of BSF was proposed from the effect of BSRR on the ultra thin cells. The specific BSF materials were investigated Zinc Telluride (ZnTe), Cupric

Telluride (Cu2Te), Arsenic Telluride (As2Te3) and Lead Telluride (PbTe). This BSF was numerically applied to the ZnCdTe cell and found that there is no positive effect in the thicker cells but showed encouraging results with thinner ZnCdTe absorber layer. With a 100 nm BSF layer the highest efficiency was found with submicron ZnCdTe thickness. The highest conversion efficiency of 21.17% was achieved in the case of ZnTe BSF with 0.6μm ZnCdTe layer. For the other cases, i.e.As2Te3 BSF with

0.5μm ZnCdTe layer, Cu2Te BSF with 0.5μm ZnCdTe layer and PbTe BSF with 0.6 μm ZnCdTe layer, 21.019%, 20.877% and 20.626% conversion efficiency were achieved respectively. The simulation result also showed that at higher operating temperatures the cells stability is acceptable. Structures with BSF layer showed better stability in general, and in particular the structure (ITO/ Zn2SnO4/ ZnCdS/ ZnCdTe/

ZnTe/ Ni) achieved the best conversion efficiency of 21.17%. The cell with Cu2Te and ZnTe BSF showed better stability, the cell conversion efficiency decreased with increasing operating temperature with a TC of - 0.15%/°C. However, the performances of the cells with Cu2Te and ZnTe BSF have shown better overall stability throughout the range than without BSF layer.

6.2 FUTURE SCOPES

The main and demanding objective of this work was to modify window and absorber layer in CdS/CdTe solar cells. In CdTe based solar cell over 90% semiconductor is its absorber CdTe. Thus, if it can be reduced, it will lead to cheaper and more affordable solar cells. Thus, reduction of the CdTe absorber layers thickness by incorporating Zn 104

to the extreme limit should to be investigated. Several ideas such as increased doping concentration, minority carrier life time, ohomic back contact and reduced recombination losses with BSF are proposed to overcome the possible drawbacks and thus to bring out the underlying performance of ultra thin CdTe cells. The insertion of BSF layer is yet to be carried out with their total potential, especially for the submicron cells. Several ideas such as increased doping concentration, minority carrier life time, BSF insertion more elaborately will be the future perspective for this work.

The proposed window layer (ITO/Zn2SnO4/ZnCdS) of the ultra thin ZnCdTe cells need to be investigated more to enhance the short circuit current (Jsc). If the transmittance of the window layer could be increased the Jsc and overall cell performance would be increased further. The ZnCdS layer could be converted to oxygenated ZnCdS layer (ZnCdS:O) for few more additional benefits such as reduced inter diffusion, better lattice match while reduced ZnCdS layer would increase the SR in the blue region. If higher optical reflector material could be inserted at the back contact of the submicron cells, which would reflect back the unabsorbed photon due to shorter diffusion length, these unabsorbed photons would get second chance to be absorbed in the submicron cells. With 100% optical reflector the effect of submicron absorber would be doubled (i.e., 0.3μm will act as 0.6μm). Thus, in the insertion of optical reflector material such as Au/Pt at the back of the submicron cells might enhance the cell performance further.

Some effective ideas are given for the extension of the works to bring out the under lying potentials of ZnCdS/ZnCdTe solar cells as proposed in this study. If the proposed ideas are fulfilled with their utmost potential, the best performance would be achieved with submicron ZnCdTe absorber layer.

6.3 GENERAL CONCLUSIONS

To be cost-competitive with conventional electricity generation, the use of poly- crystalline CdTe thin films solar cells has great potential. Due to the short optical absorption length in CdTe, the CdTe thickness of 0.5 µm is sufficient by 105

incorporating Zn to absorb more than 90% of the radiation with energy greater than its bandgap (1.53 eV). Until now, the widely used CSS deposition technology requires the thickness of CdTe films typically 4-10 µm to minimize the probability of pinholes and shunting. This work, focused on the thickness reduction of absorber CdTe to the extreme limit with expected film quality improvement by insertion Zn in CdTe and the proposed solar cells can be fabricated with RF sputtering deposition.

From the numerical analysis of the ultra thin cells, the back contact formation for the designed ultra thin cells shows higher sensitivity with the typical back contact metals to p-ZnCdTe. Moreover, the overall performance of the ZnCdS/ZnCdTe cell was found to be affected by the back surface recombination for thinner (~0.6 µm) ZnCdTe absorber. To avoid the effect of such recombination at the back contact of ultra thin cells a modified cell configuration with BSF is proposed. The highest conversion efficiency of 21.17% was achieved in the case of ZnTe BSF with 0.6μm ZnCdTe absorber layer. Structures with BSF layer showed better stability in general, and in particular the structure (Glass/ ITO/ Zn2SnO4/ ZnCdS/ ZnCdTe/ ZnTe/ Ni) achieved the best conversion efficiency of 21.17% and proved to be comparable to other reported cells.

The future works are directed toward to the better window layer (ZnCdS: O) layer, improvement of ZnCdTe carrier density and minority carrier life time. New BSF materials could be searched for the latent performance of the submicron cells. The optimization of each layer and complete cell fabrication process with suitable fabrication technology should be given more attention. Finally, the intended aim of this work related to the development of a high efficiency ultra thin ZnCdS/ZnCdTe solar cell with BSF, has been successfully achieved through numerical simulation. The precise fabrication of these proposed cells requires huge resources and the optimization of the fabrication process steps. Based on these work, it is anticipated that good quality ultra thin ZnCdS/ZnCdTe solar cells with BSF will be presented in near future. 106

REFERENCES

[1] Hansen, J., Nazarenko, L., Ruedy, R., Sato, M., Willis, J., Del Genio, A., Kock, D. Lacis, A., Lo, K., Menon, S., Novakov, T., Perlwitz, J., Russell, G., Schmidt, G. and Tausnev, N. Earth’s energy imbalance: confirmation and implications, Science, 308: 1431–1435. (2005) [2] Karl, T. R. and Trenberth, K. E. 2003. Modern global climate change. Science, 302:1719– 1723. Intergovernmental Panel on Climate Change. 2007. In Climate Change 2007: The Physical Science Basis, Summary for policymakers. [3] Fifth Assessment Report, Carbon and Other Biogeochemical Cycles, Climate Change IPCC: Chapter 6, pp. 40-510.: http://www.ipcc.ch/report/ar5/wg1. (2013) [4] Lewis, N. S. Scientific challenges in the development of sustainable energy. Presented at the 31st IEEE PVSC, Walt Disney World, FL. (2005) [5] Zwiebel, K. The terawatt challenge for thin-film PV. NREL/TP-520-38350. (2005) [6] Steven S. Hegedus and Antonio Luque. Status, Trends, Challenges and The Bright Future of Solar Electricity from Photovoltaics, Handbook of Photovoltaic Science and Engineering, John Wiley & Sons. (2003) [7] Chapin, D. M., Fuller, C. S. & Pearson, G. L. A new silicon p-n junction photocell for converting solar radiation into electrical power. J. Appl. Phys. 25(5): pp. 676. (1954) [8] Lique, A. & Hegedus, S. Handbook of Photovoltaic Energy Conversion and Engineering: John Wiley & Sons LTD, Chichester, West Sussex, England. (2003) [9] Compaan, A., Tabory, Ch., Li, Y., Feng, Z. & Fisher, A. CdS/CdTe solar cells by RF sputtering and by laser physical vapour deposition. Proc. 23nd IEEE Photovoltaic Specialists Conference (Louisville, KY), IEEE, New York. pp. 394–399. (1993) [10] Batzner,D. L., Romeo, A., Zogg, H., Wendt, R. & Tiwari, A. N. Development of efficient and stable back contacts on CdTe/CdS solar cells. Thin Solid Films 387: pp. 151-154. (2001) [11] Wu, X., Keane, J. C., Dhere, R. G., DeHart, C., Albin, D. S., Duda, A., Gessert, T. A., Asher, S., Levi, D.H. & Sheldon, P. Proceedings of 17th European Photovoltaic 107

Solar Energy Conference, Munich Germany. pp. 995-1000. (2001) [12] Green, M. A., Emery, K. D., King, Hishikawa, L. Y. & Warta, W. Prog. Photovoltaics 15, pp. 35-41. (2007) [13] Contreras, M. A., Ramanathan, K., AbuShama, J., Hasoom, F., Young, D. L., Egaas, B. & Noufi, R. Prog. Photovoltaics 13: pp. 209-213. (2005) [14] Noufi, R. & Zwiebel, K. Proc. 4th WCPEC. Pp. 317. (2006) [15] M. M. Aliyu, M. A. Martin, N. Amin and M. Y. Sulaiman, Prospects of Ternary CdZnTe in a Graded Bandgap Thin film Solar Cells, World Renewable Energy Congress XI 25-30 September, Abu Dhabi, UAE. (2010) [16] L.A.Kosyachenko, X.Mathew, V.Ya.Roshko, E.V.Grushko, “Optical absorptivity and recombination losses: The limitations imposed by the thickness of absorber layer in CdS/CdTe solar cells”, Solar Energy Materials and Solar Cells, Vol: 114, pp. 179- 185, (2013). [17] Ray, B. II-VI Compounds. Oxford: Pergamon Press. (1969). [18] Chu, T.L., Shirley, S., Chu, Britt, J., Chen, G., Ferekides, C., Schultz, N., Wang, C. & Wu, C. Q. High Efficiency Thin Film Cadmium Telluride Solar Cells, AIP Conf. Proc., pp. 268-288. (1992). [19] Ferekides, C., Britt, J., Ma, Y. & Killian, L. High Efficiency CdTe Solar Cells by Close Space Sublimation. Proceedings of 23rd IEEE Photovoltaic Specialist Conference, pp. 389-393. (1993). [20] Nowshad Amin, Kamaruzzaman Sopian, Makoto Konagai, “Numerical modeling of CdS/CdTe and CdS/CdTe/ZnTe solar cells as a function of CdTe thickness”, Solar Energy Materials and Solar Cells, Vol: 91 pp. 1202-1208, (2007). [21] D.I. Kurbatov, V.V. Kosyak, M.M. Kolesnyk, A.S. Opanasyuk, S.N. Danilchenko, Yu. P. Gnatenko, “Structural and electrical properties of ZnS/CdTe and ZnTe/CdTe heterostructures”, Materials Chemistry and Physics, Vol: 138, pp. 731- 736, (2013). [22] Capper, P. P. Capper (Ed.), Narrow-gap II-VI Compounds for optoelectronic and Electromagnetic Applications. New York: Chapman and Hall. (1997) [23] Hussain, O.M., Reddy, P.S., Naidu, B.S., Uthanna, S. and Reddy, P.J. Characterization of thin film ZnCdS/CdTe solar cells. Sem. Sc. and Tech. 6: 690-694. (1991). 108

[24] Ray, S.C., Karanjai, M.K. & DasGupta, D. Deposition and characterization of

ZnxCd1-xS thin films by the dip technique. Thin Solid Films. 322: 117-122. (1998) [25] M. Abdel Rafea, A.A.M. Farag, O. El-Shazly, E.F. El-Wahidy, N. Roushdy, “Crystallite size estimation and photosensitivity characterization of nanocrystalline

Zn1_xCdxS (0 6 x 6 0.9) based heterojunctions prepared by simple dip-coating”, Microelectronic Engineering, Vol: 122, pp. 40–45, (2014). [26] S.M. Thahab, Adel H. Omran Alkhayat, Salah M. Saleh, “The optical properties of CdxZn1−xS thin films on glass substrateprepared by spray pyrolysis method”, Optik, Vol: 125, pp. 5112–5115, (2014). [27] S.M. Thahab, Adel H. Omran Alkhayat, Salah M. Saleh, “Influence of substrate type on the structural, optical and electrical properties of CdxZn1-xS MSM thin films prepared by Spray Pyrolysis method”, Materials Science in Semiconductor Processing, Vol: 26, pp. 49–54, (2014). [28] Waqar Mahmood, Nazar Abbas Shah, “CdZnS thin films sublimated by closed space using mechanical mixing: A new approach”, Optical Materials, Vol: 36, pp. 1449–1453, (2014). [29] Yamaguchi, T., Yamamoto, Y., Tanaka, T., Demizu, Y. & Yoshida, A., “(Cd,Zn)S thin films prepared by chemical bath deposition for photovoltaic devices”, Thin Solid Films. Vol: 281-282: pp. 375-378, (1996). [30] Wang, A., Zhao, J., and Green, M.A. 24% Efficient Silicon Solar Cells, Appl. Physics Letters, 57(6): 602-604. (1990) [31] Xuanzhi Wu. High-efficiency polycrystalline CdTe thin-film solar cells, Solar Energy, 77: 803 – 814. (2004) [32] Ramanathan, K. M., Contreras,A., Perkins, C. L., Asher, S., Hasoon, F. S., Keane,J., Young, D., Romero, M., Metzger, W., Noufi, R., Ward, J. Duda, A. Properties of 19.2% Efficiency ZnO/CdS/CuInGaSe2 Thin-Film Solar Cells, Progress in Photovoltaic’s: Research and Applications, 11:225-230. (2003) [33] Kastur Lal Chopra & Suhit Ranjan Das. Thin Film Solar Cells. Plenum Press, New & London, pp.91. (1983) [34] Bube, R.H. Photovoltaic Materials. England: Imperial Cillege Press. (1998). [35] Jenny Nelson. The Physics of Solar Cells. Imperial College Press, UK. (2006). 109

[36] Meena Dadu, Kapoor, A., Tripathi, K.N. Effect of operating current dependent series resistance on the fill factor of a solar cell, Solar Energy Materials a Solar Cells, 71:213–218. (2002). [37] Emery, K. Handbook of Photovoltaic Science and Engineering. Edited by A. Luque and S. Hegedus. John Wiley & Sons. Ltd. (2003). [38] Hegedus, S. S. and Shafarman, W. N. Thin-film solar cells: device measurements and analysis. Prog. Photovoltaics 12: 155-176. (2004). [39] Sites, J. Tavakolian, R. H. and Sasala, R. A. Analysis of appearent quantum efficiency. Solar Cells, 29: 39-48. (1990). [40] Gloeckler, M. & Sites, J. R. J. Appl. Phys. 95(4438): pp. 108. (2004). [41] Riordan & Hoddeson. Crystal Fire: The Invention of the Transistor and the Birth of the Information Age. Newyork: W. W. Norton & Co. (1997). [42] Loferski, J.J. Proceedings of the 12th IEEE Photovoltaics Specialists Conference. pp. 853. (1976). [43] Gokcen, N.A. & Loferski, J.J. Solar Energy Materials. 1(24): pp. 271. (1979) [44] Cusano, D.A. Solid State Electronics. 6: pp. 217. (1963) [45] Bonnet, D. Proceedings of 19th European Photovoltaic, Solar Energy Conference, Paris, pp. 1657-1662. (2004) [46] Ohyama,H., Aamoto,T., Kumazawa,S., Higuhi, H. Arita,T., Shibutani,S., Nishio,T., Nakajima, J., Tsuji, M., Hanafusa, A. Hibino, T., Omura, K., & Murozono, M. 16.0% Efficiency Thin-Film CdS/CdTe Solar Cells. Proc. of 26th IEEE PVSC, pp. 343-346. (1997). [47] Adirovich, E. I., Yuabov, Y. M. & Yagudaev sov G. R. Phys. Semicond, 3: pp. 61. (1969). [48] Chu, T.L., Chu, S.S., Ferekides, C., Wu, C.Q., Britt, J. and Wang, C. High Efficiency CdS/CdTe Solar Cells from Solution-Grown CdS Films, Proc. of 22nd IEEE PVSC, ps. 952-956 (1991). [49] Britt, J. & Ferekides,C. Thin film CdS/CdTe Solar Cell with 15.8% Efficiency. Appl. Phys. Lett. 62: pp. 2851. (1993). [50] PVTECH2011. www.pvtech.org/news/wednesday July. 27 (2011). [51] Clean Technica. http://cleantechnica.com/2013/02/27. (2013) [52] Ullal, H. S., Zweibel, K. & Roedern, B. V. Proceedings of the 26th IEEE Photovoltaic Specialists Conference, pp. 301. (1997). 110

[53] Yoshihiro Hamakawa. Thin Film Solar Cells; Next Generation Photovoltaic & Its Applications. Springer, New York. (2004). [54] Osuwa, J. C., Oriaku, C. I.and Kalu, I. A. Variation of optical band gap with post deposition annealing in CdS/PVA thin film, Chalcogenide Letters, 6:433 – 436. (2009). [55] Ashour, A., El-Kadry,N. andMahmoud, S. A. On the electrical and optical properties of CdS films thermally deposited by a modified source. Vacuum, 269:117- 120. (1995). [56] Moutinho, H.R., Albin, D., Yan,Y., Dhere, R.G., Li, X., Perkins, C., Jiang, C.S., To, B.and Al-Jassim, M.M. Deposition and properties of CBD and CSS CdS thin films for solar cell application, Thin Solid Films, 436:175–180. (2003). [57] Jenny Nelson. The Physics of Solar Cells. Imperial College Press, UK. (2006) [58] Greentechsolar. www.greentechmedia.com/First-Solar-Reclaims-CdTe-PV Efficiency -Record. (2013) [59] Streetman, G. Solid State Electronic Devices. Prentice-Hall of INDIA. (1995) [60] Dobson, K.D., Visoly-Fisher, I., Hodes, G. & Cahen, D. Stability of CdTe/CdS Thin-Film Solar Cells (review paper), Solar Energy Materials and Solar Cells. 62: 295-325. (2000) [61] Sarlund,J., Ritala,M., Leskela,M., Siponmaa,E., Zilliacus, R. Characterization of etching procedure in preparation of CdTe solarcells.Solar Energy Materials and Solar cells 44: 177-190. (1996). [62] Ferekides, C.S. Viswanathan, V. Morel, D.L. RF sputtered back contacts for CdTe/CdS thin film solarcells. Proceedings of the 26th Photovoltaic Solar Energy Conference. ps. 423-426. (1997). [63] Wu, X. High-Efficiency Polycrystalline CdTe Thin film Solar Cells. Solar Energy. 77: pp. 803-814. (2004). [64] Menner, R., Dimmler, B., Mauch, R.H. & Schock, H.W. II-VI compound thin film forwindows in heterojunction solar cells. J. Cryst Growth. 86: 906-911. (1988). [65] Abou-Elfotouh, F. & Coutts, T. RF magnetron sputtering of polycrystalline CdTe thin film solar cells. Int. J. Solar Energy, 12, pp. 223–246. (1992). [66] Chandrasekhar, S., Martinuzzi, S. & Nataren, F.Z. Improved efficiency of CdZnS thin-film solar cells. Canadian Journal of Physics. 63: 716-718. (1985). 111

[67] Ram, P.R., Thangaraj, R. & Agnihotri, O.P. Thin film CdZnS/CuInSe2 solar cells by spray pyrolysis. Bulletin of Materials Science. 8: 279-284. (1986).

[68] Reddy, K.T.R. & Reddy, P.J. Studies of ZnxCd1-xS films and ZnxCd1-xS/CuGaSe2 heterojunction solar cells. Journal of Physics D: Applied Physics 25: 1345-1348. (1992) [69] Bhattacharya, R.N., Contreras, M.A., Egaas, B., Noufi, R.N., Kanevce, A. &

Sites, J.R. High efficiency thin-film CuIn1-xGaxSe2 photovoltaic cells using a CdxZn1- xS buffer layer. Applied Physics Letters. 89: 253503/1-253503/2. (2006). [70] Oladeji, I.O., Chow, L., Ferekides, C.S., Viswanathan, V. & Zhao, Z.

Metal/CdTe/CdS/Cd1-xZnxS/TCO/glass: a new CdTe thin film solar cell structure. Solar Energy Materials Solar Cells. 61: 203–211. (2000). [71] Hossain, M.S., Amin, N. & Razykov T. Prospects of back contacts with back surface fields in high efficiency ZnxCd1-xS/CdTe solar cells from numerical modeling. Chalcogenide letters. 8(3): 187-198. (2011). [72] Yamaguchi, T., Matsufusa, J. & Yoshida, A. Optical Transitions in RF Sputtered

CuInxGa1-xSe2 Thin Films. 1992. Japanese Journal of Applied Physics 31: L703-L705. (1992). [73] Liou, J.J. Advanced Semiconductor Device Physics and Modelling. Artech House, Boston. (1994). [74] Oladeji, I.O. & Chow, L. Synthesis and processing of CdS/ZnS multilayer films for solar cell application. Thin Solid Films. 474: 77-83. (2005).

[75] Burton, L.C. & Hench, T.L. ZnxCd1-xS films for use in heterojunction solar cells. Applied Physics Letters. 29: 612-614. (1976). [76] J. Gaduputi, Characterization of Cadmium Zinc Telluride films and Solar Cells on Glass and Flexible Substrates by RF Sputtering M.Sc Thesis, University of South Florida, USA. (2004). [77] Ferekides, C.S. and Morel, D.L. Development of II-VI Based High Performance High Bandgap Device of Thin Film Tandem Solar Cells, Proceedings of the 29th IEEE PV Specialists Conference. Pp. 724-727. (2002). [78] Perrenoud, S. Buecheler, and A. N. Tiwari. Flexible CdTe solar cells and modules: challenges and prospects. Proc. SPIE, Vol. 7409 xcv ((2009) [79] G. Sivaraman, Characterization of Cadmium Zinch Telluride Solar Cells, M.Sc Thesis, University of South Florida, USA. (2003). 112

[80] V.Kumar, G.S.Sandhu, T.P.Sharma, and M.Hussain, Growth and Characterization of CdZnTe-Sintered Films, Research Letters in Materials Science, 63702. (2007). [81] Burgelman, M., Verschraegen, J., Degrave, S. & Nollet, P. Modeling thin film PV devices. Prog. Photovolt. Res. Appl. 12: pp. 143–153. (2004). [82] Burgelman, M. Thin Film Solar Cells: Fabrication, Characterization and Applications. J. Poortmans and V. Arkhipov (ed.). Cadmium Telluride Thin Film Solar Cells: Characterization, Fabrication and Modeling. West Sussex: John Wiley & Sons. (2006). [83] AMPS-1D Manual. For Windows 95/NT. The Electronic Materials and Processing Research Laboratory at the Pensylvania State University, University Park. PA 16802. (2007). [84] Sze, S.M. Physics of Semiconductor devices, 2nd ed: John Wiley & Sons, pp. 30- 103. (1981). [85] Kanevce, A., Gloeckler, M., Pudov, A.O. & Sites, J.R. Conduction-Band-Offset Rule Governing J-V Distortion in CdS/CI(G)S Solar Cells. Mat. Res. Soc. Symp.Proc.865: 221-227. (2005). [86] Gloeckler, M., Fahrenbruch, A.L. and Sites, J.R. Numerical modelling of CIGS and CdTe solar cells: setting the baseline, 3rd world conference on photovoltaic energy conversion. (2003). [87] Fahrenbruch, A.L. Modeling CdS/CdTe Solar Cells. www link at http://www. colostate.edu/orgs/pvlab/Pages/pdfDocs.htm, pp. 5-6. (08/07/2008) [88] Green, M. A. Solar Cells: Operating Principles, Technology, and System. Applications, Prentice-Hall, Inc. 149: pp. 25-56. (1982). [89] Bosio, A., Romeo, N., Mazzamuto, S. & Canevari, V. Polycrystalline CdTe thin films for photovoltaic applications, Progress in Crystal Growth and Characterization ofMaterials 52: 247-279. (2006). [90] Acevedo, A.M. Thin film CdS/CdTe solar cells: Research perspectives. Solar Energy 80: 675-681. (2006). [91] Arturo, M.A. Thin film CdS/CdTe solar cells: Research perspectives. Solar Energy 80: 675-681. (2006). 113

[92] Matin, M.A., Aliyu M., Quadery, A.H., Amin, N. Prospects of Novel Front and Back Contacts for High Efficiency Cadmium Telluride Thin Film Solar Cells from Numerical Analysis Solar Energy Materials & Solar Cells 94: 1496-1500 (2010a). [93] Matin, M.A., Amin, N., Zaharim, A. & Sopian, K. Ultra thin high efficiency CdS/CdTe thin film solar cells from numerical analysis. Proceedings of the 8th WSEAS international conference on Non-linear analysis, non-linear systems and chaos, 338-344 (2009a). [94] M. S. Hossain, M. A. Islam, M. M. Aliyu, T. Razykov, K. Sopian and N. Amin,

“Growth Optimization of ZnxCd1-xS Thin Films by Radio Frequency Magnetron Co- Sputtering for Solar Cell Applications”, Thin Solid Films, Vol. 548, pp. 202-209, (2013). [95] Chu, T.L., Chu, S.S., Britt, J., Ferekides, C. & Wu, O.Q. Cadmium zinc sulfide films and heterojunctions. J. Appl. Phys. 70: 2688-2693. (1991a). [96] Yin, S.Y., Fahrenbruch, A.L. & Bube, R.H. Photovoltaic properties of ZnCdS/CdTe heterojunctions by spray pyrolysis. J. Appl. Phys. 49: 1294-1296. (1978). [97] Valkonen, M.P., Kanniainien, T., Lindroos, S., Leskela, M. & Rauhala, E. Growth of ZnS, CdS and multilayer ZnS/CdS thin film by SILAR technique. Appl. Surf. Sci. 115: 386-392. (1997).

[98] Gunasekaran, M. & Ichimura, M. Deposition of Cd1-xZnxS (0 ≤x ≤1) Alloys by photochemical deposition.Japanese Journal of Applied Physics. 44: 7345-7350. (2005). [99] Zhou, J., Wu, X., Teeter, G., To, B., Yan, Y., Dhere, R.G. & Gessert, T. CBD-

Cd1−xZnxS thin films and their application in CdTe solar cells. Phy. Stat. Sol. 241(3): 775-778. (2004). [100] A. Fahrenbruch, Solar Cells 21, 399 (1987) [101] H. Uda, S. Ikegami, and H. Sonomura, Sol. Energy Mat. Sol. Cells 35, 293 (1994) [102] B. McCandless, Y. Qu, and R. Birkmire, in Proc. 1st World Conf. Photovoltaic Energy Conversion, pages 107–110, (1998). [103] J. Ponpon, Solid State Electron. 28, 689 (1985) [104] T. Chu, S. Chu, K. Han, and M. Mantavadi, in Proc. 20th IEEE Photovoltaic Specialist Conf., pages 1422–1425, (1988). 114

[105] T. Gessert, A. Duda, S. Asher, C. Narayanswamy, and D. Rose, in Proc. 28th IEEE Photovoltaic Specialist Conf., pages 654–657, (2000). [106] N. Romeo, A. Bosio, and R. Tedeschi, in Proc. 2nd World Conf. Photovoltaic Energy Conversion, pages 446–447, (1998). [107] Orgassa, K., Schock, H.W. & Werner, J.H. Alternative back contact materials for thin film Cu(In,Ga)Se2 solar cells. Thin Solid Films 387: 431-432. (2003). [108] Nowshad Amin, M.A. Matin, M.M. Aliyu, M.A. alghoul and K. Sopian, “Prospects of back surface field effect in ultra thin high efficiency CdS/CdTe solar cells from numerical analysis”, International journal of photoenergy, Vol: 2010, pp. 1-8, DOI: 10.1155/2010/578580, (2010).

[109] Razykov, T.M. Physical properties of ZnxCd1-xS films fabricated by CVD in hydrogen flow for use in solar cell. Solar energy materials. 12: 233-238. (1985) [110] T.M. Razykov, presented at the 7th International conference on Thin Films, 7-11 December, 1987, New Delhi, India (to be published in Thin Solid Films 164, 301 (1988). [111] Razykov, T.M., Kadyrov, B.KH. & Khodyaeva, M.A. Energy band models of n-ZnxCd1-xS-p-CdTe and n-ZnxCd1-xS-p-Si (0≤x≤1) heterojunctions. Phys. Stat. Sol. a 91: K87-K91. (1985). [112] Singh, V.P. & Singh, S. Some physical properties of ZnCdS solid solutions. Czech. J. Phys. b 26: 1161-1166. (1976).

[113] Burton, L.C. The ZnxCd1-xS/Cu2S heterojunction: review and recent measurements. Solar Cells. 1: 159-174. (1980). [114] Gale R. P., Fan J. C., Turner G. W. and Chapman R. L. A new high efficiency GaAs solar cell structure using a heterostructure back-surface field, Proc. of 17th IEEE photovoltaic specialist’s conference,ps. 1422-1425. (1984). [115] DeMoulin P. D. and Lundstrom M. S. Projections of GaAs solar-cell performance limits based on two-dimensional numerical simulation, IEEE Trans. Electron Devices 36: 897-905. (1989). [116] DeMoulin, P. D., Lundstrom, M. S. and Schwartz, R. J. Back-Surface Field Design for n/p GaAs Cells. Solar Cells: Sci., Technol., Appl. Econ, 20:229-236. (1987). [117] Kraft,D.,Thissen,A., Broetz, J., Flege,S., Campo, M., Klein,A. and Jaegermann, W. Characterization of tellurium layers for back contact formation on close to technology treated CdTe surfaces.J.Appl. Phys. 94: 3589-3599. (2003). 115

[118] Plotnikov, V.V., Kwon, D., Wieland, K.A. & Compaan, A.D. 10% Efficiency Solar Cells with 0.5 μm of CdTe. Proc. of the 34th IEEE PVSC, 1435-1438. (2009). [119] Rioux, D., Niles, S.W., and H¨ochst, H. ZnTe: a potential interlayer to form low resistance back contacts in CdS/CdTe solar cells. J. Appl. Phys., 73(12): 8381-8385. (1993) [120] Späth, B., Fritsche, J., Klein, A. and Jaegermann, W. Nitrogen doping of znte and its influence on cdte/znte interfaces. Appl. Phys. Lett. 90: 62110-62115. (2007). [121] Makhratchev, K., Price, K. J., Ma, X., Simmons, D. A. Drayton, J. Ludwig, K. Gupta, A. Bohn, R.G. Compaan, A. D. ZnTe:N back contact to CdS/CdTe solar cells. Proc. 28th IEEE Photovoltaic Spec. Conf. ps. 475-478. (2000). [122] Gessert,T. A., Asher,S., Johnson, S., Duda,A., Young,M. R. and Moriarty, T. Formation of ZnTe:Cu/Ti Contacts at High Temperature for CdS/CdTe Devices.4th World Conf. Photovoltaic Energy Conv.,ps. 432-435. (2006). [123] Tang, J. Mao, D. Ohno, T. R. Kaydanov, V. and Trefny, J. U. Properties of ZnTe:Cu thin films and CdS/CdTe/ZnTe Solar Cells,Proc. 26th IEEE Photovoltaic Specialists Conf. (97). p. 439-442. (1997). [124] Oriaku, C.I. Ezema, F.I. and Osuwa, J.C. Fabrication and Optical Constants of CdMnS Ternary Thin Films deposited by Chemical Bath, Pacific Journal of Science and Technology. 10(1): 413-416. (2009). [125] Wu, X., Zhou, J., Duda, A., Yana, Y., Teeter, G., Asher, S., Metzger, W. K., Demtsu, S., Wei, S.-H. and Noufi, R. Phase control of CuxTe film and its effects on CdS/CdTe solar cell .Thin Solid Films. 515 (15): 5798-5803. (2007). [126] Hegedus, S. S. and McCandless, B. E. CdTe contacts for CdTe/CdS solar cells: effect of Cu thickness, surface preparation and recontacting on device performance and stability. Solar Energy Materials and Solar Cells88:75-95. (2005). [127] Farag, B. S. and Khodier, S.A. Direct and indirect transitions in copper telluride thin films. Thin Solid Films201: 231-240. (1991). [128] Visoly-Fisher, I., Dobson, K. D., Nair, Bezalel, J. E., Hodes, G. and Cahen, D. Factors Affecting the Stability of CdTe/CdS Solar Cells. Advanced Functional Materials 13:289-299. (2003). [129] Romeo, N., Bosio, A. Romeo, A. and Mazzamuto, S. Industrial Upscaling of CdTe/CdS Thin Film Solar Cells. In Materials Research Society, 1165: 263-273. (2009). 116

[130] Hossain, M.S., Amina, N. & Razykov, T. Prospects of Back Contacts with Back Surface Fields in High Efficiency ZnxCd1-xS /CdTe Solar Cells from Numerical Modeling. Chalcogenide Letter 8(3): 187-198. (2011a). [131] Matin, M. A.,Tomal, M. U., Robin, A. M., and Amin, N. “Numerical Analysis of Novel Back Surface Field for HighEfficiency Ultrathin CdTe Solar Cells”, International Journal of Photoenergy, Volume 2013, Article ID 652695, 8 pages, (2013). [132] Zweibel, K. Will We Have Enough Materials for Energy-Significant PV. (online). http://www.nrel.gov/docs/fy04osti/35098.pdf. (5 January 2011).