Title
On Single Photons and Single Molecules: From Nano-Quantum Optics to Nanobiophotonics
Abstract
Light-matter interaction at the nanometer scale lies at the heart of elementary optical processes such as absorption, emission or scattering. Over the past two decades, we have realized a series of experiments to investigate the interaction of single photons, single molecules and single nanoparticles. In this presentation, I will report on recent studies, where we reach unity efficiency in the coupling of single photons to single molecules in a microcavity and describe our efforts to exploit this for the realization of polaritonic states involving a controlled number of molecules and photons. Furthermore, I will show how the underlying mechanisms that play a central role in quantum optics, help detect, image and track single biological nanoparticles such as viruses and small proteins with high spatial and temporal resolutions and in a label-free fashion.
Bio
Vahid Sandoghdar was born in 1966 in Tehran, Iran. He obtained his B.S. in physics (1987) from the University of California at Davis and Ph.D. in atomic physics (1993) from Yale University. After a postdoctoral stay at the Ecole Normale Supérieure in Paris, he moved to the University of Konstanz in Germany to start a new line of research that combined single molecule spectroscopy, scanning probe microscopy and quantum optics. In 2001, he took on a professorship at the Laboratory of Physical Chemistry at ETH in Zurich, Switzerland. In 2011, he became director at the Max Planck Institute for the Science of Light in Erlangen and Alexander von Humboldt Professor at the Friedrich-Alexander University of Erlangen-Nuremberg in Germany. Sandoghdar is one of the pioneers of Nano-Optics, which merges various research methods to investigate the interaction of light and matter at the nanometer scale. His work was honored with the Quantum Electronics and Optics Prize of the European Physical Society for “ground-breaking research on the efficiency of light-matter interaction in quantum optics and biophysics, leading to single-molecule strong coupling and label-free detection of small proteins.” He is the founder of the Max-Planck-Zentrum für Physik und Medizin, a joint research center that addresses questions in fundamental medical research with physical and mathematical methods.
More: https://www.optics.arizona.edu/colloquium
yeah so hello everybody um Welcome to our um Optical science colloquium it’s very nice to see so many of you here um today in the room and also so many of you on Zoom so today we have a very great speaker from Germany Professor vahit sandoka um Professor sandoka is a director at the max plank Institute for the science of light in arangan and he’s also an Alexander Fon hbor professor at the University alang and nberg um vahid is actually one of the pioneers of Nano Optics and this is a field that investigates light meta interactions at the nanometer scale and he is the founder of the max plank Centrum for physic and medicine and this is basically like a joint Research Center that combines fundamental medical research with a physical and mathematical methods vahid obtained his PhD in atomic physics from Yale and after that he was a postto at the EOL normal Superior in Paris and uh at the University of constance and before he moved to arangan on this very prestigious humbard professorship he had a professorship at the eth surich um vahid Won Won of course like many many prices um too many prices to tell you all here but uh one price that actually sticks out is the price for um Quantum electronics and Optics from the European physical Society vahid we are very happy to have you here today and the stage is yours no now you hear me thank you thanks very much Florian uh it’s a great pleasure to be here uh I usually don’t cross the pond just for a lecture uh but I was very eager to come to Arizona to tucon because this is one of the uh hotspots of uh Optical sciences and and uh I didn’t want this opportunity to uh to to to let go of this opportunity so I I’ve had uh several discussions and meetings and I’ve learned uh quite a b uh quite a bit this morning I’m looking forward to learning more uh tomorrow um so uh I’ve tried to put a colloquium together and I was told that should as usual in a colloquium I have to give more of a general introduction but as you will see our work is from a fundamental point of view extremely simple so anyone who’s here past first semester physics should be able to follow uh everything that I tell you uh and if you have questions then you feel free to interrupt me in the middle uh I’ll be happy to walk you through whatever you didn’t uh you didn’t follow uh but before I start the science uh since I come from far away and uh air Langan is not necessarily a well-known place in Germany and uh in fact I was just asked whether air was part of former East Germany or former West Germany uh so I thought that I would show you a little bit where airling is this is Germany places that you would know would be Berlin Frankfurt Munich Hamburg uh Airline is right here is about 2 hour uh distance to Munich or to Frankfurt uh it’s a town of about 100 th000 uh people but uh seed to many uh global companies including zans and zans Healthcare and you might be wearing uh products from Adidas or Puma they’re all basically in airl Langan but uh the most important thing in airl Langan is our Institute of course uh this was established in 2009 so it’s still fairly new uh one of the newest Max Blan institutes uh uh and um the building is from 2016 uh if you happen to be in Germany please do come and visit us uh uh and uh floran mentioned briefly we have now initiated a new uh effort this is another building that is being inaugurated this year uh on the campus of the university hospital and uh it’s an Annex of our building in a way uh or of our Institute and the goal is to put a whole bunch of physicists and mathematicians into this building right next to Medical Doctors and try to understand what their problems are uh so that we can try to address them um and my own group uh it’s kind of scientific past Is Anchored In more Quantum Optics and uh and nearfield microscopy and nearfield Optics but we have moved to be very very interested in uh problems of biophysics and and uh medicine and uh you will see a little bit of that in my talk um but let me start with something very simple and this is as complicated as the talk will get uh but uh it’s a it’s a problem that is actually very simple and very subtle at the same time and um uh I encourage enage you to kind of go think about it more or read about it more uh uh if if you get a chance um so the problem is as follows you have an object and you illuminate it and then you detect along the same path so you’re Illuminating along with a well- defined K vector and then you put your detector right along that line of sight and uh you you know that you’re going to detect less light because some of the light is going to be scattered for example by this object and uh you also know that some of it might be absorbed by the object but from energy point of view you know that you’re going to have less light in the forward Direction and the power that would be associated with this interaction could be written as the incident Power Times some kind of a sigma which is kind of a phenomological uh area divided by the area over which you had find light this is a very very basic kind of thing that you’ve learned in your lectures now it turns out that when you look into it and do this more rigorously this is actually an interference problem and uh they’re both completely equalent so you can look at this as an energy problem but you can also look at it as an interference problem in the interference problem you’re coming with plain waves and you’re scattering like a dipole let’s say if this is a small particle and these two fields are coherently adding up in the forward Direction and interfering with each other and the phase uh turns out to be such that you get destructive interference in the forward Direction and that’s kind of the same thing as saying you have less uh less light if you were to put your detector right here this is a kind of a very simple way of looking at it because when it comes to phases and and all of these things the problem can become quite uh subtle and this is typically uh you know introduced under the title of optical theorem in uh any kind of a problem uh context that involves the scattering whether it’s quantum mechanics or light or x-rays or whatever uh so the intensity that you detect at your detector in the forward direction is going to be uh the sum of the two Fields squared which means that if you expand it you’re going to get something like this so you have the incident intensity uh you have the scattered intensity and then you have the cross term and this cross term is called the extinction term all right so now let’s say that your object is very small so that the scattering is R scattering and again we know from uh oops from uh our elementary electromagnetic theory that the electric field that’s scattered is proportional to the electric field that is incident with a proportionality factor that is called polarizability and this is the expression that you would have for a polarizability of a spherical particle uh there are a few parameters that are involved uh there’s obviously the index of uh the medium and index of the particle um but more importantly what you probably don’t care about when you go through this is that it’s proportional to volume so polarizability has unit of volume and this is actually quite important when you think about uh sensitivity limit and things like that that I will be dwelling on for the next few minutes um now polarizability is also the fundamental property that goes into uh the scattering cross-section and absorption cross-section of course because the dialectric function or the index of a fraction are complex quantities polarizability is also a complex quantity so it has a real and imaginary part and the imaginary part is what gives you the absorption and the absolute value is what gives you the scattering cross-section so you can kind of go back and forth between uh these Concepts in different context now the point that I want to kind of emphasize is that even in a normal kind of Imaging conditions where you look at a shadow or where you do bright field microscopy or Imaging or anything uh what you have is an interference phenomenon because what you’re doing is if you have an object you illuminate it and uh if you were to write the electric field at this point you would you could write it as the electric field of the incident light plus the electric field that interacted with this object and got to that point and again you would have this kind of uh equation now it becomes interesting for us uh who are interested in nanooptics uh to ask the question what happens if I shrink this teapot to a single teapot molecule uh and uh the point the the the answer is well I’m no longer going to see the shape of the molecule or the shape of the teapot because of the defraction limit that I have at Optical frequencies but I will scatter some light from it so I will see a shadow from the single molecule and because there’s some light that is being uh uh taken out or scattered by this uh single molecule now if you have a single molecule uh or let’s say if you have a single atom that can be approximated as a two-level atom the way you do in your quantum mechanics course uh then the cross-section the extinction cross-section of that atom can be derived in one page simple calculation uh to be essentially Lambda Square divided by two so Lambda is the wavelength of transition from the ground state to the excited state and it’s typically in the range of a few hundred nanometers right in the visible range so Lambda square is actually a pretty large area that is the cross-section of a single atom or a single molecule okay um so when you send your light onto a single molecule you could say that it’s just a single molecule so I’m not going to see the shadow of it because it’s too little for it to make a difference uh on the other hand if you think about this cross-section you say well but this cross-section is not that little I can focus my light down to something of the order of Lambda so I should have something that is of the order of one if I think about Sigma over a the formula that I showed you in the beginning um and indeed we’ve shown in an experiment that is now more than 10 years ago um that you can detect a shadow of a single molecule what you see here is a shadow of a single molecule uh so the experiment was done as depicted here you have molecules that are sitting kind of apart from each other on a surface and we focus light through this sample and collect the light uh with another objective and send it to a detector um simplifying the detection process that you have to be somewhat careful to make this measurement because the shadow that I’m seeing is of the order of part per million of the intensity that I’m launching onto the molecule so I sent something like a microwatt And I detect a picowatt uh of loss when a single molecule comes in my beam okay now it’s so it’s detectable but of course it’s a small effect um so the thing to to kind of notice or to remember or be impressed by is that uh the signal that you get is uh uh coming because some light was scattered by the molecule so you might think of this as an absorption process but it’s the light that is scattered by the molecule that interferes with the light that was incident to give you this dip if you think in terms of those that interference term um now that was for single molecule but uh if you now back out and and say okay I have a nanop particle uh this is going to become a topic of the next uh 10 slides or so which connects to our more biophysical type of research um this technique is called inter formetric scattering microscopy in short I scat and uh the first time that we reported this was in 2004 and we were working on this because we were were trying to detect very small gold nanop particles and at the time in the late 90s early 2000s there was a lot of interest in detecting very small gold nanoparticles for different reasons uh this was basically the beginning of what you would today know as a field of plasmonics uh and a gold nanoparticle is kind of like the smallest plasmonic unit that you would have and we wanted to detect PL do plasmonics at this fundamental level so you wanted to detect very small nanop particles and do spectroscopy on it all that kind of stuff um but there was also a general interest in detecting very small particles uh without using fluoresence because uh fluoresence microscopy at the single molecule level was reported in uh the beginning of the 90s many of you were not born yet but it’s really not that long ago that’s when I got my PhD so that’s not that long ago um and and and during that time that is since the time that I got my PhD that was when people began to see single molecules but via fluorescence and then but there are only a few species in the world that actually fluores so it’s not the most universal way of detecting uh objects so you wanted to detect things possibly via its scattering because everything scatters light now it scatters maybe only a little but everything scatters but not everything fluoresces so this was kind of the background of uh this type of work and uh so in this work we reported detection of very small gold nanoparticles down to 5 nanometers but more importantly we quickly realize that we can detect any other kind of nanoparticle as well and in particular we could detect things like viruses because a virus now after the pandemic everyone knows viruses and knows that there are kind of nanop particles uh Corona virus is about 100 nanometers or so you no uh the virus that we were working on was about half the size and uh we were doing a very simple experiment we’re putting the virus at a uh a cover slide and now what you’re doing is you’re Illuminating this and instead of detecting in transmission we’re detecting in reflection so you’re detecting the light that is scattered by the virus in the reflection uh Arrangement and because there is an interface here there’s a little bit of light that is also reflected uh here and these two uh electric Fields interfere just as they did before in the forward Direction now in the backward Direction but it does the same thing and it gives you a scattering signal and this is now one of the First videos that we were recording of a single virus moving as a kind of a Brownian motion on a lipid membrane so this is the cartoon the virus is docked to a lipid membrane via its receptors but because the lipid membrane is a fluid kind of carpet the virus moves around and then we’re detecting how it’s moving yes you have a question is there a zoom issue that you have a we have like M okay fancy okay so uh there there was another technique that’s very old called alter microscopy where light is shined and then scatter light goes in and then you can detect it right that’s basically dark field microscopy I I’ll I’ll I’ll say I’ll I’ll connect to that in a second okay so this is kind of the summary of what basically you can do using intricut I told you about a transmission Arrangement where you do a standard Extinction but I emphasize that that’s also an inter formetric detection uh so you have an object you focus on it and then you detect in transmission you can do that with a point detector you can do it with a camera uh so that would be the Widefield version of it and this would be a confocal version of it and then you can do it in reflection this is what I just told you uh but you can also have an external reference if you want to do interferometry uh so this becomes a little bit more like holography that you uh come with a reference and bead it with a light that has interfered with your uh object um and uh uh this is the uh mode that is most common it’s kind of quite similar to this one except that it’s an Imaging mode it’s a wide field modes you illuminate the large area and collect all the light that is scattered back uh uh reflected back from the substrate and beat it with the light that’s scattered from the uh particle and then you get an image on this um now these methods there are all kind of sharing the common feature that there is an interference between the reference and light that was scattered by a nanop particle uh and in principle they’re very similar and very closely related to a whole family of interferometric microscopy that in a way date back to the 19th century so if you read the literature already in the 19th century people were doing interferometric uh microscopy uh except that at the the time they we’re looking at large objects so the thing that is really new now is that we’re pushing the sensitivity limit to detect smaller and smaller nanoparticles but the act of doing this type of experiment is common to all of these different types of microscopy so if you’re doing phase contrast you know you’re doing interference differential interference contrast interference reflection microscopy quantitative phase Imaging digital holography they’re all basically writing the same equation except that now our object is a nanoparticle so we can talk about ra scattering type of thing um and this is where your common to dark field comes in dark field detects at an angle that goes off axis away from the illumination so you don’t pick the reference intensity or the reference field you only detect the intensity of the light that was uh scattered by the object and uh I told told you that the polarizability proportional to volume which means the intensity is proportional to volume squared or the diameter to the sixth power so now we see that if you were to detect via dark field your signal would go down very rapidly as you go for example from a 100 nanometer particle to a 10 nanometer particle it would be a million times weaker so the dynamic range that you need to do darkfield microscopy is huge and basically your detectors are not capable of doing that if you were detect if you were to detect the extinction term then your dynamic range can be in that case a thousand times less uh but in terms of the short noise limit you probably all know that you know there’s only a factor of two Improvement when you do a homine or heterodon type of detection compared to a normal uh absorption detection all right so that was kind of the basics and now I’m just going to show you a whole bunch of applications that became possible possible because we could do this I showed you that we detected single virus and a single virus is a pretty small object especially if you realize that the index of refraction of a virus is only of the order of 1.6 it’s in water 1.3 so the index contrast is actually not that large so the fact that we can detect light from a 50 nanometer particle in such weak index contrast shows that the method is extremely sensitive but we wanted to be even more sensitive to detect single protein and uh we uh kind of pushed things a little further uh uh and managed to demonstrate that in 2014 and basically I’m showing you uh the gist of that work here if you were to uh look at a uh cover glass uh that is maybe functionalized uh in your icecat microscope in reflection then you would see something that is kind of speckly the contrast here is only of theor of 1% it’s not very large but it’s speckly and speckly because you’re sending laser light it’s coherent light and your surface is not perfectly atomically flat so it has slight roughness a slight index of refraction modulation and that leads to a speckle type of uh background uh that makes detection of very small particles on top of this very difficult because you can’t tell whether one of these is a particle or is just a speckled but the point is that the speckle is actually a pretty constant background it’s a property of the substrate doesn’t change with time which means that if you were to take record two images consecutively of the same object and subtract them you could mask the whole thing and flatten your field now if between these two uh uh frames proteins are added we can uh re get to the signals from Individual proteins just by a simple uh subtraction now the process is short noise limited which means that uh if you want to detect very small signals you have to integrate long enough uh so if your protein is large you can do a single frame subtraction and if your protein is very small then you have to batch some of these frames for example 100 frames and then subtract it from the next 00 frames and uh to detect smaller uh proteins but that’s a slight kind of technical detail the point is that in a very simple manner you can get S you can be sensitive enough to detect scattering from Individual uh proteins and this is now very interesting because it gives you a single protein kind of sensitivity using light uh and without using fluorescence um here example of detecting large proteins I mean these are still actually small proteins but uh uh the largest that I’m showing you here 340 kilodalton uh and then 80 150 and 65 kilodalton and you see here that um the shadow here is quite strong and the shadow here is kind of weak and that’s because the strength of the signal was proportional to the polar’s ability and as your particle become smaller the polar visibility becomes uh weaker what was interesting was that uh we could show that uh the uh uh signal that we’re getting is essentially linearly proportional to the molecular weight so the molecular weight is telling you how many basically atoms uh there are what’s the mass of the protein and uh it turns out that because of this proportionality it means that the proportionality to the volume is translating into proportionality to mass which becomes very helpful because then you can now basically weigh proteins uh by doing this kind of an optical measurement so you’re not only counting how many proteins are coming but you’re also figuring out how large the protein is um more recently we showed that uh uh we can push the sensitivity down to proteins as small as 10K K doton and what we do here is basically the same experiment except that we’re using machine learning algorithms to uh pick out small signals on top of noisy backgrounds more efficiently than we could do with our simple uh background subtraction that I showed you if you’re interested in that you can consult this paper but I’m not going into details of that so you see now that you know our ladder is even much smoother and better than uh before again large protein 220 kilodalton down to something of the 10 kilodalton and they all line up quite nicely as a as a line um now this method is uh now known even as mass photometry uh this uh was uh uh the result of hard work of my uh uh former postto Philip kukura who is a professor at Oxford University and uh he and his team uh pushed this to show the first time uh around 2018 that this linearity really works as a way as a kind of a ladder to measure uh protein mass and in particular it’s very interesting to uh show that you can uh tell the difference between a protein and a protein dimer and a protein trimer and tetramer and uh the point that for example I Didn’t Know Myself when we started this research is that when you buy a protein sample and dilute it and diluted uh it’s never completely always single proteins but these proteins like to aggregate in different ways and uh people who do mass spectrometry and doing protein research actually quite often are bothered by this because they don’t know exactly what’s happening and this method now allows us to learn about proteins in their native environment without having to take them into vacuum and doing Mass spectrometry depend on the nature of the protein it’s just a matter of the size that’s basically to the accuracy that we have right now seems to be the case at some point these things must must break down but uh at the accuracy of a few percent that we have this seems to be the case and also the other question that usually people ask and it’s a good question is the shape of the protein because that goes into the polarizability in principle and all of that doesn’t seem to matter that much right now although uh uh we haven’t looked at every single protein yet so this could come out but you need a little more uh uh let’s say Precision in terms of making those measurements yes within the object that’s being imaged can one utilize the fact that there are different frequencies that are interfering to resolve absolutely so this is all done uh very very off resonant so we’re using blue or green light whereas the first absorption of the protein is in the UV at 200 something or 300 something if you use the uh Resonance of the molecule then uh you get sensitivity uh to that kind of the same way as I showed the single molecule absorption experiment right in the beginning there we were doing the experiment right at the absorption of that molecule uh the difficulty but this is I think completely viable is that doing these experiments in the UV is more challenging than doing them in the optical domain for the usual things of detectors and objectives and and all of that but you would have that Advantage Instead This to like a bomatic yes beam and get a similar advantage that’s right so we we showed that as well in that paper in 2011 you can uh you can look at at the same molecule at two different way links and that way you can do a little bit of background kind of uh suppression because of the property the spectral properties of your sample yes all right uh so again I mean this method is now commercialized uh there is a company called refin that uh sells these instruments and have visited many groups who own uh these devices um so what we did at about the same time was to uh try try to go more towards medical kind of fundamental medical research and uh we were trying to look at the very interesting uh problem of secretion so when you have a biological cell this biological cell is constantly breathing and doing its work and part of what it does is communicate with its environment and this communication is a chemical exchange so in particular a lot of cells secrete proteins and vesicles uh constantly and uh we wanted to uh look at the secretion of these proteins directly by putting a cell a single cell on the substrate and parking the field of view right next to it and now as the proteins secrete uh in real time we can watch them basically parachute down on the substrate uh right next to it and uh this movie here gives you a sense of what’s happening so every time that you see a shadow that’s a footprint of a single protein that is landing and you see that they come at different times and different places uh this is just a little field of view uh at a few microns away from the uh cell by so by doing this you get uh a sense of what the uh cell is doing over a very long time and if we wait long enough you will see here that uh Hell Breaks Loose if the movie goes long enough uh at this point the cell lied that means it kind of exploded uh because we were treating it in terms of uh pH a little badly and uh uh but this was something like 30 minutes afterwards so over a period of 30 minutes we had precise information about how many proteins were coming out at uh what time now you can take the proteins that are coming out because you can measure the contrast and then you know that the contrast is related to molecular weight you basically can get a histogram of the type of things that are coming out so this was the histogram that we saw we knew from uh the literature that this protein that this cell was uh secreting a protein called IG and the IG contribution is this much but there was more stuff that was uh being secreted and we don’t we didn’t know at the time what all of this was but we could tell that there is that much stuff at that kind of a mass distribution that was coming up so this was a very simple demonstration a few years ago we’re pursuing this type of thing but uh this was at the single protein level it turns out uh so one of the fun things about doing this type of thing of as a physicist kind of getting into a field that you’re not qualified to work on is that you’re constantly learning and uh and uh so one of the things that we learned at some point was that there are these uh there’s a zoo of objects that are being secreted so not just single proteins but proteins that are encapsulated in vesicles so vesicles are or liposomes are uh objects like this so you have a lipid kind of Bayer kind of like a soap bubble and uh this could uh be empty but usually in in out of the cell this is never empty but it could have proteins DNA RNA inside but it could also have proteins that are uh uh embedded in the membrane and then you could have a so-call lipoprotein so proteins and lipids all mixed up with each other uh you could have all kinds of objects and of course virus is one of them as well so all of these things could be secreted and they’re nanop particles and uh in fact they’re quite large they’re usually in the range of a few tens of nanometers to a few hundreds of nanometers so we realize that all these years we could have detected these things extremely sensitively better than anyone else could do so at some point we uh actually did this and uh this paper came out a couple of years ago uh and uh it is showing something that is a spoof on an existing method and the existing method is called uh nanoparticle tracking analysis this uses darkfield microscopy so you take a suspension that has nanop particles in it and you illuminate it uh for example like that and uh these these nanop particles scatter light and uh you collect the scattered light at an angle and uh then you can see how individual nanop particles are jiggling around uh because you’re seeing individual nanoparticles uh you can uh extract the trajectories so when you have the trajectory of them you can do an analysis on that trajectory and figure out the diffusion constant the diffusion constant tells you uh basically something about the Brownian motion of this object and the diffusion constant is related to the size of the particle according to this formula so this is boltzman constant this is temperature this is viscosity that’s pi and that’s d d is the size of the particle so by uh doing this type of measurements figuring out D from the trajectories you can figure out the size of the particle uh this is the commercial device that you can purchase as well and uh is quite helpful but again because it does dark field microscopy its signal is proportional to the square of V or six power of uh of of the of the particle diameter and it’s dynamic range is not very good so usually you’re not very sensitive to smaller particles what we have done is basically done the same experiment except with ice scat instead of dark field so we’re doing the same thing uh but now we illuminate and then detect in reflection uh and this is the kind of an interference uh uh pattern that we get and uh again we establish the trajectories and uh uh figure out the diffusion constant but now in addition to the diffusion constant we also have the strength of the uh ice cat signal which tells us the polarizability and that is also related to the size of the uh particle so we get two independent measures of the size of the particle and we plot them in this kind of a scatter plot so this is derived from diffusion and this is derived from the ice Gat signal now in this experiment we had mixed uh three different types of nanop particles uh 15 nanometer 20 nanometer and 30 nanometer gold nanop particles in water and you can see that we find three different popul ations in this kind of a scatter plot now if you were to put the same sample through NTA which use dark field this is the kind of a signal that you would get so you’re not resolving these uh three and uh some of you might be familiar with Dynamic light scattering which is also a scattering uh experiment but it’s an ensemble experiment and does a statistical analysis of the light that is scattered and this is the signal that you would get from that again you’re not resolving the three uh populations in this me so this is right now very exciting and we’re trying to uh uh to uh extend this to all kinds of different applications these are other examples of how we demonstrated kind of the power of this technique going to the size regime and a combination of sizes that are not accessible to uh previous methods um to close this part of kind of more biologically inspired work I show you now how we go away from uh detecting isolated nanoparticles like proteins or vesicles or viruses to really a crowded region of a cell of a live cell uh for a long time we were bothered by the fact that if you take a cell and illuminate it with coherent light you get a lot of speckle uh you you see almost nothing because it’s full of Speckles and of course in the cell there’s also a lot of Dynamics which means that you get a dynamic speckle uh that looks like a nightmare what we realized at some point was that if we do our experiment in the confocal mode uh we actually don’t have this problem and here is now a movie that I’m going to run it’s a confocal icecat image of a cell and the structures that you see here are the so-called endoplasmic reticula these are Nano tubes if you want of about 100 nanometer uh uh diameter they’re soft kind of tubular things that are Dynamic they’re constantly being generated and ripped apart and moving and everything and uh because they’re so kind of scattering weak usually in a microscope you don’t see them and people do fluorescence microscopy to be able to visualize all these things and now you can see that a simple kind of scattering microscope can visualize all of this and uh so you can see the Dynamics of all of these guys but you also of course see other particles that are moving around uh these are larger particles or agglomerates and some of them are smaller vesicles that are moving around sometimes these vesicles are carrying proteins uh and so there’s a whole kind of Rich uh landscape that is very exciting uh to investigate uh to kind of bring it back to the pandemic and the virus situation here is an example where we are looking at this is now a big program that we have in in our lab trying to understand the virus cell life cycle because it turns out that as I mean maybe you also noticed that during the pandemic we actually didn’t know much about viruses although viruses have been around for years and years and virologists have been working on them and you quickly found a vaccine but we still don’t know how a typical virus comes and interacts with the cell and how it goes in and how it’s reproduced and comes out how it infects the neighboring cell so a lot of these steps are actually still unknown and one of the things that we would like to do is we want to through microscopy uh learn more about what’s happening so this cell was infected and these are viruses that have been generated in the infected cell we infected them and uh what I’m just going to show you uh briefly is how these viruses are now being moved around with the so-called actin tail and by visualizing these viruses now without the need of fluorescence in this kind of a label-free microscope we are able to look at the development of the cell and of infection over long time scales and with the kind of ision that has not been accessible uh before because if you want to do this type of experiment with fluoresence you have to genetically manipulate your sample and your fluoresence gets bleached and all kinds of difficulties that you have but we can do this type of experiment over hours or over days on the same cell and really Chase every single virus and see where it begins and what happens to it uh at the end so we’re very excited about this so uh I would like to thank all all the people who’ve been working on this uh project from the very very first uh ice cat signal that we saw uh when we were still at the University of constance the people who worked in Zurich uh some of them are uh uh are leading their own groups by now and uh the people who are working in the lab uh today so this was the part that uh had to do with this biologically motivated uh work and microscopy uh the physical setup is very simple I encourage you to test this you can do this also in a in a student lab uh it’s very simple to get the first signal if you want to get to very high-end type of analysis the actual image processing becomes a little more complicated but we’ve we’re uh building an open source platform and trying to put as many of the things that we know or basically everything that we know we somehow formulate them so that uh others can use them and if you have questions you’re always welcome to contact us now I started the talk with a single molecule but that was a molecule that was at room temperature and molecules at room temperature are uh not that uh they’re they’re they’re still Quantum objects but they don’t they lose a lot of their quantumness because of the uh interaction with phonons uh but we can go back to that fundamental question question of what happens when you illuminate light on a single molecule uh I told you that the cross-section is of the order of Lambda Square ID two uh now it turns out that we showed in this theoretical paper a long time ago that uh really if you were to mod match the incoming Beam with the beam that is or with the light that is scattered by the molecule then you can get 100% Extinction from a single molecule okay and again that kind of makes sense because I told you the cross-section is of the of Lambda Square divid by two and you also know that you can focus light down to something of the of Lambda Square divided by two so from that Sigma over a argument it makes sense that the effect of a single molecule or single atom should actually be huge right not that 10 to minus 6 that I told you that was because it was at room temperature but if you go to a vacuum or you go to an that is completely unperturbed by its environment you should be able to get essentially 100% but for that you have to match the incoming mode so that your interference becomes uh perfect so anyhow so that’s from a theoretical point of view so it’s not a ridiculous thing to do and uh we started doing this kind of type of work uh uh now almost 20 years ago um and to do that we have to go to low temperatures uh so we’re doing this work at about one and a half Kelvin because at that temperature these molecules behave like isolated single atoms the interaction with the Matrix in a kind of a hand waving way is frozen out and uh you get really down to a cross-section that is comparable to Lambda Square divided by two and this is really the kind of in a simplified way this is the experiment that we’re doing we’re uh uh building our own micros score objective in a chry stat because you can’t take standard microscope objective and put it in the CH that it kind of breaks uh but you can make a simple High NAA objective by taking a a solid immersion lens kind of half ball lens and a simple a sphere and put them together and now you have a high na uh objective and we prepar the molecules on the flat surface and then you have another a spere to collect the light and if you now uh uh tune the trans the frequency of the laser through the transition of the molecule you see that when the molecule is resonant where the laser is resonant with the molecule you get something of the of 10% Extinction so it’s not 100% yet but it’s much better than 10 to- 6 that I told you before and uh I’ll tell you in a second why it’s not 100% uh but this was very exciting it was the first time that you know someone showed that uh Extinction from a single Quantum emitter can be really that large and we’ve kind of exploited that effect in a number of studies that I won’t talk about because they’re no kind of older um so this is the gist of it now if you think about the molecule the molecule is not a two-level system but it has vibration levels for the ground state and for the excited state and that’s where fluoresence comes from right so you excite the molecule from this level to this level that’s what we usually do but when it decays it can Decay via the same Channel or via these other channels which have a lower uh wave uh um lower frequency or larger wavelengths so they’re red shifted this is the fluorescence channel that one usually detects but we showed that you can also detect it by doing uh basic Extinction so this is the extinction signal that I showed you and simultaneously we can record uh these photons and then you get the fluorescence signal so a little bit about this system this is a very special system molecular system uh this is a family of uh uh so-called polycyclic aromatic hydrocarbons uh these are basically kind of if you want a little bit like uh you take a sheet of graphine and you go very carefully go around and cut uh with atomic Precision this is what you get these are all all the corners are uh carbon and then extended with a hydrogen uh passivated with hydrogen all around but depending on the shape of the molecule the transition is shifted a little bit somewhere in the visible or near infrared so these are the actual molecules that we use now we embed them in a molecular Crystal that is usually made of these molecules and the molecular Crystal has a band Gap that is more in the UV so it doesn’t absorb so it’s transparent for visible light but these molecules have their transition somewhere in the visible so we’re talking to these guys and this is just a matrix to hold the molecule uh now the uh interesting thing is that when you go to temperatures below 2 Kelvin these transition the transitions of these molecules become extremely narrow uh they become as narrow as a few tens of megahertz you are probably not familiar with uh with with these units as much but it’s the same kind of trans bandwidth that you have if you had an Alkali atom suspended in vacuum so although the molecules embedded in a molecular crystal in a matrix in a solid its interaction with the environment is essentially frozen out to be no longer important so you really see the intrinsic transition from ground state to the excited state that’s basically what this uh line width is telling you and this is a so-called Fier limited line which means that the line width is really just the inverse of the lifetime that you would get that means there is no extra def phasing that comes in to to uh to broaden the line all right uh so there there you know work on molecules has been very uh uh slow uh we have been probably the kind of most persistent group over the past 20 25 years to try to do Quantum Optics with molecules but there is now a kind of a Revival and people are finding it interesting and uh lots of people are working in uh from different directions so uh if you’re interested or curious about that uh we have a review article um that was published a couple years ago so I can recommend reading that um so that generally the the situation is as follows you have these molecules that are embedded in a crystal but uh they’re not all completely identical because the crystal is never a perfect Crystal so you can think of the crystal having slight defects or slight variations and that means that the frequencies of individual molecules are slightly shifted from each other and that’s what gives us the ability to detect single molecules because although for example your laser beam would be talking to many of these molecules or shining on many of these molecules at any given frequency of your laser only one of them is resonant and as you tune the frequency a little bit you see strong lines popping up you have a question can you see Cooperative effects right I’ll come to Cooperative effects at the very end of the talk if I uh get there um I’m been very slow so I probably have to wrap up in a few minutes yeah I’ll try to do it five um so if you zoom into one of these things then you see that there is a there is this narrow line that I told you about but you also see that from a kind of a data storage point of view uh you have an incredible uh kind of capacity there because you have all these molecules that you can uh have within a very small frequency range uh these are only a few hundreds of gigahertz apart um all right so um I told you about this uh situation before and the fact that we go up on this transition and then uh but when you’re up there you can come down like this but you can also come down like that and this gives us the fluorescence Channel but as far as the coherent interaction with the with light is concerned these channels are not good because these levels uh have a very short Lifetime and these levels are strongly coupled to The Matrix this level is not but these levels are strongly coupled to The Matrix and that means that this channel spoils the coherence of my interaction um and the so-called branching ratio is about 30 to 50% so it’s not that bad but it’s not 100% that means that if I’m up here about 30% of the time I come down like this but about 70% of the time I come down like that and that 70% kind of spoils the game that’s why I don’t have that 100% chance to have 100% uh interaction or scattering efficiency um now the way to go about this to fix the problem is to enhance this transition so this transition becomes faster doesn’t it happens before this gets a chance to do its job and the way you enhance a transition is to uh put your emitter in a cavity uh this is something that has been uh field of research so-called cavity Quantum electrodynamics uh for the past 40 years or so uh and uh the essence of it is that if you put a photon here you bounce the photon back and forth and every time that it goes around it gets a chance to interact with the molecule uh now if it goes around by basically finesse times you get an enhancement by the cavity finesse uh so the figure of Merit that you have for the enhancement is the Finesse but then you also usually lose because the mode area in a typical cavity is larger than you would have if you were to focus in a defraction limited microscope so uh but as long as this factor is larger than one you’re winning in terms of enhancement and this Factor you can rewrite and you can actually show that it’s the same thing as Q over V which is a more common way of understanding uh cavity enhancement or percel effect and such um it’s also related to cooperativity uh but basically what you want to do is you want to have a cavity that uh and and and if you increase Q or lower volume you are uh doing well um so this is kind of the name of the game that we uh wanted to play about 15 years ago when we started this project uh we wanted to have a cavity that has a very small radius of curvature uh on one side in its mirror so that we can make it very small uh to make that mode volume very small and then uh you would have a flat mirror on one side so they can put a small Crystal with these molecules in it and this would be our cavity and then you just use a lens to couple into it and and all that um so the way we go about this is we can take a uh an optical fiber or the end of an AFM tip and uh we use focused IMB Milling to make micro mirrors uh uh at the end of it this is the kind of a profile that you would get and then you can code this uh in a multi- layer way this is what we do at a company and uh this gives you a very high reflectivity and uh Now by bringing this tiny mirror with a radius of curvature of only a few microns uh to within only a micron or two of the substrate we make a micro cavity that has a very large uh finesse because it has a low uh volume and uh this is some of the early experiments that uh that we did this was published in 2019 uh this is the Resonance of a cavity like that uh in transmission and reflection now if you tune the Resonance of the cavity to one of the molecules that is in that Crystal then you see that the uh resonance changes completely so you’re burning kind of a hole right in the middle of the transmission because there’s one single molecule uh in the cavity in a kind of a simplest way of thinking about this is that you’ve changed the resonance condition of the cavity because you put one molecule in it and usually you change the resonance by putting something of a certain index you shift the resonance so the index of of a single molecule if you want is large enough to kind of change the resonance condition or you can think of it as an Extinction of a single molecule now that is being enhanced by this cavity and now you’re getting this almost 100% Extinction from a uh single molecule um what we could also do was to show that we could replace the laser by a single Photon stream uh so that instead of sending lasers onto a single molecule in a cavity you send individual photons onto it uh I’ll skip this part because I want to get to the uh more interesting end without keeping you too long um so and and you might know that in this uh cavity uh QED type of experiment you have a so-called weak coupling regime where you have this percel enhancement that I told you about but you can also go to a strong coupling regime that’s where if the situation like a two like two coupled pendula and now the two resonators one of them is the molecule the other one is the cavity each one of them has a resonance that means that if can store the photon in it but now if the two kind of storage situations capacities become comparable the photon can uh bounce in the cavity come back to the molecule and go to the the cavity and come back to the molecule so in the weak coupling regime the photon comes interacts with the molecule and leaves again but in the strong coupling regime the photon can bounce back and forth many times and uh give this energy to the cavity and come back to the molecule and as it does that in the time domain you begin to see oscillations and we showed uh few years after the first work but by now already three years ago that we can also enter this strong coupling uh regime with a single molecule um now if you do that you’re in a regime where you have a very strong nonlinearity so nonlinear Optics is of course very interesting for all kinds of reasons but you also know that non Optical nonlinearity is a pretty weak effect it’s a higher order effect so usually you need a large Crystal you need a lot of intensity to be able to see much but in fact a single atom or a single molecule is a pretty strong nonlinear uh uh system but if you just put a single molecule in your laser beam usually doesn’t do much but in this case a single molecule is going to do a complete switch and uh to demonstrate that uh what we’re showing here is that you have a molecule in the cavity as I just showed you and now I’m having two beams two laser beams that are detuned from each other only by about 300 mahz uh so one of them is completely res resonant with the molecule in the middle the other one is a little off and if I were to look at the resonance if if the so-called the one that is off I call Pump uh is off then I see this but if I turn the pump on then I change the transmission of the uh system so the transmission goes from uh zero all the way to one so by adding a second laser beam I’m changing the transmission of the first laser beam and that’s basically an optical uh switch that I have and the interesting thing is that this is happening when there’s only on average one Photon in the pump so the pump is called pump but it’s really not strong at all it’s just a single Photon kind of beam that uh I have so uh this is all very interesting because it really gets you to a regime where you can play with single photons and single molecules in a very very efficient way there are lots of kind of uh uh detail subtle issues uh that prevent us from immediately turning this into some kind of a device but uh to kind of bring this this far this has been very exciting to have that much control over the interaction of single molecules and single photons now uh the last thing that I want to tell you but I’m going to cut it very short is really just question of cooperativity uh basically doing work on single molecules and single photons was very interesting uh it occupied lots of groups and lots of systems for the past 20 years people have worked on single Quantum dots single atoms single color centers single molecules uh different ways of generating single photons but I would say that there although there’s still there’s still a lot to be done the field is pretty mature uh you’re not going to impress too many people if you have another single Photon Source uh but getting scaling up is kind of the bottleneck or the next challenge is how to have many single Photon sources or many single Quantum emitters and to make them interact with each other very efficiently to build something like a quantum network that actually Works efficiently this is a very interesting field of research but also very difficult uh to do part of it has to do with the fact that already one Photon and one Quantum emitter it’s difficult to get them to interact very efficiently and if you want to build a network you lose exponentially if you have losses at every node uh nevertheless you know you try hard you try to kind of develop the technology and uh the kind of things that that I showed you would be the building blocks and now you have to kind of put the building blocks uh together one way to do this is to put it on a chip so that you can scale up uh so we’re now building kind of uh a photonic circuits uh this would be for example a micro disk resonator with a certain resonance and then you have different wave guides to couple light in and out and then you can put molecules uh on top uh preparing the samples and all of that is always a challenge because you have different materials and Material Science always but so far has always killed everyone uh you know if you talk about Quantum dots or color centers they’re all killed at the End by Material Science issues uh it’s also a big problem for us because we’re putting Organics on inorganic kind of circuits and there are lots of issues that you have to solve it’s not a dead end but it’s a very slow uh process because physicists don’t know Material Science and uh so uh you put molecules and try to couple evanescently and for example you know again you see that there is a single molecule coupled to the Resonance of this uh micro resonator and now you can put micro electrodes and uh tune the frequencies of different molecules so that you can bring them to the same frequency so that they can talk to each other via the mode of the uh Optical resonator and uh this is the kind of thing that you would do you in this case you see that you have lots of molecules coupled to the microresonator you take two of them and then by tuning them you can bring them somewhere around here to have the same frequency and we showed in this work uh yeah it’s already appeared this is an old slide appeared a couple of months ago in Optica uh and we go through the details of how two individual molecules are identified and then brought together in frequency and you can show that they inter the light that it scatters from each one of them actually interferes constructively in the forward Direction and destructive in the backward Direction and that kind of thing um and the particularly exciting work that I’m not going to tell you anything about uh today is uh something that we’re trying to uh write up uh but it’s done in this geometry that I told you about before so in the open micro cavity and there we also uh getting to a situation where light that is scattered from one molecule is going to be interacting with another molecule that is in the common kind of in the same mode uh and I just want to uh flash two uh fairly recent uh papers that I’m not going to have time to talk about if you’re interested in more nearfield type of uh experiments uh this we think is an exciting uh tip this kind of arrangement to bring a Quantum Omer for example in the near field of an optical antenna and uh this is work that was published just a couple of weeks ago uh where we show that in a kind of a fluidic Arrangement you can use concepts of optical antennas to collect more photons and therefore do more uh more efficient uh uh single Photon single molecule type of uh experiment and and with that I would like to also thank the people who are working in the quantum Optics part of the group right now lots of people have contributed in the past they wouldn’t have fit in one slide so I don’t show them uh but these are the people who are currently working in the group and we are would love to hear from you if you’re interested in doing postto work in any of these different uh directions uh in at The Institute and with that I would like to thank you for listening i’ be happy to answer any questions yeah thank you very much for this very fascinating talk um I think we still have time for one or two questions and the speaker will also be around later before we take the questions I should also mention that there is an information uh session tomorrow about um you know studying in Germany especially at the plank Institute for the science of light and this is uh tomorrow at um 12 I think in uh the large lecture hall and pizza will be provided so okay so I think we have time for one or two questions no questions all questions answered can I have a quick question you have a question I have a question about the uh molecule experiment that people are showing by the way this is super cool so uh were the molecules actually sitting on top of the reservat in that case or in the first case they’re sitting on top of the resolor on top and next to it so you have a you have a disc uh the micro disk uh of titanium dioxide on a glass substrate and then on that we grow an organic Crystal and that organic Crystal includes the molecules so they’re coming close to the boundaries of the resonator I’ll ask my question thank you what are the spin properties of these molecules do they have any yeah so these molecules don’t have any spin uh which is you know again a material problem uh so the the the I mean there are molecules that have spin uh they’re spectral properties are not as nice as these ones and so far we haven’t worried about spin there are a few people who are trying to push that uh the molecules have a triplet state which would have a spin but it’s not in the ground state uh so color centers have nice situation that the ground state is a triplet State uh but here we are looking at a few different situations where triplet give us gives us spin properties but those are normal triplet states that are at a higher frequency okay so then let’s thank our speaker again so we have this certificate for you um and and we thank you again for this very nice talk thank you very much thank you thank you