Computer modelling for molecular science – with Sir Richard Catlow

thank you [Music] [Applause] humankind has always needed and liked models and the scientific Community is no exception and indeed as I hope I'll show this evening model building is one of the oldest and most vital of scientific activities now modern computer technology has given us an absolutely fantastic possibilities to build realistic models of complex systems and computer modeling now pervades the whole of science and engineering and what I hope to do is to give you just a glimpse of the way in which this technology is being used and its impact in on science and engineering but I also hope to show the the what we're doing now with computers is just the modern way of doing what scientists have done since the beginning of scientific thought so let's look at some models here is the ptolemaic model of the universe it is of Highly complex it's a three-dimensional model it's a highly complex model with intersecting trajectories and spheres and it rationalized some of the astronomical observations in the ancient world it's an earth-centered model I think as you'll see on the slide what has though some sympathy with the king of Aragon I think it was in the 12th century who said when he saw this model if I had been the Almighty I think I would have tried to make something simpler and in fact the King was right because the ptolemaic model was Swept Away by the heliocentric copernicum model in the 16th century which a much simpler model and which rationalized a wider range of astronomical data but it was a model and it was a useful model let me now just look at some models which in the Royal societies collection you'll see there a model of the transit of Venus that's very important astronomical observation made in the 18th century and then a very early model helping to understand terrestrial magnetism and here are some other models these are beautiful models used as Machi mania's teaching AIDS models of marine creatures again I think from the from the last century so modeling I say is all pervasive and let's now move to modeling models of molecules and and crystals now what we have here is probably the first ever molecular model this very same model was shown here in the Royal Institution in a discourse given by Hoffman in April 1865. the discourse was on the combining power of atoms what we would Now call valence it's a remarkable model it's a model of methane carbon in the center black four hydrogen atoms and in fact Hoffman had got the valence the combining power of carbon right it does indeed bond to four hydrogen atoms in the methane molecule he got the stereochemistry of the shape wrong in fact it's tetrahedral but he didn't have the means then of knowing that but that is a real landmark in the development of chemistry let me show you another model that's a landmark this is a model of license time an enzyme the structure of this enzyme are the molecular level was determined here in the Royal Institution by David Chilton Phillips and his then student Louise Johnson who became one of very distinguished molecular biologist it is a landmark in the development of molecular biology and I should say that David Phillips and Louise Johnson were very much helped and encouraged by Lawrence Bragg the then director and I'll come back to some of the other many achievements of brag Bragg later on but say these are molecular models and on the screen you'll see more molecular models you'll see Lord Kelvin uh with models of close packed structures there again we'll come back to those later and you'll see Harry croto the Nobel Prize winner in chemistry with models of Carmen crota won the Nobel Prize therefore demonstrating that carbon could form these beautiful open structures including the famous Buckyballs and I should mention that both Kelvin and Harry Crowther lectured regularly at the ri and in fact were very good friends of the institution so perhaps we've got a feeling now for the importance of models in science and a long-standing history why do we need models in science why do we need models in general and we need them because models scale objects and processes to a size that we can understand and be familiar with they scale length and time scales that we cannot imagine tiny unimaginable fractions of a second which is what the time scales of which some processes take place or immensely long time scales which again we'll see in a few minutes is some processes in cosmology take place over billions of years but these models allow us to scale them and they allow us to scale enormous objects like the universe and Tiny objects like atoms nuclei and subatomic particles to a size that we can understand and be familiar with so that's why we really need models and let's just think about computer modeling computer modeling Bridges fundamental Theory with experiment because of the triumphs the many triumphs of science we understand many of the fundamentals in the universe we understand what I'm not saying we understand but we can predict and calculate how electrons behave in atoms and molecules we do that using quantum mechanics now Richard Feynman famously said nobody understands quantum mechanics and I think he's right but in nevertheless provides as a Machinery to predict how electrons behave in atoms and molecules and if we look at planets and stars we can use gravity and relativity if we're talking about engineering aspects fluid flow we have the navier Stokes equations so we have this fundamental knowledge and we can now deploy this using computers to predict complex reality so let's start looking at some complex reality um I say computer modeling Bridges Theory and experiment as we're showing here and I'll give you an example of a very bit of complex reality there with the mesoporosilica system and I'm going to come back to later on but let me know start with modeling the biggest things modeling in fact the universe and I'm going to show you now some wonderful modeling results I'll show them in visual form obtained by Carlos Frank a world leading cosmologist at the University of Durham who was used throughout his career computer modeling to try and understand the evolution of the universe so what we'll see here is how initially a low density Clovis is early on in the universe starts to condense I just absolutely love this video it starts to condense you can now see by the way hundreds of millions of years of passing here uh you can now see structure beginning to emerge and we're now at the beginning of star formation and you can see gravity beginning to pull these elements together um absolutely wonderful so it's pulling more material in we're forming more stars those are the red bits and we see real structure beginning to emerge and we've then got kind of embryonic Galaxy forming but more is going to happen this is now going to start to mop up more material around it sometimes it'll throw the material away um and again we now see a really well-formed Galactic structure emerging and this is all okay I think since I started this video probably about two or three billion years has passed uh and now we have this beautiful galaxy and this was obtained by computational modeling by Carlos's uh group who've done such wonderful work in this field and we now get we kind of move away and we get a feeling for this part of the universe and we can see several of the galaxies that have condensed over this period so that's modeling the biggest things things are the Scale of the Universe modeling processes that are taking place over billions of years now let's move to the global scale and we're now going to talk about modeling the Earth in particular I'm going to talk about modeling the Earth's oceans I should express my gratitude here to the UK oceanographic Center for helping me with this now here we have the Earth just to say that the ocean is understanding the oceans are hugely important as it says here a massive store of heat and carbon biodiversity we need to understand the oceans we need to understand how the oceans evolve how they're going to be influenced by climate change and in particular we need to understand the currents in the ocean now this is a model of ocean circulation we're looking at the North Atlantic you'll see at the kind of top there a stream is beginning to emerge from the Gulf it's traveling across the North Atlantic and shortly it will reach Northwest Europe and it's a good thing it meets Northwest it reaches Northwest Europe because it helps us to keep us warm particularly in the winter now we'll move on we'll look at some of the currents taking place in the Pacific and I said these models they all fit in I should say that the same with Carlos's models they all fit in with experiment but the end help us understand the experimental observation and this kind of information crucial in understanding how the oceans evolve and how the oceans interact with the atmosphere extreme weather forecasting absolutely vital computer modeling plays a key role here this is an example of a prediction that was made of a current hurricane some years ago and if we can predict when these extreme weather events take place then mitigating action can be mitigate the action can be taken so again computer modeling playing a key role here and I should say that men meant the predictions we make about climate change are based on detail computer models let me just before I leave this I'd like to show you another video which I really like this is one about how the ice evolves and the Sea temperature evolves in What's called the Greenland Scotland Ridge and again it's a beautiful video we're starting I think uh the it's in time a kind of late Autumn and let's see what happens but what you'll see there can you see that white bit that is now the ice sheet beginning to grow you see the red bit retreating as the ocean cools and again this I should say really helps us understand what is going on in this very complex system now I think we're getting to spring now the ice sheet Retreats and you'll see the red bit which is the hot the warmer ocean beginning to expand and I think we got to the end of a another season now in the ice sheet begins to grow but again these models absolutely vital in helping us to understand um understand the the evolution of the ice sheet evolution of ocean temperatures and again how they are affected by climate change let's now move down a scale and we'll start talking about modeling at our scale and the think here of modeling relating to planes medical applications Hearts Bubbles and bones and I should Express here my gratitude to the UK supercomputer facility in Edinburgh the Archer facility now let me show you this video first this is a video of what is happening around what's called a cavitation bubbles these can occur in engineering applications they can occur actually in our blood and it's really important to know what happens when they meet a shock wave and this video will show you that in Exquisite detail you see the way of propagating outwards you see all the Ripples and Eddies and this kind of information again is valuable in both engineering and in some case in in kind of Health applications so very detailed modeling then I should say this was carried out by the Dr tutsinanis at the center for computational engineering Sciences in cronfield University here's a modeling of turbulence understanding turbulence absolutely crucial in aircraft design and here we're seeing turbulence in a supersonic mixing so this is kind of important for the development of supersonic Aviation and really the thing to take away from this is the immense complexity as these flows meet each other cause this turbulent flow cause all kinds of eddies and as you see here chocolates are generated due to the turbulence compressibility but really the thing to take away from this is the immense detail at which this modeling has given us so just a final example and one again relating to biomedicine this is a just a snapshot I'm not going to show you a video here this is what the inside of our blood vessels look like you have these little protuberances here they play an important role in controlling the transport of material in the blood and understanding how they behave understanding their Dynamics is really important this is just a snapshot of a dynamic simulation and finally from the medical applications well not finally nearly finally we look at simulation of heart of a heart now again this is a very very important and widely used technology to understand the Dynamics of Heart of essentially at heart pumping and I understand perhaps how that Dynamics can be perturbed when damage is done to the heart so we'll just see this is again a video of a simulation and I say the simulation give us information about what factors can affect this process on which Our Lives all depend and then finally let me show you this this is uh simulations applied to bone structure again engineering applications understanding the stability the strength of bone understanding how it's Modified by disease this again this work carried out at the Archer supercomputer very very important in helping to develop treatments for bone abnormalities well those are just a glimpse of applications in the field of engineering and in the field of Health what I want to move on to now for probably most of the rest of the discourse is modeling molecules and materials because this is what I do it's what I know most about and what we want to do is to learn about the structure teachers at the atomic and molecular level of molecules and crystals we want to understand their Dynamics we want to understand their reactivity now again we have the fundamental knowledge to allow us to do that over the last 60 years or so we've built up huge databases of how atoms interact with each other and we can use those we can feed them into the computer we have quantum mechanics which tells us allows us to calculate how electrons behave within atoms now say we can use that fundamental knowledge we can feed it into the computer and it can produce as detailed and increasingly accurate models of structure Dynamics and reactivity of matter at the atomic level let me just give you one initial example and this relates to of course to the crisis that we've been through in the last two years this the covid pandemic and I'd just like to actually take a step back and say that the global scientific community and those who select you who support the scientific Community should feel very proud of what the scientific Community achieved during the cobit pandemic almost all components of that Community responded I think superbly to the challenge and the fact that we've emerged uh we hope from the pandemic owes a lot to the commitment of scientists and this is one example this is an example of understanding at the molecular level how one of the key proteases that's vital in the reproduction of the virus can be inhibited by inhibited drugs and let's just have a look at the video the first thing to notice is that it's a dynamic entity I'm going to come back to that later you know actually molecules are static assemblies like this lovely model of lysosome the dynamic entities here you'll see the inhibitor model in the active site of the enzyme we can really get detailed information of how it behaves and we can also predict if we model modify that molecule will it improved the efficacy of the inhibitor and this was carried out by scientists at King's College in collaboration with Scientists at University of Southampton but now let's move on to modeling of materials at the atomic level because this is my specialty and we can do a great deal as I hope to show you in a few minutes we can model structures of really complex molecules and materials at the atomic and molecular level we can model less surfaces of course a lot of what happens the importance of a material is what's happening on its surface and we can we can do a terrific job there now we can model defects I'm going to show you later that although we look at these lovely models of crystals with a nice regular Arrangements real crystals always contain defects missing atoms additional atoms and those defects can play a very very important role in controlling the properties of the crystal we can model how molecules enter and diffuse into porous materials we can understand in fact the process is taking place during the synthesis of molecules and materials nanochemistry and I'm going to come back to later absolutely vital to modeling has been absolutely vital role there and I'll give you examples later and then reactivity and catalysis again I'll show you how we can understand chemical reactions in detail using the Contemporary power of modeling but let's now think a bit about crystals now let me introduce you to this model this is a model of sodium chloride it is the original model of sodium chloride that was developed by these two great Geniuses William and Lawrence Bragg when they developed the science of crystallography using x-ray diffraction and sodium chloride was one of the first structures that they solved it's a beautiful regular cubic arrangement of atoms the white ones let's say are sodium and the red ones are chlorine so this is I say the first model of this of this widely it is important and widely studied Crystal and I say it was determined by William and Lawrence Bragg both of whom subsequently became directors of the Royal Institution but I'm asking a question here why do atoms when you try pack them together in three dimensions why do they order why do they form these lovely regular arrays and the reason why they do is simply because that is the most efficient way of assembling them in three dimensions and I think we're now we're going to try and illustrate this with a very simple model here now we got these uh black balls that Thomas is showing you imagine those are atoms and we've got a little bit of a template for them to nucleate on and we'll start dropping them onto this surface and let's see what they do apart from dropping on the floor let's see what they do what you see they start to slot into place and they start to form more kind of nice regular layers now they're doing that not because Thomas Mike have told them to do that they're doing that because actually this is the most efficient way for them to assemble so there we have well done [Applause] foreign structure so that's a dense structure really we're just packing the atoms together as tightly as we can we'll see in a few minutes other structures which are much more open but nevertheless still the ordered way of arranging the atoms is the most energy efficient so we understand in general terms why we form these regular three-dimensional systems but can we predict them now here we have actually quite a famous piece of science journalism this was a News and Views article in nature uh in 1988 uh it was made by John Maddox he was a very celebrated editor of nature um and he said one of the continuing scandals in physical sciences is that it remains impossible to predict the structures of even the simplest crystalline solids from a knowledge of their composition now when I read that in 1988 I was pretty annoyed because I didn't think it was true that we were making progress in being able to predict Crystal structures so I got in touch with Maddox and said well I don't think you quite got this right and he very kindly said okay well you write me an article in nature to show what you can do in predicting Crystal structures so together with my colleague at UCL David Price who wrote this article which was published in 1990. so Maddox had come up with a pretty inaccurate piece of science it was a fantastic piece of journalism absolute because it says stimulating me and David Price to respond with a detailed article in nature and he actually stimulated the whole field lots of people said we're going to show that Maddox is wrong um and there's another article there which my colleague Scott Woodley and I produced her the 20th anniversary of the Maddox challenge so I say rather dodgy science but absolutely great journalism and it did a good job it stimulated the field now I want to talk give you illustrations of structure prediction by introducing you to a class of crystalline solids that I've been fascinated with since early in my career and in which a great deal of work was done here in the Royal Institution in the 1990s um these are zeolites the zeolites are aluminosilicate so what that means is the built up out of simple building blocks silicon surrounded by four oxygens at a corners of a tetrahedron and aluminum surrounded by four oxygens again at the corners of a tetrahedron and what do we do well we start to fit those together and we fit them together just by these tetrahedra sharing corners and in fact almost the whole of the hugely diverse silicate and aluminos silica chemistry is based on that principle so we fit them together by sharing corners but then we build up these beautiful networks that contain cages pores and voids and there's some of the loveliest structures in the whole of solid state structural science but why are we interested in them we're interested them just because they are such fascinating structures but also immensely important in Industry they're important for three reasons the first you see they have these channels in them now these channels are of molecular Dimensions they're about the same size as a block particularly of lots of kind of molecules in chemicals and we can use them for separation and it's a little known fact that something like 30 percent of the chemicals industry is about separations these are wonderful separations they're just sieves they can separate molecules of different sizes and shapes so they're using separation Technologies even more important is their use in catalysis they for instance they will break down the long chain molecules that are in the tar fraction of oil that break them down into the short chain molecules that are in gasoline petrol and it's a commonly quoted but true statement that every drop of Petrol in a car has seen the inside of a zeolite so they're used in very extensive in the petrochemicals industry but they're used in Fine Chemicals as well and the reason why they're so important is that the catalysis is they contain sites within them acid sites and metals that can promote chemical reactions but then the nature of the chemical reaction is controlled by the shape of these pores it's known as shape selective catalysis there's a third application that's very important as well it's the oldest and probably simplest application and that's iron exchange many of these zeolites when you make them they contain metal ions in their cavities and you can exchange them for other metal ions and they're used for example in water softening if you have a water softener in your house you look at what its contents are it will say aluminosilicates and those are zeolites they use as softeners in detergency as well so really important industrial materials but absolutely fascinating for structural scientists now can we predict their structures well I'm going to give you one example here this is work of my colleague Scott Woodley in fact who was a postdoc here at the ri for several years is now as you see a professor of computational chemistry and physics at UCL and he really founded the use of a an approach known as genetic algorithms for predicting the structures of crystals it's it's a lovely idea actually it's kind of based on evolutionary theory so what you do if you want to predict the structure of a crystal you start off just by generating lots of different structures lots of different perhaps plausible structures and then you allow them to breed you lie with them to exchange features between themselves but you only allow the ones to breed which have the best features you have some simple we call it the cost function you have a simple way of assessing whether the structure is likely to be a good structure you allow the good structures to breed and exchange information and as you run through this simulation the structures get better and better and at the end you do a more sophisticated calculation to predict the final configuration of the structure so it's a it's a neat idea and they say it's based on kind of evolutionary theory well Scott very much pioneered this when he was at the ri for predicting the structure of crystals I'm just going to give you one example it's an early example it's one of these zeolites These are the early stages of the simulation it is kind of feeling its way it's trying to find different ways of arranging it's now found a really good way of arranging them and it drives it downhill in energy we'll have a look at it again I like this one so much so I say this is early on in the simulation it's trying to find what's the best way of arranging these it's slowly getting better as it passes through subsequent generations and he ends up with this structure which is in fact a very accurate model of the structure of the minerals soda mineral and synthetic material sodalite so that's an example of structure prediction and over more recent years these techniques have been used to predict structures that we simply didn't know and then have been verified experimentally let me show you some other examples another technique again which I think is easy to understand is called simulated annealing and what you do here is using the computer you take a model and you heat it up you heat it up till it melts and then you cool it down and it will cool down into a number of different configurations and then you try and find the lowest energy configuration and that's your prediction for your structure and beautiful work was done here by Caroline Mello who you'll see there who was again a very good friend of the ri and visited and worked here for 15 years or so ago and she applied this technique to this plus a material that are known as metal organic Frameworks and what you do is you have I don't want to burn you with Too Much Chemistry on the left hand side you have can you see that octahedron that's got a metal atom in the center and it's surrounded by six oxygens then the next to it you see an organic molecule and what you can do you can start to fit these together a very lovely area of structural chemistry and you can make lots and lots of fascinating porous structure well Caroline applied these methods and successfully predicted structures which we then then she was able to show rationalized experimental data here are more of the structures that she predicted really beautiful examples of structural chemistry complex structures here perhaps is the most complex beautiful open structures really illustrates the beautiful beauty of her solid-state structural chemistry this work we should do with her colleague Gerard fere was published some years ago but again this was using these computational simulated annealing techniques which helped to solve a structures rationalize the experimental data now let's go back to our friends the brags what you're seeing here is a rather older version of Lawrence Bragg when he was director here at the Royal Institution and what we're going to do well I'm not going to don't do any of the hard work around here what Mike is going to do is show you one of Bragg's very famous demonstration missions which helps us understand what happens when atoms pack together in crystals imagine each of those bubbles is an atom and you see that they are they do form quite nice regular arrays but then you see that the r defects in those arrays can you see some of them I said earlier wrong didn't I that although crystals are regular Arrangements of atoms in three dimensions nevertheless we do get defects in them we get missing atoms and I think we may be able to see kind of lines of defects as well when different parts of the different parts of this ghost the the the soap bubbles on the raft are interacting with each other but we certainly see that although there are lots of ordering Arrangements of atoms there or we can now see a really nice line emerging there when two different regions are interacting and this actually is a pretty good model for what happens in real crystals um and in fact what we're doing now is developing what we would call a void in the crystal um you do get voids formed at Chrysalis particularly if you irradiate them but that's a lovely demonstration and it's bringing home yes that the these are representation of atoms they do indeed form regular Arrangements but there are defects missing atoms and then line defects in this Arrangement now Mark's done a fantastic job let's see how well Lawrence Bragg himself could do so let's kind of bring Lawrence into the theater he'll hear his voice in a minute and he will try and rival Mike with the bubble raft okay shall we make a start so there you see Metals produced form a kind of crystalline raft which is quite regular because the bubbles are so uniform in size here is a latest stage in the formation of such a raft the pattern has Rose in three directions just like the model made of Spears when there are bubbles forms it's generally in portions of pattern which are not parallel or as we should say are distinct crystals these meet at Crystal boundaries it can be seen in this draft that the rows are in different directions in neighboring crystals this pattern of crystals in this large eraft is very similar to the pattern of crystals seen in an etched metal thank you well I think we can give Lawrence a brag pretty good marks there he did nearly as well as mine but it also shows that as well as a scientific genius what a fantastic Expositor of science Lawrence brag was any of that was done here in the ri I think in the 1950s um well what brag was shown there that we get defects and what we would call microstructure so I want now to move to how we can understand this using contemporary computer modeling and I'll highlight some work here done by Regina mapanga anputinuepe who at the University of Limpopo in South Africa with whom I have had a collaboration now for well over best part of 30 years they run a very effective materials modeling group up in the northeast of South Africa so Regina visited us for a period probably about 15 years ago and started this work in collaboration with Dean and fee sale who again both worked in the 1990s at the ri now this work concerned a very interesting material as manganese dioxide it's important actually among for a number of reasons it's potentially used in well he's in fact used in batteries I'll come to batteries later on again for the structural scientists it's a really fascinating material again it's based on octahedra manganese sits in the middle surrounded by six oxygens but you can fit them together in lots of different ways and you're seeing two different ways here but we wanted to generate a realistic model of this just in the way that Mike and Lawrence Bragg did a realistic model for this structure because it won't be perfect and we used in fact a method developed by Dean sale what he did was first can you see on the top left hand side there you have a model for the crystalline material and what you do computationally is you heat it up and it melts and at the bottom on the left hand side you have a snapshot of the molten structure which has got a disordered arrangement of atoms and of course the atoms are moving around then you take and go to the right hand side now you take that molten structure and you start to cool it down and you apply a little bit of pressure and it starts to crystallize and then when you get to the low temperature it will have crystallized but then you look at the model the crystalline model and what do you find you find first that it's got these features here these are defects we call these what we call vacancies they're just missing atoms that when when it cool down under pressure there wasn't time for it to arrange as well as it could have done so some atoms are missing from the place where they should be and that's what happens in real crystals and let's look at the next one this is what we call a ground boundary you remember in Lawrence Bragg's video there you sure you saw those kind of bit disordered regions between two regular crystalline regions we call those ground boundaries simulated by the bubble raft but here we get an accurate model simulated by computers and these again these can control a lot of the properties of the materials and you can see here this lovely grain boundary structure that's predicted by these models so using the same kind of Concepts that might demonstrate in the Lawrence Bragg demonstrated we are using the computer able to generate these realistic models of the microstructure of this material now before I move on let me show you this this is a model of a glassy materials I've talked a lot about crystals crystals are ordered Arrangements of atoms in three dimensions we just learned but they do get elements of disorder within them now glasses are disordered Arrangements of atoms in three dimensions and this is a model for glassy silica silicon dioxide it was developed by one of my colleagues Ben arm vessel probably about 30 years ago you see it's still got tetrahedra so it's still based on Silicon surrounded by four oxygens linking together as we showed with the zeolites but you don't get a regular ordered pattern there's still structure there but it's a disordered pattern that actually turns out to be pretty a good model we think now for amorphous silica how is it how is it generated well it was generated by a computational version of the real way in which you make glasses you make a glass by starting from a crystalline material this is ancient technology you heat it up till it melts and then you quench it quickly and it doesn't have time for the atoms to organize themselves into the regular crystalline Arrangement that's what you do on the computer you take your model for crystalline silica you heat it up it melts and then you cool down computationally your model and it forms this actual turns out to be pretty good model for glassy silica it's not order but I still think it's really quite a beautiful model now let's move on to another really important area of contemporary science and that is Nana science and again I want to give you illustrations what from a few years ago from this guy here side Hamad who was a PhD student here about 15 20 years ago he's now professor of chemistry back in Sevilla and he worked on developing models for Nana structured zinc sulfide again I don't want to burn you with Too Much Chemistry zinc sulfide is an important material it's a semiconductor it has a number of applications and a lot of interest about its structure the Nano scale and he used this kind of simulated annealing method heating it up cooling it down quickly and what he found was at the Nano level the zinc and sulfur atoms to re-arrange to form these lovely open structures see the one on the bottom right hand side I think that's got 50 zinc and sulfur atoms in but it's still a kind of bubble like structure that actually came was a great surprise to us because if you look at the crystal structure if you take a Crystal zinc sulfide look at how the atoms are arranged in fact it forms this much denser structure with each zinc surrounded by four sulfurs and sulfurs are only about four zincs but if you do a calculation on that at the Nana scale it's less stable than these lovely open bubble-like structures so what we learned from that is that matter at the Nana scale can be very very different structurally and regarding its properties from MATA in a big bulk Crystal so let's take this even further you started to grow these nanoparticles and he found it got even more fascinating you've got what he called double bubbles so you've got a little kind of spherical Arrangement inside a bigger one and here's probably the biggest one that he grew and again it's got this kind of onion like structure and subsequently there's been experimental work that is strongly supported uh these models so uh I think both Sid and I feel quite proud of that work let me just give you one other illustration I've talked about the zeolites which are what we call microporous materials they've got pores of the sides the size of reasonable size molecules the molecules we have in in petrol now in the early 1990s another class of silica systems were discovered and these were called These are mesoporous catalysts these have got much bigger pores which will allow really big molecules like pharmaceutical molecules to enter into them and work here at the Royal Institution both help to understand the structure of these systems can you see they got this nice honeycomb structure that was known but the way in which the silicon and oxygen atoms are arranged in the walls that was predicted here by Rob Bell who's now at UCL and then you can implant a metal nanoparticle inside that mesoporous silica and the structure of that and its properties were determined by Stefan Bromley now a professor in University of Barcelona but anyway let's now move on to batteries topic dear to my heart and a very important one for the Royal Institution I'm going to talk about high energy density batteries and let me pay tribute initially to John good enough can you see that see him there John this year celebrates his uh 100th birthday a real towering scientist John absolutely together with others Revolution revolutionized the field of high energy density density batteries I'll give more details later on and was awarded the Nobel Prize I think it was three years ago for his work particularly on this lithium Cobalt oxide that we'll talk about in more detail in a few minutes just a few words on the history of batteries it's a fascinating history and of course it links very much to the ri the first phenomena the length of the development of batteries with the observation by galvani that when he had these two metals iron and copper close together and you put a muscle nearby I think it was a frog's muscle the Frog's muscle twitched now what galvani thought was that this was an electrical phenomenon and it was the muscle that was generating the electricity then later Volta uh said no that isn't what's happening it's the two two metals in contact that are generating the electricity and that's the electricity that's causing the muscles to Twitch and then Volta in a absolute breakthrough piece of science developed his pile his volte pile and I think we have a an original voltaic pile here at the Royal Institution so an absolutely breakthrough development in science and technology and so important for the ri because without volta's battery they'll take pile Faraday possibly the greatest experimental Genius of all time wouldn't have been able to do many of his experiments most of his experiments on electricity so it shows there how scientists are different dependent on the technology of the day and Davey wouldn't have been able to carry out his electronics experiments that identified several elements so both Davey and faradi depended to quite a large extent on this breakthrough development by Volta unbatters of course have developed over the years starting from the voltec pile the lead acid battery and the current high energy density batteries are based on lithium and the development of the lithium batteries as a real again another landmark development in science particularly in uh the science of electrochemistry now uh let's think about battery materials um why why are they so important now the reason why high energy density batteries are so important we need them increasingly for transport but of course they are what power personal electronics at the personal electronics Industry would not have been possible without the dis the development of the lithium carbot oxide battery by John good enough and if I have a lecture to teenagers about chemistry I take out my mobile and I say what has he got in it and we eventually get to the fact that it's got a battery in it and I say that battery depends on developments in solid state electrochemistry remember John good enough 40 years ago at the University of Oxford and without that you wouldn't have your mobile and since teenagers cannot conceive of life without mobiles it really brings it home to them but more about battery materials as I said battery containing lithium power laptops the living ions need to move very rapidly through the materials and computer modeling tells us how they move now let's just look before we go on at the lithium Cobalt oxide battery there's a real breakthrough piece of solid state Science and Technology um what it consists of it's really quite simple you can see on the right hand side there you've got lithium in in fact you think can be present in in graphite and when the buffer graphite when the battery discharges uh the lithium kind of loses one of its electrons the electrons go around the circuit the lithium ion that's lost its electron migrates through the electrolyte and then Nestles between the Cobalt oxide layers a really neat piece a solid state electrochemistry I said developed by good enough in Oxford I was a young research fellow there I wasn't involved in this work but we knew it was going on we all thought it was a really neat piece of work none of us realize including John goodness himself the impact that it was going to have but there is a big drive for developing other materials to particularly to replace that Cobalt oxide because there's some problems with both A supplier Cobalt of cobalt problems with toxicity and there can be some issues with safety but don't worry about your mobile so here are I'm going to start highlight right some work here of sightful Islam formerly of the University of bath now the University of Oxford and he's been working with others on a range of other materials that could do the same job as the Cobalt oxide and again beautiful structural chemistry cyclist done fantastic computer modeling on these systems he was also a Christmas lecturer here let's just look at one of them that's the lithium iron phosphate system which he has deployed is computer modeling techniques and this again experimentally was developed by good enough rather more recently than his original Cobalt oxide and it's a very effective material and one of the many contributions that siphon made was trying to understand how the lithium ions diffuse so can you see them this is modeling showed how they're diffused by this lovely curved path and it's really important to know how the lithium ions diffuse and the energetics involved in those diffusion processes but that's been sorted out some Years Ago by ciphon his colleagues for this important material now I'm getting towards the end and the first thing I want to do is to kind of present three challenges can we model chemical reactions what I've shown you so far have been essentially lots of structures but I really we haven't got to the heart of chemistry which is about how atoms react I'm going to then say can we actually see atoms Braggs produce this model by diffraction now the model we know is right from the diffraction we don't actually see diffraction experiment doesn't allow you to see atoms you deduce by the analysis and can we model and understand the dynamic nature nature of matter at the atomic level let's take the first one and here I'm going to highlight some work from my colleague Alexa Sokol again was a PhD student here in the Denver Faraday lab now a principal research scientist at UCL and we're just going to take a simple example and that is the activation of methane now methane natural gas very important molecule of course it's used worldwide as a fuel but a lot of incentive for trying to convert methane into other molecules rather than just burning it and for that you need to activate it and it was shown some years ago that you could do that by turkey an oxide material magnesium oxide introducing a bit of lithium into it which activates one of the oxygens on the surface so what we're going to see now is what happens when you bring a methane molecule down onto this activated oxygen and can you see the the one the methane molecules above it's got the tetrahedral not the planar shape and that gray feature is the activated oxygen and we're going to bring the methane molecule down and watch how the energy goes up initially it's coming down guy energy is going up going up it's now reached the top of the barrier and we're now seeing the heart of chemistry that is that a hydrogen atom from that methane is being transferred to the oxygen on the surface so we've understood in detail that kick simple but important chemical reaction and then the ox the hydrogen that's been transferred stays on the surface and a methane that's lost to hydrogen now moves away and that's how a very active species and will react to form other molecules so this is a simple example but yes we can use our modeling techniques to understand reactivity can we see atoms this is work of a student a UCL student a few years ago Scott Rogers and he was working on a catalyst which had gold on the surface of titanium dioxide you need to worry about that just think about the gold and we used Advanced or he used Advanced electron microscopy to image those gold atoms on the surface and again I think you will see just have a look look on the top right hand side you'll see single gold atoms those are those bright kind of yellow features so when then you'll see a gold cluster we have imaged this electron microscopy does imaged atoms we have seen atoms which this now is uh widely used technique for for probably well over I can't remember 20 years well over 20 years but it's got better and better and better and we can image single atoms fantastic achievement um I think Bragg would have liked to have seen that and I'm sure that John Dalton would as well and here I'm just going to show you the Dynamics of Mata at the atomic level don't really worry what's going on this is a catalytic process methanol to oliphines taking place in one of these zeolites what I want you to take from this is just how Dynamic the system is can you see the zeolite framework is constantly kind of breathing and you can see the molecules moving around inside the pores so these models can be a bit deceptive matter is really Dynamic at the atomic level this is beautiful work a good friend of mine veronique van spelberg from the University again let me show you another one these are methanol molecules inside the same porous Catalyst look at that blue thing that's an additional proton that one of them's picked up and you'll see that you get a struggle between the molecules one on car coming along trying to pinch it on the other they have a real tussle for a few minutes there eventually I think the Rogue one gets away with the proton come on is he gonna yeah he's made it he's made it but not for long not for long the other one comes and gets it back but that actually is a real example of how these additional protons can move around uh in the methanol molecules again beautiful work by the Fran spay book group now I've almost finished but I wanted to finish coming back to the Royal Institution and coming back to the foundation of the Royal Institution um if you look at the original charter it doesn't have one I put there but it does have the word common purposes of mankind now I've said science for the common purposes of mankind the word science wasn't used then I think the original is useful mechanical inventions but the idea is the same throughout my career I although I'm consider myself a basic scientist I'll add parenthetically that Faraday and Davey wouldn't have understood the difference between basic and applied scientists but although I consider myself a basic scientist I've always hoped that my work is relevant to the common purposes of mankind and in the remaining minutes I'm just going to tell you very briefly about two areas that we're currently working on the first is the CO2 challenge now CO2 you know about the challenge CO2 places the climate it's less well known that CO2 is a carbon source and we're going to need it we're going to need CO2 to produce sustainable hydrocarbon fuels it will become a carbon Source in the future because we won't be able to use fossil fuels for as a carbon source and we're going to need incidentally carbon capture and storage and emerging science at applying catalytic methods to take CO2 to react it with green hydrogen to make useful fuels and molecules so let me just show you very briefly the work of a student working with me and Dr logsdale at the University of Cardiff eagle kawalec and he is working us on how to develop catalysts that will convert CO2 into useful molecules in particular into methanol and by reacting with hydrogen now you can do that with an existing Catalyst copper zinc oxide but it doesn't work very well for this catalysis Palladium metal Palladium doesn't work terribly well on its own but it we think that this Ally of Palladium with zinc will work a lot better and so he's been exploring that computationally now I can't take you through the details of the calculations and the chemistry but let me just show you what he's doing he's taking this CO2 molecule he's looking at how it can pick up hydrogen atoms he's looking how it might diso associate on the surface he's identifying these key intermediates you'll see on the bottom right hand side and we're really getting to grips with this chemistry and that will help us improve these catalysts help us develop improve methods for converting CO2 into useful molecules and my final example is from a UCL student uh Jamal Nasir Jamal is working on Catalyst for removal of air pollution particularly nitrogen oxides and he his work is examining these poor again zeolytic catalysts which when you introduce copper into them really do work they will convert when they react with ammonia they will convert these polluting nitrogen gases into N2 nitrogen molecules which of course is present in the atmosphere and again I can't go into details but Jamal has done a fantastic job don't worry about this slide what I want to show is just a fantastic job sorting out all this chemistry and that will allow us to improve the Catalyst the gong as doggone so I will just complete by thanking the people here on this slide who've helped me put this talk together let me give special thanks to the ri team for all their help with the demos with the talk and with these beautiful models and thank you for listening foreign [Music] [Applause]

2023-03-13 22:24

Other news