the solar cell

THE SOLAR CELL Solar Cell Technology

By

Cliff Orori Mosiori B.Ed (Sc), M.Sc (Physics), Ph.D (Physics) Department of Electrical & Electronic Engineering Rift Valley Insititute of Science and technology Nakuru,

Kenya

Email: [email protected]

i

ii

PREFACE A solar cell is any device that directly converts the energy in light into electrical energy through the process of photovoltaics. A solar cell can also be called a photovoltaic cell but a solar reference cell is simply a small area surface (2 cm x 2 cm) solar cell packaged in a metal housing under a glass window intended for use indoors to set simulated sunlight levels. In its simplest form, the solar cell consists of a junction formed between n-type and p-type semiconductors, either of the same material (homo-junction) or different materials (hetero-junction). Three key elements in a solar cell form the basis of their manufacturing technology. The first is the semiconductor, which absorbs light and converts it into electron-hole pairs. The second is the semiconductor junction which separates the photo-generated carriers which are electrons and holes, and the third is the contacts on the front and back of the cell that allow the current to flow to the external circuit. The two main categories of technology are defined by the choice of the semiconductor: either crystalline silicon in a wafer form or thin films of other materials. Faced with ever-increasing demand, the earth's sources of non-renewable energy are not expected to last long. Among the many contenders vying to replace fossil fuels, photovoltaic solar cells offer many advantages, including needing little maintenance and being relatively environmentally-friendly; the major drawback to date has been cost. In order for photovoltaics to be viable for large-scale energy conversion, their efficiency must be improved whilst making them cheaper. Solar radiation represents the largest energy flow entering our terrestrial ecosystem. After reflection and absorption in the atmosphere, some 105 TW hit the surface of Earth and undergo conversion to all forms of energy used by humans with the exception of nuclear, geothermal, and tidal energy.

This resource is enormous and corresponds to almost 6,000 fold the current global consumption of primary energy (13.7 TW). Thus, solar energy has the potential of becoming a major component of a sustainable energy iii

portfolio with constrained greenhouse gas emissions. Solar radiation is a renewable energy resource that has been used by humanity in all ages and for various uses. Passive solar technologies have been used by ancient civilizations for warming and/or cooling habitations and for water heating and during Renaissance. Concentration of solar radiation was extensively studied and in the 19th century the first solar-based mechanical engines were built. The discovery of photovoltaic effect by Becquerel in 1839 and the creation of the first photovoltaic cell in the early 1950s opened entirely new perspectives on the use of solar energy for the production of electricity. Since then, the evolution of solar technologies continues at an unprecedented rate.

Nowadays, there exist an extremely large variety of solar technologies and photovoltaics have been gaining an increasing market share for the last 20 years. Nevertheless, global generation of solar electricity is still small compared to the potential of this resource. Based on current cost of the advanced solar technologies, they make solar cells or energy hardly competitive on an energy market because it is still dominated by cheap fossil fuels. From a scientific and technological viewpoint, the greatest challenge is to find new solutions for solar energy systems so as to make them less capital intensive and more efficient. A lot of investigations on low-cost and /or high-efficiency photovoltaic device concepts are being developed. Solar thermal technologies are reaching a mature stage of development and have the potential of becoming competitive for large energy supply in future. Extended research efforts in energy storage devices such as batteries and other electric storage systems, thermal storage and the direct production of solar fuels are in top gear and this book discusses the solar cell base on the above highlighted facts.

(Authors)

iv

Table of Contents THE SOLAR CELL .................................................................................................. i PREFACE............................................................................................................... iii ACKNOWLEDGEMENT ........................................................................................ x CHAPTER ONE ...................................................................................................... 1 INTRODUCTION TO SOLAR CELLS ................................................................... 1 Crystalline silicon solar cells .................................................................................... 3 Thin film solar cells ................................................................................................. 4 Challenges of Making Reliable Solar Cell Measurements ......................................... 7 Electrical Model of the PV Cell .............................................................................. 13 Photometry ............................................................................................................ 14 CHAPTER TWO ................................................................................................... 19 SOLID STATE PHYSICS...................................................................................... 19 Pauli Exclusion Principle ....................................................................................... 21 Thermal Physics ..................................................................................................... 26 Equilibrium Carrier Statistics ................................................................................. 28 Doping and Temperature dependence of Fermi level .............................................. 31 Band Bending ........................................................................................................ 35 Non- Equilibrium Carrier Statistics......................................................................... 39 Drift and Diffusion ................................................................................................. 42 CHAPTER THREE................................................................................................ 43 SOLAR CELL SEMICONDUCTOR MATERIALS ............................................... 43 Semiconductor transport carriers ............................................................................ 52 Carrier generation-recombination process ............................................................... 56 Optical phenomena in thin films ............................................................................. 57 Thin film solar cell applications.............................................................................. 61 The p-n junction ..................................................................................................... 62 Photovoltaic cells ................................................................................................... 64 Performance of photovoltaic cells ........................................................................... 66 Photovoltaic cell operation ..................................................................................... 67 Strengths and limitation of photovoltaic cells ......................................................... 69

v

CHAPTER FOUR ................................................................................................ 71 SOLAR CELL THIN FILM GROWTH TECHNIQUES......................................... 71 Growth Kinetics and Diffusion of thin films ........................................................... 72 Physical vapour deposition (PVD) .......................................................................... 74 Chemical Deposition Techniques ........................................................................... 82 Chemical Vapour Deposition (CVD) ...................................................................... 87 Liquid deposition techniques .................................................................................. 90 CHAPTER FIVE ................................................................................................... 91 CHARACTERIZATION OF THIN FILM FOR SOLAR CELLS ........................... 91 Characterization fundamentals................................................................................ 91 Measurement of Magnetic propeties ....................................................................... 92 Measurement of thin film Thickness ....................................................................... 92 Chemical measurement Techniques ........................................................................ 93 Measurement of structural properties ...................................................................... 94 Measurement of electrical Properties .................................................................... 104 Measurement of optical properties ........................................................................ 118 Thin film post-treatment methods ......................................................................... 120 CHAPTER SIX .................................................................................................... 122 THE SOLAR CELL ............................................................................................. 122 Prototype solar cell............................................................................................... 122 Solar Energy Conversion Technology .................................................................. 132 Solar technologies ................................................................................................ 139 Technological challenges ..................................................................................... 155 Hybrid Solar Lighting .......................................................................................... 171 Principles of Operation......................................................................................... 172 New Developments in Hybrid Solar Lighting Technology .................................... 176 CHAPTER SEVEN.............................................................................................. 179 OPERATION OF SOLAR ENERGY ................................................................... 179 Principle of Operation of Solar Energy ................................................................. 179 Principles of Solar Energy .................................................................................... 181 Solar Spectrum..................................................................................................... 182 Solar Receiver Technologies ................................................................................ 187

vi

Solar Photovoltaic Technologies .......................................................................... 189 Applications of Solar Energy................................................................................ 197 CHAPTER EIGHT .............................................................................................. 202 SOLAR CELL ASSESSMENT ............................................................................ 202 Factors considered in Solar System Designs ......................................................... 202 System Design of Solar PV Systems ..................................................................... 206 Solar Array Design............................................................................................... 207 Resource Measurement Assessment ..................................................................... 210 Measurement Tools .............................................................................................. 212 Photovoltaic Cell Circuit and Device Parameters .................................................. 215 CHAPTER NINE ................................................................................................. 225 MATERIALS FOR SOLAR CELLS .................................................................... 225 Alternative Sources of Energy .............................................................................. 225 Generations of Solar Cells .................................................................................... 228 Photoelectrochemical solar cells ........................................................................... 237 Dye-Sensitised Mesoscopic Solar Cells -Michael Grätzel ..................................... 238 Mode of function of dye-sensitised solar cells....................................................... 240 DSSC research and development .......................................................................... 253 Tandem cells ........................................................................................................ 258 Mesoporous oxide film development .................................................................... 264 Solid-state dye-sensitised cells ............................................................................. 268 Solar Cell Modeling History ................................................................................. 269 Drift Diffusion Model .......................................................................................... 271

vii

viii

ix

ACKNOWLEDGEMENT

We are very thankful to all who have helped us proof-read and also make corrections on some sections of this book. We are also deeply indebted to all solar cell and solid state physicists’ giants who came before us, those who taught this course or similar courses in other Universities in the world in recent years. We cannot emphasize enough here that there are many extremely good books on Solar Cells and Condensed Matter Physics already in existence that we have quoted explicitly or implicitly. There are also many good resources online that include guide books and lecture notes to Semiconductor Physics and Solar Cells that we owe a lot because

of

their

contributions

to

wholeheartedly acknowledge all them.

x

this

book.

We

xi

CHAPTER ONE

INTRODUCTION TO SOLAR CELLS A solar cell is any device that directly converts the energy in light into electrical energy through the process of photovoltaics. It can also be called a photovoltaic cell. The development of solar cell technology begins with the 1839 research of French physicist Antoine-César Becquerel. Becquerel observed the photovoltaic effect while experimenting with a solid electrode in an electrolyte solution when he saw a voltage develope when light fell upon the electrode. Solar panels are devices that convert light into electricity. They are called solar after the sun or "Sol" because the sun is the most powerful source of the light to use. They are sometimes called photovoltaics, which means "light-electricity". Solar cells or PV cells rely on the photovoltaic effect to absorb the energy of the sun and cause current to flow between two oppositely charge layers. Photovoltaic (or PV) systems convert light energy into electricity.

The term "photo" is a term from the Greek "phos," which means "light." "Volt" is named for Alessandro Volta (1745-1827), a pioneer in the study of electricity. Photovoltaics literally, mean light-electricity, thus known as "solar cells". PV systems are already an important part of our lives and the simplest systems power many of the small calculators and wrist watches we use every day. More complicated systems provide electricity for pumping water, powering communications equipment, and even lighting our homes and running our appliances. In a surprising number of cases, PV power is the cheapest form of electricity for performing these tasks. Photovoltaic cells convert light energy into electricity at the atomic level.

French physicist Edmond Becquerel first described the photovoltaic effect in 1839, but it remained a curiosity of science for the next three quarters of a 1

century. At only nineteen, Becquerel found that certain materials would produce small amounts of electric current when exposed to light.The effect was first studied in solids, such as selenium, by Heinrich Hertz in the 1870s. Soon afterward, selenium photovoltaic cells were converting light to electricity at one percent to two percent efficiency. As a result, selenium was quickly adopted in the emerging field of photography for use in light-measuring devices. Major steps toward commercializing photovoltaic cells began in the 1940s and early 1950s, when the Czochralski process was developed for producing highly pure crystalline silicon.

Because of the increasing demand for energy and the limited supply of fossil fuels, the search for alternative sources of power is imperative. Given that there is a vast amount of energy available from the sun; devices that convert light energy into electrical energy are becoming increasingly important. Solar or photovoltaic (PV) cells convert light energy into useful electrical power. These cells are produced from light-absorbing materials. A photovoltaic substance is a material used in the creation of solar cells that convert sunlight directly into electricity. The long-term goal of photovoltaic (PV) devices has been to reduce our dependency on fossil fuel generated electricity. The main benefits of electricity produced by PV’s are: Reliable operating systems; low operating costs; and virtually zero environmental pollution.

Solar energy is available in abundance in most parts of the world. The amount of solar energy incident on the earth’s surface is approximately1.5 x 1018 kWh/year, which is about 10,000 times the current annual energy consumption of the entire world. The density of power radiated from the sun is what is referred to as solar energy constant and has an average value of 1.373, kW/m2. Solar cell is a device which converts photons in Solar rays to direct-current (DC) and voltage. The associated technology is called Solar Photovoltaic (SPV). A typical silicon PV cell is a thin wafer consisting of a very thin layer of phosphorous-doped (N-type) silicon on top of a thicker layer of boron-doped 2

(P-type) silicon. An electrical field is created near the top surface of the cell where these two materials are in contact (the P-N junction). When the sunlight hits the semiconductor surface, an electron springs up and is attracted towards the N-type semiconductor material. This will cause more negatives in the Ntype and more positives in the P-type semiconductors, generating a higher flow of electricity. This is known as Photovoltaic effect. Three key elements in a solar cell form the basis of their manufacturing technology. The first is the semiconductor, which absorbs light and converts it into electron-hole pairs. The second is the semiconductor junction, which separates the photo-generated carriers (electrons and holes), and the third is the contacts on the front and back of the cell that allow the current to flow to the external circuit. The two main categories of technology are defined by the choice of the semiconductor: either crystalline silicon in a wafer form or thin films of other materials. Historically, crystalline silicon (c-Si) has been used as the light-absorbing semiconductor in most solar cells, even though it is a relatively poor absorber of light and requires a considerable thickness (several hundred microns) of material. Nevertheless, it has proved convenient because it yields stable solar cells with good efficiencies (11 - 16%), half to two-thirds of the theoretical maximum) and uses process technology developed from the huge knowledge base of the microelectronics industry.

Crystalline silicon solar cells Two types of crystalline silicon are used in the industry. The first is monocrystalline produced by slicing wafers (up to 150mm diameter and 350 microns thick) from a high-purity single crystal boule. The second is multi-crystalline silicon, made by sawing a cast block of silicon first into bars and then wafers. The main trend in crystalline silicon cell manufacture is toward or both monoand multi-crystalline Si, a semiconductor homo-junction is formed by diffusing 3

phosphorus (an n-type dopant) into the top surface of the boron doped (p-type) Si wafer. Screen-printed contacts are applied to the front and rear of the cell, with the front contact pattern specially designed to allow maximum light exposure of the Si material with minimum electrical (resistive) losses in the cell. The most efficient production cells use mono-crystalline c-Si with laser grooved, buried grid contacts for maximum light absorption and current collection. Some companies are product ionizing technologies that by-pass some of the inefficiencies of the crystal growth/casting and wafer sawing route. One route is to grow a ribbon of silicon, either as a plain two-dimensional strip or as an octagonal column, by pulling it from a silicon melt. Another is to melt silicon powder on a cheap conducting substrate. These processes may bring with them other issues of lower growth/pulling rates and poorer uniformity and surface roughness. Each c-Si cell generates about 0.5V, so 36 cells are usually soldered together in series to produce a module with an output to charge a 12V battery. The cells are hermetically sealed under toughened, high transmission glass to produce highly reliable, weather resistant modules that may be warrantied for up to 25 years. Modules are designed to meet rigorous certification tests set by international standards agencies. Thin film solar cells Solar cells usually operate more efficiently under concentrated light. This has led to the development of a range of approaches using mirrors or lenses to focus light on to specially designed cells and use heat sinks, or active cooling of the cells, to dissipate the large amount of heat that is generated. Unlike conventional flat plate PV arrays, concentrator systems require direct sunlight or clear skies and will not operate under cloudy conditions. They generally follow the sun's path through the sky during the day using single-axis

4

tracking. To adjust to the sun's varying height in the sky through the seasons, two-axis tracking is sometimes used. Concentrators have not yet achieved widespread application in photovoltaics but solar concentration has been widely used in solar thermal electricity generation technology where the generated heat is used to power a turbine. Unlike the crystalline and thin film solar cells that have solid-state light absorbing layers, electrochemical solar cells have their active component in a liquid phase. They use a dye sensitizer to absorb the light and create electronhole pairs in a nano-crystalline titanium dioxide semiconductor layer. This is sandwiched in between a tin oxide coated glass sheet (the front contact of the cell) and a rear carbon contact layer, with a glass or foil backing sheet. Some consider that these cells will offer lower manufacturing costs in the future because of their simplicity and use of cheap materials. The challenges of scaling up manufacturing and demonstrating reliable field operation of products lie ahead. However, prototypes of small devices powered by dyesensitized nano-crystalline electrochemical PV cells are now appearing (120 cm2 cells with a high efficiency). The high cost of crystalline silicon wafers which make up 40 – 50 % of the cost of a finished module, has led the industry to look at cheaper materials to make solar cells. The selected materials are all strong light absorbers and only need to be about 1micron thick, so materials costs are significantly reduced. The most common materials are amorphous silicon (a-Si, still silicon, but in a different form), or the polycrystalline materials: cadmium telluride (CdTe) and copper indium (gallium) diselenide (CIS or CIGS). Each of these three is amenable to large area deposition (on to substrates of about 1 meter dimensions) and hence high volume manufacturing. The thin film semiconductor layers are deposited on to either coated glass or stainless steel sheet.

5

The semiconductor junctions are formed in different ways, either as a p-i-n device in amorphous silicon, or as a hetero-junction (e.g. with a thin cadmium sulphide layer) for CdTe and CIS. A transparent conducting oxide layer (such as tin oxide) forms the front electrical contact of the cell, and a metal layer forms the rear contact. Thin film technologies are all complex. They have taken at least twenty years, supported in some cases by major corporations, to get from the stage of promising research (about 8% efficiency at 1cm2 scale) to the first manufacturing plants producing early product. Amorphous silicon is well-developed of the thin film technologies. In its simplest form, the cell structure has a single sequence of p-i-n layers. Such cells suffer from significant degradation in their power output (in the range 1535%) when exposed to the sun. The mechanism of degradation is called the Staebler-Wronski Effect, after its discoverers. Better stability requires the use of thinner layers in order to increase the electric field strength across the material. However, this reduces light absorption and hence cell efficiency. This has led the industry to develop tandem and even triple layer devices that contain p-i-n cells stacked one on top of the other. In the cell at the base of the structure, the a-Si is sometimes alloyed with germanium to reduce its band gap and further improve light absorption. All this added complexity has a downside though; the processes are more complex and process yields are likely to be lower. In order to build up a practically useful voltage from thin film cells, their manufacture usually includes a laser scribing sequence that enables the front and back of adjacent cells to be directly interconnected in series, with no need for further solder connection between cells. As before, thin film cells are laminated to produce a weather resistant and environmentally robust module. Although they are less efficient (production modules range from 5 to 8 %), thin films are potentially cheaper than c-Si because of their lower materials costs and larger substrate size.

6

However, some thin film materials have shown degradation of performance over time and stabilized efficiencies can be 15 - 35% lower than initial values. Many thin film technologies have demonstrated best cell efficiencies at research scale above 13%, and best prototype module efficiencies above 10%. The technology that is most successful in achieving low manufacturing costs in the long run is likely to be the one that can deliver the highest stable efficiencies with the highest process yields. Amorphous silicon is the most well-developed thin film technology to-date and has an interesting avenue of further development through the use of "microcrystalline" silicon which seeks to combine the stable high efficiencies of crystalline Si technology with the simpler and cheaper large area deposition technology of amorphous silicon. However, conventional c-Si manufacturing technology has continued its steady improvement year by year and its production costs are still falling too. The emerging thin film technologies are starting to make significant in-roads in to grid connect markets, particularly in Germany, but crystalline technologies still dominate the market. Thin films have long held a niche position in low power (less than 50 W) and consumer electronics applications, and may offer particular design options for building integrated applications.

Challenges of Making Reliable Solar Cell Measurements Photovoltaics is normally associated with images of rooftop mounted solar panels or a vast expanse of solar panel arrays spread out over a desert floor probably because so much emphasis is placed on photovoltaics as an alternative way to generate electrical power. Solar cells generate little electrical power, but do generate information of great interest for the photovoltaic researcher. Solar cells are the smallest photovoltaic devices and are used either as irradiance sensors or as samples for studying new photovoltaic materials and or processes. Solar cells range in size from a few square millimeters up to 156 mm square or more for a silicon wafer.

7

A research or prototype solar cell can have a rather crude looking construction compared to the sleek panels on display at any solar energy convention. It may simply be a thin film of photovoltaic material sandwiched between two glass microscope slides with silver paint for contacting. A research solar cell usually requires probing and typically lacks the encapsulation so important for protecting solar modules from the degrading atmospheric and weather effects.

A solar reference cell is simply a small area (2 cm x 2 cm) solar cell packaged in a metal housing under a glass window intended for use indoors to set simulated sunlight levels. A solar reference cell can be framed in such a way that it resembles a miniature version of its associated solar panel and, in place of a pyranometer, can be used outdoors for use as an accurate irradiance sensor with the same spectral and angle of incidence responses as the panel.

Review of Basic Electricity To understand Photovoltaics, it is necessary to know something about electricity. There are three very important concepts, Voltage, Current, and Power. Voltage, measured in Volts, is a measure of the strength of the electricity. It is analogous to pressure of water in a hose. Current, measured in amps, is a measure of the number of electrons flowing through a wire in a particular time. It is analogous to gallons of water per second flowing in a hose. Note that voltage and current are totally different concepts.

If the nozzle on the hose is closed, the flow (current) is shut off, but the pressure (voltage) is high. If the nozzle is opened, flow (current) is high, but pressure (voltage) drops. If the nozzle is partially open, there is both flow and pressure. Just as flow of water through a hose causes pressure drop due to friction, current flow in a wire causes voltage drop. The third concept is that of power. Power is simply voltage times current. It is also the amount of energy that is delivered in a unit of time. An important difference between water flow 8

in a pipe and electricity is that electric current always travels in a closed loop, a "circuit".

The Photovoltaic Panel The first major component in a Solar Sprint car is the Solar or Photovoltaic (PV) panel. This is its power source. The more power that can be delivered to the motor, the better the car will accelerate. The P-N Junction Photovoltaic panels convert sunlight into electricity. The most important feature of the solar cell is its layered structure. The bulk of the cell is made of "P-type" silicon. P-type silicon is mostly pure silicon that contains a small amount of impurity (or dopant), typically Boron, which gives the material a special electrical characteristic, a deficit of electrons. On top of the P-type substrate is a layer of "N-type" silicon. This layer is nearly pure silicon, but containing a small amount of a different dopant. The characteristic of this layer is that it has a surplus of electrons. The interface between these layers is known as a P-N junction, and this is the central feature of the solar cell. It has very special electrical properties. P-N junctions have the characteristic that they behave like a one-way door for electricity. Because of this characteristic they are used in all types of electronic equipment. If voltage is applied across the P-N junction with the positive connected to the P side, a small residual potential must be overcome, then the current through the junction increases rapidly. The junction is "forward biased". If voltage is applied in the reverse direction across the P-N junction, current flow is blocked. The junction is "reverse biased". One can plot a graph of current voltage for a diode, and it will appear as shown below and it is known as the "I-V" characteristic of the device.

9

When photons of light pass into silicon, they have some probability of being absorbed. When they are absorbed, the effect is to knock an electron into a higher energy state. This free electron is likely to roam around the silicon's crystallne structure for a while. It also creates a "hole", and in fact the hole can roam around the silicon, too. If the photon is absorbed far away from the P-N junction, the hole and electron simply recombine and produce heat. If it occurs near the junction, the effect is that additional current is injected into the junction. The amount of current is directly proportional to the amount of light that falls on the cell. The extra current through the junction causes the voltage across it to increase. Take the I-V characteristic of the P-N junction, shift it downward by an amount corresponding to the light-injected current, and flip it, and the result is the following I-V curve that is often presented for PV cells and the graph represents the behavior of a solar cell at particular intensities of solar radiation.

10

These are very important curves. Note the point at which a curve intersects the vertical axis. This is known as the short circuit condition, and it defines how the cell operates if a wire is connected between its terminals, shorting it out. The current flow here is known as I sc. Because there is no voltage, the cell delivers no power. Now note the point at which a curve intersects the horizontal axis. This is where the cell operates if it is unconnected. This is known as the open circuit condition, and the voltage produced is denoted Voc. Because the current is zero, no power is delivered. For each point on the graph, the voltage and current can be multiplied to calculate power. Note that the power is maximum at a single operating point. This is known as the "Maximum Power Point", or MPP. If one is to get the most out of their solar cells, it is essential to operate around the MPP. The quality of a PV cell is often rated with a characteristic called its "Fill Factor". This is defined as the maximum power produced divided by the product of Isc and Voc. One can see that the Fill Factor will always be less than 1. As mentioned, the Solar Sprint PV panel is actually 6 cells connected in series. The following graph shows an approximation of the I-V and power output curves for these panels.

11

Sun Angle It has been seen that the PV panel's output current is proportional to the solar radiation that strikes it. It is important to understand that this radiation level is reduced if the panel is not pointing directly at the sun. Like other parts of this problem, the amount that it is reduced can be calculated. The cell’s output must be multiplied by the cosine of the angle of incidence of the incoming light. When designing a Solar Sprint car, a possible design feature would be the ability to adjust the angle of the PV cell. The following figure shows how the 12

angle of incidence affects PV cell output. Note that small variations in angle do not reduce output very much. Even a 60 degree angle of incidence only makes the output decline by one half.

Electrical Model of the PV Cell We can define an electrical circuit that acts just like our ideal solar cell, and draw its schematic diagram. It is simply a current source in parallel with a diode or P-N junction. A current source is a device that produces a constant current. In this case, the current is proportional to the intensity of light that falls on the cell.

The diode has the I-V characteristic that we graphed above. If we do not connect the cell to anything, current from the current source just circulates back around through the diode. A voltage is created across the diode. This is the open circuit condition. If we short the cell, all of the current is diverted away from the diode and flows through the short. Voltage output is zero. This is the short circuit condition. Real solar cells have characteristics that degrade their performance compared to the ideal. In particular, their wiring has some resistance to the flow of current. This is represented electrically by a resistor in series. Likewise, there is some resistance that is in parallel with the current source and diode that drains some amount of power from the cell. To be more complete, then, the equivalent circuit is shown below.

13

Photometry The human eye is not equally sensitive to all wavelengths of visible light. Photometry attempts to account for this by weighing the measured power at each wavelength with a factor that represents how sensitive the eye is at that wavelength. Photometry is the science of the measurement of light, in terms of its perceived brightness to the human eye. It is distinct from radiometry, which is the science of measurement of radiant energy in terms of absolute power; rather, in photometry, the radiant power at each wavelength is weighted by a luminosity function that models human brightness sensitivity. Typically, this weighting function is the photopic sensitivity function, although the scotopic function and others may also be applied in the same way. Photometry is typically based on the eye's photopic response, and so photometric measurements may not accurately indicate the perceived brightness of sources in dim lighting conditions where colors are not discernible, such as under just moonlight or starlight. The standardized model of the eye's response to light as a function of wavelength is given by the luminosity function. Note that the eye has different responses as a function of wavelength when it is adapted to light conditions and dark conditions. Mesopic vision occurs between these limits and is not well characterized for spectral response. Photopic vision is characteristic of the eye's response at luminance levels over three candela per square metre. Scotopic vision occurs below 2 × 10-5 cd/m2.

14

Many different units of measure are used for photometric measurements. The study of the chemical effects of ultraviolet radiation lead to the characterization using the total dose or actinometric units expressed in photons per second. The use of the human eye as a detector leads to photometric units, weighted by the eye's response characteristic. Measurement of the effects of electromagnetic radiation became a field of study as early as the end of 18 th century. Measurement techniques varied depending on the effects under study and gave rise to different nomenclature. Total heating effect of infrared radiation as measured by thermometers lead to development of radiometric units in terms of total energy and power. People sometimes ask why there need to be so many different units, or ask for conversions between units that can't be converted.

Because of the ways in which light propagates through three-dimensional space; spreading out, becoming concentrated, reflecting off shiny or matte surfaces; and because light consists of many different wavelengths, the number of fundamentally different kinds of light measurement that can be made is large, and so are the numbers of quantities and units that represent them. For example, offices are typically "brightly" illuminated by an array of many recessed fluorescent lights for a combined high luminous flux. A laser pointer has very low luminous flux (that is, it could not illuminate a room) but is blindingly "bright" in one direction. There are two parallel systems of quantities known as photometric and radiometric quantities. Every quantity in one system has an analogous quantity in the other system. Some examples of parallel quantities include: 

Luminance (photometric) and radiance (radiometric)



Luminous flux (photometric) and radiant flux (radiometric)



Luminous intensity (photometric) and radiant intensity (radiometric)

In photometric quantities every wavelength is weighted according to how sensitive the human eye is to it, while radiometric quantities use un-weighted 15

absolute power. For example, the eye responds much more strongly to green light than to red, so a green source will have greater luminous flux than a red source with the same radiant flux would. Radiant energy outside the visible spectrum does not contribute to photometric quantities at all, so for example a 1000 watt space heater may put out a great deal of radiant flux (1000 watts, in fact), but as a light source it puts out very few lumens because most of the energy is in the infrared, leaving only a dim red. A comparison of the watt and the lumen illustrates the distinction between radiometric and photometric units. The watt is a unit of power. Watts are units of radiant flux while lumens are units of luminous flux. We are accustomed to thinking of light bulbs in terms of power in watts. This power is not a measure of the amount of light output, but rather indicates how much energy the bulb will use. Because incandescent bulbs sold for "general service" all have fairly similar characteristics, power consumption provides a rough guide to the light output of incandescent bulbs. Watts can also be a direct measure of output. In a radiometric sense, an incandescent light bulb is about 80 % efficient: 20 % of the energy is lost. The remainder is emitted as radiation, mostly in the infrared. Thus, a 60 watt light bulb emits a total radiant flux of about 45 watts. Incandescent bulbs are, in fact, sometimes used as heat sources, but usually they are used for the purpose of providing light. As such, they are very inefficient, because most of the radiant energy they emit is invisible infrared. A compact fluorescent lamp can provide light comparable to a 60 watt incandescent while consuming as little as 15 watts of electricity. The lumen is the photometric unit of light output. The package of a 60 watt incandescent bulb indicates that it provides about 900 lumens, as does the package of the 15 watt compact fluorescent. The lumen is defined as amount of light given into one steradian by a point source of one candela strength while the candela, a base SI unit, is defined as the luminous intensity of a source of 16

monochromatic radiation, of frequency 540 terahertz, and a radiant intensity of 1/683 watts per steradian. (540 THz corresponds to about 555 nanometres, the wavelength, in the green, to which the human eye is most sensitive). The number 1/683 was chosen to make the candela about equal to the standard candle, the unit which it superseded. Combining these definitions, we see that 1/683 watt of 555 nanometre green light provides one lumen. The relation between watts and lumens is not just a simple scaling factor. We know this already, because the 60 watt incandescent bulb and the 15 watt compact fluorescent can both provide 900 lumens. The definition tells us that 1 watt of pure green 555 nm light is "worth" 683 lumens. It does not say anything about other wavelengths. Because lumens are photometric units, their relationship to watts depends on the wavelength according to how visible the wavelength is. Infrared and ultraviolet radiation, for example, are invisible and do not count. One watt of infrared radiation is worth zero lumens. Within the visible spectrum, wavelengths of light are weighted according to a function called the "photopic spectral luminous efficiency." According to this function, 700 nm red light is only about 4% as efficient as 555 nm green light. Thus, one watt of 700 nm red light is "worth" only 27 lumens. Because of the summation over the visual portion of the EM spectrum that is part of this weighting, the unit of "lumen" is color-blind: there is no way to tell what color a lumen will appear. This is equivalent to evaluating groceries by number of bags: there is no information about the specific content, just a number that refers to the total weighted quantity. Photometric measurement is based on photo-detectors, devices that produce an electric signal when exposed to light. Simple applications of this technology include switching luminaires on and off based on ambient light conditions, and light meters, used to measure the total amount of light incident on a point. More complex forms of photometric measurement are used frequently within 17

the lighting industry. Spherical photometers can be used to measure the directional luminous flux produced by lamps, and consist of a large-diameter globe with a lamp mounted at its center. A photocell rotates about the lamp in three axes, measuring the output of the lamp from all sides. Lamps and lighting fixtures are tested using gonio-photometers and rotating mirror photometers, which keep the photocell stationary at a sufficient distance that the luminaire can be considered a point source. Rotating mirror photometers use a motorized system of mirrors to reflect light emanating from the luminaire in all directions to the distant photocell; gonio-photometers use a rotating 2-axis table to change the orientation of the luminaire with respect to the photocell. In either case, luminous intensity is tabulated from this data and used in lighting design.

18

CHAPTER TWO SOLID STATE PHYSICS

We know that matter is composed of compounds and elements where elements are the basic materials found in nature. When elements are combined to form a new material, we form a compound. To understand how semiconductors function, one must have a good knowledge of their atomic structure. The smallest particle that an element can be reduced to and still retain its properties is called an atom. Although atoms of different elements have different properties, they all contain the same sub-atomic particles.

There are a number of different sub-atomic particles but only three of these are of interest in basic electronics ie the proton, the neutron and the electron. The protons and the neutrons are contained in the nucleus of the atom, and the electrons orbit around the nucleus where electrons and protons are the only particles that have the electrical properties while neutrons have no electrical charge. An atom has the same number of electrons and protons and so it is electrically neutral. If an atom does have more electrons, it is called a negative ion and if it has more protons, it is called a positive ion. The periodic table below shows elements in group one to group O. To understand the location and energy of each electron in an atom, one must have the knowledge of the following four quantum numbers.

19

Table 1: The Periodic table

Principal Quantum Number (n) The principal quantum number characterises the average distance of an electron from the nucleus and corresponds to the principal energy level in which electron resides. Obviously, n gives some idea about the position of the electron around the nucleus. n can have positive integer values starting from 1, i.e. n = 1, 2, 3,...... The principal energy levels or shells having different values of n are designated by the letters K, L, M, N and so on. The maximum numbers of the electrons that can be accomodated in a shell corresponding to n is equal to 2n2.

Azimuthal Quantum Number (l) It is also called as orbital angular momentum quantum number and gives a measure of the angular momentum of an electron in the orbit. Physically, l indicates the shape of the classical orbit. For a given value of n, l can take all positive integer values from o to (n – l). The particular l value defines the subshell, and the sub shells with l = 0, 1, 2, 3,..... are designated as s, p, d, f, g,...... respectively. For a given principal quantum number, the energies of various subshells are in the order of s < p < d < f ....... Obviously, an electron in the s-sub shell has lower energy than in the p sub shell with same n. The magnitude of this angular momentum if n = l(l + 1) . 20

Magnetic Quantum Number (ml) Magnetic Quantum Number is also called as orbital magnetic number and this determines the preferred orientation of the orbitals in space with respect to an applied magnetic field. We know that the magnetic moment of an electron due to its orbital motion gives rise to a magnetic field which can interact with an external magnetic field. The electrons orient themselves in certain preferred region of space around the nucleus under the influence of the external field. For a given value of l, ml can take integer values between – 1 to + 1 including 0, i.e. total allowed values of ml are (2l + 1). The magnitude of the component of angular momentum along the direction of the magnetic field is ml.

Spin Quantum Number (ms) The spin of the electron produces a spin magnetic moment, which can either be parallel or anti parallel to the surrounding magnetic field. For an electron there are two spin states. Electron is spinning about its own axis in the atom.. Spin quantum number is concerned with the spinning of the electron about its own axis. Thus ms can take only two possible states, +½ or – ½.

Pauli Exclusion Principle This Principle states that in an atom no two electrons can exist in the same quantum state, i.e. in an atom there cannot be two electrons with the same value of all the four quantum numbers. With the help of this principle, one can write the configuration of electrons. We may note that all the electrons with the same value of n constitute a shell and a shell can have a maximum of 2n2 electrons.

Crystals A crystal is made of small units reproduced many times and built into a regular array. Once we understand how atoms bond together, we can examine what types of matter can be formed. An obvious thing that can happen is that atoms can bond together to form regular crystals. The macroscopic morphology of a crystal can reflect its underlying structure. It is also possible that atoms will bind together to 21

form molecules and the molecules will stick together through weak Van der Waals bonds to form so-called molecular crystals.

Here, atoms are attracted to each other but not so strongly that they form permanent bonds. Liquids and gases have disordered configurations of molecules whereby the molecules are free to move around into new configurations. Somewhere midway between the idea of a crystal and the idea of a liquid, there is the possibility of amorphous solids and glasses. In this case the atoms are bonded strongly into position in a disordered configuration. Unlike in a liquid, the atoms cannot flow freely.

Many more possibilities exist. For example, one may have so-called liquidcrystals whereby the system orders in some ways but remains disordered in other ways. For example, in the crystalline (ordered) system in one direction can also remain disordered within each plane. One can also consider cases where the molecules are always oriented the same way but are at completely random positions (known as a “nematic”). There are huge varieties of possible liquid crystal phases of matter. In every case it is the interactions between the molecules (“bonding” of some type, whether it be weak or strong) that dictates the configurations.

One can even engineer artificial types of order which do not occur naturally. Each one of these types of matter has its own interesting properties. In a liquid crystalline phase, translational and rotational symmetry are broken to form a combination of discrete and continuous subgroups. For instance, a nematic liquid crystal is made up of elementary units which are line segments and these line segments point on average in the same direction but their positional distribution is as in a liquid state. Hence, a nematic phase breaks rotational invariance to the sub-group of rotations about the preferred direction and preserves the full translational invariance. Nematics are used in LCD displays. In a smectic phase, the line segments arrange themselves into layers thereby partially breaking the 22

translational symmetry so that discrete translations perpendicular to the layers and continuous translations along the layers remain unbroken.

In the smectic-A phase, the preferred orientational direction is the same as the direction perpendicular to the layers. In the smectic-C phase, these directions are different. In a hexatic phase, a two-dimensional system has broken orientational order but unbroken translational order locally and it looks like a triangular lattice. A quasi-crystal has rotational symmetry which is broken into a 5-fold discrete subgroups and translational order is completely broken. Polymers have extremely long molecules and they can exist in solution or in a chemical reaction where cross-links thereby forming a gel.

A glass is a rigid, `random' arrangement of atoms. Glasses are some-what like `snapshots' of liquids and are probably in non-equilibrium phases. First of all, the study of solids is one of the most successful branches of physics – both in terms of how completely we understand them and also in terms of what we have been able to do practically with this understanding (For example, the entire modern semiconductor industry is a testament to how successful our understanding of solids is). More importantly, however, the physics that we learn by studying solids forms an excellent starting point for trying to understand the many more complex forms of matter that exist.

It is the nature of some solid materials to form themselves into structures called crystals which have characteristic geometric shapes. These are the crystalline substances. Exact opposites are those other solids which form into shapeless structures and are said to be plastic, non crystalline, or amorphous. Quartz is a familiar example of crystalline substances where as a block of rubber is amorphous. Elements and compounds both can be crystalline. So can be metals and non metals. Virtually all minerals are crystalline.

23

Externally, a crystal has several flat faces which are arranged symmetrically with respect to each other. Internally, it has a certain orderly arrangement of atoms in a repeating system called a lattice. Both externally and internally, a single crystal of a given true crystalline material looks like all other crystals of that material. A single crystal may be large to the point of hungeness or it may be so small as to be visible only with the aid of magnifier. Inside the crystal lattice, certain loosely bound electrons (called valence electrons) in the outer most shells of one atom align themselves with similar electrons in adjacent atoms to form covalent bonds which hold the atoms together in the orderly structure of the lattice.

Orbitals in Silicon atom

Germanium (Ge) and Silicon (Si) are two very typical semiconductors. Thus, in any covalent bond there are shared electrons, so called because they are shared by neighbouring atoms. In Germanium the atom has 32 electrons distributed in four orbits whereas in silicon there are 14 electrons distributed in three orbits. The outermost orbit is called the valence orbit contain 4 electrons in each case.The crystals are tetrahedral in structure and each atom shares one of its electrons with its neighbour. Such a sharing of its electrons between two neighbouring atoms is called a covalent bond.

At very low temperature near absolute zero all the electrons in the atoms are tied up strongly by these bonds, but with the rise of temperature, a few electrons break 24

away from some of the covalent bonds and get themselves freed creating vacant spaces, deficient of electrons known as holes. A hole is equivalent to a net positive charge-equal to that of electron. Whenever a free electron is generated, a hole is created simultaneously, i.e., free electrons and holes are always generated in pairs. Obviously, the concentration of free electrons and holes will always be equal in an intrinsic semiconductor. This type of generation of free-electron hole pairs in semiconductors is called as thermal generation. More will be discussed in chapter two.

Bloch’s Theorem Bloch’s Theorem helps in the calcualtions of the k-space expansion through which band calculations are made. In order to calculate the intrinsic carrier concentration, we first calculate the number of electrons excited into the conduction band at any temperature T in K and which in turn are free to migrate in the crystal. In carrying out these calculations it will be assumed that the electrons in the conduction band behave as if they are free with effective mass me* and energy will be measured from the top of the valence band.

In addition to the density of states function we also require the probability function for calculating the electron density in the conduction band. The appropriate probability function is the Fermi function f(E) as stated in the next chapter that Fermi level E F is in the middle of the band gap in the case of intrinsic semiconductors.If (E – EF) < 4kT, i.e., the lower edge of the conduction band is about 4kBT above EF, and Fermi function f(E) reduces to; F(E) = exp (EF – E)/ The number of electrons per unit volume having their energy in the range dE in the conduction band can be obtained from ne(E) dE = Ne(E) f(E)

25

Much more details are covered in the next chapter on brilloun zones, density of states, density of states in k-space, generic formula for density of state, density of state effective masses, For conduction band, valence band; average over directions as well as light hole and heavy hole bands. Conductivity details and effective mass defined in transport theory and calculations involving isotropic materials will not be covered in this book.

Thermal Physics In thermal equilibrium, carriers are not sitting still but undergo collisions with vibrating Si atoms (Brownian motion). They electrostatically interact with charged dopants and with each other. Their characteristic time constant of thermal motion i.e. thier mean-free- time between collisions: Ʈc = collision time [s] In between collisions, carriers acquire high velocity: vth = thermal velocity [cm/s] Characteristic length of thermal motion; λ = mean free path [cm] λ = vth Ʈc Put numbers for Si at 300 K; Ʈc ≈ 10−14 ≈ 10−13 s vth ' ≈ 107 cm/s λ ≈ 1 ≈10 nm.

Independent Particle Approximation The famous Fermi-Dirac statistics is based on this approximation. It is used in the solid state physics and in even the fairly advanced device physics is based on independent particle approximation. Independent particle approximation is the most fundamental approximation in almost all of the device physics. The other terms used to refer to it is the single particle and the one particle approximation. Band structures and also the density of states (DOS) concept are based on this assumption. Other theories that rely on one particle approximation are the semi26

classical transport theory and even the quantum mechanical portions of semiclassical theory on scattering rates and particle velocity calculations.

Band to band scattering also assumes independent particle approximation but for band to band scattering and although it’s possible to write full multi-particle approximation thorugh state perturbation theory, tractable Harmiltonian can not be written without independent particle assumption and that is why this chapter briefly summaries this concept. It is also worth to note that even the fairly advanced quantum kinetic transport theory is based on on independent particle approximation theory and calculations.

In this approximation, the theoretically assumed way to write more complete equations will be surveyed. It might omit the screen effect but it will be still very close

approximation

approaches.

To

understand

independent

particle

approximation well, basics of Fermi distributions using Fermi Dirac distribution is necessary so that Fermi distributions for metals and semi-metals, insulators and semiconductors can be easily be understood. Boltzmann distribution basics are also essential. Impurity ionization: Probability of an impurity atom being in ionized state is given as; ʃ(Nd+) = 1/(1+gd exp ((Ed-Ef)/τ) Here, gd is the degeneracy of donor level which basically means in how many distinguishable ways the donor can give out an electron. Here we calculate the probability for a bunch of orbitals and not for single orbital with added restriction that only single electron can be absent. One can usually always define a new level Ed and thus get rid of gd so that the actually gd which is quite ambiguous and difficult to calculate is avoided. Usually Ed for donor it is 24, for acceptor it is 45, and is default to 1 for deep level traps when adjusted accordingly. In armorphous materials, deep level traps which originate from dangling bonds need to be more properly treated as armophoteric states and correlation energies between two more

27

elctrons included. Calculating their statistic results is a bit more complex to express for three charged states of those traps.

Equilibrium Carrier Statistics One of the important physical laws that one can exploit to understand the charge carrier density profiles in a semiconductor is the charge neutrality condition. From basic electromagnetics, we know that ∆E = ρ/E0 (Gauss law) needs to be correct under all classical circumstances.

If we know that E= 0 everywhere, then

obviously we can enforce ρ = 0 everywhere in a semiconductor. It should be stressed that the condition of zero electric field is sufficient but not a necessary pre-requisite for charge neutrality. Similarly in any P-N junction device, if one moves far away from the metallurgical junction on either side, this condition of charge neutrality can be enforced easily.

Somewhat closer to the metallurgical junction who is also known as quasi-neitral region, the density of injected minority carrier is not negligle because of a fast dielectric relaxation in high conductivity semiconductors, majority carries that almost neutralize semiconductor. Such a condition is known as quasi-neutrality and it can be shown that the electric field is quite negligible in quasi-neutral regions. One might think that the equilibrium concentration is decided by the fact that in equilibrium, the thermal generation rate should balance the thermal recombination rate. As these rates are dependent on concentrations one might think that enforcing this balance condition, one might be able to calculate the equilibrium concentrations.

Whenever electron and hole concentrations are calculated from a single flat Fermi level, their concentrations would always satisfy this condition of equal upward and downward transtions at all temperatures. The position of the Fermi level is not decided by this condition of equal upward and downward transtions but by the condition of charge neutrality. The second important point is that enforcing condition of charge neutrality is not as simple as it might seem on first sight. We 28

would explore this second complexity in more details when dealing with doped semiconductors. This second complexity does not arise in intrinsic semiconductor. One can exploit the expression for Fermi Direc distribution together with the condition of charge neutrality to calculate these concentrations in equilibrium. Let us say that we have; no- = Nc /(1+ exp((Ec - Ef)/KT)) p0+ = Nv /(1+ exp((Ef - Ev)/KT)) for electron concentration and hole concentration respectively. Now enforcing charge neutrality condition, no- = p0+ = ni, one can always calculate the position of intrinsic Fermi level, E f = Ei as well as the value of intrinsic carrier concentrations using either successive approximation or the graphical approach to find the exact point of neutrality. Usually this problem can be solved much more easily by assuming non-degenerate Boltzmann statistics and writing; ni = Nc exp(-(Ec –Ei)/KT) = Nv exp (-(Ei - Ev)/KT). These relations would eventually provide us the position of Ei with respect to Ec and Ec as well as the absolute value of ni. Suppose we dope this semiconductor with p-type impurities. A certain fraction of impurity atoms might ionize by accepting electrons from the valence band of the semiconductor.

A very intuitive way of understanding this is to think about what happens to the equilibrium between upward transition and downward transition rates. These rates were balanced for n0- = ni and p0+= ni. Now when we add p1+ to hole concentration the upward rate is more or less the same but the downward transition rate increases significantly. Hence addition of acceptor or impurities does not affect even the electron concentration. So the second important point that we mentioned above is not to enforce partial step - step charge neutrality condition but look at the complete process. 29

This alternative explanation is mathematically more helpful because rate balance argument does not provide a unique solution for n0- = ni and p0+ = ni since it does not follow from a single flat Fermi level. But once p concentration is increased by adding a p1+ to it, it would not follow from same Fermi level as used for electrons. These upward and downward rates are automatically balanced whenever we calculate these concentrations from a single flat Fermi level. All we need to do is to find a new location of Fermi level that satisfies statistical concentration expressions as the charge neutrality condition. So to enforce the charge neutrality, we should have NA+n = p where each term depends on the position of the Fermi level.

For those who feel that the use of Fermi-Direc distribution is more accurate than the use of Boltzmann approximation, this graphical or successive approximation method of solving charge neutrality problem can of course be applied with equal ease with Fermi-Dirac expressions and because it does not necessarily need one to go for Boltzman approximation. So one can always solve such an equation exactly either graphically or using successive approximation method to find the position the Fermi level. One can go on and consider even more involved problems in which we have two or more types of impurities etc. Irrespective of the complexity of the problem, one can always solve the complete charge neutrality equation numerically once doping concentrations are been given.

In most of the practical situations we have one kind of dopant. But, in the most general case of heavy doping, the most fundamental assumption of independent particle approximation in entire device physics starts falling apart. Moreover many other effects like band-tailing needs to be taken care of thus most of the simple device physics generally works accurately only with non-degenerate doping where Boltzmann approximation is as good. So the neutrality equation would have only 3 terms. So let us discuss the case when we have donors as dopants. Moreover in almost all situations, you can always make one of the two

30

assumptions- either you can assume p ≈ 0 or you can assume Nd+ ≈ Nd+ and to maintain the mass action law np = ni2 would ensure; p ≈ ni2/Nd+ > ni (T = 0 K) as T increases, ionization increases. At the same time Ef approaches Ei but carrier density also increases. This increase is initially a direct result of more ionization. Eventually intrinsic carrier concentration becomes more than dopant concentration and the increase in carrier concentration is because of the increase in intrinsic carrier concentration. As Nd increases E f tends to Ec, carrier density increases, but ionization fraction decreases when temperature is increased.

Non-Uniform Doping and Quasi-Neutrality For Gaussian profile of doping the free carriers would also be Gaussian distributed but the free carriers profile would be broader than the doping profile. Hence semiconductors would not be locally neutral. When semiconductors are uniformly doped one can invoke condition of charge neutrality to calculate the relative position of Fermi level with respect to the position of conduction band edge. But when doping is in-homogeneous (as in abrupt pn junctions) then the semiconductor, in theory is not locally neutral. Generally the carriers want to diffuse away from high concentration regions and leave behind the ionized dopants. This creates a built-in field that tries to stop further diffusion.

If semiconductor was locally neutral at all locations, then we have an abrupt step in the condition and valance band edges then Fermi the level would be flat all across. As we know that this does not happen. Bands bend more smoothly over a distance given by the depletion width and hence within this depletion region, a semiconductor is not locally neutral. For example, assume that infinitely long uniform semiconductor is doped with Gaussian profile donor dopants. Consider 32

simple case of complete ionization. Let us first make ‘quasi neutrality’ approximation. Quasi neutrality approximation assumes that semiconductor is still ‘locally’ neutral. Hence; ρ (x) = 0 n(x) ≈ N+d (x) ≈ Nd(x) One can now easily calculate the profile of E c (x) and hence the profile of electric field. One can then evaluate the expression; ρ (x) = ɛ∆.E.

If this is really small then that justifies the quasi neutrality assumption. If this is not the case then one can use successive approximation and approximate by the equation below and calculate the new band edge profile; n(x) = Nd(x) – ρ (x)

Abrupt pn Junction and Depletion Approximation There are two theoretical inconsistencies with the charge profiles provided by the depletion approximation. Let us now discuss a special case but very important on non-uniform doping i.e. an abrupt pn junction. Carrier concentration inside depletion region is not zero. This can easily be calculated (as second step of successive approximation) as the band edge profiles are known relative to the position of the Fermi level. Near the edges of the depletion region ∆.E ≠ 0 and hence the region immediately outside the depletion region can not be locally neutral.

Also the electric field cannot take an abrupt triangular form. The profile of electric field has to be smooth so that its derivative gradually goes to zero. Intergral form of Gauss’ law is a bit deceptive here and one should actually use the differential form. The calculated profile of n(x) as second successive approximation would be smooth. Hence the second derivative of Ec(x) or the first derivative of electric file needs to be smooth. This is why regions outside the depletion regions are called quasi-neutral regions.

33

Inhomgeneity in Band Edges In electromagnetic theory, vacuum level is treated as the energy of an electron at infinity away from any charge distribution i.e basically a point where electric field is zero. All other energy levels are refered to this energy. But in semiconductors we define vacuum level as the energy of an electron just outside the semiconductor. To understand the technicalities involved, let us look at an ideal long abrupt pn junction. We have built-in electric filed in the depletion region from n side pointing towards the p side perpendicular to the junction. Ideally, there is no component of electric field which is parallel to the junction.

If we look at the region just outside the semiconductor but somewhere around junction. Componemt of electric field that is perpendicular to the junction at tangential to semiconductor surface is continuous as we move from inside the semiconductor to just outside the semiconductor. Hence electrostatic potential just outside the semiconductor and near p region is smaller compared to the electrostatic potential just outside the semiconductor and near n region. And this potential difference is exactly the same as the potential difference between equivalent points inside the semiconductor.

We call the energy of electron just outside semiconductor as vacuum level and vacuum level changes as we move from outside p region to outside n region. This change is exactly the same as the change in the conduction band edge or the valance band edge. Hence the energy difference between this vacuum level and the conduction band edge, for example is the same whether we are looking at homogeneous p type material or homogeneous n type material or homo-pnjunction.

There is one thing that gets very confusing with this definition. If we stick with ideal description of an infitely long pn junction then there is no componemt of electric field parallel to the junction even outside semiconductor. Hence even if I move infinitely way from the semiconductor at the outside, we would find that we 34

still have a perpendicular electric field in the region of the junction. Hence even infinitely away, electrostatic potential above the p region is different from that above in the n region. The only problem is that, in reality we do have parallel componemt of electric filed and this componemt makes the potential at infinity equal on two sides but we do not have to worry about all these as long as we remember that vacuum level in semiconductor is not defined infinitly away.

Band Bending Assume that we have a semi-infinitely long uniformly doped semiconductor stariting from x = 0 to x = infinite and let us assume that non-uniform doping can be handled in exactly same fashion. Suppose it is given that at x = 0, we will have an electric field of E inside the semiconductor. This electric filed would change the potential inside the semiconductor and would try to move the mobile carriers. We want to know V(x) and equivalently Ec(x) given an Ec (0). One can simply use the Gauss’s law to calculate this profile but before we apply Gauss’s law we need to figure out in which regime the semiconductor is in.

Suppose we have an n type semiconductor and electric filed at x = 0 is pointing towards positive x direction. This means negative changes would be pulled towards x = 0 surface and this surface needs to be more negative. Aslo electrostatic potential near x = 0 surface needs to be higher compared to dep inside semiconductor towards x→∞. This means electron electrosatitc energy near x = 0 surface needs to be lower compared to deep inside semiconductor towards x→∞. Hence bands need to bend downwards near x = 0 surface.

A small amount of band bending can very easily provide any amount of negative charges required by the Gauss’s law on a very thin layer of semiconductor around x = 0. So the electric filed inside semiconductor would drop to zero almost abruptly at x = 0+ and semiconductor beyond that point would be neutral. Bands in an n type semiconductor would bend downwards near x = 0 surface to make that region negatively charged. 35

Suppose we have an n type semiconductor and electric field at x = 0 is pointing towards negative x direction. We further assume that electric field at x = 0 is not very strong. This means negative charges would be pushed away from the x = 0 surface and this surface needs to be more positive. Also electrostatic potential near x = 0 surface needs to be lower compared to deep inside semiconductor towards x→∞. This means electron electrostatic energy near x = 0 surface needs to be higher compared to deep inside the semiconductor towards x→∞. Hence bands need to bend upwards near x = 0 surface.

A demand of positive charges by the Gauss’s law can only be full filled in limited quantity by the semiconductor. Even if all the negative charges are depleted, semiconductor can only provide Nd+ positive charges per unit volume. A depletion approximation can be evoked with loss in accuracy. One can then use successive approximation to fix this loss in accuracy. Under depletion approximation, electric field would drop linearly towards zero as we go inside the semiconductor starting x = 0. Rate of decrease of electric field would depend upon the doping concentration.

Situation under inversion is similar to situation under depletion except that the intensity of electric filed at x = 0 is very strong. Hence the charges required by the Gauss’s law are quite significant. Bands are required to bend so much that the surface near x = 0 inverts. After inversion, any amount of positive charge can easily be supplied.

Schottky- Mott Limit Assume that we are looking at a free surface at x = 0 of uniformly doped semiinfinite starting from x = 0 to x = ∞ and it is an n type semiconductor. Let us first assume that the two materials are at atomic contact and there are no surface states. We want to claim that the vacuum level is well defined, unique and continuous across the interface of two materials. This idealized limiting situation is called Schotty-Mott limit. In such a scenario electron injection barrier or the hole 36

injection barrier across Schotty-Mott (MS) junction is solely determined by the work function of metal and the electron affinity (EA) or the ionization potential (IP) of the semiconducting material.

It does not depend upon the doping of the material or the profile of the band bending. This is an important point to keep in mind that the barrier heights do not depend on how bands bend. Similarly, for semiconductor- semiconductor heterojunctions, the barrier heights are determined by the difference in the EA or IP or the two materials. Now how is the band profiles determined inside semiconductor for such an MS junction, for example?

To clarify the physics involved, let us first consider a case where semiconductor surface is in depletion. Since semiconductor doping is known the position of band-edges relative to Fermi level inside the bulk of the material. Once the barrier height is fixed location of band edges at the surface is fixed. These two boundary conditions complete determine the exact profile of the band. Note that under depletion approximation, free carrier density inside semiconductor is negligibly small. Hence the amount of charges would be balanced by the equal and opposite charges appearing on the metal surface.

Effects of Surface States Let us now consider a surface that can have many surface states which can either be donor like or acceptor like for the time being. Let us forget about n and p free charges and only look at fixed ionized charges at the surface due to these surface states. One can make Kittel-Kromer like diagrams to determine the position of Fermi level at the semiconductor surface to ensure that surface has zero fixed ionized charges. This position of Fermi level is called charge neutrality level (CNL). Usually if Fermi level is above CNL, then surface would have more fixed negative charges and if Fermi level is below CNL, then surface would have more fixed charges. We can draw a plot of surface fixed charge density versus location of Fermi level. 37

On the other hand, one can also calculate the depletion charges inside the depletion layer (assuming semiconductor is in depletion regime) for any position of Fermi level at the surface. Actual position of Fermi level would be decided by the equality of these two charges. Surface states are distributed inside the band gap of the semiconductor and these distributions of acceptor like states and donor like states usually peaks very sharply for certain energy level.

If we plot surface fixed charge density versus position of Fermi level, then we would see that CNL would be very close to this energy level where state density peaks. Further charge versus Fermi level diagram would show that surface changes from very negative to zero to very positive as Fermi level moves very little around this energy level. Hence location of Fermi level gets pinned. Depletion charges are not very strong function of Fermi level position. For location of Fermi level near CNL we can calculate the approximate charges inside depletion layer. As Fermi level moves very little around CNL, surface can easily supply whatever charges are demanded by the Gauss’s law. This is why we say surface states pins the Fermi level.

Effects of Surface states on Vacuum level In case of surface states, vacuum level would show a discontinuity between two materials interface. The energy discontinuity would primarily depend upon the thickness of the interfacial layer, permittivity of this interfacial layer and the surface density of states. One can understand the vacuum level discontinuity by considering the physically separated surface in thermal equilibrium. Physically separated surfaces can be brought to thermal equilibrium by connecting two materilas by a wire so that Fermi level would equate.

First consider the Schottky-Mott limit. Total potential between the bulk semiconductor and the metal surface would be equal to the differences in Fermi level positions before connecting them with a wire. This potential would divide 38

between free space interfacial layer and the semiconductor depletion layer. As the thickness of the depletion layer reduces, entire potential would drop across the semiconductor depletion layer. Metal charges are then balanced by the depletion charges.

On the other hand in the case of very high density of surface states, a potential drop across the semiconductor remains almost constant. As the thickness of the interfacial layer reduces, the electric field inside interfacial layer keeps shooting up to keep the potential drop along the interfacial layer constant. Before connecting the wire, a surface state charge balances the depletion charges. After connecting the wire, additional surface state charges required to balance the metal charges are now easily accommodated without a significant change of the position of the Fermi level. The amount of metal charges and the additional surface state charges that appear after connecting the wire is determined simply by the additional potential drop that needs to be accommodated in the interfacial layer.

Non- Equilibrium Carrier Statistics To obtain a non-equilibrium steady state carrier distribution, or a time variation of the carrier distribution among various single-particle states, we in general need to solve the kinetics of the carriers. Under many practical situations, the distributions might be quite similar to the equilibrium distribution. This assertion can easily be justified through semi-classical transport theory or through Monte –Caro simulations. In real world situations we might be injecting very high energy elcrtons into a semiconductor. Or electrons might have been accelerated by a very high electric field. As these electrons collide with the lattice, both energy and momentum relaxes.

Usually momentum relxation time is smaller than the energy relaxation time in either band-to-band or within band relaxation time. In extreme cases we might have a delta distribution in energy (which may be random distribution in momentum space) in steady state situation. So in reality, in cutting edge device 39

geometries, carrier distribution can be quite different from equilibrium distributions.

Since band–to-band energy relaxation time is much longer than within band relaxation, one can assume a different carrier concentration in conductions band and in valance band. One can further assume that at every local region of these carriers are in equilibrium with the lattice. So average momentum is zero and average energy is 3/2KT. This concentration can be parametrized by a quasiFermi level. Usually the ensemble momentum (actually velocity) relaxation time is of the order of ps whereas ensemble band to band relaxation time depends a lot on the specific semiconductor under considerations.

It can be very long or relatively long but on average band transition time varies from ms to ns and hence even the smallest band to band relaxation time is much bigger than the largest momentum relaxation time. That is why we can say that momentum distribution of electrons would get randomized much faster than the holes and electron band to band relaxation. So when excess carriers are injected into the semiconductor, holes and electrons might not be able to relax and reach at their thermal equilibrium concentrations. But momentum distribution would surely reach the thermal equilibrium configuration. Since we do not care about actual distribution, it is fairly a good approximation for the mentioned reasons.

We can add parameters by assuming that energy relaxation with lattice is not as fast and the carrier temperature is different from lattice temperature. Electron density at any local region in a semiconductor can usually be treated as an ideal bunch of non-interacting particles freely assuming their local parabolic band and Boltzmann assuming non degenratte concentration of their gas states. Now we know that the average energy per particle pr unit volume for a classical (Boltzmann), ideal (non-intracting particles) and free gas is 3/2KT provide electrons are thermal equilibrium with the environment with which thet are exchanging energy. 40

Another improvement that is added to the average momentum is by shifting the distribution momentum space or by patching up technique using the shifted Maxwell approximation. Fermi distribution is among the energy states that can easily be converted to an even distribution in k-space however this distribution cannot give any current. Hence we can add up an odd part that would just be equivalent to a drift version. Usually drift is too small to affect the carrier densities in any way and therefore it is only used to calculate the current. Carrier density can still be calculated with the usual Maxwell distribution, may be with a different temperature. The recombination geration lifetime between dopant level and bands is probably very small compared to band band relaxation life time so it would be given by quasi Fermi distribution.

Solution to General Solid State Problem Band edge bending which is not the Fermi level is solely dependent on the electric field distribution. While drawing the band diagrams, the inherent assumption involved is to that assume that the external electric field does not vary over many unit cells. So we divide structure on many small parts, consider each part as equivalent to essentially an infinite crystal. On each small part, the electric field is different. The band bottom represent a single particle state with zero Blochvector, so energy of this single electron state is only the potential energy and moreover the spatial spread of this k = 0 state is going to be the same in each subsection and

hence only electric field would decide the potential energy and

therefore, band edges at each sub-section.

Although when you go up in the band, both the kinetic and potential energy changes, some people still approximately say that potential energy remains the same while kinetic energy changes. This is something similar to consider the electrons as a free Fermi gas of independent particles. Although this statement is never used in device physics, it might help.

41

Drift and Diffusion One needs to use the kinetic theory to come with the carrier flow equations. Thermodynamics can not help here. For example all that thermodynamics can tell us is that at equilibrium Fermi level has to be flat, but it cannot tell us that if there is a particular carrier profile which means a particular quasi-Fermi level profile then what is the owing current. Usually kinetic theory, it tells us that, current flows either because of the gradient in density (which may or may not be related to the gradient in intrinsic Fermi level doesn't matter) and gradient.

42

CHAPTER THREE

SOLAR CELL SEMICONDUCTOR MATERIALS A semiconductor material is said to be in thin film form only when it is built up as a thin layer on a solid support called substrate by a controlled condensation of the individual atomic, molecular, or ionic species. These semiconductors can be in bulk, wafer and thin film forms. Semiconductors materials exhibit a number of useful but also unique properties related to their electronic structure.. This can either be done directly by a physical process (CVD), or through a chemical process (CBD) or an electrochemical reaction.

Thin films are also regarded as two dimensional materials fabricated by the process of condensation of atoms, molecules or ions. This is what makes them have unique properties significantly different from their corresponding bulk materials. These unique properties are as a result of the changes that occur in their physical dimensions, geometry and microstructure. Thick films are prepared either by a direct application of solution dispersion or by a paste of the material on a substrate and then letting them dry irrespective of their thickness. These films have properties characteristically different from those of thin films.

In thin films there are deviations from the properties of the corresponding bulk materials that arise because of their small thickness, large surface-to-volume ratio and their unique physical structure which is a direct consequence of the growth process. Most of the new thin film technologies coming up are based on the use of materials with very thin film geometry because they tend to lower costs and also lower material consumption. It is currently known that a relatively small group of elements and compounds have an important electrical property called ‘semiconduction’. This is a property where they are neither good electrical conductors nor good electrical insulators but instead their ability to conduct electricity is intermediate and that is why these materials are called semiconductors.

43

Energy bands in solids From quantum mechanics it is made clear that the electrons of an isolated atom can have only discrete energy levels. Atoms constitute a crystal. When a number of atoms are brought together to form a crystal their discrete energy levels split to form a band. Most semiconductors occur in solid form. These solids are made up of atoms which in turn contains electrons. As the inter-atomic spacing decreases these bands merge together to form a single band. When the distance between these atoms approaches equilibrium distance the band splits again into two bands separated by a region called the forbidden gap, Eg. The upper band is called the conduction band (CB) and the lower band is called the valence band (VB). In conductors (i.e. metals) or the conduction band is either partially filled or overlaps the valence band so that there is no band-gap.

This courses the uppermost electrons in the partially filled band or electrons at the top of the valence band to move to the next higher available energy level when they gain kinetic energy. Through this current conduction readily occur in conductors. In semiconductors the bonds between neighbouring atoms are only moderately strong and therefore thermal vibrations will break some bonds and a free electron along with a free hole is produced. That is why the band gap of a semiconductor is not as large as that of an insulator. Therefore some electrons will be able to move from the valence band to the conduction band leaving holes in the valence band.

If an electric field is applied then both the electrons in the conduction band and the holes in the valence band will gain kinetic energy and conduct electricity. The valence electrons form strong bonds between neighbouring atoms in insulators. These strong bonds are difficult to break. This means that there are no free electrons to participate in current conduction. These courses the band gap in an insulator to be large. When you consider the band gap in an insulator, thermal energy or any external applied electric field cannot raise the uppermost electron in

44

the valence band to the conduction band. Using the energy parameters we get that in the donor case; EC- EF = KT ln {NC/ND} and in the acceptor case; EF- EV = KT ln {NV/NA} where the symbols have their usual meanings. Electron and hole densities in extrinsic semiconductors can therefore be expressed as: n = ni exp{ (EF-Ei)/KT} p = ni exp {(Ei- EF)/ KT} It is also noted that when both the donor and acceptor impurities are present simultaneously, the Fermi level adjusts itself to preserve charge neutrality according to the mass action law.

Semiconductor materials Between conductors and insulator materials lie semiconductors that have revolutionized electronic technology in the world and contributed a lot to clean energy. In solid state electronics materials can be classified as insulators, semiconductors and conductors based on their electrical conductivity. The electrical conductivity of semiconductors at room temperature is in the range of 102 - 10-9 Ω-cm-1 (i.e. electrical resistivity, ρ from 10-2 - 109 Ω-cm).

Conductivity of semiconductors is generally sensitive to temperature (why they are used as temperature sensors), illumination (used for solar cells), magnetic field and minute amount of impurity atoms (why they are doped). The energy gap, Eg of semiconductor materials is about 1. In the periodic table of elements we have elements that exist as semiconductors. These elemental semiconductors are composed of single species of atoms such as silicon (Si) and germanium (Ge). They occur in column IV of the periodic table shown in chapter one.

45

Crystal structure Accurate knowledge of solid structures is important in solid state physics because these structures influence the physical properties of solids. Solids are classified into two types in terms of their structures. They are either, crystalline solids that include single and polycrystalline solids, or they are non-crystalline solids. A great number of useful semiconductors have diamond or zinc-blende lattice structures which belong to the tetrahedral phases. In the above mentioned structure, each atom is surrounded by four equidistant nearest neighbour atoms that lie at the corners of a tetrahedron. The bonds between any two nearest neighbour atoms are formed by two electrons with opposite spins. Diamond and zinc-blende lattices are considered as two interpenetrating face-centred cubic lattices. Most of the III-V compounds are crystalline in the zincblende/ diamond structure while others are crystalline in the wurtzite or rock-salt structures. The wurtzite structure is tetrahedral with four equidistant nearest neighbour atoms similar to a zincblende structure. In the rocksalt structure each atom has six nearest neighbours. Some compounds such as zinc sulphide and cadmium sulphide can crystallize in both zincblende and wurtzite structures.

Energy band structure in semiconductors In the figure (a) the maximum in the valence band occurs at p = 0 while the minimum in the conduction band occurs along the [100] direction at p = pC. The electron at the conduction band (minimum) with zero kinetic energy can have crystal momentum different from zero (i.e. p = pC). Materials are classified as direct or indirect semiconductors depending on the band structure. When an electron makes a transition from the valence band to the conduction band it requires then, not only an energy change greater than E g but also some change in the crystal momentum, P.

46

This makes such semiconductors to be indirect semiconductors because a change in crystal momentum is required in their transitions. In figure (b) the maximum in the valence band and the minimum in the conduction band occurs at the same crystal momentum, p = 0 and in this case, an electron making a transition from the valence band to the conduction band can do so without a change in their momentum, P value. Such semiconductors are classified as direct semiconductors because their transition from the valence band to the conduction band does not require a change in crystal momentum for the electron.

Direct and indirect transition

semiconductors Conduction in these materials is due to the intrinsic processes without the influence of impurities. These are pure crystals of semiconductor materials. It occurs due to the electrons that are excited from the top of the valence band to the bottom of the conduction band by thermal energy. The promotion of some of the electrons across the band gap leaves some vacant electron sites in the valence band. These are called holes. The number of electrons excited across the gap can then be calculated from the Fermi-Dirac probability distribution as: F(E) = 1/ [ 1+ exp (E-EF)/KT]

47

The Fermi level (EF) for an intrinsic semiconductor lies midway in the forbidden gap. The probability of finding an electron here is 50 % even though energy levels at this point are forbidden. This means then that (E - EF) in the denominator of equation above is equal to Eg/2, where Eg is the magnitude of energy gap.

Figure 3.2: The Fermi level in an intrinsic semiconductor The unity (1) factor in the denominator can be ignored in comparison to the exponential term because of the fact that (E - EF) value is larger than the thermal energy (KT) at room temperature. Therefore the probability F(E) of an electron occupying energy level E becomes; F(E) = exp [-Eg/2KT] Using the bottom of conduction band as EC and the top of valance band as E V, the electron density in the conduction band can then be given by; n = NC exp [(EC-EF )/KT] in a similarly way, hole density in the valence band can be given by; p = NV exp [- (EF – EV)/ KT] where, NC, NV, are the effective density of states in the conduction and valence bands respectively.

As each excited electron leaves behind one hole an intrinsic semiconductor contains an equal number of holes in the valence band and electrons in the 48

conduction band (i.e. n = p = ni) where ni is intrinsic carrier density. Since the mass action law is given by, np = ni2

This means that the product of the two types of carriers will remain constant at a given temperature. This law is valid for both intrinsic and extrinsic semiconductor under thermal equilibrium condition. Using equation earlier discussed, the intrinsic carrier density can be written as; ni = (NcNv)1/2 exp [-Eg/2KT] where Eg ≡ (EC – EV). When an external field is applied the electrons that are excited into conduction band by thermal means accelerate using the vacant states available in the conduction band. At the same time the holes in the valence band move but in a direction opposite that of the electrons.

Extrinsic Semiconductors The process of deliberate addition of controlled quantities of impurities to a pure semiconductor is called doping. If a dopant (impurity) is introduced into an intrinsic semiconductor then it changes and becomes an extrinsic semiconductor. This is because impurity energy levels are introduced. If the doping process results in n-type semiconductors due to the addition of negative charges (electrons) then the impurity atoms are called donors and the energy levels of these electrons are called the donor energy levels (ED). If the doping process results into a p-type semiconductor then the impurity atoms are called acceptors and the energy levels of these holes are called the acceptor energy levels (EA).

49

(p-type)

(n-type) n-type and p–type doped semiconductor

The addition of impurities (doping) greatly increases the conductivity of most semiconductors as illustrated in the figure below;

Movement of electrons and holes during conduction

In general when an extrinsic semiconductor is analysed in terms of all the energy parameters used to describe semiconductors [(Ec, ED, EF, EA, Ei, and EV, it is observed that their interactions can be demonstrated as shown in the figure below;

50

Energy interactions in (a) n-type, (b) p-type doped materials

To excite electrons from the donor level into conduction band (CB) or holes from acceptor level into valence band (VB), energy known as ionization energy is required. The ionization energy (Ei) of donor electrons is approximately the same as the ionization energy of acceptor holes in the same crystal. When it is compared to the energy gap, the ionization energy of an impurity atom is very small. This means then that at room temperature a large fraction of the donor level electrons as well as acceptor level holes are excited into the conduction and valence band respectively. This fraction is much larger than the fraction of electrons excited from the valence band or that of holes created by these electrons due to the intrinsic process. The product of the number of electrons in the conduction band and the number of holes in the valence band must be constant for any semiconductor according to the law of mass action.

For an n-type semiconductor this condition drastically reduces the number of holes and therefore electrons in the conduction band become the majority charge carriers. In a similar way, the number of electrons in the p-type semiconductor is reduced causing the holes in the valence band to become the majority charge carriers. From the figure it can be seen that the Fermi level (EF) with higher donor concentration (ED) will move closer to the bottom of the conduction band and in the same manner the one with a higher acceptor concentration will move closer to 51

the top of valence band. Therefore when the semiconductor is under complete ionization condition (i.e. when n = ND and p = NA) we obtain the Fermi level in an extrinsic semiconductor as follows: In donor case, EC – EF = KT ln [NC/ND] In acceptor case, EF – EV = KT ln [NV/NA] Also electron and hole densities in extrinsic semiconductors can then be expressed as: n = ni exp[(EF-Ei)/KT] and p = ni exp [(Ei-EF)/KT]

If both the donor and acceptor impurities are present simultaneously, then, the Fermi level must adjust itself in such away so as to preserve charge neutrality.

Semiconductor transport carriers Carrier transport in semiconductors thin films is temperature sensitive. At room temperature electrons in a semiconductor are considered to be moving in all directions. The thermal velocity then at this temperature for any individual electron is caused by random scattering from collisions with the lattice atoms, or with impurity atoms, or with other scattering centres or all of them simultaneously. If an electric field (Ē) is applied to the semiconductor thin film, then each electron will experience a force from this field and it will be accelerated along the field (say on x-axis) and a hole in the opposite direction (- x-axis) of it. The electrons under this field will have a velocity called the drift velocity, Vd.

The relationship between the drift velocity (Vd) of electrons and the electric field (Ē) applied to a semiconductor is then obtained as: νd=µn Ē where μn is the electron mobility and it is given by: 52

µn = qτc/mn where τc is the mean free time (the average time between collisions) and mn is the effective mass of electrons. Research show that electron mobility is greater than hole mobility because of their effective masses i.e. an electron is smaller than a hole. This equation also applies to drift velocity for holes as: vp = μp Ē where μp is the hole mobility. When a transport carrier is under the influence of an applied electric field (Ē) it produces a current called the drift current (J). Therefore electron current density (Jn) caused by electron mobility is written as: Jn = -qpvn = qnµn Ē and, the holes current density (Jp) due to hole mobility is likewise given as: Jp =qpvp = qpµp Ē where, q is the electronic charge, n and p are the concentration of electrons and holes per unit volume respectively. The total current density (J) in the semiconductor sample is given by the sum of electron and hole mobility current densities: J = Jn + Jp = ( qnµn + qpµp ) Ē This total current summed up can also be written as: J =σĒ where, σ is given by; σ = qnµn + qpµp where, σ is the conductivity of the semiconductor which depends on the concentration of charge carriers n and p. The numbers of charge carriers are dependent on temperature in an exponential way and therefore conductivity increases exponentially and hence the corresponding resistivity which is a reciprocal of σ is given by: ρ = 1/σ = [1/{qnµn + qpµp}]

This equation can be used to describe the conductivity of extrinsic semiconductor especially the doped ones. The only difference is that the number of electrons in the conduction band and the number of holes in the valence band are not equal in 53

the case of an extrinsic semiconductor. One of the two (p or n) dominates depending on the type of the extrinsic process but the mass action law is maintained (np=ni2).

If both the donor and acceptor impurities are present simultaneously, then the impurity that is present in a greater concentration determines the type of conductivity in the semiconductor. Therefore, it can be written for n-type semiconductor as; ρ =1/qnμn and for p-type semiconductor as: ρ=1/qpμp

Sheet resistance, Rs is an important parameter in thin films and it is used to characterize both wafers and thin doped layers. It is easier to measure the sheet resistance rather than the resistivity of thin film materials. The sheet resistance of a uniformly doped layer) with a resistivity, ρ, and thickness, t, is given by their ratio; Rs = ρ/t while the unit of the sheet resistance is Ohms, it is usually referred to as Ohms per square (Ωcm-1). This is because when the resistance of a rectangular piece of thin film material with length, L, and width, W, is to be obtained, it is taken to be equal to the product of the sheet resistance and the number of squares: R = Rs L/W where the number of squares equals the length L, divided by the width, W.

Transport carriers move from a region of high concentration to a region of low concentration when a spatial variation of carrier concentration exists in any semiconductor material. Diffusion currents emanates from this random thermal motion of carriers due to a concentration gradient. This motion causes a certain current to flow without the influence of an electric field and therefore it diffuses

54

freely. This current is called diffusion current. This current for electron carriers (n) can be given by: Jn = [qDn] dn/dx where, q is the electronic charge, Dn is the diffusivity and the term dx/dn is the spatial derivative of the electron density The diffusivity, Dn according to Einstein relation is given as: Dn =[KT/q]µn From the above equation the diffusion current is proportional to the spatial derivative of the electron density. When both electric field and concentration gradient are present, both drift and diffusion currents flow. The total electrons current density will then be the sum of these two current components: Jn = qnµnE + [qDn]dn/dx and for total holes current density: Jp = qpµpE – [qDp]dp/dx The total conduction current density (Jcond) is given by the sum of the two equations as: Jcond = Jn + Jp Optical excitation is the injection of carriers by shining light on a semiconductor.Carrier injection is a process of introducing excess carriers to a semiconductor thin film by using various methods among them optical excitation and it is a process that affects the carrier transport. When this happens and if the photon energy (hυ) of the illumination is greater than the energy band-gap (Eg) of the semiconductor, then photons are absorbed by the semiconductor and electronhole pairs are generated. Photodiodes are designed and function based on optical excitation and that is why their electron and hole concentrations are increased above their equilibrium values.

Injection level is determined by the magnitude of the excess carrier concentration relative to the majority carrier concentration. The mean free time and the mobility of the carriers in a semiconductor are affected by lattice and impurity scattering 55

mechanisms. Lattice scattering is caused by thermal vibrations of lattice atoms which disturb lattice periodic potential and allow energy to be transferred between the carriers and the lattices. Since the lattice vibration increases with increasing temperature, lattice scattering becomes dominant at high temperatures and hence mobility decreases with increasing temperature. On the other hand impurity scattering results when a charge carrier travels past an ionized dopant.

Carrier generation-recombination process These electron-hole pairs recombine and cease to exist. Therefore recombination is a reverse process and through it electron-hole pairs are annihilated. Transport carriers in thin films are generated and they later recombine. Generation is the process of elevating an electron from a state in the valence band (VB) to the one in the conduction band (CB). This process results in the production of electronhole pairs. For thin films in thermal equilibrium these two processes must balance and therefore their rates must be equal. Sometimes this process releases energy (hv) and this energy can be emitted as a photon (called radiative recombination) or dissipated as a heat to the lattice, called auger or non-radiative recombination.

An electron-hole pair is generated only when a bond between two neighbouring atoms is broken due to the thermal vibration of lattice atoms. If this electron in the conduction band makes a transition downward to the valence band, an electronhole pair is annihilated. When a film is in thermal equilibrium the generation rate, Gth must be equal to the recombination rate, Rth. It has been found that three different recombination processes occur in semiconductor materials depending on their band structures. Direct recombination occurs in the direct band-gap semiconductors and here the probability that the electrons and holes will recombine directly is high. This is because the bottom of the conduction band and the top of the valence band are lined up and no additional crystal momentum is required for the transition across the band-gap.

56

On the other hand there is indirect recombination which occurs in indirect bandgap semiconductors. Here the electrons at the bottom of the conduction band have non-zero crystal momentum with respect to the holes at the top of the valence band. This makes the dominant recombination process to be indirect through a localized energy state at the forbidden energy gap. These localized energy states act as transition media between the conduction band and the valence band and also as recombination centres.

Impurities or dopants are also efficient recombination centres. There is a recombination referred to as surface recombination. This recombination process has a strong effect on the characteristics of many semiconductor devices. Since there is an abrupt discontinuity of lattice structure at the surface, the introduction of surface re-combinations greatly enhance the recombination rate at the surface region. For a good photovoltaic cell all these types of recombination must be reduced for optimum cell efficiency.

Optical phenomena in thin films When a semiconductor is illuminated photons are absorbed to create electron-hole pairs if their energy is equal to the band-gap energy (hυ= Eg). If hυ is greater than Eg, then an electron-hole pair is generated and the excess energy (hυ – Eg) is dissipated as heat. Both processes are called intrinsic transitions (or band-to-band transitions). If hυ is less than Eg, a photon will be absorbed only if there are available energy states in the forbidden band gap due to chemical impurities or physical defects. This process is called extrinsic transition.

This is also generally true when an electron at the conduction band edge combines with a hole at the valence band edge resulting into an emission of a photon with energy equal to that of the band gap. When a film is illuminated by a light source with hυ greater than Eg and a photon flux of Φο is absorbed, the fraction of the photons absorbed is proportional to the intensity of the flux. Therefore the number of photons absorbed within distance (d) is given by; 57

Φ(d) = ΦΦ e–αd where α is the absorption coefficient and is a function of hυ, d is the film thickness, the negative sign indicates decreasing intensity of the photon flux due to absorption. The absorption coefficient decreases rapidly at the cut-off wavelength λc given by; λc = 1.24µm /Eg since the optical band-to-band absorption reduces to become negligible for hυ < Eg, or λ > λc. The complex index of refraction is then defined as; nc = n – ik where n, is the refractive index and k, is the extinction coefficient. The complex index of refraction is related to the velocity of propagation by; v = c/nc The absorption coefficient α is related to the extinction coefficient k by; α = 4πk/ λ where, λ is the wavelength of the light in a vacuum. The dielectric constant (ε) can be defined as real which describes any losses by conductivity σ as; n2 –k2 = ε nk = σ/v where, υ is the frequency, ε is also defined to be complex as: ε ≡ ε – iε This then implies that; n2 – k2 = ε1 2nk = ε1

The knowledge of n and k determines ε1 and ε2, and vice versa. A plot of α versus photon energy hv has generally the same shape as a plot of ε2, because n usually does not vary greatly with energy. The dominant feature of the energy dependence of the absorption coefficient α(hυ) is the absorption near the region of inter-band transitions from valence to conduction bands and can be summarized as discussed in the next section.

58

Allowed direct transitions The optical absorption coefficient for a direct transition is given by; α = A(hv –Eg) ½ where, n = ½ for allowed transition and Α = 3.38 × 107 n-1 (me/m0) ½ (Eg/hv) where, Eg is the energy gap at k = 0, mo is the free-electron mass, υ is the frequency of the radiation, and n is the refractive index. When the minimum of the conduction band and the maximum of the valence band in a semiconductor occur at the same k value, absorption begins at hυ = Eg and the electron is transferred vertically upwards between the two bands without a change in momentum and therefore non-vertical transitions are normally forbidden in this case.

Forbidden direct transitions: In the case of simple parabolic bands: α = A(hv- Eg)3/2 where A is a slowly varying function of energy.

Indirect transitions Indirect transitions occur with a lower probability and gives rise to an absorption edge that is less steep than that for direct transitions where; αe(hv) = A(hv – Eg – Ep)2/[1- exp(-Ep/KT)] for hυ > Eg - Ep, αe corresponds to the emission of a phonon of energy Ep in order to conserve momentum and corresponding to the absorption of a phonon we have; αa(hv) = A(hv – Eg +Ep)2 / [exp( Ep/KT) -1] When the conduction band minimum and the valence band maximum occur at different k values, optical transitions from the latter to the former require the participation of phonons in order to conserve momentum because of the change in their electron wave-vector. Phonons are either emitted or absorbed. The absorption coefficient for an allowed indirect transition is given by; α(hv) = αe(hv) + αa(hv) 59

where the first term represents the contribution of phonon emission and is taken to be zero for hυ < Eg + Ep while the second term represents the contribution of phonon absorption and must be taken to be zero if hυ < Eg – Ep where, Ep is the phonon energy and a is a constant.

Forbidden indirect transition For low to moderate doping (> ND in thermal equilibrium, a onesided abrupt (p+-n) junction is formed and conversely when ND >>NA, an (n+-p) junction is formed.

While the p-type region contains a large concentration of holes with few electrons, the opposite is true for the n-type region. When we have large carrier concentration gradients at the junction, carrier diffusion occurs. Holes from the pside diffuse into the n-side and electrons from the n-side diffuse into the p-side. As holes continue to leave the p-side some of the negative acceptor ions (NA-) near the junction are left uncompensated since the acceptors are fixed in the semiconductor lattices while the holes are mobile.

Similarly some of the positive donor ions (ND+) near the junction are left uncompensated as the electrons leave the n-side. The charge due to the ionized donors and acceptors causes an electric field which in turn causes a drift of carriers in the opposite direction. The diffusion of carriers continues until the drift 62

current balances the diffusion current thereby reaching thermal equilibrium and therefore at thermal equilibrium net current flowing across the junction is zero. For each type of carrier the drift current due to the electric field must exactly cancel the diffusion current due to the concentration gradient. For the net holes current density; Jp = Jp(drift) + Jp (diffusion) = 0 Similarly, for the net electrons current density; Jn = Jn (dift) + Jn (diffusion) = 0 For a condition of zero net electron and holes currents, the Fermi level must be constant throughout the sample.

Show directions for drift and diffusion currents

The built-in potential Vbi, termed as the diffusion potential can be expressed by the relation: qVbi = Eg – [Vn + Vp] Vbi = {KT/q} ln[nn0pn0/ni2] = {KT/q} ln[NAND/ni2] It is noted that at equilibrium, nnopno = npoppo [where the subscripts (n) and (p) refer to the type of semiconductor and (o) refers to the thermal equilibrium value]. Therefore, Vbi = KT/q ln[ppo/pno] = KT/q ln[nno/npo] Thus, the hole and electron densities on either side of the junction are related by: pno = ppo exp[-q(Vbi/KT)] 63

and npo = nno exp[-q(Vbi/KT)]

A p-n hetero-junction The use of a hetero-junction (HJ) with a large band-gap window material and a small band-gap absorber material is a means of minimizing surface recombination losses that might otherwise dominate in direct band-gap materials. Thin film technology uses HJ to expand semiconductor material possibilities for solar and photovoltaic cell applications enormously. This is a junction formed between any two semiconductors having different energy band gaps. If the conductivity type is the same in any of these two semiconductors, then it is called an isotype heterojunction while in an anisotype the conductivity type is different in the two semiconductors.

Hetero-face photovoltaic cells [where a p-n homo-junction is interfaced with a lattice matched material of larger band gap] have achieved extremely high solar efficiencies e.g. in CdS/CdTe. The carrier transport properties of HJs are generally dominated by phenomena in the interface of p-n region. The current transport in the depletion layer is usually attributed to recombination, tunnelling, or a combination of both involving energy levels near the interface. The requirements for the formation of a good quality hetero-junction are: the lattice constant of the two materials should be nearly equal, the electron affinities should be compatible, and their thermal expansion coefficients should be close.

Photovoltaic cells The choice of semiconductors for photovoltaic conversion is based on a number of requirements. Some of the requirements include; 1. A direct band gap with nearly optimum values for either homojunction or hetero-junction devices. 2. A high optical absorption coefficient to minimize the requirements for high minority carrier lengths. 64

3. The possibility of producing n- and p-type material so that the formation of homo-junction or hetero-junction devices is feasible. Most suitable window materials have an n-type window character and a ptype absorber needed in a hetero-junction device. 4. A good lattice and electron affinity match with large band gap (window) materials such as CdS or ZnO so that hetero-junctions with low interface state densities can be formed.

The above requirements are fulfilled by a number of II-VI compounds. Accurate knowledge of the band gap (Eg), refractive index (n) and absorption coefficient (α) of semiconductors is important for the design and analysis of various optoelectronic devices. If a p-n junction of a photovoltaic cell is exposed to the solar spectrum, current flows at zero applied external voltage. In such a case a photon that has energy less than the band-gap makes no contribution to the cell output and the one with energy equal to band gap contributes an energy Eg to the cell output. Energy greater than Eg is wasted as heat. What actually happens in the photovoltaic cell when illuminated is demonstrated in the current-voltage (I-V) characteristics of the junction given by the equation: I= Is (e qV/nKT -1)-IL where the source IL results from the excitation of excess carriers by solar radiation, Is is the diode saturation current.

Current-voltage characteristics of a cell under illumination

65

Performance of photovoltaic cells In the modern world where cheap and clean forms of energy are needed there is a growing need for thin film materials with good photoelectrical properties in the infrared and ultra violet regions to be used for photovoltaic cells. Photovoltaic cells absorb light (photons) from the sun and convert them into electricity. When a photon with energy greater than the band-gap of the semiconductor passes through the cell, it may be absorbed by the material and this takes the form of a band-to-band electronic transition producing an electron-hole pair.

Short-circuit current density (Jsc) The short-circuit current density (Jsc), is related to the light-generated current density (JL) through the shunt resistance (Rsh) and series resistance (Rs). For highquality photovoltaic cells with high Rsh and low Rs, the values of Jsc and JL are identical. JL at any wavelength is determined by the optical excitation rate G(λ), the photon flux Φ(λ), the absorption coefficient α(λ), the thickness of the absorber t, and the reflectivity R(λ) of the active layer. JL is also determined by the diffusion length, L, and hence also by the carrier lifetime τ, and mobility μ.

Absorption coefficient (α) The absorption coefficient (α), the optimum thickness (t) of the absorber layer and the diffusion length (L) of minority carriers are closely connected. In the intrinsic range, α(λ) and it’s slope near the band edge is many times larger in direct band gap than in indirect band gap. This requires that a certain thickness of the absorber layer is needed to absorb a substantial fraction (~ 90%) of the incident photon. Higher current densities are obtained for larger thicknesses as a result of a more absorption. Surface recombination can be reduced by proper doping of the region near a surface to produce a drift field counteracting minority carrier diffusion to the surface.

66

Grain boundaries Grain boundaries in general have severe negative effects upon device performance because the boundaries generally serve to limit both minority carrier lifetime and electron transport across the layer in which they act. Grain boundaries in the absorber layer act like internal surfaces with high recombination velocities causing a reduction in both Jsc and Voc.

Fill factor (ff) The fill factor of a photovoltaic cell can be degraded by high series resistance and low shunt resistance. The series resistance of the cell arise from the grid bar metal resistance, grid finger metal resistance, front surface contact resistance between the metal grid and the semiconductor, sheet resistance of the semiconductor layer at the surface, base layer bulk resistance and back surface contact resistance between the metal and the semiconductor. The series resistance can be reduced by proper design of the grid pattern and suitable choice of metals to provide the ohmic contacts.

Photovoltaic cell operation The photocurrent generated in a photovoltaic cell also depends on whether the cell is operated in the front-wall mode. By using high-reflection coatings as back electrodes, the thickness of the absorber material required to absorb the photons can be reduced and the collection efficiency can be increased without increasing the minority carrier diffusion length. In thin film cells the surface of the cell is very rough and the surface topography is made up of pyramids and therefore the actual area of the junction is much larger than the corresponding geometrical flat plane area. Since such a surface is non-reflecting, JL increases. However the reverse saturation current also increases with the junction surface lowering Voc and FF. In hetero-junctions a dislocation field at the junction interface arises caused by lattice mismatch between the two materials counteracting at a space. This means that the charge state, the capture cross67

section, and the density of these interface states strongly influence the photocurrent through the junction and hence determine the magnitude of the field.

It is also known that matching of electron affinities is important. The potential barrier height is reduced by the degree of band mismatch and the total shortcircuit current density depends on the intensity and spectral distribution of the incident radiation striking the thin film. Different spectra give rise to different carrier generation profiles and therefore different photocurrent magnitudes.

Local atmospheric conditions It is now clear that the design and optimization of a photovoltaic device of any particular material requires a delicate balance between several conflicting requirements. Since the basic material properties differ from material to material, no single design is applicable to all photovoltaic systems and therefore each device design has to be optimized depending on the material properties. Common loss mechanisms experienced in most photovoltaic cells range from photon losses, carrier losses, voltage losses and power losses.

Doping of layers Heavy doping is favourable for the top layer especially when it is the window layer. It is desirable that it should have low surface recombination velocity and sufficient thickness to prevent undue series resistance. At the same time the top contact grid should have sufficient number of grid lines to reduce Rs but maintain high optical transmission and good ohmic contacts with a good anti-reflecting coating or a textured surface to reduce reflection losses. The material should also be transparent in the active range of the cell, solar spectrum and impermeable to water vapour and oxygen.

The rear contact The rear contact should be transparent and highly conducting. The base material should have a long minority carrier diffusion length if it absorbs the incident 68

photons significantly and sufficient thickness to prevent shorts through grain boundaries. The layer should be properly doped to allow development of the desired space charge region and a good electron affinity match with the top absorber layer.

The absorber layer The absorber layer should have low surface recombination, long minority carrier diffusion length and sufficient thickness to absorb completely near band gap photons after reflection from the rear contact. The sheet resistance should be low to avoid series resistance and the lattice parameters should be closely matched with those of the base layer/substrate.

Strengths and limitation of photovoltaic cells Desirable properties like a band gap (Eg) in the region of 1.1eV to 1.5 eV for absorber layers, high absorption coefficient (α), large minority carrier diffusion length (L), low density of recombination centres, matching electron affinities and lattice parameters are easily achieved. One more advantage of thin film photovoltaic cells is that it is possible to further develop the tandem structure approach into a more sophisticated version.

Because of this tandem idea a new field called ‘The Integrated Tandem Solar Cell’ system (ITSC) has been developed in which the built-in electrical connections provide a permanent series connection between the different cells. Thin films have gained a tremendous advantage in the photovoltaic energy industry because their photovoltaic cells record low costs of fabrication, large area devices, and even have the possibility of integrating them with other solar energy conversion devices like light sensors, lasers, xerographic imaging devices and other optoelectronic devices. Another equally important advantage of thin films is the aspect of their relation to photovoltaic technology. Their unique growth processes makes it very possible to fabricate materials with desirable properties.

69

A factor which now limits the performance of thin film photovoltaic cells is the polycrystalline nature of thin films in general. The grain boundaries present in polycrystalline films provide recombination surfaces for minority carriers and thus degrade the performance of the device. These grain boundaries affect the device operation by allowing inter-diffusion of certain atomic species or diffusion of a particular element (dopant) from one surface to another, creating shorting paths. This means then that the electrical activity of the grain boundaries needs to be suppressed either by passivation (applying a coating), by selective anodization (coat a metal with an oxide coat) of the grain boundaries rendering them ineffective. This also works with heavy doping which provide a direct field away from the grain boundary surfaces.

70

CHAPTER FOUR

SOLAR CELL THIN FILM GROWTH TECHNIQUES Thin films have been prepared on different substrate materials like glass, mica, alkali halides, MgO, and flexible substrates like polymers using different techniques. This is because the properties of a semiconductor thin film are highly affected by the type of substrate material as well as the technique used to prepare them. A substrate is a surface onto which a thin film is deposited. The type, nature and finish on a substrate surface are extremely important because they influence properties of thin films deposited onto them. Substrates commonly used to deposit polycrystalline films include glass, quartz, and ceramic substrates.

In epitaxial growth single crystal substrates of alkali halides, mica, MgO, Si, Ge, are suitable. In laboratory work the most commonly used substrates are glass slides of Pyrex, which like other glasses and ceramics are non crystalline and have a composition of 80.5 % SiO2, 12.9 % B2O3, 3.8 % Na2O, 2.2 % Al2O3 and 0.4 % K2O. In preparation of high thermal thin films, fused quartz and Vycor glasses that have higher silicate content (~96 %) and are stable at temperatures above 500 °C are used.

Cleaning is a requirement for thin film deposition. A variety of procedures exist for this purpose. Glass substrates must be thoroughly cleaned before deposition. For example, gross contaminants are first removed by a lukewarm ultrasonically agitated ionic detergent. This ensures also that etching is uniform. It is noted that a hot detergent solution may produce non-uniform etching of soda-lime glasses. The glass is then rinsed thoroughly and several times in de-ionized water after the hot detergent and later subjected to a vapour degreaser using pure ethanol. Cleaned glass may be stored immersed in pure alcohol or dried by blowing with dry nitrogen before use.

71

Other methods for cleaning glass slides and any cleaved surfaces utilize heat, electron and ion bombardment. Fire polishing or even heating in the softening point produces a clean glass surface. Thermal treatment yields etched rough surfaces on alkali-halide crystal above about 400 °C. Atomically smooth surfaces are obtained for metals by a lengthy annealing close to their melting point.

Sputter method of cleaning is an efficient method to remove contaminants and oxide layers from a substrate. The surface is sputtered by use of ionic bombardment and this is how an atomically smooth surface is realized by sputtering, de-gassing and extended annealing. The use of hot substrates is required for many thin film depositions. In this method, a quartz iodine radiation lamp is used as an extended source, a nichrome wire heater sandwiched between mica sheets, and W or Ta strip heaters are convenient heat sources. Oliva and coworkers proposed and cleaned the surfaces by soaking the substrate into an anionic soap and de-ionized (DI) water and then degreased them with CCl4, acetone and isopropyl alcohol and finally rinsed them with distilled water in each step.

Growth Kinetics and Diffusion of thin films Kinetics tells us how fast it will happen for a thin film to be formed. Here we will concentrate on mass transport and how atoms diffuse through a solid. Diffusion in one dimension relies on the Fick's 1st and 2nd Laws. During the studies of diffusion in thin films, it is necessary that the common constants are used appropriately. These constants include the following; (i)

self diffusion, (DA)

(ii)

vacancy diffusion, (DV)

(iii)

chemical diffusion (element A in B), (DAB)

(iv)

grain boundary diffusion, (Dgb)

(v)

surface diffusion, (Ds)

72

Other parameters in thin film deposition analysis like defects, concentrations, temperature may vary with time. When we have high diffusivity paths, it requires that consideration of the following are essential since these are all more open structures with higher jump frequencies and lower energy barriers; grain boundaries, three dimensional dislocation networks surfaces. Usually the cross sectional areas of these thin films under consideration are small compared to the rest of the film. That is why the constants Do and ED are different for these paths and vary as DS > Dgb > D

Arrhenius Plot Arrhenius plots are used for determining activation energies and this plot rely on the Fick’s first law. It also noticeable that diffusion can be changed by stress fields, electric fields, other energy gradients (interfaces) as the thin films nucleation takes place. Examine what happens when we apply a field. When there is no electric field, graph in figure 4.4 is observed for crystal with a constant Gd. To determine the rate of diffusion, Nernst-Einstein equation of diffusion is used. Starting with Fick's First Law, we have:

Arrhenius plots

DEPOSITION TECHNIQUES There are many techniques for depositing thin films but they can be classified into groups that includes chemical vapour deposition (CVD), chemical deposition 73

(CBD), physical vapour deposition (PVD) and liquid deposition techniques (LD) under the following headings.

Physical vapour deposition (PVD) Physical vapour deposition involves three major groups of depositions that include vacuum evaporation, epitaxial deposition and sputtering: Plasma-Enhanced Metal–Organic Chemical Vapor Deposition Is one of the most attractive ones for synthesis of high perfection ZnO films or inorganic thin films at low and moderate temperature. The deposition system consists of continuously pumped horizontal quartz tube placed between copper plates to which RF power (13.56 MHz) is put to excite plasma in the reaction chamber and to dissociate Zn(AA)2 (AA – acetylacetonate) vapuor. The base unit can be equipped by a resistive-heated evaporator where Zn(AA)2 powder placed in quartz ampoule. Because diethylzinc (DEZ) and dimethylzinc (DMZ) are usually applied as precursors in CVD process, these precursors are less desirable due of its high toxicity. Crystalline sapphire (001), silicon (100) and SiO2/Si substrates can be deposited by this method.

Plasma-Enhanced Metal–Organic Chemical Vapor Deposition system

Vacuum evaporation To evaporate into a thin film material, it is required that a material is heated to a sufficiently high temperature to produce the desired vapour pressure. These 74

vapour atoms then traverse a vacuum and they are made to condense on a substrate surface to form a thin film after they have been scattered by collisions with residual gas atoms in the vacuum system. The rate of deposition of the vapour atoms depends on the vapour source, substrate geometry and the condensation coefficient. Vacuum evaporation requires a system with a known vacuum pressure and its residual gas analysis.

Schematic diagram of high vacuum evaporation (HVE) chamber It is done through the following deposition techniques:

Resistive heating The choice of a support material is determined by the evaporating temperature, resistance of the alloying and the chemical reaction of the evaporant. Indirect heating uses crucibles of quartz, graphite, alumina, beryllia and zirconia. A material is resistively heated on a filament or boat made of refractory metals such as wolfram (W), molybdian (Mo), tantal (Ta) and niobium (Nb) with or without ceramic coatings. Flash evaporation It is rapid evaporation method where a multi-component alloy or compound is obtained by continuously dropping fine particles of the material onto a hot surface, or a mixture of the components in powder form into the evaporator. This causes numerous discrete evaporations to occur.

75

Arc evaporation It is technique where an arc is stroke between two electrodes of a conducing material generating sufficiently high temperatures to evaporate refractory materials such as Nb, Ta and carbon.

Exploding-wire technique This technique consists of an exploding wire by a sudden resistive heating of the wire with a transient high current density approaching 106 A/cm2.

Laser evaporation In this technique an enormous intensity of a laser is used to heat and vaporize materials by keeping the laser source outside the vacuum system. A focused laser pulse is directed onto target of material in a vacuum chamber (Figure 4.8). The laser pulse locally heats and vaporizes the target surface, producing an ejected plasma or plume of atoms, ions, and molecules. The plume of material is deposited onto an adjacent substrate to produce a crystalline film.

This technique possesses several favorable characteristics for growth of multicomponent materials, such as stoichiometric transfer of the target material to the substrate, compatibility with a background gas, and atomic level control of the deposition rate. In this method, oxidation of Zn primarily occurs in the ZnO ablation plasma plume, thus alleviating the difficulties encountered with MBE of ZnO, where oxidation proceeds via surface reactions. The beam is focused onto the surface of a fine powder of the material to be evaporated. Since the laser penetration depth is small, evaporation takes place at the surface only.

76

Schematic of PLD system

RF heating The RF is also called induction heating. Heat is either supplied to the evaporant directly or indirectly from the crucible material. By suitable arrangement of the RF coils, the induction-heated material can be levitated thereby eliminating the possibility of contamination of the film by the support material. Electron-bombardment heating This technique consists of a heated W filament to supply electrons which are accelerated by applying a positive potential to the material for evaporation. The electrons lose their energy in the material very rapidly at a rate determined by their energy and atomic number of the material. Thus the surface of the material becomes a molten drop and evaporates.

Epitaxial deposition Epitaxy means the growth of film with a crystallographic relationship between film and substrate. In homoepitaxy (also called autoepitaxy or isoepitaxy, the film and the substrate are same material while heteroepitaxy is where the film and the substrate are of different materials.

77

Hot wall epitaxy A method where one or two annular vapour sources are used and the vapours transported through a heated cylindrical enclosure that is held at a temperature higher than the substrate. Deposition of homogenized and equilibrated multicomponent vapours takes place on the substrate. This technique has been used to obtain good epitaxial films of several IV - VI and II - VI compounds.

Molecular Beam Epitaxy (MBE) The wafer on which growth occurs is held at an elevated temperature so that arriving Zn and O atoms have sufficient energy to move around on the surface of the wafer and find their correct bonding positions. This technique causes an epitaxial growth onto a single crystal substrate obtained by the condensation of one or more directed beams of atoms from an effusion source in an ultrahighvacuum (UHV) system (i.e.108 – 10-10 torr).

In this technique for the case of ZnO, the growth is performed under clean, low pressure conditions where the potential for contamination is minimized. If the temperature is too high, these atoms may be re-evaporated from the surface, while if the temperature is too low, the crystal quality of the ZnO layer being grown will be poor. The source materials for the growth are very pure Zn metal, which is evaporated from an oven toward the wafer and atomic oxygen derived from plasma or ozone source. MBE is capable of layer-by-layer growth with excellent control of the purity and crystalline quality of the resulting film.

MBE is a very sophisticated and expensive setup used primarily for basic epitaxial growth studies and also for specialized microelectronic applications. In molecular beam epitaxy; evaporation at very low deposition rates, typically in ultra-high vacuum, very well controlled, grow films with good crystal structure, expensive, often use multiple sources to grow alloy films and deposition rate is so low that substrate temperature does not need to be as high.

78

Schematic of MBE system

Graphoepitaxy It is a vapour deposition technique whereby a thin polycrystalline film is formed on amorphous substrates. It also makes it possible to grow large-area crystalline films and other conducting substrates for solar cell applications. Sputtering techniques Sputtering can be regarded as an ejection process. The ejection process takes place as a result of a momentum transfer between the impinging ions and the atoms of the target surface. Vapour species may be created by kinetic ejection from the surface of a material (called target or cathode) by bombardment with energetic and non-reactive ions. The sputtered atoms then condense on a substrate to form a thin film. Since the number of sputtered atoms is proportional to the number of ions, the sputtering process provides a very simple and precise control on the rate of film deposition. A large number of sputtering variants include:

Glow discharge sputtering A sputtering technique where ions are provided by a normal glow discharge created at a residual pressure of about 102 torr of the required gas (Ar) by applying 1 to 3 kV DC between a cathode (target) and an anode (where a substrate is 79

placed) separated by about 5 cm. Because of the collisions with gas atoms, the sputtered atoms reach the substrate with randomized directions and energies.

Magnetron sputtering It is a sputtering technique done when the applied electric and magnetic fields are placed perpendicular to each other during sputtering. In a planar cathode system, the magnetic field is applied parallel to the cathode to confine the primary electron motion to the vicinity of the cathode. This increases the ionization efficiency which helps to prevent the electron bombardment on the films. The usefulness of this technique is found in depositing a number of solar cell materials such as oxides of TO, ITO, and also depositing CdS and Cu2S layers.

Magnetron Sputter Deposition is unique in that it uses DC or RF with an aim of increasing ionization of Ar because higher sputter rates at lower Ar pressures (down to 0.5 mTorr) ensures fewer gas collisions and more line of sight and hence increasing the

probability of electrons striking Ar. Most common

configuration: crossed electric and magnetic fields where magnets of about 200 Gauss are behind target.

RF sputtering The RF technique can be used with any sputtering geometry in glow discharge or magnetron modes. It is an indispensable technique for deposition of thin films of semiconductors and insulators. RF sputtering is done at low pressures (103 torr) by enhancing gas ionization with the help of an inductively coupled external RF field. If the cathode is an insulator material, a high frequency alternating potential is used to neutralize the insulator surface periodically with plasma electrons which have a much higher mobility than the positive ions.

80

Schematic representation of the RF technique Good for insulating materials and in DC systems, positive charge builds up on the cathode (target) but needs 1012 volts to sputter insulators. Caution should be taken to avoid charge build up by alternating potential. Sputter deposition occurs when target is negative and when the substrate and chamber make a very large electrode - so not much sputtering of substrate.

Position of sputter depsosit Ion beam sputtering A sputter deposition achieved by using an ion beam source. It involves primary and secondary ion beam depositions. In the primary ion beam deposition process, the ions of the required material are produced and condensed on a surface to form a thin film. In the secondary ion beam deposition process, the Ar+ ions from a 81

beam source are used to sputter a target in vacuum and condense the sputtered species on a substrate. Ion deposition has been used to deposit films of ITO on Si for Semiconductor-Insulator-Semiconductor (SIS) solar cells.

Ion plating Ion plating technique involves ionization of the vapour by bombardment with accelerated electrons from a thermal source and depositing the ions onto a substrate with or without post-ionization acceleration. A combination of thermal evaporation onto a substrate (cathode) is simultaneously bombarded with positive ions (e.g. Ar+) from a glow discharge. Ion assisted deposition has;with evaporation or sputtering (or chemical vapor deposition), bombard surface with ionsnot necessarily same type as in film;ions typically NOT incorporated in filmrelatively low voltages (50 - 300 eV), leads tophysical rearrangement, local heating can change film properties for better or worse;disruption of columnar growth requires about 20 eV of added energy per depositing atom.

Reactive sputtering This type of sputtering is accomplished by introducing the reactant in vapour form into inert gas plasma. A chemical reaction can then take place either on the cathode, in the plasma, or at the anode depending on the pressure and chemical activity of the reacting species. The given surface and temperature conditions apply. It is majorly applied in preparing controlled composite oxide films for solar cells, for anti-reflecting coatings and patterned thin films for solar cell grid structures.

Chemical Deposition Techniques There are three techniques classified as chemical deposition techniques and they include chemical bath deposition (solution growth process), spray pyrolysis and screen printing techniques.

82

Chemical bath deposition method This is a technique that is also called solution growth process and was first used in 1946 to prepare PbS films for infrared applications and recently for large area and large scale applications to obtain doped and un-doped multi-component thin films. This method is based on a controlled precipitation of a compound from a solution on a suitable substrate. The substrate is either immersed in an alkaline or acidic solution which is containing a metal ion, chalcogenide source, the added acid or base and a complexing agent. Several complexing agents are utilized in depositing thin films such as NH3, tri-ethanolamine, disodium ethyle diamine tetra acetate.

A typical chemical bath solution is comprised of three parts: at least one salt of metal M+n, a source of chalcogenide X-m (where X = oxygen, sulfur, selenium, etc), and a complexing agent in aqueous solution. The metal salts are chosen from moderate to high solubility salts and the desired end product. The chalcogenide source is chosen based on the desired end product and also on the rate at which Xm

is generated. The complexing agent provides ligands and prevents precipitation

of metal hydroxides. This prevents rapid bulk precipitation of the desired compound. It is chosen based on the rate at which free metal ions are generated in solution.

Substrates are immersed vertically in the reaction bath which is stirred continuously with a magnetic stirrer while temperature of the bath is monitored by a contact thermometer to maintain a constant temperature. The film growth rate initially is negligible because an incubation period is required for the formation of critical nuclei from a homogeneous system onto a clean substrate surface. Once this occurs the rate rises rapidly until the rate of deposition equals the rate of dissolution. If the substrates are suspended in the bath before forming the complex ions, the film thickness increases in a manner similar to that of the sensitized surface because the nuclei for the formation of the film are provided by the solution itself. The rate of deposition and the terminal thickness both depend on 83

the number of nucleation centres, super-saturation of the solution and presence or absence of stirring.

Schematic representation of CBD set up

The film growth kinetics depends on the concentration of ions, their velocities, their nucleation, growth rate processes and the type of salts/compounds used either for metal or chalcogenide ions. The rates and terminal thickness are higher for sulphides than for the corresponding selenides films under similar deposition conditions. When bath temperature increases, the dissociation of both complex ions present and the chalcogen bearing compound also increases. It is also noted that an increase in the concentration of metal and chalcogen ions coupled with higher kinetic energy of the ions results in an increased interaction. This yields a higher rate of deposition of the metal chalcogenide film.

The super-saturation is controlled by both the bath temperature and the complexing agent concentration. Impurities in the starting chemicals can be incorporated into the films only if the impurities form insoluble chalcogenides under the same conditions of deposition. It is possible to form multi-component chalcogenide films over a wide composition range. The composition in such films can be varied by controlling the initial salt concentration, complexing salt

84

concentration and temperature of the bath. Chemical bath deposition is commonly used because it is a simple, cost effective and economically reproducible method that can be applied in large area deposition and low temperatures. This technique is full-fledged now and gives a great future promise in thin film technology.

Picture of chemical bath deposition (CBD) setup

Screen printing In this method, pastes containing the desired material to form a film are screenpainted by conventional methods onto a suitable substrate to define conductor, resistor, and device patterns.

Spray pyrolysis This technique involves spraying a solution usually aqueous containing soluble salts of the constituent atoms of the desired compound onto a substrate maintained at elevated temperatures. The sprayed droplets reaching the hot substrate surface undergoes pyrolytic (endothermic) decomposition and forms a single crystallite or a cluster of crystallites of the product. The other volatile by-products and the excess solvent escape in the vapour phase. The substrate provides the thermal energy for the thermal decomposition and subsequent recombination of the constituent species. Because the film formation is carried out in air by a simple

85

apparatus in SPD, the technique is one of the most attractive film preparation methods.

SPD is essentially the same film processing technique as so-called pyrosol technique, in which a source solution is sprayed on the heated substrate to be deposited as a film. In other words, when a source solution is atomized, small droplets splash and vaporizes on the substrate and leaves a dry precipitate in which thermal decomposition occurs. Organometallic compounds and inorganic salts are used as source materials, which are dissolved in water, ethanol or other solvents to prepare source solutions. Because the source materials dissolve in a solvent as an ion, oligomer, or cluster depending on their chemical properties, the surface morphology of deposited films is easily controlled by choosing species of the source materials.

Schematic representation of a spray pyrolysis deposition (SPD)

This technique was first developed by the research group of Sotiris E. Pratsinis at ETH Zurich, Switzerland. Since then it has been used to create new and sophisticated materials for catalysis and other applications. Johnson Matthey has developed its own Flame Spray Pyrolysis Facility which produces a range of nanopowders using the flame spray pyrolysis technique.Flame spray pyrolysis can 86

be used to produce a wide array of high purity nanopowders ranging from single metal oxides such as alumina to more complex mixed oxides, metals and catalysts. It has the capacity to produce up to 100 g h−1 of nanopowder product depending on the material composition and a number of process variables enable the preparation of well-defined target materials.

Flame spray pyrolysis is a one step process in which a liquid feed (a metal precursor dissolved in a solvent) is sprayed with an oxidising gas into a flame zone. The spray is combusted and the precursors are converted into nanosized metal or metal oxide particles depending on the metal and the operating conditions. The technique is flexible and allows the use of a wide range of precursors, solvents and process conditions, thus providing control over particle size and composition. Some of these materials deposited by this process find uses in catalysis, electronics, thin film applications and other areas. Additionally the transferable knowledge gained can be applied to the synthesis of pgm catalysts and supported pgm catalysts by the flame spray method.

Chemical Vapour Deposition (CVD) A chemical reaction is initiated at or near the substrate surface producing the desired material as a solid-phase reaction product which condenses on the substrate. All CVD techniques involve essentially the exposure of the substrates to one or several vaporized compounds or reagent gases, some or all of which contain constituents of the desired deposited substance. The chemical reaction may be activated by the application of heat, an RF field, light, X-rays, an electrical arc, a glow discharge, electron bombardment, or catalytic action of the substrate surface. All these processes require that they attain deposition conditions which enable the reaction to take place near or on the substrate surface (heterogeneous reaction) in order to avoid powdery deposits, which result when the reaction occurs in the gas phase (homogeneous reaction). When compared with PVD technique, CVD may 87

be carried out at low pressures or in high vacuums, depending on the requirements. Some of the CVD techniques are as follows:

Close-spaced vapour transport (CSVT) In CSVT, a temperature gradient is maintained between the closely spaced source and substrate. Close spacing in chemical transport systems is required for growing epitaxial layers of pure and compound semiconductors. A gas is used to react with the source to form a volatile compound that is subsequently decomposed at the surface of the substrate to form a thin film. This method has been used to grow epitaxial films of CdTe and GaAs using water vapour as the transporting agent.

Plasma deposition In this technique, glow discharge plasma (dc or rf) is used to break the vapours up into different species that react to deposit as a film. The most exciting application of this technique is for the preparation of a-Si:H films for solar cells from pure, or diluted in Ar, SiH4 glow discharge. This technique also called the glow discharge deposition technique and it is essentially a plasma-assisted CVD technique.

Exchange reactions In this technique the reaction is usually performed in aqueous solution at temperatures ranging from 90 to 100 °C. The reaction is carried out in an inert atmosphere using an Ar or N2 blanket over the bath. It has been used to prepare films of cuprous chalcogens by a topotactial ion exchange reaction. The technique has been utilized to grow Cu2S and Cu2Te films.

Electrodeposition Electrolysis can be taken as the occurrence of chemical changes owing to the passage of an electric current through an electrolyte. The deposition of any substance on an electrode as a consequence of electrolysis is called electro-

88

deposition. This technique has been used to deposit different alloys, metals and semiconductor materials.

Electroless or Autocatalytic plating The electroless technique involves the reduction of metal ions to form deposits without the use of an external power supply. A catalytic surface is used to initiate the deposition and then the metal used must be catalytic to further the deposition process. The electrons for the reduction process are provided by a chemical reducing agent in the solution. Since the metal being deposited has to be catalytic in nature, only a limited number of metals can be deposited by this technique.

Anodization Anodization is a field-assisted form of thermal growth where the metal to be anodized is made an anode and immersed in an oxygen-containing electrolyte which may be aqueous, non-aqueous, or fused salt. Growth takes place at constant voltage or at constant current. The thickness obtained by this method depends on the applied voltage by the external force. This defines a quantity called the anodization constant as the thickness of the layer grown per unit voltage. If the voltage across the film increases then there is an increase in the oxide thickness and hence the limiting thickness is reached when the breakdown of the film occurs.

Electrophoresis Electrically charged particles suspended in a liquid medium are deposited on an electrode resulting on the as-deposited films forming loosely adherent coatings of a powder. It is a process that requires further post deposition treatment to produce an adherent, compact and mechanically strong surface coating. The post deposition treatment usually consists of pressurized compaction and a heat treatment to dry out traces of the suspension medium and sinter the particles within the film.

89

Liquid deposition techniques Liquid phase epitaxy and melt spinning are the two major liquid deposition techniques and they show a great promise in preparing large area thin layer materials in several micrometer dimensions for photovoltaic cell applications.

Liquid phase epitaxy (LPE) In this technique, contact is achieved either by tipping the furnace with solution or by dipping the substrate into the solution in a vertical furnace. The solution is saturated with the growth material at the desired growth temperature and then allowed to cool in contact with the substrate surface at a rate and for a time interval appropriate for the generation of the desired layer. It is a method that involves the precipitation of a material from a cooling solution onto an underlying single crystal substrate and especially used to prepare quality III-V compounds. Under optimum conditions, the layer grows as an extension of the single crystal substrate. Here the solution and the substrate are kept apart

Melt spinning technique (MS) This is a technique used to produce quenched ribbons at high speed and of several meters per second. A molten material is brought using a nozzle into contact with a spinning wheel and the liquid drop is pulled out of the nozzle in the form of a ribbon. The rate of solidification of the melt and the ribbon thickness are dependent on the nozzle dimensions, the molten drop in contact with the spinning wheel surface, the linear velocity of the wheel and the heat transfer processes.

90

CHAPTER FIVE CHARACTERIZATION OF THIN FILM FOR SOLAR CELLS Characterization fundamentals Scientists use the many or some of following thin film characterization devices when studying materials: optical microscopy; scanning electron microscopy (SEM); transmission electron microscopy (TEM); scanning probe microscopies (STM, AFM). This structure can accurately determine the following; internal structure; density. To also determine the structure at microscopic and atomic scales, the following characterization techniques are used; X-ray diffraction (XRD); stylus profilometry; quartz crystal monitors (QCM); ellipsometry; low energy electron diffraction (LEED); reflection high energy electron diffraction (RHEED). In most advanced research laboratories, the available probes used in thin film characterization include the following; light (electromagnetic radiation); electrons; ions (and nuclei and protons); neutrons; neutral atoms; "touching" the sample forces The composition of a thin film usually determines the electrical, structural, optical, mechanical, magnetic and capacitance or dielectric properties as well as the uses of these films. The following three thin film compositional properties are necessary to be known accurately; elemental composition; impurities; chemical states. These three properties are accurately measured, determined and characterized by the following techniques; Auger Electron Spectroscopy (AES); Energy

Dispersive

Analysis

of

X-rays

(EDAX);

X-ray

Photoelectron

Spectroscopy (XPS); Secondary Ion Mass Spectrometry (SIMS); Rutherford Backscattering (RBS). To determine optical properties there is a need to investigate the following properties of the sample as a function of wavelength; refractive index; absorption; dielectric properties. What electrical properties of the sample are significant is another good study question. This leads the thin film analysis to investigate; device properties; 91

material properties in which resistance / conductance and capacitance are considered; resistance - four point probe and capacitance. Magneto-optical Kerr effect (MOKE) and ferromagnetic resonance (FMR) are used to investigate the magnetic properties of the sample including hysteresis loops. Mechanical properties of the sample like internal stress in films / substrates and friction, adhesion are measured and characterized by the stress curvature measurements, pin on disk friction test and the adhesion tests. Measurement of Magnetic propeties Magnetic properties of thin films can often be described by a hysteresis curve using a Magneto-optical Kerr effect (MOKE). This technique uses ellipsometry and a magnetic field on the sample. Here magnetic materials cause an additional rotation of the polarization of reflected light and thus typically the device nulls out the normal ellipsometry effect and measure the change in intensity of reflected light as the magnetic field changes. This method is commonly used on thin films because it is sensitive to thin films. The curve produced is reflected below;

Hysteresis curve using a Magneto-optical Kerr effect (MOKE)

Measurement of thin film Thickness Since purity and thin film composition depends on concentration and depositional conditions the ratio of dM/dAs also depends on r, , and thus so does film

92

thickness (d). Consider flat substrate, perpendicular to source of the thin film deposition material as shown below for this geometry:  cos = h/r , r = (h2 + l2)1/2

Analysis of depositional geometry Therefore, it can be clearly seen that a surface source has slightly poorer thickness uniformity. For better uniformity, then a thin films needs to: decrease sample size (l); inrease distance to substrate (h); need bigger chamber; need better vacuum; wastes evaporant; use multiple sources; move substrate during deposition; use rotating amsk to reduce evaporant near center;

Chemical measurement Techniques Thin film chemical characterization is commonly done by the Auger Electron Spectroscopy among other analysis thechinques.

Auger Electron Spectroscopy (AES) It is used for determining; elemental composition (qualitative, quantitative); some chemical state information; element mapping with Scanning Auger; surface sensitive (30Å); depth profiling using Ar ion sputtering. The samples to be tested by the AES must be; conducting solids; metals, semiconductors, films; samples must be clean; have no high vapour pressure material nature, The AES is limitated in the following ways; does not detect H or He; detection limit about 1 at %;

93

sampling region diameter: 100 Å - 1 mm; requires ultra-high vacuum; beam damage on some samples. However, the Auger electrons are difficult to see because of their smalll peaks on large background. They are only determined when a derivative of spectrum to accentuate the peaks is taken. Such an analysis is reflected on the graph below.

Auger electron spectrum of stainless steel Measurement of structural properties These structural properties influence the physical and sometimes behavioural properties of all thin films and therefore they have to be known by measuring them. Crystal structure, grain boundaries, lattice constants, grain size and surface morphology are some of the very important properties in solid state physics of thin film semiconductors.

X-ray diffraction X-ray diffraction was used only for the determination of crystal structure in the earlier days but in recent research it has other uses. Today this method is applied not only for structure determination but also for such diverse problems as chemical analysis, stress measurement, study of phase equilibria, the measurement of particle size and the determination of the orientation of one 94

crystal or the ensemble of orientations in a polycrystalline aggregate. X-ray diffraction is a powerful tool for the investigation of the fine structure of matter.

Bragg’s law Bragg's law in physics show that when X-rays hit an atom they make the electronic cloud move as does any electromagnetic wave and these cloud reradiates waves with the same frequency and if there is an interference then it is constructive only when the phase shift is a multiple of 2π, a condition expressed as; nλ = 2d sin ϴ where n is an integer determined by the order given, λ is the wavelength of the Xrays, d is the spacing between the planes in the atomic lattice, and θ is the angle between the incident ray and the scattering planes. When a monochromatic x-ray beam is incident on the surface of a crystal, it is reflected. This reflection takes place only when the angle of incidence has a certain value. These values depend on the wavelength and the lattice constants of the crystal. Therefore it seems reasonable to attempt to explain the selective reflectivity in terms of interference effects as in physical optics. The model is illustrated in figure below where a crystal is represented by a set of parallel planes corresponding to the atomic planes. The incident beam is reflected partially at each of these planes which act as mirrors and these reflected rays are then collected simultaneously at a distant detector. The reflected rays interfere at the detector and according to physical optics the interference is constructive only if the difference between the paths of any two consecutive rays is an integral multiple of the wavelength. When the inter-planar distance is denoted by d and the glancing angle between the incident beam and reflecting planes is θ, then from the figure below we arrive at the following condition for constructive interference:

95

Diffracting path of x-rays by a crystal There will be a path difference between the ray that gets reflected along AC' and the ray that gets transmitted, and then reflected along AB and BC respectively. This path difference is; (AB + BC) – (AC’) The two separate waves will arrive at a point with the same phase and hence undergo constructive interference if and only if this path difference is equal to any integer value of the wavelength, i.e. (AB + BC) – (AC’) = nλ where the same definition of n and λ apply as above. Mathematical then it can be seen that; AB = BC = d/sin ϴ and, ac = which then follows that solving the above expression we arrive at,

96

which then can be simplified to; nλ = 2d sin ϴ

which is the mathematical form of Bragg's law. The angles determined by equation 3.4 and for a given d and λ are the angles at which reflection takes place only. Since sin θ cannot exceed unity (1), we may write: nλ/ 2d = sin ϴ < 1 which shows that nλ must be less than 2d for constructive interference.

Philips X-ray diffractometer

Diffraction directions To determine the possible directions, i.e. the possible angles 2θ, in which a given crystal can have a different beam of mono chromatic x-rays, we combine the Bragg’s law and the plane spacing equation as follows: The Bragg’s law is; λ =2d sin θ

for n =1

And if the crystal is cubic then: 1/d2 = {(h2+k2+l2)/a2} Combining these equations we have 97

Sin2 θ = [λ2/4a2] (h2+k2+l2)

This equation predicts, for a particular incident wavelength λ and a particular cubic crystal of unit cell size a, all the possible Bragg angles at which diffraction can occur from the planes (hkl). If the crystal is tetragonal, with axes a and c then the corresponding general equation is; Sin2 θ = [λ2/4a2] (h2/a2+k2/b2+l2/c2)

Spacing between planes of the same Miller indices In connection with x-ray diffraction from a crystal, one needs to know the interplanar distance (d) between planes labeled by the same Miller indices (hkl). The actual formula depends on the crystal structure. The distance between two planes (d) is simply the length of the normal line drawn from the origin to the plane. This distance d dhkl is given by: 1/d2 = (h2+k2+l2)/a2 2

2

2

2

2

(for Cubic) 2

1/d = (h +k )/a +l /c

(for Tetragonal)

1/d2 = {4(h2+hk+k2)/3} +l2/c2

(for Hexagonal)

1/d2 = h2/a2 + k2/b2+ l2/c2

(for orthorhombic)

Similarly we can obtain the relation of the inter-planar distance for the other crystal systems. The volume of the unit cell can be given by: V= a3

(for cubic)

2

V= a c

(for tetragonal)

V= √3/2 [(ca2)]

(for Hexagonal)

V=abc

(for orthorhombic)

Calculation of the particle size We can estimate the particle size of very small crystals from their diffraction curves by taking the width of these curves and using the relation: t = λ/βcos θ where,

98

β = ½{2θ1 - 2θ2} = θ1 – θ2 is the width at the half maximum intensity measured in radians as shown in Figure below and λ is the wave length.

The using of diffraction curve to determine the particle size

Transmission Electron Microscope In summary, the TEM technique has the following characteristics; uses microstructural analysis; interfacial analysis; crystal structure; magnifications up to 1,000,000 X => atomic resolution; small region elemental analysis; samples thinned to about 0.1 micron (1000 Å); minimum size: 1 mm; maximum size varies with instrument; sample preparation is very time consuming

Schematic of Transmission electron microscope

99

Imaging of transmitted beam intensity

Scanning Probe Microscopies This characterization technology has the following techinques that helps to determine various character properties of thin films and materials; Scanning Tunneling Microscopy (STM); Atomic Force Microscopy (AFM); Magnetic Force Microscopy (MFM); Scanning Thermal microscope; Scanning Near-Field Optical Microscope; and many others. All of them use an aperture put very close to the sample to be analyzed or measured and a potential is applied and get a current between tip and sample. Tunneling Current is a very sensitive function of tunneling gap.

Tunneling graph for high resolution level

100

Atomic Force Microscope (AFM) In this technique, tip is used which is deflected by atomic forces and this tip with a radius of curvature typically less than 400 Å tests and desplays the surface morphology of the thin film. The forces that derive its results can be attractive or repulsive.

o

Detection of surface by AFM

Scanning electron microscopy (SEM) It can provide important information about the surface features of an object, its texture, the shape, size and arrangement of the particles making up the object that are lying on the surface of the sample or have been exposed by grinding or chemical etching, the elements and compounds the sample is composed of and their relative ratios in areas about 1 μm in diameter. Scanning electron microscopy (SEM) is a basic tool for material characterization especially microstructural/morphological properties.

Furthermore,

it

can

arrangement of atoms in the specimen and their degree of order.

show the Scanning

electron microscope resolutions are currently limited to around 25 Angstroms. Figures below shows a schematic diagram of scanning electron microscopy.

101

A schematic diagram of scanning electron microscopy

A schematic diagram of scanning electron microscopy connected to computer

102

Scanning electron microscopy

The scanning electron microscope generates a beam of electrons in a vacuum. That beam is collimated by electromagnetic condenser lenses, focused by an objective lens, and scanned across the surface of the sample by electromagnetic deflection coils. The primary imaging method is by collecting secondary electrons that are released by the sample. The secondary electrons are detected by a scintillation material that produces flashes of light from the electrons.

The light flashes are then detected and amplified by a photomultiplier tube. By correlating the sample scan position with the resulting signal, an image can be formed that is strikingly similar to what would be seen through an optical microscope. The illumination and the shadowing show quite natural looking surface topography. There are other imaging modes available in the SEM. Specimen current imaging using the intensity of the electrical current induced in the specimen by the illuminating electron beam to produce an image. Electrons in SEM do not travel very far and on average, electrons incident on a material may either scatter back or knock out other (secondary) electrons. The number of secondary electrons produced is relatively insensitive to atomic number of the atoms in the material. The number of backscattered electrons, however, is sensitive to atomic number of the material. The graph below shows 103

the density behavior of these electrons. The number of electrons leaving the surface varies with the incident electron energy. This can influence the charging of the sample.

Backscattered electrons Backscatter imaging uses high energy electrons that emerge nearly 180 degrees from the illuminating beam direction. The backscatter electron yield is a function of the average atomic number of each point on the sample, and thus can give compositional information. Scanning electron microscopes are often coupled with x-ray analyzers. The energetic electron beam - sample interactions generate x-rays that are characteristic of the elements present in the sample. Many other imaging modes are available that provide specialized information.

Measurement of electrical Properties The carrier concentration in a semiconductor may be different from the impurity concentration, because the ionized impurity density depends on the temperature and the impurity energy level. To measure the carrier concentration directly, we use the Hall Effect. The Hall Effect and the electrical resistivity have been and continue to be the key parameters used in investigations of the basic electrical conduction processes in semiconductor materials. Hall measurements are widely used in the initial characterization of semiconductors to measure carrier concentration and/ or carrier mobility. Hall measurement is also one of the most

104

convincing methods to show the existence of holes as charge carriers, because the measurement can give directly the carrier type. The importance of the Hall Effect comes from the simple relation between the Hall coefficient and the free carrier concentration in the solid.

The Hall Effect When an electric field is applied on a p-type semiconductor sample along the Xaxis (i.e. electric current) and a magnetic field applied along the Z-axis, the Lorentz force, F; F = qυ×B (qυxBz) due to the magnetic field will exert an average upward force on the holes flowing in the X-direction. The upward directed current causes an accumulation of holes at the top of the sample that give rise to a downward-directed electric field Ey as shown in figure below.

Basic step to measure carrier concentration using the Hall Effect

Since there is no net current flow along the Y-direction in the steady state, the electric field along the Y-axis exactly balances the Lorentz force, that is: qEy = qvxBz Ey = vxBz

105

Once the electric field Ey becomes equal to υxBz, no net force along the Ydirection is experienced by the holes as they drift in the X-direction. The establishment of the electric field is known as the Hall effect. The electric field is called the Hall field, and the terminals voltage Vн= Eyw is called the Hall voltage. The holes current density Jp is given by: Jp = qpvx Therefore, the hole drift velocity is: vx =Jp/qp Using the above equations the Hall field Ey becomes: Ey = [Jp/qp]Bz = RHJpBz where; RH=1/qp The Hall field Ey is proportional to the product of the current density and the magnetic field. The proportionality constant RH is the Hall coefficient. A similar result can be obtained for an n-type semiconductor, except that the Hall coefficient is negative; RH=1/qn

where q is the electronic charge and p or n the density of carriers. The Hall coefficient RH determines the carrier type; it is negative for n-type & positive for p-type. More elaborate theory predicts that: RH = rn/qn, rp/qp

Where r is constant, usually between 0.5 and 1.5 that depends on the specific details of conduction in a given material. For spherical energy bands: rn = rp = 3π/8

by making measurements at high enough magnetic field, r can be reduced to 1 regardless of the conduction mechanism. For the case of mixed conduction, the Hall coefficient has a contribution from each. For a small field case; 106

RH = (1/q)(p-b2n)/[bn+p] where b is the ratio of electron-to-hole conductions mobility. When the material is intrinsic, the Hall coefficient is very large at low temperatures. As the temperature rises, the Hall coefficient decreases rapidly (exponentially) and the same behavior is noticed with the lightly doped semiconductors. In the extrinsic and heavily doped semiconductors the Hall coefficient is independent of temperature. The Hall voltage is given by: VH = 108RHBI/w where VH is in volts, RH in cubic centimeters per coulomb, B in gauss, I in amperes and the thickness w in centimeters. The Hall angle is given by: Θ= tan-1[108(RHB/ρ)] where RH is in cubic centimeters per coulomb, B in gauss and ρ in ohmcentimeters.

Hall Effect measurement system

There are a number of spurious voltages which will be included in the value read at the Hall terminals, but most of them can be eliminated by making a series of readings with the various combinations of current and magnetic field. This is the established method of making dc Hall measurements. Equation dos not separate out the Ettingshausen voltage, but this voltage is ordinarily quite small and can be neglected. The equations which we present in the Hall Effect theory assumes that 107

a rectangular sample with uniform end contacts and the Hall electrodes several sample widths away from either contact.

If the later condition is not met, the contacts will partially short out the Hall voltage. However, the smaller the L/d ratio, the greater the sensitivity to errors in determining the sample dimensions, and the smaller the Hall voltage measured. In the simple theory the Hall contacts are assumed to be infinitely small so that they do not distort the current flow. Experimentally, contacting can be by sharp tips or very small (about 1 mil) alloyed-wire contacts. More often, though, ears on the sample are used.

The ears serve two purposes: first, they allow a large area to be used for contacting without severely distorting the lines of current flow in the sample. Second, a contact made directly to the side of the bar will in general be noisier than one using an ear. If it is inconvenient to cut out rectangular barw, the thin samples of arbitrary shape such as platelets or slices can be used if contacts are made at four places around the periphery. If Vs / Is is measured with and without a magnetic field, then: RH = [(ΔV/I)d]/B where ΔV is Vs measured with the field minus Vs measured without the field and I is held constant for both measurements.

Common Hall-bar configuration

108

Exactly apposite each other, since a reversal of magnetic field will cancel out any initial unbalanced voltage. However if the unbalanced voltage is greater than the Hall voltage, error can be introduced, since VH will be the small difference between two large numbers.

Use of an arbitrarily shaped sample with randomly placed contacts

Further if electronic instrumentation is used, large unbalances may cause amplifier saturation. Two contacts on one side plus a potentiometer between them may be used, but there will be some distortion of the current lines because of the current flow between the two contacts. Current can be fed into corners of the sample, and the relative amount adjusted to provide balance at the Hall electrodes. Because of the complex shape of Hall bars they are cut most conveniently with an ultrasonic machine, although a fine-nozzle sandblaster or spark erosion can also be used.

If the sample is very thin, it may be etched to shape. For epitaxial and diffused layers isolated from their substrates by p-n junctions, etching is defined by standard photolithography techniques down through the junction to the substrate will suffice. If the material has high carrier mobility and a long life line, the surface should be treated to increase the surface–recombination velocity. Surface conduction or inversion layers can also radically change the measured value of the Hall coefficient.

109

Material, such as high–resistivity silicon is particularly susceptible to surface conditions, and series of treatments in boiling water can change the calculated mobility by factors of 10. The equipment which are needed for Hole measurement are a homogeneous DC magnetic field source of few kilogausses range, voltmeter (potentiometer) or galvanometer to measure the Hall voltage, electric field source to provide a DC current and an ammeter to measure the current flow. The average diameter (D) of the contacts, and sample thickness (d) must be much smaller than the distance between the contacts (L). Relative errors caused by non-zero values of D are of the order of D/L.

Contact geometry for Hall measurement- Square or rectangle

Resistivity Resistivity, carrier and impurity concentration are all interrelated and for most impurities in semiconductors and the interrelation is well known. Resistivity is one of the important properties of materials. Resistivity measurement is a geometry-dependent and quite sensitive to boundary conditions. Because of this sensitivity, many correction factors have been calculated. Most semiconductor materials have rather high temperature coefficient of resistivity, so if precise measurements are desired, and if the ambient varies widely, suitable correction should be made.

110

The basic methods of resistivity evaluation are as follows:

Direct method The oldest way of finding the resistivity ρ(ohm-centimeter) is to use a rectangular sample of known dimension to measure the resistance R and use the relation, R = ρl/A, where l is the sample length and A its cross section. A disadvantage is that ρ will also contain a contact-resistance term, which for semiconductors can be appreciable.

Two-point probe The effect of contact resistance can be eliminated by use of the two-point probe if the specimen cross-section is relatively uniform. A single movable probe can be used and the voltage measured between it and current lead or other suitable reference.

Two-point resistivity probes

By making several readings, dv/dx can be plotted and the resistivity calculated from: ρ = A/I dx/dv, where x is the distance along the surface. The voltage probes have fixed spacing and are moved in unison along the surface.

111

Four-point resistivity geometry

Four-point resistivity probes

Linear four-point probes In semiconductor industry the most generally used technique for the measurement of resistivity is the Four- Point Probe. The usual geometry is to place the probes in a line and use equal probe spacing. Current is passed thought the outer two probes and the potential developed across the inner two probes. Measured through any of the other five combinations of current and voltage probes can in principle be used. The resistivity is: ρ= 2πS (I/V) where S is the probe spacing in centimeters end S1= S2= S3.

Linear four-point resistivity probes 112

The limitation of current for accurate measurement in general is small fractions of amperes. Further, it is after convenient to present the current to 2πS milliamperes or microamperes so that the resistivity in ohm-centimeters will be numerically equal to the measured voltage in milli-or micro volts, respectively. Since the fourpoint probe offers the most convenient mode of resistivity measurement, variety of corrections have been developed which mostly fall into three categories cylindrical sample, slices and rectangular parallelepipeds. In general, the solution is such that simple multiplicative correct factors can be applied to give satisfactory accuracy.

Non-collinear probe spacing The probe array need not be linear and in principle can be of any configuration. The most commonly used is square. So, several others which have advantages for special application have been investigated.

Van der Pauw Method It is possible to determine R directly by placing four contacts on the periphery of the sample.

Delta Four-Point Probe This configuration has been developed in an effort to allow direct measurement of the high-resistivity layers which are in direct contact with low-resistivity substrates.

Over-under probe The placement of the probes is that, current flows through probe 1 and 3 and voltage is measured between probes 2 and 4.

Four-Point Probe Instrument The electric circuitry for a four-point probe can be quite simple, and requires only a probe, ammeter, voltmeter and source of current. 113

A sample of arbitrary shape used with van der pauw method

Basic electrical Circuitry

Errors in Four-point measurement (i)

Sample size: an error in thickness measurement translate directly into

resistivity error (ii) Substrate leakage: if the sample being measured is isolated from a substrate by a p-n junction, substrate leakage current can introduce errors. (iii) Probe spacing: any errors in determining the probe spacing translates into the same error in resistivity. (iv) Light: light shining on the surface may introduce spurious photo voltages which will cause instrumentation problem. (v) Temperature effect: A few percent errors can be introduced by unknowingly heating the sample during the measurement itself which usually occurs in lowresistive samples where large currents are required in order to obtain readily measurable voltages.

114

(vi) Thermoelectric effect: temperature gradients in the sample caused by excessive probe current will generate a thermoelectric voltage. Using low current will minimize the effect. (vii) Probe injection: for long-life time material, the contacts may inject enough carriers to cause conductivity modulation and seriously affect resistivity readings. (viii) AC pickup: DC sets are likely to have errors introduced through contact rectification of miscellaneous induced stray currents. So, operation in a shielded room is required. (ix)

Instrument Current: From a pure instrumentation standpoint, higher

currents make the voltage measurement easier and less susceptible to noise. (vi)

Applied Voltage: if the electric field becomes too high, a mobility

decrease occur which will make the resistivity reading too high.

Electrical resistivity, ρ, (also known as resistivity, specific electrical resistance, or volume resistivity) is a measure of how strongly a material opposes the flow of electric current. A lower resistivity indicates that a material readily allows the movement of electric charge.

The SI unit is ohm meter [Ωm]. Electrical conductivity, σ, or specific conductance is the reciprocal quantity, and measures a material's ability to conduct an electric current

115

Its SI unit is siemens per meter (S·m-1); where, ρ is the static resistivity (in ohmmeter, Ω-m), E is the magnitude of the electric field (in volts per meter, V/m); J is the magnitude of the current density (in amperes per square meter, A/m²). Many resistors and conductors have a uniform cross section with a uniform flow of electric current and are made of one material. In such a case, the above definition of ρ leads to:

Where, R is the electrical resistance of a uniform specimen of the material (in ohms, Ω),

is the length of the material (in meters), A, is the cross-sectional area

(in square meters, M²). The reason resistivity has the dimension units of ohmmeters can be seen by transposing the definition to make resistance the subject;

The resistance of a given sample will increase with the length, but decrease with greater cross-sectional area. Resistance is measured in ohms.

Conductivity type Doped and un-doped semiconductor thin films are known to be either n-type or ptype. This conductivity is determined by a technique called hot probe technique and the test is referred to as the hot point probe test. In this test a hot and a cold probe are connected with a voltmeter. The hot probe creates excess majority carriers which diffuses away leaving a net charge build-up.

Hot point probe showing (a) n-type and (b) p-type conductivity 116

This method is based on the fact that by definition a p-type thin film has holes as the excess majority carriers. When a hot probe is placed in contact with this film, the holes absorb this energy and as a result tend to migrate away from the hot site leaving electrons as the majority carriers. This causes the fact that if a voltmeter is connected between the two probes a voltage drop will be registered with its polarity opposite that of the majority carriers and hence for p-type film there will be a negative sign on the voltmeter. While it is accepted that the mobility of electrons is much greater than the mobility of holes the presence of so many holes is the significant factor. This is true for an n-type thin film.

Non-collinear four-point probe formulas and Guide

117

Measurement of optical properties Energy band gap (Eg) This is the energy gap between the conduction band and the valence band of a material. The energy band for any sample can be determined by drawing graph of (αhυ)2 versus the photon energy (hυ) plot. The intercept of the curve with the photon energy (hυ) axis gives the value of the direct band gap. The absorption coefficient (α) can then be calculated using the equation; α = 2.303A/d

where, α is the absorbance coefficient value at a particular wavelength (λ) and d is the thickness of the semiconductor film. Ternary and some other thin films are known to be direct band gap materials and for all direct allowed transitions, the absorption coefficient (α) is related to the band gap (Eg) by the relation: α = (hv – Eg)1/2 / hv

and therefore their optical band gap can be obtained by extrapolating their linear portions of the plots (αhv)2 versus hv to α = 0.

Refractive index (n) Knowledge of the refractive index of compound semiconductor is essential for designing lasers, opto-electronic devices, and photovoltaic cell applications. An empirical linear relation governing the variation of the optical refractive index n with the energy gap Eg for CdS as; n = 4.08 - 0.62Eg while another relation that avoids the shortcomings of the above relation to be as; n = 3.59 - log e(Eg)

Absorbance (α) Optical absorption gives the relationship between the absorption coefficient (α) and the photon energy hv, for direct allowed transition as; 118

α = (hv − Eg)1/2 Using the fundamental relations of photon transmission and absorbance, I = Ioeαt Where, t, is thickness and A = log Io/I , where, α = (2.303A)/t

Extinction coefficient (k) Extinction coefficient (k) can be computed from the equation; α= 4πk/λ which then gives the extinction coefficient equation as; k = αλ/4π Where (λ) is wavelength, (α) is the absorption coefficient.

Transmittance (T) Transmittance is the fraction of incident light at a specified wavelength that passes through a sample. It is a term related to absorptance, (α) or absorption factor, which is defined as the fraction of light absorbed by a sample at a specified wavelength. In equation form we have;

where I0 is the intensity of the incident light and I is the intensity of the light coming out of the sample and Tλ and Aλ are transmittance and absorptance respectively. In these equations, scattering and reflections are considered to be close to zero. The transmittance of a sample is sometimes given as a percentage.

119

Schematic diagram of a single-beam spectrophotometer

Schematic diagram of a working single-beam spectrophotometer Thin film post-treatment methods To improve a thin film’s physical properties many methods have been reported in order to post-treat semiconductor thin films. The use of these methods is necessary because it sometimes improves both the film’s physical properties and surface quality.

120

Annealing This is a process of heating a solid to a temperature just below the melting point and then cooling it slowly. Annealing process removes crystal imperfections and strains in the solid because thermodynamic equilibrium exists at temperatures not far below the melting point of a crystal. Air and vacuum thermal annealing are common methods in thin film treatment. Different gas treatments such as hydrogen, nitrogen and argon also have been used while annealing. More recent methods like optical annealing using light and laser have been used in thin film treatment. Annealing is intended to eliminate most of the damage produced by implantation process and to create a uniform impurity distribution in the polycrystalline materials.

Etching We have wet and dry etchings. In wet etching, the material is dissolved when immersed in a chemical solution. It is a simple technology which usually gives good results if one can find the correct combination of etchant and mask material suitable for this application. Wet etching works very well for etching thin films on substrates and can also be used to etch the substrate itself. In dry etching the material is sputtered or dissolved using reactive ions or a vapour phase etchant.

121

CHAPTER SIX

THE SOLAR CELL Solar cells are the smallest photovoltaic devices ever fabricated that are used either as irradiance sensors or as samples for studying new photovoltaic materials and/ or processes including reaction mechanisms. Solar cells range in size from a few square millimeters up to 156 mm square and even more for the case of silicon wafers. Photovoltaic’s is normally associated with images of rooftop mounted solar panels or a vast expanse of solar panel arrays spread out over a desert floor. This is probably because so much emphasis is placed on photovoltaics as an alternative way to generate electrical power. Solar cells not only generate little electrical power but also generate a lot of information of very great interest for the photovoltaic researchers or thin film analysists.

Prototype solar cell A research prototype solar cell usually requires probing but lacks the encapsulation so important for protecting solar modules from the degrading atmospheric and weather effects. A research prototype solar cell therefore, can have a rather crude looking construction compared to the sleek panels on display at any solar energy convention or any home where it is installed. It may simply be a thin film of a photovoltaic material sandwiched between two glass microscope slides with a silver paint used as electrical contacts. On the other hand, a solar reference cell is simply a small area (2 cm x 2 cm) solar cell packaged in a metal housing under a glass window intended for use in-doors to set simulated sunlight levels.

A solar reference cell can be framed in such a way that it resembles a miniature version of its associated solar panel. In place of a pyranometer, it can be used outdoors as an accurate irradiance sensor with the same spectral and angle of incidence responses as the real solar panel. There are five electrical performance

122

parameters to be discussed below that are used to characterize any solar cell and to compare it to other solar cells of the same or different materials. The challenge to making reliable electrical performance parameter measurements is setting up standard testing conditions, knowing what the actual test conditions are and at the same time accounting for all sources of error in order to express these parameters with associated measurement uncertainties.

The National ASTM standard E948 and international IEC standard 60904-1 both specifies a set of common test conditions and methods for measuring the electrical performance parameters of photovoltaic cells. The aptly named Standard Testing Conditions (STC) includes the following: (1). Temperature of the device under test (DUT) must be 25 ± 1 °C; (2). Spectral distribution of the light must be AM1.5 ± 25 % and (3). Irradiance measured at the plane of the solar cell must be 1 Sun ± 2 %. The sun is take as a unit of irradiance, one of which is equivalent to 1000 W/m2 or 100 mW/cm2. The amount of atmosphere through which sunlight passes to reach a given location on Earth is called Air Mass (AM) and it varies with that location’s air pressure, elevation, latitude, date and time of day. In a given day, AM is minimum when the sun is at its zenith and largest near the horizons. The test condition 1 Sun of AM 1.5 represents the average situation for the United States but for some combinations of locations and dates 1 Sun of AM1.5 may not occur. Given an AM of 1.5, testing out-doors may proceed only under a clear sky i.e. when there are no clouds within an angle of 30° around the sun.

A practical alternative method is to perform the photovoltaic measurements indoors using an I-V Test Station. This test station is based on Newport’s Oriel® Sol3A™ Class AAA solar simulator which collimates light produced by an intense Xenon source into various beam sizes from 2 to 12 inches square for testing purposes. The acceptable standard practice is to place a calibrated 123

reference solar cell along the simulator beam at the simulator’s working plane and adjust the current to the Xenon bulb until the short-circuit current produced by the reference cell matches the value published in its calibration certificate. The light at the working plane can be made “sunlike” by passing it through an aptly named Air Mass (AM) filter. This simulates the extended path length effect on the sunlight passing through the atmosphere at an obtuse angle of incidence. This ensures the spectral distribution at the working plane resembles the AM1.5 spectrum as referenced in tables like the international standard IEC 60904-3.

The acceptable standard practice is where one mounts a solar reference cell on a temperature controlled chuck and illuminate it with 1 Sun of simulated AM1.5 sunlight. After the cell temperature equilibrates to 25 °C, a variable electronic load placed across the cell is controlled such that the voltage across the cell is swept in small incremental steps. The electronic load can be a variable resistor, but is more often a programmable precision power supply.

Schematic of solar simulator

When the measured photocurrent is plotted against the bias voltage, the result obtained is a characteristic “I-V curve” for the solar cell. Three of the five 124

performance parameters, the short-circuit current Isc, the open-circuit voltage Voc, the maximum power point Pm, are derived from mathematical fits to different portions of the I-V curve. The parameters Voc and Isc are the intercepts of leastsquare fitted lines. Pm is the point at which the derivative with respect to voltage is zero for a fifth-order polynomial fit to power, the product of current and voltage. Fitting helps to reduce measurement noise and the Oriel I-V Test software automatically performs this fitting and analysis.

The parameters, Isc, Voc, and Pm are then used to calculate FF and PCE. Isc and Voc are the intercepts where the I-V curve crosses the current and voltage axes respectively, and the “knee” point at (Vm, Im) is where the solar cell delivers maximum power Pm. The FF or fill factor is the ratio of the area determined by Pm to the area determined by Voc and Isc. Power Conversion Efficiency is calculated by the following formula:

where, E m is the measured irradiance at the working plane of the solar cell, and Area is the surface area of the cell. For the purposes of research and early stage solar cells the PCE is the parameter of interest but for solar cells intended to be used as irradiance sensors, the short circuit current is what matters most.

A plot of bias voltage against power/current 125

Measurements of photovoltaic devices are subject to a number of errors. Some correctable errors arise because conditions deviate from the nominal STC conditions during the I-V sweep. The STC are expressed as ranges centered on the nominal conditions so that, the DUT temperature is allowed to be within 24 - 26 °C and so that irradiance may actually be between 0.98 - 1.02 Sun. Test conditions may be in tolerance but it worth to note that the performance parameters derived from I-V data under those conditions will still be in error. The measured current, Im can be corrected for each off-nominal condition according to;

where αIsc is the normalized temperature coefficient for Isc, M is the spectral mismatch factor, Em is the actual total irradiance measured with a solar reference cell. Eo and To are the nominal values for total irradiance (1 Sun) and temperature (25 °C). Only the measured current data (and not voltage data) is corrected since voltage is imposed across the cell by a power supply in sweep mode. The factor (Eo/Em) corrects the raw current data for the actual measured total irradiance during the time of test. It is recommended that Em is measured just before or just after an I-V sweep to achieve maximum accuracy.

The spectral mismatch factor (M) corrects the measured current for spectral mismatch error which will arise when the respective spectral responses and test spectra for the DUT and the solar reference cell do not match. Calculation of M requires four sets of data: the respective spectral responses and spectral irradiances for both the DUT and the reference cell. If the DUT and reference cell share either common spectral responses or spectral irradiances, then the spectral mismatch error will be zero but in general this is never the case.

In order to minimize the spectral error, M should be as close to unity as possible by carefully choosing a reference cell with a spectral response that closely 126

resembles that of the DUT and/or making the spectral distribution of the simulated light resemble the reference spectral distribution. For example, a GaAs solar cell should be calibrated with a GaAs reference cell, or dissimilar devices like an organic solar cell and a m-Si reference cell are made a good match by adding a KG5 window to the reference cell.

A rule of thumb is that if the spectral error exceeds 2%, a better matching solar reference cell must be found since a conservative estimate for the error in M is 20% of the absolute difference between M and unity. The reference cells can be equipped and calibrated with various colored glass filters. Another correctable error encountered in photovoltaic measurements employing simulated sunlight is spatial non-uniformity in the solar simulator beam. Simulated sunlight is typically more concentrated in the center than at the edge of the illuminated area and maps into a domed surface, the height of which can be used as a metric for spatial nonuniformity. Residual spatial non-uniformity causes irradiance error proportional to the relative areas of the solar reference cell and device under test and also the relative locations of the two. A factor can be calculated and applied to correct for this error.

Simulated sunlight more concentrated at the center

Some errors associated with characterizing PV cells are unavoidable and impossible to correct and the combination of these types of errors represents a 127

baseline limit to the accuracy with which the electrical performance parameters can be known. One unavoidable error source is associated with the calibration of the solar reference cell used to calibrate the cell under test.

Another unavoidable error source arises from the uncertainty in the cell’s temperature at the space charge region of the cell. A solar cell’s temperature is normally measured with either thermocouples or resistive temperature detectors (RTDs) which can either be attached to the surface in shadow (back) or to the exposed surface (front) of the cell. Temperature measured with a sensor attached to the back of the cell will be artificially low because the sensor itself tends to act like a heat sink. This effect is known as thermal shunting and can be minimized by choosing a temperature sensor with as little mass as possible to cut down on heat transfer between the cell and the sensor.

Thermal shunting can also occur when the temperature sensor is attached to the front surface and shadows the cell so that the temperature it measures will be artificially lower than that of the surrounding exposed cell area. Attachment of the temperature sensor to the back of the cell is preferable to attaching it to the front of the cell to avoid the additional error from shadowing the cell. A further challenge to knowing temperature of the cell accurately comes from the light induced temperature gradient in the cell because one side is exposed to light and the other is not. This effect is aggravated if the solar cell is mounted onto a temperature controlled chuck which cools the back of the cell by conduction more efficiently than the exposed surface of the cell exchanges heat with the surrounding air by convection.

The result is that the surface in shadow is typically a few degrees Centigrade cooler than the exposed surface of the cell. The resulting uncertainty in temperature of the cell translates into uncertainty in the performance parameters. The temperature gradient is apparent when the parameter Voc is plotted against temperature (at the back of the cell) when the cell is heated or cooled show an 128

error. As the best estimate of the open circuit voltage, the two lines then can be extrapolated to 25 °C using the temperature coefficient, and the mid-point of the two Voc intercepts at 25 °C can be calculated. In practice, guaranteeing that I-V data is collected at 25 °C becomes the greatest challenge in producing reliable photovoltaic measurements.

A set of four possible measurement scenarios depending on one’s knowledge of and ability to control the cell temperature and an associated relative measurement uncertainty arise. The first scenario is the ideal case and is the scenario prescribed in the PV standards. Calibration of solar reference cells fall into this category since a temperature sensor is built into the reference cell package and it can easily be mounted to a temperature controlled chuck. Illuminating the cell only during the I-V scan will perturb the cell temperature from the nominal 25 °C.

The bias voltage should be swept from V oc to 0 Volts rather than the other way around since Voc is more sensitive than Isc to the slight increase in temperature during the I-V scan. The short-circuit current, Isc, should then be collected after collecting Voc. The temperature drift during the I-V sweep can be eliminated by soaking the cell in 1 Sun of AM1.5 for several minutes. Eventually the cell temperature will reach equilibrium with the temperature controlled chuck. The temperature of the chuck can be set such that the cell equilibrates to 25 °C, and the I-V sweep can then be performed without disturbing the temperature of the cell.

Characterizing experimental solar cells often falls into the last scenario where there is no direct control or knowledge of the temperature. Due to their delicate construction, attaching a temperature sensor to the cell might cause permanent damage and precludes the ability to attach a temperature sensor. Bottom contacts (for the anode and cathode) require that the cell be probed from the bottom preventing mounting to a temperature controlled chuck so the DUT is essentially suspended in air during testing. The best one can do in this situation is to blow 129

temperature controlled air over the DUT and/or store the sample for several hours in a laboratory held to an average temperature of 25 °C before collecting the I-V curve with a solar simulator equipped with a fast shutter.

Plot of I-V showing minimum current and open voltage

Plot of I-V showing two scans

The fast shutter makes it possible to illuminate the cell briefly (less than 1 s) and avoid heating the cell appreciably to capture Voc at 25 °C. This process can be repeated to generate the entire I-V curve by stepping down the bias voltage from 130

Voc and measuring the current in response to momentary illumination. The cell should be allowed to sit at room temperature for several minutes before stepping the voltage down. A more practical method is to take successive (at least two) fast I-V curves. If the Voc corresponding to the second scan differs little from the V oc from the first scan, the cell was not perturbed significantly from 25 °C during the brief exposure to light.

In a similar manner, the respective maximum power points of the two scans can be compared for an estimate of STC temperature related error in Pm and in PCE. Care must be taken that the bias rate or sweep rate is not so fast that the cell can’t respond. This can be checked by comparing the FF from two scans taken with different bias rates. The second scenario is dealt with by applying the temperature correction to the I-V data. The uncertainty of the corrected results will depend on the uncertainty of Tcell. In the third scenario, the DUT is allowed to equilibrate with the chuck maintained at 25 °C. Temperature and temperature related effects can be inferred from monitoring Voc as is done in the fourth scenario. The small area solar cell is indispensable for the small scale study and optimization of new photovoltaic materials and processes, before scaling up to manufactured solar panels.

The national ASTM and international IEC standards for measuring the electrical performance parameters of photovoltaic cells allow for the calibration of solar reference cells or for comparing cells made from differing materials and/or processes. Photovoltaic cell measurements are to be performed under Standard Testing Conditions (STC). Reported results for the electrical performance parameters of a solar cell will be erroneous if the conditions during measurement deviated significantly from STC. Error sources that are not correctable are the main contributors to measurement uncertainty. Finally, a set of four possible measurement scenarios, depending on one’s knowledge and ability to control the cell temperature, is controlled in some ways to minimize measurement uncertainty for each scenario. 131

Solar Energy Conversion Technology Solar radiation represents the largest energy flow entering our terrestrial ecosystem. After reflection and absorption in the atmosphere, some 105 TW hit the surface of Earth and undergo conversion to all forms of energy used by humans with the exception of nuclear, geothermal, and tidal energy. This resource is enormous and corresponds to almost 6,000 fold the current global consumption of primary energy (13.7 TW). Thus, solar energy has the potential of becoming a major component of a sustainable energy portfolio with constrained greenhouse gas emissions. Solar radiation is a renewable energy resource that has been used by humanity in all ages and for various uses.

Passive solar technologies were already used by ancient civilizations for warming and/or cooling habitations and for water heating and in the Renaissance. Concentration of solar radiation was extensively studied and in the 19th century the first solar-based mechanical engines were built. The discovery of photovoltaic effect by Becquerel in 1839 and the creation of the first photovoltaic cell in the early 1950s opened entirely new perspectives on the use of solar energy for the production of electricity. Since then, the evolution of solar technologies continues at an unprecedented rate. Nowadays, there exist an extremely large variety of solar technologies and photovoltaics have been gaining an increasing market share for the last 20 years. Nevertheless, global generation of solar electricity is still small compared to the potential of this resource.

The current cost of solar technologies and their intermittent nature make them hardly competitive on an energy market still dominated by cheap fossil fuels. From a scientific and technological viewpoint, the great challenge is finding new solutions for solar energy systems to become less capital intensive and more efficient. Low-cost and /or high-efficiency photovoltaic device concepts are being developed. Solar thermal technologies are reaching a mature stage of development and have the potential of becoming competitive for large energy supply. Intermittency is being addressed with extended research efforts in energy 132

storage devices such as batteries and other electric storage systems, thermal storage and the direct production of solar fuels.

All these are valuable routes for enhancing the competitiveness and performance of solar technologies. The aim of this chapter is to evaluate the potential of solar energy for low-carbon intensive and large-scale energy production and to provide a picture of the state of research in the most significant solar technologies. More than a comprehensive review, this chapter is intends to be an attempt at identifying interdisciplinary and fundamental topics with high breakthrough potential for the improvement of the performance, reliability, and competitiveness of solar technologies.

Solar Radiation Solar radiation is an electromagnetic wave emitted by the Sun’s surface that originates in the bulk of the Sun where fusion reactions convert hydrogen atoms into helium. Every second 3,891,026 Joules of nuclear energy is released by the Sun’s core. This nuclear energy flux is rapidly converted into thermal energy and transported toward the surface of the star where it is released in the form of electromagnetic radiation. The power density emitted by the Sun is of the order of 64 MW/m2 of which ~1370 W/m2 reaches the top surface of the Earth’s atmosphere with no significant absorption in the space. The latter quantity or amount of power is called the solar constant.

The spectral range of the solar radiation is very large and encompasses nanometric wavelengths of gamma- and x- rays through metric wavelengths of radio waves. The energy flux is divided unevenly among the three large spectral categories. Ultraviolet (UV) radiation (λ < 400nm) accounts for less than 9 % of the total; visible light (VIS) (400nm < λ < 700nm) for 39 %; and infrared (IR) for about 52 %. As shown in the figure below the pattern of the solar spectrum resembles closely the radiation of a perfect black body at 5800K. In the figure 133

shown, AM0 indicates the Air Mass Zero reference spectrum measured and partially modeled outside the terrestrial atmosphere. Radiation reaching the Earth’s surface is altered by a number of factors namely, the inclination of the Earth’s axis and the atmosphere that causes both absorption and reflection (albedo) of part of the incoming radiation.

Solar energy emitted to the earth’s surface

The influence of all these elements on solar radiation is visible in the ground-level spectrum, labeled AM1.51 in the figure where the light absorption by the molecular elements of the atmosphere is particularly evident. Accounting for absorption by the atmosphere, reflection from cloud tops, oceans, and terrestrial surfaces, and rotation of the Earth (day/night cycles), the annual mean of the solar radiation reaching the surface is 170W/m2 for the oceans and 180W/m2 for the continents. Of this, about 75% is direct light, the balance of which is scattered by air molecules, water vapor, aerosols, and clouds.

134

Shaded surfaces represent natural energy destruction; arrows represent human use for energy services

This quantity represents the upper limit to the work obtainable from solar radiation conversion, a limit that is imposed by the second law of thermodynamics and is independent of any conceptual device. Of the 162 PW of solar radiation reaching the Earth, 86 PW hit its surface in the form of direct (75 %) and diffused light (25 %). The energy quality of diffused radiation is lower (75.2 % of energy content instead of 93.2 % for direct light, with consequences on the amount of work that can be extracted from it. A woping 38 PW hit the continents and a total energy of 0.01 TW is estimated to be destroyed during the collection and use of solar radiation for energy services. This estimation includes the use of photovoltaics and solar thermal plants for the production of electricity and hot water. Similar estimates are shown for wind energy (0.06 TW), ocean thermal gradient (not yet exploited for energy production), and hydroelectric energy (0.36 TW).

Potential of Solar Energy The global solar energy potential ranges from 2.5 to 80 TW. The lowest estimate represents around 18 % of the total current primary energy consumption (13.7 TW) and exceeds 10 % of the estimated primary energy demand by 2030 (21.84 135

TW). More optimistic assumptions give a potential for solar energy exceeding 5 fold the current global energy consumption. Despite the relatively low power density of the solar flux, solar energy has the potential of supplying a nonnegligible fraction of our energy needs. In the case of the US for example, the total electricity demand (418 GW) in 2002 could be satisfied by covering a land surface of 180 km square with photovoltaics.

Current electricity generation from PVs is only of the order of 2.6 GW compared to 36.3 GW for all renewable energies, hydroelectric power excluded. Developed countries are steadily increasing their investments in solar power plants, and IEA projections for 2030 give an enhancement of solar electricity generation up to 13.6 GW (80 % of which will be from photovoltaics, and the rest (2.4 GW) from solar thermal plants). However, this amount will not exceed 6% of the total electricity production from non-hydro renewable energies. It is worth noting that passive solar technologies for water heating, not included in these statistics, represent a fairly large amount of power. IEA estimates a power production of 5.3 GW in 2002 and an increase up to 46 GW by 2030.

The major causes of the slow deployment of solar technologies are: (i) The current relative high capital cost per kW installed compared with other fossil fuel based and renewable technologies; (ii) The intermittent nature of the energy input, and hence the requirement for energy storage systems to match the energy supply with the electricity demand and to decrease the capital cost. In a medium term, energy storage will be a key requirement for intermittent renewable energies to become more competitive versus fossil fuels. If we want solar energy to significantly contribute to the world’s energy supply, massive increases in manufacturing capacity are needed. From the research standpoint, more effort has to be put into improving efficiencies while reducing the manufacturing costs. This is a great technological challenge that requires investment of larger financial and intellectual resources to find innovative solutions.

136

Progress in photovoltaic cell efficiencies Despite the notable progress made in the improvement of the efficiencies of all these technologies, achieved values are still far from the thermodynamic efficiency limits of ~31 % for single junctions, 50 % for 3-cell stacks, impurity PVs, or up- and down converters, and 54 – 68 % for hot carrier- or impact ionization-based devices. Furthermore, the efficiencies of commercial modules are only about 50 % to 65 % of these “champion” cells. Closing these gaps is the subject of ongoing research. The solar-to-electric efficiency of solar thermal technologies varies largely depending upon the solar flux concentration factor, the temperature of the thermal intermediary and the efficiency of the thermal cycle for the production of mechanical work and electricity.

Parabolic troughs and power towers reach peak efficiencies of about 20 %. DishStirling systems are the most efficient with about 30 % solar-to-electric demonstrated efficiency. The performance of these systems is highly influenced by the plant availability. In the case of parabolic troughs and power towers, thermal storage increases the annual capacity factor from typically 20 % to 50 % and 75 %, respectively.

Over the past 30 years on solar cell efficiencies

137

Environmental Aspects of Solar Energy Solar energy is promoted as a sustainable energy supply technology because of the renewable nature of solar radiation and the ability of solar energy conversion systems to generate greenhouse gas-free electricity during their lifetime. However, the energy requirement and the environmental impact of PV module manufacture can be further reduced, even though recent analysis of the energy and carbon cycles for PV technologies recognized that strong improvements were made both in terms of energy and carbon paybacks. In the case of pc-Si in Southern Europe, energy payback calculations are not straightforward because today’s

PV

industry usually recrystallizes

silicon

recycled

from the

semiconductor industry.

For thin films, the energy required to deposit the active layer is negligible compared to forming crystalline silicon wafers. Instead, the major energy sink is the energy embodied in the glass or stainless steel substrate, the film deposition process, and facility operation. These energy costs are similar for all thin-film technologies (CIGS, CdTe, α-Si), varying only in the film deposition processes. An estimate for the frameless α-Si module electricity requirement is 330kWh/m2 (4.3kWh/W).

In a rooftop- or ground-mounted, grid-connected PV system the BOS components and module frames represent a non-negligible fraction of the total energy requirement. For a rooftop-mounted system another 120 kWh/m2 should be added to the overall life cycle energy requirement, resulting in a payback time of about 3.5 years. Support structures for ground-mounted systems would add about another year to the payback period. Despite the wide range of payback times that can be found in the literature, all estimates remain higher than for other renewable sources such as wind. It is interesting to note that analysis of fossil-fuel energy production has suggested that it has similar energy payback periods to PV technologies if the costs for mining, transportation, refining, and construction are included in the calculation of the life cycle of fossil fuels. 138

The CO2 savings (other pollutants are also avoided, including NOx, SO2 and particulates) from displacing fossil fuels with photovoltaic systems depend upon the regional fossil fuels mix and the solar irradiance values range from 270 g to greater than 1050 g of CO2/kWh. The world average is about 660 g of CO2/kWh. Assuming an average of 5.5 hours of sunlight per day, a 1kW PV panel would give a yearly CO2 savings of 1330 kg. The CO2 payback time from avoided emissions also depends on the local energy mix and the panel efficiency. Assuming an energy cost for a sc-Si panel of 600 kWh/m2 and the average of 660 g of CO2/kWh, the manufacture of a 1m2 Panel produces about 400 kg of CO2. If we assume 12% efficiency and a solar irradiance of 1kW/m2, it takes 3300kg of CO2 to produce a 1kW PV plant which is paid back in avoided emissions at 1330 kg/year for a total time of 2.5 years. Higher cell efficiencies lower both the energy and CO2 payback time, as do manufacturing techniques that are more energy efficient.

Safety and environmental issues The major safety and environmental issues related to the manufacture of photovoltaics are: (i) the safe handling of gases used for surface treatment or the growth of thin films (e.g. AsH3, SiH4, GeH4, PH3, B2H6, and H2Se), and (ii) the toxicity of some semiconductor components (e.g. Cd). It is generally believed that safe usage of potentially hazardous materials in PV manufacturing is possible and that the electronics industry has already made significant progress in dealing with similar materials. Nevertheless, further investigation could lead to the replacement of toxic components and thus eliminate most of the concerns about the environmental risks of photovoltaics.

Solar technologies A wide variety of solar technologies have the potential to become a large component of the future energy portfolio. Passive technologies are used for indoor lighting and heating of buildings and water for domestic use. Also, various

139

active technologies are used to convert solar energy into various energy carriers for further utilization:

Photovoltaics directly convert photon energy into electricity. These devices use inorganic or organic semiconductor materials that absorb photons with energy greater than their band-gap to promote energy carriers into their conduction band. Electron-hole pairs, or excitons for organic semiconductors, are subsequently separated and charges are collected at the electrodes for electricity generation.

Solar thermal technologies convert the energy of direct light into thermal energy using concentrator devices. These systems reach temperatures of several hundred degrees with high associated energy. Electricity can then be produced using various strategies including thermal engines (e.g. Stirling engines) and alternators, direct electron extraction from thermionic devices, Seebeck effect

in

thermoelectric generators, conversion of IR light radiated by hot bodies through thermo-photovoltaic devices, and conversion of the kinetic energy of ionized gases through magneto-hydrodynamic converters. Photosynthetic,

photo-(electro)-chemical,

thermal,

and

thermochemical

processes are used to convert solar energy into chemical energy for energy storage in the form of chemical fuels, particularly hydrogen. Among the most significant processes for hydrogen production is direct solar water splitting in photo-electrochemical cells or various thermochemical cycles such as the twostep water-splitting cycle using the Zn/ZnO redox system. This analysis focuses on the fundamental physical processes that govern the operation of the solar devices with the intent of identifying common trends or synergies between different technologies that could help identify novel research opportunities.

140

Photon Energy Conversions flow

Photon-to-Electric Energy Conversion Photovoltaic devices allow the direct production of electricity from light absorption. The active material in a photovoltaic system is a semiconductor capable of absorbing photons with energies equal to or greater than its band-gap. Upon photon absorption, an electron of the valence band is promoted to the conduction band and is free to move through the bulk of the semiconductor. In order for this free charge to be captured for current generation, decay to the lower energy state, i.e. recombination with the hole in the valence band, has to be prevented through charge separation. In photovoltaic devices made of inorganic semiconductors, charge separation is driven by the built-in electric field at the p-n junction. As a consequence, their efficiency is determined by the ability of photogenerated minority carriers to reach the p-n junction before recombining with the majority carriers in the bulk of the material. Thus, bulk properties such as crystallinity and chemical purity often control the device efficiency.

141

Schematic of a p-n junction (left), organic bilayer structure (right)

The operation of organic photovoltaics (OPVs) is fundamentally different. The optical and electronic properties of organic semiconductor materials are determined by the molecular orbitals that are built up from the summation of individual atomic orbitals in the molecule. The molecule’s properties, and in particular its band-gap, are determined by the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO). Light absorption in either small molecules or in conjugated polymers leads to the formation of an exciton, i.e. an electron-hole pair that is bound together by Coulomb attraction that must be dissociated.

A built-in electric field can be created by sandwiching an organic semiconductor between two semiconductors with different work functions, but this method is not effective in splitting excitons. Instead, efficient exciton dissociation occurs at the interface between a donor material, where the exciton is created, and an acceptor material with an empty energy level that is lower than the LUMO of the donor. Exciton dissociation at the hetero-junction produces electrons on one side of the interface already separated from the holes produced on the other side of the interface. This creates a photo-induced interfacial chemical potential energy gradient that efficiently drives the photovoltaic effect, even in the absence of a built-in electrical potential.

142

The efficiency of these devices is determined by the requirement that excitons reach the donor-acceptor interface charges are transferred before recombination occurs, and charges are subsequently transported to the electrodes before electrons back-transfer from the LUMO of the acceptor to the HOMO of the donor. In both inorganic and organic photovoltaic technologies, many strategies are under investigation for achieving efficient light absorption, charge separation, transport, and collection. The present analysis covers these fundamental processes and spans over a large range of technologies based on inorganic semiconductor materials such as silicon (c-Si, pc-Si, or α-Si), III-V compounds (e.g. GaAs, InP), chalcogenides (e.g. CdTe, CIGS), and various organic-based thin films:

(i) Photo-electrochemical cells, or Dye-Sensitized Solar Cells (DSSCs), where the light absorption occurs in organic dyes adsorbed on the surface of a wideband-gap nanostructured metal oxide semiconductor substrate, usually TiO2; upon excitation, electrons are injected into the conduction band of the oxide semiconductor and holes are scavenged by a redox couple in solution, such as iodide/triiodide (I-/I3 -). Solid-state photo-electrochemical cells use an organic semiconductor or ionic medium as a replacement to the liquid electrolyte;

(ii) Multilayer organic planar devices, in which molecules are deposited sequentially to form a stacked device; (iii) Organic bulk hetero-junctions, where organic donor and acceptor materials are blended on the nano-scale; (i) Organic/inorganic composites such as hetero-junctions combining light-

absorbing conjugated polymers and large-band-gap nanostructured inorganic material such as TiO2 or ZnO;

(v) Artificial photosynthetic macromolecular structures where light-absorption and charge separation are realized by separated complexes within the same structure. Advanced thin-film technologies called “3rd generation photovoltaics”,

143

are considered as a promising route to increasing the efficiency and/or lowering the cost of photovoltaics.

Efficiency-cost trade-off for the three generations of solar cell technology: [wafers (I), thin films (II), and advanced thin films (III)]

Photon absorption and carrier generation One of the most critical requirements for a single junction cell is that the band-gap energy must be optimized to transfer maximum energy from the incident light to the photo-generated electron-hole pairs. The simultaneous optimization of the cell voltage proportional to band-gap, the photo-generated current density, decreases with band-gap and of the fill factor where optimal value of band-gap is in the range of 1.1 - 1.4 eV. The band-gap energy of silicon (1.12eV) is almost ideal and allows absorption of photons in the near-infrared (NIR), visible, and ultraviolet spectrum.

However, the indirect band-gap of crystalline silicon causes relatively poor light absorption ( less than 104 cm-1) for photons with energies below 3.4 eV, which is equivalent to the silicon direct band-gap energy. Typical sc-Si wafers must be 100 - 300 μm thick for achieving efficient light absorption. Thin-film photovoltaic 144

materials have a major advantage over silicon since most of them have direct band-gap resulting in higher optical absorption. This allows thin film PV devices to use very thin layers of active material (about1μm) that can thus be of lower quality.

Today’s most successful materials for thin-film photovoltaics are α-Si where the optical absorption is increased by impurity scattering, CdTe with a band-gap of 1.48 eV, and CIGS, whose band-gap can be tuned around the nominal value of 1.04 eV by controlling its composition and that has the highest absorption constant (3 – 6 x105 cm-1) reported for any semiconductor. More effort is required to find new semiconductor materials combining optimal band-gap, inactive grain boundaries, stability properties, and ease in processing.

Spectrum splitting through multijunction cells with band-gap energies designed to match the solar spectrum is a very effective route to achieve increased efficiency since this method reduces the energy loss driven by the thermalization of hot electrons generated by the absorption of photons with energy greater than the band-gap. Many configurations and materials have been investigated for tandem and multijunction cell concepts. Among the most interesting approaches using silicon are the: Amorphous silicon-germanium alloys (a-Si,Ge:H) where the band-gap can be varied from 1.75 eV down to below 1.3 eV;

Microcrystalline and amorphous silicon tandem cells (μc-Si:H (1.12 eV)/α-Si:H (1.75 eV), with enhanced stability properties against light-induced degradation and with maximal and stable efficiencies of 14.7 % and 10.7 %, respectively;

Multijunctions

incorporating

material

alloys

such

as

amorphous

or

polycrystalline silicon carbide (α-Si:C) and silicon germanium (α-Si:Ge). III-V materials have ideal band-gap energies for highly efficient photon absorption (e.g. 1.0 - 1.1 eV for InGaAsN, 1.4 eV for GaAs). In addition, fine-tuning of both 145

lattice constant and band-gap can be achieved by modifying the alloy composition resulting in a large flexibility that is exploited for growing multijunction cells.

Lattice-matched and

metamorphic three-junction GaInP/GaInAs/Ge cells

currently hold the efficiency records under concentrated sunlight (39 % efficiency at 236 suns and 37 % efficiency at 310 suns, respectively). The cost of growing processes such as molecular beam epitaxy and metal-organic vapor phase epitaxy directed these technologies toward space applications but their inclusion in concentrator systems together with manufacturing scale-up might have a sensible impact on their cost for terrestrial applications. To achieve this goal however, concentrating technologies will require more technical development.

Nanoscale features are widely used in solar technologies to increase light absorption. In particular, quantum dot sensitization has large potential for matching the absorption spectrum of a photovoltaic cell to the solar spectrum. Nanoparticles can be built from a large variety of semiconductor materials and their band-gap can be tuned by changing the particle size and shape. Recent experimental results have demonstrated the feasibility of multiple carrier generation through impact ionization in PbSe nanocrystals for photon energies 3 fold larger than the nanocrystals band-gap energy.

Impact ionization can potentially increase the power conversion efficiency of a solar cell based on PbSe nanocrystals by 35 – 40 %. Extremely Thin Absorber devices are another example of systems taking advantage of nanoscale structures. The interest of this design is the tolerance to higher levels of defects and impurities than in flat thin-films devices because photoinduced charge separation occurs on a length scale of a few nanometers. On the other hand, making PIN junctions (p-type semiconductor/insulator/n-type semiconductor) with such high contact area is difficult and this has hampered the performance of these cells.

146

Multijunction cells and multiple electron-hole pair generation are two among a set of novel approaches that could be denoted as “3rd generation PVs” that aim at increasing the thermodynamic efficiency limit of solar cell devices. Research efforts in these technologies are increasing with a feasibility of some of them still yet to be proved experimentally. The following is a non-exhaustive list of these advanced solar conversion options:

hot carrier cells, where the rate of photoexcited carrier cooling caused by phonon interaction in the lattice is slowed down to allow time for the carriers to be collected whilst they are still “hot” thus reducing thermalization losses. Efficiency limits for these systems are of the order of 54 – 68 %.

up/ down converters, converting respectively high energy photons ( less than 2 band-gap) into two lower-energy photons with energy greater band-gap, and vice versa. The thermodynamic efficiency limit for a solar cell with a band-gap energy of 2 eV and with an optimized up-converter attached to its rear surface is 50.7 % for a nonconcentrated AM 1.5 spectrum;

multiband cells where one or more electronic energy levels are created in the forbidden band of the bulk semiconductor material through superlattice structures based on a periodic structure of alternating layers of semiconductor materials with wide and narrow band gaps, high concentration impurities such as rare-earths in wide band-gap semiconductors or by using semiconductors with multiple narrow bands such as I-VII and I3-VI compounds.

thermophotovoltaic devices involving the photovoltaic conversion by a receiver cell of radiation from an emitter which could be heated by various sources including sunlight. A prime difference from normal 147

solar photovoltaics is that emitted energy unable to be used by the receiver can be recycled allowing conversion efficiency up to 54 %.

surface plasmon on metal nanoparticles used to enhance the light absorption of thin semiconductor layers by coupling the light with the waveguide modes of the semiconductor layer.

solar antenna (or “rectenna”) arrays use a micro-scale antenna to convert broadband electromagnetic radiation into an AC field and optical frequency rectifiers to provide a DC electric output with theoretical efficiency limit is above 85 % under direct sunlight.

One of the key advantages of organic photovoltaics (OPVs) is that organic small molecules and polymer materials have very high absorption coefficients exceeding 105 cm-1 that permit the use of thin films with thicknesses of only several hundred nanometers. Current OPV devices exhibit high (above 70%) quantum efficiency however, obtaining absorption in the NIR spectral range has proven to be challenging. Band-gaps of the active organic materials must be reduced to approach the nominally optimal value of 1.4 eV while retaining good charge carrier mobility, open-circuit voltage and the efficiency of charge separation.

Conjugated polymers with band-gaps approximately 1.6 eV have been reported and various Ruthenium complex dyes absorbing up to 900 nm (“black dye”) have been used in photoelectrochemical cells. Combinations of active donor and acceptor materials in organic hetero-junctions and of different dyes in photoelectrochemical cells can be used to broaden the absorption spectrum. In this context then, light-harvesting antennae are an interesting alternative sensitizer. For example, the chromophore-loaded zeolites have demonstrated the ability to incorporate high densities of various dye molecules complementing each other in their spectral features. 148

These systems also increase the dye stability by preventing their aggregation and display very efficient Föster energy transfer processes. Alternatively, some 3rd generation approaches that have been developed can be also applied to organicbased devices. Inorganic semiconductor (e.g. InAs or PbS) nanocrystals may be combined with organic materials for larger flexibility in terms of spectral absorption, in particular in the NIR spectral range. Sensitization properties in the infrared spectrum (1000 – 1600 nm) have been reported for organically terminated PbS quantum dots used in conjunction with MEH-PPV. Additionally, quantum confinement can enhance the strength of absorption.

Organic nanocrystals have also been used in bulk heterojunctions with enhanced light absorption properties such as PCBM, a C70 derivative used as acceptor material in conjunction with MDMO-PPV as the donor material. Tandem cells can be made of organic active layers to increase the photon absorption efficiency and the open-circuit voltage up to 5.7 % efficiency tandem cell where two CuPc/C60 cells where connected in series by using a thin layer of Al (aluminium) nanoparticles as a recombination site in the center of the device. This strategy represents a major challenge to materials scientists to develop donor-acceptor combinations that have comparable efficiencies across the solar spectrum. The constraint that each sub-cell in a tandem structure must generate an equal current under 1 sun illumination intensity will most likely limit the number of cells to three cells only.

In addition, cells should be very thin in order to reduce the series resistance and thus maximize their fill factor. Innovative strategies are hence required to couple the sub-cells while concentrating the incident radiation directly into the regions that are most photoactive. Using surface plasmons from metal particles (which are collective electron surface motion excited by incident light at their resonance frequency) between the sub-cells is regarded as one of the possible technique to solve this problem. Parallel sub-cell connection is an alternative way to series connection which would not require current matching between the individual 149

cells. This configuration requires highly transparent contacts with good lateral conductivity to extract the photocurrent without incurring a significant voltage drop across the device diameter. Optically transparent nano-patterned metal films are under investigation as a replacement for ITO electrodes for high light transmission and high parallel conductivity connections in multijunction OPV cells.

Charge transfer and separation Charge carriers generated upon photon absorption in inorganic semiconductors are free to move independently. Carriers that reach the depletion region across the p-n junction before any re-combination occurs get separated by the built-in electric fields. The efficiency of charge separation depends upon the competition between recombination processes and charge transport, which will be discussed in the next section. Photo-excitations in organic semiconductors result in the formation of excitons or electron-hole pairs that are bound together by Coulomb attraction and must be dissociated. Dissociation can happen in the presence of high electric fields at a defect site in the material or at the interface between two materials that have sufficient mismatch in their energy levels.

In most organic-based photovoltaics, exciton dissociation is mostly governed by interfacial mechanisms. The exciton dissociation is very effective resulting in the transfer of electrons from the donor to the acceptor material and holes from the acceptor to the donor material with efficiencies approaching 100 % on subpicosecond time scales such as in the case of MDMO-PPV/PCBM blends. This mechanism is not well understood and there are many open questions on the kinetic requirements and on the role of interface polarizability, electric field, exciton transport rates, and interfacial electronic states. Before dissociation occurs, the exciton created in either donor or acceptor material has to diffuse to the interface before recombining.

150

As a consequence of the weak interaction between molecules, the exciton diffusion length is very small on the order of 10 nm. Together with light absorption efficiency, the exciton diffusion efficiency is the most critical limit to the performance of organic PVs and constrains the cell structure. Transparent, wide band-gap and electron conducting exciton blocking layers (EBL) are used to confine the excitons close to the donor-acceptor interface and thus reduce the effective length that excitons have to travel.

Materials with long diffusion lengths can be used to enhance the exciton diffusion efficiency where C60 is an example of such a material. The use of C60 as the acceptor material in a double hetero-junction has caused the exciton diffusion length to increase from 3 nm to 40 nm. Organic materials can also be engineered to enhance the exciton diffusion by introducing heavy metal ions into the structure of a molecular dye.

The distance that excitons have to travel to the interface to be short and therefore, planar devices must consist of very thin active layers thus limiting their optical density. Bulk heterojunctions, where donor and acceptor materials are blended on the nanoscale with a very large resulting interfacial area solve this problem by distributing the interface throughout the device. A complication to this process is that the preferential dissociation sites for excitons are not always the same. In photo-electrochemical cells based on inherently conducting polymers (ICP), exciton dissociation may occur at the ITO-polymer interface if liquid electrolytes are used or at the ICP-electrolyte interface when the electrolyte is a solid polymer. The influence of the cell structure and of the polymer morphology on the preferential dissociation site is not understood.

In nano-crystal-polymer blends, the interaction between organic and inorganic materials strongly impacts the cell performance. The morphology and the interfacial trap states must be controlled to increase the charge transfer efficiency. In this respect, promising results have been shown by binding phosphonic-acid151

functionalized oligothiophenes to the surface of CdSe nanocrystals leading to facilitated electronic interaction and passivation of trap states. The donor/acceptor band offset or the nature of the redox couple in the case of photoelectrochemical cells must be optimized to yield the highest possible photovoltage. The resistance between layers must be minimized to achieve high filling factors.

Schematic of flat (left) and bulk (right) hetero-junction structures (fundamental steps of the photovoltaic process

Charge transport The c-Si cells need a relatively large thickness because of mechanical constraints and the long light absorption length associated with the indirect band-gap of Si. Consequently, good material with high chemical purity and structural perfection is required to fight recombination. The surfaces used must be effectively passivated to reduce recombination and impurities and imperfections in the bulk state must be avoided as they can absorb extra energy of the conduction-band electrons and convert it into heat energy.

Bulk recombination is caused by lattice imperfections derived from doping the bulk Si (e.g. with phosphorus and boron). Main recombination mechanisms are the Shockley-Read recombination

which is the electron-hole recombination

through imperfections giving up the recombination energy as photons or phonons and the Auger recombination in which the recombination energy is given up to 152

another free carrier. Various deposition and growing or fabricating technologies, surface treatments and contact designs has allowed the incremental enhancement of the electronic properties of Si wafers. Charge transport in thin-film photovoltaics is mainly limited by grain boundary and defect states. In α-Si:H thin films films, large carrier diffusion lengths require a low density of defect states in the gap.

These defect states are most commonly associated with dangling bond defects. Charge transport in CdTe cells is limited by junction interfaces and structural defects. The efficiency of these devices is enhanced by post-deposition heat treatments at temperatures above 400 oC. In the case of CIGS devices, the electrical conductivity is largely determined by native defects which includes vacancies, interstitials and antisite defects whose density can be decreased by optimizing the Cu/In ratio. In organic materials the charge transport is determined by the intermolecular overlap of the frontier orbitals (HOMO and LUMO) of adjacent molecules. Charge transport in these materials is based on carrier hopping processes between molecules.

Carrier mobilities in OPVs are much lower than in inorganic semiconductor materials. Hole mobilities reported for conjugated polymers range from 10-7 cm2V-1s-1 to 10-1 cm2V-1s-1 and electron mobilities from 10-9 cm2V-1s-1 to 10-4 cm2V-1s-1, compared to 475 cm2V-1s-1 for holes and 1500 cm2V-1s-1 for electrons in crystalline silicon. Even though mobility in organic materials is usually sufficient to allow extraction of charges in OPVs, transport time for carriers to reach the electrodes has to be decreased to compete with back recombination. Electron and hole transport in polymers can be enhanced by chemical doping through covalent attachment of functional groups.

The carrier mobility in organic based cells may be increased using heterojunctions of organic materials and inorganic semiconductors with much higher electron mobilities. Hole mobility in organic bulk heterojunctions is largely influenced by 153

their morphology which must be optimized under the constraint of allowing adequate percolation of the donor and acceptor phases. In polymer-based devices, the alignment of the polymers is also critical for charge transport. In particular, polymers with molecules π-stacking on one another are required to increase hole mobility by facilitating hopping between adjacent molecules.

In devices using small-molecule materials, crystalline molecular organic phases can be synthesized to increase charge carrier transport. Phase separation processes in the synthesis of bulk heterojunctions often create resistive bottlenecks and culde-sacs where the free charges are trapped prior to collection at the electrodes. Increasing the order of bulk heterojunctions enhances the charge separation and transport since producing ordered structures allows control of the phase separation of the donor and acceptor to exciton diffusion length scale (about 10 nm). This is to avoid dead ends in one of the phases by providing straight pathways to the electrodes for electrons and holes after exciton dissociation and to align conjugated polymer chains.

Various technologies have been explored to control the morphology of organic and organic-inorganic bulk heterojunctions. In particular, incorporation of polymers into a mesoporous inorganic substrate (TiO2, ZnO, CdS) has been achieved using various techniques. Obtaining a highly aligned polymer structure with high hole mobility needs further research however, some self-assembled polymer structures in mesoporous substrates have already been produced. Bioinspired photosynthetic systems are macromolecular structures that are composed of light-harvesting pigment-protein complexes connected to reaction centers where the captured excitation energy is converted into electrochemical potential energy by photo-induced electron transfer.

The light-harvesting complex consists of covalently linked chromophores antennae that absorb light and transfer excitation energy to a central site at which charge separation occurs. The synthesis of such light-harvesting molecular arrays 154

presents many challenges. Current research underway at the time of writing this book aims at developing self-assembling and self-ordering modules that would allow the formation of functional and robust units. After photon absorption in the light-harvesting complex which forms the donor and the subsequent charge separation, efficient charge transport to the reaction center which is the acceptor is achieved

by

using

donor-bridge-acceptor

series

incorporated

in

the

macromolecular structure.

The configuration of the molecular chain (the bridge) separating the donor from the acceptor has to be optimized to achieve efficient electron transport within the macromolecule. For example, energy levels and redox components have to be tuned to foster incoherent (hopping) versus coherent (super exchange) electron transport since the former is best suited for long-distance molecular transport. More generally, understanding electron transfer in artificial photosynthetic systems needs further systematic investigation of the influence of the donoracceptor distance and orientation, the free energy of the reaction, and electronic interaction.

Technological challenges Inorganic PV devices The main factors limiting inorganic PV efficiency include; the mismatch between the solar photon spectrum and the semiconductor band-gap, optical losses due to reflection off the cell surface or shadowing by the conductor grid that collects the electric current, recombination of electron-hole pairs and the resistance of the metal-semiconductor contact. Crystalline silicon technologies have a potential for further incremental improvement in performance and cost reduction. The most successful c-Si cell designs of the last 15 years with efficiencies above 20 % are the photolithographically-based Passivated Emitter Solar Cell (PESC), the Back Point-Contact (BCP) cell of 22.3 % efficiency) and the Passivated Emitter Rear Locally-Diffused (PERL) of 24.7 %. 155

As it stands now, future developments in c-Si cells are likely to be decrease wafer thickness and thus improved light-trapping schemes. Enhanced buried contact designs and heterojunctions with Intrinsic Thin-layer (HIT) are two of the newer commercial designs. Ribbon technologies such as Edge defined Film-fed Growth (EFG), dendritic web, and string ribbon, have the potential to produce thin cells at lower cost. Deposition of Si on foreign substrates is a possible route for obtaining low-cost cells with acceptable performances. Polycrystalline silicon on SiC graphite coated substrates, expansion matched conducting ceramic substrates, and glass substrates are examples of technologies under investigation, with the latter now in commercial production in most developed countries.

In thin film technology, technical issues that include module efficiency, manufacturing scale-up, yield or throughput and module reliability are still under investigation. Research in thin-film technologies often aim at incremental improvement in all these aspects. However, there are still some fundamental material properties and processes that need to be understood. For example, the light induced degradation processes of α-Si:H devices and the Stabler-Wronski effect are major challenge to enhance steady-state efficiency of this type of cells above 6 – 9 %.

The mechanisms of this degradation process are not yet fully understood and researchers try to overcome it either by developing new deposition approaches such as the “hot wire” approach developed at NREL, or by reducing the thickness of the α- Si layers such as in the case of the “micromorph” cells. CdTe/CdS/SnO2 devices have shown high flexibility in terms of deposition techniques since they are deposited by many techniques that include spray pyrolysis, electrodeposition, vapor deposition, and close space sublimation and high performance capabilities, with power conversion efficiencies reaching 16.5 %.

Thanks in particular to post-deposition treatments that allow them to increase their grain size, passive grain boundaries and improve the electronic quality of the 156

CdTe. The interdiffusion of the CdTe and CdS layers seems to improve the junction quality but the reason for this has yet to be understood. Similarly the inclusion of p-type dopants such as Cu, Hg, Pb or Au is required to modify the CdTe contact surface, but have deleterious effect upon device durability that is only poorly understood. Finally, the sensitivity of these materials to moisture is a major limitation to cell stability that is currently overcome only through encapsulation.

CIGS/CdS solar cells have major strengths related to the ease with which large grains can be grown and passivated, the material’s tolerance to deviations from perfect stoichiometry and design flexibility imparted by the ability to vary the alloy composition. However, little fundamental understanding of these materials or devices is currently available, making progress in this technology largely empirical. Among the most relevant issues to be addressed include the optimization and simplification of the manufacturing processes, the moisture sensitivity, the development of lightweight flexible substrates, the replacement of CdS windows to improve environmental acceptance and in the long term, the limited supply of materials such as In, Ga and Se.

Organic PV devices OPVs are compatible with plastic substrates and can be fabricated using high throughput, low-temperature printing techniques compatible with roll-to-roll manufacturing: inkjet, screen, offset, or flexographic printing. Flexibility in the synthesis of the basic molecules allow for alteration of a wide range of properties including molecular weight, band-gap, molecular orbital energy levels, wetting properties, structural properties and doping. These characteristics have to be tuned in order to maximize the overall efficiency of organic photovoltaics and in particular:

157

The open-circuit voltage (VOC) is believed to mainly depend on the difference between the LUMO of the donor and the HOMO of the acceptor. However there is not sufficient fundamental understanding of the influence of other parameters such as the materials’ work function;

The short-circuit current (ISC) has to be increased through photon absorption enhancement with NIR-sensitive, low-band-gap organic semiconductors or active acceptor materials, and by increasing the charge mobility and thus the charge collection. Degradation pathways for organic photovoltaics seem to stem largely from changes in morphology, loss of interfacial adhesion and inter-diffusion of components as opposed to strictly chemical decomposition. Thus, careful design and material engineering can substantially improve device lifetimes. Little has been published so far about the stability of these devices under illumination and UV exposure or the stability of foil-based devices with plastic substrates.

Photoelectrochemical cells with liquid electrolytes usually suffer short lifetimes because of the lack of environmental stability of the solvents traditionally used as the electrolyte host. Ionic liquid electrolytes, having extremely low vapor pressure and stable electrochemical properties are stable electrolyte systems for photoelectrochemical cells and effective as a replacement for liquid electrolytes based on redox-couples. In DSSC, the interface between the nanostructured TiO2 and the electrolyte seems to be crucial for the stability of the device. The use of blocking layers of insulating metal oxides has been shown to prevent the adsorption of pollutants on the TiO2 surface and to reduce back-electron transfer. Also UV-cut-off filters may be used to prevent band-gap excitation resulting in TiO2 mediated oxidation processes.

158

Photon-to-Thermal-to-Electric Energy Conversion There are a number of options available at different stages of development. The most developed technologies are the parabolic dish, the parabolic trough, and the power tower. The parabolic dish is already commercially available. This system is modular and can be used in single dish applications with output power of the order of 25 kWe or grouped in dish farms to create large multi-megawatt plants. Parabolic troughs are already a proven technology and will most likely be used for deployment of solar energy in the near future. Various large plants are currently in operation like in California with 354 MW output or in the planning process in the USA and in Europe. Power towers with low cost and efficient thermal storage promise to offer dispatchable high capacity factor power plants in the future.

Parabolic troughs Parabolic trough systems use single-axis tracking parabolic mirrors to focus sunlight on thermally efficient receiver tubes that contain a heat transfer fluid (HTF). The receiver tubes are usually metallic and embedded into an evacuated glass tube that reduces heat losses. A special high-temperature coating reduces radiation heat losses and the working fluid (e.g. thermo-oil) is heated above 400 o

C and pumped through a series of heat exchangers to produce superheated steam

which powers a conventional turbine generator to produce electricity. It is also possible to produce superheated steam directly using solar collectors. This makes the thermo-oil unnecessary and also reduces costs because the relatively expensive thermo-oil and the heat exchangers are no longer needed.

However, direct steam generation (DSG) is still in the prototype stage and more research is required to solve the thermo-mechanical issues related to working pressures above 100 bar and the presence of a two-phase fluid in the receivers. The efficiency of a solar thermal power plant is the product of the collector efficiency, field efficiency and steam-cycle efficiency. The collector efficiency depends on the angle of incidence of the sunlight and the temperature in the 159

absorber tube and can reach values up to 75 %. Field losses are usually below 10 %. Altogether, solar thermal trough power plants can reach annual efficiencies of about 15 %; the steam-cycle efficiency of about 35 % and has the most significant influence.

Central receiver systems such as solar thermal tower plants can reach higher temperatures and therefore achieve higher efficiencies. Current research in parabolic trough systems aims at improving performance, lifetime, reducing manufacturing, operation and maintenance costs as well as improved designs. These activities concern all critical components of the system namely the support and tracking structure, the reflector which could be glass mirrors, polymeric reflectors and other alternative reflectors, and the receiver tubes (absorbers, glass/metal seals, etc).

A more fundamental research field concerns the development of new heat transfer fluids with good stability properties at higher temperatures and compatible with thermal storage systems discussed. In the case of DSG plants, a phase-change thermal storage may be better adapted than current thermal storage concepts. A technical assessment of parabolic troughs and power towers by NREL including forecasts for cost and performance can be found in reference.

Power towers In a power tower plant, hundreds of two-axis tracking heliostats are installed around a tower where they focus sunlight with concentrations ranging from 100 to 10,000 suns. The absorber is located on the top of the tower and can reach temperatures from 200 oC to 3000 oC. Hot air or molten salt are usually used to transport the heat from the absorber to a steam generator where superheated steam is produced to drive a turbine and an electrical generator. Power towers are suited for large-output applications, in the 30 to 400 MWe range and need to be large to be economical.

160

Thermal storage can be easily integrated with this type of solar systems allowing the enhancement of the annual capacity factor from 25 % to 65 % and the stabilization of the power output through fluctuations in solar intensity until the stored energy is depleted. Since early 1980s, power towers were built in Russia, Italy, Spain, Japan, France, and the USA, with power outputs ranging from 0.5 MWe to 10 MWe (like the Solar Two in Southern California) and using various combinations of heat transfer fluids (steam, air, liquid sodium, molten nitrate, molten nitrate salt) and storage media (water/steam, nitrate salt/water, sodium, oil/rock, ceramic).

The Solar Two plant has proven the feasibility of molten-salt power towers, achieving turbine operation at full capacity for three hours after sunset thanks to the two-tank molten salt storage system. The main design challenge for this system was identifying the materials that work with molten salt, since this fluid has relatively high freezing point (220 oC), low viscosity, wets metal surfaces extremely well, and is corrosive. Solar Tres is the first commercial power tower and is a follow-up of the technologies developed for Solar Two. Its size is three fold larger, the electrical power output is 15 MWe, and the thermal storage capacity 600 MWh.

The efficiency of a solar-powered steam turbine electric generator used in the power tower concept is a critical function of the temperature TR of the receiver, which is influenced not only by the incident energy but also of several factors including the heliostat optical performance, the mirror cleanliness, accuracy of the tracking system and wind effects. For an ambient temperature of 340 K, the efficiency is 35 % for TR = 800 K and 62 % for TR = 3000 K. Solar Two achieved 13 % peak efficiency with TR = 570 oC. The development of heat transfer fluids working at high temperature is a key issue for increasing the overall efficiency of power towers.

161

Alternative systems for converting thermal energy into electricity have been explored, such as liquid metal magnetohydrodynamic generators (LMMHD). These systems were first investigated in the early 1980s and offer significant increases in the system thermal efficiency over the 33 % considered attainable with conventional turbo-machinery with sodium at a temperature of 650 oC, the theoretical efficiency is 39.5 %. A peak efficiency of 46.5 % is predicted for lithium at 760 oC. The thermodynamic efficiency at maximum power with an ambient temperature of 300 K and a black-body source temperature of 6000 K is 64 % and occurs at a receiver optimal temperature of 2900 K. Other potential advantages are that the sodium/steam heat exchanger is eliminated in liquid metal systems and where LMMHD systems employ the same working fluid as the solar receiver, no re-circulating pump is required as pumping power is provided directly by the cycle.

The development of new heat transfer fluids (HTFs) is crucial for increasing the operating temperature of a solar thermal plant and hence the efficiency of the steam cycle. Stability at high temperature, low flammability, low vapor pressure at high temperature, low corrosivity in standard materials, low freezing point, high boiling point, and low cost are the main required characteristics. Various thermal storage options are currently being considered for parabolic troughs and power towers. Some of them have already been demonstrated but many need further research, particularly concerning the optimization of the HTF materials. Here are some among the most significant technologies:

Concrete – This system would use a HTF in the solar field and pass it through an array of pipes imbedded in concrete. The highest uncertainty is the long-term stability of the concrete material itself after thousands of charging cycles.

Indirect two-tank molten-salt – In current applications, a synthetic oil (e.g. biphenyl-diphenyl oxide) is used as HTF in the solar field and for 162

heating molten salt through a heat exchanger in the thermal storage system. Two separate tanks are used for this system. The excess heat of the solar collector field heats up the molten salt which is pumped from the cold to the hot tank. If the solar collector field cannot produce enough heat to drive the turbine, the molten salt is pumped back from the hot to the cold tank and heats up the heat transfer fluid. The molten salt as it was used in the Solar-Two solar tower pilot demonstration plant is a binary mixture of 60 % sodium nitrate (NaNO3) and 40% potassium nitrate (KNO3) salt. The feasibility of this system was proven and the concept seems to have low technological risk despite the relatively high freezing point (~225 oC) of the salt.

Thermocline storage – Thermoclines use a single storage tank. A lowcost filler material made of quartzite and silica sand acts as the primary thermal storage medium and replaces approximately two-thirds of the molten salt that would be needed in a two-tank system. With the hot and cold fluid in a single tank, the thermocline storage system relies on thermal buoyancy to maintain thermal stratification. The thermocline is the region of the tank between the two temperature resources with a temperature difference of about 60 oC. Thermoclines can maintain their integrity over a three-day no-operation period. Molten-salt HTF – Using a lower temperature molten salt as the HTF in the solar field allows the same fluid to be used in both the solar field and the thermal storage field leading to significant cost reduction for the thermal storage especially when used in the thermocline configuration. This also allows the solar field to be operated at higher outlet temperatures increasing the power cycle efficiency and further reducing the cost of thermal storage. Major technical barriers to this option include the challenges of high freezing temperature salts and

163

with higher operating temperatures leading to higher heat losses and requiring new materials and components.

Organic molten-salt HTF – Organic salts, or ionic liquids, have the advantage of being liquid at room temperature. Additionally they can be synthesized to have specific properties desirable for solar applications, namely low freezing point, high thermal stability, low corrosivity, good heat transfer and thermal properties and low cost. The development of organic salts is relatively new and more research is required to optimize all these characteristics particularly to lower the materials cost.

Dish-engine systems Dish-engine systems can be used to generate electricity in the kilowatts range. A parabolic concave mirror concentrates sunlight where the two-axis tracked mirror must follow the sun with a high degree of accuracy in order to achieve high efficiencies. At the focus is a receiver which is heated up over 700 °C. The absorbed heat drives a thermal engine which converts the heat into motive energy and drives a generator to produce electricity. If sufficient sunlight is not available, combustion heat from either fossil fuels or biofuels can also drive the engine and generate electricity.

The solar-to-electric conversion efficiency of dish–engine systems can be as high as 30 % with large potential for low cost deployment. For the moment, the electricity generation costs of these systems are much higher than those for trough or tower power plants and only series production can achieve further significant cost reductions for dish–engine systems. A number of prototype dish-engine systems are currently operating in Nevada, Arizona, Colorado, and Spain. High levels of performance have been established; durability remains to be proven although some systems have operated for more than 10,000 hours.

164

Most research and development efforts aim at incrementally enhancing the reliability, performance, and cost-effectiveness of all major components of these systems like the concentrators, receivers, and engines. The development of highefficiency low-cost thermal engines or alternative heat-to-electric conversion systems is the most critical requirement for these systems to become economical. Various thermodynamic cycles have been considered for dish-engine systems. Stirling and open Brayton (gas turbine) cycles have shown the best performances to date. Stirling engines have a potential for high efficiency and external heating makes them easily adaptable to solar dishes. Modern high-performance Stirling engines use hydrogen or helium working gas at temperatures over 700 oC and pressures as high as 20 MPa, resulting in thermal-to-electric conversion efficiencies of about 40 %.

The main disadvantage of these types of engines is their manufacturing cost mainly determined by the materials used for the hot part heat exchanger which usually requires stainless steel, high-temperature alloys or ceramic materials and by the design of the cooling system. In a dish-Brayton system, solar heat is used to replace or supplement the fuel at the entrance of the gas turbine. Current designs include pressure ratios of of about 2.5, turbine inlet temperatures of about 850 oC, and recuperation of waste heat with predicted efficiencies over 30 %. DishBrayton systems are still at an early stage of development.

Alternative dish systems replace the thermal engine with high-efficiency above 30% in multijunction photovoltaic cells working with concentrated sunlight or with thermoelectric or thermionic devices for direct current extraction from the high temperature receiver. Thermoelectrics can convert thermal energy into electrical energy through the Seeback effect. Lattice thermal conductivity which is the major contribution to thermal conductivity can be minimized by increasing the phonon scattering by introducing heavy atoms, disorder, large unit cells, clusters, and rattling atoms. Using these principles, a variety of high ZT materials have been developed with different operation temperature ranges. Skutterudite 165

material systems and complex chalcogenide structures are some representative examples of such materials.

Low-dimensional (nanostructured) thermoelectric materials have high potential for achieving figure of merit values of 2 and higher. In fact, quantum and size effects may enhance the Seeback coefficient and the electrical conductivity while boundary phonon scattering would lead to a large reduction in thermal conductivity. In contrast with thermoelectric devices, electron transport in thermionic converters is not diffusive but ballistic, the emitter and collector electrodes being physically separated to prevent thermal diffusion along a solid lattice.

Thermionics are an old technology that was widely developed from the 1950’s to the mid-years of 1970’s. The development efforts were virtually abandoned because of the lack of electrode materials with combined low working functions of about 1 eV and low vapor pressures for long-term stability; the low performance in their heat-to-electric conversion efficiency which was below 10% and the high operating temperatures of above 1400 K. Others were the technical challenges related to manufacturing the devices with an interelectrode spacing below 10 μm needed to decrease the negative-space-charge effect in vacuum thermionic converters.

High-pressure cesium diode thermionic converters with refractory-metal showed better performances with output-power densities ranging from 1W/cm2 to 30 W/cm2 and efficiencies from 5 % to 20 % with emitter temperatures TE between 1500 K and 2000 K where the theoretical efficiency limit for thermionic conversion was 30 %. The development of nano-technologies led to realization of devices with inter-electrode spacing as small as 1 - 10 nm, allowing tunneling of hot electrons and thus higher conversion efficiencies.

166

As in the case of thermoelectric devices, extensive research efforts are now focused on low temperature applications such as cooling at temperatures below room temperature or power generation from waste-heat at temperatures of the range of 400 – 500 K. For high temperature applications, effective high-energy electron selectivity, and decreasing the large (up to 50 %) power losses caused by the charge-space effect are among the main fundamental issues that are currently being addressed to enhance the efficiency of thermionics.

Solar chimneys In contrast with the previously described thermal technologies, solar chimneys convert into electricity not only direct normal irradiance but global irradiance. A solar chimney power plant has a high chimney up to 1000 m and is surrounded by a large collector roof up to 130 m in diameter that consists of glass or resistive plastic supported on a framework. Towards its centre, the roof curves upwards to join the chimney creating a funnel. The sun heats up the ground and the air underneath the collector roof and the heated air follow the roof until it reaches the chimney.

There, it flows at high speed (35 mph) through the chimney and drives multiple wind turbines at its bottom. The ground under the collector roof behaves as a storage medium and can even heat up the air for a significant time after sunset. The efficiency of the solar chimney power plant is below 2 % and depends mainly on the height of the tower. However, the whole power plant is not without other uses as the outer area under the collector roof can also be utilized as a greenhouse for agricultural purposes. As with trough and tower plants, the minimum economical size of solar chimney power plants is also in the multi-megawatt range.

167

Photon-to-Chemical Energy Conversion Photo-conversion processes are used for producing a large variety of chemicals with clear energetic and environmental advantages compared to conventional technical processes. Among the large variety of identified processes and technologies, we consider here three main categories of solar-to-chemical conversion processes:

photo-(electro)-chemical

processes,

thermochemical

processes, and photosynthetic processes in natural systems.

Photochemical and photo-electro-chemical systems use light-sensitive materials in aqueous suspension or in the form of bulk electrodes, respectively for absorbing photon energy and producing electrons with sufficient energy for splitting water. In thermochemical technologies, concentrated solar flux is used to produce the high-temperatures necessary to drive endothermic reactions such as syngas production from natural gas, water thermal decomposition, and water splitting through high-temperature chemical cycles. Some biological systems produce hydrogen in their metabolic activities. These systems are capable of absorbing light, separating charges, and acting as catalysts for water redox.

Photo-(electro)-chemical water splitting The efficiency limit of a photo-electrochemical cell used for splitting water is 41% for a photoanode band-gap that just matches the free energy of the water splitting reaction. The efficiency decreases to about 18 % for a band-gap of 2 eV and to 0.05 % for 3 eV in the case of the band-gap energies of TiO2 and SrTiO3, respectively. Accounting for the thermodynamic potential needed for water splitting, overvoltage losses, and the energy required for driving the reaction, 1.6 1.8 V have to be provided to

produce hydrogen from water. In

photoelectrochemistry, this voltage is provided by a semiconductor material with band-gap in the range of 1.6 eV < Eg < 2.2 eV immersed in aqueous solution.

168

Combinatorial techniques are often used to screen a large variety of photo-anode and photocathode materials whether they including binary, ternary, and quaternary compounds. These techniques also allow the study of doping effects and to test them against photon absorption efficiency, photocatalytic activity, and resistance to corrosion. Some significant examples are tungsten trioxide (WO3), ferric oxide (Fe2O3), silver-chloride (AgCl) and silver-sulfide (Ag2S) clusters (zeolite A), titanium dioxide (TiO2), and zinc oxide (ZnO). Light-absorption of photosensitive electrodes is often insufficient or inappropriate to match with the electric current needs of integrated counter-electrodes or photovoltaic bias in photo-electrochemical systems.

Therefore, light absorption and photocurrent have to be enhanced by means of sensitizers (e.g., ruthenium-based dye sensitizers on nanocrystalline TiO2 – DDSCs or Grätzel cells, dye-loaded zeolite antennae on various receptor semiconductor materials) or doping agents (platinum or ruthenium in WO3, various metal dopants in Fe2O3 or ZnO), respectively. Most stable semiconductors in aqueous solutions are oxides which usually have band-gap energies that are too large. Lowering the band-gap of stable oxide-based semiconductors is therefore an important research activity. Photo-electrode stability is also addressed by separating the generation of hydrogen and oxygen with a photocatalyst membrane by modifying the nature of the electrolyte or by covering materials with ideal solar absorption but poor stability with adequate protective coatings.

Tandem or multi-junction cells are particularly interesting in this sense since the absorption efficiency can be enhanced by complementary spectrum absorption, while a corrosion-resistant material can be used as the top-junction (e.g., WO3 on GaInP2). The energetics and kinetics of the redox reaction are determined by the chemistry of the semiconductor/electrolyte interface and the development of highactivity resistant catalysts is hence crucial for the success of this technology. Molecular catalysts are a promising lower-cost and higher turnover alternative to heterogeneous catalyst (e.g. platinum particles). One leading strategy is 169

mimicking natural hydrogenase enzymes with functional catalyst systems with low-valent redox-rich iron and cobalt platforms. Main barriers are poor stability, large over-potential, and slow-turnover.

Thermal and thermochemical processes Concentration of solar light in solar thermal plants reaches 5,000 suns and produces temperatures of several hundred degrees. Therefore, it can be used for some specific thermal or thermochemical processes where high-temperature is required. The spectrum of applications is wide and includes: Hybrid solar/fossil processes such as carbo-thermic reduction of metal oxides using coke or natural gas as chemical reducing agent. The deployment of such processes could substantially reduce greenhouse-gas emissions in industrial applications (up to 59% of CO2-equivalent emission reductions for zinc production); solar thermal decomposition, or “pyrolysis” of fossil fuels or biomass, solar steam reforming of natural gas (950 - 1200 K) with efficiency ranging from 65 – 75 %, or solar steam gasification of heavy hydrocarbons or coal.

Theoretical efficiencies are about 40 % at 2500 K and 2 mbar. H2/O2 separation is the main technical barrier; quenching is the simplest technology but introduces large irreversibility.

Hydrogen production through thermochemical cycles were

proposed to bypass the H2/O2 separation problem of direct water dissociation since hydrogen and oxygen are produced at different steps. Two-step systems with theoretical efficiencies above 30 % are based on metal oxide redox reactions: (1) solar reduction of the metal oxide at elevated temperatures either in thermal or carbon thermal, where C or CH4 are reducing agents; (2) exothermic oxidation of the metal below 800 K.

To time of writing this book, the highest potential cycles in this category are the ZnO/Zn cycle with theoretical efficiencies of about 50 % at 2300 K, and Fe3O4/FeO cycle with theoretical efficiencies of about 60 % at 1900 K had been 170

recorded. In both cases, separation of the product gases from the intermediate chemical agents is limiting the yield of the latter and the overall energy efficiency. Multi-step cycles are more complex to implement but allow the use of relatively moderate temperatures. Their overall energy efficiency is limited by irreversibilities associated with heat transfer and product separation. Some examples are the 3-step iodine-sulfur cycle at 1140 K, with an efficiency limit about 50 %, and the 4-step UT-3 cycle, based on the hydrolysis of CaBr2 at 1020 K and of FeBr2 at 870 K with a comparable efficiency limit. Although 2000 to 3000 cycles have already been identified, only very few have been investigated in the last 20 years. Many fundamental and technical problems have still to be solved, including the development of materials with excellent thermal and corrosion resistance; the increase of the yield of intermediate chemical agents, limited by the efficiency of gas separation technologies that also limits the overall energy efficiency; and the matching of the reaction rates of the individual steps of the cycles.

Hybrid Solar Lighting Artificial lighting accounts for the largest component of electricity use in commercial U.S. buildings and security systems. Hybrid solar lighting provides an exciting new means of reducing energy consumption while also delivering significant benefits associated with natural lighting in commercial buildings. The hybrid lighting technology was originally developed for fluorescent lighting applications but recently has been enhanced to work with incandescent accentlighting sources, such as the parabolic aluminized reflector (PAR) lamps commonly used in retail spaces.

Commercial building owners use the low-efficiency PAR lamps because of their desirable optical properties and positive impact on sales. Yet the use of this inefficient lighting results in some retailers’ spending 55 – 70 % of their energy budgets on lighting and lighting-related energy costs. Hybrid lighting has the 171

potential to significantly reduce energy consumption while also maintaining or exceeding lighting quality requirements. Implementation of the hybrid solar lighting technology across the United States would represent significant energy savings to the country and would provide building managers with a near-term, energy-efficient, higher quality, economically viable alternative to incandescent lamps. Artificial lighting accounts for almost a quarter of the energy consumed in commercial buildings and 10 – 20 % of energy consumed by industry.

Solar lighting can significantly reduce artificial lighting requirements and energy costs in many commercial and industrial buildings and in institutional facilities such as schools, libraries, and hospitals. Future R&D is aimed at enhancing the performance and reliability of the technology as well as extending the application of the system to work with newly emerging solid-state lighting sources. The hybrid solar lighting technology delivers the benefits of natural lighting with the advantages of an electric lighting system which include flexibility, convenience, reliability, and control and these overcomes the constraints that marginalized the use of day lighting in the 20th century.

Principles of Operation The hybrid solar lighting system uses a roof-mounted solar collector to concentrate visible sunlight into a bundle of plastic optical fibers. The optical fibers penetrate the roof and distribute the sunlight to multiple “hybrid” luminaires within the building. The hybrid luminaires blend the natural light with artificial light to maintain a constant level of room lighting. One collector powers about eight fluorescent hybrid light fixtures which can illuminate about 1000 square feet. When sunlight is plentiful, the fiber optics in the luminaires provides all or most of the light needed in an area. During times of little or no sunlight, a sensor controls the intensity of the artificial lamps to maintain a desired illumination level.

172

Unlike conventional electric lamps, the natural light produces little to no waste heat having an efficacy of 200 lumens/ Watt (l/W) and is cool to the touch. This is because the system’s solar collector removes the infrared (IR) light from the sunlight: the part of the spectrum that generates much of the heat in conventional bulbs. Because the optical fibers lose light as their length increases, it makes sense right now to use hybrid solar lighting in top-story or single-story spaces. The current optimal optical fiber length is 50 feet or less. The hybrid solar lighting technology can separate and use different portions of sunlight for various applications. Thus, visible light can be used directly for lighting applications while IR light can be used to produce electricity or generate heat for hot water or space heating. The optimal use of these wavelengths is the focus of continued studies and development efforts.

Advantages of Hybrid Solar Lighting Electric lighting is the greatest consumer of electricity in commercial buildings and the generation of this electricity by conventional power plants is the building sector’s most significant cause of air pollution. Hybrid lighting can help conserve electricity in proportion to the amount of sunlight available. Hybrid solar lighting technology could benefit federal buildings particularly in the Sunbelt where cooling is a significant source of energy use. Full-spectrum solar energy systems provide a new and realistic opportunity for wide-ranging energy, environmental and economic benefits. Because hybrid solar lighting has no infrared component, it can be considered as a high-efficiency light source. Other advantages include; Roof penetrations are small and minimal, reducing the potential for leaks. IR and UV energy in sunlight is separated from the visible light, rather than being transmitted

into

buildings.

Heating,

ventilation,

and

air-conditioning

(HVAC) loads are thus reduced by 5 to 10%, compared to buildings having conventional electric lighting systems. Hybrid solar lighting systems are readily adaptable to commercial buildings with multiple floors, relatively low ceiling 173

heights, and interior walls, though currently fiber optic output is optimized on the top two floors. A single system can distribute enough sunlight to co-illuminate several rooms in an office building.

Large portions of valuable plenum space; the area between the roof and drop ceiling; are not needed, so there is little competition with other building services such as HVAC ducts, sprinkler systems, and electrical conduits. Hybrid solar lighting can be used both for direct ambient lighting and for indirect lighting, task lighting, and accent lighting. In retrofit applications, hybrid solar lighting is easily incorporated into existing building designs, and the optical fibers can be rerouted to different locations as lighting needs change. By intentionally misaligning the solar collector from the sun, occupants can even dim or curtail distributed sunlight.

Cost Considerations The concept of hybrid lighting has existed since the early 1970s but it has been difficult to make the technology practical. When that target price is reached, a building owner in Hawaii could pay for implementing the new technology in just 2 – 3 years with the savings on electricity bills alone. In other parts of the country where sunlight is less abundant and utility costs are lower, this payback would obviously take longer. The payback period for hybrid solar lighting lengthens in proportion to the efficiency of the electric lamps used in combination with distributed sunlight. Because linear fluorescent lamps are very efficient (65 - 90 lm/W), the models indicate that a hybrid configuration used with such lamps will require more than 10 years to pay for itself in most regions of the country during the early years of commercialization. As prices fall, hybrid solar lighting has the potential to become cost-competitive in most in-door lighting scenarios. A hybrid configuration is likely to extend the typical life of incandescent and/or halogen lamps. When incandescent lamps are dimmed, filament temperatures decrease and 174

as filament temperatures decrease, life expectancy increases. Although the lamps will last longer, a penalty in efficiency occurs because cooler filaments are generally less efficient at radiating visible light.

In contrast to roof penetrations for skylights, penetrations for hybrid solar lighting are few and small reducing the potential for leaks. As R&D improves system performance, increases system lifetime and reduces system price and as also the secondary benefits of the technology are demonstrated), hybrid solar lighting will move into the larger market of existing buildings that use all fluorescent lighting.

Potential Candidates for Hybrid Solar Lighting The first commercial use for hybrid solar lighting will probably be; On the upper two floors of buildings having the following characteristics; Sunbelt location in areas where daytime electricity prices are highest; Occupied every day including weekends and Lighting of high quality (or color rendering) and replacing lessefficient electric lamps that are currently in use. Hybrid solar lighting is being targeted first for commercial buildings because in these buildings lighting can account for the largest part of the electricity bill. Residential uses may be farther down the road because the cost advantages there are not as great. However, federal sites would benefit from using hybrid solar lighting technology because decreased use of conventional electric lamps and the air-conditioning loads associated with them would result in energy savings. Additional benefits would include improved lighting quality and positive environmental impacts from reducing the need to generate electricity. New construction projects create outstanding opportunities for holistic building designs that utilize highly efficient systems that can create satisfying working and living environments with minimized operating costs. A successful balancing these factors against cost and scheduling issues is the key to creation of a well designed efficiently operating building. 175

New Developments in Hybrid Solar Lighting Technology In solar technology, the controller compares the actual direction it is pointing to the actual computed position of the sun and then determines if the collector needs to be moved to match its position with that of the sun (tracking system). The motors then move at a speed proportional to the difference in the actual and computed positions. This process is performed continuously throughout the day in order to track the sun accurately. The control board operates on a 12 V or 24 V dc supply and uses less than 2 Watts or a photovoltaic solar cell can also be used to power the board. The development of a re-designed and less expensive tracking mechanism for the solar collector was also begun in 2005. Among suggested design modifications from an initial “manufacturability” analysis was the use of a high-precision linear actuator in combination with a gear-train drive to reduce cost while still providing high tracking accuracy. Aging of the polymers used in optical fibers to distribute concentrated sunlight is still a concern as is the need for more efficient methods of coupling converging sunlight into fiber optic bundles. New luminaires that provide seamless spatial and chromatic uniformity during transitions between natural and electric illuminants must be developed for several lighting applications. Ultimately, the advent of hybrid solar and solid-state lighting systems that use light-emitting diodes (LEDs) capable of chromatically adapting to match the spectrum of sunlight throughout the day is expected. Prototypes are already proving that the hybrid solar lighting concept is viable both technically and economically. The latest prototype provides lighting practitioners with unprecedented design flexibility and control over where and how sunlight is used inside buildings. This prototype uses a bundle of 127 small optical fibers, each of which can distribute 350 lumens to several different hybrid luminaires on a sunny day making possible numerous day lighting applications. For example, some hybrid luminaires being developed allow hybrid solar lighting to be used with linear and compact florescent lamps and with incandescent/halogen lamps.

176

Hybrid Lighting Partnership A broad-based public–private alliance is working to commercialize hybrid solar lighting. The Hybrid Lighting Partnership includes the organizations in most countries. The partnership will improve quality of life by providing more efficient and affordable solar energy, cleaner air, lower utility bills during peak demand periods and a healthier more productive work environment. The partnership’s mission is to develop and deploy hybrid solar lighting worldwide early in the 21 st century. By 2020 in the United States, hybrid solar lighting is expected to provide; annual energy savings of more than 30 billion KWh; reductions in carbon emissions exceeding 5 megatonnes of carbon per year, and total economic benefits exceeding $5 billion.

Future of Hybrid Solar Lighting Electricity use for artificial lighting in commercial building space costs building owners nearly $17 billion a year as reported by the personal communication, Energy Information Administration sector. Despite the high energy consumption and cost of electric lighting, natural lighting from conventional options such as skylights and windows illuminates only a tiny fraction of existing commercial buildings. This limited use of natural lighting is a result of the architectural limitations of skylights and windows and the uncontrollable nature of the sunlight itself. The future is bright for hybrid solar lighting. The nationwide field trial program will provide system performance data and user feedback essential for successful use of this solar energy technology. During the field trial program R&D will continue at ORNL in collaboration with industry and university partners to lower component costs, improve the longevity of optic fibers, and advance system control. New solid-state (LED) hybrid luminaires are also being researched for increased energy efficiency. Exciting new areas of R&D for ORNL and its

177

industry and university partners include utilizing hybrid solar lighting technology for space heating, water heating, and hydrogen production.

178

CHAPTER SEVEN

OPERATION OF SOLAR ENERGY Developing countries can cover all their demands for energy by solar systems with only 0.1% of the surface of their land area. World population is expected to double by the middle of the 21st century according to the global energy report of 1998. This will consequently result in a 3 to 5 fold increase in world economic output by the year 2050 and a further 10 to15 fold increase by the year 2100. As a consequence, primary energy requirements are also expected to increase by approximately three folds by the year 2050 and five folds by the year 2100 respectively. This is expected to exert tremendous pressure on primary energy supplier all over the world. Energy has an established positive correlation with any economic growth or decline. Providing adequate, affordable and clean energy is a prerequisite for eradicating poverty and improving productivity especially in the developing countries.

There has been an inevitable increase in the use of fossil fuels. Alongside this usage, a country’s economic growth presents associated side effects of threat to the nation’s energy security as well as environmental degradation through climate change due to pollution. A feasible alternative to these indiscriminate burning of fossil fuels lies in the accelerated use of renewable energy or solar energy. In tropical countries like Kenya, Uganda and Ghana in Africa, which have sunshine almost throughout the year in most parts, solar energy is one of the most viable options. Energy from the sun has been used to provide electricity for many years all over the world. This form of renewable energy occupies less space compared to the space occupied by hydropower projects.

Principle of Operation of Solar Energy The amount of solar energy incident on the earth’s surface is approximately1.5 x 1018 kWh/year which is about 103 times the current annual energy consumption of

179

the entire world. The density of power radiated from the sun (referred to as solar energy constant) is 1.373 kW/m2. Solar energy is available in abundance in most parts of the world.

A Solar cell is a device which converts photons present in Solar rays to directcurrent (DC) and voltage (V). This associated technology is called Solar/ Photovoltaic (SPV) technology. A typical silicon PV cell is a thin wafer consisting of a very thin layer of phosphorous-doped (N-type) silicon on top of a thicker layer of boron-doped (P-type) silicon. In this set up, an electrical field is created near the top surface of the cell where these two materials are in contact and form a P-N junction. When sunlight hits the semiconductor surface, an electron springs up and is attracted towards the N-type semiconductor material. This causes more negative (electrons) in the n-type and more positive (holes) in the P-type semiconductors, generating a higher flow of electricity. This is known as Photovoltaic effect.

The amount of current generated by a PV cell depends on its efficiency, its size or surface area and the intensity of sunlight striking these surfaces. For example, under peak sunlight conditions a typical commercial PV cell with a surface area of about 25 square inches will produce about 2 watts peak power.

Working mechanism of a silicon solar cell 180

Principles of Solar Energy Solar Irradiance The Sun is the fundamental driving force for energy behind the Earth's climate system and global energy content. It is of crucial to understand fully the conditions of how sunlight arrives at the top of the atmosphere and its transformation through into the earth’s surface. The amount of solar power available per unit area of surface is known as irradiance. Irradiance is a radiometric term used to refer to the power of electromagnetic radiation at a surface per unit area.

It is used when the electromagnetic radiation is incident on a surface. Irradiance fluctuates according to the weather conditions and the sun’s location in the sky at that particular time. This location constantly changes through the day due to changes in both the sun’s altitude or elevation angle and its azimuth or compass angle. Figure 7.2 below shows the two angles used to specify the sun’s location in the sky.

The sun’s elevation angle and the sun’s compass angle

Solar Constant The solar constant is the amount of incoming solar electromagnetic radiation per unit area. It is measured on the outer surface of Earth's atmosphere on a plane perpendicular to the rays. The solar constant includes all types of solar radiation 181

and not just the visible light. It is estimated to be roughly 1,366 watts per square meter (W/m²) according to satellite measurements although this fluctuates by about 6.9 % during a year (from 1,412 W/m² in early January to 1,321 W/m² in early July) due to Earth's varying distance from the Sun. For the entire planet Earth with a cross section area of 127,400,000 km², the power is (1366 W/m2 x 1.274×1014 m2) 1.740×1017 W, plus or minus 3.5 % error. The solar constant does not remain constant over long periods of time. The average value cited, i.e 1,366 W/m², is equivalent to 1.96 calories per minute per square centimeter, or 1.96 langleys (Ly) per minute.

Solar Window The solar window represents the effective area through which useful levels of sunlight pass throughout the year for a specific location. It is used to determine potential shading problems when designing a photovoltaic system as illustrated in figure below.

Design of a solar window system

Solar Spectrum A solar spectrum is simply a spectrum of electromagnetic waves. The sun radiates power over a continuous band or spectrum of electromagnetic wavelengths. The power levels of these various wavelengths in the solar spectrum is composed of 7% ultraviolet radiation (UV), 47% visible radiation (VIS) and 46% infrared (heat) radiations (IR). 182

Ultraviolet, Visible and Infrared Radiations The sun’s total energy is composed of 7 % ultraviolet radiation, 47 % visible radiation and 46 % infrared (heat) radiation as stated above. Ultraviolet (UV) radiation causes many materials to degrade and is significantly and naturally filtered out by the Ozone layer in the upper part of the atmosphere. Photovoltaic cells mostly and primarily use visible radiation. The distribution of colours within light is important because a photovoltaic cell will produce different amounts of current or energy depending on the various colours it reflects on its surface and those that it absorbs. Infrared radiation contributes to the production of electricity from crystalline silicon and some other materials while other materials contribute current from visible light. In most cases, however, infrared radiation is not as important as the visible portion of the solar spectrum.

Solar radiation spectrum

Solar Insolation Solar insolation is determined by summing solar irradiance over time and is usually expressed in units of kWh/m2 /day.The results of the earth’s motion and atmospheric effects at various locations have leads to essentially two types of solar insolation data. These are daily and hourly. Solar irradiance is related to power per unit area where as solar insolation is related to radiant energy per unit area. Thus; 183

Solar irradiance = power/area = P/A and, Solar insolation = radiant/area = R/A

Average Daily Solar Radiation To provide long-term average daily solar radiation data, an average of daily solar radiation is calculated for each month over a period of typically 30 years or more. This data is therefore enough and useful both in predicting long-term performance solar energy and also in analyzing the economics of solar energy systems to be installed. However, the actual average daily solar radiation for a given month may vary significantly from the long-term average for that month.

Peak Sun Hours The number of peak sun hours per day at a given location is the equivalent number of hours at peak sun conditions of at an average of 1 kW/m2 that produces the same total insolation as actual sun conditions. Figure 7.5 below shows how Peak Sun Hours is determined by constructing a graph having the same area as that for the actual irradiance versus time of irradiance.

Peak Sun Hourly Solar Radiation data

184

Typical Meteorological Year (TMY) data Typical Meteorological Year data is the result of statistical analysis of Solar and Meteorological (SOLMET) rehabilitated weather data for past years. The TMY consists of a selection of each of the twelve months so that it best represents the average of that particular month over past years. The TMY is therefore a composite year with representative months selected from different years from the past data available. For Example, consider a representative month as January. Every January month of past years is compared with the average data of all the past years (January month) and the one closest to the average is considered. The selection is weighted 50 % on solar radiation and 25 % each on ambient temperature and wind speed. TMY data is useful for photovoltaic system, design and analysis.

Direct and Diffuse Solar Radiation The component of the radiation coming from all directions in the sky is diffused. When the sun is directly overhead, it has a diffuse component of about 10 % when skies are clear. It is common to consider separately the ‘direct’ (or beam) radiation coming from solar disk and the ‘diffuse’ radiation from elsewhere in the sky with their sum known as ‘global’ radiation. Percentage increases with increase in Air Mass are common withnessed. Sunlight coming from the sun is reduced by about 30 % before it reaches the earth surface due to the following; Scattering by atmospheric particles; Scattering by aerosol, dust particles etc.; bsorption by atmospheric gases

Advantages and Limitations of Solar Energy Renewable energy sources in general and solar energy source in particular has the potential to provide energy services with zero or almost zero dangerous emissions. The solar energy is abundant and no other source in renewable energy is like solar energy. Every technology has its own advantages and disadvantages but solar energy has very many advantages and little disadvantages. As the solar 185

insolation and atmospheric conditions vary significantly from place to place, efficiency of solar energy also differs accordingly.

Advantages of Solar Energy a) It is an abundant Renewable Energy b) This technology is Omnipresent and it can be captured for conversion on a daily basis. c) It is a Non-polluting technology, which means that it does not release green house gases. d) It is a Noiseless technology as there are no moving parts involved in energy generation. e) This technology requires Low-maintenance because of lack of moving parts. f) It can be installed on modular basis and expanded over a period of time. g) Most viable alternative for providing electricity in remote rural areas as it can be installed where the energy demand is high and can be expanded on modular basis.

Limitations of Solar Energy a) As the technology is in an evolving stage, the efficiency levels of conversion from light to electricity is in the range of 10 to 17%, depending on the technology used. b) The initial investment cost of this technology is high. At present the technology is basically surviving because of subsidy schemes available by the government. c) Solar energy is available only during daytime. Most load profiles indicate peak load in the evening/night time. This necessitates expensive storage devices like battery, which need to be replaced every 3 to 5 years. Generally, the cost of the Battery is 30 to 40% of the system cost. 186

d) As the efficiency levels are low, the space required is relatively high. For instance, with the existing levels of technologies, the land required for putting up a 1 MW solar PV power plant is between 6 to 9 acres. However, research is going on to increase the efficiency levels of the cell. e) Solar energy is heavily dependent on atmospheric conditions. f) Solar insolation varies from location to location, so there are certain geographic limitations in generating solar power. g) With

the

existing

module

and

inverter

manufacturing

technologies, it may not be worthwhile in terms of costs to deploy solar energy for certain loads which require very high starting power (e.g. air conditioners).

Solar Receiver Technologies The types of receiver technology used for collecting solar energy are classified as follows:

Flat Plate Arrays For most applications, flat plate arrays are in fixed orientation. They can operate in either fixed orientation or in a sun-tracking mode. Flat plate arrays use both diffused and direct sunlight. However, with the advent of low-cost passive suntrackers, flat plate tracking arrays are becoming more popular. Figure 7.6 below depicts flat plate collector where one is mounted with a solar tracker.

187

A flat plate collector and pole mounted solar tracking array respectively

Tracking Arrays Tracking arrays aid the solar array follow the path of the sun and maximizes the solar radiation incident on the photovoltaic surface at all times during the radiation process. The two most common tracking orientations are; One-axis tracking: In this tracking mechanism, the array tracks the sun east to west. It is used mostly with flat-plate systems and occasionally with concentrator systems. Two-axis tracking: In this tracking mechanism, the array points directly at the sun at all time. It is used primarily with PV concentrator systems. A compromise between fixed and tracking arrays is the adjustable tilt array, where the array tilt angle is adjusted periodically (usually seasonally) to increase its output. This is mostly done manually.

Concentrator Arrays Concentrator arrays must track the sun because they rely on the ability to focus direct sunlight on the cells of the solar panel. Concentrators are best used in the areas with high direct beam radiations. Concentrator arrays use optical lenses and mirrors to focus sunlight onto high-efficiency cells. Figure 7.7 below shows three forms of concentrator devices. The major advantage of concentrating device is that they use relatively small areas of expensive photovoltaic material. The larger aperture areas are made up of less expensive plastic lenses or other materials. 188

Concentrator arrays tracking system

Solar Photovoltaic Technologies The heart of the solar energy generation system is the Solar cell. It consists of three major important elements namely; A semiconductor material which absorbs light and converts it into electron-hole pairs, The junction formed within the semiconductor which separates the photo-generated carriers (electrons and holes), The contacts on the front and back of the cell that allow the current to flow to the external circuit. Two main streams of technologies have been evolved for the manufacture of Solar Cells/Modules over the past few years namely;

Flat plate Technology: The Flat Plate Technology is further classified in two ways namely Crystalline Technology and Thin Film Technology. Concentrated Technology: The Concentrated Photovoltaic Technology has been classified according to the type of cell and the Optical system.

Crystalline Technology Crystalline Silicon (c-Si) was chosen as the first choice for solar cells. This material formed the foundation for all advances in semiconductor technology in most developed counrties. This semiconductor technology led to development of stable solar cells with efficiency up to 20% where two types of crystalline silicon

189

are used in the solar industry and they are the Monocrystalline Silicon and the Multicrystalline Silicon types.

Mono-Crystalline Silicon Mono-Crystalline Silicon cells are produced by growing high purity single crystal Si rods and slicing them into thin wafers. Single crystal wafer cells are very expensive and they are cut from cylindrical ingots and do not completely cover a square solar module. This results in substantial waste of refined silicon. Their efficiency remains between low at about 17 - 18 % because of the level of purity yet to be attained at a very high percentage.

Multi-Crystalline Silicon Poly-crystalline silicon cells are made from sawing a cast block of silicon first into bars and then form wafers. This technology is also known as Multi crystalline technology. However, Poly-Si cells are less expensive to produce than single crystal silicon cells as the energy intensive process for purification of silicon is not required in poly-Si cells. They are less efficient than single crystalline cells. The efficiency of poly crystalline silicon cells ranges from13 – 14 %.

Thin Film Technology In Thin Film Solar technology, a very thin layer of a chosen semiconductor material that has a range of some nanometer level to several micrometers in thickness is deposited onto either coated glass or stainless steel or a polymer substrate to produce a thin film coat. Various thin-film technologies are being developed to reduce the amount of light-absorbing materials required to construct the solar cell. This has resulted in reduction of processing costs and energy content. However, conversion efficiencies are also lower in these cases by an average of 7 – 10 %.

Their modules produced are of lesser efficiency at the same level of energy requirement and as a consequence, longer collector areas are required and 190

consequently too more requirement of land to install them. This technology is therefore apt where non productive land is available for example deserts of Rajasthan or Sahara desert. They have become popular compared to wafer silicon due to their lower costs, flexibility, lighter weights and ease of integration. Figure 7.8 below depicts the Thin Film Cell modules.

Thin film solar cell module

Amorphous Silicon Thin Film Technology Silicon thin-film cells are mainly deposited by chemical vapor deposition which is typically plasma-enhanced PE-CVD process, from silane gas and hydrogen gas. Depending on the deposition parameters, this can yield either, (i) Amorphous silicon (a-Si or a-Si:H); (ii) Protocrystalline silicon, (iii) Nanocrystalline silicon (nc-Si/ nc-Si:H) also called microcrystalline silicon. It has been found that protocrystalline silicon with a low volume fraction of nanocrystalline silicon is optimal for high open circuit voltage. The solar cells made from these materials tend to have lower energy conversion efficiency than bulk silicon but are also less expensive to produce. The quantum efficiency of thin-film solar cells is also lower due to reduced number of collected charge carriers per incident photon.

Cadmium Telluride Thin Film Technology A Cadmium Telluride (CdTe) solar cell is a solar cell based on cadmium telluride, an efficient light absorbing material (P-type) for thin-film cells. Compared to other thin-film materials, CdTe is easier to deposit and more suitable for largescale production. CdTe technology has been significantly been refined over the 191

past few years. It is uniquely capable of producing high-volume, low-cost modules, making widespread and affordable solar electricity a reality. The physical characteristics of CdTe are such that it is almost perfectly matched to the solar spectrum. This allows CdTe modules to absorb more of the available solar energy in low and diffuse light situations such as dawn and dusk or under cloudy skies and convert it into electricity more efficiently than conventional cells. As a result, CdTe thin film modules will generally produce more electricity under real world conditions than conventional solar modules with similar power ratings.

Concentrated Photovoltaic Technology In Concentrated Photovoltaic (CPV) systems, solar energy collected over large area is focused on each cell having smaller area to achieve higher power output and improved conversion efficiency. Thus the expensive semiconductor material required for power generation is reduced giving a substantial cost advantage. Although Si based SPV technology is fairly mature, CPV technology is still evolving and has a huge potential. The pPrimary reason for using CPV is that, the same amount of semiconductor material used can produce higher amount of energy than conventional materials thus reducing the cost of power generation significantly. In CPV systems, optical materials like mirror or lenses are used to collect sunlight on large area and focused it onto each cell having a smaller area. Despite these advantages of CPV technologies, their application has been limited because of the costs of focusing, sun tracking and cooling arrangements. Figure below depicts a Concentrated PV Module.

Concentrated PV Module

192

Solar Thermal Technologies Concentrating Solar Thermal technologies (CSP) is used to produce heat or electricity where their systems use lenses or mirrors and tracking systems to focus a large area of sunlight into a small beam. The concentrated light is then used as heat or as a heat source for a conventional power plant. A concentrating collector system can have a stationary collector or tracking one to track the sun. In stationary systems, the reflector and absorber are in fixed positions usually oriented directly to true south. Tracking devices shift the position of the reflector and the receiver to maximize the amount of sunlight concentrated on to the receiver.

Tracking collectors are either single-axis or double-axis. Single-axis tracking devices move the collector on one axis East to West or North to South while Dualaxis tracking devices track the sun on all axes. The entire collector, containing the reflector and receiver, generally moves as a unit in both types. Systems with dualaxis tracking facility concentrate most of the solar energy and generate very high temperatures.

Each concentration method is capable of producing high temperature and with correspondingly high thermodynamic efficiencies but they vary in the way in which they track the sun and focuses light. Due to new innovations in technology, concentrating solar thermal is becoming more and more cost-effective. These are the most complex in structure and so expensive. Wide ranges of concentrating technologies are prevalent and a list of few common technologies is given below and they are; Parabolic Trough, Dish Stirling, Concentrating Linear Fresnel Reflector and Solar chimney.

Parabolic Trough A parabolic trough consists of a linear parabolic reflector that concentrates light onto a receiver positioned along the reflector’s focal line. The receiver is a tube positioned right above the middle of the parabolic mirror and is filled with a 193

working fluid. The reflector follows the sun during the daylight hours by tracking along a single axis. The working fluid is heated to 150-350°C as it flows through the receiver and is then used as a heat source for a power generation system. Trough parabolic systems are mostly developed through CSP technology. Figure below depicts a parabolic trough system.

A parabolic trough system

Concentrating Linear Fresnel Reflectors Concentrating Linear Fresnel Reflectors are CSP plants which use many thin mirror strips instead of parabolic mirrors to concentrate sunlight into two tubes with working fluid as in parabolic trough. The advantage with Concentrating Linear Fresnel Reflectors is that flat mirrors are cheaper than parabolic mirrors and utilize space better. Concentrating Linear Fresnel Reflectors can be deployed in large plants or smaller ones. Figure below depicts the Concentrating Linear Fresnel Reflector System.

Concentrating Linear Fresnel Reflectors 194

Dish Stirling A Dish Stirling which is also called a dish engine system consists of a stand-alone parabolic reflector that concentrates light onto a receiver positioned at the reflector’s focal point. In this engine system, the reflector tracks the sun along two axes while the working fluid in the receiver is heated to 250 - 700 °C and then used by a Stirling engine to generate power. Parabolic dish systems provide the highest solar-to-electric efficiency which is currently about 25 %. It is also advantageous that their modular nature supports scalability. Figure below depicts Dish Stirling System and some of their major installations are the Stirling Energy Systems (SES) and Science Applications International Corporation (SAIC) dishes at UNLV and the Big Dish in Canberra, Australia.

Solar Chimney A solar chimney consists of a transparent large room which is sloped gently up to a central hollow tower or chimney in which the sun heats the air in this greenhouse-type structure which then rises up the chimney thereby driving an air turbine as it rises. This air turbine then creates electricity. Solar chimneys are very simple in design and could therefore be a viable option for projects in the developing countries in the world. Figure below is an example of Solar Chimney.

Solar Chimney

195

Solar Power Tower A solar power tower consists of an array of dual-axis tracking reflectors or heliostats that concentrate light on a central receiver atop a tower and the receiver contains a fluid deposit which could contain seawater. The working fluid in the receiver is heated to 500 – 1000 °C and then used as a heat source for a power generation or energy storage system. Power tower development is less advanced than trough systems but they promise higher efficiency and better energy storage capability. The two solar power towers in Daggett, California and the Planta Solar 10 (PS10) in Sanlucar la Mayor in Spain are working examples of this technology. Concentrating Solar Thermal Power (CSP) is the main technology proposed for a cooperation to produce electricity and desalinated water in the arid regions of North Africa and Southern Europe by the Trans-Mediterranean Renewable Energy Cooperation DESERTEC. Developing countries can also adopt this technology to manufacture clean energy. The figure below is a representation of Solar Power Tower.

Solar Power Tower

196

Applications of Solar Energy Three most important and widely used applications of Solar PV have been considered in this section. These are the Solar home lighting systems, the Solar water pumping systems and the Solar power plants as photovoltaic applications and Water heating system as a thermal application.

Solar home lighting system Home lighting systems are powered by solar energy using solar modules and the generated electricity is stored in batteries and used for the purposes of lighting or domestic uses whenever required. These systems are most widely used in nonelectrified rural areas and also as a reliable emergency lighting system for important domestic, commercial and industrial applications. The Solar Home Lighting system is a fixed installation designed for domestic application in which the system comprises of Solar PV Module or Solar Cells, a charge controller, a battery and a lighting system comprising of lamps, bulbs and fans. The solar module is installed in the open on roof/terrace - exposed to sunlight and the charge controller and battery are kept inside a protected place in the house. The solar module requires periodic dusting for effective performance. The schematic of the Home lighting system is shown in figure below.

Schematic of Solar Home Lighting System

197

Solar water pumping system Water pumping systems are powered by solar energy either as a stand-alone system. The power generated by solar module is used for operating DC surface centrifugal mono-block pump set for lifting water from bore or open well or water reservoir for minor irrigation and drinking water purpose. The system requires a shadow-free area for installation of the Solar Panel. Typical specification of solar water pumping system has a SPV water pumping system with a photovoltaic array of a capacity in the range of 200 to 3000 watts. A SPV water pumping system is expected to deliver a minimum of 65,000 liters per day with a 900 watts panel and 135,000 liters per day when with a 1800 watts panel from a depth of 7 meters on a clear sunny day. In case of deep well submersible pumps, the water output will be about 45000 liters from a 1200 watts panel.

These are only ideal cases which might not be achieved in practice. The discharge from the pump would vary with the intensity of the sunrays from morning till evening and expected to be maxima at around noon-time. The water output from the pumping would considerably drop with the increase in the depth from which water needs to be pumped. The SPV water pumping system can be used to irrigate 0.5 - 6 hectares if the water is to be pumped from a depth of 10 meters.

Schematic of Solar Water pumping System

198

Solar Power Plant Stand-alone solar power plant for the power generation comprises of Solar PV module array, Module mounting structures, Charge controller, Battery bank, Inverter and Load circuitry. Power supply in most of the cities and towns is unreliable which has forced some people to use small generators and accumulators. These generators are operated with fossil fuels like kerosene, petrol and diesel which in turn cause air, water and soil pollution. It also leads to an increase in dependence on oil imports but a solar power plant is a good cheaper and clean option for electrification in areas that are located away from the grid line or where other sources are neither available nor can be harnessed in a technoeconomically viable manner.

A solar power plant of the size 10 – 100 kW (kilowatt) depending on the load demand is preferable particularly with a liberal subsidy and low-interest soft loan from financial institutions. The idea is to raise the quality of life of the people subjected to poverty in such areas. This coupled with low-gestation remote areas of many states that need electrification. A typical stand-alone Solar PV power plant is shown below with a control panel (inset of photograph) with all the peripheral components housed. A typical solar power plant for village electrification should be clean, silent and an eco-friendly source of power in the range of 1 kWp to 10 kWp capacity, Module Rating of 75 Wp or more with a potential of 24 V and a battery of 300AH, or 48 V 150 AH and a low maintenance lead acid tubular plate.

199

A typical stand-alone Solar PV power plant

Solar Water Heating Systems Solar water heating system generally comprise of solar thermal collectors, a fluid system to absorb the heat from the collector toughened glass shield, insulated storage tank, cold water supply tank and insulated piping. In a Solar water heating system water is heated by the use of solar energy. These systems use solar energy to heat either water or a heat-transfer fluid, such as a water-glycol antifreeze mixture in collectors generally mounted on a roof.

The sun rays penetrate through the glass and fall on the absorber. The heat of the sunrays is absorbed by the cold water inside the absorber thereby increasing its temperature. The storage is either through the thermosyphon or the forced flow system. In the Thermosyphon system, up to 3000 liters per day can be installed however, for higher capacities it is necessary to use forced flow system. The water temperature can be raised up to 850C. The solar water heating system can be used for bathing, washing, boiler feed water pre-heating and other similar purposes. The investment made can be recovered in 4 to 6 years time and the life of the system is around 10 - 15 years if maintained properly since operation and maintenance cost are negligible. A typical schematic diagram of solar water heating system is shown below.

200

Schematic of Solar Water Heating system

The following are some of the factors about Solar Thermal system in South Africa; Hot water availability is in 60 to 120 degree temperature range; 100 liters per day system cost is Rs. 22,000 /-; Saves about Rs. 2200/- worth units of electricity annually; Payback period range from 3 to 5 years and emission of 1.5 tons of CO2 annually.

201

Prevents

CHAPTER EIGHT

SOLAR CELL ASSESSMENT The availability of solar radiation directly determines the revenue of solar energy power plants all over the wolrd. Knowledge of this resource is therefore crucial to determine the economical viability of any country. Solar radiation is the main fuel resource for solar energy systems. Direct normal irradiance is the amount of solar radiation received directly from the sun ignoring radiation from the rest of the sky falling onto a plane perpendicular to the direction of the sun. It can be converted and used for electricity generation via concentrating solar thermal power plants or concentrated PV.

Direct irradiance has the advantage that it can be concentrated directly with mirrors to reach high temperature or high radiative flux. However, the disadvantage is that it is only available in cloud-free situations or areas. Therefore, energy systems that use direct irradiance are only possible in sunny regions where cloud-free conditions are prevalent. These regions are mostly along the tropical countries.

Factors considered in Solar System Designs Solar Radiation Solar Energy is a perennial and pervasive source of energy. Solar electricity is ideal for remote electrification in the current context and therefore Stand-alone SPV power plants are the ideal choice for rural remote villages where conventional grid extension is not viable either due to inhospitable terrain or due to poor density of load. Solar technologies using concentrating systems for electrical production require sufficient direct beam radiation which is the beam radiation from the sun that passes through the planet's atmosphere without deviation and refraction.

202

Consequently, appropriate site locations are normally situated in arid to semi-arid regions where there is intense sunlight most of the time. Acceptable production costs of solar electricity occur where radiation levels exceed about 1700 kWh/m² year. Most Suitable regions include the south west United States, Northern Mexico, the North African desert, the Arabian Peninsula, major portions of India, Central and Western Australia, the high plateaus of the Andean states, and Northeastern Braziland and countries along the tropical in AFrica. Promising site locations in Europe are found in Southern Spain and several Mediterranean islands.

Atmospheric effect on Solar Radiation Using the definition in astronomy, air mass is the optical path length through the earth's atmosphere for light from a celestial source travel. As it passes through the atmosphere, light is attenuated by scattering and absorption and the more atmosphere through which it passes, the greater the attenuation. Consequently, celestial bodies on the horizon appear less bright than when they are at the zenith. The attenuation, known as atmospheric extinction is described quantitatively by the Beer-Lambert-Bouguer law. For utilization of solar energy, it is necessary to know the amount of depletion of incoming solar radiation by the atmosphere. It has been reported that for clear sky conditions, the fractions of direct solar radiation which is depleted due to various reasons are: Atmospheric scattering 9 % , Surface reflection 6 % . Other gases, smoke, dust etc. 3 %

Air Mass Air mass normally indicated in practical use is the relative air mass i.e. the path length relative to that at the zenith at sea level. By definition then, the sea-level air mass at the zenith is one. Air mass increases as the angle between the source and the zenith increases reaching a value of approximately 38 at the horizon. Air mass can be less than one at an elevation greater than sea level. However, most closedform expressions for air mass do not include the effects of elevation and so adjustment must usually be accomplished by other means. 203

Solar Air Massesangle adjustmennt

Daily and Seasonal Temperature Variations One of the most popular myths about the use of solar energy is that on cloudy days there will not be any electricity generation. However, this is not true. Consider the systems below;

Solar PV Systems In a cloudy day Solar PV panels produce electricity from diffused sunlight. The amount of energy that can be collected is certainly less than the amount that can be captured on a sunny day where the process of collection depends on degree of sunlight. This energy can be stored in batteries to cater to needs during the night.

Concentrating Solar Thermal Heating Systems Concentrating collectors produce electricity from direct sunlight. So they work best in climates that have a high amount of direct solar radiation. They do not function on cloudy days when available solar radiation is mostly diffused. The amount of useful heat they produce is mainly a function of the intensity of solar radiation available, the size of the reflector, how well they concentrate solar energy onto the receiver, the characteristics of the absorber, and the control of the flow rate of the heat transfer fluid. 204

Physical Parameters The following are the list of a few physical parameters that needs to be considered while selecting suitable location for the installation of solar energy system. These parameters are most appropriate for large scale solar systems like solar power plants.

Availability of Land and Foundation needs The land must be plain and continuous. None fertile or barren land should only be considered. Rocky terrain shall be preferred so that the cost of foundation will be cheaper.

Orientation and Obstructions The proposed land for SPV power generation must have a clear south facing without any obstruction in Southern hemisphere if you are in the Northern hemisphere.

Proximity of Power Evacuation Proximity of high tension sub-station is an important factor for the proposed site as the cost of laying transmission line is significant.

Water Availability Water is required for the construction purpose and for periodic cleaning of solar panels as a part of daily operation and maintenance.

Any industries of pollution nearby It is suggested that any site be selected should not have any polluting industries in the neighborhood. Otherwise the smoke and dust emitted by these industries forms a deposition on top of panels resulting in array losses.

Power supply for construction Availability of adequate power supply for construction work is necessary. 205

System Design of Solar PV Systems Load Analysis In the design of any system, several factors are considered. These factors include the following; Accurate Sizing Accurate sizing of the load involves analyzing the various components in the load list in terms of energy level requirements. It includes the current drawn by each component, Operating voltage range of that component and its expected duty cycle.

Peak current loads For equipment loads that are variable or pulsating, identify "peak" current levels unless definite patterns or duty cycles are determinable. Consider it when designed any PV system.

Worst case scenarios Assessment of worst case scenario is extremely important during the design process because any small increment in load apart from already assessed load can lead to system un-balance or a cycling down of battery capacity. So, assessing worst case load scenario is important. Worst case load scenario could be consequent to any load variations due to seasonal conditions.

Plan for the future The system must be designed focusing on the future needs. The system must be scalable to cater to the needs of expansion within its projected life cycle.

206

Compatibility issues The system must be compatible with existing systems to meet the load requirements. All the loads must be cross checked to ensure their compatibility of operation throughout the upper and lower voltage ranges of the solar system.

System design margins Additional design margins to be considered and kept at minimum level to make the system more cost effective because of the early consideration of worst case load scenario and possible system expansions.

Solar Array Design A Solar array is one of the major subsystems of any solar power generating system. Solar arrays are formed by connecting solar “modules” in a series and/or a parallel arrangement. These arrays produce direct current with respect to the incident solar radiation on them. The following are some of the factors that need to be considered in designing solar array for power generation:

Collector size The required solar collector area depends on the solar insolation level of a particular region. A region with poor insolation level will need a larger collector area than one with high insolation levels. Once insolation level of a region is known, the required collector size and energy output can be computed with some precision.

Selection of most appropriate module Solar modules are often rated on the basis of peak watts, and their electrical characteristics are described on a current-voltage curve popularly known as I-V curve. However, the most important factor is the module's behavior under expected operating conditions. One very important concern is module's charging

207

voltage generation under expected high temperature. It must be adequate to charge the battery after providing for system losses.

Dirt and Other contaminating effects Dirt and other contaminants including bird-droppings on the face of the solar array can reduce the power output power. Site conditions should be assessed to gauge the problems associated with contaminants. The mitigating solutions like special mounting considerations, more frequent cleaning of the solar modules is recommend. If the tilt angle of the array is less than 30°, buildup of dirt and other contaminants can be expected to be faster.

Orientation and Tilt issues The specific orientation and tilt of the solar array should be adopted to optimize system power during the worst-case periods of the year and when the average solar insolation is lowest and load requirements are highest. It may be desirable in certain locations to increase the array tilt to aid the clearing of snow and ice.

Design of Balance of Systems (BOS) Design of Balance of Systems is a very important factor in system design. Balance of systems includes the Charge controller, Battery, Cables etc. Balance of systems must be designed in such a way that it is neither too small nor too large. Battery selection and sizing is critical to overall system performance and reliability. The battery serves as an energy buffer storing excess energy produced by the solar array during the day and releasing that energy as required during night and periods of inclement weather when the array is unable to support the load. The following are some important factors that need to be considered in designing a battery bank.

The battery should be capable of handling both the physical and electrical rigors of the application, while providing the desired life expectancy and reliability. Key areas to be considered include: 208

(i) Cycle life, (ii) Capacity to withstand extended undercharged condition, (iii) Capability to withstand extended overcharging when not regulated, (iv) Charge efficiency and degree of self-discharge, (v) Need for equalization, (vi) Performance and life effects of temperature extremes, (vii) Tolerance of abuse,

(viii) Maintenance requirements.

Reserve Capacity The capacity of the battery should be sized to override the expected uncertainties in solar insolation and any seasonal periods when the array power is unable to fully match the load requirements.

Temperature and Ageing deration Battery performance is not static but will vary with age and environmental conditions. Battery performance should be de-rated to compensate for loss of capacity due to ageing and the reduction in available capacity due to low temperature. These factors will vary with type of battery. An additional consideration for certain applications will be the life-shortening effects of sustained high-temperature environments.

Regulation and Charge control A system regulator or Charge controller may be necessary to prevent excessive overcharge during peak periods of solar radiation which could damage some batteries particularly flat plate lead acid batteries and sealed maintenance-free batteries. A regulator or controller may also be desirable to reduce battery water consumption and extend required maintenance intervals.

209

Resource Measurement Assessment Global Solar Irradiance – Pyranometer The primary instrument used to measure global solar irradiance is the pyranometer. It measures the sun’s energy received from all directions (steradian) in the hemisphere, above the plane of the instrument. Global solar irradiance is the sum total of direct solar radiance and diffused solar irradiance. The most common pyranometer design uses a thermopile (multiple thermocouples connected in series) attached to a thin blackened absorbing surface shielded from convective loss and insulated against conductive losses. When placed in the sun, the surface attains a temperature proportional to the amount of radiant energy falling on it. The temperature is measured and converted through accurate calibration into a readable format of the global solar irradiance.

Pyranometers may also be used to measure the global solar irradiance on inclined surfaces. An example would be measurements from a pyranometer placed in the same plane as a tilted solar collector. As can be seen from the sketch in Figure 8.2, this measurement now includes solar energy reflected from surrounding surfaces.

Most clear day

210

Most cloudy day

Direct normal Solar Irradiance – Pyrheliometers The direct normal component of the solar irradiance can be measured by an instrument called Normal Incidence Pyrheliometer, (NIP). This device, shown in Figure 8.3 is essentially a thermopile pyranometer placed at the end of a long tube that is aimed at the sun. A two-axis tracking mechanism is incorporated to maintain the sun’s disc within the acceptance cone of the instrument.

A Normal Incidence Pyrheliometer (NIP) measures direct component of solar radiation

211

Diffused Solar Irradiance Pyranometers can also be used to measure the diffuse component of the global horizontal radiation. It can be done by providing a "shadowing" device just large enough to block out the direct irradiance coming from the sun’s disc. To avoid moving a shadowing disc throughout the day, a shadow band is often incorporated. This band must be adjusted often during the year to keep it in the ecliptic plane. Since the shadow band blocks part of the sky, corrections for this blockage must be used.

Recently, rotating shadow band pyranometers have come into general use. With this design, the shadow band rotates slowly about the pyranometer blocking the direct irradiance from the sun every time it passes in front of the pyranometer. The signal from the pyranometer reads global horizontal irradiance most of the time with reductions down to the diffuse irradiance level when the shadow band passes between the sun and the pyranometer. This design gives the advantage of using a single pyranometer to measure both global horizontal and diffuse horizontal solar irradiance.

Measurement Tools (i) Sunshine Recorders In addition to the pyranometer and the normal incidence pyrheliometer, which measure global and direct solar irradiance respectively, there is a traditional measurement often reported in meteorological observation. This is the "duration of sunshine". The traditional standard instrument used to measure this parameter is the Campbell-Stokes sunshine recorder. This instrument consists of a glass sphere that focuses the direct solar radiation and burns a trace on a special pasteboard card. These recorders have been replaced in most installations by photo detector activated ‘sunshine switches.’ The data produced by these instruments are of minimal use to engineers because there is no measure of intensity other than “threshold” intensity.

212

(ii) Cloud -Cover Observations Another source of solar irradiance data is from periodic ground observations of cloud-cover. These are made on an hourly-basis at weather observation stations around the world. Examining the SOLMET weather data tape format, the analysis of these observations are carried out in the United States of America. Cloud-cover data along with other weather data have been used to predict solar irradiance levels for the locations which do not have solar irradiance measurement capabilities.

Satellite Observations A similar type of measurement correlation using satellite images appears to provide accurate solar irradiance data over a wide region to a resolution of about 10 km. The results obtained with the use of satellite images made half-hourly in the visible (0.55 - 0.75 micrometer) and IR (9 - 12 micrometer) regions of the spectrum. Recently, efforts have been made to predict accurately solar irradiance from ground reflectance. The models for producing reliable solar irradiance data from satellite have been developed and validated.

Solar Radiation Databases When designing a solar energy system, the best way to predict its energyproduction performance is by calculating the minute-by-minute solar irradiance levels over the lifetime of the system with respect to exact location of the system. Since weather patterns are difficult to predict and are somewhat random with regard to time and place, the system designer is forced to accept historical data, recorded at a different location and with values reconstructed from incomplete data records. Because of the inherent variability of future solar irradiance, historical records are an extremely useful analytical tool appropriate for a wide range of applications. However, the designer must keep in mind that system performance predicted using even the best historical data may not necessarily represent the future output of the system.

213

Typical Meteorological Year Data Sets - TMY A typical meteorological year data set is made up from historical weather observations for a set of twelve typical months at a specific location. Each typical month is chosen from a multi-year set of data for a specific month and selected because of having the ‘average’ solar radiation for that month. For example, solar radiation data for January of maybe 30 different years is searched to determine the year in which the January was typical or average. Next, 30 different February data sets are searched to determine the typical February. As is usually the case, the typical January and the typical February may not be from the same year. Typical months are determined for the remaining months and some data smoothing done for the transition between months. An hour-by-hour database of readings for all recorded weather parameters from each of the ‘typical’ months is then generated and is called a Typical Meteorological Year.

European and Worldwide Solar Radiation Databases A solar radiation database atlas has been developed under the auspices of the European Union. This atlas offers a unique instrument dedicated to the knowledge and exploitation of the solar resources for Europe in a broad sense from Ural to Azores and from Northern Africa to Polar Circle covering the period 1981 - 1990. A computer program permitting calculation of hourly values of solar radiation data throughout the world is available and has been validated at many sites. The program is continually being updated to include more weather station data reducing the amount of extrapolation necessary between sites.

Solar Cell I-V and C-V Measurements The search for alternative sources of power is imperative because of the increasing demand for energy and the limited supply of fossil fuels. Given that there is a vast amount of energy available from the sun’s radiations devices that convert light energy into electrical energy are becoming increasingly important. Solar cells or photovoltaic (PV) cells convert light energy into useful electrical power and these cells are produced from light-absorbing materials. When the cell 214

is illuminated, optically generated carriers produce an electric current when the cell is connected to a load.

A variety of measurements are made to determine the electrical characteristics of PV cells. Characterizing these cells often involves the measuring the current and capacitance parameters as a function of an applied DC voltage. These measurements are usually done at different light intensities and temperature conditions whereby important device parameters can be extracted from the current-voltage (I-V) and capacitance-voltage (C-V) measurements such as the conversion efficiency and the maximum power output.

Electrical characterization is also important to determine losses in the PV cell. Essentially, electrical characterization is needed to determine ways into which one can make the cells as efficient as possible with minimal losses. To make these important electrical measurements, the Model 4200-SCS Semiconductor Characterization System can simplify testing and analysis. The 4200-SCS is a measurement system that includes instruments for both I-V and C-V measurements as well as software, graphics, and mathematical analysis capability. The software includes tests for making I-V and C-V measurements specifically on solar cells and deriving common PV cell parameters from the test data.

Photovoltaic Cell Circuit and Device Parameters A photovoltaic cell may be represented by the equivalent circuit model shown in Figure 8.6. This model consists of current due to optical generation (IL), a diode that generates a current [IseqV/kT], a series resistance (rs) and shunt resistance (r sh). The series resistance is due to the resistance of the metal contacts, ohmic losses in the front surface of the cell, impurity concentrations and junction depth. The series resistance is an important parameter because it reduces both the shortcircuit current and the maximum power output of the cell. Ideally, the series resistance should be 0Ω (rs = 0). The shunt resistance represents the loss due to surface 215

leakage along the edge of the cell or due to crystal defects. Ideally, the shunt resistance should be infinite (rsh = ∞).

Idealized equivalent circuit of a photovoltaic cell

If a load resistor (RL) is connected to an illuminated PV cell, then the total current becomes: I = IS(eqV/kT – 1) – IL, where, IS = current due to diode saturation IL = current due to optical generation. Several factors determine the efficiency of the solar cell, including the maximum power point (Pmax), the energy conversion efficiency (η), and the fill factor (FF). These points are illustrated which shows a typical forward bias I-V curve of an illuminated PV cell. The maximum power point (Pmax) is the product of the maximum cell current (Imax) and voltage (Vmax) where the power output of the cell is greatest. This point is located at the “knee” of the curve. The fill factor is a measure of how far the I-V characteristics of an actual PV cell differ from those of an ideal cell. The fill factor is defined as: FF = (Imax Vmax)/ (IscVoc) where, Imax = the current at the maximum power output, Vmax = the voltage at the maximum power output, Isc = the short-circuit current and Voc = the open-circuit voltage. Another important parameter is the conversion efficiency (η), which is defined as the ratio of the maximum power output to the power input to the cell: η = Pmax / Pin 216

where, Pmax = the maximum power output Pin = the power input to the cell defined as the total radiant energy incident on the surface of the cell. These described parameters of the solar cell can be determined through electrical characterization of the device.

Knee of the I-V curve

Measuremement Using the 4200-SCS on a Solar Cell To simplify testing, a project has been created for the 4200-SCS that makes both I-V and C-V measurements on a solar cell and also extracts common measurement parameters such as maximum power, short-circuit current, opencircuit voltage, etc. The project is called “CVUPvcell” and is included with all 4200-SCS systems running KITE version 7.0. A screen shot of the project is shown in Figure 8.8. This project has five tests, called ITMs (Interactive Test Modules) that perform a forward bias I-V sweep (fwd-ivsweep), reverse bias I-V sweep (rev-ivsweep), C-V sweep (cv sweep), 1/C2 vs. V plot (C-2vs V) and C-f sweep. Many important device parameters can be determined from currentvoltage (I-V) measurements of the solar cell as mentioned earlier.

The I-V characteristics are measured using one of the Model 4200-SCS’s Source Measure Units (SMUs) which can source and measure both current and voltage. Two types of SMUs are available for the 4200-SCS; the Model 4200-SMU which 217

can source/sink up to 100 mA, and the 4210-SMU which can source/sink up to 1A. If the output current of the cell exceeds these current levels then the output current may have to be reduced. One way of reducing the output is to reduce the area of the cell.

Making connections to the PV Cell A solar cell connected to the 4200-SCS’s SMU for I-V measurements is shown in figure below. A four-wire connection is made to eliminate the lead resistance that could otherwise affect the measurement accuracy. With the four-wire method, a voltage is sourced across the PV cell using one pair of leads (Force HI and Force LO), and the voltage drop across the cell is measured across a second set of leads (Sense HI and Sense LO). The sense leads ensure that the voltage developed across the cell is the programmed output value and compensates for the lead resistance. The resistence or impedenca plots are shown.

Forward Bias I-V Measurements Forward bias I-V measurements of the PV cell are generated under controlled illumination. The SMU is set up to source a voltage sweep and measure the resulting current. This forward bias sweep can be accomplished using the “fwdivsweep” ITM. The user can adjust the sweep voltage to the desired values. The parameters VOC and ISC can easily be derived from the sweep data using the Model 4200-SCS’s built-in mathematical analysis tool, the Formulator. For convenience, the “CVU_Pvcell” project has the common parameters already calculated and the values automatically appear in the Sheet tab every time the test is executed.

The figure below shows some of the derived parameters in the Sheet tab. These parameters include the short-circuit current (ISC), the open circuit voltage (VOC), the maximum power point (Pmax), the maximum cell current (Imax), the maximum cell voltage (Vmax), and the fill factor (FF). Using the Formulator, the conversion efficiency (η) can also be calculated if the power input to the cell is known. The 218

current density (J) can also be derived using the area of the cell. Figure 8.11 shows an actual I-V sweep of an illuminated silicon PV cell generated by the 4200-SCS using the “fwd-ivsweep” ITM. Because the system’s SMUs can sink current, the curve can pass through the fourth quadrant and allow power to be extracted from the device. Sometimes it may be desirable to plot log I vs. V. The Graph tab options support an easy transition between graphically displaying data on either a linear or a log scale.

I-V Sweep of Silicon PV Cell Generated with the 4200-SMU

The series resistance, (rs), can be determined from the forward I-V sweep at two or more light intensities. First, make I-V curves at two different intensities. Knowing the magnitudes of the intensities is not important. Measure the slope of this curve from the far forward characteristics where the curve becomes linear. The inverse of this slope yields the series resistance: rs = ΔV /ΔI Using additional light intensities, this technique can be extended using multiple points located near the knee of the curves which then develops a line is generated from which the series resistance which can be calculated from the slope. An important measurement feature of the system’s SMU as an ammeter is that it has very low voltage burden. The voltage burden is the voltage drop across the ammeter during the measurement. Most conventional digital multimeters 219

(DMMs) will have a voltage burden of at least 200mV at full scale. Given that only millivolts may be sourced to the sample, this can cause large errors. The 4200-SCS’s SMU never produces more than a few hundred microvolts of voltage burden, or voltage drop, in the measurement circuit.

Reverse Bias I-V Measurements The leakage current and shunt resistance (r sh) can be derived from the reverse bias I-V data where typically the test is performed in the dark. The voltage is sourced from 0 V to a voltage level where the device begins to break down. The resulting current is measured and plotted as a function of the voltage. Depending on the size of the cell, the leakage current can be as small as in the picoamp region. The Model 4200-SCS has a preamp option that allows making accurate measurements well below a picoamp. When making very sensitive low current measurements of nanoamps and smaller, low noise cables and place the device in a shielded enclosure to shield the device electrostaticallye are used.

This conductive shield is connected to the Force LO terminal of the 4200-SCS. The Force LO terminal connection can be made from the outside shell of the triax connectors, the black binding post on the ground unit (GNDU), or from the Force LO triax connector on the GNDU. One method for determining the shunt resistance of the PV cell is from the slope of the reverse bias I-V curve. From the linear region of this curve, the shunt resistance can be calculated as: rs = ΔVReverse Bias / ΔIReverse Bias

220

Actual Reverse Bias Measurement of Silicon PV Cell Using 4200-SMU

Capacitance Measurements In addition to determining the I-V characteristics of a PV cell, capacitance-voltage measurements are also useful in deriving particular parameters about the device. Depending on the type of PV cell, the AC capacitance can be used to derive such parameters as doping concentration and the built-in voltage of the junction. A capacitance-frequency sweep can be used to provide information about the existence of traps in the depletion region. The Model 4200-CVU, the Model 4200SCS’s optional Capacitance-Voltage Unit can measure the capacitance as a function of an applied DC voltage (C-V), a function of frequency (C-f), or a function of time (C-t). To make a C-V measurement, a solar cell is connected to the 4200CVU as shown in ealier. Like I-V measurements made with the SMU, the C-V measurement also involves a four-wire connection to compensate for lead resistance.

The HPOT/HCUR terminals are connected to the anode and the LPOT/LCUR terminals are connected to the cathode. This connects the high DC voltage source terminal of the CVU to the anode. The shields from the coax cables need to be 221

connected together as close as possible to the solar cell. Connecting the shields together is necessary for obtaining the highest accuracy because it reduces the effects of the inductance in the measurement circuit. This is especially important for capacitance measurements made at the higher test frequencies. To reduce the effects of cable capacitance, it is also important to perform a SHORT cal, OPEN cal, and Cable Correction.. Given that the capacitance of the cell is directly related to the area of the device, it may be necessary to reduce the area, if possible, to avoid capacitances that may be too high to measure. Also, setting the 4200-CVU to measure capacitance at a lower test frequency (10 kHz) and/or lower AC drive voltage will allow making higher capacitance measurements.

C-V Sweep C-V measurements can be made either forward-biased or reverse-biased. However, when the cell is forward-biased, the applied DC voltage must be limited otherwise the conductance may get too high. The maximum DC current cannot be greater than 10mA otherwise the DC voltage output will not be at the desired level. Figure 8.13 illustrates a C-V curve of a silicon solar cell generated by the 4200-CVU using the “cvsweep” ITM. This test was performed in the dark while the cell was reverse-biased.

Instead of plotting dC/dV, it is sometimes desirable to view the data as 1/C s vs. V. The doping density (N) can be derived from the slope of this curve because N is related to the capacitance by: N(a) = 2 /qESA2[d(1/C2)/dV] where: N(a) = the doping density (1/cm3), q = the electron charge (1.60219 × 10– 19

C), Es = semiconductor permittivity, 1.034 × 10–12 F/cm for silicon), A = area

(cm2), C = measured capacitance (F), V = applied DC voltage (V). The built-in voltage of the cell junction can be derived from the intersection of the 1/C2 curve and the horizontal axis. This plot should be a fairly straight line. This graph was generated using the “C-2 vs V” ITM. The “Linear Line Fits” graph option can be 222

used to derive both the doping density (N) and the built-in voltage on the x-axis. The doping density is calculated as a function of voltage in the Formulator. The user must input the Area of the device in the Constants area of the Formulator.

C-V Sweep of Silicon Solar Cell

1/C2 vs. Voltage for Silicon Solar Cell

223

C-f Sweep The 4200-CVU can also measure capacitance as a function of frequency. The curve in Figure 8.15 was generated by using the “cf sweep” ITM. It is possible to adjust the range of sweep frequency as well as the bias voltage.

C-f Sweep of Solar Cell

Measuring the electrical characteristics of a solar cell is critical for determining the device’s output performance and efficiency. The Model 4200-SCS simplifies cell testing by automating the I-V and C-V measurements and provides graphics and analysis capability. In addition to the tests described here, the 4200-SCS can also be used to make resistivity measurements on the materials used for the PV cells, a process that is refered to as “Four-Probe Resistivity and Hall Voltage Measurements with the Model 4200-SCS.

224

CHAPTER NINE

MATERIALS FOR SOLAR CELLS Alternative Sources of Energy The sun provides enormous amounts of energy powering oceans, atmospheric currents, and cycles of evaporation, drives river flow, hurricanes and tornadoes that destroy natural landscape. The San Francisco earthquake of 1906 with magnitude 7.8 released an estimated 1017 joules of energy which is equivelant to what sun delivers in one second. Earth’s resource of oil mounts up to 3 trillion barrels containing 1.72 x1022 joules of energy that the sun supplies in 1.5 days. Humans annually use about 4.6 x 1020 joules annually which sun supplies in one hour. The sun continuously supplies about 1.2 x 1025 Terawatts of energy which is very much greater than any other renewable or nonrenewable sources of energy can provide.

This energy is much greater than the energy required by human beings which is about 13 Terawatts. By covering 0.16 % of Earth’s land with 10 % efficient solar cells would provide 20 Terawatts of energy about twice of fossil fuel consumption of the world including numerous nuclear fission reactors. Solar energy is in abundance but only very little is used directly to power human activities. About 80 % - 85 % of our total energy comes from fossil fuels. These resources are nonrenewable, fast depleting, produce green-house gases and other harmful environmental pollutants. Threat to climate is one of the main concerns in adopting any resource as a primary source of energy. Fossil fuels emit a large volume of greenhouse gas like CO2 into the atmosphere that disturbs the ecological balance.

These emissions have been increasing due to over-utilization of fuels to meet the ever expanding needs of human society. The solutions for this problem include the use fossil fuels in conjunction with carbon sequestration, nuclear power and

225

solar power. Carbon sequestration is an extremely difficult method since a large volume of space is required to store the emitted greenhouse gases and its maintenance is a very crucial issue.

Nuclear power seems to be a good option but the feasibility of deploying several thousands of 1Gegawatt power plants all over the world to meet the 10 Terawatt demand of the society is skeptical. The Uranium resource for these power plants on earth also gets exhausted in this process in about 10 years after which the processing of sea water has to be adopted which is also exhaustible and difficult. On the other hand shifting the focus on renewable sources of energy is the ideal choice and solar power is by far the most prominent energy source owing to its versatility, inexhaustible and environmental friendly features.

Annual Production of Oil

The exhaustible nature of fossil fuels has also pushed us into the adoption of renewable sources of energy as the future of fossil fuels seems deem. The figure below shows the plot of the annual production of oil per year with a 2 % annual growth and decline rate. We observe that these estimates show a very steep decline of this resource after the year 2016 thus demanding the need for an 226

alternative source of energy. The burgeoning solar cell market is maturing to become a very profitable investment to industries resulting in an annual growth of 41 % in the last five years.

High costs and conversion efficiency have been the major bottle-necks in the potential of solar power becoming a primary source of energy. Nowadays major research done with the motive of improving the efficiency of these cells has brought this dream closer to reality. New methods of harnessing the full spectrum of the sun’s wavelength, multi-junction solar cells whether homo-junctions or hetero-junctions, and new materials for making solar cells are paving way for solar power to be the emerging power resource for the world at large.

Global PV Installations by Year

227

Best Research Cell Efficiencies

Generations of Solar Cells Solar cells are categorized into three generations based on the order of their prominence and the time they were fabricated. Research is being conducted on all the three generations concurrently to improve their efficiencies while the first generation solar cells comprise the major share of commercial production about 89.7 % in 2007. Large area, high quality and single junction devices form the ‘first generation’ solar cells. Reduction in production costs of this technology is nullified owing to high energy and labour costs, material costs mostly for the silicon wafer, strengthened low-iron glass cover sheet and costs of other encapsulants.

This trend is continuing as the photovoltaic industry is expanding. Although it has a broad spectral absorption range the high energy photons at the blue and violent 228

end of the spectrum is wasted as heat. Producing solar cells using high-efficiency processing sequences with high energy conversion efficiency are thus favoured provided they do not increase the complexity of the solar cell. The theoretical limit on efficiency for single junction silicon solar cells is 33 % and this is also being attained very rapidly.

First Generation Solar Cell Efficiencies

To address these problems of energy requirements and production costs of solar cells, a switch from ‘first generation’ to ‘second generation’ of thin-film cell technology was imminent. Eliminating the silicon wafer a major reduction in material costs has been possible in the thin-film technology since they also have an advantage of increasing the unit size from silicon (about 100 cm2) to glass plate (~1m2). Over time the second generation solar cells are expected to bridge the gap between them and the first generation cells with respect to energy conversion efficiency.

The materials generally used in this thin film technology are cadmium telluride, copper indium gallium arsenide, amorphous silicon and micromorphous silicon. With the increase in dominance of this technology, the costs of the constituent 229

materials also goes up for top cover and other encapsulants to give it a longer life.These materials reduce mass and cost by forming substrates for supporting glass and ceramics. Not only do they reduce costs but also promise very high energy conversion efficiency.

A trend towards shifting to second generation from first generation is showing up but the commercialization of this technology has proved difficult. Fortunately with the development of new materials over the coming decades, the future of thin-film technology seems to be promising. Research for improving solar cell performance by enhancing its efficiency and pushing it closer to the thermodynamic limits has led to the development of third generation solar cells. To improve upon the poor electrical performance of the thin film technology by maintaining low production costs, this technology includes among others nonsemiconductor technologies like polymer based cells and biometrics.

Second Generation Solar Cell Efficiencies

These devices comprising the third generation solar cells are quantum dot technologies, tandem/multi junction cells, hot-carrier cells, up conversion technologies and solar thermal technologies like thermo-photonics. 230

Third Generation Solar Cell Efficiencies

Ultimate Limits in Efficiency The steady evolutionary progress of the PV industry is the result of increase in automation of production of thin film solar cells with increased efficiency and lowering costs. The need for revolutionizing break-throughs in the PV industry is sometimes halted by the advancements in the PV materials and manufacturing technology. This leads to improvements in the cost competitiveness and the expansion of the PV market. By shattering the old limits of efficiency and cost by bringing about innovations, by exploiting new understanding of physics and material science will become a fast paced revolution.

The maximum theoretical limit for a single junction solar cell without sunlight is about 31 % established by the Schokley-Quiesser limit. Under the highest possible amount of sunlight i.e. 50,000 suns a single junction solar cell can have a maximum efficiency of about 41%. This efficiency value can be increased by using multi-junction solar cells by capturing more of the solar spectrum. The true limit of efficiency is the thermodynamic limit of 68 % for PV with one sun concentration and is about 87 % for maximum solar concentration. Research and development in the PV industry are developing technologies based on concepts 231

such as multiple exciton generation, optical frequency shifting, multi energy level and hot carrier devices. Carbon nanotubes, organic materials and other nanofabrication technologies enable the development of these concepts into practice.

Data showing potential magnitude of future improvements

Thin film solar cells The crystalline silicon is often referred to as the first generation photovoltaic technology, while the second generation photovoltaics consists of thin film solar cell materials such as amorphous silicon (a-Si), cadmium telluride (CdTe), copper indium gallium diselenide (CIGS) and thin film crystalline silicon. The driving force for the development of thin film solar cells has been their potential for the reduction of manufacturing costs. While silicon solar panels are assembled from individual cells processed from about 100 cm2 silicon wafers, thin film semiconductor materials can be deposited onto large surfaces which is beneficial for volume production. Also as direct band gap semiconductors the thin film semiconductor materials have much higher absorption coefficient than silicon and 232

therefore less than 1 mm thick semiconductor layer is required, which is 100 1000 times less than for Si. The amount of expensive semiconductor material is reduced or on the more expensive semiconductors can then be used in the thin films.

A photovoltaic substance is a material used in the creation solar cells that convert sunlight directly into electricity. The long-term goal of photovoltaic (PV) devices has been to reduce our dependency on fossil fuel generated electricity. Presently, most PV solar cells are produced from either single- or polycrystalline silicon with efficiencies of around 13 – 17 %. This first generation of photovoltaics is heavily reliant on the supply of pure silicon in the form of single crystals. Technology has now been developed to produce polycrystalline silicon specifically for the PV industry with lower purity and lower costs.

However, the main disadvantage of this material is the lower cell efficiency. Therefore, critical to the long-term commercial success of PV technologies, are advances in module efficiencies and improvements in cost and reliability. In an effort to circumnavigate these issues, two emerging technologies have been developed based on either Copper Indium Gallium Selenide (CIGS) or Cadmium Telluride (CdTe) solar cell devices. The advantages of these thin-film PV’s include lower cost, reduced material usage, lower energy requirement for manufacture and shorter payback time.

Crystalline Silicon The most prevalent bulk material for solar cells is crystalline silicon also known as solar grade silicon. Bulk silicon is separated into multiple categories according to crystallinity and crystal size in the resulting ingot, ribbon, thin film, thick film or wafer.

233

Monocrystalline Silicon Monocrystalline silicon is often made using the Czochralski process, but the single crystal wafer cells are finally expensive. Since they are cut from cylindrical ingots, they do not cover a square solar cell module without a substantial waste of refined silicon. Therefore, most monocrystalline silicon panels have uncovered gaps at the four corners of the cell. Ribbon Silicon is a type of monocrystalline silicon formed by drawing flat thin films from molten silicon and has a multicrystalline structure. Although less efficient, ribbon silicon cells are more cost effective because their production does not require sawing from ingots.

Poly- or Multicrystalline Silicon Large blocks of molten silicon are carefully cooled and solidified to create cast square ingots which are less expensive to produce than monocrystalline silicon cells however multicrystalline cells are also less efficient than the latter. Advantages of Crystalline Silicon include up to 30 % efficiency and long lifespan.

CIGS Copper indium diselenide (CIS) or CIGS when gallium is added is perhaps the most promising thin film material at the moment. It holds the record laboratory efficiency of 18.8 % among all the thin film materials and shows excellent stability. The efficiency of the CIS cells tends to rise during the first 10 - 30 minutes of light exposureand thus the first pilot line products are now being marketed by Siemens with efficiencies above 10 %. Small amount of cadmium and selenium is present also in the CIGS solar cell. The concerns associated usually with the CdTe apply strictly speaking also to the CIGS technology to some extent. Unlike silicon solar cells which are based on a p-n junction, the structure of CIGS is a complex heterojunction.

Photovoltaic devices based on Cu(In,Ga)Se2 -(CIGS) perform with 19% efficiency, the highest of any thin film technology recorded so far. However, the possibilities for even higher performance are significant. CIGS cells use either a 234

glass, steel or polymer substrate onto which a layer of molybdenum is sputtered (Back Contact) and a 1.3 - 2.5 mm thick CIGS layer can be deposited in various ways including sputtering, co-evaporation, printing and electroplating.

A thin cadmium sulfide buffer is added next, which is prepared by chemical bath deposition. It is then over-coated with a transparent conductor e.g. aluminium doped zinc oxide. Advantages of CIGS include modules that can be on glass and also flexible steel or polymer substrates. They are able to produce lower-cost PV with less material usage, highest efficiency of thin-film technologies up to 19.9 % for single cells and 13 % for large modules.

CdTe Cadmium Telluride (CdTe) is an ideal semiconductor for use in photovoltaic applications as the band gap of about 1.5 eV almost perfectly matches the solar spectrum. Thin-films of CdTe can be deposited onto a substrate by a variety of different methods including evaporation, electrodeposition and printing. Photovoltaic modules based on thin-film cadmium telluride technology are of great interest for large-scale grid connected applications, primarily due to their cost benefits over other PV technologies.

Cadmium telluride has a nearly optimal band gap and is easily deposited with thin film techniques where over 16 % efficiencies have been demonstrated in the laboratory. An often discussed issue on the CdTe technology is the cadmium content and level in the cells. While this question can be easily and most certainly be solved by recycling the Cd, the public acceptance of the technology may be somewhat a trickier question. Solar electricity usually marketed as an environmentally friend form energy may not be tolerated to imply environmental concerns about hazardous metals used for the manufacture of the solar cells.

235

Schematic representation of Cadmium Terrude solar cell

Production plants for commercial presentation of the CdTe cells are being built and the expected module efficiencies are in the range of 8 – 9 %. Advantages of CdTe includes a simplified manufacturing abundance supply of cadmium element produced as a by-product of industrial metals such as zinc and absorbs sunlight at close to their ideal wavelength with above 10.6 efficiency modules.

a-Si Amorphous silicon (a-Si) solar cells have been commercially available for many years now and hold 13 % of the worlds PV market. The special market for a-Si cells has been low-power sources for consumer electronics where processability weights more than the absolute solar conversion efficiency. The problem with aSi has been the degradation of efficiency by the Stäbler-Wronski effect and the stabilized efficiencies of only in the 4 – 5 % range obtained for single junction cells. However, decreasing the thickness of the active a-Si layer can enhance the stability of the a-Si cells.

The best commercial a-Si cells utilize a stacked three-layer structure with stabilized efficiencies of 6.4 %. The three-junction concept also enables tuning the band gap of individual layers to achieve higher efficiencies. A three-junction

236

a-Si tandem-cell technology (tendem structure) utilizing a-SiGe alloy in the two lower junctions holds the record confirmed module efficiency of 10.4 %.

Thin film Si One way to reduce the amount of Si in the solar cell is to use optical confinement of light trapping; an idea is utilized in the thin film crystalline silicon technology. When crystalline Si layer with flat surface is deposited on a Lambertian reflector surface, the thickness of the Si layer can be reduced to about 5 - 50 mm. Such a thin Si layer needs a supporting substrate which can be either low-grade silicon or foreign materials like graphite or glass. Presently the crystalline thin film Silicon technology is actively developed in the laboratories and the first crystalline silicon thin film product with 12.2 % efficiency is about to be brought out commercially by Astropower.

III-V Semiconductors Semiconductors such as GaAs, GaAlAs, GaInAsP, InAs, InSb, and InP are interesting solar cell materials because they have near-optimal band gaps. These materials are extremely expensive and have found applications mainly in space solar cells where performance is a more important criterion than their cost and in some extent in concentrating systems where the active surface area of the cells can be reduced significantly and therefore allowing expensive materials to be used.

Photoelectrochemical solar cells The oldest type of photovoltaic cell to be fabricated was the photoelectrochemical solar cell already used by Becquerel when he discovered the photovoltaic effect in 1839. In any of the photoelectrochemical solar cell, a semiconductor-electrolyte junction is used as a photoactive layer. While energy conversion efficiencies exceeding 16 % have been achieved with the photo-electrochemical solar cells utilizing semiconductor photoelectrodes, instability of these solar cells by photocorrosion has left them without practical importance. Furthermore, the photo237

electrochemical solar cells using the same semiconductor materials as in the commercial solar cells such as Si, CuInSe2 or GaAs does not offer any real advantages over the established solid-state solar cells.

Dye-Sensitised Mesoscopic Solar Cells -Michael Grätzel Photovoltaic devices are based on charge separation at an interface of two materials of different conduction mechanism. To date, this field of photovoltaics has been dominated by solid-state junction devices usually made of silicon, profiting from experience and the abundance in material availability resulting from the semiconductor industry. The dominance of the photovoltaic field by inorganic solid-state junction devices is now being challenged by the emergence of a range of new device concepts including devices based on nanocrystalline and conducting polymer films. These devices offer the prospect of very low-cost fabrication and present a range of attractive features that will facilitate market entry.

It is now possible to depart completely from the classical solid-state junction device by replacing the phase contacting the semiconductor by an electrolyte, liquid, gel or solid thereby forming a photoelectrochemical cell. The phenomenal progress

realised

recently

in

the

fabrication

and

characterisation

of

nanocrystalline materials has opened up vast new opportunities for these systems. Contrary to expectations in other semiconductors, devices based on interpenetrating networks of mesoscopic semiconductors have shown strikingly high conversion efficiencies competing with those of conventional devices. The prototype of this family of devices is the dye-sensitised solar cell (DSSC) which realises the optical absorption and charge-separation processes by the association of a sensitiser as light-absorbing material with a wide-band-gap semiconductor of nanocrystalline morphology.

238

Historical background The history of the sensitisation of semiconductors to light radiation of wavelength longer than that corresponding to the semiconductor band-gap has been long. It represents an interesting convergence of photography and photoelectrochemistry both of which rely on photoinduced charge separation at a liquid-solid interface. The silver halides used in photography have band-gaps of the order of 2.7 – 3.2 eV and which therefore, are insensitive to much of the visible spectrum are now used in dye-sensitised solar cells. The first panchromatic films able to render the image of a scene more realistically into black and white followed the work of Vogel in Berlin after 1873 in which he associated dyes with the halide semiconductor grains.

The first sensitization of a photoelectrode followed shortly thereafter using similar chemistry. However, the clear recognition of the parallels between the two procedures, the realisation that the same dyes can in principle function in both systems, and the verification that their operating mechanism is by injection of electrons from photo-excited dye molecules into the conduction band of the ntype semiconductor substrates, dates back only to the 1960s. In the years that followed, it became recognised that the eye could function most efficiently if chemi-absorbed on the surface of the semiconductor.

The concept of using dispersed particles to provide a sufficient interface area emerged then and was subsequently employed for photoelectrodes form ealier 1960s. Titanium dioxide (TiO2) quickly became the semiconductor of choice for the photoelectrode on account of its many advantages for sensitised photochemistry and photoelectron chemistry. It is a low-cost, widely available, non-toxic and bio-compatible material and as such, it is even used in healthcare products as well as industrial applications such as in paint pigmentation. Initial studies of the TiO2-based DSSC employed tris-bipyridyl ruthenium (II) dyes which are paradigm sensitisers for photochemical studies, functionalised by the

239

addition of carboxylate groups to attach the chromophore to the oxide substrate by chemisorption.

The standard pn-junction solar cell configuration from the 1960s Progress thereafter was incremental, a synergy of structure, substrate roughness and morphology, dye photo-physics and electrolyte redox chemistry contnued until the grand announcement in 1991 of a sensitised electrochemical photovoltaic device with an energy conversion efficiency of 7.1 % under solar illumination. That evolution has continued progressively since then with certified efficiencies now over 11 %.

Mode of function of dye-sensitised solar cells Device configuration At the heart of a DSSC system is a mesoporous oxide layer composed of nanometre-sized particles that have been sintered together to allow electronic conduction to take place. The material of choice has been TiO2 (anatase) although alternative wide-gap oxides such as ZnO, SnO2 and Nb2O5 have also been investigated. Attached to the surface of the nanocrystalline film is a monolayer of a sensitiser dye solar cell.

240

A schematic representation of the structure and component of the dyesensitizedsolar cell Photo-excitation of the dye results into the injection of an electron into the conduction band of the oxide thus generating a dye cation. The original state of the dye is subsequently restored by electron donation from the electrolyte. This step is often referred to as the regeneration reaction. The electrolyte usually comprises an iodide/triiodide redox couple dissolved in a liquid organic solvent although attention is increasingly focusing on alternatives for the solvent including ionic liquids, gelled electrolytes and polymer electrolytes.

Injected electron by the oxidised dye

241

The re-generation of the sensitiser by iodide intercepts the re-capture of the injected electron by the oxidised dye. The iodide is in turn regenerated by the reduction of triiodide at the counter electrode with the electrical circuit being completed via electron migration through the external load. The high surface area of the mesoporous metal oxide film is critical to efficient device performance as it allows strong absorption of solar irradiation to be achieved by only a monolayer of adsorbed sensitiser dye. Whereas a dye monolayer absorbed on a flat interface exhibits only negligible light absorption, the use of a mesoporous film dramatically enhances the interfacial surface area over the geometric surface area by up to a 1000- fold for a 10 µm thick film. This leads to a high visible-light absorbance from the many successive monolayers of adsorbed dye in the optical path.

Another advantage of the use of a dye monolayer is that there is no requirement for exciton diffusion to the dye/metal oxide interface. Also the non-radiative quenching of excited states often associated with thicker molecular films is avoided. The high surface area of such mesoporous films does however have a significant downside as it also enhances interfacial charge-recombination losses.

A recent alternative embodiment of the DSSC concept is the replacement of the redox electrolyte with a solid-state hole conductor which may be either inorganic, thereby avoiding the use of a redox electrolyte. Such solid-state sensitised heterojunctions can be regarded as functionally intermediate between redox electrolyte

based

photo-electrochemical

DSSCs

and

the

organic

bulk

heterojunctions. Devices efficiencies for such solid-state DSSCs are as yet limited to about 4 %, in contrast to efficiencies of over 11% achieved for the more widely studied redox electrolyte-based DSSCs.

Device fabrication DSSCs are fabricated on transparent conducting oxide (TCO) glass substrates enabling light irradiance through this substrate under photovoltaic operation. The 242

conductive coating used is fluorine-doped SnO2 (FTO) preferred over its indiumdoped analogue (ITO) for reasons of lower cost and enhanced stability. Prior to deposition of the mesoporous TiO2 film, a dense TiO2 film may be deposited to act as a hole blocking layer preventing recombination (shunt resistance) losses between electrons in the FTO and oxidised redox couple.

The TiO2 nanoparticles are fabricated by the aqueous hydrolysis of titanium alkoxide precursors, followed by autoclaving at temperatures up to 240 0C to achieve the desired nanoparticle dimensions and crystallinity (anatase). The nanoparticles are deposited as a colloidal suspension by screen printing or by spreading it with a doctor blade, followed by sintering at about 450 0C to ensure good inter-particle connectivity. The film porosity is controlled by an addition of organic filler such as carbowax to the suspension prior to deposition. This filler is subsequently burnt off during the sintering step.

Schematic picture representing the two main DSSC cell (upper) and module (lower) architectures: a) the sandwich structure b) the monolithic structure The classic sensitiser dye usually employed in DSSCs is a ruthenium (II) bipyridyl

dye,

cis-bis(isothiocyanato)-bis(2,2’-bipyridyl-4,4’-dicarboxylato)243

ruthenium(II) also known as ‘N3’ or in its partially deprotonated form (a di-tetra butylammonium salt) as ‘N719’. The incorporation of carboxylate groups allows ligation to the film surface via the formation of bidendate and ester linkages whilst the (–NCS) groups enhance the absorption of visible light.

Adsorption of the dye to the mesoporous film is achieved by simple immersion of the TiO2 film in a solution of dye. This results in the conformal adsorption of a dye monolayer to the film surface. The counter electrode is fabricated from FTOcoated glass with the addition of a Pt catalyst to catalyse the reduction of the redox electrolyte at this electrode. Electrical contact between working and counter electrodes is achieved by the redox electrolyte with capillary forces being sufficient to ensure the electrolyte efficiently penetrates the film pores.

Energetics of operation In contrast to silicon and organic devices, the high concentration of about 0.5 M of mobile ions in the electrolyte effectively screens out any macroscopic electric fields and charge separation is here primarily driven by the inherent energetics (oxidation/reduction potentials) of the different species at the TiO2/dye/electrolyte interface rather than by the presence of macroscopic electrostatic potential energy gradients. These charge-transport processes are primarily diffusive, driven by concentration gradients in the device and generated by the photoinduced charge separation.

Photoinduced charge separation occurs at the dye-sensitised TiO2/electrolyte interface. Electron injection requires that the dye must be in the excited state to be more reducing than the TiO2 conduction band enabling transfer of an electron from photoexcited dye into the metal oxide. This energetic requirement is equivalent to the dye excited-state LUMO orbital having a lower electron affinity than the electrode conduction-band edge. Regeneration of the dye’s ground state by the redox couple requires the dye cation to be more oxidizing than the iodide/triiodide redox couple. Charge separation in DSSCs is regarded as a two244

step redox cascade resulting in the injection of electrons into the TiO2 electrode and the concomitant oxidation of the redox electrolyte.

Both charge-separation reactions are thermodynamically downhill and can be achieved with near unity quantum efficiency in efficient devices. Following the charge separation, charge collection from the device requires transport of the photogenerated charges to their respective electrodes. For an efficient DSSC solar cell under AM1.5 illumination solar irradiance has thier charge fluxes in the order of 20 mA cm–2. The high ionic concentrations in the device effectively screen out any macroscopic electric fields thereby removing any significant drift component of these transport processes.

The transport of both electrons and redox ions are therefore both primarily driven by diffusion processes resulting from concentration gradients. Under optimum conditions, these charge-transport processes which involve electrons towards the FTO working electrode and tri-iodide towards the counter electrode can be efficiently driven with concentration gradients and therefore only small free energy losses are achieved (less than 50 meV). At the counter electrode, the triiodide is re-reduced back to iodide and the platinum catalyst enables this reaction to proceed with minimal overpotentials (again less than 50 meV). A similar ohmic contact is achieved at the TiO2/FTO interface. It is apparent that the energetics at the TiO2/dye/electrolyte interface is of primary importance in determining the overall device energetics.

Power output from the DSSC requires not only efficient charge collection by the electrodes but also the generation of a photovoltage corresponding to a free energy difference between the working and counter electrodes. In the dark and at equilibrium, the Fermi energy of the TiO2 electrode which corresponds to the free energy of electrons in this film after thermalisation equilibrates with the mid-point potential of the redox couple resulting into zero output voltage. Under these conditions, the TiO2 Fermi level lies deep within the band-gap of the 245

semiconductor and the film is effectively insulating with a negligible electron density in the TiO2 conduction band. Photo-excitation results into an electron injection into the TiO2 conduction band and concomitant hole injection into the redox electrolyte.

The high concentrations of oxidised and reduced redox couple present in the electrolyte in the dark a significant change in chemical potential of the electrolyte which remains effectively fixed at its dark resting value. In contrast, electron injection into the TiO2 conduction band results a dramatic increase in electron density (from the order of 1013 cm–3 to 1018 cm–3 ), raising the TiO2 Fermi level towards the conduction-band edge and allowing the film to become conducting.

This shift of the TiO2 Fermi level under irradiation increases the free energy of injected electrons and is responsible for the generation of the photovoltage in the external circuit. The mid-point potential of the redox couple is given by the Nernst equation and is therefore dependent on the relative concentrations of iodide and iodine. The concentrations of these species required for efficient device functioning are in turn constrained by kinetic requirements of dye regeneration at the working electrode and iodide regeneration at the counter electrode.

Concentrations of these species are in the range 0.1 – 0.7 M iodide and 10 – 200 mM iodine, constraining the mid-point potential of this electrolyte to ~ 0.4 V vs. NHE. It should furthermore be noted that in the presence of excess iodide, the iodine is primarily present in the form I3 resulting in this electrolyte often being referred to as an iodide/triiodide redox couple. Determining the energetics of the TiO2 conduction band is more complex. As with most oxides, the surface of TiO2 may be more or less protonated depending on the pH of the surrounding medium. This result into a change in surface-charge which cause the surface potential to exhibit a Nernstian dependence on effective pH and henc shifting by 60 mV per pH unit.

246

In many bulk metal oxides, surface-charge can result in significant bending of the conduction and valence bands adjacent to the surface. However, in the mesoporous TiO2 films employed in DSSCs, the nanoparticles are too small to support any significant band-bending. As a consequence, the whole conduction band of such mesoporous films shifts with the surface potential. At pH 1 in aqueous solutions, the conduction band edge for TiO2 shift negatively as pH is increased. DSSCs employ organic rather than aqueous electrolytes complicating quantification of the effective pH.

Nevertheless studies in organic solvents have demonstrated shifts of the conduction band of mesoporous TiO2 films of up to 1 V depending on the concentration of potential-determining ions primarily small cations such as protons or lithium cations in the electrolyte. For this reason, the concentration of such potential-determining ions in the electrolyte plays a key role in determining the energetics of the dye-sensitised interface, and thereby device performance. Additives added to the electrolyte to determine such energetics include Li+, guanadinium ions, N-methylbenzimidazol and t-butyl pyridine which functions as a base. Further influence on this interfacial energetics can be achieved by variation of the extent of protonation of the sensitiser dye.

The choice of suitable sensitiser dye energetics is essential to achieve suitable matching to the metal oxide and redox couple. The excited-state oxidation potential must be sufficiently negative to achieve efficient electron injection into the TiO2 conduction band whilst the ground-state oxidation potential must be sufficiently positive to oxidise the redox couple. The redox properties of adsorbed sensitiser dyes may differ significantly from those measured in solution mainly due to the high surface-charge densities and dipoles present at this interface.

The function of a DSSC involves the sequence of electron transfer and chargetransport processes which result in photovoltaic device function. In addition to the forward electron transfer and transport processes, there are several competing loss 247

pathways. These loss pathways include decay of the dye excited state to ground and charge recombination of injected electrons with dye cations and with the redox couple. Each charge-transfer step results in an increased spatial separation of electrons and holes, increasing the lifetime of the charge-separated state but at the expense of reducing the free energy stored in this state.

This functionality exhibits a close parallel to function of photosynthetic reaction centres. As in natural photosynthesis, kinetic competitions between the various forward and loss pathways are critical to determining the quantum efficiencies of charge separation and collection and are therefore key factors determining energy conversion efficiency. Here we have confined ourselves only to consideration of the impact of these dynamics on the performance of DSSCs. Efficient electron injection requires the rate of electron injection to exceed excited state decay to ground. Typical rates of dye excited-state decay to ground are in the range 107– 1010 s–1. The rate of electron injection depends on the electronic coupling between the dye excited-state LUMO orbital and accepting states in the TiO2 and on the relative energetics of these states. In model system studies of dye-sensitised metal oxide films, electron injection rates of greater than 1012 s–1 have been achieved for a range of sensitiser dyes consistent with efficient electron injection.

Schematic energy diagram of a bilayer donor-acceptor device under short circuit conditions

248

The working principle of the dye-sensitized nanostructured solar cell

It should be noted that fast electron injection dynamics require both strong electronic coupling of the dye LUMO orbital to the metal oxide conduction-band states and a sufficient free energy difference to drive the reaction. As such, electron-injection dynamics are dependent on the energetics of the TiO2 conduction band and therefore on the concentration of potential-determining ions (e.g. Li+) in the electrolyte. Omission of such ions from the electrolyte can result in an insufficient energetic driving force reducing the quantum yield of charge injection and thereby reducing device photocurrent.

It should be noted that the heavy metal ion in the ruthenium bipyridyl dyes widely employed in DSSCs results in ultrafast (up to 1013 s–1) intersystem crossing from the initially formed singlet excited state to a triplet state. This relaxation process reduces the excited-state free energy by about 400 meV. While several studies have reported subpicosecond electron injection from the singlet excited state of these dyes and thus an electron injection from the triplet state is significantly slower (1010 –1011 s–1) consistent with the lower free energy of this excited state, but still faster compared with decay of the triplet state to ground (107 – 108 s–1). Recent studies of complete devices have suggested that this relatively slow

249

electron injection from the dye triplet state is the dominating injection pathway in efficient DSSCs.

Efficient dye regeneration requires the rate of re-reduction of the dye cation by the redox couple to exceed that of charge recombination of injected electrons with these dye cations. This recombination reaction has been shown to be strongly dependent on the electron density in the TiO2 electrode and therefore light intensity and cell voltage. This accelerates by at least an order of magnitude between short-circuit and open-circuit conditions. It is furthermore dependent on the spatial separation of the dye cation (HOMO) orbital from the metal oxide surface with the rate constant decaying exponentially with distance which is consistent with electron tunneling theory.

The regeneration reaction is dependent on the iodide concentration, electrolyte viscosity and dye structure. For the N719 sensitiser dye employing employing a low-viscosity electrolyte such as acetonitrile, the regeneration reaction has a halftime of about 1 µs, sufficiently fast to compete effectively with the competing recombination reaction and ensuring that the regeneration reaction can be achieved with unity quantum efficiency. Efficient charge collection by the external circuit requires the time constant for electron transport within the TiO2 matrix to be faster than charge recombination of injected electrons with the redox couple.

Electron transport is a diffusive process that is strongly influenced by electron trapping in localised sub-band-gap states resulting into dynamics being strongly dependent on position of the TiO2 electron Fermi level and therefore raising the Fermi level towards the conduction-band edge resulting in increased trap filling. Electron-transport times under solar irradiation are of the order of milliseconds. If a low-viscosity solvent such as acetonitrile is used, transport of the oxidised redox couple to the counter electrode is not rate-limiting. However, the use of higher viscosity solvents more suitable for practical device applications for reasons of 250

device stability can result in significantly lower ionic diffusion constants with the resultant iodide/ triiodide concentration gradients causing significant free energy (series resistance) losses and potentially interfacial charge recombination.

Charge transfer processes between dye and the TiO2 lattice: 1). MLCT excitation, 2). Electron injection and 3).Charge recombination

Given the relatively slow time scale for charge transport in DSSCs compared with most other photovoltaic devices and the extensive interfacial area available for charge recombination in the device due to its mesoscopic structure, it is remarkable that the quantum efficiency of charge collection can approach unity. The key factor enabling this high efficiency is the slow rate constant for the interfacial charge recombination of injected electrons with the oxidised redox couple. This reaction is a multi-electron reaction, most simply being described by the equation which must therefore proceed via one or more intermediates states.

The mechanism of this reaction has been extensively studied and while the details remain somewhat controversial, it is apparent that without a suitable catalyst such as platinum, one or more of these intermediates steps exhibits a significant activation barrier resulting in a slow overall rate constant for this reaction. The low rate constant for this recombination reaction on TiO2 contrast to the facile electrochemistry of this redox couple on the platinised counter electrode, is a key factor behind the remarkable efficiencies achieved to date for DSSCs.

251

The flux of this recombination pathway increases with increasing electron density in the TiO2 electrode and therefore with the TiO2 Fermi level or cell voltage. In the dark, it responsible for the diode-like leakage current observed in current– voltage scans while under illumination it is the primary factor limiting the voltage output of the device. The kinetic competition between charge transport and recombination in DSSCs has been analysed in terms of an effective carrier diffusion length Ln, given by;

Ln = √(τDeff) where, Deff is the effective electron diffusion length, and τ the electron lifetime due to the charge-recombination reaction. Deff increases with light intensity due to the increased electron density in the TiO2 film whilst τ shows a proportional decrease resulting in Ln being largely independent of light intensity. Typical values for Ln are 5 – 20 µm, and even 100 µm near the optimum power point for cells with >10% conversion efficiency, consistent with the high carrier-collection efficiencies observed in efficient DSSCs.

It is important to emphasis that the energetics and kinetics of DSSC function are not independent considerations. The kinetics of the interfacial electron-transfer dynamics depend strongly on the energetics of the TiO2/dye/electrolyte interface and on the density of electrons in the TiO2. Raising the energy of the TiO2 conduction band reduces recombination losses as for a given TiO2 Fermi level, the electron density in the TiO2 film will be lower and therefore may give a high cell output voltage but at the expense of a lower free energy driving force for charge separation which may result in a lower quantum efficiency for charge generation and therefore a lower output current. In practice, modulation of these energetics and kinetics to achieve optimum device performance remains one of the key challenges in DSSC research and development.

252

Free energy driving force for charge separation

DSSC research and development Two prototypes of the monolithic Z-type interconnected DSSC modules has already been fabricated by Aisin Seiki in Japan using carbon as a back contact to cut costs. Comparative field tests of these modules and polycrystalline silicon (pcSi) have been running for several years. The test results revealed the advantages of the DSSC as compared with silicon modules under realistic out-door conditions for equal rating under standard test conditions (STC), the DSSC modules produced 20 – 30 % more energy than pc-Si modules. The superior performance of the DSSC can be ascribed to the following factors; The DSSC efficiency is practically temperature-independent in the range 25 – 65 0C while that of monocrystalline and pc-Si declines by about 20 % over the same range. Out-door measurements indicate that light capture by the DSSC is less sensitive to the angle of incidence, although this needs to be further assessed. The DSSC is more efficient than pc-Si in diffuse light or cloudy conditions

253

Device structure of a conductive polymer (MDMO-PPV)/ fullerene ([6,6]-PCBM) blend solar cell . While it is up to the commercial supplier to set the final price for such modules, it is clear that the DSSC shares the cost advantages of all thin-film devices. In addition it uses no high-vacuum technique and hence less cost-intensive steps and only cheap and readily available materials. Although it might be thought that the ruthenium-based sensitiser adds high material cost, its contribution to their costs is relatively low because of the small amounts used. Also purely organic sensitisers can now achieve practically the same efficiency as Ru complexes. Given these advantages at comparable conversion efficiency, module costs are realistically lower even for production plants having well below GW capacity. The DSSC has thus become a viable contender for large-scale solar energy conversion systems on the basis of cost, efficiency, stability and availability as well as environmental compatibility. Building-integrated DSSC panels have been installed in the walls of the Toyota ‘Dream House’ and the British company G24i has recently announced the building of the first 20 MW DSSC manufacturing plant in Wales.

Panchromatic sensitisers The ideal sensitiser for a single-junction photovoltaic cell converting standard global AM1.5 sunlight to electricity should absorb all light below a threshold 254

wavelength of about 920 nm. In addition, it must also carry attachment groups such as carboxylate or phosphonate to firmly graft it to the semiconductor oxide surface. On excitation, it should inject electrons into the solid with a quantum yield of unity. The energy level of the excited state should be well matched to the lower bound of the conduction band of the oxide to minimise energetic losses during the electron-transfer reaction.

Its redox potential should be sufficiently high that it can be regenerated through electron donation from the redox electrolyte or the holes conductor. Finally, it should be stable enough to sustain about 100 million turnover cycles corresponding to about twenty years of exposure to natural light. A singlejunction device with such a sensitizer could reach a maximum conversion efficiency of 32 % in global AM 1.5 sunlight.

Much of the research in DSSC dye chemistry is devoted to the identification and synthesis of sensitisers matching these requirements while retaining stability in the photo-electrochemical environment. The attachment group of the dye ensures that it spontaneously assembles as a molecular layer on exposing the oxide film to a dye solution. This molecular dispersion ensures a high probability that once a photon is absorbed, the excited state of the dye molecule will relax by electron injection into the semiconductor conduction band.

However, the optical absorption of a single monolayer of dye is weak, a fact which originally was cited as ruling out the possibility of high-efficiency sensitised devices as it was assumed that smooth substrate surfaces would be imperative in order to avoid the recombination loss mechanism associated with rough or polycrystalline structures in solid-state photovoltaics. This objection was invalidated by recognising that the injection process places electrons in the semiconductor lattice, spatially separated from the positive charge carriers by the dye molecules which are insulating in the ground state and hence provide a barrier

255

for charge recombination. By now, the use of nanocrystalline thin-film structures with a roughness factor of more than 1000 has become standard practice.

The best photovoltaic performance in terms of both conversion yield and long term stability has so far been achieved with polypyridyl complexes of ruthenium and osmium. Sensitisers having the general structure ML2(X)2 where L stands for 2,2’- bipyridyl-4,4’-dicarboxylic acid, M is Ru or Os, and X represents a halide, cyanide, thiocyanate, acetyl acetonate, thiocarbamate or water substituent are particularly promising. Thus, the ruthenium complex cis-RuL2(NCS)2, known as N3 dye has become the paradigm heterogeneous charge-transfer sensitiser for mesoporous solar cells.

The absorption spectrum of fully protonated N3 has maxima at 518 and 380 nm with extinction coefficients of 1.3 × 104 M–1 cm–1 and 1.33 × 104 M–1 cm–1 respectively. The complex emits at 750 nm while the excited-state lifetime being 60 ns. The optical transition has metal-to-ligand charge transfer (MLCT) characters, excitation of the dye involves transfer of an electron from the metal to the orbital of the surface-anchoring carboxylated bipyridyl ligand from where it is released in a time-scale of femtoseconds to picoseconds into the conduction band of TiO2 generating electric charges with unit quantum yield.

Spectral response of dye-sensitized solar cell for different dyes compared with the spectral response of bare TiO2 electrode 256

Discovered in 1993, the photovoltaic performance of N3 has been unmatched for eight years by virtually hundreds of other complexes that have been synthesised and tested. However, in 2001 the ‘black dye’ tri(cyanato)-2,2’2”-terpyridyl4,4’4”- tricarboxylate)Ru(II) achieved an efficiency of 10.4 % at AM1.5 solar-topower conversion efficiency in full sunlight. Conversion efficiencies have meanwhile been improved further and the current record validated by an accredited laboratory is 11.1 %. The overall maximum-power efficiency (ȵmp) of the cell is calculated from the integral shortcircuit photocurrent density (Isc), the open-circuit photovoltage (Voc), the fill factor (ff) and the incident solar irradiance (Eos = 1000 W m–2) as;

ȵ mp = Isc × Voc × ȵ fill /Eos

Performance improvement on DSSC cells The solar-to-electric power conversion efficiency of DSSC laboratory cells validated by an accredited PV calibration laboratory has reached 11.1 % under standard reporting conditions (i.e. air mass 1.5 global sunlight at 1000 W m–2 intensity and 298 K temperature), rendering the DSSC a credible alternative to conventional p-n junction photovoltaic devices. A couple of years ago, solid-state equivalents using organic hole conductors reached an efficiency of only 4.2 %, and recently at 5.1 %, whereas nanocomposite films comprising only inorganic materials such as TiO2 and CuInS2, have achieved efficiencies between 5% and 6% in the ETA cells.

Judicious molecular engineering of the ruthenium dye structure will now allow further increase of light harvesting in the 700 – 900 nm regions. Ruthenium complexes of quaterpyridyl derivatives give great promise in this respect. The goal is to obtain a DSSC having optical features similar to those of GaAs. A nearly vertical rise of the photocurrent close to the 920 nm absorption threshold would increase the short-circuit photocurrent from currently 20.5 mA cm–2 to

257

about 28 mA cm–2, raising the overall efficiency to over the 15 % mark without changing the currently used redox system.

Ultimately the combination of two sensitisers, one being the red or black ruthenium complex and the other an organic dye that show strong absorption in the near-infrared region, may be required to achieve this desired spectral response. Such dye ‘cocktails’ are presently under intense investigation and the first results look promising. A road map to achieve this goal by the year 2009 had been elaborated and was serve to coordinate synthetic efforts of several groups on the international scale.

Tandem cells A feature that makes the DSSC particularly attractive for tandem cell application is that its optical transmission and short-circuit photocurrent can be readily adjusted by changing the film thickness, pore size, the nature of the dye and the dye loading. This, along with the ease of forming layered structures, for example by producing the mesoscopic oxide films using screen printing or doctor blading methods renders the DSSC particularly well suited for the fabrication of tandem solar cell structures that capture the solar emission in an optimal fashion.

Note that for a two-level tandem cell a conversion are around 1.65 eV and 1 eV respectively. It has been recently demonstrated that a tandem device comprising a DSSC as a top cell for high-energy photons and a copper indium gallium selenide (CIGS) thin-film bottom cell capturing the red and near-IR solar emission respectively produces AM1.5 solar-to-electric conversion efficiencies greater than 15 %.

258

Band gap changes in tendem cells

The performance of the tandem cells has been clearly superior to that of some individual cells despite the fact that the short-circuit currents of any two cells compared are not perfectly matching. Likewise, no effort yet has been made to minimize the optical losses. This leaves no doubt that further rapid efficiency gains reaching well beyond the 20 % mark can be expected from the fructuous marriage of these two thin-film PV technologies. Combining a relatively low-cost thin film CIGS substrate cell with a DSSC superstrate cell may be a cheaper method of achieving efficiencies above 15 % than the use of a high-efficiency CIGS cell alone.

Band gap transitions in tendem cells

Stability Unlike amorphous silicon which suffers from degradation due to the well-known Staebler–Wronski effect, the intrinsic stability of the DSSC has been confirmed 259

by extensive accelerated light-soaking tests carried out over the last decade. One major issue that has been settled during this period is that the sensitisers employed in the current DSSC embodiments can sustain twenty years of out-door service without significant degradation. However, as new and more advanced dye structures emerge and in order to avoid repeating these lengthy tests every time the sensitiser is modified, kinetic criteria have been developed to allow prediction of long-term sensitiser performance.

Critical for stability are side reactions that occur from the excited state or the oxidised state of the dye which would compete with electron injection from the excited dye into the conduction band of the mesoscopic oxide and with the regeneration of the sensitiser. These destructive channels are assumed to follow first or pseudo-first order kinetics and are assigned the rate constants of k1 and k2. Introducing the two branching ratios, we have; P1 = kinj/(k1 + kinj) and P2 = kreg/(k2 + kreg) where, kinj and kreg are the first order or pseudo-first order rate constants for the injection and regeneration process respectively. The fraction of the sensitiser molecules that survives one cycle is given by the product P1 × P2. A simple calculation shows that the sum of the branching ratios for the two bleeding channels should not exceed 1 × 10–8 in order for the lifetime of the sensitiser to reach at least twenty years. The turn-over frequency of the dye in the working DSSC averaged diurnally and seasonally at about 0.16 s–1.

For most of the common sensitisers, the rate constant for electron injection from the excited state of the dye to the conduction band of the TiO2 particles is in the femtosecond range. Assuming kinj = 1011 –1013 s–1, any destructive side reaction should have about k1 < 102 s–1. Ruthenium sensitisers of the N719 or K-19 type readily satisfy this condition as the decomposition from the excited-state level occurs at a much lower rate than the 102 s–1 limit. Precise kinetic information 260

gathered for the second destructive channel involving the oxidised state of the sensitiser, the key parameter being the ratio k2/kreg of the rate constants for the degradation of the oxidised form of the sensitiser and its regeneration are now available. The S+ (oxidized) state of the sensitiser can readily be produced by chemical or electrochemical oxidation and its lifetime can be independently determined by absorption spectroscopy. A typical value of k2 is around 10–4 s–1 while the regeneration rate constant is at least in the 105 s–1 range, hence the branching ratio is well below the limit of 10–8 which can be tolerated to achieve the 100 million turn-overs and a 20-year lifetime for the sensitiser.

Many long-term tests which have been performed with the N3-type ruthenium complexes have confirmed the extraordinary stability of these charge-transfer sensitisers. For example, a European consortium supported by the Joule program confirmed cell photocurrent stability during 8,500 hours of light soaking at 2.5 Suns, corresponding to about 56 million turnovers of the dye without any significant degradation. These results corroborate the projections from the kinetic considerations made above. A more difficult task has been to achieve stability under prolonged stress at higher temperatures, i.e. 80 – 85 C. The introduction of hydrophobic sensitisers has been particularly rewarding in allowing the DSSC to meet the specifications laid out for outdoor applications of silicon photovoltaic cells.

In addition, these dyes show enhanced extinction coefficients due to the extension of the conjugation onto one of the bipy ligands by styrene moieties. Taking advantage of these properties and using a novel robust electrolyte formulation, an 8% efficient DSSC has been realised that show strikingly stable performance under both prolonged thermal stress and light soaking. While impressive progress has been made in the development of stable nonvolatile electrolyte formulations, the conversion yields obtained with these systems are presently in the 7 – 10 % 261

range, below the 11.1 % reached with volatile solvents. Future research efforts will be dedicated to bridge the performance gap between these systems. The focus will be on holes conductors and solvent-free electrolytes such as ionic liquids. The latter are a particularly attractive choice for the first commercial modules owing to their high stability, negligible vapour pressure and excellent compatibility with the environment.

Organic dyes When considering organic dyes for use in DSSCs, materilas such as porphyrins and phthalocyanines have attracted particular attention where porphyrins has lead because of the analogy with natural photosynthetic processes and the latter because of their photochemical and phototherapeutic applications. Porphyrins cannot compete with the N3 or black dye sensitiser due to their lack of red light and near-IR absorption. Phthalocyanines on the other hand show intense absorption bands in this spectral region, however, problems with aggregation and the unsuitable energetic position of the LUMO level which is too low for electron transfer to the TiO2 conduction band have turned out to be intractable for the moment.

The use of coumarin or polyene-type sensitisers show strikingly high solar-toelectric power conversion efficiencies reaching up to 9.2 % in full sunlight have been achieved in liquid-electrolyte cells. The rational design of such sensitisers based on quantum mechanical considerations has made great progress during the last few years and replacement of ruthenium sensitisers by such molecularly engineered dyes is likely to occur in the near future.

262

Organic semiconductors categorized according to their mechanical or processing properties as insoluble, soluble or liquid crystalline

Quantum dots as sensitisers Semiconductor quantum dots are another attractive option for panchromatic sensitisers. These are II–VI and III–V type semiconductors particles whose size is small enough to produce quantum confinement effects. The absorption spectrum of such quantum dots can be adjusted by changing the particle size. Thus, the band-gap of materials such as InAs and PbS can be adapted to attain the value of 1.35 eV which is ideal for a single-junction solar quantum converter.

During the last decade, a wealth of information has been gathered on the physical properties of quantum-dot materials and research is being pursued very actively. One problem with this approach is the photocorrosion of the quantum dots that will almost certainly happen if the junction contact is a liquid redox electrolyte. However, quantum dots are expected to display higher stability in solid state heterojunction devices. The advantage of Q-dots over conventional dyes as sensitisers is that their very high extinction coefficients allow thinner films of the mesoporous oxide to be used.

This reduces the dark current, increasing Voc and the overall efficiency of the cell. An exciting discovery made in recent years is that multiple exciton generation 263

(MEG) can be obtained from the absorption of a single photon by a quantum dot if the photon energy is at least two times higher than its band-gap. The challenge is now to find ways to collect the excitons before they recombine.

Energy diagram of a single layer conjugated polymerphotovoltaic, device under short circuit conditions. VB valence band, CB conduction band, Eg and gap energy, P+, P-, positive and negative polarons respectively

As recombination occurs on a picosecond time scale, the use of mesoporous oxides as electron collectors presents a promising strategy because the electron transfer from the quantum dot to the conduction band of the oxide electrode occurs within femtoseconds. This opens up research avenues that ultimately may lead to photon converters reaching external quantum efficiencies values of several hundred percent. A calculation based on Henry’s model shows that the maximum conversion efficiency of a single-junction cell could be increased from 34 % to 44 % by exploiting MEG effects.

Mesoporous oxide film development When the dye-sensitised nanocrystalline solar cell was first presented, perhaps the most puzzling phenomenon was the highly efficient charge transport through the 264

nanocrystalline TiO2 layer. Mesoporous electrodes differ greatly from their compact analogs because; the inherent conductivity of the film is very low; the small size of the nanocrystalline particles which do not support a built-in electrical field; and the electrolyte penetrates the porous film all the way to the back contact making the semiconductor/electrolyte interface essentially threedimensional.

The mechanism of charge transport in mesoporous systems is under keen debate today and several interpretations based on the Montrol–Scher model for random displacement of charge carriers in disordered solids have been advanced. However, the ‘effective’ electron diffusion coefficient is expected to depend on a number of factors such as trap filling and space-charge compensation by ionic motion in the electrolyte. The theoretical and experimental effort will continue as there is a need for further in-depth analysis of this intriguing charge-percolation process. The factors controlling the rate of charge-carrier percolation across the nanocrystalline film are presently under intense scrutiny.

Intensity-modulated impedance spectroscopy has proved to be an elegant and powerful tool to address these and other important questions related to the characteristic time constants for charge-carrier transport and reaction dynamics in dye-sensitised nanocrystalline solar cells. On the side material science, future research will be directed towards synthesizing structures with a higher degree of order than a random assembly of nanoparticles. A desirable morphology of the films would have the mesoporous channels or nanorods aligned in parallel to each other and vertically with respect to the TCO glass current collector.

This would facilitate pore diffusion, give easier access to the film surface, avoid grain boundaries and allow the junction to be formed under better control. One approach to fabricate such oxide structures is based on surfactant templateassisted preparation of TiO2 nanotubes. These and the hybrid nanorod–polymer

265

composite cells as developed have confirmed the superior photovoltaic performance of such films as compared with random-particle networks.

Molecular engineering of the interface The high contact area of the junction in nanocrystalline solar cells renders mandatory the grasp and control of interfacial effects for future improvement of cell performance. The nature of the exposed surface planes of the oxide and the mode of interaction with the dye is the first important information to gather. The prevalent orientation of the anatase surface planes is (101) and the sensitiser is adsorbed through two of the four carboxylate groups, at least one of them being anchored through a bidentate configuration bridging two adjacent titanium sites. Molecular dynamic demonestrations employing a classical force field and charge excitement have been carried out to predict the equilibrium geometry of the adsorbed sensitiser state.

Schematic presentation of the different organic photovoltaic device architectures A) Single layer, b) donor-acceptor bilayer, c) donor-acceptor bulk heterojunction (or blend) and d) laminated donor-acceptor structure

More sophisticated first principle density functional calculations have also been undertaken to model the surface interactions of TiO2 with simple adsorbates as well as the surface reconstruction effects resulting from the adsorption. The latter approach is particularly promising and will provide an important tool for future theoretical investigations. Synthetic efforts focus on the molecular engineering of 266

sensitisers that enhance charge separation at the oxide solution interface. The structural features of the dye should match the requirements for current rectification i.e by analogy to the photofield effect in transistor the gate for unidirectional electron flow from the electrolyte through the junction and into the oxide is opened by the photo-excitation of the sensitiser.

Current-voltage (I-V) curves of raspberry dye-sensitized solar cells

The reverse charge flow, i.e. re-capture of the electron by the electrolyte could be impaired by judicious design of the sensitiser. The latter should form a tightly packed insulating monolayer blocking the dark current. The gain in open circuit voltage can be calculated from the diode equation

Voc = (nRT/F) ln[Isc/Io) –1] where n is the ideality factor, whose value is between 1 and 2 for DSSCs and Io is the reverse saturation current.. Thus, for each order of magnitude decrease in the dark current, the gain in Voc would be 59 mV at room temperature. Work in this direction is indispensable to raise the efficiency of the DSSC significantly over the 15 % limit with the currently employed redox electrolytes.

267

Solid-state dye-sensitised cells Research on solid-state DSSCs has gained considerable momentum in recent years as this embodiment is attractive for realising flexible PV cells in a roll-toroll production. The most successful p-type organic-conductor employed to date is spiro-OMeTAD with a work function of about 4.9 eV and hole mobility of 2 × 10–4 cm2 s–1. First reported in 1998, conversion efficiencies of solid-state cells incorporating this hole conductor have increased dramatically over the last few years from a fraction of a percent to over 4 %. The main drawback of these cells has been the fast interfacial electron-hole recombination reducing the electron diffusion length to a few microns as compared with 20 – 100 μm for electrolytebased DSSCs.

Examples of the molecular structures of different organic pigments used in efficient organic solar cells

268

Common conjugated polymers and macromolecules used in organic photovoltaics The dye monolayer can block this back reaction to some extent because it is electrically insulating. Hence, current efforts are directed towards molecular engineering of the interface to improve the compactness and order of the monolayer and prevent charge carriers from recombining. Another difficulty that has been encountered is the filling of the porous network with the holes conductors. This impediment may be overcome by developing oxide films having regular mesoporous channels aligned perpendicular to the current collector. On the other hand the Voc values obtained with solid-state DSSCs are high reaching nearly 1 V owing to a better match of the hole conductor work function than that of the electrolyte with the redo- potential of the sensitiser. The future of these solid hole-conductor systems thus, looks very bright if the recombination and pore filling problems can be solved.

Solar Cell Modeling History The photovoltaic community has demonstrated and proposed a wide variety of solar cell modeling structures using a wide range of photovoltaic semiconductor materials. Numerical modeling has proved to be a valuable tool in understanding the operation of these devices. There are several numerical solar cell simulation 269

programs in use. The first solar cell program was developed by Mark S. Lundstrom as part of his PhD Thesis. Other programs developed at Purdue University at later times include; Thin-Film Semiconductor Simulation Program (TFSSP), Solar Cell Analysis Program in 1 Dimension (SCAPlD), Solar Cell Analysis Program in 2 Dimension (SCAP2D), PUPHS, and PUPHS2D

These have been used to model a number of solar cells - thin-film cells (Si:H, CdS/CIS, CdS/CdTe, Si, Ge, & GaAs) in one spatial dimension and of high efficiency Si and GaAs solar cells in two-dimensions. One-dimensional simulations are adequate for conventional geometry solar cells especially at low solar intensities and for semiconductor materials that are not well characterized. At high intensities, 2D effects can become important even in conventional geometry solar cells and many high efficiency cell designs require 2D simulations or even 3D simulations. The interdigitated back contact solar cell is an example of a 2D geometry and the point contact solar cell is an example of an inherently 3D geometry. While the basic approach to modeling any of these devices is essentially the same, special purpose codes have been developed for each material. This makes modification tedious since many different codes must be updated and tested each time.

A Device Emulation Program and Toolbox (ADEPT) has been developed to address this problem by unifying the common components of all these codes. In addition, ADEPT was developed to be a tool to examine novel materials and device structures. Today, there are numerous solar cell programs developed by researchers from all over the world and there are also commercial simulation tools that can do solar cells modeling. Among them, the best programs are Silvaco and Crosslight. Synopsis has announced that it also has some solar cell simulation capabilities. In SILVACO, TFT is an advanced device technology simulator equipped with the physical models and specialized numerical techniques required to simulate amorphous or polysilicon devices including thin film transistors.

270

Specialized applications include the large area display electronics such as Flat Panel Displays (FPDs) and solar cells.

In Cross light, Advanced Physical Models of Semiconductor Devices (APSYS) is based on 2D/3D finite element analysis of electrical, optical and thermal properties of compound and silicon semiconductor devices. Emphasis has been placed on band structure engineering and quantum mechanical effects. Inclusion of various optical modules also makes this simulation package attractive for applications involving photosensitive or light emitting devices and solar cells. For Si rear-contacted cells (RCC) with textured front surface, RT techniques are utilized to compute the enhanced optic absorption. Conversion efficiency could be improved with about 20.7 % percent for certain textured devices and good agreement with the experimental can be obtained. Other Si cells like passivated emitter, rear totally diffused (PERT), and passivated emitter, rear locally diffused (PERL) cells can also be modeled with APSYS.

Drift Diffusion Model The semi classical transport of charges can be explained using Boltzmann Transport equation (BTE), however, the direct analytical solution of the BTE is difficult combined with the field solvers for device simulation. Therefore the predominant model providing solutions for the Drift Diffusion equations is used for traditional semiconductor device modeling. In this model the electric fields and spatial gradient of the carrier density is localized i.e. the current at a particular point only depends on the instantaneous electric field and concentration gradient of carriers at that point. The drift diffusion equations can be obtained by solving the Boltzmann Transport equation by solving for the moments of this equation. For steady-state and 1D geometry, the use of relaxation time approximation for the BTE results in the electron and hole carrier concentrations, thus have to be updated for every iteration.

271

Using a principle derived from natural photosynthesis, mesoscopic injection solar cells and in particular the DSSC have become a credible alternative to solid-state p-n junction devices. Conversion efficiencies over 11 % and 15 % have already been obtained with single-junction and tandem liquid-electrolyte cells respectively on the laboratory scale but there is ample room for further development. Future research is expected to focus on improving the short-circuit current by extending the light response of the sensitisers in the near-IR spectral region. Substantial gains in the open-circuit voltage are expected from introducing ordered oxide mesostructures and controlling the interfacial charge recombination by judicious engineering on the molecular level.

Hybrid cells based on solid-state inorganic and organic holes conductors are an attractive option and in particular for the flexible DSSC embodiment. Nanostructured ETA cells using purely inorganic components will also be developed. Mesoscopic dye-sensitised cells are well suited for a whole realm of applications ranging from the low-power market to large-scale applications. Their excellent performance in diffuse light gives them a competitive edge over silicon in providing electric power for stand-alone electronic equipment both in-doors and out-doors. DSSCs are already being applied in building-integrated PV and this will become a fertile field of future commercial development.

272

273

274

275