AUSTRALIAN FRONTIERS OF SCIENCE, 2008

The Shine Dome, Canberra, 21-22 February

Metal-organic frameworks: From hydrogen storage to electronic and magnetic function
by Professor Cameron Kepert

Cameron Kepert has a degree from The University of Western Australia and completed a PhD from the Royal Institution of Great Britain, University of London, on a Hackett Scholarship. In 1995 he moved to the Inorganic Chemistry Laboratory at the University of Oxford as a Junior Research Fellow, where he commenced research into coordination framework materials. He was appointed to the University of Sydney in 1999 and currently holds the position of ARC Federation Fellow. He is the recipient of a number of awards, including the Malcolm McIntosh Prize for Physical Scientist of the Year, the Le Fèvre Memorial Prize, the Royal Society of NSW Edgeworth David Medal, and the Royal Australian Chemical Institute Rennie Medal. Cameron’s interests include nanoporous coordination framework materials, molecular conductors, molecular magnets and negative thermal expansion materials.

(Click on image for a larger version)

By way of broad introduction I want to take a step even further back beyond nanoporous materials and talk about host-guest systems. As chemists we are not so much talking about pathogens and plants; we are more talking about the sorts of systems that we have heard about this morning, and even on a smaller scale than that, where we have some small guest molecules that interact in some way with some larger host structures. It is these interactions that lead to a whole lot of very interesting and very useful sorts of chemical processes, whereby the host structure can, for example, recognise the presence of the guest so we can get molecular sensing and the host can actually then cause the guest molecule to behave in certain ways. We can achieve the separation of different guests, and we can even cause chemical reactions to occur within these host structures. So we can control chemistry, using hosts.

To go back a little while, before chemistry really got its act in gear in trying to make new systems: of course, there were already in existence a large number of highly inspirational host systems that occur naturally. Zeolite materials, these porous aluminosilicates, as shown at the top left of the slide, have holes within their structure of the order of size of one nanometre, large enough for some interesting molecules to go into these holes. Illustrated below them we have clays that expand and contract as water goes between the layers. We can even have beautiful nanotubular systems, such as the one shown at the bottom left, occurring naturally, that are formed in these geological time periods.

Humankind has extrapolated on these materials to great effect. Chemists have taken zeolites and generated a whole lot of other, new materials that are similar to these, and they form the basis of a large number of really very large-scale industries petrochemical cracking, gas separation, the exchange of radioactive cations and so on. So chemists have made very good use of extrapolating on the materials at the left of this slide.

If we turn towards what biology has achieved, we see that while biology has also had some very long time periods, it has had the very powerful mechanism of natural selection to select against the performance of these host systems. At the right of this slide you see some examples from biology: we have enzymes that catalyse reactions very selectively, we have molecules like haemoglobin that can transport oxygen, and really amazing systems where we can pump protons around using irradiation by light.

And again, chemists have made use of these sorts of systems for applications, for example in bioleaching you can extract metals in mining by using bacteria and we heard yesterday that yeasts have all sorts of useful properties.

So where chemistry comes on board is to say, 'Well, we have the opportunity to actually create our own systems, and to have the fun of discovering how those systems behave. We can try and design' and I won't use the term intelligent design 'and create these new host systems and try and get some interesting new properties.'

What I am going to talk about today is really research that has come about, I guess over the last 10 years, where in some loose sense we are starting to bridge the two areas shown at the two sides of the slide. We are making materials that in some way resemble zeolites in having robust host structures that have host-guest chemistry, but we are now building some organic components into these structures, which is providing a lot of versatility and some very interesting properties.

(Click on image for a larger version)

So what do these materials look like? To take a simple coordination complex, this has some metals and some organic units in it.

(Click on image for a larger version)

If, instead of allowing this to terminate and just form a discrete molecule, we use a bridging ligand such as is indicated here, an organic unit that can bridge these centres, we can form a three-dimensional framework that has large pores inside the material and we can think about doing some chemistry inside of these. So this material really does look a lot like a zeolite, but we now have organic units within the structure.

(Click on image for a larger version)

If you grow crystals of these materials, these holes are filled up with solvent molecules, and if you have the right type of structure, you can actually then remove both the unbound and in this case also, importantly, the bound guests on these metal atoms within this material and you end up with a structure that is very robust, where you can start to think about putting other molecules back into those holes.

(Click on image for a larger version)

We have recently pioneered a technique that has given, arguably, the most accurate structural information on a nanoporous host-guest system, where we take single-crystal diffraction which you have just been hearing about this morning and we expose these crystals to a variety of different guest molecules during a crystallography experiment, and follow the uptake and release of the guests.

You can see here that we are binding a wide range of different guest molecules on these copper atoms. This is all done on one crystal. We can follow exactly how this host-guest chemistry occurs.

(Click on image for a larger version)

We can also look at what is happening in the larger pores, and see these highly dynamic molecules that sit in particular pockets within those pores. So we are beginning to get some incredibly accurate understandings of the host-guest chemistry of these systems.

These and other studies are pointing towards the fact that these materials do have a very reversible and highly selective guest exchange chemistry. You can use them to separate different types of molecules, and you can also use them to catalyse different molecules within these pores. We are just starting to see some commercialisation efforts along those lines.

So that is well and good. You can use these materials in much the same way that you use zeolites. But the really interesting question is: what can you do with a molecular approach that you can't do with some of the more conventional materials that have been around for a few decades?

We can think about what some of the really interesting guest molecules are at the moment.

(Click on image for a larger version)

Arguably the most topical is also about the most simple of guest molecules, which is actually the hydrogen molecule, H2. The reason this is such a topical molecule is that hydrogen has been put forward as part of a potentially clean energy cycle in which you convert chemical energy into electrical energy very efficiently in a hydrogen fuel cell, and this releases water no greenhouse gases. We can potentially use this in an energy cycle and replace fossil fuels.

The Academy of Science has just released a roadmap on where Australia might go in the hydrogen area, recognising that there are two major scientific challenges that must be met in order to make this a reality. Firstly, can we produce hydrogen gas cleanly and cheaply? there is some very good research being done on that at the moment. The problem I want to address is: how do you store hydrogen gas? Because it is a very volatile molecule, it is actually very hard to store this molecule in a small volume with a small mass.

The US Department of Energy has put forward the various requirements shown here if this is really going to be feasible. The one I really want to focus on is 6 wt %. That doesn't sound like very much, but hydrogen doesn't weigh very much, so it actually is quite a challenge.

(Click on image for a larger version)

In just the last four years it has been shown that these metal-organic frameworks completely surpass any porous materials that have been looked at, in terms of storing hydrogen. The reason for this, in a nutshell, is that incredibly high surface areas can be achieved. Because they are such low-density frameworks, we can go to more than 5000 square metres of surface area per gram of material that is five soccer pitches worth of surface area per gram and these very highly porous materials can take up to above 7 wt % of hydrogen onto the surface of these pores. So already we have surpassed the Department of Energy requirements that were put forward for 2010.

However, if you look carefully at the data you notice something: this only occurs at low temperature (this is down at liquid nitrogen temperature), and at one atmosphere of pressure we are only getting less than 2 wt %. We actually have to go up to about 50 atmospheres of pressure before we really reach saturation of the hydrogen gas on the surface.

The reason is that hydrogen is basically a very non-sticky molecule. It is very small, and hard to get to bind as a diatomic molecule to the surface of these frameworks.

(Click on image for a larger version)

So where we have come in is to start thinking about frameworks that have very specific types of functional groups on their surfaces, and in fact the material that I have just shown you has bare metal sites. An experiment that we did a couple of years ago, using powder neutron diffraction, was to show that as you gradually introduce hydrogen gas into this material, the very first site that the hydrogen gas goes to is directly above and below, attached to copper atoms in the structure. We have since done some other measurements on these systems, and we see that the energy of this interaction is almost optimal for what you need if you are going to have hydrogen storage at ambient, non-extreme conditions.

(Click on image for a larger version)

But there aren't all that many bare metal sites in this system, so as you continue to fill up the pores you then put hydrogen gas into the regions shown here.

(Click on image for a larger version)

At more extreme conditions, we then start to fill up the large pores.

(Click on image for a larger version)

And eventually, if you keep pushing hydrogen in at higher pressure and lower temperature, you can get up to about 6 wt % of hydrogen in this system, which is the requirement that is stated. But we are still a little way away from achieving that under the conditions that we would like.

(Click on image for a larger version)

To look at where we are at the moment: this slide shows the US Department of Energy target, which you can see is actually not as good as what we are currently very privileged to be able to use. Porous solids previously were low on this chart we just didn't have enough surface area to give us enough loading of hydrogen gas. Just in the last four years we have moved up around the centre of the chart with these metal-organic frameworks, and there is very exciting potential to get this even higher and to start to approach the Department of Energy target, in terms of both the gravimetric density and the volumetric density.

(Click on image for a larger version)

To think about some other interesting guest molecules: oxygen gas is one for which we can turn to biology for some inspiration, because biology has gone to quite some lengths to find ways to bind the dioxygen molecule. For example, in the blood of animals like octopi there are two metal sites that bind oxygen in the form of a peroxide anion, and this happens reversibly: you can reversibly attach and remove the oxygen.

We have recently been looking at some of these systems that you can mimic synthetically in the laboratory. Shown here is one with two cobalt atoms, where the oxygen attaches.

(Click on image for a larger version)

Building this into a framework material, we have something that is porous and a material that binds oxygen extremely strongly. If you have an excess of this material at room temperature, it will remove more than 99 per cent of oxygen gas from the air, and at the same time take up no appreciable amounts of nitrogen gas. It is an extremely selective process, because we are making use of a chemical reaction to recognise the presence of oxygen and extract it. It is actually a much more selective process than is currently used industrially for the separation of nitrogen and oxygen gas.

(Click on image for a larger version)

Other guest molecules of interest to us and I really want to use this talk to give you a broad overview of what might be achievable are those that are chiral: right- or left-handed. Can we start to make frameworks that are able to separate right- from left-handed molecules? These metal-organic frameworks have just in the last eight years provided the first ever nanoporous crystalline homochiral systems, where you can start to do that.

This is a recent one from my group, where you have chiral organic linkers with chiral functional groups hanging off. If you bridge these into a robust framework, you get a beautiful honeycomb-type structure.

(Click on image for a larger version)

This structure has one-dimensional pores that are large enough to do some interesting chemistry in, and we have chiral functional groups pointing out into helical channels. It is materials like these that have provided the first examples of enantioseparation separation of right- and left-hand molecules and we have also recently seen some work where chiral catalysis occurs in these types of pores.

So with this metal-organic approach you can do what zeolites do, and you can also, using the versatility of molecular chemistry, go beyond that and make use of the design aspects. You can build in particular types of surface functionality, and aspects such as chirality, to give you the types of properties that you are interested in.

But there are also other things that you can do using this molecular approach, which is highly versatile.

One of those things is to build molecular switching sites into these framework lattices, trying to target some entirely new materials phenomena by developing materials that have properties entirely different from what we have seen in the past.

(Click on image for a larger version)

We have been very interested in building in electronic switching, and have done that by using so-called spin-crossover centres, where we have metal sites that are able to switch, as a function of temperature, pressure and irradiation, between different electronic states. When you do that, you get a colour change, and a change in the geometry bond distances and so on so it is a very easily observed phenomenon. And we see some interesting repercussions when the framework actually undergoes these changes. We get a change in the pore chemistry as a result of that switching transition.

(Click on image for a larger version)

One example is shown here. We have a framework host where we have a variety of different guest molecules methanol, ethanol and propanol in the pores. You can see that these different guest molecules are perturbing our host structure in slightly different ways, and that is having quite a marked effect on the way that the host lattice is switching between a so-called high-spin electronic configuration and a low-spin electronic configuration at low temperature. It is measurements such as these that are giving us a brand-new insight into the subtleties of spin-crossover this switching phenomenon whereby a very small structural change leads to these effects.

If you like looking for applications for things, then we also have crystals which, if we expose them to a vapour, we can get to change colour at room temperature. So we can think of making single-crystal breathalysers, if that is the sort of thing you are interested in.

On the flip side of the coin, represented at the bottom left of this slide, we also have the first host systems that we can access by using external stimuli such as light, and so we have the first systems whereby the guest environment can be modified using external stimuli. So we may be able to control processes such as guest uptake and release, and maybe even the reactivity of those guests within those pores.

The last thing I want to talk about is another interesting property, which people have been trying to build into a nanoporous host for some time that is, spontaneous magnetisation, or magnetic ordering. Can we make materials with very small holes in them that are also magnets?

(Click on image for a larger version)

It has proved to be a very difficult thing to do, because magnets require magnetic exchange coupling between the metal sites, but porous materials need to have large open areas, and so these two properties had largely been thought to be mutually exclusive. But we have actually achieved this, quite recently, using one-dimensional ribbons that contain metal atoms. The ribbons are shown on this slide as running into the screen and being bridged by organic units, and you can see that we now have one-dimensional channels that are filled with water molecules within the host structure.

(Click on image for a larger version)

It is an incredibly robust material. You can remove the water molecules with there being almost no change whatsoever to that host structure, a very small change in the angles. The remarkable thing is that you go from a system that is an anti-ferromagnet, where all of your spins are aligned anti-parallel you have no magnetisation in this system to a ferromagnet whereby all of the spins are now aligned parallel. This is reversible: you can go from an anti-ferromagnet for the hydrated to a ferromagnet for the dehydrated, and back again.

That is quite a remarkable property, and it actually goes against what you would think, conventionally, that the guest molecule should do. Conventional thinking would be that it is the direct exchange coupling occurring through the ligands that is important, not through the weaker interactions that we have here at the top left of the slide.

We are using systems like this to try and get some better fundamental understandings of molecular magnetism.

(Click on image for a larger version)

Also you might imagine, if you are interested in applications, that there is some prospect that this particular mechanism could be used for molecular sensing, whereby the absorption and desorption of guest molecules had some fundamental impact on the properties of the host lattice that you could then measure.

(Click on image for a larger version)

I hope I have given you the idea that these are exciting times and there really are a lot of potential applications and new properties that we can target with these systems, and I would be interested if people came up with some more that we could think about.

It just remains for me to thank our past and present group members and a number of collaborators and funding bodies.