USF Libraries
USF Digital Collections

Development of cadmium selenide as an absorber layer for tandem solar cells

MISSING IMAGE

Material Information

Title:
Development of cadmium selenide as an absorber layer for tandem solar cells
Physical Description:
Book
Language:
English
Creator:
Jeedigunta, Sathyaharish
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
substrate temparature
transmission
fill factor
CSS
Dissertations, Academic -- Electrical Engineering -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Cadmium Selenide is a binary compound. It has a band gap of 1.7 eV. This is one of the suitable materials for an absorber layer in the top cell of a tandem solar cell. CIGS with a low Gallium content has a band gap of 1 eV suits well as an absorber layer for the bottom cell. CIGS cells have already attained an efficiency of 15% 1,2. Since years, research has been done in developing the bottom cell. The results of the bottom cell are promising. So the fabrication of an efficient top cell in a tandem solar cell is a challenge. To achieve a high tandem efficiency of above 25 %, the top cell has to contribute at least 2/3 of the total efficiency, which necessitates the top cell to have at least 16 to 18 % efficiency 3. Development of a defect free absorber layer is a crucial step in this process to achieve the above goals besides optimizing other layers. Selenium vacancies in CdSe make the absorber layer n-type. CdSe is deposited by closed space sublimation. Deposition of CdSe at higher substrate temperatures in comparison to the standard conditions was studied. ZnSe acts as an insulating layer. It is thermally evaporated in an Evaporation system. Copper acts as a metal contact on top of the insulator resulting in a MIS structure. Copper is also deposited by Thermal Evaporation. Devices are fabricated on different substrates like SnO2: F, AZO etc. Fabricated cells are characterized by J-V and Spectral response measurements. Devices fabricated on SnO2: F substrates show typical open circuit voltages of around 220 mV, short circuit current densities of 10.02 mA/cm2 and fill factors around 33 %. N-type CdS when deposited on SnO2: F below the absorber layer further improved Voc's to around 330 mV. Annealing of these devices improved Voc's to about 350 mV but Jsc's remained 7.21 mA/cm2.
Thesis:
Thesis (M.S.E.E.)--University of South Florida, 2004.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Sathyaharish Jeedigunta.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 88 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001469377
oclc - 55623482
notis - AJR1131
usfldc doi - E14-SFE0000279
usfldc handle - e14.279
System ID:
SFS0024974:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001469377
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 040524s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000279
035
(OCoLC)55623482
9
AJR1131
b SE
SFE0000279
040
FHM
c FHM
090
TK145
1 100
Jeedigunta, Sathyaharish.
0 245
Development of cadmium selenide as an absorber layer for tandem solar cells
h [electronic resource] /
by Sathyaharish Jeedigunta.
260
[Tampa, Fla.] :
University of South Florida,
2004.
502
Thesis (M.S.E.E.)--University of South Florida, 2004.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 88 pages.
520
ABSTRACT: Cadmium Selenide is a binary compound. It has a band gap of 1.7 eV. This is one of the suitable materials for an absorber layer in the top cell of a tandem solar cell. CIGS with a low Gallium content has a band gap of 1 eV suits well as an absorber layer for the bottom cell. CIGS cells have already attained an efficiency of 15% [1,2]. Since years, research has been done in developing the bottom cell. The results of the bottom cell are promising. So the fabrication of an efficient top cell in a tandem solar cell is a challenge. To achieve a high tandem efficiency of above 25 %, the top cell has to contribute at least 2/3 of the total efficiency, which necessitates the top cell to have at least 16 to 18 % efficiency [3]. Development of a defect free absorber layer is a crucial step in this process to achieve the above goals besides optimizing other layers. Selenium vacancies in CdSe make the absorber layer n-type. CdSe is deposited by closed space sublimation. Deposition of CdSe at higher substrate temperatures in comparison to the standard conditions was studied. ZnSe acts as an insulating layer. It is thermally evaporated in an Evaporation system. Copper acts as a metal contact on top of the insulator resulting in a MIS structure. Copper is also deposited by Thermal Evaporation. Devices are fabricated on different substrates like SnO2: F, AZO etc. Fabricated cells are characterized by J-V and Spectral response measurements. Devices fabricated on SnO2: F substrates show typical open circuit voltages of around 220 mV, short circuit current densities of 10.02 mA/cm2 and fill factors around 33 %. N-type CdS when deposited on SnO2: F below the absorber layer further improved Voc's to around 330 mV. Annealing of these devices improved Voc's to about 350 mV but Jsc's remained 7.21 mA/cm2.
590
Adviser: Morel, Don
653
substrate temparature.
transmission.
fill factor.
CSS.
690
Dissertations, Academic
z USF
x Electrical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.279



PAGE 1

Development Of Cadmium Sele nide As An Absorber Layer For Tandem Solar Cells by Sathyaharish Jeedigunta A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Electrical Engineering Department of Electrical Engineering College of Engineering University of South Florida Major Professor: Don Morel, Ph.D. Christos Ferekides, Ph.D. Y.L.Chiou, Ph.D Date of Approval: March 26, 2004 Keywords: css, fill factor, trasmission, substrate temperature Copyright 2004 Sathyaharish Jeedigunta

PAGE 2

Dedication To my parents and brother.

PAGE 3

Acknowledgments I would like to convey my deepest apprecia tion to my major professor, Dr.Don Morel for giving me an opportunity to be a me mber of his research group. His excellent guidance was of great help during my thesis work. I am very much thankful for his support and guidance. I would also like to thank Dr.Chris Fereki des and Dr.Y.L.Chiou for their assistance. I am more than happy to work with my fellow researchers and express my thanks to all of them who have help ed me in every stage of my work. My special thanks to Prasanna Mahawela, Harish Sankarnarayan an, Sridevi Vakkalanka, Shirish Pethe, Venkatesh and Madhavi. I am thankful for th e encouragement provided by Vijay, Raghu and other friends. I am thankful to god who has given me his support and blessings. I am very much indebted to my parents and my brother for their encouragement.

PAGE 4

i Table of Contents List of Tables v List of Figures vi List of Abbreviations ix Abstract xii Chapter 1 Introduction 1 1.1 Introduction 1 1.2 Economics of Photovoltaics 4 1.3 Outline 6 Chapter 2 Background 7 2.1 Photovoltaics 7 2.2 Benefits of Photovoltaics (PV) 7 2.3 Photovoltaic Effect 8 2.4 History 8 2.4.1 Problems with Amorphous Silicon 9 2.4.2 GaAsan Interesting Material 9 2.4.3 Thin Film Polycrystalline Solar Cells 9

PAGE 5

ii 2.4.4 Advantages of Thin Film Polycrystalline Solar Cells 10 2.4.5 Disadvantages of Thin Film Polycrystalline Solar Cells 10 2.5 Structure of Solar Cell 11 2.6 Principle of Operation of Solar Cell 12 2.7 Typical I-V Curve 13 2.7.1 Effect of Series and Shunt Resistance 14 2.7.1.1 Effect of Series Resistance 15 2.7.1.2 Effect of Shunt Resistance 15 2.8 Requirements of the Materials used in Solar Cells 16 2.8.1 Back Contact 16 2.8.2 Absorber Layer 17 2.8.3 Window Layer 17 2.8.4 Front Contact 18 Chapter 3 Literature Review 19 Chapter 4 Device Structure and Fabrication 32 4.1 Tandem Cell Structure 32 4.2 Structure of a CdSe Based Top Cell 34 4.2.1 Substrate 34 4.2.1.1 Cleaning Procedure 34 4.2.2 Back Contact 35 4.2.2.1 Deposition of SnO2: F 36

PAGE 6

iii 4.2.2.2 AZO as a Back Contact 36 4.2.2.3 Structure of Cell with AZO at the Back Contact 37 4.2.2.4 Deposition Procedure for AZO 38 4.2.2.5 Transmission of AZO 38 4.2.3 Absorber Layer 39 4.2.3.1 Properties of CdSe 41 4.2.3.2 Preparation of the CdSe Source 41 4.2.3.3 Deposition Procedure for CdSe 41 4.2.3.4 Parameters 42 4.2.3.5 Transmission of CdSe 43 4.2.3.6 Experiments Performed on the Absorber Layer 43 4.2.4 Window Layer 44 4.2.4.1 Properties of ZnSe 45 4.2.4.2 Deposition Procedure for ZnSe 45 4.2.4.3 Evaporation System 46 4.2.4.4 Transmission of ZnSe 47 4.2.5 Front Contact 47 4.2.5.1 Transmission of Copper 48 4.3 Deposition of Cadmium Sulphide 48 4.3.1 Transmission of CdS 50 4.3.2 Structure of Cell with CdSe on CdS/SnO2: F 51

PAGE 7

ivChapter 5 Results and Discussions 52 5.1 Results 52 5.1.1 AZO as a Back Contact 52 5.1.2 CdS on SnO2: F at the Back Contact 54 5.1.3 Absorber Layer at Higher Substrat e Temperatures 57 5.1.4 Effect of Time on Device Performance 63 Chapter 6 Conclusions 66 References 70

PAGE 8

v List of Tables Table 3.1 Device Parameters 24 Table 4.1 Properties of CdSe 41 Table 4.2 Properties of ZnSe 45 Table 5.1 Device Parameters for Devices with AZO as a Back Contact 52 Table 5.2 Device Parameters for Devices with CdS on SnO2: F at Back Contact 55 Table 5.3 Effect of Substrate Temperature on De vice Performance with 200 ZnSe 58 Table 5.4 Effect of Substrate Temperature on Device Performance with 220 ZnSe 59 Table 5.5 Effect of Substrate Temperature on Device Perfor mance with 180 ZnSe 61 Table 5.6 Effect of Time on Device Performance 63

PAGE 9

vi List of Figures Fig 1.1 Module World Market Growth 4 Fig 2.1 Substrate Structure 11 Fig 2.2 Superstrate Structure 11 Fig 2.3(a) Equivalent circuit of an ideal solar cell (b) Equiva lent circuit of solar cell with series and shunt resistances 12 Fig 2.4 Typical I-V Curve 14 Fig 2.5 Effect of Series Resistance 15 Fig 2.6 Effect of Shunt Resistance 16 Fig 3.1 XRD Pattern of CdSe Film Deposited fro m an Electrolyte Containing CdCl2 20 Fig 3.2 XRD Pattern of CdSe Film Deposited fro m an Electrolyte Containing CdSO4 at a) T=18C b) T=75C 21 Fig 3.3 I-V Curve 22 Fig 3.4 Photo-current Action Spectra for Di fferent Electrodes a) OTE/SnO2 b) OTE/CdSe c) OTE/SnO2/CdSe 23 Fig 3.5 I-V Curve of CdSe and CdZnSe 25 Fig 3.6 a) AFM results for Ts=300 K b) AFM results for Ts=400 K 26 Fig 3.7 I-V characteristics of a) undopedPMeT/CBD CdSe b) doped PMeT/CBD CdSe c) doped PMeT/CBD CdSe (in presence of Silicon Tungstic Acid) 28 Fig 3.8 Variation of Band gap Vs Deposition Temperature 29 Fig3.9 Variation of Resistivity Vs Depositi on Temperature 30 Fig 3.10 J -V characteristics of CdSe thin f ilms with temperature as variable a) with bias voltage < 0.6V b) with bias voltage > 0.6V 31

PAGE 10

viiFig 4.1 Tandem Solar Cell Structure 32 Fig 4.2 Structure of CdSe Based Top Cell 34 Fig 4.3 Transmission Response of SnO2: F 35 Fig 4.4 Wurtzite Structure of Zinc Oxide 37 Fig 4.5 Structure of Cell with AZO at the Back Contact 37 Fig 4.6 Transmission Response of AZO 38 Fig 4.7 Wurtzite structure of CdSe 39 Fig 4.8 AFM image of CdSe 40 Fig 4.9 Closed Space Sublimation Setup 42 Fig 4.10 Transmission Response of CdSe 43 Fig 4.11 Structure of ZnSe 45 Fig 4.12 Evaporation System 46 Fig 4.13 Transmission Response of ZnSe 47 Fig 4.14 Transmission Response of Copper 48 Fig 4.15 Structure of CdS 49 Fig 4.16 Transmission Response of CdS 50 Fig 4.17 Structure of Cell with CdSe on CdS/SnO2:F 51 Fig 5.1(a) and (b) J-V Curve for Device # sh18-g 53 Fig 5.2 Spectral Response for Device # sh18-g 54 Fig 5.3(a) and (b) J-V Curve for Device # sh39-1-a 55 Fig 5.4 (a) and (b) J-V Curve for Device # sh40-1-e 56 Fig 5.5 Spectral Response for Device # sh39-1a and # sh40-1-e-a 57 Fig 5.6 J-V Curve for the Devices # sh44-1-f, # s h45-1-e, # sh48-2-l and sh52-1-l 58

PAGE 11

viiiFig 5.7 (a) and (b) J-V Curve for Devices # sh49-2-f, # sh50-2-e and # sh53-1-l 60 Fig 5.8 (a) and (b) J-V Curve for the Devices # sh49-3-b and # sh51-2-c 61 Fig 5.9 Spectral Response for Devices # sh50-2-e, #sh49-3-b and # sh52-1-f 62 Fig 5.10(a) and (b) J-V Curve for Device # sh44-1-f 64 Fig 5.11 Spectral Response for Device # sh44-1-f 65

PAGE 12

ix List of Abbreviations eVelectron volt Q.E-Quantum Efficiency Voc-Open circuit voltage Isc-Short circuit current Jsc-Short circuit current density Vmp-Maximum voltage Imp-Maximum current Rs-Series Resistance Rsh-Shunt Resistance I-V-Current-Voltage PV-Photovoltaic MW-Mega Watt A.C-Alternating current D.C-Direct current mV-milli volts mA-milli amps CIGS-Copper Indium Gallium Diselenide CdSe-Cadmium Selenide ZnSe-Zinc Selenide

PAGE 13

xSnO2-Tin Oxide CdS-Cadmium Sulphide Si-Silicon GaAs-Gallium Arsenide AZO-Aluminum doped Zinc Oxide ITO-Indium Tin Oxide CdTe-Cadmium Telluride CdZnTe-Cadmium Zinc Telluride Mo-Molybdenum Cu-Copper HF-Hydro Fluoric TMT-Tetra Methyl Tin Cd-Cadmium ZnO-Zinc Oxide In-Indium Al2O3-Aluminum Oxide TCO-Transparent Conducting Oxide -Angstrom CSS-Close Space Sublimation CBD-Chemical Bath Deposition Tsub-Substrate Temperature Tsou-Source Temperature RT-Room Temperature

PAGE 14

xitdep-Deposition time C-Centigrade -Absorption coefficient -micron R-Roughing valve F-Foreline valve V-Vent valve G-Gate valve

PAGE 15

xii Development of Cadmium Selenide as an Absorber Layer for Tandem Solar Cells SathyaHarish Jeedigunta ABSTRACT Cadmium Selenide is a binary compound. It has a band gap of 1.7 eV. This is one of the suitable materials for an absorber laye r in the top cell of a tandem solar cell. CIGS with a low Gallium content has a band gap of 1 eV suits well as an absorber layer for the bottom cell. CIGS cells have already attained an efficiency of 15% [1,2]. Since years, research has been done in developing the bottom cell. The results of the bottom cell are promising. So the fabrication of an efficient top cell in a tandem solar cell is a challenge. To achieve a high tandem efficiency of above 25 %, the top cell has to contribute at least 2/3 of the total efficiency, which necessitate s the top cell to have at least 16 to 18 % efficiency [3]. Development of a defect free absorber laye r is a crucial step in this process to achieve the above goals besides optimizing ot her layers. Selenium vacancies in CdSe make the absorber layer n-type. CdSe is deposited by closed space sublimation. Deposition of CdSe at higher substrate temp eratures in comparison to the standard conditions was studied. ZnSe acts as an insulating layer. It is thermally evaporated in an Evaporation system. Copper acts as a metal c ontact on top of the insu lator resulting in a MIS structure. Copper is also deposited by Th ermal Evaporation. Devices are fabricated

PAGE 16

xiiion different substr ates like SnO2: F, AZO etc. Fabricated cells are characterized by J-V and Spectral response measurements. Devices fabricated on SnO2: F substrates show typical ope n circuit voltages of around 220 mV, short circuit current densities of 10.02 mA/cm2 and fill factors around 33 %. Ntype CdS when deposited on SnO2: F below the absorber layer further improved Voc’s to around 330 mV. Annealing of these devices improved Voc’s to about 350 mV but Jsc’s remained 7.21 mA/cm2.

PAGE 17

1 Chapter 1 Introduction 1.1 Introduction As the world’s population increases at an alarming rate, the need to meet their requirements is an important aspect to be considered. The modern day computerized world needs less labor. Majority of the tasks are performed by machines driven by power. To perform all these tasks w ithout hindrance, a continuous supply of power is needed. The conventional sources of energy like hyd ro, thermal and nuclear are able to meet the current requirements to some exte nt. But whether the conventional sources of energy are sufficient to meet the demand in coming centuries remains a question. The consumption of non-renewable energies has caused a huge damage to the environment. Electricity generated from fossil fuels has led to high concentr ations of harmful gases like carbon monoxide, carbon dioxide, etc in the atmosphere. This in turn led to many problems such as ozone layer depletion and global warming Many hydroelectric plants have been constructed in recent past. This is one of the better ways of generating elect ricity with minimum waste pr oducts. Construction of these projects involves a lot of capital to be invested. Moreover most of the rivers are not

PAGE 18

2 perennial and hence cannot generate power continuously through out the year. Another disadvantage of hydroelectric plants is th eir detrimental effect on the ecosystem. Construction of large hydroelectric plan ts has caused tremors in those areas. The nuclear energy is another alternative source of energy, which has a potential to produce many megawatts of power. Nuclea r energy if used for constructive purposes can generate power. On the other hand if it is used for destructive purposes can be benignant to human life. The op tion of nuclear energy raises sa fety issues like disposal of its waste products. So these factors make ma ny nations to rethi nk about constructing nuclear plants. Hence an alternative form of ener gy is an obvious choice to meet the requirements. Thus there is an immediate need for the development of non-conventional energy sources. Tidal, solar, geo-therma l and wind are few non-conventional energy sources. These are inexhaustible energy sources and he nce are renewable. They cause fewer emissions. Their use can reduce pollutio n. They stand out as a viable source of clean and limitless energy. The tidal energy would be an attractive al ternative if it were more reliable. However the vast regions across the world are covered by land and hence it is not a feasible option. Windmills do not meet the practical applicati ons of present day. They can be used only for some specific purpose a nd also cannot generate many megawatts of power.

PAGE 19

3 Sun is the center of all the activities in the world. It is calculated that in every second about 6 X 1011 kg of H2 is converted to He in sun. An enormous amount of energy (approx 4 X 1020 J) is produced in this process [9]. This energy is emitted as an electromagnetic radiation. It is projected that this constant amount of energy can be obtained for at least 1010 years. Hence among the above renewable energy sources, solar energy is the most readily av ailable source of energy. But how can we tap this bountiful source of energy? Research in this field has led to the development of first photo-voltaic ce ll (Bell laboratories, 1954) [4]. In 1960’s, photo voltaic modules were used to power sp ace satellites. Since then many researchers have spent decades in this field in ma king solar energy a viable alternative. The first homojunction sili con solar cell developed dur ing 1954 had an efficiency of only 6 %. But the progress in silicon technology resulted in single crystal silicon solar cells with high reliability and efficiencies reaching above 25 %. These silicon photovoltaic modules are very expensive. Less expensive polycrystalline silicon modules surpass the previous ones. Polycrystalline materials result in less efficient solar cells than single crystal silicon. Currently about 95 % of commercially av ailable photovoltaic modules are made from crystall ine silicon cells and only 5 % is contribu ted by thin film technology [5]. Thin films solar cells have become a very good competitor for the single crystal and polycrystalline silic on solar cells because of th eir cheaper raw material as well as processing costs. Substrates used in th in film solar cells are either a metal or a glass or a polymer. These are relatively ch eaper when compared to silicon wafers.

PAGE 20

4 1.2 Economics of Photovoltaics The average cost per kilowatt-hour from c onventional energy sources is 6-7 cents [6]. Present day photo-voltaics generate 2030 cents per kilo watt-hour. Hence the cost per kilo watt hour should be reduced by 4 to 5 times to make PV an economical candidate. Various factors like module efficiency, lif etime and cost per unit area effects the PV energy cost. Lower efficiency module costs less than an equivalent higher efficiency module to produce the same amount of electricity. According to the statistics of 2001, PV market is generating 381 MW of power which is being divided among US, Japan a nd few European nations. U.S companies export 75 % of PV products to developing count ries. Market growth in last decade was about 15 25 %. Market growth will depend on the continuing decline of PV costs relative to conventional supplies. Fig 1.1 Module World Market Growth

PAGE 21

5 Support programs in many countries are involved in acceler ating the market growth. In Germany, a law was passed which se ts a rebate rate of 0.5 Euro/Kwh of PV generated electricity. Japan’s very ambitious 70,000 roof program resulted in 63 % of its production in 1999. Interest in this roof top app lication started in USA during 1970’s [7].Japan also came into the race in mid 1980’s with the construction of a test bed for over 200 residential systems on Rokko Island in 1986.After several years of evaluating technical issues to their grid connection, a subsidized installation program was launched by the Japanese government in 1993.By 2010 Japan targets to meet a 1.5 million rooftop applications. US government announced a “million roof ” program in mid 1997 targeting this number of systems by 2010.At the end of 1997, Eu ropean Union also agreed to a similar target by 2010 with half the systems to be installed in Europe and half outside. Since then other countries like Netherlands, Italy a nd Australia have a nnounced targets for photovoltaic roofs. The demand for electricity is often highe st during the day. Solar energy can be drawn more in those parts of world near the equator. Other economical issue that has to be taken into consideration is the installati on of inverters to get an A.C output. The grid also requires battery setup for the energy to be stored so that it can be used whenever it is required.

PAGE 22

6 1.3 Outline After having discussed a bout the various kinds of energy sources-conventional and non-conventional, silicon cells, economics of photovoltaics, further topics stress on background, history of photovoltaics, solar cells A review on literature helps readers to better understand the ongoing research. Major fo cus is on experiments performed and on results obtained. A few conclusions are drawn with some suggestions towards achieving the future goals.

PAGE 23

7 Chapter 2 Background 2.1 Photovoltaics Photovoltaics (PV) is the di rect conversion of sunlight to electricity. When photons of energy greater than the band gap are incident on a semiconductor, electrons are excited from valence band to the conduction band. If these excited electrons can be collected before they recombine, electricity can be produced. Electri city obtained can be used right away, can be converted to AC or can be stored for a later use. PV systems are modular and hence they can be engineered fo r any size of application from a smallest wrist watch to a huge power station. 2.2 Benefits of Photovoltaics (PV) PV has many benefits. A few of them are: Reliability: Photovoltaics operate reliably for long periods of time without any maintenance. Low costs: They have low operating costs. Environment friendly: They are pollution fr ee and have no by products or wastes. They are clean and silent. Secure energy: Photovoltaics prov ide inexhaustible domestic energy.

PAGE 24

8 2.3 Photovoltaic Effect The technique of conversion of solar ener gy to electricity occu rs by “photovoltaic effect” which was first observed by Antoine Ce aser Becquerel in 1839[8].It is defined as “an emergence of an electric voltage between two electrodes attached to a solid or liquid system upon shining light on to this system”. 2.4 History Few decades after the discovery of phot ovoltaic effect, a group of scientists observed that solid selenium also exhibited photo conductivity when light was incident on it. Initially the research was focused on se lenium and cuprous oxide cells. But they did not show any promising performance .They ha d efficiencies of ar ound 1 %. Nearly after four decades of no significant progress, pe ople started to expl ore new materials for photovoltaic applications. During 1954 three scientists, Chapi n, Fuller and Pearson were successful in developing a homojunction silicon so lar cell with an efficiency of 6 % [9]. Further research resulted in attaining efficiencies of around 10 %.Silicon sola r cells found their first commercial application in space crafts. The first silicon solar cells were made from single crystal s ilicon. The manufacturing process of single crystal silicon from its raw material involves various in termediate steps making single crystal silicon expensive. Single crystal silicon is an indirect band gap semiconductor and hence it requires more material to absorb light. During the same period polycrystalline Si, amorphous silicon, GaAs and thin film solar cells like CuxS/CdS were developed. CuxS/CdS solar cells had an efficiency of 6 %.

PAGE 25

9 Further progress improved their efficiency to 9 %.But these cells degraded because of copper diffusion. 2.4.1 Problems with Amorphous Silicon Amorphous silicon is another viable candidate as a solar cell material. But these cells showed comparatively lesser efficiencies This could be because the crystal structure of the material. Amorphous silicon solar cells performed poorly when compared to its counterparts because of their instability. Th eir performance degraded considerably over a period of time. 2.4.2 GaAsan Interesting Material The emerging technologies aim at brin ging an optimization between the performance of cells and their processing costs. GaAs is a direct band gap material with a band gap of 1.43 eV. During 1970’s reasonabl e efficiencies of around 10 % were reported for GaAs cells. Solar cells with Ga As are expensive due to high material and processing costs. Therefore a wide range of promising polycrystalline compound semiconductor materials become a choice in solar cell processing. 2.4.3 Thin Film Polycrys talline Solar Cells Polycrystalline materials are composed of many localized single crystals. These single crystals are called grains. The transition regions in polycrystalline solids existing between various single crystals are calle d grain boundaries. These regions contain structural and bonding defects which act as centers for impurity collection.

PAGE 26

10 Most of the thin film po lycrystalline semiconductor ma terials have a direct band gap. Hence it is easy to excite electrons fr om the valence band to the conduction band. Most of these materials have high absorption co efficients so that few micron thick film is enough to absorb major portion of the so lar spectrum. The range of operating temperatures is less and hence more uniform pol ycrystalline films can be obtained. It also involves inexpensive processing te chniques and material costs. 2.4.4 Advantages of Thin Film Polycrystalline Solar Cells a) Low processing costs. b) Relatively less material costs. c) Substrates used mostly are glass, meta l sheet or polymers which are inexpensive. d) Thin film solar cells can be produced on large area substr ates giving a hi gh output with a low unit cost. 2.4.5 Disadvantages of Thin Film Polycrystalline Solar Cells a) Thin film polycrystalline so lar cells have lesser efficiencies when compared to its equivalent counter parts. b) Polycrystalline materials have grain bounda ries. They have crystal defects. These defects are the centers of impurities. These defects act as recombination centers. Recombination of charge carriers re sult in poorer electronic properties. c) Reproducibility of large area uniform films is not assured. d) Stability of these cells is a nother issue to be yet resolved.

PAGE 27

11 2.5 Structure of Solar Cell Sun light Fig 2.1 Substrate Structure A solar cell can have two structures, supers trate structure or a substrate structure. The difference between these two structures is that in a substrate structure light is incident at the front contact whereas in a s uperstrate structure light first passes through the substrate. The two stru ctures are shown in Fig 2.1 and Fig 2.2 respectively. For a tandem solar cell these two st ructures are placed one over th e other so that the top cell absorbs the lower wavelengths and the bottom cell absorbs th e higher wavelengths. Sun light Fig 2.2 Superstrate Structure Back contac t Absorber layer Window layer Substrate Front contact Back contact Absorber layer Window layer Front contact Substrate

PAGE 28

12 2.6 Principle of Opera tion of Solar Cell A solar cell is a p-n junction. Solar energy is converted into el ectricity when light is incident at the junction. When light is incident at the juncti on, photons with energy lesser than the bandgap makes no contribution to the output. Photons with energy greater than the bandgap will contribute energy equa l to the bandgap to the output while the excess energy will be wasted as heat. The absorbed energy generates charge carriers across the junction which can be separated by the electric field pr esent at the junction. Equivalent circuit of an ideal solar cell is shown in Fig 2.3 (a). An equivalent circuit of the solar cell with th e losses due to series and the shunt resistances is shown in Fig 2.3 (b). Fig 2.3(a) Equivalent circuit of an ideal sola r cell (b) Equivalent ci rcuit of solar cell with series and shunt resistances

PAGE 29

13 Under no illumination, current across the junction is given by the ideal diode equation I=I0[(eqV/AKT-1)] (2.1) Under illumination due to the generati on of charge carriers, a current IL flows through the junction. The current acro ss the junction is given by eq(2.2) I=I0[(eqV/AKT-1)]-IL (2.2) Where I0 is the reverse saturation current IL is the light generated current A is the diode ideality factor 2.7 Typical I-V Curve A typical I-V response of a solar cell is shown in Fig 2.4[8]. In eq (2.2) when current I is made zero, Voc is given by Voc = (AkT/q)*[ln (IL/I0 +1)] (2.3) Similarly under short-circuit conditions, when voltage V is zero in eq(2.2) the short circuit current is given by Isc =-IL (2.4) Vmp and Imp are the maximum voltage and current that can be extracted at the output. The squareness of the curve determ ines the output parameters of a device.

PAGE 30

14 Fig 2.4 Typical I-V Curve The maximum power from a solar cell is given by Pmp=Imp*Vmp (2.5) The efficiency of a cell is given by = Pmp/Pin (2.6) = Imp*Vmp/Pin (2.7) = Isc*Voc*FF/Pin (2.8) FF= Imp*Vmp/Isc*Voc (2.9) Pin is the incident power. FF is the Fill factor. 2.7.1 Effect of Series and Shunt Resistance A practical solar cell has losses due to both series and shunt resistance. The practical diode equation including the se ries and shunt resistance is given by ln((( I+IL)/IS)-(((V-IRS)/ISRSH))+1)=(q/kT)*(V-IRS) (2.10)

PAGE 31

15 2.7.1.1 Effect of Series Resistance The series resistance is manly due to ohmic losses and resistance of the bulk semiconductor. The effect of series resi stance on the I-V curve is shown in Fig 2.5. Fig 2.5 Effect of Series Resistance The series resistance for a practical cell shou ld be in the order of tenths of ohms. When RSH = and V is Zero eq(2.10) can be written as ln((( I+IL)/IS)+1)= (q/kT)*(-IRS) Therefore I=IS*exp (-qIRS/kT)-IL Hence the increase in the series resistance results in lesser short circuit currents. 2.7.1.2 Effect of Shunt Resistance The Shunt resistance is due to leakage curr ents. This results in high reverse saturation currents. From eq(2.3) higher value of I0 result in lower Voc’s. Hence the effect of shunt resistance is reduced Voc’s. Th e effect of shunt resistance on I-V characteri stics is shown in Fig 2.6.

PAGE 32

16 Fig 2.6 Effect of Shunt Resistance 2.8 Requirements of the Materials used in Solar Cells Semiconductor materials used to fabricate solar cells should be sensitive to the color of light. Certain materials absorb more light and while others absorb less. The amount of absorption depends on the absorp tion coefficients of the materials. The materials used should have good optical propertie s so that optical loss es due to reflection, scattering can be reduced. The requirements of materials used in different layers of the solar cells are discussed below. 2.8.1 Back Contact A back contact should have low resistiv ity and be able to form a good ohmic contact to the absorber layer. It should ha ve high mechanical stab ility and must have good adhesion to the substrate. SnO2: F, AZO (Al doped ZnO), ITO (Indium Tin oxide) are few TCO’s.SnO2: F is mainly used as a back contact in this work.

PAGE 33

17 2.8.2 Absorber Layer The absorber material should have a direct band gap so that light can be absorbed efficiently. Thickness of the absorber layer de pends on the absorption coefficients of the material. Bandgap of the absorber determines the output voltage of the cell. Higher bandgap results in higher outpu t voltage but lower short circ uit currents and viceversa. Materials used as absorbers should have high minority carrier lifetimes so that they can be collected before they recombine. Defect free absorber layer is required for a good performance of solar cell. It should have a good lattice match with the window layer. A large variety of inorganic, organic, crystalline, polycrystalline and amorphous materials can be used as an absorber la yer [10].During 1980’s CdSe has drawn lot of interest of many researchers as a photo-el ectrochemical cell [11].But the efficiency obtained was only 6 % [12].CdSe is a binary compound with a direct band gap of 1.7 eV. It has a high absorption coefficient of 104cm-1.CdSe has a lattice constant of 6.05 . 2.8.3 Window Layer A window layer should have a large band gap with minimum absorption and maximum transmittance. The ideal bandgap of a material to be used as a window layer should be greater than 2.5 eV so that it can transmit maximum amount of light to the absorber. There should be good lattice match between the window and the absorber layer materials.

PAGE 34

18 ZnSe has a band gap of 2.7 eV. It has 90 % transmission in visibl e region. It has a lattice constant of 5.66 . The lattice mism atch between CdSe and ZnSe is around 6.5%. 2.8.4 Front Contact A Front contact should have low reflectan ce and high transmittance. It must have low resistivity and be able to form a good ohmic contact with th e window layer. Copper is deposited as a front co ntact in all the devices.

PAGE 35

19 Chapter 3 Literature Review Most of the II-VI compounds have drawn in terests of many people in research because they find their applications in op toelectronic devices, photo electrochemical cells, thin film transistors [23], gas sens ors [24, 25], acousto-op tical devices [26], vidicones [27], photographic photoreceptors [28] and gamma ray detectors. Among II-VI compounds CdTe, HgI2, CdS, CdSe etc are prominent because of their properties like direct band gap, high absorption coeffi cients etc. CdTe, CdS and CdSe find its applications in photovoltaic devices and HgI2 finds its application in detectors. Different techniques have been used in depositing th ese materials and in fabricating devices. Close Space Sublimati on (CSS), Chemical Bath deposition (CBD), Sputtering, Thermal Evaporation, Molecula r Beam Epitaxy (MBE) and Metal Organic molecular beam epitaxy (MOMBE) are some of the vacuum deposition techniques. Lot of research has been done in past years a nd literature reflects the progress made in this field. In this section, a brief review of litera ture on CdSe for optoelectronic applications done by various research groups is mentioned.

PAGE 36

20 E.Benamar [13] et al employed a method called “cathodic electrodeposition” for depositing polycrystalline CdSe films on ITO/ glass substrates. Electrocodeposition is a low cost technique which can be used eff ectively for large area films. The process involved a potentiostatic reduction of Cd Se from acid aqueous bath. The aqueous solution had 0.2 M CdSO4 (Cadmium sulphate) or CdCl2 (Cadmium chloride) and 7 X 104 M H2SeO3 (Selenious acid) for simultaneous c odeposition of Cd and Se ions. The process involved the reduction of Cd ions a nd Se ions at the cathode thereby depositing a film on the substrate. They claim that when CdSO4 is used as a source for Cd ions, CdSe deposited had cubic structure. By using CdCl2 both cubic and hexagonal forms of CdSe was observed. Fig 3.1 XRD Pattern of CdSe Film Deposite d From an Electrolyte Containing CdCl2

PAGE 37

21 Fig 3.2(a) Fig 3.2(b) Fig 3.2 XRD Pattern of CdSe Film Deposite d from an Electrolyte Containing CdSO4 at a) T=18C b) T=75C Using CdSO4 as an electrolyte, CdSe was de posited on ITO substrates at 18 C and 75 C. From Fig 3.2(a) it can be observed that th e peak at an angle of 2 = 30 flattens when the temperature was increased from 18 C to 75 C. In Fig 3.2(a) and (b) CdSe films on ITO substrates gr ow with a preferential orient ation of (111). SEM studies showed that modular spanning of 14m or less in extent is present. From Fig 3.3 they have obtained the values of Voc as -0.7V and Isc as 1.7mA.A FF of 0.49 % and an efficiency of 1 % is obtained. They attributed the results to the kinetic properties of the interface.

PAGE 38

22 Fig 3.3 I-V Curve Another group lead by Chouhaid Nasr [ 14] et al at Ra diation laboratory, University of Notre Dome conducted comp arative studies on photoelectrochemical behavior of SnO2/CdSe and OTE/SnO2/CdSe nanocrystalline films. ITO coated glass was used as an optically transparent electrode (OTE).SnO2 was coated on OTE and was dried in air on a hot plate. The f ilms were then annealed for 1 hour at 673 K. CdSe films were deposited both on OTE and OTE/SnO2 .CdSO4 and SeO2 (Selenium dioxide) were chemicals used in the reaction .They conducte d both electrical and optical measurements which suggested that coupling of OTE/SnO2/CdSe had better performance when compared to SnO2/CdSe in terms of conversion e fficiency and stability.OTE/SnO2 electrode absorbed light be low 400 nm whereas OTE/SnO2/CdSe absorbed till 700 nm with a peak value at 470 nm. In a nanocrystalline film, there was no space charge region and hence the charge separation took place by the kinetics of reactions. Tests conducted proved that the charge separation process was more effec tive in a coupled system (OTE/SnO2/CdSe). In a

PAGE 39

23 coupled system photogenerated electrons migrate to the conduction band of SnO2 and hence prevent recombination with holes present in CdSe film. But when CdSe was deposited directly on OTE there was more reco mbination taking place before the carriers could reach the back contact. Hence it drops the photo current by a large magnitude. Fig 3.4 compares the photo-current action sp ectra of different electrodes. Typical cells showed a Voc and a Jsc of 0.550V and 27Acm-2 respectively. FF was 0.47.An efficiency of 2.25 % was achieved fo r an incident power of 0.3mWcm-2.They concluded that coupling of CdSe and SnO2 improved the performance of OTE/SnO2/CdSe rather than SnO2/CdSe alone. Fig 3.4 Photo-current Action Spectra for Different Electrodes a) OTE/SnO2 b) OTE/CdSe c) OTE/SnO2/CdSe Meera Ramarahiani [15] studied the charac teristics of Zinc doped polycrystalline CdSe films. CdSe films were first depos ited on Titanium substrates at room

PAGE 40

24 temperature from a solution of SeO2, CdSO4 and H2SO4 by a process called electrocodeposition. These films were then a nnealed at 400C for 2.5 hrs. Zn ions were then incorporated on CdSe by dipping the film in an aqueous solution of ZnSO4.This was followed by heating in air at 100C for half an hour. Incorporation of Zn ions at the surface produced favorable states in the ba nd gap which improved the charge transfer kinetics at the interface thereby reducing th e recombination process. Upon heating these films, Zn diffused through the grain boundaries and reduced the recombination centers for majority carriers. When some of the Cd atoms were replaced by Zn atoms, the bandgap of the material increased. Hence Voc was improved but Isc was reduced to some extent. This could be because of the formation of Cd1-xZnxSe.Cd1-xZnxSe has a bandgap lying between the bandgaps of both CdSe and ZnSe. This material with a higher bandgap than CdSe contributes to the increase in Voc. Photoelectrode was prepared by applying a common epoxy. Copper leads were spot welded at the back cont act for electrical connections. Deposition time was optimized to be 45 min and a film of approximately 4 m was grown on Titanium substrate. Followi ng are the parameters of a typical device. Table 3.1 Device Parameters VOC(mV) ISC(A) PMAX(Nw) FF CdSe 42.5 16.2 214 0.33 Cd1-xZnxSe 53.5 20.8 497 0.45 I-V characteristics ar e shown in Fig 3.5. They claimed that I-V characteristics of Zn incorporated CdSe film showed a positive shift and concluded that doping of Zn on the surface of n-CdSe films improved the performance.

PAGE 41

25 Fig 3.5 I-V Curve of CdSe and CdZnSe Cristian Baban et al [16] studied the st ructural and optical characteristics of thermally evaporated CdSe thin films on gla ss substrates. The influence of preparation conditions on the fundamental absorption of CdSe thin films was also studied. Polycrystalline powder of CdSe was therma lly evaporated by a quasiclosed volume technique. Thin film samples were de posited at various s ource and substrate temperatures. XRD studies reve aled that the films had a hexagonal wurtzite structure with (002) orientation. AFM te sts confirmed that the grain size was between 20 100 nm (shown in Fig 3.6). They concluded that as source and substrate temperature increases, crystal size also increases.

PAGE 42

26 Fig 3.6 AFM results for a) Ts=300 K b) AFM results for Ts=400 K The films were annealed at 200C for 30 minutes after deposition. Annealed samples showed an increase in crystal size. The transmission response of both annealed and as deposited samples was observed. They concluded that as deposited samples had lesser transmission when compared to annealed ones. They attributed this to the increase in the crystal size. This might be because as the crystal size increases, the number of grain boundaries decrease resulting in the decrease of scattering effects thereby improving the transmission. D.Pahinettam Padiyan et al [17] observed the influence of thickness and substrate temperature on electrical and photoelectrical pr operties of vacuum deposited CdSe films on glass substrates. Optical st udies of the films confirmed that the band gap decreases with an increase in the thickness of the film s and substrate temperature. CdSe powder with 99.99 % purity was taken as a source material. The films were deposited at three

PAGE 43

27 different temperaturesRoom temperature (RT), 100C and 200C. As claimed in the previous paper, this group also stated that the grain size incr eased with an increase in the substrate temperature. The band gap energies at different temperatures were observed to vary and they attributed this to th e grain size. By conducting photoelectrical measurements, they concluded that the dark a nd light currents for as deposited substrates increased with an increase in the film thickness and substr ate temperature. C.Sene et al [18] studied the effects of silicontungstic acid on CdSe films grown on polymer substrates. They employed CBD fo r depositing films on various substrates like glass, gold, ITO and an organic polymer material (PMeT (poly (3methylthiophene))).This polymer was either a p-type or a quasi metallic conductor. The focus was on organic inor ganic photovoltaic junctions. CdSe deposited by CBD on PMeT formed a p-n junction. Electropolymer ization of CBD CdSe formed a Schottkytype junction. P-N junction obtained a conve rsion efficiency of 0.03 % while the Schottky-type junction had a conversion e fficiency of 1.3 %. The presence of silicontungstic acid in the chemical bath increased the conversion efficiency of Schottkytype junction to 2.7 %. The op tical and structural propertie s of CdSe film depended on the type of substrate that was used. Phot ovoltaic properties of CdSe/doped PMeT and CdSe/ undoped PMeT junctions were studied. They stated that the use of silciontungstic acid resulted in forming hi ghly efficient junctions.

PAGE 44

28 Fig 3.7 I-V characteristics of a) undopedPMeT/CBD CdSe b) doped PMeT/CBD CdSe c) doped PMeT/CBD CdSe (in presence of Silicon Tungstic Acid) From Fig 3.7 it can be observed that undope d PMeT/CBD CdSe j unction has obtained higher voltage. Doped PMeT/CBD CdSe ha s resulted higher curr ent density when compared to undoped PMeT/ CBD CdSe while doped PMeT/CBD CdSe in presence of Silicon tungstic acid has resulted in a higher current density than the other two junctions. S.S.Kale et al [19] studied the thickne ss dependent properties of CdSe thin films deposited by chemical deposition method. The eff ect of thickness on electrical, structural and optical properties was st udied. The chemicals used in the process were 0.1 M CdSO4 solution, 0.13 M Na2SeSO3 (Sodium Selenite) solution and Ammonia. The

PAGE 45

29 deposition temperature was varied betw een 273 K and 358 K and the corresponding thicknesses of films were measured to be between 600 and 2400 respectively. They concluded that as the deposition temperature decreased from 358 K to 273 K, thickness decreased from 2400 to 600 a nd grain size decreased from 80 to 40 . As the temperature was increased, disassociati on of the complex and the anion, the rate of release of selenium and thickness of the film s increased. Larger grains were obtained. From Fig 3.8 it can be noticed that there is a linear variation of band gap with the deposition temperature. Band gap increased from 1.90 to 2.4 eV as the deposition temperature was decreased from 358 K to 273 K. They attri buted this to quantum size effect. Fig 3.8 Variation of Band gap Vs Deposition Temperature When the electrical properties were studi ed they observed that the resistivity increased from 103 to 104 cm as the deposition temper ature decreased from 358 K to 273 K. In general the resistivity of a na nocrystalline material is more than a

PAGE 46

30 polycrystalline material. Hence the high resist ivity at low depositi on temperatures could be because of its smaller gr ain size. The variation in resistivity with deposition temperatures is shown below. Fig 3.9 Variation of Resistivity Vs Deposition Temperature R.M.Abdel-Latif at University of Minia [20], Egypt studied a bout the electrical properties of evaporated CdSe thin film s. Aluminum with high purity (99.99%) was evaporated from a tungsten filament onto a cleaned glass substrate through a mask to form a base electrode. Pure CdSe was de posited through a mask from a molybdenum boat. A top electrode of Au was deposited over this. The substrate was maintained at room temperature.

PAGE 47

31 Fig 3.10 J -V characteristics of CdSe thin film s with temperature as variable a) with bias voltage < 0.6V b) with bias voltage > 0.6V The J-V characteristics of the devices are shown in Fig 3.10. As the deposition temperature was increased, there is an increase in the variation of the current density with the voltage. They have conclude d that for Al-CdSe-Au thin films at low voltages, current varies exponentially with voltage. They have further stated that the films had an electron concentration of 1.1 X 1018 cm-3.

PAGE 48

32 Chapter 4 Device Structure and Fabrication 4.1 Tandem Cell Structure A tandem solar cell usually has two cells having different band gaps placed one over other with an encapsulan t between them. They are designed in such a way that the top cell has a larger band gap and hence can absorb lower wavelengths. The bottom cell with a relatively smaller band gap can abso rb higher wavelengths that pass through the top cell. A 4 terminal tandem so lar cell structure is shown below Fig 4.1 Tandem Solar Cell Structure Glass Top Window layer CdSe or CdZnTe Transparent conductor Encapsulant Top window layer CIGS Mo Glass

PAGE 49

33 To obtain efficiencies greater than 25 %, the top cell should ha ve a bandgap of 1.7 eV. This would mean that the top cell has to co ntribute to 2/3 rd of total efficiency. The bottom cell of the tandem structure having a bandgap of 1 eV has already been standardized. CIGS solar cells with a bandgap of 1 eV have attained an efficiency of 18 %. Though there are a few of I-III-VI2 compounds like CuGaSe2 with a band gap of 1.7 eV and CuInS2 with a bandgap of 1.55 eV, they cannot be used for the top cell because of their relatively lower efficiencies [21, 30]. Other materials being investigated as viable candidates for the top cell are Cd1-xZnxTe and CdSe. Cd1-xZnxTe is a ternary alloy with a tunable bandgap. CdSe is also foun d to be another suitable material to serve as the absorber layer in the top cell [3]. P.Gash in et al fabricated a 3 terminal monolithic tandem structure of n-ZnSe/p-ZnTe/n-CdSe and reported an efficiency of 10.8 % [22].CdSe can be processed with lesser co mplexity when compared to any I-III-VI2 compound. This thesis focuses on the developmen t of a top cell with CdSe as an absorber layer. The top cell of the proposed tandem sola r cell should have a s uperstrate structure. But for initial experiments we starte d of with a substrate structure.

PAGE 50

34 4.2 Structure of a CdSe Based Top Cell The structure of a CdSe based top cell is shown below. Fig 4.2 Structure of CdSe Based Top Cell 4.2.1 Substrate In our process we use Corning 7059 glass su bstrates because of their stability to withstand high temperature processing above 70 0C.Typical dimensions of the substrate are 1.25 X 1.35 X 0.05 inches. 4.2.1.1 Cleaning Procedure Before any deposition process, the s ubstrate goes through a regular cleaning procedure. This is a very important step in processing. Any contaminants present on the substrate could affect the subsequent depos ition steps and the device performance. The substrate is first dipped in dilute HF (1 part of HF in 10 parts of water) for 5 seconds, rinsed under a jet of DI water. It is again di pped in dilute HF for 3 seconds and is finally rinsed under a jet of DI water. It is then blown dry with nitrogen. Cu Cu Cu Cu ZnSe CdSe SnO2: F 7059 glass In

PAGE 51

35 4.2.2 Back Contact A good back contact should be transpar ent, conductive and have the proper contact energy. SnO2: F and AZO are the two TCO’s used as back contacts. SnO2: F has a band gap of 3.5 eV. It is a very good TCO because of its high transmission of 90 % and low sheet resi stance of 7-10 ohms/sq. The transmission response of SnO2: F is shown in Fig 4.3. Trasnsmission Response of SnO2:F0 10 20 30 40 50 60 70 80 90 100 300400500600700800900Wavelength(nm)% Transmission SnO2:F Fig 4.3 Transmission Response of SnO2: F

PAGE 52

36 4.2.2.1 Deposition of SnO2: F SnO2: F is deposited by MOCVD.TMT (Tetra Methyl Tin) is used as the source for Tin. Halocarbon 13B1 is used as the source fo r Fluorine which acts as the dopant. The substrate is heated to 470 C. For th e first 8 minutes-TMT, Halocarbon 13B1 and oxygen are reacted to deposit a doped layer of SnO2. This layer is n-type. For the next 5 minutes, the substrates are annealed in the presence of oxygen. For the last 5 minutes an undoped layer of SnO2 is deposited. In order to overcome problems due to diffusion an undoped layer of SnO2 is deposited. The thickness of the film is approximately 800 . 4.2.2.2 AZO as a Back Contact AZO with a bandgap of 3.3 eV has also been used as a back contact. The following properties of AZO make it a viable back contact. 1) Raw materials required are cheap and abundant. 2) They have a very high transmittance reaching close to 90 % in the visible region. 3) They are readily produced for large scale coatings. 4) They have high stability in hydrogen plasma. 5) They have low growth temperature. 6) They have a very low resistivites of order 1.4 X 10-4 -cm. The conductivity of ZnO is primarily dom inated by electrons generated from oxygen vacancies and Zn interstitial atoms. The electrical conductiv ity in AZO films is more than

PAGE 53

37 ZnO films due to the added contribution from Al+3 ions on substitutional sites. ZnO films have polycrystalline hexagonal wurtz ite structure as shown in Fig 4.4. Fig 4.4 Wurtzite Structure of Zinc Oxide 4.2.2.3 Structure of Cell with AZO at the Back Contact Fig 4.5 Structure of Cell with AZO at the Back Contact Aluminum doped Zinc oxide is deposited on Corning 7059 glass substrates. CdSe is deposited on AZO by close space sublimation in the chamber shown in Fig 4.11. ZnSe and Copper are deposited by Thermal Evaporatio n in the chamber shown in Fig 4.14. In Cu Cu Cu Cu ZnSe CdSe AZO 7059 glass

PAGE 54

38 4.2.2.4 Deposition Procedure for AZO AZO films are deposited by R.F Magnetr on Sputtering. A cleaned 7059 glass substrate is loaded in the chamber. The chamber is pumped to low micro Torr range. The substrate is heated to 125 C through a speci fic heating profile by a variac. A ZnO (2 % wt Al2O3) target is sputtered in the presence of argon. To begin with, 2000 thick films were deposited. Later on the thickness of the films was increased to 5000 . Transmission of about 90 % with a resistivity of 4 X 10-4 -cm has been achieved. 4.2.2.5 Transmission of AZO The transmission response of AZO film s having a thickness of approximately 5000 is shown in Fig 4.6.We can observe that in the wavelength region between 450 and 900 nm there is a high transmission of about 90 %.Because of its good optical properties, AZO is a good choice for the TCO layer. Fig 4.6 Transmission Response of AZO

PAGE 55

39 4.2.3 Absorber Layer Due to its high electronic quality and a fixed optical bandgap of 1.7 eV, Cadmium Selenide is chosen as one of th e materials that can be used as an absorber layer in the top cell. It belongs to II-VI group of the periodic ta ble. Cd has 2 electrons in its outer most orbit while Se has 6 electrons. Every Cd at om transfers its 2 electrons to 6 valence electrons of Se to form CdSe. A vacancy of a Selenium atom frees two electrons of Cadmium making CdSe n-type. 4.7 Wurtzite structure of CdSe CdSe crystallizes either in wurtzite (Hex agonal) or cubic (Zin c blende) structures. The wurtzite structure of CdSe is shown in the Fig 4.7.CdSe deposited as a part of this work has Hexagonal structure which was demonstrated from the AFM image shown in Fig 4.8.

PAGE 56

40 Fig 4.8 AFM image of CdSe From the AFM image we can notice the he xagonal planar structure of the CdSe film. The dark regions in the image are gr ain boundaries. Usually, th e larger the grain size the better the quality of the films. Larger grain sizes also result in the improvement of the transport propert ies of minority charge carriers.

PAGE 57

41 4.2.3.1 Properties of CdSe Table 4.1 Properties of CdSe Bandgap 1.74 eV Electron Affinity 4.56 eV Lattice constant 6.05 4.2.3.2 Preparation of the CdSe Source High purity (99.999%) CdSe powder is used in the preparation of the source. Enough care is taken so that there are no cont aminants present that could affect the device performance. A circular disk of 2 mm th ick and 1 inch in diameter is prepared by pressing CdSe powder in a circular mold. 4.2.3.3 Deposition Procedure for CdSe CdSe can be deposited by various met hods like Thermal evaporation, Chemical vapor deposition, Sputtering, CBD, Electrodepos ition and CSS. CSS is not only a simpler process but can also produce films of good electronic proper ties. The deposition rate is high so that 1-2 m thick films can be deposited in a short time. The source is kept on a graphite holder and is separated from the s ubstrate by quartz spacers. Each of the two quartz lamps is fixed to a reflector. Lamps are placed one over the other so that most of the heat is localized. The setup is shown in Fig 4.9. The reactor tube is pumped to less than 1 Torr by a mechanical pump and is purged with Helium for couple of times. Fina lly it is back filled with Helium to a

PAGE 58

42 pressure of 3 Torr. The source and the substr ate temperatures are set using a temperature controller. Both the lamps are turned ON. Wh en source and substrate temperatures reach their set values, sublimation begins and the time for the sublimation is noted. Both the lamps are turned OFF as soon as the deposition time is over. Fig 4.9 Closed Space Sublimation Setup 4.2.3.4 Parameters CdSe was deposited under different conditions. The quality of the films was optimized by varying parameters like pr essure, source temperature, substrate temperatures and spacing. The optimized parameters are: Tsou-670C Tsub-560C Ambience used-He Pressure-3 Torr Source Lamp Gas lines Thermo Couple Substrate Lamp

PAGE 59

43 tdeposition 14 minutes. The thickness of the film obtained for the above conditions is around 1.5 microns. 4.2.3.5 Transmission of CdSe The transmission response of CdSe films deposited under standard conditions is shown in Fig 4.10.From the transmission respon se measurements; we can notice that the films have approximately 80 % transmittan ce between 750 and 900 nm. CdSe absorbs high energy photons and transmits the low ener gy photons to the bottom cell. The cutoff wavelength is at 725 nm which corresponds to the band gap of CdSe. Fig 4.10 Transmission Response of CdSe 4.2.3.6 Experiments Performed on the Absorber Layer The source is sublimated at different te mperatures from 630 C to 670 C. The films obtained at 630 C were not uniform but as the source temperature is increased to 670 C, uniformity of the films improved. This could be because at high temperature

PAGE 60

44 the source gets the required energy to disassoci ate into fine particles and then sublimate into a uniform layer of CdSe. Similarly the substrate temperature was ma intained at 560 C for standard runs. Under standard conditions the film thic kness was about 1.5 m. Further experiments were performed by elevating the substrate te mperature in steps of 10 C starting from 560 C. Substrates at higher temperatures have hi gher energies associated with them, and therefore atoms from the source do not tend to stick. Hence the thickness of the film decreased with the increase in the substrate temperature. Sticking of the material depends on the type of material that is being deposited and the type of substrate on which the film is growing. The coefficient of sticking varies from material to material. 4.2.4 Window Layer In order to transmit the maxi mum amount of light to the ab sorber layer, we need a window layer with a high bandgap. Zinc Seleni de with a band gap of 2.7 eV can be used as a window layer. It is a II-VI semiconduc tor widely used in many opto-electronic applications. The crystal stru cture of ZnSe is shown in Fig 4.11.It crystallizes into a Zincblende structure.

PAGE 61

45 Fig 4.11 Structure of ZnSe 4.2.4.1 Properties of ZnSe Table 4.2 Properties of ZnSe Bandgap 2.7 eV Electron Affinity 4.1 eV Lattice constant 5.668 4.2.4.2 Deposition Procedure for ZnSe High purity (99.999%) ZnSe powde r is used as the sour ce which is thermally evaporated. A specific voltage-time profile is followed for heating the source. ZnSe films are deposited at two different substrate temperaturesat room temperature and at 250 C. The devices fabricated with Zn Se deposited at 250 C did not show good performance whereas the devices fabricated w ith ZnSe at room temperature have shown good performance. All the films in this work are grown at room temperature.

PAGE 62

46 4.2.4.3 Evaporation System Fig 4.12 Evaporation System The Evaporation system shown in Fig 4.12 is equipped with a diffusion pump and a Mechanical pump to evacuate the chamber to the micro Torr range. A cold trap separates the chamber and the diffusion pump. Liquid Nitrogen is filled in regular intervals to trap moisture. There are two Molybdenum boats-one for ZnSe and the other for Copper. These boats are resistively heated.

PAGE 63

47 4.2.4.4 Transmission of ZnSe The transmission response of ZnSe is s hown in Fig 4.13.The thickness of the film is 200 . There is a significant absorption in the blue region of the spectrum which can be clearly observed. We can notice from Fig 4.13 that ZnSe has approximately 90 % transmission in the visible region which ma kes it a good choice for the window layer. Fig 4.13 Transmission Response of ZnSe 4.2.5 Front Contact The material used as the front contact must be conduc tive, transparent and make an ohmic contact with ZnSe. Copper is deposited as front contact by Thermal Evaporation in the chamber shown in Fi g 4.13.Copper pellets (99.999% pure) are added to the front boat in the chamber. A mask of 0.1 cm2 area dots is used for depositing copper on the ZnSe film. The chamber is initially pumped to the micro Torr range. The source is heated using a variac fo llowing a specific vo ltage-time profile.

PAGE 64

48 4.2.5.1 Transmission of Copper The transmission response of copper is shown in Fig 4.14.The thickness of the film is 60 . Since Copper is a metal, most of the light incident on the surface will be reflected if the film is too thick. So at lower wavelengths most of the incident light is reflected, and hence the transmittance is less. Fig 4.14 Transmission Response of Copper 4.3 Deposition of Cadmium Sulphide CdS is a direct band gap material. CdS film s have a wurtzite st ructure as shown in Fig 4.15 and has a band gap of 2.4 eV. CdS can be deposited by different methods like close space sublimation, chemical bath deposition and sputtering. CBD is popular among the three since it gives uniform films. The process is simple and inexpe nsive compared to an expensive vacuum deposition.

PAGE 65

49 Fig 4.15 Structure of CdS For the results presented below, CdS is deposited by CBD. Thiourea, Ammonium Hydroxide and Cadmium acetate are chemicals used in the reaction. Cadmium acetate is used as a source for cadmium. Thiourea is used as a source for Sulphur. The growth rate of CdS depends on deposition temperat ure and the pH of the solution. The precipitation temperature is around 87 C. For a deposition time of 80 minutes, a thickness of around 1000-1200 is deposited on the substrate. CdS of thickness around 300500 is deposited on SnO2: F to fabricate the same structure as shown in Fig 4.17.

PAGE 66

50 4.3.1 Transmission of CdS 0 10 20 30 40 50 60 70 80 90 100 350425500575650725800875 Wavelength(nm)% Transmission CdS(500 A) CdS(1000 A) Fig 4.16 Transmission Response of CdS CdS of thickness around 500 and 1000 is deposited by CBD on SnO2: F substrates. SnO2: F is used as reference for the transmission measurements. From Fig 4.16, we can notice that the transmission of thin CdS is above 90 % whereas the thick ones have a transmission of 85 %.This lo ss is due to the absorption in CdS.

PAGE 67

51 4.3.2 Structure of Cell with CdSe on CdS/SnO2: F Fig 4.17 Structure of Cell with CdSe on CdS/SnO2: F A uniform film of CdS is deposited by CBD. When SnO2: F substrates are placed in the chemical bath during CBD, CdS is depos ited on both sides of the substrate. CdS on the other side of SnO2: F is chemically etched by a cott on swab dipped in dilute HCl. A narrow strip of CdS is etched on the SnO2: F in order to put an Indium contact. CdSe In Cu Cu Cu Cu ZnSe 7059 glass SnO2: F CdS

PAGE 68

52 Chapter 5 Results and Discussions 5.1 Results Devices fabricated are measured to study their responses in light and dark illumination conditions. 5.1.1 AZO as a Back Contact Devices with AZO as a back contact ar e fabricated and anne aled at 100 C for 10 minutes. The performance of the devices befo re and after anneali ng is shown in Table 5.1. The J-V response for the device s h18-g before and after annealing is shown in Fig 5.1(a) and 5.1(b) respectively. The effect of annealing on the J-V response of the devices can be clearly seen. Table 5.1 Device Parameters for Devices with AZO as a Back Contact Sample # Voc(V) FF Voc(V) FF Jsc (mA/cm2) sh18-g 0.210 30 0.370 50.2 5.19 From the J-V curve we can notice that the Voc and FF have increased after annealing the sample. From the dark J-V curve we can notice that the devices turn on Before Annealing After Annealing

PAGE 69

53 slowly after annealing. The devices had lowe r shunt resistance before annealing. After annealing the shunt resistance has improved but the series resistance remained almost the same. The low shunt resistance before annealin g could be due to high leakage currents. Annealing could have improved the inte rface between CdSe/ZnSe which resulted in increased Voc and FF. It was observed that annealing of AZO films at high temperatures has resulted in increased resi stivity. As the deposit ion of CdSe involves high temperature processing, this could have deteriorated the prope rties of AZO films resulting in high series re sistance of the devices. (a) (b) Fig 5.1(a) and (b) J-V Curve for Device # sh18-g Sh18(7059/AZO-5000 /CdSe-670 C,560 C,3 T He-14 min/ZnSe-380 /Cu-40 )

PAGE 70

54 The spectral response of sh18-g is shown below in Fig 5.2. Fig 5.2 Spectral Response for Device # sh18-g From Fig 5.2 we can observe that the Q.E is less than 30 %.Though in the wavelength region between 500 nm and 700 nm Q.E is flat, there is a considerable absorption in the blue region. This is beca use of the absorption by ZnSe. The response cutsoff at 725 nm corresponds to the fundame ntal band gap of CdSe. As the Q.E’s are approximately around 30 %, Jsc’s are relatively low. 5.1.2 CdS on SnO2: F at the Back Contact CdS is deposited on SnO2: F. n+ CdS makes the other side of the junction more n type. This increases the band bending which results in improved ope n-circuit voltages. Table 5.2 shows the results of the cells with Cd S of thickness 500 and 1000 deposited on SnO2:F.Devices were annealed in air at 100 C for 10 minutes. Devices with thinner CdS show poor perf ormance before annealing. This can be because thinner CdS may not have uniform ly covered the rough surface of the back contact. After annealing Voc and FF have improved. This could be because surface

PAGE 71

55 properties might have improved resultin g in a better interface between SnO2: F/CdS. Annealing could also have improved the junction between CdSe/ZnSe by reducing the interface traps. Table 5.2 Device Parameters for Devices with CdS on SnO2: F at Back Contact J-V response for the device fabricated with 500 thick CdS is shown in Fig 5.3(a) and Fig 5.3(b). (a) (b) Fig 5.3(a) and (b) J-V Curve for Device # sh39-1-a Devices with thicker CdS have shown be tter performance even before annealing. Thicker CdS may have uniformly covered th e rough back contact. This makes the other Before Annealing After Annealing Sample Thickness of CdS in ) Voc(V) FF Voc(V) FF Jsc(mA/cm2) Sh39-1-a 500 0. 270 18 0.320 50.2 7.52 Sh40-1-e 1000 0.330 55.1 0.350 47.5 7.21 Sh39-1-a(7059/SnO2: F800 / CdS-500 /CdSe-670C,560C,3T He-14min/ZnSe-200 /Cu-30 )

PAGE 72

56 layers grow more uniformly increasing the junction and surface properties. Annealing has not shown a considerable improvement in the performance. The increase in Voc after annealing could be due to the passivation of interface traps. This could have reduced the leakage currents. J-V response for the device with 1000 CdS is shown in Fig 5.4(a) and Fig 5.4(b). (a) ( b) Fig 5.4 (a) and (b) J-V Curve for Device # sh40-1-e From the J-V curve we can also confir m that the Fill factors are comparable before and after annealing. Spectral response of both the devices is shown in Fig 5.5. sh40-1-e Dark J-V curve-0.02 -0.02 -0.01 -0.01 0.00 0.01 0.01 0.02 0.02 -0.50-0.30-0.100.100.300.500.70Voltage [Volts]Current Density [A/cm2] sh40-1-e sh40-1-e-a sh40-1-e-Light J-V curve-0.02 -0.02 -0.01 -0.01 0.00 0.01 0.01 0.02 0.02 -0.50-0.30-0.100.100.300.500.70Voltage [Volts]Current Density [A/cm2] sh40-1-e sh40-1-e-a Sh40-1-e(7059/SnO2-800 /CdS-1000 /CdSe-670 C,560 C,3 T He-14 min/ZnSe-200 /Cu-30 )

PAGE 73

57 sh39-1-a (7059/SnO2:F-800 A/CdS-500 A/CdSe-670C,560C,3T He-14min/ZnSe-200 /Cu-30A) sh40-1-e(7059/SnO2:F-800 A/CdS-1000 A/CdSe-670C,560C,3T He-14min/ZnSe-200 A/Cu-30A) 0 10 20 30 40 50 60 400450500550600650700750800850900 Wavelength(nm)Q.E (%) sh39-1-a sh40-1-e Fig 5.5 Spectral Response for Device # sh39-1-a and # sh40-1-e-a For the device with CdS thickness around 500 in the shorter wavelengths Q.E is around 35 %.The device with CdS thickness around 1000 has collection problems and hence has lower Q.E in higher wavelengths. 5.1.3 Absorber Layer at Higher Substrate Temperatures During the J-V measurements when the light is incident through SnO2: F none of the cells showed any response. So to allow more light to pass through the back contact and reach the absorber layer, the thickness of the absorber layer has been decreased. To make films thin, the substrate temperature was raised from 560C. This section shows the experiments performed on the absorber la yer by sublimating at higher substrate

PAGE 74

58 temperatures. But the devices when measured for light J-V did not show any response when light is incident through SnO2:F. So their response to J-V measurements when light is incident through Copper is studied. Voc’s and Jsc’s are given in Table 5.3 Thickness of the ZnSe film is 200 . Table 5.3 Effect of Substrate Temperature on Device Performance with 200 ZnSe Sample Tsub( C) Voc(V) FF Jsc(mA/cm2) sh52-1-f 560 0.130 35.2 9.48 sh48-2-l 570 0.150 16.5 6.27 sh44-1-f 580 0.270 33.3 5.42 sh45-1-e 590 0.160 48.6 8.58 Light J-V curves for the above substrates are shown in Fig 5.6. Light J-V curves-0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05 -0.25-0.100.05 Voltage [Volts]Current Density [A/cm2] sh52-1-f sh48-2-l sh44-1-f sh45-1-e Fig 5.6 J-V Curve for the Devices # sh44-1-f, # sh45-1-e, # sh48-2-l and sh52-1-l

PAGE 75

59 From the J-V plots, it can be seen that these devices had hi gher series and low shunt resistances. As the substrate temperat ure is increased from 560 C to 590 C, VOC’s have improved with the highest VOC around 270 mV for a substr ate temperature of 580 C in sample sh44-1-f. On the other hand, JSC is highest for sample sh52-1-f of 9.48 mA/cm2. At 590 C, the current density is reason able. At 590 C Voc’ s tend to fall. All the above devices have low fill factors because of their series and shunt resistances. But the above devices did not follow any partic ular trend in their performance. As a part of further experimentation, th e thickness of ZnSe is increased to 220 to observe its effect on device performance. Thickness of ZnSe film is 220 . Table 5.4 Effect of Substrate Temperature on Device Performance with 220 ZnSe Sample Tsub(C) Voc (V) FF Jsc (mA/cm2) sh49-2-f 580 0.190 50.0 8.4 sh50-2-e 590 0.230 72.4 9.11 sh53-1-l 610 0.110 37.9 7.48 The high fill factor in sh492-f and sh50-2-e is due to th e sharp turn in light J-V shown in Fig 5.7. In a MIS type structure, Zn Se acts as an insulating layer. When copper is deposited on ZnSe there is a chance of diffusing into the insulating layer which might lead to formation of homogenous p-type materi al [30]. This could re sult in higher built-in potential hence improving the open-circuit voltages. But if the amount of copper

PAGE 76

60 deposited is not sufficient to make a p-type contact to the absorber layer, or if the thickness of ZnSe is not in proportion with the thickness of copper then there cannot be any significant improvement in Voc’s. As the substrate temperature is increase d, the structural properties of the films may change. Generally it is expected that the films obtained at higher substrate temperatures are more crystalline in nature and hence have better properties [16]. But on the other hand if the temperature is increased further, films may become amorphous affecting the photovoltaic prope rties of the devices. (a) (b) Fig 5.7 (a) and (b) J-V Curve for Devices # sh49-2-f, # sh50-2-e and # sh53-1-l Dark J-V curves-0.02 -0.50-0.35-0.20-0.050.100.250.40 Voltage [Volts]Current Density [A/cm2] sh49-2-f sh50-2-e sh53-1-l Light J-V curves-0.1 -0.50-0.35-0.20-0.050.100.25 Voltage [Volts]Current Density [A/cm2] sh49-2-f sh50-2-e sh53-1-l

PAGE 77

61 Table 5.5 Effect of Substrate Temperature on Device Performance with 180 ZnSe Sample # Tsub(C) Voc (in V) FF Jsc (mA/cm2) sh49-3-b 560 0.220 33.8 10.02 sh51-2-c 600 0.220 71.8 5.76 Both these devices had Voc’s of around 220 mV. From the light J-V of sh49-3-b in Fig 5.8, it can be observed that the series resistances are high and shunt resistances are low. This device has obtained a highest Jsc of 10.02 mA/cm2. If the shunt resistance values can be increased, Jsc’s may be improved further.sh51-2-c has low Jsc but Voc’s are comparable. The spectral response of devices sh50-2-e, sh49-3-b a nd sh52-1-f are shown in Fig 5.9 (a) (b) Fig 5.8 (a) and (b) J-V Curve for the Devices # sh49-3-b and # sh51-2-c Dark J-V curve-0.02 -0.01 0.00 0.01 0.02 -0.50-0.30-0.100.100.300.50 Voltage [Volts]Current Density [A/cm2] sh49-3-b sh51-2-c Light J-V curve-0.05 -0.03 -0.01 0.01 0.03 0.05 0.07 -0.20-0.100.000.100.200.300.40 Voltage [Volts]Current Density [A/cm2] sh49-3-b sh51-2-c

PAGE 78

62 sh50-2-e( 7059/SnO2:F-800 A/CdSe-670 C,590 C,3 T He-14 min/ZnSe-220 A/Cu-30 A) sh49-3-b (7059/ SnO2:F -800 A/CdSe-670 C,560 C,3 T He-14 min/ZnSe-180 A/Cu-30A) sh52-1-f (7059/ SnO2:F -800 A/CdSe-670 C,560 C,3 T He-14 min/ZnSe-200 A/Cu-30 A)0 10 20 30 40 50 60 400450500550600650700750800850900 Wavelength(nm)Q.E(%) sh50-2-e sh49-3-b sh52-1-f Fig 5.9 Spectral Response for Devices # sh50-2-e, #sh49-3-b and # sh52-1-f The Q.E of sh50-2-e is lower in comparis on with the other two devices. This is because of the absorption in the blue region. This device has higher absorption in the blue region due to thicker ZnSe. It is observed that the thickness of copper and ZnSe is crucial for the performance of these devices. If the th ickness of copper is more than the standard thickness, then there is a chance of copper diffusing thr ough the junction and shorting the devices. If the thickness of copper is less, it was observed that the open circuit voltages obtained was less. If the optimum thickne ss of copper is deposited on ZnSe, it can enhance the photo-voltaic properties of the devices. Hence optimization of the thickness of copper is important .We have varied th e thickness of copper from 60 to 30 .

PAGE 79

63 The highest Q.E of around 55 % is obtained for sh49-3-b at mid wavelengths. The carrier collection in the mid wa velengths is good resulting in higher Jsc’s compared to the remaining devices. 5.1.4 Effect of Time on Device Performance It was observed that the devices fabr icated have shown better photovoltaic properties if they are re-measured after a c ouple of days. One of the devices exhibiting this behavior is shown in Table 5.6.In the de vice sh44-1-f, improveme nt in Voc is seen. This might be because of Copper. If the devices are left for a couple of days after they are completely fabricated, Copper might diffuse into the shallow areas of the insulating layer making ZnSe a p-type contact to the absorber resulting in the improvement of Voc. It was also observed that the copper dots spread. Devices showed better performance after sp reading. There could be some chemical reaction taking place in the surface which might aid in improving the device parameters. The J-V response for the above device is shown in Fig 5.13(a) and (b). Table 5.6 Effect of Time on Device Performance Sample # Voc (in V) Voc (over time in V) FF Jsc(mA/cm2) sh44-1-f 0.140 0.270 31.7 5.42

PAGE 80

64 (a) (b) Fig 5.10(a) and (b) J-V Curve for Device # sh44-1-f The initial measurements show that the seri es and shunt resistances in the dark are better. The later measurements showed a lot of shunting under illumination. Under illumination, there is no significant change in the series resistance. The spectral response of the device sh44-1-f measured after it is left over a period of time is shown in Fig 5.13. sh44-1-f-Dark J-V curve-0.02 0.01 -0.50-0.30-0.100.100.300.500.70 Voltage [Volts]Current Density [A/cm2] sh44-1-f sh44-1f(over time) sh44-1-f-Light J-V curve-0.05 -0.03 0.00 0.03 0.05 -0.40-0.200.000.200.40 Voltage [Volts]Current Density [A/cm2] sh44-1-f sh44-1f(over time) Sh44-1-f (7059/SnO2:F-800 A/CdSe-670 C,580 C,3T He-14 min/ZnSe-200 A/Cu-30 A)

PAGE 81

65 Fig 5.11 Spectral Response for Device # sh44-1-f Q.E is less in the shorter wavelengths and it tends to rise as it reaches the mid wavelengths with the highest Q.E of approxima tely 28 % which is still a lesser value in comparison to the other devices.

PAGE 82

66 Chapter 6 Conclusions As discussed in the first chapter, the goal of achieving efficiencies greater than 25-30 % in a solar cell is possible only by fabricating a Tandem cell. A 4-terminal tandem solar cell has two cells of suitable band gaps placed one over the other with a proper encapsulant between them. According to the initial investigation, it was chosen that low band gap CIGS suits well as an absorber layer in the bottom cell. Significant efforts have been put in finding the suitable ab sorber for the top cell. In order to attain high efficiencies the band gap of the mate rial should be between 1.5 eV and 2.0 eV. Few of the I-III-VI2 compounds meet the requirement of having a band gap lying between 1.5 eV and 2.0 eV [30]. But th ese devices have not shown promising results. Further investigation resulted in opting for CdSe and Cd1-xZnxTe as a choice of absorber layer in the top cell. The band gap of CdSe is 1.7 eV while Cd1-xZnxTe has a tunable band gap. As a part of this study, the possibility of using CdSe as an absorber layer in the top cell of the tandem solar cell has been inve stigated. The experiments performed partly confirmed the possible application of having CdSe as an attractive II-VI compound for solar cell applications. The various experiment s performed showed that the absorber layer is deposited to suit the high temperature proc essing without causing a ny adverse effect to

PAGE 83

67 the back contact.SnO2: F forms a good choice for the back contact with lower sheet resistance. Even after undergoing a subseque nt high temperature processing during the deposition of CdSe, the sheet resistance valu es have not increased by a large magnitude. Hence SnO2: F is chosen as a back contact in standard CdSe solar cells. The experimentation done has confirmed that n-type CdSe absorber layers have good electronic and structural properties. It was also observed that the deposition parameters like substrate temperature, sour ce temperature, spacing, pressure in the reactor tube or the ambience largely determine the growth and the properties of CdSe films. Ideally in a MIS structure, ZnSe should act as an insulating layer. It has good optical and physical properties which match with the absorb er. Copper is deposited as a metal contact to ZnSe. There are some potentia l problems with this metal contact. Copper diffuses into deep regions near the junction and thereby shor ting the devices. On the other hand it is suspected that depos ition of an optimum amount of copper results in diffusion into the shallow regions of ZnSe and ther eby forming a homogenous material Cu/ZnSe: Cu enhancing the device performance by maki ng ZnSe p-type [29]. This behavior of Cu/ZnSe junction shows a deviation from MIS type devices. Therefore the amount of copper present is an important concern. Since it is a shiny metal, it has a high reflectivity and thereby loosing some of the incide nt photons due to losses in reflection.

PAGE 84

68 Hence the photo-currents ge nerated are less. If copper can be replaced by a less reflective ptype metal cont act to CdSe, then the chance of obtaining high photo-currents at the output is more. Under standard deposition conditions, th e thickness of the absorber layer is around 1.5 microns. It is desirable to have a comp letely depleted absorber so that most of the carriers generated with in the diffusion length of the semiconductor can be collected before they recombine. In order to achieve a completely depleted absorber layer, the thickness of the absorber laye r has to be decreased to 0.5 0.7 microns without adversely affecting the device output parameters. Further examinatio n in obtaining a completely depleted absorber layer has to be achieved. Further experimentation has to be done to optimize the parameters and the thickness of th e subsequent layers in achieving the same. The Metal Insulating Semiconductor (MIS) st ructure can be replaced if we can dope the insulating layer to p-type and th ereby forming a p-n junction between the window and the absorber layer. There should be minimum absorption loss in the layers above the absorber to enhan ce the device output parameters A transparent front contact has to be deposited to replace copper and thereb y we can overcome the losses at the front contact to a considerable extent. In some cases, it was noticed that there is a variation in the device parameters from one device to the other device on the same substrate.

PAGE 85

69 This might be due to the difference in the thickness of the different layers deposited. This issue has to be taken into consideration. In general, for the same conditions the devices showed good re peatability and reproducibility. Most of the devices have generated reasonable Voc’s and Jsc’s. The highest values achieved so far for Voc and Jsc are 220 mV and 10.02 mA/cm2. CdSe when deposited on CdS/SnO2: F has resulted in a Voc of around 350 mV, Jsc of 7.21 mA/cm2 and a fill factor of 47.5 %.

PAGE 86

70 References [1] D.Tarrant et al,"I-III-VI Mul tinary solar cells based on CuInSe2”, proceedings of the 23rd IEEE photovoltaic specialists conference, 372, (IEEE 1993). [2] M.Nishitani et al,”Photovolta ic properties of Cu (In, Ga)Se2 thin film solar cell fabricated by coevaporation pr ocess”, proceedings of the 24th IEEE photovoltaic specialists confer ence, 222, (IEEE 1994). [3] P.Mahawela, S.Jeedigunta, C.S.Ferekide s and D.L.Morel,”Dev elopment of II-VI high bandgap devices for high efficiency tandem solar cells”. [4] D.M.Chapin, C.S.Fuller and G.L.Pearson,” A new silicon p-n junction photo-cell for converting solar radiation into el ectrical power”, J.Appl.Phys, 25,676(1954). [5] NREL Research photovoltaic research -“Creating electricity from an unlimited resource-sunlight”. [6] National Renewable Ener gy Laboratories, “Photovoltaic s program plan FY 1991FY1995”, National photvoltaics program, US Department of energy, Washington, DC (October 1999). [7] Adolf Goetberger et al,” Photovoltaic materials, histor y, status and outlook”, August 2002. [8] M.A.Green,” Solar cellsoperating principles, Technology and system application”, Prenice Hall, Engle Wood Cliffs, NJ (1982). [9] S.M.Sze, "Physics of semiconducto r devices”, Wiley, New York (1981). [10] Stephen J.Fonash,” Solar cell device physics”. [11] M.Russak, J.Eichman, H.Witzke, S.Deb a nd S.Chen, J.Electroch em.Soc:Solid-state science and technology, Vol 127,725(1980). [12] D.Bonnet and E.Rickus, Proceedings of the 14th IEEE PV Specialist Conference, 629(1980).

PAGE 87

71 [13] E Benmar, M.Rami, M.Fahoume, F.Chraib i, A.Ennaoui,” Elect rodeposited cadmium Selenide films for solar cells”, Ann.Chim.Sci.Mat, 1998, 23, pp.369-372. [14] Chouhaid Nasr, Prashant V.Kamat, Su rat Hotchandani,” Photo-electrochemical behavior of couples SnO2/CdSe nano crystalline semiconductor films”, Journal of Electroanalytical Chemistry 420(1997)201-207. [15] Meera Ramrakhiani,” Zinc doped pol ycrystalline CdSe film for solar energy conversion”, Material Science and Engineering B35 (1995)493-496. [16] Cristian Baban, G.I.Rusu,” On the structur al and optical characteristics of CdSe thin films”, Applied Surface Science (p 6-12). [17] D.Pahinettam Padiyana,*, A.Marikanib, K.R.muralic,”Influence of thickness and substrate temperature on electrical and photoe lectrical properties of vacuum deposited CdSe thin films”, Materials Chemistry and Physics 78(2002)51-58. [18] C.Sene, H.Nguyen cong, M.Dieng, P.Charti er,”Hybrid photovoltaic structures based on polymer/Chemically grown Cadmium Selenide films: Effect of S ilicontungstic acid on the junction properties”, Materials Research Bulletin 35 (2000)1541-1553. [19] S.S.Kale, C.D.Lokhande,”Thickness depend ent properties of ch emically deposited CdSe thin films”, Materials Chemistry and Physics 62(2000)103-108. [20] R.M.Abdel–latif,” Direct current conductivity of ev aporated Cadmium Selenide films”, Physica B 270(1999) 366-370. [21] V.Nadenau, D.Hariskos a nd H.W.Schock, proceedings of 14th European Photovoltaic solar energy conference, Stephens and Assoc,Bedford,U.K.,1997 p.1250. [22] P.Gashin,A.Focsha,T.Potlog,A.Sima shkevich and V.Leondar,”Solar Energy Materials and Solar cells”,46,323,(1997). [23] A.VanCalster, F.Vanfleteren, I.DeRy cke, J.De Baets, J.Appl.Phys.64 (1988)3282. [24] N.G.Patel, C.J.Panchal, K.K Makhijja, Cryst.Res.Technol.29 (1994)1013. [25]V.A.Smyntyana, V.Gerasutenko, S.Ka shulis, G.Mattogno, S.Reghini, Sensors Actuators 19(1994)464. [26] B.Bonello, B.Fernandez, J.Phys.Chem.Solids 54(1993)209. [27] J.C.Schottmiller, R.W.Fran cis, C.Wood, US patent 3884688, 1975. [28] B.J.Curtis, H.Kiess, H.R.Brunne r, K.Frick, Photogr.Sci.Eng.24 (1980)244.

PAGE 88

72 [29] P.Mahawela, S.Jeedigunta, C.S.Fereki des and D.L.Morel,”Analysis of carrier transport phenomena in high band gap II-VI based MIS photovoltaic devices”. [30] D.Braunger,Th.Durr,D.Hariskos,Ch.Kobl e,Th.Walter,N.Wieser and H.W.Schock,” Improved Voc in CuInSe2-based Solar Cells”, Proceedings of the 25th IEEE PVSC, Washington, D.C.,1996,p.1001.