Season 2, Episode 4: Martin Green, UNSW Sydney
This week Torsten sits down with one of the most influential voices in the solar research community, Martin Green, aka “the father of modern PV,” for a retrospective look at the early history and rapid progress of solar technology.
Martin’s groundbreaking work has shaped the industry as we know it, and his story is one of innovation and grasping opportunities as they arose. His early interest in microelectronics in the 1960s led to developing world-record-setting solar cells in the 80s. Humble despite these amazing achievements, Martin jokes, “One of my colleagues used to say I was famous in my own lunch box.” But with a ferry in Sydney Harbor set to be named after him, his impact is undeniable.
Martin details his pioneering work with what is now known as TOPCon and with PERC technology – both major breakthroughs in solar cell architecture. Although going against conventional wisdom and in a constant battle to secure funding, his team persevered. His thoughts on the naysayers? “We were told at the time we'd never make it into production. So, yeah, don't listen to the advice that you get given."
Looking ahead, Martin sees potential for even lower solar costs. "We might be looking at a future where solar is not just cheap, but insanely cheap." An icon in the field, his story is a must-listen for anyone curious about how solar became a leading solution in the fight against climate change.
🎧 Catch this engaging episode for a glimpse into the mind of a true solar pioneer, packed with insights, expertise, and a forward-looking vision.
🎧 Listen to the full episode here.
📹 Subscribe to our YouTube page for the latest video updates.
Show Notes:
"Solar Cells: Operating Principles, Technology and System Applications" by Martin Green
From his early days in electrical engineering to discovering solar panels in 1969.
Developing groundbreaking solar cell structures, including tunnelling and PERC (Passivated Emitter Rear Contact).
How his team set multiple world records for solar cell efficiency, starting in 1983.
Insights into how technologies like PERC, TopCon, and thin-film cells reshaped solar production.
Predictions for cost reductions in solar energy—possibly reaching five cents or even one cent per watt.
Why the U.S. lags behind in solar adoption, and the striking contrast with countries like China and Australia.
Transcript
[00:00:00.300] - Torsten
Welcome, Martin.
[00:00:01.390] - Martin
Thank you, Torsten.
[00:00:02.790] - Torsten
I studied physics in Karlsruhe in Germany, and my professor back then, Peter Wurfel, that's in 1993, so roughly 31 years ago – he mentioned that there's this book by this Australian professor, Martin Green, and it would be a good read in addition to his own lecture. So I thought, if Peter proposes a book from Australia so far away, it must be a good one. And that made me actually write not only my very first emails, but also spend one a half year in Australia to do my master thesis.
Martin, your former coworker, Andrew Blakers, was faster, so he saved you from my presence. This is why I did not go to Sydney to see the Opera House, but Canberra where I did spend a wonderful time and got a great training in SolarCell Physics, also by Andres Cuevas. I'm sure you know him as well. But I surely think you have better beaches.
Do you regard yourself as famous? And I guess the correct answer should be yes. And what's it like to be so famous.?
[00:01:01.700] - Martin
One of my colleagues used to say I was famous in my own lunch box. I'm famous in the photovoltaic community, but not as well known outside of it. Although here in Sydney, we, of course, have our iconic Sydney Harbor. And it was just announced earlier this month that one of the new Sydney ferries that sails on this iconic Harbor is going to be named the Martin Green. I'm probably about to become more famous in Sydney than I have been in the past.
[00:01:34.150] - Torsten
The Australian Center for Advanced Photovoltaics recently launched the ACAP Industry Consortium, and JA Solar and Aiko have signed on. What does this mean for Solar Research at the UNICEF, your home university?
[00:01:50.050] - Martin
I started ACAP in 2013 and was Director for 10 years, but I've just handed over to Professor Renate Egan. And Renate has organized funding to 2030. So that's unusual, to be able to secure funding for such a long period for a center like this. But one of the conditions is we have to have more industry money coming in as time goes on. So the money given to us by the government decreases and we're supposed to build up our industry funds. We have very good contacts within the manufacturing industry, particularly in China, since it's mostly located there. That's where a number of my students actually founded that industry.
And then in 2000, we started the world's first undergraduate engineering program in photovoltaic engineering. And over a thousand Chinese nationals have taken that program and then go back to work in the photovoltaic industry in China. Australian universities are very dependent upon overseas fee-paying students to make up their budget. Chinese students have been quite interested in coming to Australia to study photovoltaics. So we've done quite well in that area.
[00:03:04.680] - Torsten
So what's the payoff for companies like JA and Aiko and for all the others that will join the consortium?
[00:03:13.030] - Martin
You're in contact with our research center, being kept up to date with the progress in the different areas that we're working on. Student recruitment is a big factor. So we're facilitating student recruitment. And in fact, part of their membership fee will pay the stipend of a PhD student. It might be like one company all own one PhD student, but we'll have this pool of PhD students, many of which are Chinese nationals as well as our PhD cohort. They'll be able to do things like spend vacations at one or other of the consortium members as well. So it'll be a chance for recruiting some of the top people coming through our program.
[00:04:01.090] - Torsten
Here at The Solar Journey, we like to look into how people first got involved in PV. You also were a person at some point where you didn't do PV, hard to imagine. But what can you tell me, ask about your early studies and what drew you to PV?
[00:04:17.040] - Martin
I went through electrical engineering in the late '60s. In the final years of my program, I got very interested in microelectronics. People were sticking four transistors on a chip, and even some were getting up to 10. That was about the state of development then. But obviously, it was a very exciting area and there was going to be plenty of progress in the future. I guess in the semiconductor course, we had a lecturer that had difficulty explaining the concepts clearly. They tended to be the courses that I did best in, because if you could understand what the lecture was saying, there was no need to worry about it. You knew you'd be able to study up for the exam and get through that subject. But if you didn't understand the material that was getting presented at all, I was forced to put in my own work, try to learn what the lecturer was talking about. I ended up topping the course in semiconductors because I had put so much effort into trying to understand the basic concept. That combination of interest in semiconductors and microelectronics – I won something called the "Third Year Traveling Scholars", which gave me 80 Australian dollars, which is about 60 US dollars.
And the aim of that was to see the world and explore the topics that you're interested in, visiting industry players in that area, all those things. So I hopped in my old car and drove down to Sydney and Melbourne. I was studying in Brisbane. And it's about a thousand kilometers between Sydney and Brisbane, and another thousand to Melbourne. So a fair drive. And then I visited microelectronics places in Sydney and Melbourne, where most of the action was happening in Australia. But I also saw my first solar panel on that trip. That was in 1969. And the interest in terrestrial use of solar was just starting to develop then. And so it was quite early to see a terrestrial panel.
[00:06:18.140] - Torsten
And they made it themselves?
[00:06:19.860] - Martin
No, it might have been Phillips because Phillips was very early into developing terrestrial panels. So they certainly had terrestrial panels in 1969.
[00:06:30.810] - Torsten
When did you start your science career in PV?
[00:06:34.720] - Martin Australia is a member of the British Commonwealth of Nations. And there's a scholarship scheme that allows students to do a postgraduate study in Commonwealth countries. So I'd put my name down for a Commonwealth scholarship in the countries where I thought they'd be more advanced in microelectronics than in Australia, which was the United Kingdom and Canada. And the Canadian scholarship came. So bird in the hand, that's the type of thing. I accepted it and got ready to go to Canada. I can't remember exactly why I had to stay in Australia for a while. Maybe it was passports coming through or something like that.
So I did a masters and that was very important in my solar career, because I wrote one of the early programs, like PC1D or one of the more sophisticated programs that are out now, that solve the semiconductor equations that determine the carrier transport and so on, within the solar cells and transistors. I actually wrote it for transistors. But it was probably in the first four or five programs that were written that did that thing. It wasn't the first. So I did a masters thesis. And what I used that program for was to look at the differences between first order theory and what the computer was saying.
I was just looking for areas where there was a difference between the standard theory and the computer simulations, which was really great because it forced me to understand both the standard theory as well as understanding how devices deviated from it. So that really gave me a very good understanding of the basic principles of bipolar semiconductor devices.
[00:08:24.110] - Torsten
When did you make your first solar cell, or was it the simulation which you applied to solar cells?
[00:08:31.510] - Martin
I made my first cell in 1971 when I went to Canada. So my school in Brisbane didn't have the necessary gear to make semiconductor devices. So it wasn't really possible to make them there. But when I went to Canada, they had a lab set up. And my supervisor that I chose there, he was working on tunnelling MIS structures. Tunnelling would be, this was 1971, and the Nobel Prize for physics was awarded to three gentlemen who did different things with tunnelling in various semiconductor structures. So tunnelling was a pretty hot topic there. And my supervisor was just as interested.
The MOS device was starting to become popular in microelectronics. But he was interested in what happened when the insulator layer became sufficiently thin that you got tunnelling between the metal and the semiconductor – was there anything interesting that happened when that occurred? So that was the topic that I was given. And I was able to adapt the computer programs I'd written for my masters to incorporate a tunnelling structure. So I did quite a sophisticated job of that. In fact, simulations like that that have been done recently on the tunnelling structures have taken over the industry.
[00:09:52.200] - Martin
But in some aspects, they weren't as sophisticated as what I was doing back in 1971. My job was, because I did my PhD in engineering physics, but because I had an engineering background, my topic was not just a study of currents and so on in the device, but to look for possible applications for them.
So I was looking for ways you could use these tunnelling structures, and there was all kinds of transistors structures that looked interesting, and I wrote a few papers on those. But my supervisor had a project with the Atomic Energy Commission of Canada. That was to make long life cardiac pacemaker batteries. What the project he was working on was, you used a radio isotope to blast electrons onto a solar cell, and it provided power as long as... Well, I guess 10 years, that was the type of time frame that we were looking at for the project. And he had a couple of students making conventional cells for that structure. But what I'd shown was these MIS structures, if you designed it properly, you could get structures that were the equivalent of a PN junction dioxide, except he didn't have any diffusions or anything.
So in some aspects, it was cleaner than a PN junction Diode. And when we used them as generators under these radio isotopes, we got much higher voltage from them than we did from the conventional PN junctions. And just looking back at some of my results now, it was about 670 millivolts under one sun equivalent – that was the types of voltage. And so I was getting in that era when the best conventional cells were probably doing 610 millivolts or something like that. So that's the advantage I got over the other structures that were looking at being generated. So I said, oh, I'm onto something here. The only trouble was that you could blast the electrons through the metal contacts that you're using the top of the cell, but light had more difficulty getting through the metal than electrons. So when you're using electrons as your source of photo excitation, the metal in the way wasn't a big problem. Once you tried to use it for light, you had to find a way of getting the light through the metal. So you can make the very thin, which is very difficult, or you could use grading structures.
So that got me in grading structures for photovoltaics.
[00:12:26.220] - Torsten
And then you return to Australia and the UNICEF, and you formed a team that really shaped much of the industry right through today. How did that team come together? How did it evolve when you returned to Australia?
[00:12:40.700] - Martin
Yes, I returned to Australia in 1974. So this is my 50th year at the University of New South Wales. It was at the beginning of the large programs, particularly in the US, for looking at the terrestrial applications of photovoltaics because of the oil embargoes that had happened starting in 1973. The US, under President Nixon, actually started "Project Independence," which was a big program to wean the US from Arabian oil. It was the official target. Some of the photovoltaic enthusiasts, which were mainly space (and the space cell researchers that were working in the space industry or space cell engineers), probably they pushed the cause of photovoltaics. Back in that area, it was thought nuclear was going to save the world. And one of the nuclear engineers said, "that's going to have as much impact as a flee on an elephant's back." Those were the comments of the era. Anyhow, they managed to push it through. And of course, the cells were horrendously expensive – so it was a bit of a push. But they managed to convince people that there was some chance of reducing it down to competitive costs.
It was quite a big program. In fact, just in dollars of the day, there was more money going into photovoltaic research in the US than there is now. That's without counting for inflation. So it's something like USD200 million at its peak of that program going in for a year. Quite a big program. It was managed by the Jet Propulsion Laboratory, who I understand supervised the program to put the man on the moon. I'm not really an expert on that topic, but that's what someone told me. They used some of the strategies they had used in developing technology they needed for tasks like that and applied it to photovoltaics.
They'd get people to submit proposals on a topic they were interested in, and they'd pick the three best. And they'd have those three companies competing against each other. And then one by one, drop them off. Some funding had stopped and then more funding went into the ones that were still left standing. That was a very large program covering just about every aspect of photovoltaics. A lot of the technology that's around today, like Continuous Czochralski and diamond-wire sawing and fluidized bed reactors, they were all studied under that program and actually got those technologies started.
[00:15:09.310] - Torsten
You got funding through that Nixon program?
[00:15:12.780] - Martin
No, but it made it a hot topic in Australia as well.
[00:15:16.650] - Torsten
So Australia said we should work on this area as well.
[00:15:20.050] - Martin
Yeah, well, so did Japan. Japan was probably second in and then Europe and then Australia, probably in that order. But Japan started a very big project as well, the "Sunshine Project." It was really very exciting times to be involved in photovoltaics because there was so much effort. And in the US, it was all the good labs involved in the program, all the top researchers in microelectronics had moved in there because there was money available for research in this topic area.
But NASA, who was always also doing its own programs independent of this, had decided that they wanted to improve silicon space cell efficiency. And they had determined that improving the voltage was the way to go about it. The textured cells had just become available in '74 using pyramidal texturing and so on. So the currents had jumped up. They figured the currents were just about as far as you could take them. And back surface fields had been incorporated in the early 1970s. So they figured the current, not much more you could do with that. But the voltage, like 610 was regarded as pretty good value. And you do the calculations and you should be able to get over 700 millivolts.
So NASA started a program to try and improve silicon cell open circuit voltage to 700 millivolts. They had their own researchers working on it and had about six or seven other subcontractors, just trying to, like Westinghouse, play a role. I can remember a couple of universities working on improving the voltage of silicon cells. We weren't initially part of the program, but with our our tunnelling structures, we were able to get well ahead. And of course, the voltage is dead simple to measure. We didn't need to worry about the light getting through the metal, even to measure it. You could just collect the current peripherially and work out what your effective circuit voltage was. Or we use very coarse grading structures.
We weren't worried about the current density because the voltage only depends logarithmically on that. And we had 50 or 60 millivolts advantage for the rest of the mob because we were using essentially TOPCon back then. We streaked ahead of all these NASA subcontractors, and that brought our work to international attention. I've set up this lab and there was literally no equipment at all to use. But these tunnelling structures were really very simple to make.
[00:17:48.700] - Martin
You just take a P-type silicon wafer, heat it in a furnace at 600 degrees to grow the oxide, and you don't even have to have oxygen in there to do that. It's more the temperature that determines it. And then evaporate aluminum onto the front, and that gives you your rectifying contact. And then you mess around with the back a bit to get rid of the oxide and evaporate aluminum onto the back. And then you've got a solar cell. Or maybe we put the aluminum on the back first, come to think of it. And during the firing of that, we actually grew the oxide on the front. So it was even simpler. But this other structure was dead simple to make. So we only need fairly rudimentary equipment to do that. And then we just need lamps and stuff to be able to measure the voltage. So it was very, very simple.
[00:18:36.900] - Torsten
So at this point you are on the global map for solar cell excellence.
[00:18:42.270] - Martin
Yeah, we were. In '80, we got an invited paper at the IEEE conference. And in those days, no one used to get invited papers. So we might have been the first ever invited paper at that time to talk about our high voltages. People could see we were onto something that was going to lead to much better cells down the track, I think. That helped us get local funding. By then, Australia had started to crank up its solar program, a little bit later than the US and Japan, but it started its program. So there was money available locally.
And then in those days, applications for funding were generally sent overseas for assessment. And everyone overseas had heard of what we were doing and everything. So we got good assessments back. And we were able to buy things like evaporators and furnaces and those things. And we said we're going to make a good solar cell. Now we've got these high voltages. We got very close to 700 millivolts. So we said we're going to turn this 700, really 700 millivolts, into an advantage. The next best was less than 650. And at that point in time, maybe 640 or something. We had this huge advantage in voltage.
And we said we've got to be able to make a better cell. It requires some high, fine fingers and conductive fingers and anti-reflection coatings and all that. So, much more sophisticated processing. But by then we built up the funding to allow us to buy that equipment and install it. We got our first world record in 1983 using these tunnelling structures. The best confirmed efficiency then by present standards was 16.5%, and that was Westinghouse or someone like that. And we got over 18 %.
[00:20:33.320] - Torsten
We'll come back to technology later on. I was just wondering, again, because you mentioned that nuclear was considered to save the world. But when you look at where solar is today, did you envision back then that it could be so cost-efficient and then contribute, to such an extent, to the global energy supply?
[00:20:52.610] -Martin
Large-scale application was the driver. So the guys involved in setting up the solar program had figured out that was 50 cents a watt, in the dollars of the era, the early 1980's dollar – that was the cost you needed to compete with for conventional large scale generated electricity. So 50 cents a watt was the official target of the large US program. And the aim was to get there by 1985.
[00:21:19.510] - Torsten
And there was a clear roadmap to these 50 cents? What was the concreteness that you could take it from where you were to these 50 cents?
[00:21:30.840] - Martin
So there was a number of parallel paths, like Continuous Czochralski was one, and the ways the ingots are produced. So it was very similar to the type of thing today. And ribbon growth was very popular as well. So a lot of effort and funding was put in mobile solar – and Westinghouse, where they're big proponents of ribbon growth, they had two completely different ribbon growth methods that were highly funded.
And then a lot of interested cruder grades of silicon, so metallurgical grades, silicon refining up to something useful for solar cell fabrication was another goal of the program. So I had all these parallel strands. And then as they developed, and they had groups just doing the costings of the strands. So there was independent costings of the strands going on as well.
[00:22:23.970] - Torsten
And 1985, so it was within a few years. Did you actually achieve these 50 cents per watt? Could you find that spot on the learning curve?
[00:22:35.510] - Martin
There was an Iranian hostage drama that helped dispose of President Carter and President Ronald Reagan got involved in, and he wasn't the least bit interested in renewables. So his big funding program was "Star Wars," which was a missile defense system. So all the top US semiconductor people that were working on photovoltaics got dragged off into Star Wars programs, because you get funding to support the graduate students and things.
But Australia being a little bit slower in taking up programs, we're also a little bit slower shutting them down. So we gained a bit of momentum through the first half of the '80s, whereas a lot of the US groups fell by the wayside. So they had been established and survived right through from the '70s to the '80s. Adolf Gerdts, I also got to set up the Fraunhofer Institute in Germany, and it survived through, and we survived through. So we're the three groups that I'm aware of that were making an impact in the photovoltaics area that survived right through from the '70s to the late '80s, to '86, I guess it was with Chernobyl, which was really kickstarted the solar programs again then, but it was a bit of a barren era.
We managed to get US government funding, even though it had been cut back during that era. Because we were setting all these world records for cell performance, we had what were declared to be unique capabilities, which made us eligible for US Department of Energy Funding. It wasn't a lot by US standards, but the Australian funding sources were a lot smaller Solar. So it gave our program a real boost, being able to tap in both US and Australian funding. So that was an important reason why we were able to survive through the '80s.
[00:24:26.910] - Torsten
Solar made a fantastic evolution, and you followed it through right from the beginning, or almost from the beginning. I guess Becquerel was a bit earlier than you. So we've seen this growth, and we are now at least from the production capacity, somewhere around one terawatt and with shipping around, let's say, half a terawatt.
At what level do you think there will be a saturation of the manufacturing base? They usually bring forward three terawatts will be the level annual shipment where we saturate. I just talked to Jenny Chase and she thinks that one terawatt shipping per year will be the saturation level. What's your take?
[00:25:13.720] - Martin
I like Pierre Verlinden's analysis: you get up to the three terawatts, it is enough to replace 90 terawatts, or something, a year of totally installed photovoltaics. There's this thought that photovoltaics is already, according to the International Energy Agency, providing some of the cheapest electricity in history. But the exciting thing is the costs are still going down. If we can get tandem cells working, the cost is going to go down even further.
There's still a lot of development to occur with just standalone silicon cells, I believe. There's some people talking about this third energy revolution where we have this super abundant energy, 100 times more than we use now, available from solar and wind and so on that are going to be incredibly cheap. So what does that mean for the future of the human race? So they started to ask questions like that. Maybe we will open up a completely different reality where we conceive of energy in a completely different way, which might mean that the three terawatt level just gets completely swamped.
[00:26:23.320] - Torsten
And when you look at the learning curve, we touched on it already a few times – now we see these record low prices of, let's let's say, 10 cents per Watt. Is there a bottom line? It's so hard to imagine that it keeps going and going, but in a way it does. And the slope is still at 2006 on the 40% learning curve rate. Overall, it's 25% or something, but just over the last 20 years, it's been 40%. Where does it take us? Will there be a level where it converges to a stable level?
[00:26:54.110] - Martin
We could see five cents a Watt pretty soon, actually. According to PV Insights, they're selling modules now for eight cents a Watt. So five cents is not all that far off. It's a little bit hard to say. So I'm on record in 2002 as saying a dollar is as cheap as you will ever get with crystalline silicon. So I'm a bit careful about making projections about lower costs. I did say, however, some time back that would be at 10 cents a Watt in 2025. So we're there – I was correct on that one, despite COVID.
So I got a bit worried when COVID started knocking the prices around, but we made 10 cents a Watt by 2025 quite comfortably. So, yeah, like I said, we're definitely going to see five cents a Watt. Are we ever going to see one cent a Watt? Maybe. This guy, Ramez Naam, he used the terminology, solar is going to become insanely cheap. So like I said, one cent a Watt is probably insanely cheap to my thinking. We might get there because the industry is still in its infancy. I don't really see silicon lasting forever.
If we can develop tandem cells, I think the ultimate technology would be all thin film, which has the potential to go really low cost. In one sense, it might be feasible with a 6L tandem made of thin films, or something like that.
[00:28:21.310] -Torsten
I want to sidestep a little, because we're talking about taking decisions in which, particularly as a professor, a leader of a group, you have to make decisions on which route you take. What was your basis for decision making? When you look back at your general technological route you took, which principles guided you? How did they evolve?
[00:28:40.660] - Martin
I guess we're always guided by theory. So we try to interpret the experiment results we get in terms of what was happening in the cell and what was causing this and causing that. And even if the theory was wrong, it gave you a strategy for changing things and trying something different.
But if one course of action proved to be a dead end, we then look for another course that would prove more productive, I guess. A difficult area, I guess is the solar cell processing. We do a lot of work with optics, of course, in making the cells, and that works more or less as it should. If you design something on paper optically, you make it. It generally works quite close to what you predicted. But with solar cell processing, there's so many interacting parameters that trial and error was the desire about what processing courses were productive and which were not. I was fortunate in having a strong group like Stuart Wenham and Jinwa Chou and Ewa Wang, they were leading the high efficiency cell program for most of our history. And just a group of four of us sitting around, we just brainstormed.
Jinwa and Ewa were very good experimentalists. So we do little sub-experiments around the basic processing, just to get a better idea of what that particular aspect of the cell design was contributing to, and so on. We consistently made progress for close to 30 years.
[00:30:15.020] - Torsten
The theory-based approach, that's interesting, because we can trace back right from to your master thesis, where you did the simulation program. So you're really always about the core physics. You could think that's natural, but I would assume that not every group leader really goes back to these basic simulations. And you just mentioned a number of your coworkers. How did you meet?
[00:30:38.510] - Martin
Yeah, well, Stuart did electrical engineering at the University of New South Wales, and I was one of the lecturers in the courses there. So the SolarCell course was a postgraduate course, and it formed the basis of my book I call my Red Book, "Solar Cell Operating Principles," and so on. That was basically the lecture notes from that course.
But Stuart, as an undergraduate, and he was top in his year, he got the university medal, which means you come top by miles, type of thing. He wanted to do this graduate course, and you need permission from the lecturer to do it. He came to see me and asked if he could do my graduate course. So we got to know each other. And then he topped that course as an undergraduate and got quite interested in solar cells. He was an electronics guy. Before then, he was very good at designing electronic circuits and all that stuff. So when he graduated, he got 13 job offers. But Australia had just set up its first solar cell manufacturing line. This was in 1981. It was in the process of setting it up. And my first PhD student, Bruce Godfrey, had got headhunted to be the managing director of the company setting up that line.
Stuart got 13 job offers, but he accepted this one with the solar cell company because he was interested in that technology. And this company was called Tideland Energy, and they had bought the very best technology that was available in that era, which was the screen printing technology that Spectralab had developed in 1974 and then in subsequent years as part of this large US developmental program. But they used texturing of the cell and then screen printing of the silver paste on the top and screen printing of an aluminum paste on the rear. So it was like the modern back surface field technology that prevailed until PERC took over in 2018. I think PERC finally bumped the dolphins perch. But that was basically Spectralab technology that leaked out from Spectralab in several different ways. Tidaland actually bought the technology from Spectralabs, so they were on a broker. Stuart got to interact with these people that had developed the screen printing process. So he's expert in that. He was basically the only technical person on the whole establishment that was sitting up this production line in Australia. Maybe it was 100 kilowatts production capacity, which was massive for that era.
[00:33:09.790] - Torsten
You just mentioned PERC, I made my very own first solar cell with Andrew Blakers in 1993, 1994. We did it on these two-inch wavers, and I think it took me three days to process four cells or something, right? Back then, I would have never, ever imagined that this type of technology, because it looked so complicated – would it ever make it into production? And now we are almost in the post-PERC, TOPCon phase. What was your thinking when you found PERC? Did you think it ever make it into production?
[00:33:45.180] - Martin
We were told at the time we'd never make it into production because it was too complicated. So, yeah, don't listen to the advice that you get given, would be my recommendation. But we actually started more with the TOPCon structure. So the problem with the metal tunnelling structures was you heat it up a bit and the metals were low work function metals that you needed for P-type substrates. And they're very, very active with oxygen. So everyone's saying, Oh, you're wasting your time with that because the metal is just going to eat through the oxide. So we said, well, we'll show them they're wrong.
So we did some polysilicon tunnelling structures in 1981. And so that was the first TOPCon sales to be made. And we published it as well, although subsequently people have missed it. But the processing was horribly complex. The polysilicon, we had to use someone else's gear and you had to book it up weeks ahead and all that stuff. Whereas we are used to just pumping these devices through continuously, having high throughput was always a feature of our fabrication. So throughput, we realized was important. We made our world record cells using this tunnelling structure.
And then Stuart and I were annealing them. So Andrew processed them. And Stuart and I were tweaking them in by annealing them in what was really a pizza oven. Just give him a little gentle anneal in just so we can tweak their performance in a bit. And it turned out we could do a little bit more than what we thought. So we figured, well, the metal must be getting through, probably in filaments, through the oxide, the temperature we're taking to in the length of time and everything. And we're still seeing a similar performance.
That led us to PERC, which was basically a small area contact technology. So we got away from tunnelling structures. I'd suggest that the tunnel structures on both sides back in 1976 or something, or other. When I thought of it, I thought that's got to be the ultimate solar technology with a P-type tunnel contact on one side and a N-type on the other side, a slippery side, potentially a gradient. So we actually switched from TOPCon to PERC because it simpler to process. We were finding with the TOPCon, we needed different tunnelling oxide thicknesses from the field oxide.
We'd have a thin oxide underlying the anti-reflection coating, but that could be a lot thicker than your tunnelling oxide. To really tweak the surface recombination velocity, we needed two different oxide thicknesses. Whereas we went to the PERC, we could get by with just one, and the processing became a lot simpler. There wasn't need for the same level of realignment and all that thing. We switched to PERC, and we got our second world record at the end of '93 using a PERC-type structure in a small area contact. In those days, we were getting about 650, 660 millivolts voltage. So the difference between PERC and TOPCon, you couldn't really see it when you were limited to that type of voltage. But they were miles higher, those voltages, than what anyone else was doing, and that gave us a fixed advantage.
[00:36:57.530] - Torsten
Could you really envision that the these technologies would make it into mass production? Was that really in your mind when you went to the lab and then or analyzed the function? Could you see there would be gigawatts of production lines with these technologies?
[00:37:12.820] - Martin
Well, it probably happened a bit more quickly than what I thought. We had some PERC patents that we let go in about 1995 because there was other stuff we were doing. The commercial arm of the university was getting a bit sick of all the patent costs they were paying on our account. And we were sticking in more patents. And they said, well, do you want us to pay for these new ones as well as your old ones? You're going to have to make a decision. So we decided in 1995 that PERC was too far off for those patents to be viable by the time it got commercially introduced. So we let the PERC patents go then. I guess we thought we'd find simpler ways of making the structure.
[00:37:55.750] - Torsten
You already mentioned that there were other trends coming up. There was your silicon and glass activities. And then for some time, there was a third generation PV as a big trend with lots of different subtopics. There was then the PERC and TOPCon. So things came and go. You said trends yourself, and you saw also other things that failed, just like when we had our few encounters with the silicon on glass. I worked on microcrystalline and amorphous silicon, you on the CSG, solar activities. What's your general learning when look back at all those technological trends, how to make decisions? You just said it's so hard to predict. So you actually decided to go for the new topics and didn't really envision that PERC would make it. Is there any learning in it or is it just the learning that it's hard to predict the future.
[00:38:49.360] - Martin
I guess the future might evolve completely differently from what the experts say. So in the '90s, everyone thought the future had to be thin film cells. There was plenty of people working on amorphous silicon, like you mentioned, and that was thought it was likely to be the winner. But I didn't like the technologies the US groups were concentrating on, cadmium telluride and CIGS, mainly because of material abundance issues and toxicity issues and so on. I said, this is silly. I'm not putting your effort into this.
You can't tell until its in front of your eyes, are you ever going to be able to make? I calculated about 30 gigawatts were needed. So in a year, which first all is about halfway there, they might get to 30 gigawatts. But I could see the industry was going to be much bigger than that. So that's just a drop in the bucket, compared to what's going to be needed. And then with other things, you've got the issue with the indium and selenium and so on. So it's not a great bill of health. And you're using cadmium sulfide as a emitter layer in those structures as well. So I didn't really like them.
So I said, we're having all this success with silicon. And with work on light trapping, all of a sudden the silicon cells became 50 times thicker than what everyone used to think they could become. So like a micron of silicon was all of a sudden 50 microns with the 50 times the light trapping. So we were using those in our high efficiency cells. We were using light trapping early on. We said, Oh, we get this light trapping working well, which we did in the CSG. The optics were really good, but the electronic quality of the material, we could never get up to adding liquid to get the 15% efficiency that was our target. So we just thought, Well, we can make the silicon on glass technology work, different from amorphous silicon, which we thought, iI's got the stability issue, it might never overcome them completely. Although I was probably pretty confident in the '80s that they would be overcome. Guys like Stan Ofsynchki were saying, All you have to do is add this or that to the amorphous silicon, and it'll be perfectly stable. So with those type of comments going around, you thought, Oh, someone's going to crack this.
And it never seemed to be solved completely. So we just thought we'll have a crack at this silicon on glass technology. And we got it into production. And we were following First Solar, they were a couple of years ahead of us. They were starting at 5% efficiency, then 6, 7, 8, 9. We said, all we've got to do is follow them and keep in contact with them and it will be right. But we could never really get above about 10% efficiency with that silicon on glass technology. So it was not high enough. And then when my students started setting up manufacturing in China, that pulled the rug out from under all these thin-filled technologies.
[00:41:47.740] - Torsten
My standard last question is, what does it take solar to take it to the next level? So we talked a lot about technology. It can be from any other area as well.
[00:42:00.040] - Martin
I think it's mainly political now. The US needs to be more aggressive in its uptake of solar, because the cost of solar panels in the US is just horrific by Australian standards. And then some of my friends, when they put solar on the roof of their homes in the US, it's just mind boggling how much they have to pay, like several thousand US dollars a week. Whereas there's companies in Australia advertising six kilowatt systems for 3,600 Australian, which is like 2,000 US or something like that. The same system in the US would cost you more like 20,000. So it's just mind boggling how much they have to pay for their photovoltaics there. They need to get on a bit with installing solar. They're lagging behind. I think they're below the world average in terms of the amount of solar electricity that is being fed into their system now. I think it's still below 6%, whereas the world average is now above 6%. And both China and India are above 6%. And it was much harder for them to get there than the US. I'm more for silicon, crystalline silicon junctions.
[00:43:10.480] - Torsten
Martin, thank you so much for coming onto the show. Thanks for sharing your solar journey. It's been a real honor. And thanks a lot for what you've done for the industry and also the young scientists, including me, indirectly through your book and all the other solar engineers around the world. Thanks a lot, Martin.
[00:43:31.870] - Martin
Nice talking to you, Torsten. Goodbye for now.
Comments