 Okay, our next speaker is Hema Karunadasa and I believe she's in her second year as an assistant professor in the chemistry department here at Stanford and her entire research group develops perovskite materials for various applications and Today she'll talk about her G-SET project on using them for solar cells Great. Thank you Mike for the introduction and thanks also to Sally and Richard for the invitation to speak I'm really thrilled to be here today to tell you a little bit about our very recent results in hybrid perovskite solar cells So this is a collaboration between my group in chemistry and Mike McGee's group in material science and engineering and We are a bunch of synthetic inorganic chemists and together with Mike's group in engineering We both synthesize new materials that could function as solar cell observers and we also test these materials in actual devices So to start I want to Okay, so to start I'd like to start with the structure of this material and try to explain to you Why there's so much excitement about hybrid perovskites in the photovoltaics community So perovskites are a huge family of compounds and these materials have been known for many many years In fact the first work on perovskites dates back to the 1800s And you can form perovskites with metal oxides metal sulfides metal halides And these materials have very versatile electronic structures as well So you can form perovskites that are high band gap insulators Semiconductors conductors even high-temperature superconductors these materials are everywhere and these properties of perovskites have been exploited for many decades and These materials have this the same general oral structure here I show you the structure of this lead halide perovskite material that has caused so much excitement in the photovoltaics community Where the inorganic framework just consists of a very simple framework of lead halide octahedra that share corners And this extends out in all three dimensions and this framework has a net negative charge So therefore to provide charge balance We need a very small positive ions that have to reside within the cavities of this structure So within each cavity that's defined by eight corner sharing octahedra sits a very small organic cation Well, it can be either an organic or an inorganic cation in the case where we put a very small organic cation like this Methylabonium that I've just shown as a gray sphere. These materials are called hybrid perovskites But there are many related materials which are which can be structurally derived from from this oral structure so today I will talk to you about both this 3d structure shown here as well as a Family member which is the 2d material and you can again think of the 2d material as If you took your 3d perovskite and took a very fine knife and started slicing layers of your of your 3d structure and again you get this An extended two-dimensional area of corner sharing led iodide octahedra But now you have this additional advantage for at least for structural tune ability Which is that between these sheets you can put much larger and more and potentially more elaborate organic cations and later in the talk I'll show you about I'll tell you why you might want to move Work with both the 3d as well as the 2d structures so one thing that Continues to amaze me about these materials is that they form on their own accord This is called solution state self-assembly which means you can just take a beaker put it on your bench And just throw into solution your precursor ions So you can throw in the highlights and the metals and the organic cations and if you just tune your conditions Just write these materials form on their own and the beauty of self-assembly is that whenever you get Molecules or ions to do what they want to do they do it perfectly well So when you get to the conditions just right you can isolate this hybrid perovskites as Crystalline powders thin films and in my group We love to grow single crystals because they are beautiful and also because you can do very precise measurements on Well-formed single phase single crystals. In fact every picture I show you is from a single crystal X-ray diffraction study so I want to discuss a Some of the strengths of this material that there are many I'll just highlight some of the points that I I think are worth mentioning a Very important point about Perovskite observers is that these materials can be formed through solution state methods Which makes everything extremely inexpensive the precursors are very cheap and you can use solution state methods like spin-coating drop-casting or Vapor conversion methods to make these materials So here I show you a schematic of one particular way to make nice films of hybrid perovskites Where we can simply take a tight a mesoporous titanium scaffold and drop a solution of lead iodide into this material The solution will coat this scaffold and form this nice layer of of lead iodide Let iodide forms a layered structure, which means that it likes to grow to form uniform films around any kind of scaffold and Lead iodide is yellow. So we get this yellow film we can then simply take this film and and Immersed it into a solution containing methyl ammonium iodide and the methyl ammonium iodide will work its way in through the lead iodide film and Convert that film into the perovskite So yellow film is turned black and we are left with this beautiful uniform film of the perovskite We simply heat this film drive out and any volatiles and we're left with a nice a nice film that we can then Continue to fabricate the soils the soil cell device So that makes everything very cheap and very scalable I Think this is the chart that I stole from Mike slides that that gets everyone really excited about these materials There's been a very very sharp increase in the efficiencies of these materials over a remarkably short period of time So the first Solar cell as Mike mentioned the paper came out in 2009 and this was a really neat result But it kind of stood alone and it was more of a scientific curiosity and then in just a very short time This efficiencies got to the point where everybody cared very deeply about these materials And there's certainly no sign of plateau plateauing so we can expect we can most likely expect further increases in the efficiencies of these materials 2009 was a very memorable year for me I was writing my PhD thesis when the first paper came out and I just read this paper and thought oh that's neat But then by the time I started my group in 2012 things got a lot more exciting and this is a great great subject to work on So I'll tell you some of the reasons why these materials have this great efficiencies It's very remarkable that these materials have very intrinsic properties in its Photo excited state so when we hit this materials with light and we create photo generated electrons and holes These charge carriers can move around very freely the diffusion lengths are very long and they don't recombine very easily Which means we get ourselves some time for those holes and electrons to get to their to go their own separate ways and Complete the external circuit in the solar cell device and this material seemed remarkably tolerant of defects In fact, you don't need beautiful single crystals of these materials to realize very high charge carrier mobilities Even polycrystalline powders. We have very well and even films which is very strange and very unusual for for solar cell adsorbers and This is the typical diagram We show when we talk about the energy levels of a solar cell adsorber in contact with charge carriers So the basic idea behind a solar cell is that we have a semiconductor This is your adsorber so in this case this would be the perovskite and we hit this material with light corresponding to this band gap energy and the idea is that we excite an electron from the filled valence band shown here to the empty conduction band shown here We now need to separate those electrons and holes and send them their separate ways to do that We need materials that can conduct the electrons and conduct the holes So the way that we can move this electron to electron transport material is to drop it in energy And the way we can move these holes is to is to raise it in energy Holes, that's just an absence of an electron. So raising the energy of the hole stabilizes it So by necessity this diagram shows nicely that the voltage you can get from a solar cell is Smaller than the band gap of the adsorber because just to be able to transport those carriers We had to drop the energy of the electrons or increase the energy of the holes But some very interesting Recent reports have shown that these perovskites are not just great adsorbers, but they can also transport charge very well So imagine getting rid of this electron conductor In this case your adsorber is both the adsorber and the electron conductor Which means you just bought yourself a lot more voltage a lot more obtainable voltage So this has been shown for both electrons and for holes. So perovskites are really versatile materials and A very nice attribute of these materials is again just through solution state self assembly We can start adding different ions into the lattice So we can start with the lead iodide lattice and then start doping it with bromide and we can change the adsorption energy of the material And you can see that there's this nice increase in band gap as we start increase start adding bromide into the lead iodide lattice and Being able to tune the band gap of the material is very interesting for lots of applications In with respect to solar cells Mike mentioned That his group works on tandem devices and they have indeed shown that tandems can work very well in a in a two adsorber tandem device the higher adsorber The energy the band gap of the higher Higher energy adsorber has to be about 1.9 EV So with lead iodide alone we get a band gap of about 1.5 EV So in order to get to 1.9 EV one potential strategy string to incorporate bromide into this structure I'll get back to whether we can do this or not But the fact that we can tune the band gap is very exciting for applications in a solar cell adsorbers light emitting diodes and lasers It's very cool to think that we can simply change the structure and change the light the color of the light that comes out of this material in a laser So these are I hope I've convinced you that this has these perovskites have very promising properties And I'm I love perovskites. I'm a huge advocate of these materials and I and I very much want to see this technology work But precisely for these reasons we've decided to focus on the Weaknesses of this material because we want to identify any problems that we might find in these materials So that we can we have a chance of addressing them and hopefully resolving them So for the rest of my talk I will focus on on the weaknesses the potential problems that we might encounter using these materials There's no denying the fact that there's lead in this material and a lot of it There's a lot of concern that this lead is not just toxic But these materials are water soluble and there are many people who say we've just removed the lead from our houses Are we really going to coat our roofs with lead back again? So this might this might boil down to just an encapsulation problem But for now there's some concern about the toxicity of the lead particularly your water soluble source of lead Another problem that we have seen you know in our labs Have you've seen this firsthand is that these materials are not stable to moisture? so we can form nice uniform black films of this material and Here's a nice annealed film of the 3d perovskite And if you just let it sit on the lab on a bench in the lab over the course of a few days I hope you can see this there are these white sorry yellow patches that start forming on the surface of your film So it's bad news when you have black solar cell adsorber turns yellow that means it's not absorbing sunlight So these yellow patches are lead iodide patches that form this material converts to lead iodide in ambient moisture In fact, you can see this even more clearly in the absorption spectrum This is the band gap of your perovskite and over time you see that this band gap decreases And there's this higher energy band gap material that forms this shows that the perovskite material is decomposing and lead iodide is forming instead and The final point I want to bring up is something that has become more apparent recently I told you that by adding bromide into this lead iodide lattice we can change the band gap so we can Form materials that are that have a band gap of 1.6 using a lead iodide lattice as well as a material with a band gap of 2.3 we using lead bromide lattice and by Using intermediate stoichiometries of lead and bromide we can actually access all intermediate band gaps But when we look at the voltages that have been achieved using these materials the lowest band gap material gives a low voltage The highest band gap material gives a high voltage But the intermediate band gap materials do not give higher voltages than the lead iodides material and this is very curious So we decided to investigate why this might be as well So we have efforts in both our groups to address all three points But for the rest of the talk I built I will focus on the last two points and show you our efforts on trying to make a more moisture resistant material Lead perovskite with greater resistance to moisture and I'll also talk to you about our investigations on why this this mixed hairline materials Don't deliver the high voltages that we expected them to give us So I show you again the picture of the three-dimensional perovskite. It's it's a beautiful structure But for a synthetic chemist trying to make a material modification This can also be a very frustrating structure because there isn't much we can do So if you want to maintain that the beautiful adsorption properties of this inorganic framework We really don't want to mess with the inorganic structure So that leaves us with only these very small organic groups to play with this cavity really restricts the size of the molecules That can enter this structure So if you want to improve the moisture stability We could swap out this cation for another small cation, but there isn't much we can do we wanted more space So we took a step back and looked at some related materials that we had been studying before we moved on to 3d 3d perovskites So in my lab prior to our work on solar cells We had been working on the on the layered perovskites that I showed you in one of my first few slides So these materials have just one inorganic sheet sandwiched by organic layers above and below and this is a repeating unit And I'm going to call this n equals one This has a very large band gap and this has very high exotone binding energies Which means when you hit this with light and you generate electrons and holes these charge carriers are very strongly attracted to each other in fact these have all the properties that you seek to avoid in a solar cell adsorber and We know that these materials have have other advantages We've found that some materials which have these n equals one sheets are actually very nice Fossils so when we hit this material with blue light or with UV light the material emits white light Broadband white light which looks like sunlight And we know that the reason you get this emission is because this layered material gives rise to electronic structures which which induce very strong exotone binding energies as well as very localized charge carriers and We know that if you were to simply expand the thickness of this inorganic sheet all the way to n equals infinity We know that we can decrease the band gap and we know that we can also decrease the exotone binding energy and We also know that these two are extremes in what's really the continuum of the hybrid Hybrid perovskite family so we decided to make the intermediate species and you can see that there's this nice Change in the electronic structure as we go from n equals one two three all the way to infinity the band gap decreases and Most importantly the exotone binding energy decreases So now when we generate these charge carriers, they don't recombine they can move around and they're free from each other So you and Smith in my group crystallize the first N equals three lead iodide perovskite he formed single crystals of these materials This is the crystal structure and he decided to make a solar cell with this to investigate whether three n equals three was thick Enough to generate photo current in a solar cell device I will give a brief description of Ian's work here, but Ian is giving his own talk today later on in the symposium So he constructed a planar geometry solar cell device where he basically used titanium a flat layer of titanium as the electron Transport material deposited the the layered perovskite as the light absorber and spun an organic whole transport material on top of that And he found that this material does indeed work as a solar cell first generation devices have power conversion efficiencies of about 4.7 percent and This compared to the the highest efficiencies with 3d 3d perovskites Which have now reached about 18 percent is a low number But this is just this is not an optimized device in any way And I like to discuss some advantages of the 2d material with respect to the 3d structure So for one this has a larger band gap so we can access higher voltages These voltages are higher than any voltages that have been achieved with the 3d perovskite solar cell These materials are layered the structure is layered and therefore they grow as splits This is in contrast to 3d perovskites which grow as cubes blocks and because of this layered structure These materials love to form beautiful uniform films So we can simply drop cast this material on a slide and not not you This doesn't require any kind of high-temperature annealing we can form nice continuous films and Here I show you a film of the 3d of the 2d perovskite and in contrast similar deposition techniques for that for the 3d material don't give the same nice surface coverage and By far I think the most important point here is that the 2d material is a lot more stable to moisture So with the 3d material here, I show you powder XRD patterns and over time we see when this is exposed to about 50% Humidity we see the growth of new peaks which completely dominated the spectrum after about 40 days This is lead iodide whereas in the 2d material this material is pristine untouched by lead iodide over 40 days So we think that there might be a critical value of n where we can make the inorganic layer just thick enough that it Pretends it's a it's a 3d material and has all the beautiful adsorption properties of the 3d structure While the organic functionalities can bring bring new functions into this materials So you can imagine putting flora carbons here This would look like you've taken your 3d material sliced it and coated those those sheets with teflon to improve the moisture stability at the atomic level and we can imagine even molecules that adsorb light or transport charge to bring new functionality to these materials so With the rest of my time I'm going to switch gears and talk to you about some very curious properties We've seen in in mixed hairline perovskites in 3d perovskites So here I show you again the bandgap of these materials as you dope in bromide to the lead iodide lattice and you see this nice Increase in bandgap energy as a function of bromide content into this lattice And if you look at the photo luminescent spectrum of these materials again as we expect We see that the energy of the emission goes to higher energies as we have more bromide in the material This is exactly as we would expect and we get this beautiful series But Eric found that if he kept shining the light on these materials just at once on illumination for just about a minute This material changes changes dramatically so he finds that as He keeps illuminating these materials the initial photo luminescence starts to die down and a new photo luminescence band Starts to grow in that gets very high in intensity at around 1.7 eV Curiously it doesn't matter where the initial photo luminescence band begins They all merge and the eventual photo luminescence arises from this from this band at 1.7 eV So the all bromide Material does not show a shift in PL the all iodide material does not shift but every other intermediate stoichiometric changes and moves to this material that emits at 1.7 eV and What's really surprising to me about this transformation is that it's reversible When you turn off the light you regenerate the initial material and when you turn on the light You get this new new low-energy PL and you can cycle between this Eric has done this several times here. I show you data for four of these cycles So we decided to study this material under illumination to figure out what on earth was going on And we did several experiments So here Eric sees that under illumination a new sub band gap shoulder grows in at around 1.7 eV And again when the light is turned off the shoulder decreases light is turned on the shoulder grows in and so on this can be cycled and Eric then wanted to study the structure of this material under illumination so he He decided to look at the X-ray powder diffraction pattern of these materials under illumination So he employed this very sophisticated Experimental setup where he suspended a bicycle lamp inside the diffractometer and He just turned on this lamp and this was enough to cause a change In the in the powder pattern of this material when the light is turned on He sees that every peak is split into two peaks and when the light is turned off He sees that the original peak grow back in and this Split peaks show the presence of a material with a larger lattice than the original as well as a material with a smaller lattice Than the original mixed hairline perovskite so We went to the electronic structure of this material to try and understand what might be the what might be going on here and Here I show you for the X equals 0.6 material the valence band as well as the conduction band when we hit this with light There's there are holes in the valence band and electrons in your conduction band We know that initially the photo luminescence comes from these electrons falling from here to here But in time we get this low energy photo luminescence band, which means we are forming a trap So here's my trap We have some state that where the holes fall into a state with lower energy We know that the energy of this photo luminescence corresponds to the same energy as the photo luminescence from an x equals 0.2 material So we asked the question. Maybe why not form the the x equals 0.2 material, but only under illumination So our speculative mechanism can be summarized by this picture. We have this homogeneously mixed bromide iodide lattice Lead bromide iodide lattice, but when we shine light on this material There's hairline segregation that is initiated by the illumination and here the my darker blue squares show Show areas of higher iodide content and this lighter blue square show areas of lower bloom lower iodide content so there's hairline segregation under illumination and all the holes of course will move to the lower energy Material, which is the iodide rich material and therefore we get photo luminescence only from that trap and our data so far support this this speculation if We were to fit this this shoulder that we see image in the in the light-surf material That actually corresponds quite well to what we would expect from the adsorption spectrum of the x equals 0.2 material if about 1% of that material converted and If we were to form regions of high iodide content That means we must by necessity form regions of low iodide content as well and that also explains a split diffraction peaks Because here this peak matches well with the x equals 0.2 material and this peak matches quite well with the x equals 0.8 material So hairline segregation implies hairline motion under illumination and we can we know that this is a thermally activated process because at different temperatures we see different rates of this photo luminescence band growing in and we can Calculate an activation energy for this for this process we get a number of about 2.3 ev and this number fits very well with activation energies that have been attributed to hairline migration in metal here in other metal hairline perovskates So just to end I want to summarize this section We see very interesting reversible changes in this materials upon illumination so far observations are all consistent with hairline migration initiated by stabilization of these holes through hairline segregation and We see this behavior even if you were to change the organic ions in this material We see this behavior in films formed through all the deposition Techniques mentioned in the literature for making this these materials we see this even in single crystals of the material so we know it's inherent to the material and not due to any any weird surface sites and Though I think this is incredibly cool in terms of solar cells This restricts the the voltages that you can obtain from from mixed hairline perovskates And if you really do want to form higher voltage perovskates We need to either try to prevent hairline segregation in these materials or Look at alternative strategies to increase in the bandgap So for example changing the metals or working with the lower dimensional 2d materials that I talked about previously And I'm out of time. So I'll end there I want to thank my group for the inordinate amount of hard work They put into their work the students in my group and in Mike's group who have directly contributed to the Jusep FS Listed here, but we often rely on the combined expertise of all members in both groups to move our work forward And I'm very grateful for a Jusep support of this work that has really accelerated our collaboration I'm done. Thanks for listening. I'm happy to answer questions Okay, do we have any questions for him a result of the light and segregation Is the first instance I've seen of a light dependent So ability so it seems as though the solubility is one number without light and another number with light so Have you ever seen this before and it can't be a kinetic effect because it's those back and forth So it's yes changing somehow the thermodynamics of solution Yes, I have seen cases of this previously where the the migration was Not reversible so light-induced halide segregation has been noted in lead halides in silver halides When you use silver halides for photography, for example, you induce halide segregation in these materials Which are driven by the halide vacancies in the material, but I have not seen a reversible process up to now Very interesting data Also relates to Bruce question. So like I'm saying and then this halide start to move around and segregate So do they come out of place toast and then go into the environment? Would that be a degradation mechanism? So I know that certain is well known like I'm saying and then split it and then I'll like go Yes, we have fooled around with a few of those experiments so for example if we take this material and You know disperse it in hexanes and over time if you ready this material and you do a starch test in the hexane solution You see I didn't so so the halide will eventually leave as as x dot and 2x dots will get together and form the halogen Those are extreme cases where we basically try to decompose the material in these materials because we see such beautiful Reversibility I don't think the highlight will leave the material the stability issue under moisture Have you checked that out to see how is it worse when there's light present or I think all your stuff was done in the dark Yeah, actually, we haven't considered the effect of light for the stability test. We just use this control humidity chambers That's an interesting point. We should do this in the dark and in the light. We haven't noticed any Any differences that really stand out for now if there are no more questions. Let's think came again