 Yeah, it's a pleasure to be here and thank you for coming. So numerously to my lecture, I hope I won't disappoint you. So this afternoon I'd like to present on two topics. One is our passion, the molecular photovoltaics. But as Paul mentioned, this has gone through some maturity. And for example, you see here one of the applications of this invention is these glass panels that would use electric power from visible light. And this is a filling station. It's a filling station, it's an electric power station for these electric vehicles that you're very familiar with here in California. So I hope you're still going to use our technology. So yeah, and so these applications come out of fundamental research. And so in the first part of my lecture, I want to take you the path of this invention, how it emerged. Some people ask me, well, how did you invent the solar cells? And if I woke up one morning and I had this invention in my head, that's not how it happened, okay? And so, but it's also clear that one of the most exciting developments, that recently, if not the most exciting development in photovoltaics, the recent years has been the postcard solar cell. Which is shown here, one of the recent papers we have. This is shown here on the left. And these are really amazing materials. They span of six to seven years. The efficiencies went from two to 22%. And this has never happened in photovoltaics. And in addition, they can be made by solution processing. So it comes very close to the dream of having a photovoltaic paint where you would just take a brush and put this stuff on a wall and get the power out of your, out of your thin film. And so, so the second part of the lecture I'd like to take you along the line of development of these postcard solar cells. So, yeah, so this is where we work at the Lac-Lemon, the lake of Geneva in Lausanne, and here's our lab. And so, energy certainly is a big issue. This is what the lecture will be about. And here's a photovoltaic. Photovoltaics have been growing. And last year we had this about 60 gigawatt peak power installed. And so that, why is it gonna go down next year or this year? Well, here's the problem, okay, the feed enterals. It's all said here, the story is being told, have been decreasing. And so it shows you this technology. I mean, it's mainly silicon, 90% silicon. It's still living on subsidies. And if subsidies go down, then the sales drop. So this has to change. So there's room for developing technologies that, number one, are not dependent on subsidies. And you also need just the sheer volume. It sounds a lot to 58 gigawatts sounds a lot. But just think about it. You're consuming 16 terawatt, okay? That's permanent. And those 50, that's peak power. So divide by five, you get the average power. So it's actually 10 gigawatt average power. So what's missing a factor of 1,000 here, okay? So if we would install at that present rate the photovoltaics, it would take 1,000 years to get to those 10 terawatt to make an impact, okay? So definitely there's room for new technologies and the good news for you folks is that the funding will be there, okay? Over one trillion dollars is expected to be spent in the next 20 years on solar PV along with other renewables. And so who will be the winner? Well, this is the famous triangle. So efficiency, stability and cost, we know what we can, okay? So I will come back to this triangle at the end of my lecture. And so, well, one winner is certainly at the moment for the synthesis. For the synthesis, it has a 1,000 terawatt power storage, okay? Just think about the 50 gigawatt or the 10 gigawatt we have with the PV, okay, 1,000 terawatt. So that's what inspired us, the light reaction for the synthesis, which works well since two billion years. And so we have, our initial thinking was to try to mimic that reaction by using molecular absorbers, it's molecular photovoltaics. Using molecular absorber to interact with sunlight. And so I will be putting this machine in motion in a minute. So it's a molecular sensitize, not a diode absorbs light, a photo diode is a molecular sensitizer. And we get this photo coming in, pumping the electron up, injection, collection, then the electron does work. Then we bring it back to the sensitizer. And very, very important is this redox shuttle. It's for why we got stuck with one single redox shuttle, the iodide triadide. And you see the problem with that, it's in the middle of the gap. And so all this energy is wasted here. So for many years we stuck with 12% because of that major loss here. But this has changed, and recently there was a breakthrough with the new copper complexes that work beautifully in the solid state in the liquid, and I'll share some of this with you shortly. So the outline then, we will talk about this molecular photovoltaics, and then turn to the perovskite solar cells. So it makes big difference if you put sensitize on a flat surface or on this nano or crystalline film. The yields in photocon go up 10,000 times. It's amazing. I mean, this piece of material cost $50,000 to make, okay? And that's cost about half a dollar to make, but it gives you 10,000 times more power out. So it's done it, and why is it so good? Well, it's your probe light, and the important thing is you're able, not only the probe light, but you generate charge and you collect those charges, and the product is close to 100%. The EQE is shown here, over 90%. You have it on a flat surface, 0.1%, so that tells you the story. This system is working well, the mesoscale is doing a miracle for you, okay? And so we explored what the reason for this was, and I don't want to go through too much of the theoretical details here. So what is the scaffold doing? It's a false sensitizer, the nanoparticles are stacked so you get the light absorbed, although there's a monolayer on the surface of these particles. And here's the important second function, scaffold extracts selectively electrons from the excise sensitizer. And the nanoparticles clean the electronic charges. Just think about it, under full sunlight, each of those nanoparticles has about 20 electrons inside. And so usually we think, well, how is that possible to squeeze 20 negative charges in such a small particle? There's a Coulomb blockade effect. But no, it's possible because you got those charges cleaned by the irons on the surface. And so here it shows you the staining of the film, of the nanocrystalline film that happens spontaneously. It's a self-assembly that happens spontaneously. And it darkens your film. Here we have this ruthenium dye as a sensitizer that hooks on the surface through the carboxylate group. We can image that those dyes, this has been done on a single crystal. And you can see how the sensitizer sits on the surface. You can also massage these images and find out more how, for example, you see a dimer formed by between two sensitizers. And we find that the carboxylates go on by dente. They bridge two titanium, adjacent titanium lines on the surface. It's a bi-dane-coordinated binding. And so all this information has given, oops, wow, wow, somehow it was there. I'm sorry, I have to go back to PowerPoint and went out, don't know why. Okay, so now, let's just quickly run through these slides. If I tried to do anything else, I might upset the, okay, I think we got just to this stage where I told you we can massage these SCM photos and we get information on the binding. And here's a quantum mechanic calculation that shows how the charges is injected and you have to medallic and charge transfer on excitation. And the charge is injected within a few femtoseconds. But then nothing happens. I mean, you see a flat signal here and that tells you the back reaction is a much slower timescale than the forward injection. That's the trick. Actually, there's a factor of 10 to the nine in the rate of the back electron transfer with regards to forward. And this is what made that photovoltaic system work. It's a kinetic argument, a kinetic observation. And so, injection, regeneration. So the important thing is that you can collect the charges before they recombine. This is the key thing, the transport time recombination time. If your transport is 100 times faster recombination, then you collect 99%. And for many systems, it's possible to do that. To get those ratios that you need to collect fully the charges that are injected in the nanocrystalline film. That was one of the physicists' main argument against this concept. They said, well, we believe you can absorb light, you can inject electron. But you will never collect those charges. Because there's no field to separate the positive and negative charge. But that was wrong. We were able to collect very, very efficiently those charges. And so, this paper came out in 1991. It has received over 17,000 citations since then. And it was cited so highly because of the change in product. We went from the flat PN junction to this three dimensional nanoparticle network that's served as electron selective contact. And so, yeah, I should mention to you, I just found out that I don't want to practice my citations. So that, so this thing was, I ignored for a long time that we were actually, it was the hundreds of ever, ever published scientific papers. And actually, our citations have gone up to 17,000. So the rate of citations is 2,000 a year of this paper still. That's quite a score. So that means every day you can calculate, we get eight citations. So let's go away from the citation thing. But it's also this whole thing spread. And so, these nano column dots, they're used as a replacement sensitizer. And we have the solar state system. And I should mention to you that the pulse guide solar cells started out as a quantum dot sensitized mesoscopic solar cell. It's like a DSC, but replace the dye by a quantum dot sensitizer. You get the pulse guide solar cell how it was originally. And also the patent applications, we had first one patent, it was totally ignored, we couldn't even license it to anybody. Then we took a few more and then this whole thing took off. And so we were able to license after that interest took such big proportions. And so, yeah, today the original sensitizers that were ruthenium complexes, they're replaced by the so-called donor acceptor. I don't want to dwell too long on this, but you will understand if I have a structure like that where the molecule has a donor moiety and a chromophore and an acceptor. When you excite the dye, the electron goes from top to bottom and it finds the particles there, it gets injected. But the positive hole that remains behind is on the donor is up there. So if the electron wants to come back out, which it loves to do, it loves to wrinkle back, but it's far away, okay? That hole is far away, and that makes the charge separation so long lived. So it's a molecular design of your dye that helps you to achieve those great photovoltaic assaults. And so we have this green dye which was designed in this fashion and did some molecular and amaz calculations and computational help to design sensitizers like the one we have here with this extra acceptor. So this field has taken off in the sense that it's now possible on a computer in silico, you can design your dyes and you can predict with a high accuracy of point E, point one EV, the absorption properties of those sensitizers. And so recently, as I mentioned to you, we found these copper complexes and the differences here in this slide, you see when you excite and inject then the regeneration of the sensitizer costs only 100 millilectron volts. So, it's like a silicon photovoltaic, okay? There's very little loss and not surprisingly, that has helped to move forward with that. These are now solid state, this copper complex actual hole conductor and it's in the solid state. And so solid state cells have been made. And here's the latest invention. We have co-centers sensitized, there's two sensitizers. And this is a photovoltaic cell. It doesn't look like one, but it's a photovoltaic cell. And believe it or not, it beats gallium arsenide in performance, okay? And so the color is the combined color of these two dyes. And here you see the impressive performance, the EQE goes to 90%. 90% of all the incoming photons are turned into electricon, okay? You see the high voltage we can get out of this. And so the verdict is here, 1,000 looks in ambient light. You get more power out of this device. And I'll show you later the first applications. We have ebook readers that have eternal life because they get all the time power, so there are quite a few applications that come out of this new discovery. So now let's talk about the pulse guides. And yeah, we have the dial cells type system. The pulse guide cells are kind of dull, they're always black, okay? But that's what you get if you want to absorb all the sunlight, they're invisible, they will look black. Why does it do that to me? Somebody must have bad intentions here. And so, okay, it's just to go through the routine one more time. So the sad news is I have to go through the whole thing now. And because I distrust your electronics. So we were here at this point and now let's go to the pulse guides. Go to these, so strong absorbers. All of this is typical for direct bandgap material. It's also a material that doesn't form excitons, so the excitons dissociates easily. But the real interest in this material is it comes defect free. It has a valence band maximum that is formed by undibonding orbitals. And so suppose you take a silicon cell, you break the silicon-silicon bond. It makes your defect right in the middle of the bandgap, which will be recombination center. If you break the bond between the lead and the iodide, you actually stabilize the whole system. Because you break something that is undibonding. It's kind of a mind-boggling argument, but think about it. It's true, okay? And so we have very few defects in the bandgap. And so these materials, electronically, they are very, very interesting and high quality, even though you don't grow single crystals. And so as I mentioned to you, we have these quantum dot cells. They started with the quantum dot cells. And we have here those quantum dots of the pulse guide this time. Sensitizing the TIO2, the trouble with these concepts was that they felt dissolved in the electrolyte within ten minutes. You had ten minutes time to take your JV curve, and then it was gone, okay? No more sensitizer. And so, of course, that wasn't a very sustainable solution. But it turned out if you swapped the electrolyte with the whole conductor, then the solid state whole conduct, then the whole thing became much more stable. And this is what happened. Two papers appeared almost simultaneously. And this kind of put the field on track with this whole concept to have a whole selective contact, solid state, electron selective contact. And you can see efficiencies going up and the number of papers. Now we have 2,000 or 3,000 papers published. A few years ago, we had only one paper published. It was totally ignored, okay? And so incredible evolution of this field. And so you see why I'm sure that we have 3,000 papers. We have 2,700 citations. And each citation is a paper, dear friends, okay? So we must have at least that many papers out. And so we also start infiltrating the pore space. And this evolved to, finally, there is now two configurations. And we have the mesoscopic, which always has a capping layer of perovskites on top. And this is a planar configuration. And so which one is better, close, but so far the highest efficiencies certified have been reached with the mesoscopic film. And so also the mixtures, we have these cation mixtures that today's formulations use. And the story is, we started with the formamadenium, which is not stable as a perovskite phase at room temperature. But if you put a little bit of methamoronium in a mixture, then you get a stable phase. It was a discovery that was made by my graduate student, Norman Pellet. And we have sat here in this paper. This concept has not been applied previously in perovskite-based solar cells. It shows great potential. But everybody ignores this paper, okay? So I had to beg on the table and say, hey, every talk I will show the publication. So I've done that now, so we're moving on. And so the interesting thing is if you take a cesium and formamadenium, then you have two materials that are unstable as perovskite from the yellow phase at room temperature. But you mix them, you get the black phase. That is magic, okay? You can even mechanically just grind those two materials, get a black thing out, okay? So black magic, and also these mixtures are more stable than the methamoronium mixtures. Because methamoronium is kind of unstable, it evolves easily as it forms methylamine and H-iodide. So it's not to be recommended as a stable perovskite component. And so then triple cation started, so we went from double to triple, and we found out why these things stable as mainly antropically driven. And then here's the triple paper, cesium. And today we have four, so I have to let's move forward. This is actually four cations. It's obedium, cesium, methamoronium, formamadenium. Now you really need four. That's how this thing evolved, okay? The four cation formulation gave us a very good voltage over 1.24 volts. And you see the stability amazing, I mean 85 degree plus full sunlight. And we run the cell on maximum power tracking. So we were drawing power at 85 degrees under full light. This is a double ramor stress, okay? Nobody does that usually to a photovoltaic cell, and it survived. And not only that, these cells are mislight and miseth. And that's to be expected. A photovoltaic cell has a high voltage at open circuit, will be a good light emitter. This comes from Eli Jablonovich, but really it goes back to Hoss's work, and a guy, the student of Melvin Calvin at Berkeley. And he came up with this equation, which is easy to derive. And so if your voltage is an ideal voltage, that's about for one sun, it's the band gap minus 300 millivolts, roughly speaking. So if the pulse guide has 1.6, EV band gap you will have a maximum voltage of 1.3. That's your theoretical one. And so if you get to 1.2, you should have a quantum efficiency of a light and an electron luminescence of a few percent. And that's what is observed, we get 4% quantum efficiency. So you can run these systems as a light emitting diode or as a photovoltaic, the same cell. It's the same cell, okay? It's the same amazing thing. And so here are the four percent, you can see them on the scale. And so this becomes a very good quality check for photovoltaics, this light emission property. We move forward and so today's efficiencies are 22.1. And let me just point out to you that the blue points are polycrystalline silicon, so the pulse guide have beaten actually polycrystalline silicon, the market leader in photovoltaics. So where are we now? Who's winning the race? Very early, I mean, this still is a laboratory cell and we have to worry about stability. Cost wise, it looks actually pretty good. Because, I mean, here's the annual analysis, which tells us, let's see, the lowest cost of all photovoltaic technology. And you have here the materials cost. This is the most expensive part here is the hole conductor. But this will change the cheap hole conductors now around. So at the end, I mean, you see, this analysis expressed the LCOE, the levelized cost of energy. That's what counts. How much do you have to pay for one kilowatt hour electric energy? And so here, the prediction is three cents, three cents. And that's a very competitive price. It assumes a 12% module efficiency and 15 years lifetime. So 12% module efficiency is around the corner. In fact, if you look at the scale-up, there are several approaches that allow scale-up. Here we have the Singapore, I'm just mentioning, it's a module 11%. They have now beefed it up to 12%. So these are all, this is screen-printed films. They don't use any sophisticated technology. The whole photovoltaic device is screen-printed. And here's the disole cell. They are very close in efficiency, quite at this level, but very close. And here's a bigger module in the, actually, this carbon cell comes from Wuhan. And I'm so proud to be associated. She didn't know that's my name here, it's the micro-gratso center of the topic solar cell, okay? That is expressed in Chinese, okay? So I was very proud to be named the honor director of the center. And so, yeah, actually with a fuel generation, I will be very quick. I mean, the first-guide solar cells have been used as voltage source. They have this high voltage for fuel generation. And for a while, we were holding the record. But this, I think now, it's 35 semiconductors have kicked in a little bit higher than post-good. But we have, we're holding a record for CO2 conversion to COO. We started off with this paper, 6.5%. But recently, we have a new catalyst. It's a Cooper's oxide on both, the cathode and the anode. And it's the cathode side is a Cooper's oxide or Kubrick oxide with a tin oxide overlayer. And it makes it selective for CO2 to CO conversion. You can see that here. You run the CO alone, it doesn't make much CO. But with a tin oxide ALD overlayer, it does make your CO very selectively. And so that has been now scaled, and we get 13.4%. That's solar to chemical, okay? If you count all the fuels, including the hydrogen that is formed, it's actually 16%. So impressive efficiencies for solar to fuels with this particular system. And so now, let me finish because I was told to finish in about five minutes from now. And so I will, so I'll just show you at the end some applications. And then there will be, you can challenge me by questions. And so here's some, all the applications we see commercially are from the dye sensitized cells. The first guys, as I've mentioned, they're now being scaled up and we need to consolidate all of this. And the applications, we're talking to Mike McGee today on tandem cells. And so they're very promising perspectives for power skates. But now we're talking about products out of the market. And these are IoT applications, internet of things by companies, like the G-cell company in Great Britain. And this is the famous solar backpack, okay? That was developed by the same company and even our Gorgon, one of those, it was actually I, it was my back. It was the last one he had, okay? He doesn't know that he has my back. But also I gave one to Bill Gates and he likes the technology. Actually, he was very kind to us. He mentioned the, in his statement on the Paris conference, he mentioned the sensitized technology as being very promising for the future and so he will see some applications. So as the chairman said, we have for the molecular photovoltaics, we have two applications. One is the building, the building integration and panels, translucent panels make electricity from sunlight. And believe me, I have ordered myself for my house, now these panels. So I must be convinced that they work, okay? So, this is how they look from inside. They also have a heat reflection effect. Because they have these two FTO layers that they cut the heat inside. And this is our Congress center at night. If you come to Geneva Airport, please admire our panels in the departure level, this is the filling station I showed you. This is on the highway from Byrne to Zurich. Don't overlook our fence when you drive along the highway. It does make us, it does do a sizable production of electricity. This is the Swisscom building and very close to the side. The Korean company Dongcheon is also very active. Here we get to mass production in China. I don't know what that means here, I hope it's nothing bad about me. Because I can't read Chinese. But some of you might understand the Chinese here, so you have to tell me. And so here's the installation of the panels in China. So business is going up. Here is the, this is a very significant development in Sweden. Where the company Accenture has just now raised $40 million to increase production capacity for the products. One of their products, if you go to their website, is that ebook reader and what does it say? E-reader with eternal life, okay? So once you have it, you don't need to worry about battery anymore, okay? So and in case you want to move to Sweden, okay? You are here, there are jobs available in this company, okay? Just in case, okay, just in case. So at the end, very end of my presentation, I'd like to show you something very personal. We have this, had this Expo 2015 in Milano. And there was this big disk that the company SFL from Austria had installed. And these are our glass panels, the 16 kilowatt power. And so I went there on the 11th of May. Now, this is a birthday, not only my birthday, but my wife also has same day birthday. So we'll never forget our birthdays, okay? So my wife Carol was coming with me and we got a birthday cake from the Austrian. But the real important thing was here, the truck. This truck was powered by the panels you saw. And those panels, remember these small molecules we made in a lab? Finally, powering a truck that is driving around. So a dream became true for me. Really, it was a very emotional moment at this affair. So with that, I finish my sponsors, financial support. Most importantly, you better ski in Switzerland. That loose guy you don't see in Lausanne, you're in a gray park there. Okay, so thank you for your attention. Okay, so I think there's time for some questions for Professor Gressel. I'll lead off with one. Okay, so Professor, in your presentation, you mentioned briefly the photo-electrochemical results, the new results of the copper oxide electrodes. I noticed that the pH was quite different between the two, the anodic and cathodic chambers. How practical is that to have to run it under those conditions? Yeah, you need this bipolar membrane. You saw that it was in a design, and what that does for you is it has an inner compartment where you have water, and the water is dissociating in protons and OH-. So there is a potential across that membrane that will help you to dissociate the water, but when you add all the losses up, it's like the water splitting reaction. If some of the dynamics are the same, this membrane doesn't change a thing. It's just that you have to be careful with the flux of ions across so leakage currents, and so this has been a challenge, but we have been able with using cesium as a monovalent cation to prevent the transport of cesium ions across so that you want only protons to move, or each minus of protons. The moment you have other ions, it will mess up the system because it changes this composition on one side or the other side. For CO2 reduction, you need that bipolar membrane because you want to do the oxygen evolution and alkaline solution, and the CO2 is per definition a neutral solution. And so this membrane works perfectly well if you have a pH 0 and 14. Then it's really perfect. It's more challenging when you have a neutral pH 7 and a pH 14. Could you give some idea about dollar per watt compared to other technologies? And the second question is that do you think you'll be able to scale this technology? I wish you can do that up to megawatt and gigawatt level. Which one are you talking about now? The health guys or the product? Yeah, well, I mean, it's happening as we speak. Companies are investing, and I showed you some of the scale up products, Daiso and Singapore. So the various companies, the other day I was in Clarkson, the Clarkson University in Potsdam, and somebody pulled out a health guide cell, a venture capital proposed and invested, and he wanted to show it to me. So we ignore all what is going on. There's a lot of activity out there, I can tell you that. And not only for the cells themselves, but also the tandem I mentioned to you, the tandem where you have a technology like silicon and you build it on this technology. So in my opinion, if you had asked me two years ago, I wrote a review on a health guide solar cell called the light and shade of a health guide solar cell. And there are a lot of uncertainties about stability and the lead problem and so on. Today I think if you had these solar farms and you have a contained glass with concrete support, you're okay. I mean, you have less lead than silicon panels, so per square meter. So that's all reason to be optimistic. For the questions. This is a middle of a follow-up, and you caught my attention with the 3.7 per kilowatt hour, because I know from either straight-up technology or comparisons or modeling studies, that looks pretty good, which was your point. How soon, where and when can we buy this in Europe? I guess one custom thing about your talk was going all the way from deep in the lab at an analog level all the way through to the market. And there are probably a few people who can do that whole spectrum of thinking about what's going on. Yeah, my brain was turning really at the fifth gear. Anyway, it's a good question. So, yeah, it's a good question. We should be very careful not to overstate. When I talk to investors, I always tell the truth. Okay, I never exaggerate. And since I have been investing myself in startup companies, and I know how it hurts if somebody doesn't tell you the full truth, I'll be very frank to you. This still is an early time, but the risk is coming down very quickly. And so, you have seen those stability results. Also, Mike McGee showed me very good stability. So, we are going through those test programs. IAC 6146, or 6146, I forgot now the code. So, there is a test program. And we have seen that, for example, carbon cells, you can take them through all these tests, the critical tests that the IAC calls for outdoor photo attacks. And so, it will still take some time to be fully reassured on whether the stability is what we really need for 25 years outdoor. But so far, what we can say is, the things that go wrong, it's not really the path guide. It's very often the contact materials or something else, corroding or water penetrating. So, these materials are much more rugged than we had believed, especially if you take the methylammonium out. So, my feeling is that we should be, we should keep going. It's a winning card. Good question, yes. So, regarding the very long stability that you observed, can you give us your expert opinion as to the key mechanisms that perovskites also must avoid to be stable for thousands and thousands of hours? Yeah, so, one issue was the contacts. I mentioned that. So, a lot of scientists use gold as a contact and the whole conductor, the classic one, is something we introduced 20 years ago. It's called the spiral ometard. It's a triarylamine. So, these materials are not stable enough. This whole conductor is just for research. It costs also too much, it costs $500 per gram. So, you need about one gram of this whole conductor material for one square meter. So, you're already dead with $500 per square meter. You will not be able to be competitive. But, it's a very good research material. You can buy it. It's working. It's not the final story. So, no materials are being used. I'm sure that we have ourselves something in print, which I unfortunately cannot disclose here. But, the whole conductors are really very promising, the inorganic whole conducts. I think we can be positive on this one. So, I would pay attention to those contact materials. I would take the metal ammonium cation out of the perovskite. It's a pain in the neck. At 85 degrees, it disappears. You make that iodide. So, that has to be taken out. But, if you take these other formulations, they will be very rugged. Unless you flood yourself with water. I mean, this is still, this ambient contact is an issue, but that's an issue for any photovoltaic. They're all laminated because you want to protect them. Question? Yeah. Can you talk a little bit about manufacturing processes? Are there particular kinds of things that are needed for the perovskites? The one I showed you, they're being screen printed. So, you can go to NTU in Singapore. They show you that, how they make those modules, and they get 11% on the module. And so, that's not the only technique. The slot die coating, certainly not spin coating. Okay, take the spin coat out. It's not going to work. That's why new procedures have been developed that avoid the spin coating. So, are they new really? Well, for the perovskite, yes. But I think these are existing then from the position techniques. I think there was a question back there. I just wanted to continue a little bit on your reliability question that you're answering. It's really certainly a very interesting topic. You know, among the three critical tests in damp heat, the temperature cycling and the light soak, what's actually more fundamental is light soak in many ways. If it's material starts to degrade, if you shine light on it, then it defeats the purpose of being a solar cell. What are the particular things in your utopian dream stack which would actually prevent it from, actually would enable it to be resistant to any issues with light? For example, if there's organics, there's certain kinds of organics which may deteriorate with light exposure. With the Dyson-style cells, we always had problems with the heat test. So, that is solved for the perovskite. It's a piece of cake, really, now to get a formulation that takes you over 90 days, 85 degrees. Just think about 85 degrees, 90 days. If you would go in a sauna and stay at 85 degrees for over 90 days, I wonder what is left. So, it's a harsh test, and it's even harder. The real test puts 85% humidity in addition. So, you need to absolutely be sure about your sealant. That's mainly a sealant issue, a lamination issue. So, that's a harsh test, but I think if you've got a good sealant, then the perovskite will take that. It's easy enough. As you have seen, we have done one experiment for 500 hours where we clouded it in addition with the full sun. And we had to self-work at maximum power. If there was only a 5% decline. Yeah, that was... The trouble with the test is now the referees ask me all the time to repeat the test. But it's not in the norm. It was beyond norm. So, anyway, it was a good experiment, and it shows that this can be very stable. It's not the whole answer, though. We have to be careful and do more testing. I would say it will probably take about five years to sort all of this out and have confidence between three and five. With a carbon technology, the stability is very good. So, that's why you already see modules being made. Well, I'm sure there are many more questions to ask. I know there are, but we probably need to cut the formal questioning off here. There may be opportunities to speak with Professor Gessel at the end of the presentation. Let's thank him once again for a very stimulating lecture.