BZE interviews Dr. Matthew Edwards from the University of NSW on solar technologies

Beyond Zero's Matthew Wright speaks to Dr. Matthew Edwards, an expert in photovoltaics from the University of NSW, about the latest solar technologies.

Dr. Matthew Edwards interview

download

Transcript

Matthew Wright: Today we’re interviewing Dr Matthew Edwards, and he’s from the University of New South Wales, and he has a PhD in Photovoltaic and in particular methods for doing incredibly nerdy things, that I can’t even describe it to listeners, but we can get Matthew’s mouth, so I will grab him on the line now. Hello are you there Matthew?

Mathew Edwards: Yes I am.

Matthew Wright: Now look, I was going to say screen and stencil print technologies for industrial end type silicon solar cells, but that may not mean a lot to my audience so, can you tell us a bit about yourself? Obviously your from UNSW, how did you get interested in solar?

Mathew Edwards: Well, many of the people that I am working with are from an environmental background, or an environmental perspective. I guess for me it was slightly different, I always had a real passion for electronics, and when I first learnt about solar cells I really fell in love with them as a device.

They are simple, they are basically just a diode, a very simple electronic device, no moving parts, and I really just loved the way the electrons moved about in the silicone.

The fact that it is a very new technology with the potential to help humanity and the environment was just the icing on the cake for me, and I felt that here was an area where I really can make a contribution.

Soon after that, I started working with solar cells as part of the solar racing team and UNSW, spending countless hours under a solar cars with a soldering iron, mucking around with solar cells, connecting them in the Australian outback, and when you are passionate about something, it’s really great to get to work with, and play with them, and touch the device, rather than learning theory, exams, and that’s sort of how I fell into solar research.

Matthew Wright: And these days you are actually concentrating on commercial application, getting to a point of greater efficiency and lower costs?

Mathew Edwards: That’s right.

Matthew Wright: Can you talk a bit about the kid of work you do, and describing maybe the process of getting to a silicon way fare that can have the sun shining on it, and can produce electricity?

Matthew Edwards: So at UNSW, I am the programme manager and co-founder of what we call the Photovoltaic Technology Transfer Team. What we do is, we aim to put cutting edge solar cell technology into worldwide mass production, so we work with Asian, European, and US companies.

Some examples: Hyundai, Suntek, Shin Sun, LG, Centrotherm, Guodian, DuPont, Pharoh, to name a few. We do research on reducing the cost and increasing the efficiency of solar cells. We do the work to aim to reach grid parity, and you might have seen in the media recently that we are starting to reach grid parity now in Australia, so that’s very exciting for us.

But we’re going to continue to reduce cost and increase the efficiency of solar cells so that we move beyond grid parity.

Matthew Wright: And in terms of silicon cells, so listeners get an idea, we are talking about I assume you buy ingots in, so you pour ingots in. So can you describe what that means?

Matthew Edwards: So basically you get an ingot of silicon which is grown in a large caste, or grown from a single crystal in a long cylindrical ingot, and that’s sliced up of silicon wafers, so a solar cell consists basically of a silicon wafer with a positive metal contact on the rear and a negative metal contact on the front, so in that sense it’s a bit like a battery, you’ve got one side is positive contact, the other side is the negative contact.

The light comes from the sun in little balls called photons. Some of these photons when you put the solar cell in the sun you have some of these photons of light, coming from the sun and hitting the cell. Some of these photos will bounce off, which is reflected light, and some of the photons go through the cell and come out the other side, and these are the low energy ones red light photos.

The important thing is we get some of the photos going in to the silicon wafer, and getting absorbed by the silicon. If a photon is absorbed in the silicon, the energy that was in that little ball of light is given away to the silicon, and this energy tears an electron away from one of the silicon atoms, which is in that silicon wafer.

So you might recall that electrons are negatively charged, and when they move about freely, they can create an electric current. So where there electron used to be, there is a vacant place that we call a hole. A hole is effectively positively charged. Because other electrons can jump into the hole, the hole can also move about freely, just like the free electron can.

So with lots of these photos of light being absorbed, we end up with lots of free electrons and free holes, and then what we do is structure the cell in such a way so that the electrons and holes all move away from each other, and the negative electrons all move up to the front of the cell and the positive holes go down to the rear surface of the cell. So you end up with the front contact negatively charged, and the read contact positively charged.

That’s just like a battery or a similar device and then just like a battery if you connect a load across that, like a light bulb, or a music player or whatever you like, electricity flows, and the light turns on. So that’s basically how a solar cell works.

Matthew Wright: So negative side is facing the sun basically?

Matthew Edwards: Yeah, that’s right.

Matthew Wright: In terms of the ingot, how big is the ingot? Like how many metres long, or centimetres long?

Matthew Edwards: It varies. The ones that I have seen being made in the industry, they’re really long, they can be sort of 10 metres long, so they’re quite huge, and there’s many thousands of dollars in those ingots.

Matthew Wright: And if they’re not caste, do they sit along the floor or hang on the roof…?

Matthew Edwards:The way it works is, when your making single crystal ingots, you start with a very long seed crystal, which is sort of like a rod, and around that you have a melt, so you thrown in a lot of raw silicone, and that melts and then you have the whole thing spinning, and the crystal cools down from the centre of the cylinder to the edge, so over time you grow this single crystal ingot which kind of forms around the seed crystal, if you understand what I mean?

Matthew Wright: And how long does that process take? Half a day, days or…?

Matthew Edwards: Yeah, it takes a day or two.

Matthew Wright: And the doping process, that allows the electricity to flow across the cell, does that happen at that stage, or does it happen after you have chopped the ingot up?

Matthew Edwards: Well, the way you structure a solar cell, you have the cell doped what we call ‘P-Type’ in the main bulk of the cell, and at one edge you dope the ‘N-Type’ which is the opposite polarity, so P stands for positive, N stands for negative.

The P-type doping which we use boron, is done at that process when the ingot is grown, so in that spinning melt that I was talking about, you throw in some boron, and that dopes the p-type doping. The n-type doping we do later on when we actually have the wafers cut up and we are making the cell.

So that gives us the structure that I was talking about that allows the electrons and the holes to go in opposite directions and give us a separation of charge.

Matthew Wright: And I understand these wafers are getting thinner all the time. Can you give us an idea how thick would a wafer have been 10 years ago versus today?

Matthew Edwards: 10 years ago we were sort of looking at 300-400 microns, now we are reducing them down to about 180-200 microns, and we are going thinner all the time. 300 microns is .3 of a millimetre, so down to about .2 of a millimetre, .18 of a millimetre, so we are getting thinner and thinner all the time.

And we are getting thinner wafers with our sawing processes and the nice thing about these thin wafers is that, if you have an ingot which is worth thousands of dollars, obviously the more wafers you can cut from it, the more solar cells you can produce from it. So those solar cells are getting cheaper and cheaper.

So what we are looking at, are methods of making solar cells where we don’t destroy those very thin fragile wafers that we have sliced up and that’s one of the things that we are looking at as well.

Matthew Wright: So that also effects the embodied energy of the solar panels, as you get narrower, thinner and thinner wafers, your using less energy which is why you have got the lower costs?

Matthew Edwards: The wafers themselves become a lot cheaper, because for a given an amount of silicon ingot, your allowed to cut more wafers from that ingot, so we are at a situation here we have about 70-75% of the cost of producing a solar cell, is in the silicon itself, so we are always looking at ways to use less silicon or to use cheaper lower quality, which doesn’t cost as much to grow and purify in the first place.

Matthew Wright: We are speaking with Dr Matt Edwards who is the Program Manager from the Photovoltaic Technology Transfer Team in UNSW. So back to the wafers, what is the trajectory? You said they have gone from .3 millimetres, down to .18…

Matthew Edwards: yeah…

Matthew Wright:… what’s the trajectory over the next 10 years? How thin to they go? Do they get too thin and no-one can handle them anymore?

Matthew Edwards: Well that’s the problem, they get to the stage where it becomes very difficult to handle them, and I guess we are looking at ways we can process them, and this is one of the things we have done with our laser dope selective emitter cells, we have taken away the screen print process.

The screen print process is a rough process; it puts a lot of pressure, direct pressure on the wafers. One of the reasons we want to do that is so that we can work with thin wafers. There are other things we are trying to do as well.

Basically, the thickness of the cells is limited by how gently you can process them afterwards. They can get very thin indeed, and in fact when you get down to very thin, they start bending and become flexible, they are actually easier to handle once they get to that stage, but it’s only a certain select few solar cell processes if you like, that work with silicon wafers that thin.

And the other thing that you have to think about is when you’re making wafers that thin, the more red light will pass right through the wafer and not contribute to the power generation. So there are a lot of different things to think about, but definitely we are always looking for ways to make the cells thinner and thinner, by creating advanced solar cell designs that trap that red light, but also gentle processing.

Matthew Wright:And how are they actually sawn, and how is the sawing process trying to reduce wastage when they cut the actual wafers?

Matthew Edwards: They’re sawn basically with a machine that has a very long diamond coated wire, that looks like a guitar string or something, and when I say long, I mean kilometres almost. It’s coiled around and if you have a look inside one of these machines it’s almost like a giant harp or something the way the wire is arranged. And this wire spins very fast, and the ingot itself is then lowered onto that. In time, and you have to sue a lot of slurry which contains a cutting compound in it, with time that ingot is sliced into very thin wafers which can then be used to produce solar cells.

Matthew Wright:And are they reducing the size of the cutting wire?

Matthew Edwards: Yeah, they’re reducing the size of the cutting wire using new methods as I mentioned the diamond coated wire, and basically looking at ways to reduce the amount of saw dust if you like, they call it curflos. And as the sawing technologies improve, the curflos gets lower and lower, and you can effective saw thinner and thinner, and you can get more wafers from the one ingot.

Matthew Wright: This is aver interesting interview, and I hope we can recruit a number of budding new UNSW solar PhD students…

Matthew Edwards: I hope so.

Matthew Wright:… I have spoken to a lot of solar PV people in the past, and have never had such a good rundown, so I appreciate that. We are about to get to the exciting bit, which is the fact that you guys have achieved this 19.5% efficiency goal. But before then I have a couple of questions about solar panels.

So now another thing, which might relate to your efficiency record, is, once you lay these very thin wafers and sandwich them between glass and something, you have actually got to get physical contacts onto them, and that takes away some of the area that the sun can work on the wafer. Can you talk about that? 

Matthew Edwards: Yeah, that’s right. So as I was talking before about the front negative metal contact which is on the cell, and it’s traditionally is done by screen printing, similar to the way you would screen print a t-shirt or something, but in this case, we are screen printing a paste which contains silver, and when we fire it, we are left with this metal grid of silver on the front surface of the cell, then when you have to encapsulate the cells and put them into the panel, you have to connect those contacts together, so once cell will be connected to the next with a metal ribbon.

So all of these metal contacts and interconnection sit on top of the cell so they shade the light basically. You lose about 7% of the light from these contacts with a standard silver screen printed solar cell. In our new more advanced solar cells, we only lose about 3% of the light, because we have finer, narrower front metal contact.

Matthew Wright: Now you’ve talked to us about the red light passing through, in satellite based solar cells, the very expensive ones, like the once from Mark Wanless at NREL, National Renewable Energy laboratory, they actually run multi layers, where they can capture different colours of light, or bands of light, and is there likely to be, once these cells get really thin, a new process that allows domestic grade panels to actually have 2 layers or 3 layers?

Matthew Edwards: Look everything is possible, at the moment the issue with these tandem cell structures, is that you require a stack of solar cells that have different band gaps.

At the very front of the stack you put the widest band gap, and that absorbs the ultra-violet and the blue light, and then as you get further down into the stack, you put narrower band gaps which absorb first the green, and then as you get towards the rear they start to absorb the red light.

So they have very narrow band gaps at the rear. And the way you do this engineering of a cells band gap requires exotic materials. You can’t just use silicon. Silicon itself has a fixed band gap. The really nice thing about silicon, is that its cheaper than these exotic materials, and secondly the band gap of silicon just happens to be quite well matched to the most of the spectrum of light that we are getting from our sun.

So it’s quite fortuitous that we have this material that is abundant on the earth and which just happens to be almost perfectly suited to absorbing energy from the sun.

If you want to make it better suited as they do at NREL, and as they do with these tandem structures, then you need to start using more exotic materials, galleon arsenide’s and things like this. Those materials are more expensive. So those type of cells that you’re talking about produce more power, but you have to spend a lot more to make them.

Matthew Wright: The light that can be captured by the silicon cell is a maximum theoretical of about 30% of the sun’s energy?

Matthew Edwards: Yeah, that’s about right. We have the highest standard world record solar cell efficiency at UNSW for our pearl cell, which is about 25%.

Matthew Wright: Is that a laboratory cell?

Matthew Edwards: So that’s a laboratory cell. As you say the physical limit is a little closer to 30% the practical world record is 25%.

Matthew Wright: And let’s just talk about the commercial stuff now, because I understand from a press release that you were working with Centratherm and a number of other companies, and you got a 19.4% efficient in June this year.

Matthew Edwards: That’s right.

Matthew Wright: Now, is that a commercial application, or is that another lab cell? What’s the story with that?

Matthew Edwards: No. The thing with that world record, that is the highest independently confirmed efficiency for a low cost mass producible solar cell. And those word ‘mass producible’ are very important, because the important thing about this, is that we produced it on standard P-type silicon which is commonly used in industry today, and it was actually fabricated on a standard Centratherm production line, which was modified only slightly by removing the screen printers and the firing furnaces for the screen print contacts, and replacing those with a laser and a plating bath.

So, one of the reasons that we are excited about this cell, is that it is a mass producible cell. And in fact we have various companies in pilot production of this technology.

Matthew Wright: And they are reasonably big players?

Matthew Edwards: They are, yes.

Matthew Wright: Can you describe the process, you said you just replaced it with the laser…

Matthew Edwards: Yeah…

Matthew Wright: Is that going to be more energy intensive, or using more expensive equipment, or is it actually lowering the cost?

Matthew Edwards: By way of introduction, about 90% of the world’s production, is in silver screen printed solar cells, which is this basic form of silicon wafer with this rear positive contact, and a front negative silver screen printed contact.

The top surface of these cells is heavily phosphorous doped and we call this a homogenous emitter. We need this heavily phosphorous doping on the surface of the cell to get good electrical contact with the silver screen printed contact. The problem with that heavily phosphorous doped surface, is that this is where all the blue light is absorbed on the front surface of the cell. And because of the heavy phosphorous doping we have a huge power loss on the front surface of the cell, so we lose this blue light.

So with our Lased Dope Surface Emitter, or LDSE cells, we have a very lightly doped front surface, so we regain a lot of this blue light, and then we use a laser to heavily dope only those regions of the silicon that are directly under the silicon contacts. So that’s why we call it a selective emitter, selective doping. And what this allows us to achieve is very good contact resistance still between the silicon and the metal, but with a very good response to blue, and ultra violet light.

We also want to be able to remove the screen print silver. Number one, because it is expensive, these silver pastes, because of the price of silver these days, silver pastes’ are accounting for more than 1/3 of the cost of converting a silicon wafer into a solar cell, so it’s quite ridiculous.

Number 2, they have a high shading loss, about 7% screen print contact. It’s a high temperature step number 3, and number 4 rough processing as I have already mentioned. So it is not good for our cheap, or thin silicon wafers.

So the main advantages of the LDSE over a standard screen printed commercial cell, is that we’re effectively optimizing the front surface, we’re removing the standard screen print contacts, and replacing them with light induced copper plated contacts.

And the light induced plate is really a beautiful process where it’s a little bit like electro plating, where we’re passing the light through a plating bath, but the plating bath has lights inside of it, and as the cells pass under the lights, they are illuminated, and they start producing a voltage, because that is what solar cells do. And that voltage they produce is used simultaneously to electroplate themselves effectively. So it’s a really beautiful elegant process.

And this all takes the shading loss from about 7% for a screen print contact, down to about 3% for our light induced copper plated contact, and it completely removes the use for the expensive silver, and replaces it with the cheaper copper.

And then the other advantage of course, is that we get the higher cell current, which is due to better response to blue and ultra violet light, so much higher current coming out than we would normally get compared to a normal commercially available solar cell which is screen printed.

This extra light and extra current takes the efficiency of the cell from 18%, up to 19.4% which is our world record, and leads to a cost term, cost per watt terms of 5% over a standard commercial silicon solar cell.

That 5% cost saving might not sound like too much, but for a typical large solar cell producer, producing over 1 GWs of generation capacity, per year represents a $60 million improvement to their bottom line.

Matthew Wright: And what if you achieve the 22% efficiency from the rear surface as well, what sort of saving?

Matthew Edwards: So what we are doing there, is that we are also optimizing the rear surface, so we have now both surfaces optimized, so we effectively remove the screen printing from the rear from the cell, so that gives us a very nice selective rear surface, which affects the voltage. So now we are affecting the cell current and the cell voltage, and when you do that, the efficiency increases well above 21%, and heading towards to 22%.

That’s a cost reduction in cost per watt terms 15%, which is about $220 million per year improvement to a bottom line of a large solar cell company.

Matthew Wright: Alright, we have to leave it there. Fantastic, thanks for joining us today Dr Matthew Edwards.

Matthew Edwards: No problems. Thanks for having me.