Scientist Stories: Jennifer Dionne, Nanophotonics to Visualize Dynamic Processes near Atomic Scale

Thank you so much for the introduction and the opportunity to be here. Can everyone hear me in the back okay? Maybe I'm too loud. Yeah, usually people complain that I'm too loud and ask me to lower my voice a bit. Well, today I wanted to share with you some of my passions. Hopefully in the poster session you'll get to hear about some of my passion for science. I also really love food. I love cycling, and I also love photography. And in the presentation today, I want to share with you a little bit about my love of science and photography combined. It's kind of amazing that humans collectively take 1.5 billion photos each day. That is a ton of information with which to learn about ourselves, our community, our history, and where we might be headed. Some of the best pictures can take us to places we've never been before, like this famous image of the supermassive black hole at the center of the M87 galaxy. Yet other images can teach us about our place in the universe, like the famous image of the Earthrise, which in the words of Carl Sagan tells us that we're just a pale blue dot, the only world known so far to harbor life and a place that we must cherish and sustain and learn to treat each other more responsibly and kindly in. Still other images can inspire us to strive to new heights, like this famous portrait of one of my favorite athletes, Michael Jordan, who famously said, "I have failed and failed and failed again, and that is why I succeed." And I think we heard some of those messages in the Q&A session earlier. And yet other images can teach us what lies within, like this first X-ray of Professor Ronken's wife. And I think it's images like this that tell us that the future can hold far more promise for medical advances and for advances in sustainability than we can possibly even imagine today. And yet, when we reach the molecular scale, we can't just keep zooming in on our cameras to understand what's happening. Oftentimes we wind up with animations like this animation of a microtubule with a motor protein walking along it. So here the scale bar of the protein is roughly about 10 nanometers, so the protein's 10 nanometers across. And while this animation is derived accurately from the science, it's an animation nonetheless. And when I first saw this animation, when I started as a faculty member, I thought, like, wow, wouldn't it be really cool if we could develop microscopies or imaging techniques that allow us to get videos like this, but of actual processes as they're occurring? And that's a lot of what my lab at Stanford focuses on. We try to develop new tools and new techniques to be able to image dynamic nanoscale processes. And in many cases, those techniques rely on combining photons with electrons. So my lab has developed a number of electron microscopy techniques that enable input of various gases or liquids that can react with their environment. In this case, we're looking at hydrogen molecules reacting with a nanoparticle. And by looking at either the electron spectra or the optical spectra, we can get a feeling of how the molecules are interacting at the surface of the nanoparticle and intercalating into the nanoparticle. So here I false-colored in red the region where the hydrogen molecules are interacting with the nanoparticle, intercalating in and causing a phase change. And movies such as this that we can record in real time with near-atomic-scale resolution allow us to optimize catalysts and photocatalysts to enable improved, say, energy storage materials or catalytic materials for sustainability. We've also developed combined electron and optical techniques that allow us to look at how electrons and holes are recombining in materials. These are the sorts of processes that are fundamental to solar cells, photocatalysis, and light-emitting diodes in displays. And here, the video that you saw showed how electrons and holes are recombining in 3 dimensions. So just like Professor Hell was discussing earlier, we have the ability to do real-time imaging, but also in 3 dimensions, so doing nanoscale optical tomography. And then we also do some nanoparticle development, and in particular developing nanoparticle sensors of various external stimuli. So in this case, what we've done is develop light-emitting nanoparticles that change color in response to applied force. Now, you may wonder, why do you want to have a force sensor? In my own lab, we're very interested in biological forces. We experience forces every time our ears hear, our heart beats, a wound is healing. Forces are also very critical at the interface between our immune cells and pathogens. So they're thought to be quite important in terms of understanding our immune responses. So my group is trying to develop sensors of optical force that work in the nanonewton regime. So that way we can better understand immune responses to pathogens and hopefully develop better immunotherapies. And here's just one example of nanoparticles that are changing their emission color from yellow to red with applied force or with applied pressure. You can see from the transmission electron microscope that they're pretty small, about 20 nanometers across, and then they're also very biocompatible. So as a first case study, what we've done is just We feed them to worms, C. elegans, a small roundworm. They chew them up. You can see them here in the digestive tract. And then as they're being eaten, I'll just show you this false color image. You can see the nanoparticles moving throughout the digestive tract, again, eventually being excreted. And based on the color, we can tell what the magnitude of the force is. But more than pretty pictures, I think what these dynamic videos have taught me is how the most fundamental building blocks of matter, namely molecules and photons, are interacting. And from learning about that interaction, we can develop new technologies. And what I want to share with you today is work that my group has just recently started to work on, but we've made some progress on. And using information about light and molecule interactions to develop better pharmaceuticals and better agrochemicals. So to teach you a little bit about the history of this project, I want to bring you back to 2014. I was living in the Bay Area. It was the rainy season. And for those of you who have been to California in the spring when there's a lot of rain, there are also a lot of snails. And I found these two snails mating outside my apartment. As you can see, the snails— or the snail shell has a certain twist associated with it. In this case, it's a right-handed spiral. So the snail has a chiral shell, namely a non-superimposable mirror image or twisted shell. And intriguingly, this twist propagates through all the way to the internal organs of the snail, meaning that a right-handed snail is literally only a good fit for another right-handed snail. So you tend to find snail colonies dominated by exclusively one-handedness or one chirality of the shell. Now, like the snail shell, there are many natural materials that are chiral and that have a certain handedness associated with them. That would include DNA, which of course occurs in all natural materials as a right-handed helix. Its non-superimposable mirror image would be a left-handed helix, but like I said, in all natural healthy organisms it's just the right-handed spiral. Sugars also have non-superimposable mirror images like glucose and fructose, and again, in all natural kind of healthy organisms, we tend to find just the right-handed version of the sugar. And also amino acids, the building blocks of proteins, are chiral, generally all amino acids are left-handed. Now this homochirality of nature is perhaps one of life's greatest mysteries, one that many scientists are still trying to figure out the origin of, but importantly, because life and healthy life is homochiral, the presence of an opposite-handed molecular structure can have a pretty profound impact or can signify disease. So if we look, for example, at amino acids, which, like I said, are exclusively left-handed, it turns out that once disease has onset—for example, kidney disease or schizophrenia or Alzheimer's—you tend to find an increased presence of right-handed, or D-amino acids. So if we can develop techniques to more sensitively detect the presence of right-handed amino acids, we might be able to detect disease at a much earlier onset than we currently can. And just to give you a few more examples about the function of different-handed molecules, if we take this molecule limonene— here I'm showing you the organic structure. You've seen organic molecules throughout the talks today. So here each of the vertices corresponds to a carbon atom. When you're synthesizing this molecule in the lab, You could imagine that this last hydrogen, if you're adding it on, could go with equal probability to the right or to the left, and depending upon which way it goes, the molecule will either smell like orange or it'll smell like pine or turpentine. So these opposite-handed molecules can have very distinct functions, in this case for your olfactory sense or your sense of smell. Now, the importance of structure or molecular structure in pharmaceuticals was brought to prominence in the late 1960s when the drug thalidomide was first marketed in Germany. And it was supposed to treat morning nausea in pregnant women. But unfortunately, the other enantiomer, or the other-handed chiral molecule, actually induced birth defects and in about 50% of all cases, fetal death. So because of that, in the United States and worldwide, the FDA has cracked down a lot more on better understanding what the side effects of various, uh, handed molecules might be. But still, if you look to the pharmaceutical industry and also the agrochemical industry, approximately 50% of all drugs on the market and 30% of all pesticides and herbicides are chiral. And of those, more than 90% are sold as racemic mixtures. Those would include, for example, the drug Advil. Who here has taken Advil for pain relief? Okay, a couple people, or at least a good number of people. So Advil is sold as a mixture of both the right and left-handed molecule. Probably it takes about 30 minutes for you to feel pain relief. And that's because the wrong-handed chiral version of drug basically slows down the efficacy of Advil. So it gets in the way of, say, the left-handed chiral drug interacting with your left-handed amino acids, or the building blocks of proteins, in order to block those pain receptors. And if it were sold as a pure drug, it would only take about 10 to 12 minutes for pain relief in Advil. Also, the antidepressant Prozac is sold as a mixture of both the right- and the left-handed molecules, and the wrong-handed enantiomer can give rise to some of the negative side effects like headaches and heart palpitations. And then in agrichemicals, for example, like the herbicide fipronil or the pesticides that are common, they can leave residues in soils, giving rise to colony collapse in bees. And organ failure in birds and fish and larger mammals. So why is it that so many agrochemicals and pharmaceuticals are sold as mixtures? Well, it's because, like Professor Ebison was mentioning earlier, it's actually very challenging to separate the chiral molecules. Essentially, these enantiomers, or these non-superimposable mirror-image molecules, are identical in what I'll call scalar physical properties. So because they have the same atomic composition, they have the same mass, they have the same charge, so you can't use conventional techniques to separate them. And for those of you who have done basic chemistry experiments, generally, you know, centrifugation is a great technique for separating molecules of different weights, right? You just spin them around and the heavier ones crash out first. But you can't use that because they have the same weights. They also have the same charge, so you can't use, for example, a battery or an electric field to separate these molecules. So the only way they can be separated is by coming up with another entity that is handed and that preferentially interacts with the handed molecule. So some of you may have run what's called a column. The typical way of separating chiral molecules in industry today is to run a chiral column where you take your mixture of both the right and the left-handed, um, synthetic molecules that you've made, and then you run them through a tall glass cylinder where packed within that cylinder are silica beads. So you have these micron-scale beads, and then those are functionalized with molecules that's called the chiral stationary phase. So this is a molecule that is made specifically to interact with one-handedness of the drug. So that way, as the drug or the, um, agrochemical passes through, it basically, uh, sterically interacts with the chiral stationary phase and gets slowed down. So you can see in this case that the right-handed version of ibuprofen basically would latch on to the right-handed version of this chiral stationary phase. And therefore it would pass through the column more slowly than, say, the left-handed version, which goes through more quickly because it doesn't geometrically interact with that chiral stationary phase. So for each new drug that comes to market, there are companies that have to work on developing a new chiral stationary phase. And often that's a process that can be very time-consuming and very expensive. It's done usually by empirical methods, but it's usually a trial and error process. So this is why many pharmaceuticals and agrochemicals tend to be sold as mixtures, because right now there's still no efficient and effective way of separating chiral compounds. So in my lab, we are interested in light and molecule interactions. And we know that light can also be a chiral entity. So if you think about light as an electric field and a magnetic field that's propagating through space and time, you can simply pass it through a polarizer or a wave plate that enables the electric field to trace out a helix in space or in time. And then when you look at the peaks of that helix, that corresponds to the wavelength of light. Now, if we want to make sure that we're not destroying our molecules in the process of illuminating them, we're generally going to want to work at optical frequencies, so ultraviolet up through near-infrared frequencies. If we start getting to higher frequencies, um, we might start destroying the molecule. So the key challenge with using light as a reagent, as basically your chiral stationary phase, and separating the molecules is that the wavelength of visible frequency light is much, much bigger than a molecule. So the typical wavelength of visible light is between 700 and 400 nanometers, and the typical size of a molecule is on the order of a nanometer, or up to a few nanometers for some of the proteins. So that's one of the challenges. Another challenge is that most molecules are blind to the magnetic field of light. And the key reason why that's really important is because enantiomers are basically electric dipoles and magnetic dipoles. And the key thing that distinguishes them or that makes them non-superimposable is the direction of their magnetic dipole. And to kind of conceptualize that a little bit, maybe you can think about that helix of DNA again. Where you can think about the z-axis of the helix as being the electric dipole, and then you know that magnetic dipoles correspond to current loops. So if you think about the loop of the helix, for one of the enantiomers, that helix is going to the right, whereas for the other enantiomer, that helix is going to the left, and that corresponds to completely out-of-phase magnetic dipoles in the simplest of chiral molecules. So we need to come up with ways of correcting for that size mismatch between molecules and light, and also for enabling light or the magnetic field of light to interact more strongly with chiral molecules so that way we can better distinguish them or use light to initiate kind of an optical handshake with one enantiomer. So how do we address that? Well, we use nanostructured materials I'll call them loosely metamaterials just for historical reasons, but you might find them in the literature called metasurfaces or Mie resonators. Broadly, though, I just mean nanomaterials that have optical frequency, electric and magnetic resonances, and that act as nanoantennas to focus light down to the nanoscale. So here's one example, the one that Oprah got excited about, these so-called negative refractive index materials. This is just a simulation of negative refractive index water, and you can see that the straw gets bent the wrong way. You also can't see the bottom of the glass of water. So these negative refractive index materials have a lot of interesting and unusual properties, and the key way you get that is by tuning both the permittivity and the permeability. So we saw earlier in one of the talks that the speed of light was related to the permittivity and the permeability. Basically, the product of those two, epsilon and mu square root, is the refractive index. So to tune the refractive index, you need to tune both the magnetic response, which corresponds to the permeability, and the electric response, which corresponds to the permittivity. Now, we're not going to be focusing on making negative index materials per se, but we do want to have that magnetic response. And that means that we're going to be working with materials that essentially tune what the effective optical index is. And the larger the optical index is, the smaller the wavelength of light is. Here I'm just showing you one example of a metamaterial. It's a colloidal, um, synthesized assembly of, in this case, small metallic nanoparticles that are arranged around a silica sphere. And the size of the structure is much smaller than the wavelength of light. So that way when they're dispersed in water, they essentially act like artificial atoms or artificial molecules that can control light. And here I'll just show you one of these structures. You can see the scale bar is 30 nanometers, so this structure is about 100 nanometers across in total. And you can see that the magnetic field is strongly enhanced within the core of this particle. We're just looking at a cross-section of it. And you can also see that the electrons in the metallic nanoparticles are tracing out kind of these circular displacement currents, which is what's giving rise to that strong magnetic field. So this set of structures where you essentially use either metals or dielectrics that are nanostructured can help us shrink down the wavelength of light and also can give us strong magnetic light-matter interaction. So essentially, these materials exhibit optical frequency magnetism. Now for this talk, I'm going to focus on a slightly different nanostructure. I'll call it a coaxial cable. It's a lot like the coax that you would plug into the back of your cable television, but shrunk down by many, many orders of magnitude. And it's comprised of a metallic core, a metallic cladding, a dielectric channel, and then circularly polarized light will enter into the dielectric channel and essentially excite collective oscillations of the electrons at the metal-dielectric interface, and then the light will be scattered away from the structure so that it can interact with the chiral molecule. And when you arrange these nanostructures in an array, you again wind up with a metamaterial, or basically this periodic nanostructured material, that gives you control over the local refractive index as well as the electric and magnetic fields. So the vision in terms of separation, or one vision that you could think of, is to have an array of these coaxes essentially arranged as a film. And then when you illuminate this array with circularly polarized light, what the mixture of enantiomers winds up being separated by virtue of this optical handshake that's initiated between light and the molecule. So in this case, the right-handed enantiomer gets attracted towards the coax. And I'll mention that this technique is actually quite generalizable. So we did it first with this metal dielectric coax, but now we're doing it with nanostructures that are more amenable to packing in the column. So that way you don't need to have that chiral stationary phase. You can just illuminate your column with a series of LEDs and then have the nanoparticles have those optical frequency electric and magnetic resonances to attract one-handedness of the molecule towards that structure. Okay, so can we make this particular structure and show that it exhibits optical forces that are preferential based on the handedness of a specimen nearby? So we use a technique called focused ion beam milling, essentially focused gallium ions that allow us to mill away tiny little structures in the metal. Here you can see the coax. So we have the metallic core, the metallic cladding. Here's a dielectric channel. And then we also patterned a grating around the coax, actually inspired by some of the earlier work of Thomas Edison, in order to increase the transmission that's coming through the coax. And if you zoom in a little bit more to that coaxial array, you can see the dielectric channel is just about 30 nanometers across, so about the size or roughly half the size of a virus. So now let's look at some calculations of how light would interact with this coaxial structure and how that light would interact with the chiral specimen above it. So this is a, a calculation basically solving Maxwell's equations and trying to figure out what the pulling force would be of light that has exited from the coax on a right-handed structure that for simplicity we just assume is roughly a 5-nanometer sphere with an electric dipole and a magnetic dipole that's close to what you would find with a molecule. So you can see for this right-handed structure, With right circularly polarized light, the pulling force is uniformly distributed around that dielectric channel. So we get the strongest optical force interacting with that chiral specimen when the chiral specimen is above the dielectric channel. You'll notice that the pulling force is about 2.5 piconewtons. That's not a very large force when you think about, you know, typical forces that we're used to working with, for example, Newton scale forces to, uh, weightlift at the gym.. But it turns out to be large enough to exert a significant force on molecules and to separate them. And of course, force is a vector, so if we plot not just the magnitude but also the direction of the force, we can see that regardless of where the molecule is going to be in space, it's always going to be attracted towards the coaxial structure and eventually pulled through it. Now, if we compare this with the left-handed structure, you can see that the pulling force is quite a bit weaker, and then if we look at the direction, that pulling force is also repulsive. So the molecule will actually get repelled away from the aperture, giving rise to that physical separation that we had envisioned. So how can we quantify these forces and see if our calculations hold true? Well, we use a technique called atomic force microscopy where we take, um, a nanostructured pyramid of either a dielectric or a metal, or as I'll show you later on, a molecule, and then we put it on a cantilever. It's essentially like a diving board that we're scanning over the surface of our nanomaterial or a metamaterial. And based upon the optical force, this whole cantilever plus the tip on it will either get deflected towards the sample or away from the sample. And what you do is you can bounce a laser off the end of this cantilever, and based upon whether the laser moves down or up, you know whether your tip has moved down or up, and therefore is attracted or repelled from your structure. So it's a really powerful technique. People have used atomic force microscopy to map out with very high resolution the structure of materials, with near-atomic-scale resolution at this point, based upon the sharpness of the tip. But we wanted to use it to map out these optical pulling forces. So what we did as a first proxy for a chiral molecule is we just patterned a spiral at the end of our atomic force microscope tip. So this spiral is a lot like those snail shells I showed you at the beginning of my talk. In this case, we have a right-handed spiral. And when we turn the light on, we can see the tip gets deflected down towards the sample. And when we turn the light off, the tip remains in its stationary phase. So in this case, it's a certain height above the coaxial structure. And we're just turning the laser on and off. And we can measure the pulling force because we can measure the laser deflection coming off that cantilever. Here we're just illuminating the coax at one particular wavelength. This corresponds to the resonant wavelength of the coaxial structure, which is where we'd expect to see the largest forces. And then because we have that sharp tip, what we can do is scan it over the coax. So in this case, we're scanning it over just one quadrant of the coax so we can get slightly higher spatial resolution. And when we have right circularly polarized light illuminated on this right-handed AFM tip spiral, we can see that the pulling force is quite symmetrically distributed and large, whereas for less circularly polarized light, the pulling force is much smaller. So we can quantify those pulling forces as a function of the separation from the coaxial array to the chiral tip. And what we see is that for left-handed illumination, the pulling force is always repulsive, so it's positive, meaning the tip gets deflected away from the surface. Whereas for right-handed illumination with our right-handed, um, spiral, the force is always negative, meaning it's always attracted towards the sample. So this is kind of the crux of this part of the presentation, showing that circularly polarized light can exhibit preferential forces on enantiomers. And in addition to swapping out the light, we can also just swap out the spiral pattern that we mill for the right- or left-handed versions to show that we get similar results. We're not just limited to doing this on AFM tips. We can do this on molecules, and as a test for whether it worked on molecules, we were able to attach just a single DNA to the end of our AFM tip, and then we could again sweep the separation between the coaxial fiber and where the DNA was, And we can see that when we're in the near field or when the molecule is quite close to the coax, it gets attracted to that channel, whereas when we're in the far field, so at separations greater than 300 nanometers away, there's very minimal interaction. So 20 base pairs of TNA is a 4-nanometer molecule. So this is getting us closer to the regime where we can start working with pharmaceutically and agrochemically relevant molecules. Molecules. So this was a metallic structure, and you probably saw that the resonant wavelength we were working with was in the red, so at 770 nanometers. In addition to getting this to work for smaller molecules, we need to be able to scale the wavelength so that way our structures are resonant where the chiral molecules are also resonant. And for most pharmaceuticals and agrochemicals, that tends to be in the ultraviolet regime. That's where you'll have electronic resonances in your molecules or electric and magnetic resonances. It's largely because in the pharmaceutical industry they don't want to design the drug to have, uh, a color associated with it because when you excrete the drug you might panic if it came out bright pink. Uh, so it makes our life as physicists a little bit more challenging. But we can work with materials that have resonances in the ultraviolet regime. And these are actually some of those nanoparticle-type structures. So we could imagine packing a column with these nanoparticles and then just changing the geometry of the nanoparticle in order to be able to selectively trap or slow down a particular new drug that comes to the market. And then in certain cases, people are also interested in using vibrational modes to separate molecules. So this is why I think I was paired up with Thomas Ebbesen. It's kind of a nice complement. So we can also design materials that work better in the infrared regime. So we can target vibrational modes and therefore wind up separating molecules by illuminating them with circularly polarized light and separating them by resonantly exciting a vibrational mode, causing one enantiomer to have a lower molecular weight than another by virtue of bond breaking. And then as my final slide, I'll just share with you how beyond just sorting chiral molecules, we can also use these structures to direct an enantiopure synthesis. So generally when you have an organic synthesis, it's quite challenging to develop a chiral catalyst or a chiral synthesis, one that preferentially just gets you, say, the right-handed molecule. But with these nanostructures, we can take a synthesis that normally would just be, say, the reactants yielding equal probabilities of the right and left-handed molecule and giving rise to, for example, just the right-handed molecule. And here you can see what the yield would be for a particular reaction as a function of the enantiomeric excess. So experimentally, people have actually done this with circularly polarized light alone. And they can get pretty high in antiomeric excesses on the order of about 10 to 20%, but they wind up destroying most of the molecules. So for a high enantiomeric excess, you wind up getting a low yield if you just use circularly polarized light. But with some of the nanostructures that are designed to direct the synthesis, we can get high yield with higher enantiomeric excesses. So we're approaching the point where this technique can compete with industrial techniques like chiral catalysts and HPLC or chiral columns. So with that, I'll summarize. In the beginning of my talk, I shared with you just some of my favorite images, and those kind of inspire my work in the lab. Stefan Hell was saying earlier he's not a microscopist. I do kind of consider myself a microscopist. Or maybe better, a nanoscale videographer. I really love capturing videos of the nanoscale because I think if you can see something, you can better understand it and then hopefully better improve it. And by combining electronic approaches with photonic approaches in generally what I'll call a nanophotonic approach, we can get a lens into these dynamic molecular and nanoscale processes. And then from those lenses, What I've learned is how to better control light interactions with molecules, and hopefully you're walking away with the message that light might just be the next handy reagent for pharmaceuticals and agrochemicals. We're not used to thinking about light as something you can tweak in your chemical reaction, but I hope you'll walk away with that from my talk. So by using light combined with nanostructures, we can enable both high-sensitivity chiral sensing to detect trace amounts of opposite-handed amino acids. And we can also enable low-cost separation and potentially enantiopure syntheses of chiral molecules. You can think about this, like I said, as light initiating an optical handshake with chiral molecules. Or going back to our snails, perhaps a little optical smooch that hopefully enables safer pharmaceuticals and agrochemicals and a healthier planet and population. And with that, I'd like to make sure I acknowledge, uh, all of the students and postdocs who have made this work possible. Um, those of you thinking about being a scientist, I can't imagine a better career. It's not only a lot of fun every day in the lab to be able to interact with extraordinarily bright coworkers and colleagues, but it's also a job that gives you the opportunity to travel the world many times over, Here I'm showing you some pictures with some of my students and colleagues, role models. Recently I was in Singapore, which was a lot of fun. And then these are some of my groups throughout the ages. So thank you so much, and I look forward to the Q&A session.