George Crabtree is a Senior Scientist at Argonne National Laboratory and Director of its Materials Science Division. He has won several research awards, including the Kamerlingh Onnes Prize for his work on high-temperature superconductivity. He has served as Chairman of the Division of Condensed Matter Physics of the American Physical Society, as a Founding Editor of the scientific journal Physica C, and as a Divisional Associate Editor of Physical Review Letters. He has published more than 400 papers in scientific journals and given approximately 100 invited talks at national and international scientific conferences. His research interests include materials science, superconductivity and magnetism, and energy science. Recently he served as an organiser and spokesperson for the Department of Energy’s Workshop on Basic Research Needs for the Hydrogen Economy.

The two hydrogen economies
by Dr George Crabtree

Thanks, Michael. What a pleasure it is for me to have received the invitation to come to talk to you and spend some time in Australia. I would like to thank Jim Peacock and Michael for this opportunity. It is a real pleasure to see science treated so well and respected so highly at this event. Especially, when it comes to hydrogen and the way in which the scientific aspects are being stressed. That’s really a wonderful way of thinking.

I would like to tell you a little bit about what I have called the ‘Two Hydrogen Economies’, and on my first slide I would like to acknowledge, as Michael indicated, Millie Dresselhaus and Michelle Buchanan. Millie was the Chair of the workshop on hydrogen (which the Department of Energy held a couple of years ago), and Michelle was an Associate Chair, working along-side me.


(Click on image for a larger version)

Here is a preview of what I would like to present. Firstly there is a little bit about hydrogen as a solution to world energy challenges. There are really four challenges – supply, security, pollution and climate change – and I will spend a little bit of time on each one of those. I will then come to what I have called the two Hydrogen Economies, which are very simple to understand. I have called them the incremental one and the mature one. The incremental one is what we could institute today with the technology we have, and we have heard a little bit about that already. The mature one is where we really need to be able to make an impact on the energy issues that we are all facing. Thus, as we will see, those two things are really quite different.

The last subject is a little bit about the outlook for the mature Hydrogen Economy: what do we need to get there? The thesis I will put forward is that there are a lot of basic research challenges. You certainly can’t do it with today’s technology. As can be seen in the slide, I have listed a few of these challenges, and some of these, I will talk about during the next few minutes. Finally, I will address the topic of compelling market incentives. Suppose you could do it, would anyone do it? Would it be implemented? What does it take to get it implemented?

So those are some of the things I would like to tell you about.


(Click on image for a larger version)

Let’s start by just looking at the big picture. What is the world energy use? You can see in the upper left of the slide, a graph of world energy demand since about 1970 that is projected up to about 2025. The unit is terawatts. What does that mean? That means that if you take all the energy that the world uses in one year, expressed in joules, and divide that by the number of seconds in one year, you will get watts. It turns out that the amount has a prefix of a large factor of 10 (e.g. 1012 watt) and this is terawatts. This unit is therefore used to talk about energy at large scales, when we’re not talking about oil in billions of barrels or talking about electricity in kilowatt hours. So then we refer to terawatts.

The graph shows where we have been. You can see that right now we are sitting at about 12 or 13 terawatts average power use. The projection is to double by 2050 and triple by 2100, and remarkably those projections are pretty much agreed on by many in the community.

You can also see what the ‘use’ figures are: the industrial countries use about half of the total energy; while the US uses about half of the industrial, so that’s 25 per cent of the energy. I might note that the US has 5 per cent of the population, so there is an imbalance there that is rather obvious. It may also be seen that the developing countries have the largest slope on the graph. In fact, their slope is pretty much the slope of the total. This trend is what is driving the total energy use up so high. Then the bottom line represents eastern Europe and the former Soviet Union, which actually use less energy since they faced the economic problems of the 1990s.

So we need to double our energy output by 2050, and triple it by 2100. The truth is that no-one really knows where all that extra energy will come from. You might say, ‘Well, let’s just do what we’re doing now and double it or triple it.’ What would that mean?

Here in the lower right of the slide is the world fuel mix in 2001. It doesn’t change a lot. Oil provides almost 40 per cent of our fuel; gas, more than 20 per cent; coal, a little bit more than that; nuclear and renewable are at levels of about 8 per cent each. So if you look at these figues you see that about 85 per cent is fossil energy. Thus if you just double everything – or triple it – in the coming years, you reach a few challenges.


(Click on image for a larger version)

So supply and access are the first two challenges that I will mention. Now let’s take a look at the world oil supply and demand curve from 1900 just up to about 2000, as the important focus (I note that there is a projection on the right, but let me cover that in a minute). You can see that in 1900, oil registered as virtually nothing. We didn’t use any oil compared to what we use today; coal was ‘king’. Then in the mid-’50s people around the world started to drive, and so oil usage went up. That’s what drove the big increase. Round the ’70s and about 1980 there was the first oil crisis; and you can see a peak there. Remarkably, the world used less oil in response to this situation, and this pattern gives us an idea of what may happen in future years, if indeed oil becomes very expensive. At that moment in time, oil costs (in today’s US dollars) were slightly above $85. Now, we’re at $75, so we haven’t quite hit that peak, remarkably.

Then after the first crisis, oil became cheap and use went up again. So you can see where we are today.

You hear a lot about the possibility that the world will run out of oil. Actually, probably the world will never ‘run out’ of oil, because there will always be some left. The real question is: will the world be able to supply the demand for oil that is there? You see in the future there a couple of very simple projections – you should not take them too seriously, but it does indicate what might happen in the future. As we know, the demand curve goes up at a rate of about 2 per cent a year; this is demand growth. There are two scenarios here.

First, suppose that after oil peaks (i.e. we have produced the maximum we possibly can for a given time) we project how the production curve will decline. If you decline symmetrically, that is at 2 per cent a year, you get the red curve. However, if you say, ‘Well, instead we’ll only pump 10 per cent of what is left after the peak,’ then you get the blue curve. Nevertheless, in either case you see that (at least in this projection), it won’t be too long – indeed some time around this mid-century – when oil production will peak and we will have to look for other sources of energy.

In either one of these projections you can see that if you take today’s use of oil and project beyond this, then by 2050 or thereabouts we will be producing less oil than we do now. So what does this mean?

Well, after the peak, the world becomes a different place. This may represent one of the tipping-points for energy changes. For example, we will have to find other sources of oil – and there are other sources around. Oil shale and oil sands, for example, could make up the difference. But the pattern of our use has to change.

So this is the first challenge. The second challenge is access. Suppose you have enough oil. Can you really get to it?

The graph at the lower right of this slide illustrates three bars which correspond to each of three areas of the world – the US, the OPEC countries, and the rest of the world. The blue bar represents the reserve, the cross-hatched bar represents the production, and the red bar is consumption. It can be seen right away that consumption and reserves are inversely related. In other words, the places in the world that have lots of oil - don’t use it, and the places that use lots of oil - don’t have it. That is inherently unstable, and you can imagine that, especially after the peak – and, as we are seeing now, even before the peak – it is not clear that oil will flow from the producers to the consumers.


(Click on image for a larger version)

Here is another issue. A lot of the local pollution that we find in the world comes from impurities in fossil fuels. I don’t want to spend much time on this, but I have here a chart of the six major pollutants that are monitored (at least in the US). The map of the eastern US, in the upper right of the slide, makes the point that pollution is not a global problem, it’s a regional problem or a local problem, and that it comes really from two sources: automobiles and power plants. This slide shows where acid rain was predominant in the late ’90s, and if you look at where coal-fired power plants are located, you see that they correlate very well with the plume of the power plant in terms of producing the acid rain in the area.


(Click on image for a larger version)

Here’s the final challenge, climate change. Many people say that this is the most important one.

You see here, in the upper left of the slide, an interesting curve of time that ranges from thousands of years before the present. Three things are being plotted. One is the CO₂ level in the atmosphere, another is the methane level, CH₄, and the other is the relative temperature of the earth. You see that, remarkably, these three things correlate very, very well. There can be no doubt about this data, which comes from Antarctic ice cores.

You see a couple of other things too. The pattern is very asymmetric: the world warms up a lot faster than it cools off. It can also be seen that there are two limits here - an upper and lower limit. The world’s temperature and greenhouse gases seem to shuttle between these two limits, with a period that can be discerned, and we don’t really know why this is so. However there is no doubt that now (here the graph shows us coming up to the present) we are in a period of warming, as shown by all three of these curves. Thus you might think, ‘Well, we’re approaching the limit. Now it’s time to turn around and go back down.’ But indeed that is not what is happening.

So in the lower right of the slide we have a graph of CO₂ and temperature for the last 1,000 years, and you see that around 1750 or so, when the Industrial Revolution started, both of these two things tracked up. Then if you look at the CO₂ level in the atmosphere now – this is 2000 but in 2004 – you can see that it is 380 ppm. That’s way outside the historical trend. So we don’t really know what is going to happen now. In a sense, we are doing an experiment with the Earth, I guess rather inadvertently. However this could have rather serious consequences and we all know what people are predicting in relation to these outcomes.

It will be a serious problem for a long time. In the upper right of the slide you see the relaxation times: transport of CO₂ or heat to the deep ocean, 400 to 1,000 years. So whatever we put out there now, will have to play out its course. We really can’t control that on a time scale that is shorter than that.


(Click on image for a larger version)

We all know that there are good reasons why the hydrogen economy can address the four issues we have just discussed. In the upper half of this slide is a bubble diagram of what the hydrogen economy is, showing production, storage and use. Now we all know that hydrogen is an energy carrier, and not an energy source - so we have to first make the hydrogen somehow. Then we have to store it, and then we can use it.

There are lots of ways to make hydrogen. Most hydrogen is currently made by fossil fuel reforming, and we should put carbon capture together with this process, although we don’t really do this at present. Then there are other ways to make hydrogen: the best way is to split water, as we heard earlier, and this diagram shows some sources of energy to split water.

Once you make hydrogen, you then need to store it before you can use it. This is a crucially important step, because it separates the production from the use. So, no matter how you make it, no matter how you use it, you need to pass through this storage step. (That is an important flexibility producer.) So in terms of storage, this can now be achieved as a gas, or hydride; though we could store it as liquid. However, I will say in a minute why we don’t really want to store hydrogen as a liquid.

The main methods of using hydrogen typically involve fuel cells: for automobiles or for consumer electronics as well as for stationary electricity or heat generation. The reason fuel cells are so attractive is that they are, in fact, very efficient.

However there are some problems that are illustrated in the lower half of this slide. We now make, in the world, about 9 million tons of hydrogen a year. However, you actually need about 150 million tons a year for light cars and trucks in 2040 just to supply the US, and this represents a 10- or 15-fold increase.

Certainly, we need to be able to store such energy. Currently, you can store it at about 4.4 megajoules per litre as very high pressure gas – 10,000 psi; or alternatively, as 8.4 megajoules in the form of liquid hydrogen. However, this latter method is not really a viable way to store hydrogen, because you lose about 30 or 40 per cent of the energy content just by liquefying and then regasifying the hydrogen. To expand on this point, you need about 9.72 megajoules per litre to power car travel, so let’s say, 500 kilometres on a tank is attainable. If this is so, then there is quite a gap here, of a factor of 2 or so.

So then, how about the use of hydrogen in fuel cells? Well, it now costs about $3,000 per peak kilowatt of output to make fuel cells. That might become reduced down to $300, representing a factor of 10, with mass production. However a gasoline engine costs $30. So again, there is quite a gap here - before fuel cells become competitive.


(Click on image for a larger version)

This slide shows the topic of a DOE workshop that was held a couple of years ago. You see here Millie Dresselhaus, Michelle Buchanan and myself, and some of the details. It was quite a big workshop that brought all the experts together.


(Click on image for a larger version)

We then produced this report, which you can find on the web, about the prospects for the hydrogen economy. This came out in July 2003, and about six months later, a National Research Council report from the US reported, remarkably, almost the same conclusions. This was even though it was written completely independently of the workshop report. In the right-hand panel of this slide you can see a summary that was reported in Physics Today about a year and a half ago (which Millie, Michelle and I wrote). This contains, in a few pages, most of the spirit and messages that both reports outline and it is pretty easy to read. You will find that available online too.


(Click on image for a larger version)

I want to spend the rest of the talk thinking about the Hydrogen Economy and assessing state-of-the-art facets of its development.

We will look at questions like; what is the impact on energy challenges? What are the research opportunities? What is the outlook for development? Then, because it is so conveniently divided into production, storage and use, we will just take those compartments as our structure.


(Click on image for a larger version)

We will start with use, because that is the component most familiar to our everyday lives. I am going to take, as my prototype fossil fuel, methane (CH₄) - a natural gas, because it has a definite chemical formula and you can write the evolution of extracting energy from this source in a very nice, simple thermodynamic way. However everything I say about methane can be extended to coal or oil. Oil is said to be, on average, CH₂. Coal is said to be CH_0.8. So you can just scale these numbers.

The reason that fossil fuels are such good energy sources is they contain carbon and hydrogen; and these combine with oxygen to make CO₂ and water. These are two very stable compounds, and the energy you obtain in the process is 817 kJ per mol. (The minus sign you see in the slide means you get the energy out, rather than putting energy back into the process.) So, it’s a great energy source and that’s why we use it so much.

What about hydrogen? Well, you don’t have any carbon, so there is only one stable compound that comes out and you get considerably less energy from the combustion of hydrogen. Indeed it is a little more than a factor of 2; although with hydrogen, there are no primary pollutants - you can reduce the nitrogen oxide if you carry the process out at lower temperatures; and there are no greenhouse gases. So there are ways that we can use hydrogen right now. These processes are all within technological reach.

Almost every heat engine that burns fossil fuel can burn hydrogen instead. This includes gas turbines for electricity generation and even jet engines. You can put in a mixture, any mixture you want at levels of up to 100 per cent hydrogen; and only need to make a few changes. You can also burn hydrogen directly – of course; and you can burn it in combustion, but a better way to burn it for many applications is over a catalyst such as platinum. This is because by controlling the rate at which hydrogen is fed into the system, the temperature of the combustion and also the heat output can be controlled. So here is a picture of hydrogen being burned over platinum at 500°C; that’s great for space heat or stoves.

It is also possible to burn hydrogen in a car. The present internal combustion engines run pretty well on hydrogen. Of course, these have to be modified slightly to take hydrogen instead of gasoline – and lots of car companies are developing such cars. Here is one BMW with a 4.4 litre V8 engine, 184 hp that can be driven at 133 mph in a 190 mile range. The interesting aspect of this car is that it can burn either gasoline or hydrogen – the driver flips a switch, so if you run out of one you switch over to the other fuel. This model is being test driven today but is not really for sale. Ford and Mazda are also looking at it, and in fact if you look on the web, you will see that the Hydrogen Car Co. will take your present car in for a small fee and convert it to hydrogen. So we could all be burning hydrogen for transportation right now, if we wished to do so.


(Click on image for a larger version)

What about the uses tomorrow? As I put here, in fuel cells instead of for internal combustion engines and other sources? Here is something which was referred to a little bit earlier. In transportation, there is a program in Europe – 30 buses, 10 cities – to put hydrogen powered buses in regular service.

I was in London a couple of years ago at a hydrogen conference and had the opportunity to ride the bus. It was very quiet, although you wouldn’t know that in London, and people seemed to like it. The stories I was hearing were that people would wait for the bus and see a bus coming and say, “No, I don’t want to take that bus because the next one is the hydrogen bus. I’d rather ride that.” So there does seem to be a degree of public acceptance. This covers one very big area, that of transportation.

Now another area is stationary power. Shown here is the 250 kilowatt solid oxide fuel cell that can produce local power. A very interesting point here is battery replacement. For your laptop, you could get twice the amount of running time using a power source of the same size and weight that runs on a fuel cell, and not a battery. In fact,this may indeed be one of the first market penetration opportunities, because although the price per watt is rather high, the cost is low and lots of people, including me, would love to be able to run their computers longer.


(Click on image for a larger version)

Let’s now take a look at what fuel cells are. This slide shows one of the most common types, although it is by far not the only one, and it is known as a proton exchange membrane fuel cell. It is a way, essentially, of stripping off electrons, taking them outside the power source to do the energetic work, and then return them back into the system by chemical means.

Hydrogen enters on one side of the fuel cell, it then contacts an electrode where it is dissociated and the electrons become stripped off, so what is left behind is essentially a proton. For this to occur, a platinum catalyst is required. The proton then travels through a solid membrane, which is the electrolyte. The electron, on the other hand, goes through the external circuit where it contributes to energetic work, and then it comes back to the cathode, where the proton and the electron are re-united, along with oxygen, which has to be split. Thus the chemical reaction that takes place produces water which is the only output.

So there are two half-cells. At the left is the reaction at the anode and at the right is the reaction at the cathode. Both need platinum or some other catalyst. Thus in this way, chemical energy is directly converted into electrical energy. So it is interesting to note that this is not a heat engine but a chemical engine. Therefore, it has in principle the potential to be twice as efficient as a heat engine.


(Click on image for a larger version)

There are however, some problems. The main problem is shown right here, the so-called ‘overpotential’. This graph plots the output of a fuel cell in volts versus the power that is run from it. The setting should be located at the thermodynamic value of 1.23 volts, however as soon as you draw anything at all out of it, it immediately drops to a much lower value, and as you draw more and more current out, the voltage drops even further so that it is common to work at around 0.7 or 0.8 volts instead of 1.2 volts.

Some of this reduction in voltage is due to internal resistance, as might occur in any battery - however most of it is just because of sluggish reaction kinetics at the cathode. Specifically, the oxygen reduction reaction is a complicated one, and doesn’t proceed very fast. So this is a problem – especially if you have a hydrogen car and you put your foot on the hydrogen because you want to be able to accelerate up the hill or to pass that big truck, then this is one of the issues that will prevent you from doing so.

So why is that so hard? Well, if you think about the chemical reaction taking place at the cathode, you realise that if you take four electrons, and you take four protons from different sources, and then put them in contact with a catalyst which also has an oxygen molecule on the surface, then you have to transfer all the electrons and break the oxygen-oxygen bond to make water. At present, it is not even clear by what route this actually proceeds and there are many discussions about the intermediate steps that may occur. Thus it is a difficult reaction because it has so many atoms and electrons to coordinate. Therefore catalysts are a problem.


(Click on image for a larger version)

A second big problem involves the membranes. Shown here on the left is the membrane that is used in most proton exchange membrane fuel cells. It is called Nafion®, made by DuPont, and this product has been the gold standard for 20 years. It has a complicated polymer structure. The basic idea is that using a polymer backbone, it is possible to have SO₃ groups which weakly bind hydrogen as ions, and then various vacancies can be placed inside this structure so that the hydrogen can jump from one side chain to the next and make its way through the material.

That is the standard. As it turns out, for this material there is the requirement for the proton to be surrounded by a cloud of water, so this means having to hydrate the proton for it to make its way through the structure. This means that it is not possible to operate the membrane at temperatures higher than the boiling point of water, typically at about 80°C. So at this temperature, reactions are rather slow and there is a need for powerful catalysts to make this work well enough to supply power. Ideally then, what would be better is to have a higher-temperature membrane than the ones presently available.

On the right of the diagram is another kind of fuel cell, a solid oxide fuel cell where oxygen is the mobile species, (I won’t spend much time on this). Oxygen moves through vacancies in this type of framework, typically a perovskite structure that requires high temperatures, typically about 800°C to provide such vacancies. This is too high to be compatible with other materials and of course when you warm temperatures up to 800°C and cool down the system down then there is a lot of thermal stress on the system. Thus, people would like to have, for solid oxide fuel cells, a lower operating temperature for the membranes.


(Click on image for a larger version)

So what about storage? All scientists know the famous bottles in the left-hand panel of this slide – this is how researchers store gas in the laboratory. They are way too heavy for transportation. Instead, lightweight fibre-reinforced gas tanks are required like the one shown in the centre panel here, which is so light that it can be lifted over the head of the gentleman pictured. These are needed to store gas at high pressure, so 5,000 psi or 350 bar is the standard pressure. It is possible to attain 10,000 psi, or 700 bar, and more and more it is possible to indeed see hydrogen being stored at these pressures which are at about the limit of what is achievable. So this can be done today and there is no need for a lot of science to do this.

The other way to store hydrogen that people can use today involves liquid storage as shown in the right-hand panel. Here is the BMW that I showed earlier, and in the trunk of that car is a DOER of hydrogen. It is great for stationary applications – though this approach is a bit tough in practical terms because it takes up so much space in a car. Further, as I mentioned earlier, about 30 to 40 per cent of the energy is lost in liquification. So, although people can use this approach, it is really not the right way to do it.


(Click on image for a larger version)

This graph shows why it is such a challenge to store hydrogen. So why is this so and what does this figure represent? The figure shows the volumetric energy density of various fuels, the energy per unit volume, the gravimetric energy density, and the energy per unit mass. The basic message is as follows. At the far left are batteries, including the best known today which are lithium ion batteries. This kind of battery has an energy density, both volumetrically and gravimetrically, which is pretty low. Located to the right of these are a couple of dots that indicate some of the hydride materials that can be made.

Next on the diagram is the energy density of compressed gas hydrogen, and then liquid hydrogen, which looks pretty good but isn’t really practical because of the 30 per cent loss. Further to the right is the proposed DOE goal for transportation, and out at the far right is gasoline. So it can be seen that it is really tough to beat gasoline as an energy storage medium. Certainly this is one of the reasons why people continue to use this fuel in cars. You might even be quite shocked to calculate, when you fill the tank of your car, just how much energy per unit time that is transferred. It is really a big number; and yet it is done so casually.

Anyway, this is why it is such a challenge to find alternative fuels that come anywhere near gasoline. Thus it might be asked why the DOE goal is so small compared to gasoline. The reason is that fuel cells can be twice or, very optimistically, two and a half times as efficient as a gasoline engine and so there is no need for quite the same energy density to go the same distance. So you can get away with a lower density, although even this lower density is pretty hard to achieve. In practice we’re not anywhere near where we really need to be.


(Click on image for a larger version)

In order to store hydrogen effectively for transportation, you have to store it at densities higher than liquid hydrogen. Thus you might think, ‘Well, that’s pretty tough to do,’ but actually it isn’t – liquid hydrogen isn’t very dense – and every material, every hydrogen compound, listed here on this slide, stores hydrogen at densities higher than the liquid.

So what does this slide represent? This is the volumetric density, the same as the last graph, but now the hydrogen mass density is represents the percentage of the mass that is hydrogen. You can see here 100 per cent hydrogen, so that is all hydrogen. Here is the volumetric density at 350 bar gas pressure, at 700 bar and in liquid hydrogen, and as you see, everything on this chart has densities of hydrogen higher than the liquid.

This shows some of the materials that people are looking at as potential materials for storing hydrogen. There are some very traditional ones, for example, lanthanum nickel 5 is here and this is a material that dates back to the 1970s. Also, there are some very new materials. I should point out here some of the hydrocarbons – alcohol and various things; and here is liquid natural gas. You can see that they are great ways to store hydrogen, and in fact, one of the proposals is to simply take a hydrocarbon that has a lot of single carbon bonds in it, and take out some of the hydrogen. Then you would have to make double and triple carbon bonds in order to release the hydrogen, but you can do that reversibly. In this way, it would be possible to take out hydrogen in a reasonable mass fraction. The trouble is that you can’t do that on a car very easily; it requires lots of chemistry. Although by comparison this type of reaction system can be done feasibly for stationary applications.

Here are some other alternatives. One possibility that people have looked at recently involves the so-called alanates. For example, here is lithium aluminium hydrogen 4 (LiAlH₄), which has an anion AlH₄ that looks very much like methane, except in place of carbon there is an aluminium cation that has a net negative charge. Thus there is a need for a positive ion, lithium, to make a storage compound. There are also other useful compounds such as lithium boron H₄ (LiBH₄) and sodium and so on. Essentially, the idea is to use very light elements.

Something that just came up about a year or maybe two years ago involves the use of ammonia (NH₃), in this case ammonia borane (NH₃BH₃). A reaction is used to produce NHBH by removing two hydrogens. This reaction turns out to be pretty effective. Its position on this chart is indicated by the lower arrow.

However, there are still some outstanding difficulties with hydride storage. On the one hand there is a need for a lot of hydrogen so that people can drive long distances; and this means needing to store it densely with strong bonds. Yet, on the other hand, there is also the need to be able to accelerate fast; that means ensuring weak bonds so that it is possible to get the hydrogen off quickly. This process usually needs a loose path for the hydrogen to escape from the centre. Ultimately, these two things are difficult to satisfy simultaneously.


(Click on image for a larger version)

So far we have looked at use and at storage. Now let’s take a look at production. Again we start with what we do today. Almost all the hydrogen we make today is made by reforming natural gas and this is shown on this slide as CH4. The process involves steam reforming, mixing it with hydrogen, and then there is a need to put in quite a bit of energy, represented by a plus sign: +150 kJ per mol. The result of this process include the two products:CO₂ and H₂, and half this hydrogen actually comes from the water. So in addition to liberating hydrogen from methane the process involves splitting water and obtaining hydrogen in a source in this manner. This is a pretty good deal. Though, the byproduct of CO is unwanted and so it is necessary to do a water-shift cleanup – where the CO is mixed with water. In this way, a little bit of energy is reclaimed however it is not very much. Still, CO₂ is made by this process and this goes up into the atmosphere or somewhere. Therefore, four hydrogen molecules are produced from this whole process, starting from CH₄.

So then what you want to do is burn the hydrogen, typically in a fuel cell or some other way for it to react with oxygen, to get 4 water molecules (H₂O). From this process, 948 kJ per mol of energy is released.

Now this should be compared with, for reference, just burning the methane directly. Well, here at the centre left of the slide is the same reaction presented earlier: methane plus oxygen, with the release of 817 kJ, a little bit different. However this is because some energy was expended and some energy was released, so that on balance this is taken into account. A table is presented here that shows how this occurs. Burning fossil fuel directly leads to 817 kJ. However, if it is first converted to hydrogen, and then burnt, then this leads to 817 kJ. Well, it’s nice to know that the conservation of energy still works: you couldn’t get anything else, because you have started from the same chemical state and ended in the same chemical state.

So the lesson here is that it is not possible to affect the supply of fossil fuels at all by converting methane to hydrogen. Even though one reaction burns hydrogen at the point of use – the reaction is essentially the same as just using hydrogen. So buried back here at the beginning, at the point of production, the process is using effectively the same amount of fossil fuels.

What about the other energy challenges? What about access and security? Of course, if hydrogen is coming from fossil fuels you still need to take as much out of the ground from whatever country it comes from, so there is limited access for fossil sources. It is the same for hydrogen as it is for fossil fuels. Furthermore, pollution comes from the impurities in the fossil fuels. Well, these must come out somewhere along the line, so using this approach doesn’t really affect the management of pollution levels from its source either.

What about greenhouse gases? Looking here, how much CO₂ is liberated per methane? If it is burnt directly it is 1; if it goes through the reforming and cleanup process it is 1. So this doesn’t affect the amount of greenhouse gases produced either.

This is a point that often is not appreciated: because the production of hydrogen, at least from fossil fuels, goes through this process, there really isn’t much scope for having an influence on the four major energy challenges for fossil fuels.


(Click on image for a larger version)

What are some of the other things that can be done? Well, hydrogen can be produced by splitting water. There is plenty of water in the world. This slide shows again an energy balance. When given some thought, it is possible to realise that it is necessary to contribute 474 kJ per mol to split water, and thereby extract hydrogen. The oxygen goes off somewhere else. Then the hydrogen can be burnt to reclaim the 474 kJ back – if everything was 100 per cent efficient, but of course, it isn’t.

There are lots of energy sources for splitting water. Non-fossil electricity can be used such as solar, hydro, wind or nuclear. Even solar or nuclear heat can be employed. People wouldn’t want to use fossil, for the same reasons as discussed before, specificallt because there would be little impact on the production issues.

It is also possible to liberate hydrogen by electrolysis. This can be done by photo-electrolysis, by dissociation, or by a thermochemical cycle. I will show some examples of this in a minute. As already said, the hydrogen can be used and converted into energy again through fuel cells, heat engines or combustion. So this approach is pretty versatile.


(Click on image for a larger version)

Now we will examine some of the challenges for some of the things that I have just listed. In the left-hand panel of this slide is conventional electrolysis. We know how to do this really well, because we have had 100 years of experience. We are all used to seeing a battery and two electrodes in a beaker of water. Well, they don’t really look like that in practice; they look like fuel cells. They are pretty much a fuel cell in reverse, and you see a diagram of this here.

So you could take electricity from some source, apply it to an electrolyser, which typically runs at around 100°C or a little bit lower, and produce hydrogen at one electrode and oxygen at the other one. That is pretty efficient, because we know how to do it so well. It is 80 to 90 per cent efficient in big units.

You can do a little better if you do it at high temperature – indeed you can use some of the thermal energy to replace some of the electrical energy that would be needed. So in the centre panel you can see a simple graph of how much electrical energy and thermal energy would be needed as a function of temperature. At 4,000°C, water is dissociated anyway. That is its natural state so you don’t have to add any electrical energy. But it is pretty hard to tap off the hydrogen at that temperature, and if you lower the temperature it just recombines. So what you really want to do is work at around maybe 650°C or a little higher to obtain some 20 per cent or so advantage from the thermal environment as well as increasing the efficiency by another few per cent.

This is basically a solid oxide fuel cell working in reverse, so there is even a way to achieve this, and a path forward.

The right-hand panel illustrates a much more interesting pathway for scientists. Instead of producing the electricity separately and then using a current to split water, some methods are aimed at doing this all in one step. For instance, a sun’s photon comes in to a site and then a semiconductor excites an electron up and leaves a hole in the conduction band. Then, because there is charge separation there, it is possible to produce an electric field or a voltage. A question then arises which is: why not use that process to simply electrolyse a neighbouring water molecule, so that out comes the hydrogen and oxygen?

You can do this in the laboratory, you can get 6 to 18 per cent efficiency – with materials that really won’t last at all in the real environment, but it shows that it is possible to do it. Interestingly, those efficiencies are actually somewhat respectable.


(Click on image for a larger version)

Another way to achieve this goal is to just take heat, which might come from a nuclear reactor or might come from a solar concentrator. Shown here are the three reactions of the sulfur-iodine cycle, which is one set of reactions that people talk about. The net effect is simply that water is split into hydrogen and oxygen through this sequence. It is a bit complicated but it works at a high temperature; the downside is that it is very corrosive. We can’t really demonstrate this very effectively at the present time.

There are some lower-temperature cycles that are alternates. These are not highly developed, but they are certainly possible.


(Click on image for a larger version)

One of the most interesting ways to produce hydrogen is biologically, or through bio-inspired techniques. Plants have been doing this for a long time. So what do plants do? They take a photon from the sun and water vapour from the air, and actually produce protons, electrons and oxygen. The oxygen goes out into the atmosphere – that is where all the oxygen came from originally that animals came to breathe. So we have plants to thank for that – and then the plants manipulate these protons and electrons through complicated biological systems to bring them in contact with CO₂, another photon, and make uphill reactions of carbohydrates and sugars like the one shown here. This is fuel for the plant.

Biology has been doing this, at room temperature, for a billion years. It is much better than what can presently be done synthetically/artificially. So what does biology use for its catalyst? Well, it doesn’t use platinum, it uses manganese in some form – much cheaper and much more abundant. We really need to learn how to do that; it would provide a nice source of hydrogen.

So it is possible to imagine going one step farther, just imitating photosynthesis. Why not change it slightly? Instead of making a sugar, carbohydrate, let it make hydrogen directly – indeed it is almost right there as CH₄ or an alcohol, as the output of such an artificial photosynthetic process.

Bacteria also produce hydrogen. In the lower part of the slide you see the famous example in which hydrogenase takes protons and electrons, and makes hydrogen, which goes either way. That is exactly the half-cell reaction that takes place at the anode of fuel cells.


(Click on image for a larger version)

So what is the outlook? Well, for the incremental hydrogen economy – that is, what we know how to do now – we can produce hydrogen by reforming fossil fuel (you see here such a production plant), we know how to store hydrogen as a gas or a liquid, for example, and we know how to use it in heat engines. So we can do all this now. But the thing is that although we can do it now, it doesn’t really have the impact on the energy challenges – supply, security, pollution and climate – that we really need to have.


(Click on image for a larger version)

What we need is the mature hydrogen economy. That is very different. In this model, hydrogen is produced by splitting water, not by reforming fossil fuels; and it is stored in a solid somehow, not as a liquid or a gas; and it is used in fuel cells, which are much more efficient than heat engines. So this model has a high impact on energy challenges, but we need to do some science to figure out how to get there. We need breakthrough discoveries in catalysis, membranes, nanoscale architectures and bio-mimetics. These are the things that people like us can research and develop for the next 20 years.


(Click on image for a larger version)

Here is a graph that says the same thing in qualitative terms. What is the energy payoff versus the research need? Well, at the left we have the incremental hydrogen economy that we can achieve now. At the right is where we want to get, but we have to do a lot of research to get there. Indeed when we do get there, there is the expectation of a much bigger energy payoff.


(Click on image for a larger version)

So that is the summary of the technical parts. Let’s just for a minute look at hydrogen market incentives.

The driver for the hydrogen economy really involves energy problems ‘in the large’ domain: supply, security, pollution and climate. They affect all of us, but they don’t affect one of us more than another. Importantly, there is a shortage of benefits for either individuals/consumers or businesses ‘in the small’ domain. Why would you do it?

Well, from the point of view of business, there is a big investment to convert to hydrogen, a long time horizon, an uncertain payoff, and not much incentive to do it. From the point of view of the consumer, there will probably be lower performance from a fuel cell car than that already obtained from a gasoline engine. This may mean lower reliability – and the need to take it into the shop much more often just to be tuned up. It is also going to cost consumers more as compared with fossil fuel. Where’s the incentive?

So you might say that this is a very big challenge. Well, it is, but there are some examples of how a challenge like this can be overcome. One is air pollution. In the last 20 or 30 years the air has become cleaner, and that is because some governments around the world have simply passed laws that say, ‘It is better for us as a society to have clean air, and we’ll pay the cost. We’ll all pay it.’ That only works if basically the law is a law that everyone wants; everyone has to agree that that’s a good idea. Then we can do it by everybody sacrificing a little bit.

The other interesting example is cell phones. I remember that when cell phones first came out in the United States they were called ‘car phones’, they were pretty big, you stuck them in your glove compartment and they took up the whole space. About the only thing people used them for was to call in the case of an emergency, a breakdown. At this stage not too many people had them.

Cell phones are very different now. We all carry cell phones in our pocket. I can call the States on this cell phone that I carry, and there are lots of other places I can call from Australia. I think it’s quite amazing that I can look at my email. It provides a service that we want, and we are willing to pay for. Now suddenly there are cell phone towers all over the place; they spring up miraculously.

So here are two routes by which we might create the market incentives for a hydrogen economy.


(Click on image for a larger version)

Well, will it succeed? That’s a question everyone asks, and of course no-one knows. But if you take a look at the history of coupling energy carriers with energy converters, you see an interesting pattern.

The first thing that made such a big impact on the energy system was coal plus steam engines. These were a very good pairing in railroads, steamships and factories, that started at around 1750. There was a big revolution when oil came along since it worked so well with the internal combustion engine. These systems replaced the steam engine in a lot of applications, especially for transportation, at around 1900. Also electricity works really well with motors and lights, because you can very efficiently convert electricity – it’s clean – to motion and to light for seeing. That happened around 1890 or so, and these things have really caught on.

What about hydrogen and the fuel cell? Well, hydrogen as a fuel and the fuel cell go together in just the same way as the previous energy carriers and converters do. They make sense together, they are very efficient, they are very clean – there are a lot of reasons why you might want to do this. They could replace some of these old uses, as the internal combustion engine did, and create new uses. The trick will be to find the compelling market incentives to make this happen.


(Click on image for a larger version)

So that was a little bit of a philosophical outlook. Here is the summary perspective – in a form that follows pretty much what I have already told you.

Hydrogen can meet the energy challenges as a carrier. There is an incremental hydrogen economy that we can put in place today, but it also has an incremental effect on energy challenges. We need the mature hydrogen economy and the accompanying slide shows what it means. There are research challenges to getting there.

With that I think I will close. Thanks very much.

Chair: Dr Michael Barber – I hoped this morning’s session would set the scene and indicate the thesis of the day, and Dr Crabtree has done that superbly. I am sure he would be willing to entertain a few questions.

Ben Hankamer– Given the various options that there are for future energy markets, how do you see the hydrogen economy developing? Do you think there are any other major directions, for example in the US, that are being entertained?

George Crabtree – Sure. The hydrogen economy, I think most people agree, is pretty far off. There was a wave of optimism a few years ago that maybe it was only 20 years off, but I think that is an optimistic statement. We need to find things to do. There is no silver bullet for energy; we need to take every reasonable approach.

One beautiful approach is hybrid cars. That is a no-brainer. Sure, we should do that.

The next step after hybrids is plug-in hybrids, so in fact you can charge them up at night and drive at least 10 miles, let’s say, or some short distance the next day without using any fossil energy. You have to look at the production side – that means that wherever you get the electricity from, which is probably a fossil source, is now driving your car so maybe you haven’t done so much, but it is a great option for going forward.

And when it comes to oil peaking – and I didn’t realise this until about a year ago – people say that at $30 to $40 a barrel it makes sense to convert coal to gasoline. So we’re there now, and I think the reason it is not happening is that companies are reluctant to invest the huge billions of dollars that are required to start that process, until they are sure that for the long term that will come back.

But yes, we need to take lots and lots of other looks at it.

Michael Dopita – Today there was news of the leakage of the IPCC report conclusions that are pointing towards a 3°C rise in temperature by the end of the century. In the earlier report they pointed out that the world should try to constrain itself to 2°C, and Hooper’s book on avoiding dangerous climate change points to a number of tipping points in the climate, anywhere between 1°C and 3°C. What this is saying is that we are apparently locked into dangerous climate change, even today.

The question I would like to ask is, basically: do you think that at $70 a barrel, $80 a barrel, the incentives are sufficient to drive change at a speed that would actually enable the hydrogen economy to come in on a shorter time scale and rescue us from this dreadful fate?

George Crabtree – That’s a great question. I come from the US, where everybody likes to drive. They use plenty of gasoline, and you may find it interesting that $3 a gallon is considered a high price in the US. But the question is: how much does gasoline price have to rise in order for people to cut down driving? I think that’s really the thing that counts. My own feeling is that a factor of 2 might have an effect; I don’t think a smaller amount would. People would just pay it. Will it be a factor of 2 higher soon? I don’t know. That’s quite a rise and it might be a few years before it hits that level. But that is what it is going to take.

Consistent with what I was saying earlier, there are some research challenges, not just economic ones, before we get to a viable hydrogen economy. They are going to take a while. So even with high prices of oil, I think we are talking decades.

Barbara Hardy, SA Adelaide, NGO – You have just mentioned the word I wanted to ask about, ‘economics’. Is there any hope – not in the short term, but the longer term, maybe – of externalities being included in the costs of fossil fuels and those sorts of things, so that you then bring the economics of a desirable energy future, comparing it with what we are currently enduring, which has a lot of externalities in sickness and pollution, climate change and all those sorts of things? The United Nations, I understand, is doing some work on this, but I am wondering what you think about it.

George Crabtree – What a good idea. If you put a tax on carbon in the air, or just a tax on gasoline, it would certainly go in the right direction. And that would make us realise that the health cost of pollutants, let’s say, is borne by everyone individually. Why not just take them out at the beginning, put a tax on them, to encourage the right kind of behaviour? I think that’s a wonderful idea. Many governments are reluctant to do that, unfortunately, but it does seem to be the right thing to do.

Fred Mendelsohn – Could you comment on a mixed model? In Australia we have got a lot of sunshine, but wind and solar have the disadvantage that they are not good for base power production. But you could envisage electrolysis of water, even on a fixed facility, being a good way to store the energy in between the peaks.

George Crabtree – That is a very good point. The alternatives to fossil are either nuclear or renewable, and renewable has what you just mentioned: it is intermittent. That is an issue, so what do you do about it? You have to learn how to store that energy. One way is to store it as hydrogen, through electrolysis, and that is quite a reasonable alternative.

What we don’t do very well is store large amounts of energy. Just the daily up and down of power plants and power demands – power plants are ramping and down every day but it would be nice to have them run at the same level all the time and just store the excess or draw from the storage, from the resource, when you need it. That means you would have to store, let’s say, a gigawatt-day, because the output of a power plant is about a gigawatt and a day is sort of the cycle over which you want to store it. We really don’t know how to do that very well on such a scale. You could imagine storing it through hydrogen. That would be a lot of hydrogen.

I think that is also a fundamental science challenge: where is a good way to store electrical energy? We really don’t know how to do that.

Pete Griffith – My question relates to energy storage for hybrid cars. I understand the French are producing cars with compressed air as a stored energy. Would you care to comment on that?

George Crabtree – Well, I haven’t heard that, but compressed air on a large scale I have heard about. On the question of storing a gigawatt/day, if you have an underground cavern that is big enough, and you use electrical power during the night to pressurise it (usually you want to pressurise it, I am told, not only with pure air but with a little natural gas) then when you want to take it out, the pressure drives a turbine but not only the pressure, you ignite the stuff and so you get a little combustion driving the turbine as well. There are storage locations where that is being done, and it is an interesting alternative. But it is difficult to find a cavern that is big enough, and it is expensive to install the machinery.

I haven’t heard of the idea of driving a car that way. I have a feeling that you will have a limited driving range.

Andrew Holmes – Thank you for your lecture, Dr Crabtree. This is a comment, more than a question. I am sure you are aware of Dan Nocera’s calculations about world energy production and demand. Nocera’s prediction is that we are not going to make it with all the energy sources that are available at the moment, on the prediction of your curve. That is why I would like to thank you for drawing attention to solar energy. I really feel that in this meeting and in this country we haven’t addressed that issue particularly strongly, and this is a plea to the community to think hard about photovoltaics, low-cost large-area photovoltaics, as an equally important scientific challenge for the future.

George Crabtree – Very good comment.

Michael Barber – Of course, in two years’ time we will need another A-side Symposium, and we have some new Fellows with some very interesting issues. I am sure we could have an appropriate Symposium!

Thank you, George.

Symposium program

Other speakers

Dr John Wright
Setting the scene: What is the hydrogen economy?

Professor Cameron Kepert
Hydrogen storage in nanoporous materials

Dr Sukhvinder Badwal
Fuel cells

Professor Andrew Dicks
Advanced nanomaterials for fuel cells

Dr Evan Gray
Hydrogen storage: status and prospects

Dr Ben Hankamer
Solar powered H₂ production from H₂O using engineered green algal cells

Dr Catherine Grégoire Padró
Production of hydrogen

Professor David Trimm
Catalysis and syngas for the production of hydrogen

Dr Wes Stein
Making hydrogen from the Sun

Professor Harry Watson
Hydrogen car prospects