 Thank you for attending this class. I want to try to keep the slides fairly short and have an opportunity for some discussion and some questions. So if you save your questions to the end, I think we should have plenty of time at that point to discuss things. I'll just start by saying that what I'm going to be presenting today is my own personal interest in hydrogen from renewable electricity, or at least my personal perspective on it. I wanted to share a few developments that I think are interesting from the point of view of science and technologies that can enable hydrogen production from renewable electricity in a way that could be economically viable and also have some important implications for the future decarbonized economy. So I'm affiliated with several different organizations at Stanford. I'm a professor in the Department of Material Science and Engineering. I'm also a professor of Photon Science. So I'm a faculty member at Slack as well as on campus. And I'm director of the Stanford Synchrotron Radiation Light Source, which is a major scientific user facility at Slack and at Stanford. And I also am a senior fellow of the Precourt Institute for Energy. So I've been involved with the Precourt Institute for many years, initially helping the Precourt Institute develop an industrial affiliates program a way to better interact or one way to better interact with industry in the green energy space. So just to say a little bit about my current job at Slack, the Stanford Synchrotron Radiation Light Source is, as I said, a major facility. We have about 140 staff. There are a large number of postdocs and students from Stanford who come and do research at SSRL, either as users. So they submit user proposals like people all around the world do to use our facility. Or they may become involved as collaborators with our staff and with some of the faculty at Slack. So as graduate students get involved in research and have advisors on campus, they often develop these kinds of collaborative relationships. So this is a facility where we have a particle accelerator. It accelerates electrons at close to the speed of light. And by jiggling those electrons in a strongly varying magnetic field, we can produce x-rays at 32 different experimental stations around this ring. And those x-rays vary in energy and intensity. And they're used in many different kinds of experiments involving spectroscopy, x-ray scattering, and diffraction, and imaging. An example of the kind of work that we do at SSRL that's related to energy, just a couple of examples shown in these schematics, are things like operando studies of battery materials. So within a battery, within the layered structure of the cells inside a battery, there are micro and nanoparticles. And as the battery is charged and discharged ion intercalation and deintercalation events occur in these particles. And it's possible to study these things using synchrotron radiation to obtain chemically sensitive images without having to destroy the battery. So with designing the cell appropriately, we can actually look at the process of ions going in and out of these materials, the expansion and contraction, and other mechanical effects that occur. Another application for synchrotron radiation is in catalysis. So being able to study how molecules interact with the surface of nanoparticle catalyst, how they bind to that surface by looking at the electronic signature of that binding. So looking through spectroscopy at the details of their electronic structure of the molecules, touchdown and bind to various sites on catalysts can give insights into how the catalysts actually work, what are the active sites, how can we synthesize better catalysts that are more active. So there's a learning circle that's depicted here going from synthesizing the catalysts, characterizing them, testing them on a beam line under realistic conditions, and then going back to theory to understand what does the data mean and how do we make use of it. So that gives you just a very brief flavor of what we do at SSRL. And I hope that many of you become involved as users or collaborators with us. So just as my outline here, I'll say a little bit about hydrogen as a fuel, and in particular, focus on the issue of energy density, which is one of the things that makes fuel so interesting and important technologically. Then I'll talk about a few different technologies that are in development or have been under investigation for some time and are being improved. Liquid organic hydrogen carriers, I think that's quite an exciting topic. And low temperature electrolysis with ion selective membrane electrode assemblies. These two technologies, as they're developed and made more efficient, have the potential to really push hydrogen synthesis in an economically viable direction. I'll focus on a specific interest of my group, which is coated catalysts for water oxidation. And that's part of what happens in this membrane electrode assembly electrolysis. And then combining efficient catalysts with high performance solar cells. So this is an approach to artificial photosynthesis, which could be interesting for distributed production of hydrogen off the grid. I also want to begin by acknowledging that there are many other PIs at Stanford and Slack who are involved in these kinds of topics. These are just some of the people that I know of who are working on hydrogen from renewable sources. Tom, Professor Thomas Haramio, he's also the director of the Suncat Center. Professor William Chu in my department in material science and engineering. Professor Friedrich Prinz, he's a professor in mechanical engineering with a joint appointment in MSc. Professor Matt Canan in chemistry, Hongjie Dai in chemistry. And then these researchers, Simon Baer, Demosthenes Socaris, both at SSRL, they're both experimentalists who use these techniques to study catalysis that I was just mentioning. And then Frank Abild-Petersen, he's a Slack theorist who's interested in catalysis and particularly related to green energy. So if we look at energy density and the reason why fuels become interesting or have an advantage, it really is compared to things like lithium ion batteries. The fact that we can store energy either with very high volumetric density, so this is a megajoules per liter on the vertical scale, or with high specific energy density, that is megajoules per kilogram. So this is something that might be more important, perhaps in a vehicle context. This is something that is also important in a vehicle context, but could be important in terms of other kinds of stationary storage. So we use diesel fuel, for example, in backup generation because we can store a tremendous amount of energy per unit volume. We don't have to have huge tanks of diesel in order to do that. But the ion batteries score pretty low on both of these energy density metrics. And that's primarily because you have all that other, in addition, the energy storage is occurring by shuttling these ions around, but you have all this other structure present, as I showed in that cross-section of the battery, to support that. So there's a lot of weight and there's a lot of volume that really is not effectively used in the process of storing energy. They have a lot of other benefits as well, obviously, and batteries are a huge and important area of research and development, but in this particular metric, they struggle a bit. So if we think about how these kinds of energy density metrics compare to renewable energy, if you look at the energy content of 40 liters of gasoline, so a small, by US standards, gas tank filled with gasoline, that has an energy content of about 1800 megajoules. And you need to run a 20% efficient solar panel, two meters by 10 meters in area. So this is a big panel for two weeks in order to generate that much energy. So from the perspective of doing things like backing up renewable energy that has a high temporal variability, where we're not producing it in constant quantities throughout the day, fuels are an interesting option. I also point out here that in the case of hydrogen, hydrogen is in some way the simplest fuel that we can imagine making in terms of the chemistry or electric chemistry that's involved. The specific energy density, of course, is very high because hydrogen is the first element in the periodic table. It's a small mass molecule and it's present in all of these phases as a gas at standard temperature and pressure or even as a liquid. With a relatively low density, so it has a high specific energy density. The volumetric energy density tends to be low in comparison to these other fuels, the hydrocarbon fuels. So there's a lot of interest in being able to come up with approaches to store hydrogen in a form where it has something like liquid hydrogen like volumetric energy density, not so low in volumetric energy density. And one approach for doing that is to use what's called a liquid organic hydrogen carrier. This is an image that I got off the website of a Japanese company called Chioda Corporation. They're a leader in this technology and their vision is to take different sources of energy, use them to perform electrolysis on water and create hydrogen, and then effectively hydrogenate a liquid organic molecule. That liquid organic host for the hydrogen then can be put into a tanker and can move across the ocean as it would if it were full of an organic fuel. In this case, toluene is the carrier that Chioda is looking at for the hydrogenation. And then it's possible to distribute that hydrogenated liquid organic carrier throughout the economy to dehydrogenate it wherever you want to. That might happen at a big plant to produce hydrogen in large scale for industrial processes. It may occur in a localized setting for hydrogen filling stations. In principle, it might be able to occur within a vehicle. You might be able to extract the hydrogen and then return the organic carrier. And the idea is then the toluene goes back to the electrolysis facility. In this case, it's a wind-powered facility that they're showing or solar. And you regenerate it and the process then continues. So it's a very interesting idea. It's not a new idea. The initial work on this was done in the mid-80s when toluene was studied for this. More recently, there's been interest in new generations of liquid organic hydrogen carriers. One of them is di-benzol toluene. This particular molecule is nice because it's possible to do the hydrogenation and dehydrogenation under more standard conditions. So not having to go to such high pressures for some of these processes. So we take hydrogen produced through electrolysis, add it to this molecule at not a crazy high pressure, do this over a catalyst. So a ruthenium catalyst on aluminum oxide support, for example. This is an exothermic reaction and we can get up to a 6.2 mass percent hydrogen which is a pretty significant hydrogen loading. It gets us about halfway to the volumetric energy density of liquid hydrogen. And then dehydrogenate this material and that can be done over a platinum catalyst with aluminum oxide and use the hydrogen that results from that in a fuel cell or a combustion chamber. So that's the general cycle of life for this kind of thing and it would be then reused as I said before. So I think that's a really exciting technology for a couple of reasons. The big issue with hydrogen storage is doing it in an energy efficient fashion, not having to waste a tremendous amount of energy in pressurizing the hydrogen so that it's sufficiently volumetrically dense in energy for many applications. And it's also using existing infrastructure because you can in principle use pipeline and storage infrastructure and shipping infrastructure to develop for hydrocarbons with this approach. It's a way of repurposing all of this very expensive stuff that has been developed for hydrocarbon fuels. The other technology that I wanna say a bit about is low temperature electrolysis for pressurized hydrogen synthesis. So as you saw in that previous slide, we do need to pressurize the hydrogen to some extent in order to do this reaction, 10 to 50 bar. And an exciting technology for this is low temperature electrolysis with a polymer electrolyte membrane electrolysis cell. That's what PEM stands for. This is a special kind of polymer that's selective for hydrogen transport. So we flow liquid water through one of the cells. There's a reaction that occurs on a catalyst laden coating on this surface of the membrane that involves oxidation of the water to make O2 and to make protons and electrons. The electrons, the protons are sent through this polymer electrolyte membrane fuel cell and then the electrons that are produced in this side of the half cell when pushed along by sufficient voltage will reduce those protons to make hydrogen. So this is occurring in contact with liquid water. On this side, we can actually have the hydrogen come out as a gas. So the hydrogen evolution reaction produces H2 in the gas phase and by providing sufficient over potential, so sufficiently high voltage here, we can produce hydrogen that's pressurized. And this is a really exciting innovation, the ability to use what's called the gas diffusion electrode kind of technology to make hydrogen vapor at pressure opens up a lot of possibilities for economic use of the hydrogen in the kind of scheme we just saw with the liquid organic hydrogen carriers. The most electrochemically challenging step here in the electrolysis is actually this water oxidation or oxygen evolution process. And that's because we need to transfer four electrons per oxygen molecule. So it's a multi-step electrochemical process and there are significant energy barriers to making it happen. And one of the real challenges here is finding a really good catalyst. The best and most efficient catalyst unfortunately is a very expensive material, iridium oxide. It's really the only efficient OER catalyst that is stable under acidic pHs. And if we're running this thing at high rates and making a lot of protons here, it's going to be relatively acidic conditions. So this is one of the challenges is how do we most efficiently make use of the iridium oxide and keep it from corroding away or losing it because it is so expensive. Just a little snapshot of one of these low temperature electrolysis cells. It's actually a stack of different materials. This membrane here is a complex material. The one that's shown here is actually a little bit different where we're allowing water flow and water generation out the other side of the membrane. But in any case, these are all, you know, highly engineered layers in this stack. And actually this whole technology really comes out of the PEM hydrogen fuel cell technology for vehicle fuel cells. So there's a lot of work that's been done that we can reapply to electrolysis. I'll say a little bit about this iridium oxide catalyst. One of the issues with the catalyst or any catalyst is deactivation. So can we maintain the high activity of our oxygen evolution catalyst over time? Will it become poisoned by impurities that are present in the feedstocks? And will it be mechanically robust and stable? Will the process of oxidizing water cause stresses that may cause these iridium oxide nanoparticles to delaminate from the membrane that they're attached to so one of the approaches for trying to stabilize and protect and prevent poisoning of catalysts is to use an overlay or some kind. And that overlay could be an active overlay where something about the combination of the overlay and the catalyst induces activity for the catalytic reaction on the surface of the overlay or it could be an inactive material but it allows the correct reactant species to come to this interface and producing the correct product and preventing other components that would tend to poison the surface of the catalyst from entering. So there's been a lot of interest in these overlays in the literature for various different kinds of catalyzed chemical and electrochemical reactions. One of the things that we've observed and also a group of Caltech observed right around the same time a couple of years ago is that if we coat iridium oxide with titanium dioxide which is a normally inactive material for water oxidation it actually enhances the catalytic reaction. So we can see that enhancement by measuring the current that's associated with the water oxidation. I said we need to transfer four electrons for oxygen molecules. So we can measure the rate of water oxidation by just measuring the current that passes. We can see the current is higher for a given potential when we have a TiO2 coated with a very thin layer on the order of less than a nanometer thickness of TiO2 compared to the bare iridium oxide. So this is a very interesting observation. It's not necessarily indicative of greater stability or protection from poisoning. It suggests that there's actually some kind of an enhancement in the overall catalysis as a result of putting this very thin and normally inactive material there. So this is not really understood at present and this is a topic of ongoing research in our lab. And so one of the ways that we can analyze this and study it is using this atomic layer deposition technique, which my group has developed as a way of coating materials with very thin coatings. So this is the kind of chemistry involved. It's really a chemical vapor deposition type reaction where precursor molecules come down and react on a substrate surface or on the surface of a growing material but the mechanism involves sort of a saturated chemisorption of these molecules and that means that we can make films and coatings that are conformal and pinhole free at very, very thin thicknesses on the order of a nanometer and it can be useful in coating complex surfaces because of this self-limiting kind of deposition mechanism including powders and nanoparticles. So it's an ideal way to take an inactive nanoparticle and coat it with something like iridium oxide or then coat that with TiO2 as we talked about in the previous slide. Our initial studies of the ALD TiO2 coatings on flat iridium oxide films that we've grown on flat substrate surfaces are interesting. We've studied this using a technique called X-ray photoelectron spectroscopy using angle resolved measurements. So the bottom line from this figure is that the intensity ratio, the intensities of the photoelectrons that we detect from the overlap, the TiO2 and from the underlying iridium oxide are such that as we change the angle at which we're measuring these photoelectrons as they come off the surface in this X-ray measurement, it appears that the TiO2 coating is actually a uniform thickness closed film so it has a fractional surface coverage of one meaning that it seems to coat the entire surface. So it doesn't appear to be a situation where there are pinholes and cracks and the water is able to go through those, react and the oxygen come back out. It seems like it may be transported through this TiO2 layer itself that's involved but we're currently studying that. This is an example of how we can use X-ray tools to say something quite interesting and detailed about mechanisms. So the ALD process is something that we've investigated for various different applications. One of them is taking materials like TiO2 that are normally inactive for water oxidation and using them to protect underlying reactive substrates like silicon that are good at absorbing light and generating photovoltage that we can use for electrochemistry and then we can coat those with an iridium catalyst also by ALD if we wish as I showed previously. And this just gives sort of a schematic of what the mechanism of the TiO2 deposition is. Because we can make these films very thin and pinhole free, they can be effective in protecting silicon without absorbing a lot of light or being a significant electrical resistance between the catalyst where we have to extract those electrons in order to oxidize water and the silicon itself. So this kind of structure then can be used to protect a normally reactive silicon photovoltaic material from the electrochemical environment involved in water oxidation. So these are experiments under very extreme pH conditions. So this is one molar sodium hydroxide. So very basic electrolyte solution, one molar sulfuric acid. And when we have the TiO2 films or coatings present, we can continually split water on the silicon that's so coated, but if the coating is not present, the current that we measure for water oxidation decays within a few minutes. And that's because we effectively start oxidizing the silicon and building up a very thick SiO2 layer that shuts down the flow of current and shuts down the electrolysis. These chrono-amperometry measurements show similar kind of situation if we perform them in either sulfuric acid or strong base. We can protect and continue water oxidation for many hours on silicon that would normally fail in just a few minutes. So the last thing I'll say about something about then is this idea of taking these highly active catalysts and combining them with a photovoltaic device to make a sort of a light driven artificial photosynthesis cell that's capable of generating hydrogen from water in remote locations off the grid. And in doing this, we need to combine the current voltage characteristics of the solar cell with what's called the load curve of the electrocatalyst. So this is the amount of current we can push through during electrolysis as a function of the applied potential. And ideally we want this load curve to be very, very steep and to cross the IV curve of the solar cell at some high current density where we can effectively make a lot of oxygen and hydrogen per unit time so we can have a high production rate. The solar to hydrogen efficiency can be calculated in terms of this short circuit photo current here and the voltage required, the minimum voltage required to split water which is 1.23 volts and the incident power of the sunlight. And the initial Department of Energy target for this efficiency is about 15%. That's what the initial desired efficiency is for an economically viable sort of pathway to a technology. What we've been able to do is combine silicon high efficiency cells. These are heteronjunction cells called HIT cells with a built in electrochemical flow cell underneath one of the three series connected silicon HIT cells and obtain at zero external bias. So just driven by the sunlight, a production rate over 10 milliamps per centimeter squared of current. And that corresponds to a solar to hydrogen efficiency of about 13%. So very close to that 15% goal for the Department of Energy has stated. And these cells can survive for actually for days. This shows 20 hours of stability but we've been able to test them up past 100 hours of continuous operation by using ALD to protect the backside of the silicon solar cell where all this catalyst is and the electrochemistry is taking place. We're able to get very good stability. This is just an indication of the kind of stability results that we've seen. And the fact that we can, in this case, this was, these were up to 24 hours but we've, as I said, gone up to over 100 hours now. 120 hours of stable water splitting. So five days, which is about as long as we would do this kind of measurement in my lab. We only have one simulated solar light source to use for these kinds of experiments. So we can only keep it tied up for so long. Another interesting thing about this, if we use the ALD-TIO2 to protect the silicon photovoltaic, we can actually operate over a wide range of pH. So we can operate with these relatively acidic or basic solutions which have very low resistance for operation. But we can also operate in a buffer solution at pH seven. And this is important if you want to actually develop a technology. Most people don't want to have a solar cell with a water solution next to it that's concentrated acid or concentrated base. They don't want to have to mess with these relatively dangerous chemistries. A pH seven buffer solution is much more benign. So it will work across a wide range of pH if you use the ALD protection scheme and the right catalyst for this kind of structure. So that's really all I had prepared in terms of slides. I could speak to them in more detail, but I thought it would probably be a good idea to just open things up for questions now. And we can sort of talk about this in a more, in a less formal way. So I see it utter, yeah, you can unmute yourself. Yeah. Hi, Professor, you mentioned that you could combust hydrogen like using the gas infrastructure first, the hydrocarbons. What are the benefits of using the combustion process to get your energy back or the fuel cell process? Yeah, that's a really good question. Generally speaking, fuel cells can have higher efficiency than internal combustion engines. They have some downsides though. So if we use, if we take that PEM kind of structure that I showed before for electrolysis and instead of using it to split water, we use it to react hydrogen with oxygen to make water to sort of the reverse process. One of the issues with that is you have, you do have to use quite expensive catalysts in that process. And that's something that you need to do within every vehicle. So every vehicle has to be fitted out with a fuel cell that has these relatively expensive catalysts. And there's some concern about sort of long-term reliability of fuel cells in comparison to IC engines. IC engines have been around forever. People have a century, more than a century of experience with how they can fail and how you can reduce the likelihood of that. They've just gotten much, much better over time. Personalized fuel cells are much newer and they're still sort of on the learning curve for many of these things. So that's probably the major, I would say the major distinction. In principle, the fuel cell could be significantly more efficient, but there's sort of the balance of systems costs and then the reliability issues that I think still need to be worked out. Most likely these kinds of fuel cells are gonna be implemented and will be seen in fleet vehicles, long haul trucking, maybe buses, sort of situations where you can imagine having to check in as a bus goes around its route, having to check in to a depot. And if you need to swap the membrane out, there's somebody there who can just do it and then put you right back on the road. It's not so easy to do that for in a personal transportation vehicle. Hi, Anga here from the GSB, a chemical engineer. I wanted to ask you about, there seems to be a trend going against hydrogen and more pro-green ammonia as a fuel carrier. I was curious if you can talk a little bit about that, if that's also a research topic. And I see some parallels with the catalyst and the potential for photo-criticals. Can you maybe compare and contrast? I'm curious to see what the trends are going on. Yeah, I think that that's a really exciting area. And I don't want to minimize the importance of other potential energy carriers and certainly green ammonia is one of them. It's been an area where the catalyst development is probably some years behind that for hydrogen. And so that's where real breakthroughs could come. And synthesis of ammonia, electrochemical synthesis of ammonia is intrinsically interesting from the perspective of agriculture as well. Ammonia fixation is one of the critical things that allows our contemporary society to exist. And so I think anything that people can do to understand that better and how to do that with renewable energy sources is important. In addition to the potential fuel possibilities. So I think that's a really exciting area. I don't work on it myself, but some of the people on that slide that I showed early on do. So I think in particular, Suncat, which is a center here at Stanford for catalysis research, has several PIs who are interested in that. Hi, thank you for your talk. I was curious. So when this technology gets to a place where it can start being used more just for fuel for something like long-haul trucking, what do you see as kind of the other areas of research that need to be done or things that need to be figured out to make that really viable in our world? Yeah, it's an excellent question. Well, I think that the whole production part of it, so being able to produce it at sufficiently low cost per liter that people are going to be willing to adopt it compared to other approaches. There's the storage and transportation part. And if the liquid organic hydrogen carriers can be really used at scale with existing petroleum infrastructure, that would be huge. And that's an active field of research. There's a lot of interest in that right now. I think that in terms of the economic consequences, where we're likely to see this have the biggest impact is in things like what you mentioned, long-haul trucking, heavy transportation, not probably personal vehicle transportation because there seems to be a pretty good path for batteries to work as the range increases. But I see for heavy transportation, trucking, shipping, trains, there's a lot of that, which will be hard to run on batteries just because of the necessary specific or volumetric energy density constraints. And perhaps things like backing up the grid. Although what Tom Harmeal would say, Professor Harmeal in chemical engineering is really we need to think about this sort of electrochemical production of protons and electrons, which we do when we make hydrogen as an enabler for all kinds of chemistry that we currently do with hydrocarbons. So currently we use hydrocarbons in so many different contexts. Steel making uses a lot of hydrocarbons. There's so many things around us that we depend on that are gonna be hard to decarbonize. But if you have a source of electrons and protons at the right potential, you can do a lot. So I think that that's another thing that will naturally come from this research and development is a path to replacing a lot of the chemical synthesis we do today with electrochemical synthesis driven by renewable sources. That's the hope anyway. Yeah, hello Professor. So, yeah, catalysis is essentially a surface phenomena. Like if you are using active overlay coatings, like how is that not going to hinder the actual potential of that catalysis itself? Now then we can already shift all together to using that coating only all by itself. Yeah, this is an excellent question. So it turns out that there are, the example I showed, as I said, we don't really understand it yet. Why it should boost the performance of iridium oxide to put this normally inactive material there? And it could be that the nature of the chemical bonding between the TiO2 and the iridium oxide is such that effectively the surface of the iridium oxide, the surface bonds are distorted or there are structural changes to the surface of the iridium oxide induced by the presence of the TiO2 that make the iridium oxide a better catalyst. But there are other systems in which this kind of sort of mutually cooperative effect occurs and they're often driven by strain. So if you take one metal nanoparticle and you coat it with another metal that is structurally similar but has a different atomic spacing, the coating induces strain on the inner core particle and the inner core particle induces strain on the outer coating. And that can sometimes lead to very high activity because the strained bonds that are present on the surface actually allow the reactant to stick better and to more easily reactant and for the product to come off. So catalysis depends on having something that's just sticky enough to keep the reactant molecule in place while the catalyzed reaction takes place and then releasing the product very easily. And so bond density, the chemical identity of replets right at the surface and strain actually all play a role in that. And that's how these coatings may be in some cases a real game changer but it's a very ripe area for research right now. And it's hard because you're talking about really nanoparticles. So measuring these things and understanding where every atom is on the surface of a nanoparticle and how that's happening as chemistry's going on, that's really, really hard to do. And that's one of the exciting things about working at a place like SSRL is we have some of the tools to do that. But there's also a lot of room for theory that's necessary to understand how these things work. But strain in the coatings will not result in mechanical damage? So it depends on how much strain you're talking about and with nanoparticles they tend to store strain elastically whereas as you know metals that are under strain and are bulk will form defects called dislocations that relieve the strain. Dislocation formation in nano crystals is much more difficult. So you can achieve states of strain that are really impossible in bulk materials. This is one of the reasons why science and nanotechnology got people excited 20 years ago is that you could do things like that. It's just one example of what these nanostructures allow you to do.