Lec38 6-5 太陽能電池製程技術

We now talk about the solar cell manufacturing process. Because this is very relevant with mechanical engineering. Let's look at how to create silicon wafers The first step is starting from sand. There are a lot of carbon and debris in sand. of course, we need to refine. We refine it to become a high level of silica sand. This level seems to be 96% to 98%. The proportion of carbon becomes less and the proportion of silicon reaches to 96% to 98%. So this is the raw material which we normally buy in silicon wafer plant, it is a block. Then that is a big furnace. The different methods would come up with the different effects. In the furnace, there are a lot of high-temperature heating film, the silicon block becomes melt. After it becomes melt, we use a variety of methods to produce a wafer. Silicon block is melted like casting. Of course, with traditional casting method, it cannot be defined as mono-crystalline. Mono-crystalline has to be produced by other method. But it is like this, after melting, it become an ingot. There are thermal flow. If we make it into p-type semiconductor or n-type semiconductor, we need to provide boron or phosphorus. This thing has been studied for a long time and it is very complex. But the entire manufacturing process, such as multi-crystalline is mechanical casting. Then how to maintain the quality of the interface and the temperature is related to the thermal flow calculation. Basically you'll find that the teacher who are working on wafer calculation are thermal flow experts. So you make the wafer becomes this type. The process of producing wafer is the first half work on the company. How do we produce multi-crystalline and mono-crystalline? There are many methods. I'm talking about the methods which are commonly used. For mono-crystalline, it is called Czochralski (Cz method). It is probably like this and it is pretty high. The furnace size is depend on the size of the ingot. 8 inches wafer is just like this and there will be seed crystal. This is a steel pot. You put all that stuff like silicon block in the steel pot. Here are heating plate. It is heated to become melt. At this time, you put a seed crystal in the melt to let it grow in the same direction. Then the system pulls it out slowly and cools it down with argon. Then this thing will spin to pull it out slowly and then cool it down with argon. After cooling, it becomes solid, so what you pull out is solid. After pulling out, this is the wafer. It depends on how large you want to pull it out. 4 inches, 8 inches and 12 inches are produced with the same method. It take at least 10 hours to make a wafer. So you will see that a lot of people inside the factory wear Clean room suit. Then you will observe the equipment there, there are a few machines to pull crystal. This is the machinery industry, is it? There are all thermal flow, then all machinery do. After it is pulled out, it is like this one with many columns. So at this time, the solid-liquid interface is important. Because the solid-liquid interface will affect some crystal arrangement, there are some dislocation. The study is more relevant to Department of Materials Research. Anyway, for us, in terms of the mechanical engineering department, we need to control the solid-liquid interface. You can do some simulation. This is what we simulate. It is the entire thermal flow field. And it is the solid-liquid interface. We are studying on how to control it. This is the melt, the melt will be less and less. This is the ingot which will become gradually high. It all depend on the length of ingot, temperature distribution and the solid-liquid interface. There are also some impurities such as oxygen which is released from the crucible wall. So it need to use the thermal flow field to control the oxygen distribution. Try to not make oxygen in the middle part. So you need to understand the natural convection inside. This is the mono-crystalline which not only is more expensive but also spends more time. This furnace would spend properly tens of millions and there are some furnaces in the factory. This is the view from the top. This part is the crucible. There is silicon melt with at least 1000 degrees. Then the bottom rotation pulled the seed crystal out. The seed crystal is just like this thing which is worth hundreds of thousands. Of course, it is also similar to semiconductor, but the silicon material of semiconductor has to be purer. The purity of the silicon material for the solar panels does not require to be as good as semiconductors. This is multi-crystalline and it is a big furnace. Multi-crystalline is just like casting. The large container on the package in this way. This is also a crucible. In the crucible, there are some heating films and some coolers. The different part is all melted originally and then open the insulation to cool down the system. As the insulation is gradually high, it will start to grow up. So the upper part would become liquid and the lower part would become solid, it is just not the same. If we watch this animation, it is probably like this. After the crucible opening, the lower part began to solidify, because the gas comes in. There is the cooler with low temperature. So you will see that it becomes taller and taller. The things are all closed, so the edge with impurities has to be cut out. It isn’t like mono-crystalline. This is a square ingot. It is totally different to traditional casting. Multi-crystalline is used for low level products. We talked about that the temperature field and the concentration field are both important. If you are making a p-n semiconductor, generally these are used p-type for solar energy. Then it would slowly get to the n-type with diffusion. So you just add some boron inside. Then there will be the boron concentration, the concentration of boron can also do some simulation. For example, in the process of growing, we could observe the boron concentration. In the start, it may be pulled in this way, we are now talking about is in the ingot which also means in the solid. Therefore, the concentration distribution can be calculated. The concentration here is higher and higher. It is called segregation effect. The concentration of the fluid inside the originally homogeneous, But when the fluid transfers into solid, the solidified speed of each part are not the same. Some part is faster and some part is slower. Due to slower solidification with some impurities, most would still remain in the melt. The concentration in the melt is getting higher, so higher concentration is accompanied with higher ingot. And then the boron also diffuse in the solid like heat transfer. Of course, maybe you have not learned mass transfer, as long as you have learned heat transfer, you will understand the mass transfer. In heat transfer, the heat diffusion equation become a mass diffusion equation and the temperature becomes concentration. Other parts are almost the same. Another one is the thin-film. We just talked about the ingot based. After the wafers grow, the wafers is a p- type semiconductor or n-type semiconductor. Then there is a thin-film with many layers on top. In general, then with the chemical vapor deposition (CVD), it is injecting something inside. When it contacts the bottom and create some chemical reactions, it will be deposited. The plate contains specific elements and the vapor inject with high temperature. When it touch the bottom plate, then it will produce some kind of chemical reaction and it is deposited with a layer of solid. It is called Chemical vapor deposition (CVD). This is a wafer. If you want to do a piece of wafer, these are various different ways. For example, there are several films and this is my bottom plate. When the gas came in, the high temperature will lead to chemical reaction This is the simulation and it will become like this. This is the flow field and temperature field. The gas came in from the place and then I hit went out from other place. And then the chemical reaction produces. So it can be calculated that how many things can be deposited on the substrate. Is this thing good? This is a variety of flow fields. For example, the gas came in vertically to touch it and then start the reaction. The other type are used with the gas coming in horizontally, and then it rotates to produce a chemical reaction. After we produce ingots, in general, these ingots are all p-type or n-type. If we want to produce a p-n diode, we need to use a diffusion method. This is another key point of the study. How do people mix the dopants of group V into solid? The surface produce an n-type thing. Generally, there are a lot of sample such as laser. In the manufacture procedure, there will be a diffusion process. Put another dopants inside, at this time, you will have there are the p-type dopants and n-type dopants. Based on the p-type dopants, you diffuse with n-type dopants. There will be p-type dopants and n-type dopants. In the procedure, there are a lot process like cleaning, but it will turn into a cell. The cell will turn into a module and then the module will turn into array. So in general, what we saw is an array. There are many module in the array, there are many cell in the module. This is the cell, module and array. So finally it all depends on how much an array is? These things have certain prices and specifications. How many cell can become a module? How many modules can become an array? After you bought it, you just need to put it on. We were just talking about the general producing method of solar energy. We say that if such solar energy, because it is all the thin-film type, mono-crystalline and multi-crystalline. The efficiency are different. Mono-crystalline is the best one with the efficiency at properly 20%. We use a lot of ways to increase the efficiency. This is the concentrated solar photovoltaic system. Basically, it is just like solar thermal energy. We concentrate heat. Basically, you will see that there will be a large curved mirror in this place. The curved mirror will make it focus on this small mirror. And then it is reflected back again by this small mirror. So in this small area, it have very strong sunlight. The benefit is that it just need this little area for the solar panels. We can spend more money to make the equipment more precise. Previously, people have demand for lower cost. Because it is not very much, it can be more precise here. This is the Concentrated PV System (CPV). Its advantage is that you can make a lot of layers and each layer can be aimed at different kinds of light. Because the visible light from solar energy is the biggest, Basically we use different layers to absorb different kinds of light. Unlike the previous one with only one layer, so it is called multi-junction. What is multi-junction? That is the system with many layers and many joints. This is a typical one. The first junction absorbs blue light. The second junction absorbs green light. The third junction absorbs red light. Then it can collect the three kinds of light together. Of course, I talked that this practice would be more expensive in the manufacture procedure. But because it is small area, maybe we can recover the lost. Therefore, with the practices, the efficiency can be increased a lot. Basically, here are some advantage of Multi-junction solar system. If we use concentrated solar system, then you can use the more expensive Multi-junction cell. Then it is more efficient. But there is no such perfect thing. There are still some problems such as thermal dissipation. For example, first one is whether it will burn when sunlight concentrate here. do not even say burning out. As long as the temperature increases, the efficiency of solar photovoltaic panels will become low. So the thermal dissipation is very important, the problem need to be solved. The second problem is that because the light is difficult to gather after heat dissipation or refraction. So most still need to be in the place where the light came in directly, it may be near the equator. If it is closer to the equator, then the effect will be better. It still has the limitations.