This is. Really true. Thank you very much. So today what I'll talk about as you can read from the title I'm going to stand back here so you can hear me a little bit better is what we're now calling materials alchemy and I'll tell you what I mean by that in more detail but it involves chemical transformation of both biogenic and synthetic structures and you can see some of the of the of the morphology is up there on the left is a diatom algae type structure in the middle is a butterfly scale that's now a ceramic and on the right is a rapid prototyping then a converted rocket nozzle material and so it might be before I get into the details of the successes we had I should of course acknowledge that this would have happened without a number of really creative and hard working graduate students in my group and collaborators groups in fact that's something I want to make clear we really collaborate quite a bit we're good. We're good at certain things but we're not complete in being able to do all the things that we need to do to come to fruition on some of these projects and so we've been very fortunate to interact with a number of colleagues here at Georgia Tech in different schools as well as outside of Georgia Tech and in particular our interaction with the Air Force stock and rejection is been a very fruitful one is well. OK So what do I mean by alchemy or materials alchemy in particular right so if you look up in an unabridged dictionary what the definition of alchemy means or the first definition you'll find is it's a medieval chemical science and specular philosophy whose aims were pretty dramatic transmuting base metals like lead into gold. You probably. I'm there with Alan but a couple I wasn't familiar with discovering a universal cure of disease and prolonging life indefinitely while we're not doing that. A second definition means a greater magical power of transmutation and I'm not sure I'd say greater magical but certainly we're transmuting if you wonder what that word means means to check one of the definitions is to change into another substance or element that's what really in fact specifically we've come up with strategies for changing chemistry but not shape. That's the idea because oftentimes you'll find either from a biological template you might have the structure you want but not the right chemistry for properties. Or you might be able to wrap a prototype with one chemistry but it's not really the one that you want for a particular application and so I'll show you how we've developed these shape preserving chemical conversion methods one approach is by using gases or liquids to transform one material into another without losing shape show you some examples of biogenic as well as synthetic templates and the other approach of thermodynamics is not on our side to allow reactions to work is to use conformal codings and we've developed several strategies for putting very conformal coatings on templates and then in some cases will remove the template and so it looks pretty much like the same shape we started with. OK And so one of our favorite if not our favorite organism in nature is the diatom if you're not familiar with diatoms or type of algae and there's something like tens of thousands if not one hundred thousand species thought to exist on Earth. What you see here is a collage of S.C.M. images showing the silica structures that these organisms make each of these species each of the tends to hundred thousand species is genetically predisposed to make one silica shape and you can see from each image there is a different species the shapes are really all over the place you can find almost any shape you want if you look at the different species of diatoms I have a pretty thick book in my in my office actually on different species in the shapes they make. Interestingly also because it's genetically predisposed to make a particular shape if you let these organisms reproduce. Every time they reproduce they double they make two. Four eight sixteen and so if you have say eighty reproduction cycles if you pull your calculator out to do the eighty it is more than a trillion trillion of the same shape. So when you think about the number of shapes available to choose from the precision with which they are made the same way genetically and the massive parallelism that combination is really tough to beat. However it's all silica people have tried to induce diatoms to make something other than silicon have essentially failed diatoms are very selective about what they make their their Microcell is out of and so we've come up with another strategy and that is to take what diatoms do really well which is to massively reproduce very complicated three D. shapes out of silica. And as I show in a cartoon here and show you in real life how we can convert the silica into many other things without losing the shape. So we can marry what's great about biology with what's great about synthetic chemistry to make many more compositions and are available in nature. OK so we're going to see a lot of this particular diatom Microcell we're frostily as it's called all diatoms make their silica in two halves in this case all of us here are diatoms each half looks like this it's a cylinder has a hole on one end of these fingers on the back side you can see the two mating and and the fingers and Turkle AIDS. So there's actually two cavities here inside in these pores that are a few hundred meters that run along the cell wall. OK And so some it's been a decade now since we've been starting to work on diatoms that we showed back in two thousand and two. Here's half of an olive Microcell Here's a complete one we first reacted the silica with magnesium gas turns out thermodynamically magnesium oxide will form from the silicon and leave behind silicon and so at the time we were doing this at a fairly high temperature we don't need that. Now we now realize but what I've done here is to mark ten small features on these diatoms to prove to you. The shape doesn't change so Here's the before picture with silica. Here's the after picture it rotate a little bit but now this is all magnesium oxide and every one of the fine features is exactly the same right because of this displacement reaction. And I. Actually what happened was we use more magnesium than we needed. So the extra magnesium reacted with silicon at this temperature to make a liquid. This is kind of a lucky break. And this magnesium silicon liquid poured out and so we wound up with just all magnesium oxide and a very nice reproduction of the same starting template. So that was then when we make magnesium oxide and we've since been able to make many more things you can make silicon replicas of the diatoms we can make silicon carbide and carbon rubble does not talk quite a bit about those today we can make gold replicas and many other metals as well. Titanium barium Titan eight. More complicated compositions and many more than I'm showing you here by the tricks that we've developed in doing this. And by the way this works for any other silica as well we're not limited to diatom silica including silica made on silicon that's micro fabricated for example. And we're working with the idea to group on doing that. OK so how do we use our gas solid conversion reactions and so here's that reaction I showed you a second ago where the magnesium can react with the silica to make two moles of M G o one wall of silicon if you look at the volume of the products here volumes it's about sixty five percent energy zero thirty five percent silicon that's enough of both phases that if they're uniformly distributed throughout the Microcell wall which they which we found they are they're both above the percolation limit. So if you dissolve either want to weigh selectively you'll still have the same shape it'll just be more porous than what you started with and in fact if you dissolve the magnesium away the Moldovan the silicon is less than half that of silica so it should be really or a silicon porous silicon right I mean pretty useful material in some applications I will talk about also OK So here are some X. ray diffraction patterns the diatoms we started with for flame polish and so there's some crystal and silica that she can track. If we react instead of at nine hundred but it's six fifty degrees so you don't lose the silicon it's still there in these black squares along with magnesium oxide then if you dissolve the magnesium oxide away and simply in hydrochloric acid you wind up with. All silicon material and it's pretty fine grained as you can tell the peaks are pretty broad here from X. ray diffraction It's about eight nanometer crystal size. OK So here's what you get looks pretty much the same. In fact the prostate that's been induced in this silicon replica of the silica diatom is small enough that you can't see it by by this kind of this kind of magnification S.C.M. image. What's more the surface area of the starting diatoms is about two meters squared program. More than two orders of magnitude. It's gone up over five hundred meters squared program. So it's like Swiss cheese basically diatoms silicone now instead of silica. So I won't talk any more about silicon because this is actually something we did about five years ago. I'll talk about what we've done since but I'll just say that of course of course Silicon has a number of useful applications not the least of which is as an adult material for lithium ion batteries it has a very high capacity for storing lithium that we have a patent on that and other people as well as us have been looking at this to make silicon for anodes can also be used as a gas sensor for electric properties as a photo luminescent particle for inverse opals a number of things that porous silicon is good for but instead what I'll talk about today is what we've done since which is to convert it into other things. So here again is the X. ray diffraction pattern of silicon made from silica. If you react that with methane you can crack the methane and convert it into silicon carbide and this can be done a fairly high temperatures. Then you can selectively remove the current the silicon with chlorine gas to make silicon Tetra chloride and leave behind a carbon replica. Looks the same so despite all these steps we can make a replica out of carbon as you can see from the CD X. pattern from the starting silica and if you're impressed by over five hundred meters where program surface area. Turns out there are several volume changes that are occurring as we do this. This is over thirteen hundred meters squared program surface area higher than carbon blacks in fact. And so what we've done is to use it as a very porous conductive electrode material and so you can too. You can incorporate the. Latinum a catalyst particle by using platinum CARBONELL chloride vapor the vapor will infiltrate into the very tiny pores and there's a lot of micro prosti meaning pores that are less than two nanometers in this stuff. And so the platinum that plates out can only get as big as the porous Here's an example of a two nanometer platinum particle and it's just if we put enough platinum in it's just filled with this stuff very active platinum. In fact it works pretty well then as a cathode material in fuels in pen fuel cells. Here's a schematic of a pen fuel cell so at the anode of a fuel cell hydrogen is converted into protons and electrons over here. The electrons can then go through an external circuit and do some useful work at the other electrode the hydrogen passes through the electrolyte and then it reacts with oxygen and electrons to form water. This is the problem. This reaction the oxygen reduction reaction is pretty sluggish and that's what it hits a lot of fission see in as fuel cells and so it turns out we wanted to look to see whether our platinum loaded carbon replicas of diatoms would be good as cathode materials for fuel cells. So instead of setting up a fuel cell we actually work with mail in lieu M.S.E. using an A proton bath so if you're a gas it in a rotating disco electrode and you measure the current that are platinum loaded carbon frostily could carry and you can see it reaches a steady state that's dramatically higher say three or four times higher then these other controls and that includes a bulk and carbon which is the standard carbon black and carbon from silicon carbide the blue curve here I put that in there because this has also a very high surface area fairly similar to the diatom stuff but still the diatoms outperform that material and so in fact if you look at the specific surface area. The specific micro pore volume core is less than two nanometers specific maize of core value to fifty nanometer pores. It's no surprise that the carbon Frausto replicas outperform the carbon black Vulcan carbon because all of those numbers are higher. Right. So that's not surprising. What's a little surprise. Thing is when you look at the carbon derived from silicon carbide where again the silicon was removed with chlorine the numbers are very different and yet the diatom structure still outperforms the carbon from silicon carbide Well the reason has to do with the shape. If you look at the replica of the carbon frost we mill through it. You know the platinum is inside here but it's generally hollow. And so the fluid that contains the oxygen the auction can just come in and it only has to diffuse half the distance which is about twenty seven microns because it's you know a little over a micron thick compared to the other particles which aren't hollow. Right. I mean the carbon from silicon carbide Well it has a high surface area that actually has to diffuse halfway through a greater distance even though it has the same size overall as a diatom and the Vulcan carbon while it looks pretty porous actually has a fairly low surface area and so is the shape of the diatom that we're winning here in terms of getting the higher performance cathode materials and we load it with platinum. We also look at this as a template as an electrode support for enzymes and so glucose oxidase is sort of a model enzyme that people often will will use as exam as an example. It's an odd so oxide a reduction is enzyme it turns out glucose oxidase itself is neutral ph is slightly negatively charged. It turns out also that you can easily oxidized carbon partially to make it negatively charged we want to electrostatically bind the enzyme to the surface. Well negative charge won't bind very well to a negative charge and so what we did was and working with Niels Kroger and colleagues is to cross link the glucose oxidase with proto mean pro to mean is a positively charged protein that has a very high. Iso electric light meaning it neutral ph is positively charged. So if you make a hybrid of this molecule where the glucose osseous is still active. But now is positively charged. You can attach it to negatively charged carbon surfaces and so the way you can make carbon negatively charges to carb oxalate it. If you expose it as has been shown in the literature. Sure to nitric acid depending on the time of exposure it becomes more and more negatively charged. I don't have time to get into it but we've actually amplify the negative charges too with a surface treatment to get pretty negatively charged carbon. OK So then we expose that to the glucose oxidase proto mean to bind it to the end the enzyme to the carbon we put the pattern compact into a capsule sealed it up and then just pump glucose through it and measure how fast the glucose was consumed to show how good this template is the support is as an enzyme support. So I'm plotting here as glucose concentration as a function of time for the three different sports is just diatoms that we're functionalized with the enzyme that's the carbon black and here's our carbon frost all replicas and you can see they outperform the other templates it takes about half less than half the time to eat up all the ends of all the enzyme glucose all the glucose in solution. Compared to the Vulcan carbon the standard carbon black takes about an hour compared to twenty minutes. So again these are very open very porous very accessible surfaces for enzymes to do their work and can be used and bio fuel cells or another catalytic applications. OK So that was a quick tour of some of the gas all the reactions that we conduct another exam on biogenic templates Another example is with synthetic templates with liquid solid reactions. And so here we've been working with a technique we padded called the displace of compensation a process process the idea here is to use volume changes to fill pores in. And the application we're looking at quite different than what I just talked about a second ago is solid fuel rocket nozzles if you think you've heard of extreme conditions I don't know that you can find anything more extreme than this because the temperatures in our solid fuel rocket nozzle climbed over twenty five hundred Center grade within a second from ambient what happens in a solid fuel come fuel is you have aluminum particles mixed in with a lot of organic stuff ignites and the aluminum becomes aluminum and it gets so hot the aluminum oxide melts and it pounds through that. Supersonic speed so it's extremely erosive extremely thermal shock prone very high temperature. What do you use for that for the nozzle right I mean conditions are extreme has to be high melting over twenty five hundred Center great thermal shock resistant as a much as anything can be terrible shock resistant. Erosion resistant lightweight if you want to launch something heavy non-porous and expensive except for a I can count on my hand the number of materials at a work right that people have used tungsten tungsten is very heavy not very erosion resistant at these temperatures carbon again not very erosion resistant radium rhenium very expensive very heavy. So what do you use that's going to be better than what exists now none of those materials are very erosion resistant. In fact if you if the Navy ships tend to you know that aircraft carriers etc tend to have solid fuel rockets on them because they fly fast they have very fast high thrust because they're not as explosive as liquid fuel propellant. And so you want to have something that's going to be erosion resistant so if the missile misses what's coming in and it can turn around and chase whatever it was coming at you like another missile but the orifice opens up so quickly with as erosion that it falls in the water and can't do that. So the Navy in particular is really interested in erosion resistant rock and I was a liner rocket nozzles. So we think we have one that will outperform existing stuff called it is or Conan carbide tungsten the pretty interesting composite system both tungsten and zirconium carbide melt pretty high and in fact when you put them together if you look at a phase diagram they're solid as temperature together where the first look at forms is also pretty high. They're chemically compatible tungsten and zirconium carbide form a tile and together they don't form any new compounds most remarkably And this almost never happens. I can tell you in metal ceramic composites the ceramic the coefficients of thermal expansion are almost identical between tungsten they're going in carbon refractory metals have pretty low C.T.E.. They're both very conductive And so from a thermal shock standpoint that's pretty good carbide is very hard not as hard as diamond but not far off and so it's pretty erosion resistant tungsten becomes ductile above about. A hundred C. so it's the tough phase. So we have a hard road resistant phase and a tough phase. Even if all those things weren't true if we replace the tungsten was returning carbide you win because it weighs a third the weight of tungsten. And so we think this composite system is really attractive for solid fuel rocket nozzles so how do we make it. Well the way we do it is by again a displacement reaction we can take a tungsten carbide pre-formed say shown here a particle preform that you can make by slip casting gel casting or even machining I'll show you in a second here you can infiltrate that with say as or Kone and copper liquid. Those are Tony will go after the carbon to make zirconium carbon tungsten it turns out if you look at the volume of these two products. It's about twice the volume of the starting tungsten carbide So what that means is if you start with a fifty percent porous preform you can fill the pores in with the products you make and it doesn't change shape or size actually really a neat way to make these composites we think and in fact it works as I show you here. The tungsten reacts to tungsten carbide are actually from a tungsten layer and around that is are going in carbide layer and so the shape of the tungsten carbide actually is reflected in the end in the shape of the tungsten and the composite. OK So here's how we make them one way to do it is to take a tungsten carbide billet take a tongue take a milling tool and just machine the shape of a nozzle it's pretty fragile stuff. Then what we do is we take it. We put it in a basket so here's the sample with the basket has holes in it. We heat up the sample and the melt. This is are going in copper liquid we dug it in the melt the liquid wants to suck in like water into a sponge you pull it out of the melt and we heat it up a little bit more and what happens is are Kone and copper liquid the copper is just there to lower the melting point. Once the pores get filled in filled in the liquid copper gets extruded back out. So you can tell the pores are getting filled in because the sample turns a copper color the copper comes back out again. And so here's the dunk tank that we that we do this. OK so here's an example of the machine nozzle I was showing you a second ago by a rapid prototyping C.N.C. machine. This is what it looks like after we dug it. And again you can see the copper that. Come out on the surface. You can remove that by just polishing or nitric acid treatment. And here's the product we make not only does the shape and size stay the same as I show you a second here. But even the milling tool. Are preserved so if we want to rifle the inside of this we can do that. So here's the dimensions of the preform we started with the exit is the bottom part of the bigger diameter and trances the top diameter and the nozzle height and these are in millimeters and this is the same sample dimension to looking at the same sample afterwards and you can see the dimension changes are less than about point three millimeters less than one percent. So by naked eye. You can't tell there's any difference. It looks pretty much that looks exactly the same afterwards in terms of size of dimensions as it did at the start. This is remarkable because there is no other way you can make as a kind of car by trying to think of posit and preserve the shape like that. We think is remarkable and we made an advertisement for Georgia Tech and other other shape to make cones we made tubes we can make all kinds of shapes so however you can wrap a prototype of shaves. If you cross-section this it's pretty dense you can occasionally find a poor like that like that one but that could have been from pull out from polishing if you zoom in you can see the Grays are kind of carbide matrix and the white tungsten particles. It's again pretty dense. If you zoom in even more you can see the tungsten tungsten particles when polishing the diamond you can see the diamond scratch because the tungsten is soft but the matrix is hard. In fact if you look at the tungsten map you can see the tungsten is a county map shows their county in Carbide in the Matrix. So we made things like rocket nozzle liners This way you don't even actually need to make the whole rocket nozzle out of it just the part that sees the combustion products right so we can make it even lighter weight thereby making the nozzle and that again the dimension changes are really tiny and we've done a rocket test with our colleagues at Edwards Air Force Base west often and when a rocket test you don't actually launch a rocket because you have to find it again. Right. You actually launch it upside down. You put their combustion. But though the fuel down here in the nozzle up here. It's in a frame when you ignite it and it burns through the. And then you take them all as a lot and look at it and so we actually had to back them liner with carbon to fill the space and here so the gas wouldn't go around the nozzle but instead through it. OK So if this works. The next movie next video is a movie the next slide is a movie which will show the countdown of the rocket test and if you listen closely enough you can hear him go three to one. So the nozzle is up here and it's going through this fairly short very erosive propellant it's like ninety percent aluminum. We pulled it out and that's what it looks like here's the aluminum oxide liquid that can get that condensed as a solid oxide at the end and we ship this off and measure the erosion and there wasn't any basically So they're pretty excited about this material is an erosion resistant liner. Well we're not just in a making stuff we also want to understand how this works. And so we look at the mechanism of how this reaction happens so here's the tungsten carbide we put it in this economy copper liquid as I mentioned you get a tungsten layer adjacent to the tungsten carbide and those are kind of carbide layer around that. So how is this happening what is the rate limiting step right it could either be diffusion through the liquid. In principle. It could be a chemical reaction in any one of these interfaces or it could be a solid state diffusion of say stuff through their own carbide or through the tungsten and for us to be able to understand how long it takes to react we need to know what the rate limiting step is so what we did was to take a slab of Dunc's dense tungsten carbide put it in a zirconium copper bath and then look to see what happened at the bottom at the top. So as you're sucking zirconium out of the liquid they're going in copper liquid it becomes lighter and so the liquid should rise right and so you'll get a convective loop forming where the reaction products here should be different thickness than the reaction products there. If it's liquid phase diffusion controlled. Turns out they're not so we know it's not liquid phase diffusion controlled. If you plot the thickness squared versus time for the tungsten or those are going in carbide layer different temperatures you get pretty good fits which is a pretty good indicator that it's solid state diffusion control. Planar reactant In fact if we plot the parabolic rate constant for thickening of these layers tungsten those are kind of carbide you get numbers about two fifty to sixty killed us promote all which is pretty close to the activation energy is for carbon diffusion through zirconium carbide grain boundaries or through tungsten they're very close. So it's hard to tell one from the other but it's clear that it's carbon diffusion That's rate limiting here. So once we realise that since this is a nano attack talk. We actually are now making nano composites because what we what we realise also in looking at the phase diagram is if you look at zirconium carbide here and tungsten carbide there they form a pretty long range of solid solutions. So instead of starting with just tungsten carbide we now start with actually should say is the tungsten carbide solid solutions the thought being that if it's carbon diffusion That's rate limiting if there's a common tungsten or atomically mixed in the carbide and you're sucking away the carbon from the tungsten carbide. You'll trap tungsten nanoparticles in there Clinton carbide matrix and you can make refractory nano composites. You never hear those two words to get those words together refractory and nano composites right we can do that now. And so what we have shown here X. ray diffraction patterns. Here is actually a solid solution that we made point five eight zero Tony in point four two tungsten carbide salt solution we reacted it and you wind up making zirconium carbon tungsten and so we made what we wanted in the end but in terms of what it looks like here is a T.M. image and a back scatter to S.C.M. image and you can see the bright particles here a tungsten they're trapped in a grey matrix of zirconium carbide it's even more remarkable if you look at these three zones or here's a mixture of tungsten and returning carbon the T.M.H. tungsten is dark in the T.M. image to read pretty dense The Zone here is just carbide and here is the tungsten particle itself if you do electron diffraction on each of these regions. Say look at this one in these two particular regions one and two what you'll notice is the whole thing is a zirconium car but the carbide everywhere is a single crystal. So not only do we. Bed tungsten within zirconium carbide tungsten nano particles they're embedded within a single crystals or cone and carbide So it's a very hard matrix with a tough set of particles on the inside. I don't know any other way you can do that. How do you put tungsten particles in a single crystal as are going carbide. And so we look and we can do this with different compositions as well and you get smaller particles of tungsten with with lower tungsten loading so we're really just starting to look at how this could be utilized to make refractory nano composites. Living ahead there and my tour of our techniques and that's really what my mission is today is to show you our techniques in case there are collaborative opportunities here which gears and talk about coding methods that we've developed as well. And so we look to instead of converting diatom fossils by reaction instead coding them with certain materials like metals for example. So here's a cartoon of a diatom for us. Still there are some hydroxyl groups on the surface. So if you use a solid ization treatment you can put primary means on the surface. Turns out those are useful for Electrolux deposition because you can put a catalyst like Palladium chloride on the A means. And so if you do that you silent eyes then you introduce Palladium chloride then you should be able to dunk these things and electrolytes copper Baz and put copper on the diatom you can do it but it's not a great way to do it because as you'll see here the copper you put on it isn't very well interconnected there's gaps in between and even if you go longer you just get a thicker but still porous coating this way. So when you dissolve the silica through the pores away. It doesn't hold the shape very well so you can't make a replica. So instead what we've done is to go back to the drawing board on functional ization with the help of Seth martyr's group here and we've amplify the I mean density what it told us was the catalyst was a very concentrated right you were getting copper everywhere when it was depositing So what we did was to expose the a means to a multi act relate any one of these accolade groups combine to the I mean. Now you've got four or five accolades on every I mean post you can expose anyone actually to a poly I mean any one of these it means can react with the accolade and go back go out by going back and forth between Ackerley to me. I mean you can really flood the surface with the means. Then you can put a lot of chloride plating chloride Catalyst on the surface and get a better coating. And so that's what we do we silent eyes. We Apple fire the means we put the plate employed on and then deposit the copper and actually this image the silica is gone. So this is an all metal replica copper replica the dye you go full Actually it's exactly copper in the shape of the diet. And you can tell there is no silica left if you do X. analysis as well as you don't see any peaks for silicon so we can make really nice copper replicas this way. So one of the more recent things we've done is to look at this diatom Cosco discus Asta Rumple as it's a cylinder shape Marine diatom. If you look at the poor as on the valve they have this pseudo hexagonal symmetries are here six here. If you move it over here six. It's not perfect it's really quite as I periodic because the spacing isn't exactly crystal in here it is the but it's hexagonal morphology so one of my colleagues Joe Perry looked at this and said My What if we converted this into a plasma transmitting metal like gold. You want to get a phenomena called extraordinary optical transmission. And so we hired back shot of one of the poorest and so we did the treatment I mentioned I mean a silent ised amplify the a means in this case we used or a Clorox acid sensitizer put the gold on as a plasma transmitter and then so what that is all the way. So here is the replica and gold of the cost going to discus Auster awful is a very nice replica no silicon left anywhere and in fact if you cross-section it with a fish. You can see the space where the silica had been it was does. There are there have to be some pores occasionally to allow this to dissolve the silica way but not many. And so this is a porous. It has a shape of the diatom and it has these cavities that run through the diatom valve former died. So if you're not familiar with extraordinary optical transmission it's a way of transmitting light through a porous material where the pores are smaller than the wavelength of the light. So the idea is if you hit this with in this case infrared light you launch surface plasma from the poorest kind of like dropping a stone in water the pole. Plasma in Calera tons transmit across here and if the pores are properly space they constructively interfere. Then transmit down the length of the channel. Interfere on the bottom part again and regenerate the light that you that you can put on the top that was probably a terrible physics explanation for what's going on here but that's my simple minded understanding of how this works. Bottom line is that the wavelength of light. That's transmitted is definitely affected by the spacing of the pores and some of since the pores here are on the order of several microns you're going to transmit infrared light through a gold replica of this stuff. So here's our gold replica diatom actually an optical image of it. If you plot the transmission versus wavelength four point two micron light you get about fourteen percent of the light through which is rather impressive. I mean because again the pores are smaller than the wavelength of what we're putting through and so what we're now thinking about is using these a sensor is these are flat diatoms they'll lie flat like cards. You know on a piece on a table or something. If you functionalize the surface of the of the gold with agents that are going to capture and analyze the surface plasma player Aton transmission is very sensitive the surface changes. So if you get a binding of that it will shift the extraordinary optical transmission peak and so you can scatter these things in a fluid say in water somewhere. We've got ideas of how to harvest them you can harvest them put them on a slide glass slide it will lie flat then just shine a light through it and tell if something is a particularly that and a lot. You're interested in absorb on the surface by having a selective functional ization on the surface. That's the idea. OK yeah yet another template we focused on just to show you our rays of range of options here are organic template so this is a more full butterfly this is in collaboration with Monster any of us. Errol. If you zoom in on this butterfly you'll see the scales are very reflective in the blue. If you zoom in further with the S.C.M. you can see they're very periodic they have ridges and actually they have ribs that are periodic Lee spaced very elegant intricate structure that's reflecting the lie there's no pigment or die in here it's just in for it's just light reflection color light reflection from our stamp. We wanted to we wanted to make replicas of this ceramic replicas so we could shift the wavelength that was being reflected and make a more robust material. Turns out these are really nice materials because they're made of chitin which has a lot of hydroxyl groups on it. That's a that's a functional group that we can actually deal with in fact the way we deal with it is to use a technique called surface all Jill. You've probably heard of this all gel route to make ceramics this is a way to do it. Layer by layer on a surface. So if you have a hydraulic fluid surface if you expose it to a metal oxides like titanium I suppose. One or two of the cox side will react with one or two of the hydraulics and for metal action bonds alcohols will come off and that will continue until you react away all the hydroxyl ZX then the unreactive Alcott sides can be exposed to water to form want hydroxyl groups again. Again alcohol is come off. So by going back and forth between side water Alcott side water you can build up a layer of desired thickness and fire to make the ceramic so because my students initially got tired of dipping butterflies skills over and over again we've automated this process of actually artists on a second generation of automation right now but the idea is to put a butterfly scale over a porous for it. Turn a computer on it a pump in the side for some time it will wash away the extra cost side then pump in the water and then dry and because it's automated we can push a button and put down hundreds of layers and come back in the next morning. So here's again another butterfly this is a more full Heller butterfly It has pointed tips and again it has this very intricate structure. Here's a cross-section of it so it has ridden. Ridges it has ribs it has these supporting structures. After we did this surface all gel process with fifty one cycles of surface all gel and then fired it we made these ceramic replicas there's no cotton left. This is all ceramic now Titanium the scale shape is preserved and even the very fine features are very well preserved and the three dimensional aspect as well preserved. But of course because it's a coating. And we burned away the template it isn't really a positive replica it's a negative replica and in fact if you cross-section that you can see the regions where the chitin had been but it is so thin and formal It looks like a. Replica from a distance. So we got a little bit more sophisticated with a shot. This is proof that it's titanium but we've also gone and made barium Titanite you can react to titanium hydrothermal E. and make more complicated ceramics as you see here that turns out to be useful for the next application I'll show you one of my students. Jonathan Vernon got interested in this particular butterfly. It has a very recognizable shape even in an optical microscope. If you zoom in you can see these ridges also that are observable on an optical microscope not just in sci. And so Jonathan used to work for a paper company and he thought what if we converted this into something photoluminescence where I can see the shape of an optical microscope. But I can know that it's that is a special composition because of its photoluminescence its surrounding the composition that I can interrogate shape and photoluminescence and have a tag like for anti-counterfeiting for example. So what he did was converted this into a very tight knit replica as you see here and a lantern I don't replica very tight it is known as a trash can. Ceramic you can do it with just about anything. Basically will take a lot of different elements including lantern ides So if you look at this in an optical image transmission you look at there's ceramic Interestingly enough if you look at this in U.V. XA Taishan it glows red as you'd expect from the rope and that's present and so what Jonathan did was he took white filter paper these look white by naked eye and he sucked them on to filter paper which is also white and so on bright light image white regular white image you can't tell there scales there maybe for close enough you can tell and dark image darkfield image you can see the scales the replicas that are there. And even more impressively in fluorescent imaging you can definitely see they're there. And so these can be used or other organic templates can convert of the same way. Can be used as tags for anti-counterfeiting we can control the shape which can be pretty specific. And we can control the color of foetal in essence by doping with different Latin either combinations of land the knives. And so in fact here's a stack image showing several of these scales not only to the scales are there. To see by naked eye but the ceramic replicas so thin walled they're flexible they actually conform to the surface so you can't even feel them either. It's pretty clever. All right. And very quickly if I have time. Let me just tell you about one other technique that we developed in coding so I talked about two reactions schemes gas all that was solid. A way to code biogenic templates and something that we're doing with synthetic templates where instead of using biology as the template. We're using biology to make the coding. OK So the idea here is to use a protein I talked about earlier called proto mean problem is an interesting protein it's pretty short thirty one. I mean no it's it's long very positively charged because of all the origin in amino acids that are present. It's cheap because it's harvested from space from fish sperm and so it's very readily available in the environment. In fact it instead of paying seven dollars per milligram for even a shorter peptide it's only point four cents per milligram one thousand times cheaper than commercially synthesized peptides. Well what we're interested in it because protein is pretty sticky it will stick to lots of surfaces like silica to put the protein on silica you can expose it to water soluble titanium precursor called Tai bald the proto mean will do it will dissociate the Liggins on this and make titanium but become buried in the titanium that's kind of critical for an application I'll tell you here. What you form the titanium on the surface you can add more proto me to that and then go back and forth between protein type all protein type hold and build up a coating of desired thickness. So it's a layer by layer water based process using water soluble precursors So our template in this. This is porous an article Lumina if you're not familiar with this. It's almost magical to me you take a luminance oil and you are not equally at it in the right way and you get channels that are nano Manno sized and that run through the through the former film now as alumina So this is a top down view Here's a side view showing the channels. What we've done is to take this porous an article Lumina coated with proto mean Titanium the way I just described it then because the pro to me is in caps late in the titanium when you burn away the proto mean the titanium. Becomes porous. And the porous Titanium you can dissolve through the porous Titanium the alumina and make nanotube arrays. So titanium is a pretty interesting electrode material say for solar cells. This is a way we can make a very porous but align titanium electrodes. So here's a coated substrate made this way after eight proto mean ball deposition cycles. You can you can't really tell it's coated because the coating a so conformable But here's a cross-section. And you can see the proto mean titanium coating that's present. So if you take X. P.S. on this. You can see peaks for titanium an option that's in the coating nitrogen from the proto mean. But there's no peaks for alumina So the coating completely covers the surface of the alumina. But when you fire it after that six fifty c Now the piece for aluminum alumina start to show up because the coating is porous you can see through it. Now with X. Yes. So here in fact is one of the nano tubes we made this way it's nano Crystal in titanium. After eight cycles it's only about thirty five nanometer stick so you don't get only get a few of maybe four or five nanometers pursue this way it's pretty conformal So we did was we went through the deposition we fired it to remove the proto mean we dissolve the aluminum away and we stuck it on electrode a glass substrate. And that's what we've got titanium nanotube arrays made this way. And so because the titanium wall of these tubes is porous. It's not dense titanium. You can load it up significantly with Di And so here's one of the dyes that's used in Dyson's Thai solar cells. Our colleagues at the Air Force have made titanium nano tubes that are dense by a different technique and they get this amount of dial OD'ing with our proto mean derived porous wall nanotube arrays we get more than twice the dial OD'ing and that's a very useful thing not just for solar cells but for some other applications as well. All right so I was kind of a quick and dirty tour of what we do and materials alchemy what I've shown you that we can take bio silica as well as synthetic silica and by a series of controlled gas solid reactions we can make silicon silicon carbide carbon carbon platinum enzyme carbon replicas that are good. Electrodes we've developed a look at metal in filtration process that allows us to take rapid prototype and make dense ultra high melting near a net shape near net sized parts we developed coding techniques for coating biogenic templates with metals with ceramics we've even developed a protein based technique all water based is that of alcohol based that allows us to make replicas of structure synthetic structures. So we've used these things for supports Rekha Tallis says for water purification for chemical synthesis for electrodes for solar cells batteries fuel cells. We've used them as photoluminescence agents for anti-counterfeiting and as high temperature rocket nozzles nozzle materials. So with that let me acknowledge our sponsors we've been fortunate to get significant funding from the Air Force including a biopic Center of Excellence for the bio related work a couple of mirrors that are also bio related bio optics that's led out of Harvard bio paints is led by Carson Meredith here at Georgia Tech as well as the U.S. Department of Energy and I'd be happy to take any questions or suggestions for collaborative research that you might have thanks. Yeah. The carbon he said. A court of law to me. Right. So the way the way we incorporated the plan especially into the two nanometer pours is with a gas. So we took platinum CARBONELL chlorine vapor it percolated through and and cracked the carbon cracked the platinum in the pores and so the plot and plan of particles were partially encapsulated in the pores and so as a result. You know they're partly buried in the pores and so they're not as. The fall out actually. Yeah because it's not platinum resting on a fly on a surface is actually embedded inside the carbon in the partially in the pores of the carbon. So it's mechanically heat it's locked in better than if it were just sitting flat on the surface of carbon. Just the performance as well. That's not that's why I was saying at the start we need collaborators. We do the chemical stuff we don't do the testing as much as we would like to be able to do so if there are if you have ideas on how to do the performance and durability I'm all yours to test those things. Yeah. Right. So the reaction we use the I mean actually is a Michael addition reaction that's pretty well known. I'm not sure how well that would how often I mean I don't know I mean this is really Seth martyr's thing. I mean he came up with actually one of his postdocs came up with the idea as a way to amplify a means and so. So they have done a lot of testing on this. They've done. Are. They've done. Not on diatoms but they've done a Q.C. on a flat silica surface but also they've done they've they've done something else or they've done water wedding tests because the a means or hydrophilic the accolades or not and you can see cycling between wedding and not a wedding with every layer you put down. So it's been fairly extensively documented that it is a layer by layer accolade I mean. Process but Seth deserves all the credit and his post I should say deserve all the credit for that idea. You know to raise. Well so first solar cells and I didn't show it here. We've actually put these in this and by sensitized solar cells. For other catalytic applications for titanium like water splitting type applications anti-microbial actually to I mean and a taste I did mention is anatase titanium anti-microbial with U.V. light we can load it up with enzymes also make it even more anti-microbial And so I think there's a number of potential applications where those and are not limited to tubes we were able to make tubes because we had a template porous a lot of aluminum but the same idea can be used to coat Stober silica spheres we've coated diatoms with this stuff. I think and in fact we're working with a D.B.A. group to code for lithographic we patterned silicon with silica on it with porous titanium. So I think there's a number of potential applications and and substrates that we could use to do this with. I didn't get into the details of it. I had to look at the precursor I saw it. So it turns out the way the proto method works it's essentially all X. or static protein is a very positive charge silica neutral ph is negatively charged so it sticks to silica pretty well type all the salt that we use is negatively charged gets recruited by the positively charged pro to mean it but the magic is when proto mean takes the Libyans off the lactate leggins and makes titanium I don't know how that happens exactly but it happens that titanium that you make is also negatively charged so the protein will stick to that. So we kind of got lucky and having a negative charge template possibly charge protein negatively charged salts negatively charged product and it all works so to do it WAS think oxide to have to find it negatively charges in the precursor that's going to be tough and that's that's the limitation of the proto method the number of salts that you can find with a metal is contained in a negatively charged molecule is most of those salts are going to give you positively charged cat ions right. Using core it'll give you zinc plus two which is going to help you with this method. So you have to identify it negatively charged think precursor then I think it probably would work especially if it's a lactic acid. Proto means seems to do a great job at knocking off lactic acid groups like take groups from molecules. So I have to search for one if you can find one and I think it may be worth a shot of the other. OK yeah yeah. So you have to have a heavily. To get a continuous oxide you have to have a highly hydroxyl ated surface. If it's not we when we first try to do this with diatoms which natively aren't that highly Hydrox lated you get spotty coatings so we had to amplify the hydroxyl density on the surface of diatoms using the Michael addition process actually terminated with hydroxyl groups on the last molecule you so you have to have to have a heavily hydroxide surface butterfly scales are great because a cut in itself has a lot of hydraulics and we don't have to do anything to that stuff and you can cut it really nicely. So I mean if you have a substrate. We can probably come up with a scheme for putting hydraulics was on it to coat. What your.