High Efficiency Solar Cells

[edit] Introduction

"The use of solar energy has not been opened up because the oil industry does not own the sun." - Ralph Nader, American Presidency Candidate.^{[verification needed]}

This trend of "taking the easy way out" and using oil and fossil fuels is a trend that is coming to an end. The world is becoming more conscious of the environment, and more and more people are turning to solar energy.

The technology of solar power and solar cells has come a long ways in a short amount of time. We have gone from using 1% efficient basic solar cells to advanced photovoltaic cells, capable of exceeding 40% efficiency. Such advances in technology have set the solar power industry abuzz, and it is unlikely that the technology will stop improving any time soon.

[edit] How Does It Work?

How do solar cells work?

A solar cell can also be referred to as a "photovoltaic." A photovoltaic cell is essentially a group of cells wired together for increased efficiency. Light shines on the photovoltaic, and the photons from this light are absorbed by a semi-conducting material. A variety of materials can be used for this, but silicon is the most common choice for today's photovoltaic cells.

When the photons are absorbed by the semi-conducting material, so is their energy. The energy being transferred into the semi-conducting material causes electrons to be knocked loose.

To use these now free electrons, an electric field is usually used to force the electrons in a certain direction. A flow of electrons is in reality, just a current, since current is defined as charge/second. This flow of electrons can then be drawn off the photovoltaic in the form of a DC current source.

Although this may seem like a reasonably straight-forward process, there is much room for improvement with regards to materials processing. The more efficient we can make solar cells, the cleaner the world will be.

^[1]

[edit] Semi-Conducting

A material is said to be a semiconductor^W if it is neither a good insulator, nor a good conductor, but something in between. When looking at a periodic table, the metals (on the left of the table) are naturally good electrical conductors. Non-metals (ceramics, polymers, etc.) which are found on the right side of the periodic table, are poor conductors, but good insulators, meaning they act in a reverse manner to the metals. Semiconducting materials fall in neither of these categories, and can be found in the middle of the periodic table.

The two most notable semiconducting materials are Germanium and Silicon. The reason both of these materials are good semiconducters is related to each substance's number of free electrons. Both Germanium and Silicon have four free electrons for each atom. That means that they will form covalent bonds with nearby atoms in a lattice pattern. This lattice causes a very strong crystal to be formed, with zero valence electrons, meaning that Silicon and Germanium are naturally more similar to an insulator than a conductor. The conductivity of the material can be varied by doping it, but this will be discussed later.

[edit] P-type semiconductor^W

As can be seen in the picture above, most solar cells actually use a combination of a p-semiconductor and an n-semiconductor. A p-type semiconductor is a semiconductor that has been doped positively, hence the "p". This means that atoms have been added to this material to cause the material to be positive. This seems rather counter-intuitive since when an atom accepts an electron it becomes negatively charged, not positively charged. However, when doping a material to form a p-type semi-conductor, the dopant accept electrons from the semi-conducting material, causing the dopant atoms to become negatively charged and forming holes where the electrons used to be. These holes are positively charged, and when enough holes are formed, the material behaves similar to a quantitiy of positive charge, thus forming a positive-type semi-conductor.

[edit] N-type semiconductor^W

N-type semi-conductors are formed in the same way as p-type semiconductors, the only difference being that a dopant is chosen which donates electrons to the semi-conducting material. This makes the dopant atoms positively charged and the holes negatively charged. As the number of holes increases the material begins to act like a quantity of negative charge. Although both p-type and n-type semiconductors are created using a similar process, the dopant used determines the charge of the material.

The reason p-type and n-type semiconductors are used, instead of one single type of semiconductor, is that by creating some positive and some negative charge, the flow of electrons created by the solar cell can be directed in a single direction using the fields generated by the p-type and n-type's opposing charges.

[edit] Solar Cell Manufacturing Process

The solar cell manufacturing process can be split into three different stages: 1. growing or casting the silicon crystal, 2. solar cell manufacturing, and 3. assembling and installing the solar cell system. ^[2]

[edit] Growing/Casting Silicon Crystals

The first step in making solar cells is to grow or cast the silicon crystals needed for the semiconducting layers (or wafers) that are inside of each solar cell. This is one of the most energy intensive steps of the solar cell manufacturing process, due to the fact that it involves processing molten silicon at approximately 1600 degrees Celsius.

The most common way of creating semiconductor quality silicon wafers is the Czochralski process^W. This process involves melting high-grade silicon in a Crucible^W, and dipping a Seed crystal^W on a rod into the melt. When placed into the super-saturated silicon solution the seed crystal will cause the silicon crystal to start forming on the rod. The rod is then rotated and slowing lifted out of the crucible, allowing the silicon crystal to form along the rod, as seen below. This is also where the dopants are added to the silicon, to make them either p-type or n-type.

^[3]

Although this is a commonly used process, there is one problem with it. When the molten silicon is poured into the crucible (which is usually made of quartz), some of the quartz dissolves into the silicon, causing impurities. A more recently developed method involes casting the silicon in reusable graphite moulds, forming blocks of silicon, from which the wafers are cut. However, these wafers are multicrystalline and not as desirable as the monocrystalline wafer produced by the Czochralski process. Multicrystalline wafers have lower efficiency than monocrystalline wafers because the lattice formed is not as uniform.

So, the Czochralski process would be the far superior method if only the impurities from the dissolving of the quartz crucible could be eliminated.

[edit] Modified Czochralski Process

I have developed a modified version of the Czochralski process that will eliminate the impurities occuring from the quartz crucible. Instead of using a quartz crucible, a crucible is made by hollowing out an ingot of silicon produced by the regular Czochralski process. This ingot would have relatively thick walls and bottom. Then, the high-grade silicon would be poured into the hollowed out ingot. By allowing the molten silicon to cool quickly, and by spinning the rod with the seed crystal very quickly, a second ingot could be produced. If this is done quickly enough, only a small amount of the hollow, crucible-ingot would melt off into the new ingot. And that which does melt off would be composed of the exact same material, so impurities would not occur like they would with a quartz crucible. The same crucible-ingot could continue to be used as long as the walls and bottom remained thick enough. Another great perk of this process, is that once the ingot-crucible becomes too thin, it can simply be melted down and mixed with the other molten silicon, meaning that absolutely no silicon is wasted, making this an incredibly efficient process.

Let's analyze this with some numbers. After the normal Czochralski process has been done, the silicon usually contains about 10^18 atoms of oxygen/cm^3. Since the new crucible is made out of this silicon, this will be the surface concentration of oxygen. The following equation is then used to determine the concentration of oxygen atoms in the new silicon ingot:

  (C - Cs)/(Co - Cs) = erf(x/(2*sqrt(D*t)))

It will be assumed that the molten silicon contains no oxygen at all, so the Co terms disappears, and this equation can be rearranged to be:

  C = Cs - Cs*erf(x/(2*sqrt(D*t)))

Where x is the distance from the surface, D is the diffusion coefficient and t is the amount of time over which the diffusion takes place. We already know that Cs = 10^18/cm^3 and D was found to be 3.6 x 10^-4 um^2/s^[4] and the time it takes to grow a half metre long ingot is 20 hours^[5]. Having this information, a plot can be made using the above equation of concentration of oxygen (C) versus distance from the surface (x). This plot can be seen below.

As can be seen above, the concentration of oxygen drops off very quickly, the further we go into the silicon ingot. From this plot, there will not be any oxygen past a depth of 20 micrometres. If the sample is approximately a third of a metre in diameter, only a very tiny fraction of the sample will actually contain the impurities from the oxygen. Also, to completely erase the oxygen impurities, a layer of about 20 micrometres could be shaved off the silicon ingot, leaving us with an incredibly pure silicon ingot. Although some of these numbers were general approximations, this process does seem to give very reasonable results, and could very well be used in industry.

[edit] Manufacturing the Solar Cell

The manufacturing of the actual solar cells is a complex process. The silicon wafers produced must undergo a rigorous semiconducting treatment, including etching, diffusion and screen-printing, so that they have the right texture to become operational solar cells.

The assembly of the actual parts that each solar cell consists of is often done at these plants as well, or smaller, assembly specific plants.

These plants are incredibly expensive to build and maintain. The largest plants have a capacity of 50 MW, a floor space of over 50000 square feet and can cost in excess of 100 million US dollars to build, nevermind the upkeep costs. Seeing as how most of the plants operate similarily, I think that only one single solar cell manufacturing plant is needed in today's world. By consolidating all of the smaller ones into one massive plant, the manufacturing costs would go down. Manufacturing plants are already shipping their product all over the world, so nothing would change there. Also, by having only a single plant, an industry standard for size, cost, efficiency etc. would be set. Meaning that companies could design their products around very specific specifications, because they would know exactly what they are getting.

[edit] Assembling the System

Very rarely is a single solar cell used on its own. Solar panels, for example, consist of many solar cells wired together. This step involves mounting or attaching the solar cells to the desired device, and then wiring them all together for maximum efficiency. Also, the solar cells need to be positioned so that they recieve the most sunlight, and therefore generate the most power.

An important factor in this step is configuring how much sunlight the solar cells recieve. Obviously, the most sunlight that the cells recieve, the more power they will generate. Solar cells that are place near the equator will generate more power because they have more exposure to the sun, and the sun is stronger at the equator. Much research has been done with regards to rotating or tilting the solar cells, so that they face the sun for a longer period of time. However, this is only useful if it takes less energy to turn the cells, than the extra energy that the cells will produce because they have more exposure to the sun.

[edit] Anti-Reflective Coating

All commercially used solar cells have an anti-reflective coating on them. Untreated, with just a glass coating, the silicon cell will only absorb about 67.4% of the sunlight, with the rest being reflected away. This sunlight being reflected away is a large waste of energy, so anti-reflective coatings are placed over the silicon solar cells to increase the amount of sunlight absorbed by the cell thereby increasing the energy efficiency of the cell. ^[6]

Anti-reflective coatings are typically extremely thin, in the range of micrometers or some times even nanometers. A basic sketch of how they work can be seen below:

For an anti-reflective coating to be effective, it is desirable for all the light to be captured. To do so, the reflections inside the anti-reflective coating shoul be cancelled out by the following incident light waves.

Light behaves like a wave. Waves have a period, or a time after which they repeat themselves. If one wave is added to another, which is out of phase from the second one by half a period, then the waves will cancel completely, stopping the reflected light from escaping. Since the wave has to travel across the anti-reflective coating twice (once going down, once going up), and it is desired that the reflected wave is half a wavelength out of the phase from the original wave, then the ideal thickness of the coating is given by:

    Thickness = (1/4) x wavelength[7]

Also, the ideal refractive index of the coating is given by:

    n-coating = sqrt((n-glass) x (n-air))

This value is usually around 1.3. Often silicon monoxide is used for the anti-reflective coating because it has a refractive index of about 1.3, and large amounts of silicon are usually available because it was used to make the solar cells themselves.

[edit] Improvements to this Technology

Although reflective coatings do work, they will often only increase the amount of sunlight absorbed by about 10%. However, this might change in the near future.

[edit] Wavelengths

One challenge with capturing sunlight is that it comes in different wavelengths. Reflective coatings can be configured to work for small ranges of wevelength, but with great variance, wavelengths outside these ranges are lost. A way to fix this would be to convert all the different wavelengths of the light from the sun into one known wavelength. The reflective coating could be configured for this wavelength and all of the light would be absorbed.

Since wavelength = (speed of the wave)/(frequency), the wavelength of the light can be changed by changing the speed of the wave or the frequency. The frequency would be rather difficult to change, but the speed of the wave is dependent on the material at which it is travelling through, which can be easily changed. My idea is to separate the incident sunlight into its different wavelengths. Then the light of different wavelengths will each pass through a different material, in which the speed of the wave is different. Using the knowledge that wavelength = (speed of the wave)/(frequency) it can be configured that after each wavelength of light, after entering its respective material, has the same wavelength. All of this light could then enter a solar cell with an anti-reflective coating configured for exactly this known wavelength, and all of the light will be absorbed, without exception. It will look something like this:

Physicists in Germany have recently had some preliminary success in idea. From sunlight, they were able to filter out all but the wavelengths at which green light forms, and then convert this light into blue light by essentially increasing the energy of an incoming photon above its ground state, then returning it to its ground state while emitting a single, high frequency photon, of different colour. This technology, so far, has only found to be about 1% efficient. That being said, we are only using a portion of the sun's spectrum of wavelength's, so any efficiency above 0% would be a gain.^[8]

[edit] References

1. ↑ Future Students - What are Photovoltaic Devices, University of New South Whales, Accessed on Sunday, November 16th: http://www.pv.unsw.edu.au/future-students/pv-devices/how-they-work.asp.
2. ↑ Solar Cell Manufacturing Plants - Types of Plants, solarbuzzTM, Accessed on Sunday, November 16th: http://www.solarbuzz.com/plants.htm
3. ↑ Photovoltaic Materials - Silicon, The A to Z of Materials, Accessed on Sunday, November 16th: http://www.azom.com/details.asp?ArticleID=1169
4. ↑ Phase Transformations in Metals and Alloys, D.A. Porter and K.E. Easterling, Taylor and Francis Group, 2004, Page 79
5. ↑ theImage.com, "Czochralski Process", Accessed:Thursday, November 27th/2008 at:http://www.theimage.com/newgems/synthetic/syntheticanimate2.html
6. ↑ Michael Mullaney, "Solar Power Game-Changer: "Near Perfect" Absorption of Sunlight, From All Angles", Rensselaer, Accessed:Monday, November 17th/2008 at:http://news.rpi.edu/update.do?artcenterkey=2507
7. ↑ Why do coated lenses look purple?, Tripod, Accessed:Monday, November 17th/2008 at:http://rick_oleson.tripod.com/index-166.html
8. ↑ Tuning the Sun's Rays, Michael Schirber, Accessed:Monday, November 17th/2008 at:http://focus.aps.org/story/v18/st11