Loading

Module 1: Solar Cell

Study Reminders
Support
Text Version

Set your study reminders

We will email you at these times to remind you to study.
  • Monday

    -

    7am

    +

    Tuesday

    -

    7am

    +

    Wednesday

    -

    7am

    +

    Thursday

    -

    7am

    +

    Friday

    -

    7am

    +

    Saturday

    -

    7am

    +

    Sunday

    -

    7am

    +

So, from metallurgical grade silicon, we need to create something that is referred to as electronic-grade silicon EGS it is also referred to as semiconductor grade silicon okay. So, its metallurgy electronic-grade silicon or semiconductor grade silicon you can see here the difference what was you know 2 per cent impurity, now suddenly drops to ppm parts per million of carbon and oxygen in it and parts per billion of any other metallic material present in it. So, this is a remarkable you know shift from going from you know several orders of magnitude you are changing in purity when you go from something that’s 98 per cent pure to something that is you know having parts per million of impurities parts per billion of impurities and so on. So, to do that in fact, what they do is a process where you take silicon solid. So, you have to start with silicon solid and you have to arrive back at silicon solid right. So, you are going to start with silicon solid you going to do some processing. So, it may in the middle it may become something else and then, in the end, you again get back the silicon solid except that. Now, it is purer silicon solid. So, that’s essentially the idea in this step here. So, if you consider this is a step 1, this essentially steps 2. So, step 2 is you take silicon solid you react it with HCL and you will essentially get this trichlorosilane gas trichlorosilane. So, that is the, it is in the gas phase and then once it is in the gas phase we can purify the gas. So, there is it may be it may have other constituents in the gas we do some separation some kind of you know fractional distillation etcetera and then you can separate the trichlorosilane such that you now have trichlorosilane in a purer form you don’t have anything else along with it. So, that way we separate this silicon fraction and then you again reduce it back to silicon you reduce it back to silicon. So, in many ways, this reaction is the same you simply go in the forward direction initially to create this trichlorosilane and then go in the reverse direction to get back your silicon. So, you set the condition such that it will go on the forward direction get this trichlorosilane and then go on the reverse direction get back your silicon except that in the middle there is a purification step the trichlorosilane is separated from any other constituent that is there and so you have a purer version of this trichlorosilane that is then reduced and you get back your silicon. So, when you do this process you get this electronic-grade silicon okay. So, so that’s how you arrived at this electronic-grade silicon. Most likely that sample is going to be polycrystalline nature. So, most likely you are having polycrystalline electronic-grade silicon which is a so pretty significantly pure silicon sample typically polycrystalline in nature which is going to be available to you which on which you have to do further processing to generate whatever sample you are interested in working with okay. So, this is the base background situation from where we start and then we try to create the samples that we are interested in. (Refer Slide Time: 20:40) So, from here we will now, move to the czochralski process. So, in the czochralski process what we do is we are trying to now the next two processes we are looking at we are trying to create a single crystal of this sample we have electronic-grade silicon available to us from there you would like to create a single crystal of silicon. So, there are a few things that are done there. Essentially what is the basic idea is you have to melt this material and then when it recrystallizes you have to set the conditions right that it cannot it recrystallizes as a single crystal that’s the basic idea. So, you take something that may be polycrystalline in which case the atoms are you know the order is being broken repeatedly at several locations which are those grain boundaries and then you melt it so that you now, have atoms in a molten state. And then when it recrystallizes you don’t allow them to break order you will try to encourage them to stay ordered. So, that is the basic working principle here. So, what we do here is we have a crucible a quartz crucible in which we have molten silicon. So, you have molten silicon. So, this melting point is about 1414 degree C. So, molten silicon is there this is electronic-grade silicon or semiconductor grade silicon this is what is being used here. And the crucible itself is actually of a higher melting point 1670 degree C or higher based on you know the exact phrase that is there, but that is the melting point that is present there and you have to heat it. So, you have to first heat it. So, you have heating coils here. These heating coils are essentially using this induction process sort of you know like the induction stove that we that seems to be prevalent these days it’s an induction process. So, you have an alternating current in the with radiofrequency RF frequency current that is flowing and that causes the induction and because of the induction that’s heating and because of heating the silicon melts. So, that’s how the melting occurs and we are trying to maintain about 1425 degrees C which is just above just a few degrees above the melting point of silicon. So, we melt it, but we don’t go to very high superheat we just stay just a little bit above the melting point. The reason being from this melt we want to crystallize something. So, what is crystallizing has to you know naturally be a little cooler only then it will crystallize. So, we have to maintain the liquid at a slightly higher temperature the solid at a slightly lower temperature compared to the melting point and then in that process, you can keep the solids separately. So, this is the basic idea. So, on both sides, you have these heating elements and we have some inert atmosphere because we are already taken some effort to you know prevent impurities from coming into the system. So, we want to make sure that we don’t reintroduce impurity. So, you keep an inert atmosphere so that impurities do not come in. However, if you want to do doping, so this is a good time to do it, you can add either boron if you want to create a p-type semiconductor or phosphorous if you want to create an n-type semiconductor one of those two can already be added into this melt. That the single crystal that forms will automatically already have this boron or phosphorus present in it, so the doping can be taken care of in that sense. So, we have molten silicon and then what we do is we first introduce something called a seed crystal okay. So, in the next slide, I will show you that in a step by step basis in this I am showing in the overall picture. So, we introduce something called a seed crystal which is right now, up here. So, originally that seed crystal would have just come in contact with this molten liquid. So, once it comes in contact with the molten liquid and there is cooling occurring very close to this then what happens is the liquid begins to solidify, the atoms in the liquid begin to attempt to solidify they will try to solidify on this seed crystal. So, the seed crystal is a small single crystal which you can easily cut out of a polycrystalline sample if you wish, where you know the orientation. So, you know the orientation you keep that orientation some crystal surfaces there the front of that crystal and that comes in contact with that molten liquid. When you do that the atoms which are below actually in the liquid state find it energetically more favourable to line up with the atoms that are in the solid instead of randomly lining up. So, they line up they given a chance they would line up with those solids because that is the lowest energy state for them. If they did not line up with atoms in the solid that’s a little higher energy state, if you give them a chance they would line up, and you are at the melting point. So, you are in a position to give them a chance and to increase their chance you rotate that you know very slowly you rotate the seed crystal you also rotate the crucible by rotating it you are giving the atoms some level of movement without too much movement some level of movement. So, that they can skip reorienting and then getting into the lowest energy state possible they have enough thermal energy you give them some ability to reorient they reorient and you get this lowest energy state lining up with the atoms in the seed crystal. And then you gradually pull that seed crystal out very very slowly you pull the seed crystal out. So, you are pulling it out at 25 millimetres an hour okay just two and a half centimetres every hour you pull it out the two and a half centimetre slowly gradually it is getting pulled out averaging 25 mm per hour okay. So, when you do that this single-crystal begins to grow. So, as you keep pulling it out more and more liquid keep solidifying and in all the solidifying liquid is forming a single crystal in line with the seed crystal and then it continues to grow and then eventually do you have a large single crystal and you have essentially consumed all of the molten silicon that is present. We of course, just have a sum holder here which is being referred to as a chuck and the chuck holds that seed crystal and then and it can be rotated. So, that is the basic idea that we use in this czochralski process and you get this single crystal that grows. Generally given the nature of the, you know the dynamics of this whole process the induction heating that is happening you know the kind of conditions under which we are operating. Generally, the kind of diameter they can succeed to get is about 200 to 300 millimetres diameter these days they are finding ways to make even larger single crystals larger diameter single crystals, but this is the kind of diameter you are looking at. I will also tell you that in both these processes the czochralski process as well as the next process that I am going to talk about them because there is you know liquid which is trying to solidify onto a solid there is always some surface tension related behaviour. That’s the reason why you get the circular shape circular cross-section, otherwise, you could even get a square cross-section or you know some other cross-section. But because it’s a liquid which is in a position to you know to shape itself based on its surface tension you get that circular cross-section because that gives it’s the last parameter and if it where you know 3 dimensional it will take the spherical shape, but here you are you know your position of interaction is a 2-dimensional surface the bottom of the seed crystal and the top of the liquid. So, you only have a 2-dimensional
the surface there they are in contact with a 2-dimensional location. So, there instead of a sphere it takes a circular shape and so, these single crystals that grow are typically circular in cross-section. And, so what you eventually get what you are seeing here is a cylinder it’s a cylinder cylindrically shaped a single crystal. That’s the reason if you look at solar panels where they have used slices of the single crystals you see circular cross-section there you don’t see some random shapes you see the circular cross-section. So, anyway, so that is the idea of the czochralski process. So, it is czochralski is a person credited with it he discovered it by accident and he found that you know when he was working with some melt he could pull out a single crystal by you know pulling contacting it and then pulling something out which was cooling very well. So, that is the basic idea. So, this is one way in which you can generate a single crystal. The same thing in a stepwise sequence I am just showing you here. The same czochralski process just to you know make it better make a better image in your mind about what is happening. (Refer Slide Time: 28:22) So, you would you can say you are starting with this situation here your melting polycrystalline silicon and adding dopants if you want to add dopants. So, at that point, you simply have a melt you have a melt that is here and you can build up the melt to how much over the degree that you want and it is a high purity melt okay it is semiconductor grade silicon that you have chosen. So, it is a high purity melt. So, in the next step of the schematic, we introduce the seed crystal. So, that is the seed crystal that is sitting here it is a small single crystal as I said and it comes in contact with this liquid at that one location there. And then you start slowly you start pulling this out. So, you can see here this is the growing a single crystal that has started appearing okay. So, as you slowly pull this out over some time. What is happening is these two things are happening your single crystal is growing and of course, the liquid is getting consumed. So, the liquid level is continuing to drop the single crystal is continuing to grow. So, those two things are happening. So, in the end, you will have a relatively tall single crystal up to here and then a very small amount of remaining liquid, so this is the basic process and so that we would call as a completed single-crystal process. So, this is what we have and further processing is done with this single-crystal okay. So, that is the czochralski process as an overview and you can look up more details of it, it’s a pretty extensively used process for single crystal manufacture. (Refer Slide Time: 29:57) The other process we have is called a float zone process okay. So, another process here called the float zone process. So, it has some advantages and disadvantages to the czochralski process. So, the first thing is in the float zone process we don’t melt the entire silicon. So, we don’t start with a melt we don’t start with a molten you know container a container full of molten silicon that is not our starting point instead our starting point is a rod, we pick up a polycrystalline rod. So, for example, the top part of this sample that you see here is that polycrystalline rod originally the entire sample would have been polycrystalline from top to bottom. So, I have something here called the upper chuck and then there is a lower chuck these are the two things that are holding that sample together originally from the upper chuck to the lower chuck entire thing would have been a polycrystalline silicon rod. So, that is what we would have had. We have a small seed crystal at the bottom we have a small seed crystal. Now, the interesting thing that we do is we have heating coils same process we are using we are using some RF-based heating coils, but these heating coils are moving. So, they are moving. So, we have heating coils that are moving upwards and therefore, their sphere of influence is a very small region. So, that’s the region that is the sphere of influence that is where the sample is hot. So, the sample may be cold on top may be cold at the bottom, but there is a region in the middle which is hot. So, you have a small molten zone that is what you see here in this region that is your molten zone. So, the silicon melts there alone locally okay. So, there alone it melts and it has some shape mainly because of surface tension effects it has some shape it’s trying to minimize its surface energy. So, you get that kind of a shape and it melts. So, on top, you have a polycrystalline sample, in the middle, you have this melt in the bottom you have a solid. So, given that you have a seed crystal and you have given that melt some opportunity to align itself concerning that seed crystal and that’s the direction in which it is cooling it begins to form a single crystal at the bottom. So, this becomes a single crystal okay. So, that becomes a single crystal. So, the bottom becomes a single crystal because the seed is there and you are moving away from that direction. So, you are continuing to melt towards the polycrystalline side you continuing to solidify towards a single crystalline inside and because you do this steadily it the sample becomes single crystal from bottom to top okay. So, the big advantage here is that you are not having to melt the entire sample, okay and here also you use argon because you want you don’t want to introduce further impurities into the process. So, you do keep the inert atmosphere, but the main disadvantage perhaps is that because you have this you know this small region that is that has to hold a stable you know physical state in without you know collapsing and it is you know limited by surface tension forces and so on. So, there is some limitation to the size or the diameter of this rod that you can use. So, generally, it is found at about 150 millimetres is the sort of the largest diameter a sample that you can create using this process called the float zone process where you know you just have this small region of melt that is moving up and that is primarily due to surface tension. But the nice thing is you get very high purity you get much higher purity relative to the czochralski process. In the czochralski process, you do get some impurities from that cross the quartz crucible, the quartz crucible is adding because it has SiO2 it is adding some amount of tiny amounts of oxygen into the process it also gets affected a bit by the thermal process that is occurring there. And so, therefore, some the level of purity that you get here is a little bit less than the level of purity that you get with the float zone process and therefore, from the purity perspective this is better but size-wise you get a little larger size if you work with a czochralski process. So, these are two major processes that are used to create your single-crystal silicon and you know the focus is to both these single-crystalline and also to have high purity. So, in this context, we will also look at the idea of zone refining which is primarily focused on the purity aspect of it. You already got a single crystal what can you do to purify it okay. So, in some ways, it is also related to the float zone process because the general idea is the same you will have this kind of a single crystal. Now, from bottom to top, but you can do this moving melting in this case you know as you move this heating coil the melt the molten region is moving from the bottom to the top right. So, the whole sample is not molten only once a small region is molten and that molten region is moving from one end of the sample to the other end of the sample. So, that’s an idea you keep in mind. We go ahead and look at this situation here. (Refer Slide Time: 35:09) What do you see on your screen is a phase diagram a section of a phase diagram, of some hypothetical material we will assume let’s say it is a section of a phase diagram let’s say for silicon. But of point B on your y-axis, you have a temperature. So, that is what you have here on the x-axis you have a composition in percentage B. So, percentage B is any impurity, in this case, it is some impurity that is there in silicon which you don’t want in silicon. So, you want to remove it. So in fact, you want this percentage you want to be as close to this end of your final sample to have a composition as close to this end of the x-axis as possible right. Supposing you start with a sample that has this composition, okay, you start with the sample that has this composition and you start cooling it. So, we call this the liquidus this is the liquid state this is a solid-state and this is solid plus liquid, okay and this is called the solidus this line is called the solidus okay. So, what this means is if you take this composition and you start cooling down if you stop here if you are at this temperature let’s say you are at this temperature and that temperature they sample will be entirely in liquid state okay. The same composition if it were at this temperature corresponding to this region on your y-axis then the sample will have partly solid partly liquid and we can use the lever rule to find out you know how much of it is liquid how much of it is solid. If you are at this temperature this sample will be entirely in the solid-state. So, this is how the phase diagram works. Now, what is of interest is the transition that happens when you are between here and here, where you start off being completely liquid and you arrive at a state where you are completely solid okay. So, when you do that the interesting thing about this phase diagram is it tells you what is the composition of the solid that is separating at any given instant okay at any given circumstances. Even though you start with a liquid that has this uniform composition, it is uniform composition, what happens is as a virtue of this phase diagram the first solid that comes out does not have this uniform composition. It has this composition down here okay. So, that’s the composition that the first solid that comes out of this liquid will have. Progressively as you keep decreasing the temperature the next I mean if you go down somewhat and further in the temperature you will have solid that is coming out with this composition etcetera. So, the composition of the solid that comes out will slide down this line, the composition of the liquid will slide down this line. So, in the end, you will have a solid which is a single crystal in this case. But the point is if you have done this the solid this single-crystal let’s say you started at the bottom and you finished off this process at the top of that single-crystal the bottom will have this composition, the top will have this composition. So, therefore, you will have a situation where the composition has been changing continuously through this sample right. So, therefore, the end of the sample which solidified first is a much purer sample than the end of the sample that solidified last. So, therefore, you can do this process and you can remove the last part of the sample which was impure and you will have now, a sample which may be slightly smaller in size, but has higher purity than the sample you started with. You can continue this process repeatedly and each time you do this you will keep getting one end of the sample that keeps getting purer and purer and purer. So, this is a process they use to make this overall sample as pure as possible and that is how you get this high purity you know sample. So in fact, if you do this in an equilibrium state then you will move across then the composition of the entire solid will move on up along this line. So, that the final solid has this composition, but normally we do it in n non-equilibrium sense. So, you will have a range of compositions in the solid. So, you will such that the overall average composition is this okay. So, the average composition you cannot move away because that is a composition of the solid, but it will not necessarily move down this solidus because you will be doing it in a non-equilibrium sense right. So in fact, you will have a range of compositions and you will have you know not exactly finishing off with this exact composition that you see on the that you will not follow this line directly you would be a little bit away from the line because you are doing it in a normal equilibrium sense. So, that is the practical aspect of it, but this is the concept behind it and you can keep getting a solid that is purer and purer and purer. So, that is zone refining that is used for refining single crystals. (Refer Slide Time: 40:18) So now, now that you have a solid you know single-crystal available to you in the form of tubular structure what is the next step that is done. The next step that is done is slicing, slicing of silicon ingot to get wafers. So, that is what we keep here you know silicon wafer is a term that we keep here this is what is happening at that point. So, they have taken this cylinder that we got through the single-crystal process either czochralski process or float zone process, and then we purified it all that we did they are now satisfied that the quality of that you know the single crystal is what we want it to be the composition is right the orientation is right all that we figure out. Once you have figured all that out we do this slicing. So, normally what they use is they use a diamond cutter, diamond cutting wheel and they use the inner edge of it that is why it is called the inner diameter slicing and then you slice the silicon wafers one by one you get a slice. And typically those silicon wafers are about 0.2 to 0.75 mm thick there are some limitations because if you get thinner than this you may you know you may you will have some problem in stably making it. So, you tend to get in this dimension 0.2 to 0.75 mm thick. But this kind of a cutting process will get you one wafer at a time they also have some other processes which are based on some wire saw. So, wire saw that you see here can be used where you basically have a thin wire with an abrasive slurry and then you can have several of these thin wires or the same wire you know wound in different directions etcetera so that you can cut the sample simultaneously at several locations so that at an instant you get a huge number of wafers okay. So, simultaneously the wire is cutting the same cylinder at several locations and therefore, you end up getting several slices.
So, that’s another way in which you can get several wafers at the same time. So, this is an important step. Once this is done there is still some cleaning that is required of the surface of that sample and also to make it as flat as you would like it to be. So, there is still some processing steps involved, but this is the after you have made the single-crystal the next major step is to make this wafer and this is the general idea that you just cut it slice by slice by slice and you get this way these wafers. So, that’s the next major step. (Refer Slide Time: 42:26) So, having done this what is also done in the semiconductor industry. So, that it becomes easy because once you have all these wafers it can get very confusing you will dope them and p doping you have done n doping. And usually based on the crystal faces that are available you can either have 100 planes on the, as being the planes that are you know parallel to the top surface. So, the entire cylinder consists of 100 planes going from one end to the other or you can have 111 planes going from one end to the other. So, these are the two you know crystallographic planes oriented concerning the face of the cylinder of the cylindrical sample that is most commonly present in this silicon single crystal. So, that doesn’t have to be true for other systems in this system, these are the two that are of interest that typically these samples are made in this form. So, just to make it easier for people who just look at a sample and say okay this is n-type or p-type and this is a 111 or 100 plane on the surface they usually use notches. So, this is a notch that you see here. So, you can see what was the circular cross-section you know a small bit of it has been cut off. So, you can have this you know initial notch that is present here if you have only that notch it means it’s a p-type sample which has 111 orientation on the surface. If you have this notch as well as another notch you know perpendicular to it, it is still a p-type sample, but it has a 100 orientation associated with it.
Instead of this second notch being at you know 90 degrees to the first notch if you had it on top then that is an n-type sample also with 100 orientation and finally, you can have n with a 111 orientation where you have this initial notch as well as this notch on the side at the edge at an angle to it. So, we do these notches to help you quickly identify at a glance you know what are you dealing with n-type or p-type 100 or 111 and that is useful for further you know processing or further any further activity you want to do with the wafer this is a very important piece of information and you cannot keep on going back to the x-ray diffractometer to find out what sample you have what orientation is it what is the composition is it p-type or n-type we are using some analysis technique. You know initially what it is, so you just mark it in such a way that you can see it permanently. These days they have come up with some for larger you know wafers they have come up with a slightly different kind of a notching process or at least there is some modification on it, but this is still a there in many of the samples that you are going to see. So, this is another important step. So, you made the single-crystal you did the slicing, you introduced these notches. So, this is what we have done. (Refer Slide Time: 45:05) And then that can be used. So, we will talk briefly about how it can be used in the next slide, but before that, we will also talk about amorphous silicon because that is another this so far we have spoken about single crystalline silicon. So, amorphous silicon is made by chemical vapour deposition process CVD process essentially; that means, that you create a vapour which carries this you know silicon atoms. So, typically from silane gas is used that has silicon atoms in it and then you will have some substrate on which it can deposit. The nice thing about amorphous silicon is that it can be deposited on any surface. So, you can put a semiconductor on any surface any curved surface you don’t have to worry about the shape of the surface. So, potentially, for example, let us say you are trying to make a solar cell-powered car, instead of buying flat you know rigid flat solar cells and then placing it on top of a car you can almost deposit this amorphous silicon all over your car, and so your car can be in any shape you will have amorphous silicon all over it and then you can use that for any process that you want. So, that’s the basic idea that in any shape you can do it I mean that is the idea it may not be that easy to j