TTTATTTITREASS
`Sea Am a Lena
`
`int TUTTEcans an 7nm
`
`3hed ed boda
`
`:
`
`HANWHA 1016
`
`

`

`ELECTRICITY FROM SOLAR CELLS
`
`
`
` PRACTICAL PHOTOVOLTAIC
`
` 3rd edition revised
`
`NURIAa
`Introduction by John Perlin
`
`aatec publications - Ann Arbdor, Michigan
`
`

`

`Second Printing, 2002
`
`Third Edition Revised, 2001
`
`© 1995 aatec publications
`
`aatec publications PO Box 7119 Ann Arbor, Michigan 48107
`800.995.1470 (outside the U.S. 734.995.1470)
`FAX 734.995.1471
`aatecpub@mindspring.com
`
`Library of Congress Cataloging-in-Publication Data
`
`Komp, Richard J.
`Practical photovoltaics: electricity from solar cells / Richard
`J. Komp. — 3rd ed.
`p.
`cm.
`Includes bibliographical references and index.
`ISBN 0—937948—11—X
`1. Solar cells.
`2. Photovoltaic power generation.
`TK2960.K65
`1995
`621/31°224—dc20
`
`I. Title
`
`95-—22580
`CIP
`
`All Rights Reserved, No part of this publication may be reproduced or transmitted in
`any form or by any means, electronic or mechanical, including photocopy, record-
`ing, or any information storage and retrieval system, without the written permission
`of the publisher.
`
`“The History of Solar-Generated Electricity” © 1995 John Perlin
`Manufactured in the United States of America
`
`Translations by Catherine McCargar
`Cover by Hesseltine & DeMason Design, Ann Arbor, Michigan
`
`The author and aatec publications assume no responsibility for any personalinjury,
`property damage, or other loss suffered in activities related to the information pre-
`sented in this book. Please practice caution and follow propersafety procedures.
`
`

`

`Contents
`
`Preface and Acknowledgments
`
`vii
`
`Preface to the First Edition ix
`
`Introduction: The History of Solar-Generated Electricity
`by Jobn Perlin
`xi
`
`Chapter 1
`
`I
`
`4
`
`Solar Cells: What They Are and How They Work
`What Solar Cells Are
`7
`J
`Physical Characteristics
`Electrical Characteristics 2
`Cell Performance Ratings
`How Solar Cells Work 5
`The Nature of Sunlight 5
`Solid-State Physics
`7
`Doping
`11
`Junction Photovoltaic Cells 13
`Homojunctions 73
`Factors Which Influence Cell Efficiency
`Band Gap Width
`18
`Recombination 19
`Reflectivity and Light Absorption
`Heterojunctions
`20
`
`18
`
`20
`
`ili
`
`

`

`iv
`
`PRACTICAL PHOTOVOLTAICS
`
`Schottky Barrier Junctions 27
`Advanced Semiconductor Devices 22
`Recommended Readings
`23
`
`Chapter 2
`
`Chapter 3
`
`Chapter 4
`
`25
`
`29
`29
`
`How Solar Cells Are Made
`2
`Silicon 25
`Metallurgical-Grade Silicon 26
`Semiconductor-GradeSilicon 26
`Growing Single Crystals
`ae
`Slicing the Crystal into Wafers
`Polishing and Etching the Wafer
`Forming the p—n Junction 30
`Applying the Fingers and Back Contact 32
`The Antireflection Coating 33
`Assembly into Modules
`34
`New Production Techniques for Silicon 36
`Solar-Grade Silicon
`36
`Polycrystalline Silicon
`37
`Ribbon Growth Systems
`39
`AmorphousSilicon
`40
`Recommended Readings
`
`47
`
`50
`
`53
`
`Solar Cells and Modules 49
`Series and Parallel Strings #9
`Module Construction Techniques
`Encapsulation
`53
`Module Failure Mechanisms
`Joints Between Cells
`53
`Encapsulant Problems
`54
`Shunt Diodes
`55
`Concentrator Systems
`Two-Axis Tracking
`One-Axis Tracking
`Recommended Readings
`
`56
`58
`59
`
`62
`
`63
`Using Photovoltaics
`Current Cost of Solar Electricity 63
`Sizing the Array
`69
`69
`Calculating Load
`70
`Sizing Methodology
`Installing Solar Cell Arrays 74
`Array Orientation
`74
`Wiring the Array
`76
`Combination Systems
`
`76
`
`

`

`Chapter 5
`
`Chapter 6
`
`CONTENTS
`
`Vv
`
`84
`
`77
`Inverters
`7&8
`Rotary Inverters
`Solid-State Inverters
`78
`Synchronous Inverters
`/9
`Recommended Readings
`SO
`Batteries and Other Storage ae 81
`Storage Batteries
`S&/
`:
`The Lead—Acid Battery
`8&2
`The Deep-Discharge Lead—Acid Battery
`Lead—Acid Cell Characteristics
`8&7
`A Practical Battery Storage System 94
`Maintenance 95
`Sulfation 95
`96
`Used Batteries
`Nickel—Cadmium Batteries 97
`Nickel—Iron Batteries
`99
`Nickel—Hydride Batteries
`Small Storage Systems
`1O1
`Mechanical Storage Systems
`Flywheels
`702
`Hydrogen
`103
`Recommended Readings
`106
`New Developments in Photovoltaic Technology
`Other Junction Structures
`108
`Tin Oxide and Indium Oxide
`Metal—Semiconductor Junctions
`Photoelectrochemical Cells 709
`Cadmium Sulfide/Cadmium Telluride and
`Other II-VI Heterojunction Cells
`710
`IU—V Semiconductor Combinations
`1174
`Electroplated Cells
`715
`Organic Semiconductors 117
`Clever Optical Systems
`119
`Cascade Cells
`120
`
`7200
`
`102
`
`107
`
`708
`i109
`
`Chapter 7
`
`122
`Luminescent Concentrators
`The Solar Thermophotovoltaic Systems
`Recommended Readings
`124
`
`123
`
`125
`The Future of Photovoltaics
`U.S. Governments and Photovoltaics
`The Multinationals and Photovoltaics
`Photovoltaic Power Farms 137
`
`126
`128
`
`

`

`vi
`
`PRACTICAL PHOTOVOLTAICS
`
`Foreign Solar Cell Activity 136
`Japan
`136
`Western Europe 738
`Underdeveloped Countries
`The Solar Electric Home
`141
`Solar Cells in Space
`143
`745
`The Solar Power Satellite
`7146
`A Proposed Alternative
`Recommended Readings 748
`
`139
`
`Appendix A
`
`153
`
`759
`
`Assembling Your Own Modules 149
`Testing Solar Cells
`150
`757
`A Simple Solar Simulator
`757
`A Simple Sample Holder
`Current—-Voltage Measurements
`How Cells are Connected
`154
`Soldering Solar Cells
`158
`Building a Flat Plate Module
`Laying Out the Case
`159
`Cutting the Plastic
`160
`Gluing the Plastic Case
`767 -
`Fastening the Cover
`162
`762
`Installing the Solar Cells
`Building a Hybrid Concentrator Module
`The Backing Support
`164
`Isolating and Encapsulating the Cells
`The Reflector Wings
`777
`Design J/71
`Making the Reflector Wings 171
`A Hybrid Hot Air System 173
`
`163
`
`766
`
`Appendix B
`
`Manufacturers
`
`175
`
`Appendix C
`
`Current Carrying Capacity of Copper Wire
`
`177
`
`Appendix D
`
`Conversion Factors
`
`179
`
`Glossary
`
`181
`
`Index
`
`I91
`
`About the Author
`
`198
`
`

`

`Preface
`
`Sincethe first edition of Practical Photovoltaics was published, both the
`alternative energy field and the photovoltaics industry have experienced
`considerable change. While the pressing need for energy conservation
`and alternative energies continues, the panic has abated and government
`involvement in research and developmenthas declined drastically. [Note
`to Edition 3.1: In 2001, the panic returned.] Nonetheless, the photovol-
`taics industry has become international in scope with an existence that
`goes far beyond the programs set up by the Department of Energy or
`other government agencies, More solar cell companies have started up,
`making photovoltaic cells by new processes, while the original com-
`panies have evolved and grown.
`The amorphous silicon solar cell has come into its own, powering
`millions of calculators and other small devices; polycrystalline silicon
`blocks are being made into solar cells in Germany and Japan, as well as
`in the United States. Sales of solar cell modules and systems are growing
`fast, increasing from about five megawatts per year when thefirst edi-
`tion was published in 1981 to better than 200 megawatts a year now,
`reflecting our increasing dependence onthis subtle, silent energy source.
`Solar cells are now an accepted and important product, the indus-
`try is maturing, and the future looks bright. PV cells are becoming more
`
`Vii
`
`

`

`viii
`
`PRACTICAL PHOTOVOLTAICS
`
`efficient as their cost decreases. Research is continuing, not only to
`better understand the physics and chemistry of the present crystal-
`line and amorphoussolar cells, but also to develop entirely new cell
`concepts, such as high-efficiency tandem cells, and new materials,
`such as cadmium telluride, that are now being used in commercial
`products.
`This third edition of Practical Photovoltaics has been extensive-
`ly reorganized, offering an improved, even more clear presentation
`of its contents. Throughout, material has been brought up to date
`and new data added. To reflect the changes of the past few years, the
`chapters on “New Developments in Photovoltaic Technology” and
`“The Future of Photovoltaics” have been updated and expanded,
`The chapters on using and installing systems report recent develop-
`ments and improvements. Practical Photovoltaics, Third Edition, is
`more comprehensive, useful, and current, and I look forward to its
`being as well received as its predecessors,
`
`—RJK
`
`ACKNOWLEDGMENTS
`
`I gratefully acknowledgethe aid of Dr. Dan Trivich, Dr. Edward Wang,
`and others who furnished a good deal of the information on photo-
`voltaic cells and inspired me to develop this book.
`I also am happy to thank Vicky Patton, who typed the manu-
`script; Lawrence Komp, who drew the diagrams; David Ross Stevens,
`who furnished many of the workshop photographs; and the manu-
`facturers who kindly supplied illustrations.
`Finally, I wish to acknowledge the help of my workshop partic-
`ipants, who have assisted in the development andtesting of proce-
`dures used to make practical use of photovoltaics.
`
`

`

`Chapter 1
`Solar Cells: What They Are
`and How They Work
`
`WHAT SOLAR CELLS ARE
`
`Solar cells are solid-state devices that absorb light and convert light
`energy directly into electricity. This is done entirely within their solid
`structure; solar cells have no moving parts.
`
`Physical Characteristics
`
`When you look at a solar cell, you see a metallic-blue or black disc
`covered with thin silvery lines. (Figure 1.1 shows both typical silicon
`solar cells and the silicon materials from which they are made.) The
`actual work of converting sunlight into electricity occurs in the dark
`metallic area because dark colors absorb light more readily. Thesil-
`very lines are the front contact fingers and are used to make the
`electrical contact to the front of the cell. The fingers must be very
`fine so that aslittle sunlight as possible is blocked out. The back
`contact is a solid metal layer that both reflects light back up through
`
`

`

`Z
`
`PRACTICAL PHOTOVOLTAICS
`
`Figure 1.1. Silicon ingot, cylinder, wafers, and finished solar cells.
`Get Propulsion Laboratory, California Institute of Technology)
`
`the cell and makes a good electrical contact. A detailed explanation
`of how solar cells work appears later in this chapter.
`
`Electrical Characteristics
`
`In order to use photovoltaic cells, we should have a basic under-
`standing of their electrical characteristics. When illuminated, the so-
`lar cell acts somewhat like a battery in that it produces a voltage
`between front and back. This voltage is developed across a junction
`that is built into the cell structure. This voltage can be used to pro-
`duce a current, just like from a battery, but the amount of currentis
`limited by the amount of light falling on the cell.
`We can use a simple circuit, as shown in Figure 1.2, to explain
`the electrical behavior of solar cells. A load resistor connects the
`front and back of a solar cell. The resistance of the load can be varied
`from a short-circuit zero resistance to a very high value. Two meters—
`a voltmeter and an ammeter—measure the voltage developed across
`the cell and the current passing through the load.
`
`

`

`SOLAR CELLS: WHAT THEY ARE AND HOW THEY WORK
`
`3
`
`SOLAR CELL
`
`BACK CONTACT
`
`
`LOAD
` FRONT
`@) VOLTMETER
`
`
`CONTACT
`
`Figure 1.2. A simple circuit to test a solar cell.
`
`AMMETER
`
`If sunlight shines on the cell when the load resistance is very
`high or the load is disconnected (essentially giving infinite resistance),
`the voltmeter will read a maximum voltage. This is open-circuit
`voltage. For example, the voltmeter would read 0.58 volts on the
`typical silicon solar cell. No current is being drawn from the cell
`under these conditions.
`Conversely, if the load resistance is made zero, we will short-
`circuit the cell Gwhich does a solar cell no harm whatsoever) and
`draw the maximum possible current from the cell. This short-
`circuit currentis directly proportional to the amountoflight falling
`on the cell.
`It is possible to adjust the load resistor between these two ex-
`tremes and measure the voltages and corresponding current pro-
`duced by the cell under different load conditions. The current—
`voltage curve GV curve) in Figure 1.3 showsthe results of such an
`experiment. This figure plots the current on the vertical axis versus
`the voltage on the horizontal axis. The short-circuit current is shown
`on the current axis at zero voltage. As the load resistance increases,
`causing the voltage output of the cell to increase, the current remains
`relatively constant until the “knee” of the curve is reached. The cur-
`rent then drops off quickly, with only a small increase in voltage,
`
`

`

`4
`
`PRACTICAL PHOTOVOLTAICS
`
`SHORT-CIRCUIT
`CURRENT
`
`MAXIMUM POWER
`POINT
`
`Pmax = Vm * Im Vvoltage
`
`CURRENT
`
`ao
`
`OPEN-CIRCUIT
`VOLTAGE
`
`Voc
`
`Figure 1.3. Current—voltage G—V) characteristics of a silicon solar cell.
`
`until the open circuit condition is reached. At this point, the open-
`circuit voltage is obtained and no current is drawn from the device.
`The power output of any electrical device, including a solar
`cell, is the output voltage times the output current under the same
`conditions. The open-circuit voltage is a point of no power: the cur-
`rent is zero. Similarly, the short-circuit condition produces no power
`because the voltage is zero. The maximum power point is the best
`combination of voltage and current and is shown in Figure 1.3. This
`is the point at which the load resistance matches the solar cell inter-
`nal resistance.
`Figure 1.4 shows a series of I-V curves for a solar cell under
`different amounts of sunlight. The peak power current changes pro-
`portionally to the amount of sunlight, but the voltage drops only
`slightly with large changes in the light intensity. Thus, a solar cell
`system can be designed to extract enough usable powerto trickle-
`charge a storage battery even on a cloudy day.
`
`Cell Performance Ratings
`
`To compare the performance of different solar cells, the cells are
`rated at specified amounts and types of sunlight. The most common
`rating parameter for terrestrial cells is air mass 1 (AM1). This is the
`amount of sunlight that falls on the surface of the earth at sea level
`when the sun is shining straight down through a dry, clean atmo-
`
`

`

`SOLAR CELLS: WHAT THEY ARE AND HOW THEY WORK
`
`5
`
`3.0
`
`2.0
`
`=b=u
`
`u
`cc
`oc
`
`> O
`
`1.0
`
`0.1
`
`0.2
`
`0.3
`
`0.4
`
`0.5
`
`0.6
`
`VOLTAGE (V)
`
`Figure 1.4. Series of I-V curves showing solar cell characteristics
`versus sunlight intensity.
`
`sphere. The Sahara Desert at high noon, with the sun directly over-
`head, is a close approximation. The sunlight intensity under these
`conditions is very close to 1 kilowatt per square meter (1 kW/m’). A
`closer approximation to the sunlight conditions usually encountered
`is air mass 2 (AM2), an illumination of 800 W/m?.
`
`HOW SOLAR CELLS WORK
`
`The Nature of Sunlight
`
`In order to understand how solar cells work, you first need to know
`something about the nature of sunlight. All light, including sunlight,
`is a form of electromagnetic radiation similar to radio waves or mi-
`crowaves. The sun gives off this radiation simply because it is hot.
`This black-body radiation is composed of a broad mixture of dif-
`ferent wavelengths, some of which—the visible spectrum—can be
`seen by the naked eye and many wavelengths shorter or longer than
`these. The sun is a black body; if it were cold, it would appear black
`because it would only absorb radiation.
`
`

`

`6
`
`PRACTICAL PHOTOVOLTAICS
`
`Figure 1.5 diagrams the solar spectrum. The curve labeled AMO
`illustrates the air mass O spectrum the sun emits as it appears in out-
`er space; the other curveis the air mass 1 spectrum as seen from the
`surface of the earth. The difference is caused by our atmosphere. In
`Figure 1.5, the visible spectrum is in the middle, from 400 to 700
`nanometers in wavelength [a nanometer (nm) is 10° meters or one-
`millionth of a millimeter]. At 700 nm, the visible spectrum appears
`red; at the shorter wavelength end of 400 nm,it appears violet; the
`other colors of the rainbow appear between. Our eyes are most sen-
`sitive to the wavelengths around 500 nm where the peak of the solar
`spectrum on earth occurs. Over half of the sun’s energy that reaches
`the earth’s surface is in the form of visible radiation.
`Ultraviolet (UV) wavelengths are shorter than 400 nm and are
`present in the solar spectrum in small but significant amounts. Ozone
`and most transparent materials absorb or filter out most of these
`energetic wavelengths, which is fortunate because short wavelength
`ultraviolet light can be destructive to organic materials and living
`things. The small amount of UV light that does makeit to the earth’s
`
`PHOTON ENERGY(eV)
`403.0 2.0 1.5
`1.0
`
`0.8
`
`oO
`
`A S
`
`o
`
`a,ay
`ew
`o=
`23
`oe
`
`£3a
`
`e
`x a
`&
`2
`= Fe
`=s
`
`Figure 1.5. The solar spec-
`trum. The dips in the inten-
`sity of the AM1 spectrum
`seen from the surface of the
`earth are mostly caused by
`water vapor in our atmo-
`sphere.
`
`O04
`
`1.6
`1.2
`0.8
`WAVELENGTH(microns)
`
`4
`2.0
`
`

`

`SOLAR CELLS: WHAT THEY ARE AND HOW THEY WORK
`
`7
`
`surface is responsible for the tan—or burn—we receive when ex-
`posed to the sun,
`Infrared wavelengths are longer than 700 nm and, though
`invisible, are perceived by our skin as radiant heat. As the diagram
`shows, a great deal of the sun’s energy is infrared radiation. Large
`bands of infrared are absorbed by water vapor, carbon dioxide, and
`other substances in our atmosphere, but because most of this ab-
`sorption takes place at longer wavelengths, a solar device that does
`not effectively collect wavelengths longer than 2 microns (2,000 nm)
`suffers only a small loss in efficiency.
`According to the theory of quantum mechanics, light is com-
`posed of particles called photons. Photons, which travel through a
`vacuum at the speed oflight, have no mass, but each is a packet of
`energy related to the wavelength of the light. The shorter the wave-
`length, the larger the packet. The energy of individual photons of
`different wavelengths, as shownatthe top of Figure 1.5, is expressed
`in electron volts. One electron volt is the energy an electron acquires
`when it accelerates in a vacuum across a potential of one volt. This
`unit of energy is commonly used bysolar cell physicists since it is a
`convenient size when considering individual electrons.
`
`Solid-State Physics
`
`Although the solar cell was accidentally discovered in the nineteenth
`century, and inefficient versions of selenium and cuprous oxide pho-
`tovoltaic cells were investigated and even used commercially in the
`early part of this century, it was only after the development of mod-
`ern solid-state theory and the band model of semiconductors that
`the inner workings of photovoltaic cells have been understood.
`The band model of solids, presented here in a simplified ver-
`sion, is based on a crystal with all the atoms fixed in a pattern. The
`individual atomic nuclei vibrate around a fixed spot in a three-
`dimensional lattice pattern, but they usually cannot jump completely
`out of place. Powerful electrostatic forces tie most of the negative
`electrons in a particular atom very closely to the positively charged
`nucleus. However, the outermost electrons (called the valence elec-
`trons) can be considered “delocalized”; that is, they do not belong to
`any particular atom, but to the crystal as a whole.
`
`

`

`8
`
`PRACTICAL PHOTOVOLTAICS
`
`VACUUM ZERO
`
`+.
`
`ELECTRONS
`
`WORK FUNCTION
`
`
`HOLES
`
`
`
`
`
`*
`
`=
`* ———
`
`yt Ss= —
`tgMery LEVELS
`<3 P26
`:tf
`
`
`
`
`FERMI
`LEVEL
`
`
`
`
`Figure 1.6. Energy level diagram for a metal.
`
`These valence electrons balance the remaining positive charges
`of all the nuclei. It takes a certain amount of energy to remove one of
`these electrons from the crystal because of electrostatic attraction,
`This energy is represented in Figure 1.6 as a vertical distance. In this
`figure, which illustrates the band diagram for a metal, zero energy is
`taken as the energy of an electron at rest in empty space. This is
`called vacuum zero. The energies of the electrons in the crystal are
`all below this zero; the kinetic energy of a fast-moving electron in
`vacuum would be abovethe zero.
`One conclusion of quantum mechanics is that electrons (or any-
`thing else, for that matter) cannot have a continuum of energies but
`are allowed only certain selected energies (the energy levels in the
`diagram). Another quantum mechanical principle is that no two elec-
`trons can occupy the same energy level.
`It is possible to calculate the number and positions of the ener-
`gy levels of a particular metal; if a count is made, it turns out that
`each atom contributes more energy levels than electrons. (This also
`is shown in Figure 1.6.) Since the band of valence levels is greater
`than the number of electrons, some levels must be empty. The elec-
`trons would preferentially occupy the lowest lying levels. At absolute
`zero, when none of the electrons are moving, they would fill the
`band to the fermi level, where the probability of finding an electron
`is one-half, In order to move, the electrons must have a bit more
`energy (their kinetic energy) and would occupy a level just above
`
`

`

`SOLAR CELLS: WHAT THEY ARE AND HOW THEY WORK
`
`9
`
`the fermi level. Since there are many empty levels at this energy,
`the electron has a large number of possible kinetic energies and can
`move very freely throughoutthe crystal. These moving electrons can
`produce an electric current, so a metal would be expected to be a
`good conductorof electricity and capable of carrying a large current.
`Each electron from below the fermi level that jumps to a higher
`empty level leaves behind an empty level. These empty levels in the
`“sea” of electrons are called holes. A hole is a level that an electron
`could occupy but currently does not. A hole “moves” when an elec-
`tron falls into it, creating a new hole. Thus, a hole can be seen as a
`particle similar to an electron with a positive charge, a mobility, and
`even an effective mass. Holes can carry electric current and in some
`metals are the dominantcarriers.
`If the number of available levels is exactly the same as the num-
`ber of electrons for each atom, the situation depicted in Figure 1.7
`results. This figure also shows a band of completely empty energy
`levels at some higher energy. Since the number of excited electronic
`states in any atom is theoretically infinite, there will always be such a
`conduction band, sometimes separated from the valence band by
`a band gap. Because the valence band is completely filled, there are
`no nearby energy levels for an electron to occupy. The electrons
`
`VACUUM ZERO
`
`COMPLETELY EMPTY CONDUCTION BAND
`
`+
`
`BAND GAP
`
`LLLLLLLLLLALLALALLALLL YffLLL)Res
`
`OMPLETELY FILLED VALENCE BAND
`
`Figure 1.7. Energy level diagram for a perfect insulation.
`
`

`

`10
`
`PRACTICAL PHOTOVOLTAICS
`
`cannot move and therefore cannot create any holes. Thus, no con-
`duction can take place and the material is called an insulator. The
`energy of the band gap is in electron volts, as are all the energies
`shown in the diagrams.
`If the band gap is sufficiently narrow (below 2.5 electron volts
`or sO), it is possible for a particularly energetic electron to jump from
`the top of the valence band to one of the empty levels in the conduc-
`tion band (see Figure 1.8). The electron might be thermally excited
`by the vibrations of the atoms in the crystal lattice or a photon of
`light of sufficient energy might be absorbed by the crystal and cause
`the electron to jump. Once the electron is in the upper conduction
`band,
`it is free to move and can act as a carrier of electricity. In
`addition, the hole left behind in the valence band can also be a
`current carrier. A material that has these properties is called a semi-
`conductor.
`In Figure 1.8, the fermi level is in the middle of the band gap.
`Even though there are no energy levels in the middle of the band
`gap, one can have the concept of a fermi level there. Were there an
`energy level at this place, it would be occupied by an electron half
`
`VACUUM ZERO
` a»
`
`AFFINITY
`BAND
`ELECTRON
`&eeSe,eeee,eessa
`ah
`
`WORK
`
`e
`
`FERMI LEVEL
`
`BAND GAP
`
`Y U7,neticnsie|VALENCE BAND
`
`TTLLZ
`
`Figure 1.8. Energy band diagram for an intrinsic semiconductor.
`
`

`

`SOLAR CELLS: WHAT THEY ARE AND HOW THEY WORK
`
`11
`
`the time, on average. Above the fermi level, fewer levels are occu-
`pied; below it, most are. The unoccupied levels are holes.
`As the temperature of a solid increases, the most energetic elec-
`trons and holes can be found farther from the fermi level, but the
`position of the fermi level remains fixed. Only the most energetic
`electrons and holes are actually occupying levels. (The most impor-
`tant point to rememberis that the closer a band is to the fermilevel,
`the more conduction electrons—orholes, if below the fermi level—
`that will be found in the band.) For a more complete mathematical
`discussion of these solid-state concepts, see Kittel’s Introduction to
`Solid-State Physics, 5th Edition.
`
`Doping
`
`The electrical conductivity discussed in the last section is called in-
`trinsic conductivity becauseit is an intrinsic property of the partic-
`ular material and its crystal lattice. The ideal intrinsic semiconductor
`cannot exist in reality because all materials have some impurities, no
`matter how carefully prepared. The impurity atoms will occupy var-
`ious spots in the crystal lattice and disturb its perfection. In addition,
`these foreign atoms usually contribute a different number of levels
`and/or electrons to the system than the normal atoms. For example,
`in a single crystal of silicon each silicon atom contributes exactly 4
`valence levels, 4 conduction levels, and 4 electrons since the valence
`of silicon is 4.
`If an atom of boron (which has a valence of 3) is added to the
`crystal, it will contribute only 3 electrons to the system. It will also
`contribute 3 valence levels, plus 1 level which will be created in the
`band gap slightly above the top of the valence band, This situation is
`diagrammedin Figure 1.9 deft). These extra levels, called acceptor
`levels, are capable of accepting an electron from the valence band.
`Since the acceptor level is associated with a particular impurity atom,
`the electron that occupies it is trapped and cannot move. However,
`the hole created when the electron leaves the valence band can move
`and carry an electric current. Silicon doped with boron makes a p-
`type semiconductor, since positive holes carry the current.
`The process of deliberately adding a known impurity to a pure
`semiconductor is called doping and the resulting material is an ex-
`trinsic semiconductor. Similarly, if a small amount of phosphorus
`
`

`

`12
`
`PRACTICAL PHOTOVOLTAICS
`
`Y
`
`
`
`O83ZWANDVA
`
`HSIHEVOALIHONIW310H--LENBS__VF_-
`
`
`73A37HONOGienHAS
`
`HSINHVOALINOPVINSieINESALINONIW
`
`HOLONGNODIWAS3dAL-UHOLONGNODINGSadAL-d
`
`
`
`NOH19473ONYVE—ou9973NOHL9374|NOLLONANOD
`
`
`
`
`
`‘sIOJONpUOITWAS(Padop)JO;WeIseIPpueqABIOU‘6JaunsyyIIsULNXA
`‘IojonpuootuasadAj-U-Wq5IYyJOIONpuCoIUesadAj-d-yfaT
`
`
`
`
`
`ALIMOPYNALeee(ii30H
`SLeeoeAH3INHVD
`
`STANSTHOLd300V
`
`SONAIWA
`
`anya
`
`SHAIMHV9
`
`
`

`

`SOLAR CELLS: WHAT THEY ARE AND HOW THEY WORK
`
`13
`
`is addedto thesilicon crystal, the atoms of phosphorus (which have
`5 valence electrons) will contribute 4 electrons and 4 levels to the
`valence band, and the extra electron will occupy a level near the bot-
`tom of the conduction band as shown in Figure 1.9 (right). It takes
`very litthe energy for the extra electron to jump to the conduction
`band from this donorlevel, so-called because it can donate conduc-
`tion electrons to the system. This type of material is an n-type semi-
`conductor, since negative electrons carry the electric current.
`Notice that Figure 1.9 deft) shows p-type material: the fermi
`level is located near the valence band and the vast majority of cur-
`rent carriers are holes. Figure 1.9 Gight) shows n-type material: the
`fermi level is near the bottom of the conduction band and the major-
`ity of carriers are electrons. In p-type semiconductors, the positive
`holes are the majority carriers and the electrons are the minority
`carriers. In n-type semiconductors, the negative electrons are the
`majority carriers and the holes are naturally called minority carriers.
`Finally, note the quantities marked work function and elec-
`tron affinity. The work function is the energy difference between
`the fermi level and vacuum zero; the electron affinity is the energy
`difference between the bottom of the conduction band and vacuum
`zero. These definitions hold true for all semiconductors. For metals,
`the work function and electron affinity have the same value.
`
`JUNCTION PHOTOVOLTAIC CELLS
`
`Homojunctions
`
`Figure 1.10 shows the construction of a typical n-on-p junction solar
`cell. This is the most common type of cell and is made by taking a
`wafer of p-type single-crystal silicon and diffusing phosphorus atoms
`into the top surface. This can be done by heating the wafers in a
`diffusion furnace in the presence of phosphorus-containing gas. The
`phosphorus atoms have a valence of 5. This means that each phos-
`phorus atom has 5 electrons in its outermost region compared to
`silicon atoms which have 4. The extra electron is very loosely held
`near the atom, producing donor levels in the top layer of silicon,
`making it n-type. The n-type silicon on top of the p-type silicon
`produces a junction where the two types come in contact. The back
`
`

`

`14
`
`PRACTICAL PHOTOVOLTAICS
`
`INCIDENT RADIATION
`
`bid
`
`uw FRONT ELECTRODE
`
`FRONT ACTIVE
`ELECTRODE —»
`
`“Opaque metal
`
`«— BACK ACTIVE ELEMENT
`*«Semiconductor
`
`«— BACK ELECTRODE
`
`Figure 1.10. Cross section of a solar cell showing its construction.
`
`contact and the top fingers shown are used to connect the solar cell
`to the external circuit. A more detailed description of the construc-
`tion of these cells is given in the next chapter.
`The energy band diagram for this n-on-p cell (called a homo-
`junction solar cell) is shown in Figure 1.11. This diagram is made by
`overlaying the diagram for the n-type semiconductor with the dia-
`gram for the p-type material and then aligning the fermi levels. No-
`tice that now there is a difference in the vacuum zero levels for the
`two materials, This represents the built-in potential that exists be-
`tween the materials because of the difference in their work func-
`tions. Between the n- and p-layers is a depletion layer where there
`are essentially no carriers—neither electrons nor holes, This layer,
`where the bands are bent, is the actual junction. The front and back
`metal contacts must be ohmic contacts: that is, contacts that do not
`impede the flow of electrons into or out of the semiconductor, In the
`dark,
`if the,junction device is connected to a battery so the front
`fingers are made positive and the back negative, the situation shown
`in Figure 1.12 Gop) would result. The applied potential is added to
`the built-in potential and the uphill barrier for the electrons is in-
`creased; this results in no current flow except for the tiny current of
`electrons thermally excited from the valence to the conduction band.
`At room temperature this current would be a fraction of a microamp.
`On the other hand, if the battery is reversed so that the front is neg-
`ative and the back positive, the applied potential can be made large
`enough to cancel out the built-in potential. As Figure 1.12 (bottom)
`
`

`

`SOLAR CELLS: WHAT THEY ARE AND HOW THEY WORK
`
`15
`
`I
`
`— VACUUM ZERO
`
`
`
` BUILT-IN !
`ELECTRON-
`
`HOLE PAIR
`
`
`POTENTIAL
`
` METAL BACK
`
`| i
`
`CONTACT
`
` FERMI
`
`
`, ACCEPTOR
`LEVEL
`LEVELS—- = ¥
`
`
`
`SUNLIGHT
`
`n-TYPE
`
`j DEPLETION
`1
`LAYER
`
`p-TYPE
`
`Figure 1.11. Band diagram for p—n homojunction solar cell.
`
`shows, there is no longer any barrier to the flow of either holes or
`electrons and the device becomes a good conductor of electricity.
`So, a solar cell is a cell is a junction diode or rectifier, letting
`electricity pass in only one direction in the dark. The “dark” curve in
`Figure 1.13 is a typical current voltage plot for this type of junction.
`If light falls on a solar cell, those photons with energy greater
`than the band gap width can excite an electron from the valence to
`the conduction band and create a hole-electron pair. The conduction
`electron, if created in the depletion layer, will slide “downhill” to-
`ward the front n-layer and the hole will slide “uphill” Cike an air
`bubble in water) to the p-layer. Thus, the charge carriers will be
`separated and current will flow. The ohmic front contact allows the
`electron to flow through an external circuit to the back contact where
`it can recombine with a hole, leaving everything as it was. Even if the
`conduction electron is created away from the junction, if it happens
`
`

`

`16
`
`PRACTICAL PHOTOVOLTAICS
`
`APPLIED VOLTAGE CANCELS
`
`BUILT-IN POTENTIAL
`
`
`Figure 1.12. Top: Forward
`bias rectifier. Bottom: Reverse
`bias rectifier.
`
`APPLIED VOLTAGE
`ADDS TO BUILT-IN
`POTENTIAL
`
`NO CURRENT FLOWS
`
`BATTERY
`
`to drift toward the junction it will be carried down as before and
`produce a current in the external circuit.
`If the external circuit is simply a wire and has no appreciable
`resistance, the current that flows is the short-circuit current C,.) and
`is directly related to the number of photons of light being absorbed
`
`

`

`SOLAR CELLS: WHAT THEY ARE AND HOW THEY WORK
`
`17
`
`CURRENT OPEN-CIRCUIT
`
`VOLTAGE
`
`V