 So, thank you all for coming, today we have a really great from Austria Institute, one of the leading scientists in Kohl Assens, he started his career in Heidelberg, where he was an associate professor and then in 2000 he went to Austria Institute, who is a food professor and today he is the director of the institute I think, right? And he is one of the big experimentalists in Heidelberg-Kohl Assens as I said, so because of that he has won many awards, he has won the Wittgenstein awards, he has been an Austrian of the year in 2005, he has been a scientist of the state of Tugut and also an Austrian scientist of the year 2009, he has won the Fadayev Medal for his work on tube body physics in Tugut States and last year he won the BSE Advanced Brand and he has also won the BEC Senior Award, which is an award that has given out every second year in the field of Kohl Assens to one of the leading scientists in the field and we got it last year. So this research has been one of the pioneers in BEC physics, he created the first season of BEC, he created the first biology of BEC together with David Gin, he has also worked a lot on ultracorneatomic Fermi gases, strongly impacting Fermi gases, so this whole business about BEC, BCS crossover, he was one of the guys together with Ketterdi and David Gin, doing a lot of collector modes and radio frequency spectroscopy which today we now think is a standard method to inspire me at those. He also, as I said, won the Fadayev Award, which was given due to his work, I guess, on the chemo physics. So these are, I don't know whether we have any of the chemo guys here in the audience, but these are free body states, so even if you don't have a two body bound states, you can have three body bound states and those will be picked by the chemo And for the first scene, polypility involved atomic gases, by Boudic. And now he's very active in impurity physics, as we also are here, so he works mostly on the Fermi polynomial, and I guess he's going to tell us about that today. And also he will tell us about his new directions into exotic superconductors, I think, new superconducting phases with new Fermi-Fermi mixtures and Boudic Fermi mixtures. So let's say welcome to Boudic, and let's see his voice. It's so great to be here. Thanks a lot for inviting me to this colloquium. I really enjoyed coming to Denmark. In particular, when I crossed the border yesterday with training, I could take off the mask. It really feels differently, so sometimes I get nervous because I don't have a mask here, but I think that's not a problem here. Okay, so you see a little impression of Innsbruck, a special place in the mountains. I have a winter background slide, which is this one. It's no longer realistic now. I mean, now it's still there on top of the mountains, but not in the valley. I also have a summer slide, so this is a transition that we are now making to winter to summer. All time of the year, it's a nice place to visit. And I'm at two institutions, the University of Innsbruck, and at the ICOFA Institute of Quantum Optics, the Quantum Information of the Austin Academy of Sciences. So, okay, how to start? Maybe I start with looking back into history, 20 years, roughly 20 years, and one of the really very important experiments which changed the game, which made a new era to begin. And that's an experiment that's by the John Thomas group carried out at Duke University, and this is an optical dipotrap. It's just a laser beam, an infrared laser beam forming an optical dipotrap, trapping a spin mixture of lithium-6 atoms. The lithium-6 atoms are here, a spin mixture of two spin states, and what is very important, there is a Feshbach resonance, magnetically field-dependent resonance, which is shown here. They put the field to 800-something Gauss, where the interaction is very, very strong, and the S-way is scattering like diverges. So we have interaction control in the experiments using this Feshbach resonance phenomenon. And here it was very important, and what they observed then looked like this after there is a time of light expansion, and immediately after the expansion you see almost the initial shape, the cigar shape, but then you see that the crowd does not expand ballisticly, it would just be round at the end, now it expands hydrodynamically in the sample. And that was very exciting because if you work with, with the interaction in Bose-Einstein conversations, you know, I mean, hydrodynamic expansion is one of the signatures of Bose-Einstein conversations. Here it's actually not, but it's still a precursor of some kind of superfluous state, it's strongly connected to the fact that it's strongly interacting with the Fermi-Gauss. And I want to measure, mention another experiment from 2003, 19 years ago, there is a molecule formation in the Fermi-Gauss. So that's an experiment by Debbie Chin, an integral at Gila and Boulder, and she worked with a spin mixture of potassium atoms, different spin states. And, okay, this picture here shows just the Stangana separation of the crowd into these two different spin states. So when they cross the Feshbach resonance, they see the signal gets weaker, some atoms get lost, but they are not completely lost. If you apply a radiofrequency pulse with the right frequency, then you see something else happening. And you see something is coming back, actually, not at the location of these atomic blocks, there's somewhere, something in the center. And this is a cloud of dimers formed by Feshbach association, which are then dissociated by the radiofrequency, which are imageless atoms. So this is the signature of molecule formation, Feshbach molecules formed in the Fermi-Gauss. And this is very important and exciting because what you see is if you put two fermions together and you form a composite particle, it's obviously you get a boson. So it changes the quantum statistics. Then you can ask, okay, can these bosons condense? Yes, the answer was given just a few months later, and they can condense. We observed at the end of 2003 that our super shallow dipole term could contain many more atoms than it offers quantum states to the fermionic atoms. Something else is happening. We got 10 times more particles into the trap than the fermionic nature would allow. So we interpreted this as both Einstein's compensation, and this interpretation was correct. A little bit later we saw also bimodal behavior in the same way that also David Gin and the group had seen here, bimodality in their profile and the catalytic group at MIT. And this was the starting point actually of many, many experiments over almost two decades on BC-BCS crossover physics. The system, depending on the magnetic detuning, can have both Einstein character and the other side of the resonance BCS type character. And in the center you get a so-called unitary form gas with the strongest interaction that is allowed by quantum mechanics. Okay, and then another slide on superfluidity. Later, a couple of experiments have demonstrated superfluidity. We had seen some hints in collective modes. But the experiment with pair condensation at Gila in 2004 and vortex formation at MIT in 2005 made it crystal clear. And we could also contribute a few observations related to the superfluid nature, like the quenching of the moment of inertia. You see that the cloud, which if you put the cloud in the slow rotation, no vortices are formed, but you can measure the angular momentum of the cloud. And then you will find that the superfluid part does not carry angular momentum, cannot carry angular momentum unless you form vortices. And so you observe a quenching, a reduction of the moment of inertia below the superfluid transition. And another experiment that was actually carried out in Innsbruck with the collaboration with Sandro Sengali and Lev Bitaevsky from Trento was the observation of second sound that in a superfluid you have an additional sound mode, like here is the second sound, which disappears as soon as the sample is no normal superfluid. Okay, so this is a brief history, some important milestones. I'll come back to a few things a little bit later. I just want to mention one more thing. Why with thermi-gases? There's one important ingredient, and this is the collision and stability, which they can have. Usually our enemy, if we work with strongly interacting quantum gases, is three-body decay. So two of them form a dimer. The atom is there to carry away the binding energy. This is the three-body process, three atoms, together one dimer and one atom, and a lot of binding energy is released. Exactly this process can be very strongly suppressed if you work with fermions, because in a three-body process, in a two-state spin mixture, you inevitably have two identical fermions involved. And this suppresses the three-body decay. In particular, if you have so-called broad-fetched resonance, this suppression effect, it can be extremely strong and make the situation very stable. And if you work, for example, with the resonant lithium-6 thermi-gases on top of the resonance, infinitely scattering energy, the system can be even more stable than maybe if you work with rubidium, the rubidium-Bosai-H. 9.0. So this effect is extremely important. So that means there's really a very important difference in the experiments. For robots, if you have resonant scattering properties, you usually have very fast decay. This can reveal also interesting physics, like three-body states. Also the group of yarn has to consider the three-body decay in that regime. And in the case of fermions, they can be collisionally stable. At least scattering, inelastic scattering, can be suppressed by effect of 1,000 or something like this. And you can have long-lived states in your trap. Important difference. I come back to this a little bit later. Okay, now this was my brief historic review of some basic ideas in the field. Now it's time for the first part. So experiments on impurities in the Fermi scene. So we can work with potassium atoms, bosonic potassium-14, fermionic potassium-14, and bosonic potassium-41 in the Fermi scene of lithium-6. For the system of lithium-6, it's very similar to what I've introduced before, but now I have additional purity signal. That will be the next maybe 25 minutes of the talk, and then I switch to another topic, which is progress towards a new mixture, putting this boson and potassium atoms together, with the goal to create exotic superfluorism. I think we have made important progress towards this goal. Okay, coming to the impurities. Okay, so we now have a situation where we have a large firmacy of atoms. We have the blue ones here, and we have a few impurity atoms. And so these impurities in the Fermi scene, they have been calling Fermi-Polaron, going to work back from 2009, when the three line group at MIT. So we're looking at Fermi-Polaron's quasi-particles of impurities in the Fermi scene. Okay, so we don't consider a bosonic medium. It's done very nicely in laws. I don't speak about the boson-Polaron, I don't know much about this. So we are interested in impurities, and these can be fermions or bosons in the Fermi scene. Then there are a lot of research questions, interesting research questions. So, in general, what are the quasi-particle properties? Static and dynamic quasi-particle properties. Interaction energies, energy shifts. How stable are the quasi-particles? How are they formed? What is the formation dynamics? But also, what happens if these quasi-particles move? And what happens if you have enough quasi-particles so that they're dense enough so that immediate interactions can become important? And what is the role of the impurity quantum statistics? Well, these are some questions I will touch in the next 20 minutes. But here I want to introduce a team working in the lab. This is the lab in the background. Here members of the team, the current PhD students. Isabella is no longer a PhD student. She finished last year, staying a little bit longer in this book. Cosetta and Erich, PhD students. Adriana, master student. Postdoc O1, senior scientist Emil Kirilov, myself. And we work in the lab at the Akoku Institute. And our system is quite similar to what I've introduced already. We have an infrared optical dipole trap. Just a standard 1,064 nanometer infrared light. And we have a lithium-fermicine. And it's produced first by evaporative cooling in a spin mixture, but then one spin component is removed. And we have a single spin state. And this is the lowest one in the system. About 10 to the 5 atoms. And temperatures are around 150 nano-carobin, which is about 17% of the Fermi temperature. OK. And now we add some impurities. Here are the impurities. And the impurities can be potassium-40 or potassium-41 atoms. Potassium-40 are fermions. Potassium-41 are both impurities. And we have typically 10 to 20 times less impurity atoms than we have atoms in the pharmacy. And the concentration in the center, they sit in the center of the trap, is locally about between 10% or 50%. And we see what the influence is of these. Not zero concentration. OK. And a very important ingredient. We have tunable interaction. We have a fishbuff resonance. In both cases we find fishbuff resonances. They are a little bit different field, but the character is very much the same. So we can switch from boson to fermions and see what is the difference in quantum statistics of the impurities. The most simple method to probe the system is radiofrequency injection spectroscopy. So we look at two different spin states of the impurity atom. OK. We start in one state which is weakly interacting with the medium. Weakly interacting means usual scattering length, 60 times the bore radius or something like this, just enough to provide elastic collisions to thermalize. But if you want to look at strong interactions, it's extremely weak and it has almost negligible effect, the interaction. There is another state, and in this state, the system shows the fishbuff resonance. So we can tune the radiofrequency interaction to resonant and then the spectrum of excitation gets much broader, gets broader. And again, we are interested in the spectral function which we probe by radiofrequency transfer. So we inject from the non-interacting state into this tunable, strongly interacting state and see how the response depends on the interparticle interaction, the interspecies interaction. OK. So the next one, OK, some details for the experimentalists. The system is set in such a way that without the interaction we drive pipe pulses, we transfer all the population on this state and the other one. We use black man pulses which don't have side lobes. And the spectral resolution is something like 700 hertz, which is 4% of the thermalization. Good. The next one is also more technical. OK. This is just the spin states involved. I don't go to the details. The fishbuff resonance here sits for the bosons at 157 grams, for the fermions at 335 grams. And actually, I see that it's wrong. It's the other way around. This number should be here. The numbers should be interchanged. That doesn't matter really for the principle and the frequencies we're working with are close to 40 or 60 megahertz. But these are just some technical details. OK. Now I introduce the interaction parameter, the dimensionless quantity characterizing the interaction. We call it capital X. And it's minus 1 over the Fermi wave function times the scattering length, A. OK. What does it mean? The Fermi wave number is the inverse interparticle distance with some pre-factor of order 1. The Fermi energy we calculate in the usual way. Typically, these length scales, 1 over KF, corresponds to 4,000 times the bore values. OK. To make it strongly in direction, we have to get to these values of the scattering length. That's what we can do using our fishbuff resonance, which is an S wave scattering length, too magnetically tunable, and we find that under our conditions the strongly interacting regime is just 15 mini-gouves wide. Before the experimental is to control on the mini-gouves level on more than a few hundred gauss, it's already some kind of challenge, but I think we muster this in the experiment. Good. So here is one of the first results we got in the impurity system long ago, or almost 10 years now. And we looked at this injection spectrum. And then you see, if you go to the fishbuff resonance, here's the resonance. If you approach it from the lower side where the scattering length is positive, the energy is upshifted, and then somehow the signal fades out and disappears. And here for attractive interactions of negative a, the energy is shifted down and the signal disappears. What you see here is what we call the quasi particle, the polar roll, and then it's on top of incoherent excitations. If we look at the full spectrum, and later in experiments in 2016, we see a nice polar on peak here in the spectrum on the repulsive side and a broad pedestal of incoherent excitations of typical width of the Fermi energy. The same also for the attractive polar. Okay. And do we understand what we see here? Yes. Here I'm very grateful for our long-standing collaboration that we have with you here. Marcelian from Iqfo, Spain, started around this time on the polar on properties. And here the solid lines show the results of a T-matrix approach based on single particle hold excitations to calculate the polar on energy in the single impurity. Do you know what that fits? Quite well there might be some little deviations which also have a physical reason. What happens between the two dashed lines at the end there, is that the molecule hold continuum where the impurity atom can recombine with an atom from the Fermi sea and they form a molecule that can happen in a broad spectrum which we investigate in detail in other experiments. I don't want to discuss it anymore. But we understand this disappearance of the attractive polar on peak here. We understand because it crosses into the molecule hold continuum and the polar has become unstable and decay into molecules. Here for the so-called repulsive polar on it's a different decay making mechanism but it also decays into at the end the molecular excitations down there. Good, so we understand basically what we see there. We did many more experiments but now we want to come to a question in the Fermi sea what will change if I leave the Fermi sea the same as it was before, in lithium-6 but if I change the impurity character from fermion in potassium-40 to boson in potassium-41 what will change? What do you think will change? Okay, let me see. So this is what I've shown before the fermion case. Now with the same method I show we took a spectrum on the bosonic impurities I mean then the color coding the color map is a little bit different but it's the same experiment that it looks like. The range is a little bit different and the colors are a little bit different but what you see the repulsive polar on here the attractive polar on crosses into the molecular whole continuum some excitation in the center how to see it in these graph but it's basically the same. But here we worked on the conditions where the bosonic impurities, the bosons they can condense but we worked on the thermal conditions slightly above the condensation temperature. Okay, so that means thermal impurities slightly above the condensation temperature bosonic nature we have almost the same as the fermion impurities to the extent we can see there but does nothing really change we look again to this spectrum the same graph I've shown before but now we change the temperature from 19% of the fermion temperature of nitrogen a little bit below to I think 16% or 15% or 14% and then we get a partially condensed impurity cloud and you see the spectrum well you still see the polaronic branches repulsive and attractive polar on but you see the digital branch with little energy shift and this apparently comes from the bosonic condensate that is formed so what is the situation here the situation so a new branch is that we have innermodular situation we have this fermiC in the trap okay and we have a cloud of thermal impurities in the center and in the very center even much smaller there's a BC and as you all know the BC has a much higher density about 30 or 40 times higher than the impurity cloud what happens in the center that they change role here we have a dense bosonic out with a few fermionic atoms in it and it's more the opposite situation it's the case of bosopolar ones basically but we probe still on the then on the majority which is different from the experiments here okay so a remark I have to make is that if you do this you create such a situation with large interspecies scattering length in a static situation what you would see is either a collapse of the cloud by attractive interaction or phase separation for repulsive interaction but we do this experiment fast enough so the collapse dynamics and phase separation is clearly slower so I think we are not effected or have only weak effect we study it separately okay that are the regions in the trap so it's an inhomogeneous situation where the branch in the center comes from these dense bosons the common state and the polar only branches comes from the thermal cloud which you see here the thermal cloud okay so we looked a little bit more into these bosopolar on question but it's just very rough experiments we see there's a little energy shift in the concentration there's the BC concentration it's much higher than what we have in the Fermigas and there is an energy shift which we can roughly explain in a kind of back-axle model looking at the energy interaction energy between the two species but this is just some rough demonstration of something that is probably related to the bosopolar on question we have to come back to this question maybe a little bit later we can see that the data scatter a lot we were quite affected also by fluctuations in the interaction parameter the exact magnetic field and also temperature okay but we didn't we haven't deeper into this question but now I'm going to come to another question how about induced interactions between the polarities now we have more impurities and so somehow they overlap and to what extent can I understand this in Landau's family with theory or are there phenomena beyond it okay as experimentalists what we do is we carry out experiments and we vary the impurity concentration by different loading times and so on trying to keep all other conditions constant but just changing the impurity concentration and measure energy shifts turned out to be quite hard to do this I mean you have to do many many experiments and there are different concentrations and extract them with the data but I can show some results but this is actually if we are in the repulsive regime already strongly interacting then we see if we see a clear energy shift here okay there's no doubt there's an energy shift with the impurity concentration it's 50% of the impurities and we find always for all measurements we do we can nicely fit it with linear behavior okay so we analyze it based on the linear behavior and one thing we can do is we can extrapolate to zero previous experiments were somewhere here maybe with the concentration and were probably somehow a little bit effected by density induced effects but here we can extrapolate to zero if we do it we get extrapolate that zero concentration energy shifts and we can compare it with theory and it fits perfectly okay but how if I do it a higher concentration somewhere let's say 0.4 let's go here okay let's take show the same experiment data for 0.4 concentration for variable interaction parameter then it looks like this you see there is a clear effect so the shift is less the energy shift is reduced also here it's less energy shift by these overlapping color ones so it's I can think one can state it's a clear observation of an energy shift mediated somehow by the interactions what exactly the mechanism is we don't know we have some ideas and another thing we can look at is the slope the linear slope we extracted the zero the value of zero concentration now we extract the slope okay and then for the slope we get this data here in the weakening interaction regime so these are only the weakening interaction data up to x minus plus minus one and you see there is an energy shift but it kind of is consistent with this is a theory provided to us by Pietro and then Georg on the applying Landau Fermi liquid theory in the weakening interacting regime and it kind of it seems to be consistent okay but we were of course looking into the strongly interacting regime and now I have to expand the scale a little bit then it looks like this for strongly interacting conditions that's the data said you know the shifts are much larger and this is still a big puzzle for us we discussed this today, this morning possible reasons so for us it's a big puzzle for this simple Landau Fermi liquid type approach looking at the other number of atoms in the dressing cloud and so on this does not fit here and something else is going on and one possible explanation might be that we discussed that here we have a molecular component and with this molecular component this leads to the strong upshift so there are some first indications that this would be consistent but we have to see this is work in progress but an exciting question good so a little bit into the future what are we going to do next and one open question is motion effects on the polar one what happens if I move the polar one through the cloud and what is the relevant velocity scale it would be the Fermi velocity of the medium so I can expect if I approach the Fermi velocity of the medium the ferments can no longer follow the dressing cloud and the picture should change that's actually a question I started to look into with Richard Schmidt a couple of years ago but now we are close to do many experiments like this so if you look at the Fermi speed it is at the Fermi energies we have it's typically something like 44 millimeter per second now how can we probe this and one nice tool would be Raman transitions so we can actually replace our we can replace our radio frequency transition with an optical Raman transition and if you do it with counter propagating beams you get a photon momentum transfer of two times the photon momentum you can do it with co-propagating beams no photon momentum then it should be exactly the same as for the radio frequency excitation or you can do it with counter-propagating beams then we have 2h per k now if you calculate what is the velocity change by 2h per k then you see it is 26 millimeters per second which is half of the Fermi speed so it should be relatively easy to do this experiment moving with half of the Fermi speed for the medium I would already expect quite substantial modifications or it's also possible to have an extended Raman scheme 4h per k then we have the Fermi speed and these are experiments we are preparing right now good and these are what I wanted to show just proof of principle I don't want to go through all these pictures it's just Rabi oscillations induced by Raman beams under different conditions and it works really very nicely and you see the nice contrast of the fringes of these Rabi oscillations it works very nicely and it works also very nicely if the push is to the limits we can actually look at the time scale this is a microsecond over here so we have a PIPALS duration here of 250 nanoseconds and this now allows us to enter a new regime where the PIPALS duration is shorter than the Fermi time Fermi energy time divided by h bar which is about 5 microseconds in our case previous experiments let me just mention it briefly on the Ramsey spectroscopy on this impurity cloud which we have carried out were limited by the finite duration of the pulses with very good frequency we can do maybe 25 microseconds or something like this that's what we can do but not below the Fermi time but here we can go 2 hours of magnitude faster and that's very interesting for a new generation of experiments using this Ramsey excitation scheme so that's some new developments going on in this experiment and now it's time just to conclude on the Fermi polaron I think we have quite a good understanding of static quasi-particle properties stability of the polarones the lifetime of the formation dynamics we start to understand this work in progress impurity-impurity interactions at least we have seen clear interactions we understand aspects of the impurity quantum statistics and we started experimenting with the emotional effects but there are more questions like for example are there few body effects influence in the system or another interesting topic no one has ever studied light impurities here we have heavy impurities in the firmacy of light atoms no one has studied the opposite system and I think we now have a new system with properties to do that but that's the next part of my talk okay but this was about the impurity physics and now in the remaining time I want to introduce a new system we are having a lot of fun with this new system mixing the disposing atoms with potassium atoms both phonionic okay so this is a spin mixture going back to lithium-6 or potassium-40 in a spin mixture so you can extend this so now a nice extension is you create imbalance you have more of one spin state than the other spin state bringing imbalance or you can also call it polarization and this of course introduces some new physics and has been extensively considered in experiments about 10 to 15 years ago and in extreme case of course is one impurity in the firmacy or lower purity in the firmacy it's a matter of taste okay but this is a situation of polarized pharmacosis that have been considered in quite many experiments in the lower theory and here is a phase diagram of these polarized Fermigas from Hengstof's work but this is a mean field approach which shows the qualitative features quantitatively I mean the numbers have to be changed or reduced by a factor of 2 to 3 but it shows at least what happens if there is no polarization spin mixture of course if you have zero temperature it's superfluid, if you increase the temperature there is transition to a normal so SF is superfluid and is normal but now let me introduce some spinning balance then at zero temperature I mean FR which does not mean fashionable resonance it means in this plot forbidden region it's a region where the system phase separates actually phase separates in a way that some atoms pair up they form pairs in the sample the excess number of particles unpaired particles then is expelled and they phase separate in the trap so it's an unstable situation only finite temperature can stabilize it again and then here is the so-called tri-critical point on the other side the same so this is symmetric I mean if you can spin mixture if you have more than one spin of the other spin it's a symmetric situation good so what do I want to tell you in a lot of experimental work one paper kind of summarizes many things by the catalogue group and this is the they map the phase boundaries such spinning balance and you see exactly here these what was FR before the unstable region the tri-critical point the superfluid phase and here is the normal phase and of course I mean because of the spin spin mixture it should be a symmetric situation so that's the phase diagram basically you see before and you will see again here okay this is now quite well understood there are a lot of papers, theory papers on looking for exotic phases exotic pairing phases and they find in certain regions there are some maybe very close to this point hard to see there might be regions where at very very low temperature you get other pairing phases where so F is a low type phases where particles pair up and don't get zero momentum but they are finite momentum so to say pairs with finite momentum this can be in both directions so traveling wave or standing wave affects F is a low effect but these are theory papers and these have not been observed experimentally the reason simply is you have to go to really super low temperature pairing phases may exist in the phase diagram for experimented as if you ask an experimented as can you produce an article called Fermigas at a temperature which is 1,000th of the Fermi temperature that would be rather pessimistic okay so it is very difficult to reach the spin mixture okay coming back to this illustration here we have the polarization of population imbalance but you can add more imbalance you can add a new degree of freedom mass imbalance so here you have more blue heavy ones than the other ones and yes the other way around you can play with the mass of the system and you can also have imbalance in both here is a situation where you have population imbalance and you have mass imbalance you have more heavy guys than light guys okay that's actually the particular interesting situation so okay how do we change the mass of an atom it's not so easy maybe you can play a little bit with effective masses and so on in the lattice but the most direct way is you just choose a not a spin mixture but a species mixture you put different elements together and then what can you expect and there is this phase diagram again you have seen it and this calculation shows the phase diagram now with a mass imbalance factor 7 in this case but it does not very sensitive to it should look like this on the right hand side positive p you see basically the same you forbid reach and try to get a point superfluid phase nothing very exciting but on the r side here it gets very exciting because now there is a so called lp standing for lift shift point and there is where the standard kind of cooper pairs become unstable towards finite momentum below this point somehow the interesting stuff is happening I mean those use SS for super solid phases but it's more in general phases broken symmetry like FFLO phases here so the interesting things would happen below this point that's where all theoreticians I think would agree there are no good predictions what would happen exactly where but that's the interesting stuff so here is the interesting region and exciting things happen now look at the temperature scale of course it's a matter of how you define the Fermi temperature for mass imbalance system but you see 0.2 and 0.2 does not sound very dramatic we can easily get 10% of the Fermi temperature so this is in a temperature range which we may access in experiments it looks promising but that's not the main motivation for the work on our species mixtures of dysplosium and potassium so you can say why dysplosium and potassium you need two different fermions eight different species have been brought to degeneracy in the last 20 years or so of course there are common ones lithium-6 and potassium-40 there are other ones 2087, comium-53 dysplosium-161, erbium-166 erbium isotopes which ones to choose so you need some criteria criteria you should have a decent mass ratio not too small of course if you have 1.1 it's almost one but if it's too large then again you can have your fume of states which make the system unstable so something around 5 may be nice so and then you have other criteria you would need systems where you can at least expect tunability of phosphor resonances this is not the case for petroleum because of the closed shell structure and so on and then just a few options are left and actually it turns out the obvious one, lithium-potassium we investigated this in more detail it has no broad-feshper resonances only narrow resonances and no good collision instability so this standard one does not work in Florence they investigated lithium and comium we were waiting for the results now they have understanding of the lithium-comium system unfortunately they also only narrow resonances at kilo-gram spheres but our choice was potassium plus dispersion and in these submerged shell atoms with an inner shell which is not closed many bizarre things can happen we just have to understand these would be obvious but it doesn't work for the superfoods but we have to understand the interaction properties no one there was absolutely no knowledge on this you have to start from scratch you don't know, you take full risk it can be very bad the interaction properties only losses of narrow resonances but we took the risk and we started experiments on that and it took quite some while a couple of years but now we don't understand it fully I think no one will ever understand it fully but we understand at least the interaction properties of the system before coming to that we introduced the team members these are the past team they built up the experiment carried out the first experiments now we have kind of a generation experiment and these guys here took over and now we are again after these generations change and after all the Covid time which slowed down everything we are again in a very productive phase okay so for the experimentedness I mean this cooling properties actually is prosium I mean the level scheme looks a little bit scary many many levels that's usually what laser coolers don't like but you have selection rules and you find a strong cooling transition here at 421 nanometer which can be used for slowing an atomic beam and then you can operate actually a magneto optic trap on this line 626 nanometer which has about 150 kilohertz line width and there are more narrow transitions which you can use for narrow line cooling so it's very easy to prepare the system using a 421 nanometer Zeeman slower 626 nanometer magneto optical trap then you get down to 10 micro cave or even a little bit less very convenient for loading an optical type of trap for potassium what we use in so called D1 sub Doppler cooling, gray molasses cooling which also produces similar temperatures and so the message is just if I don't go into detail the preparation of the system is quite easy the species like each other there are no major problems bringing them together and to create a mixture of a few micro cave is relatively easy that's already quite good news then how do we do further cooling actually we have a single dipole trap to the dipole trap infrared laser usually and the potassium atoms are much more polarizable than the disposal atoms disposal atoms have about 3.2 times less polarizability at this wavelength so it means the trap is just more shallow and then we have gravity and disposal is much heavier than potassium so we have a combination of two traps a relatively shallow trap for disposal where the atoms can escape and a deep trap for potassium that means in our case if we do evaporative cooling potassium would be the cooling agent and sorry the other way around disposal would be the cooling agent and the potassium atoms are just aesthetically cool and this works super well super nicely because we have dipolar collisions and so even we have a firm gas of disposal in a single spin state we usually learn in a single spin state we don't get any domestic collisions but this is not true for the strongly magnetic atoms because of the long range dipodipole interaction we always have a nice cross section for collisions and so evaporative cooling of disposal 161 in single spin state is as easy as it is for big Americans so this works very well and we get to nice conditions with disposal we get down typically to 10% of the Fermi temperature then it's for potassium even colder I don't want to go into details but this works very well okay cooling to deep degeneracy done now comes the difficult part do we have control knob here which we can turn to control the interaction tune the magnetic field to resonance study the unitarity we look into the fashion of resonances and already the first experiment is now done I think three years ago we saw many scary things resonances everywhere this is just between 50 and 61 there are super many resonances in disposal alone we typically have 50 super narrow resonances per Gauss resonances in 50 per Gauss and disposal alone and for the mixture you have also resonances not as many maybe one or two per Gauss and so then we started many experiments concerning these forests of resonances with different related fields which you can imagine is quite time consuming you don't know will you be rewarded at the end or does the experiment die in losses okay we found something interesting actually near 200 grams this is the thermalization measurement we prepare potassium and disposal together in the trap but potassium is heated a little bit so this has 3.5 micro kV this is 1.5 micro kV near thermal measurement then we see there are certain magnetic fields indicated by the arrows where there is very fast thermalization what does it mean it means we have resonances broader resonances which there are poles here resonance poles and we can understand this basically as a scenario of three broader resonances here pole there, pole there and so how can we characterize that we use actually a product formula for fashionable resonances which is valid for overlapping resonances can be written in that way here the C i these are the zero crossings where the scattering length goes from zero and P i are the poles of the resonance features and then we need a background and then how do we get this A background we get it from the cross section so we have to get some information on the thermalization and here I am actually we apply a model and I am very happy that Marcel is here because this is the model we developed in Heidelberg more than 20 years ago for the inter-speciesal stormalization between lithium and potassium and since then I mean it has been used by many groups to do that and we routinely use this now to get the A background here and that is our result then so we see we can characterize the poles and the zero crossings and we have the background scattering length so it looks like this so this is our scenario and it tells us there is a broad resonance near 217 cos good, can we do something interesting with this resonance now we look back 20 years the John Thomas experiment hydrodynamic expansion something similar, hydrodynamic expansion experiments with a mixture ok, here you see such an experiment we prepare a mixture not the general but close to the general see, this posium and potassium and we release the mixture from the trap ok, and then here we have a ballistic expansion this posium is or potassium is by a factor of 4 lighter means by a factor of 2 faster the expansion and we see it expands slowly potassium expands fast as we expect for a thermal equilibrium ballistic expansion of thermal equilibrium but this is done near a zero crossing with weak interaction or no interaction if we do the same on top of the resonance we jump on top of the resonance with our magnetic field do the same experiment the picture drastically changes now you see the potassium cloud is smaller much smaller we find in the expanding disposal cloud what happens they expand together there are many many collision where it is like crazy tens of kilohertz or so they collide and this slows the expansion of potassium down, makes it almost as larger for the disposal we call it a locked expansion a locked hydrodynamic expansion and then we look at the spatial profiles when we see something funny we see a bimodality we were extremely excited seeing this in the potassium component we saw a bimodality a narrow component the broader p-desperate and the reason why we were so excited is for spin mixtures the bimodality in the minority component is a signature of super fluidity so we were very excited when we looked at the temperature around the Fermi temperature this cannot be the case actually there is something else going on and we understand it now what happens is that I mean the potassium atoms move performs something like a Brownian motion in the expanding disposal cloud but sometimes they reach the outer region and then they are released from this hydrodynamic core and they expand ballistic this creates this p-desperate we can understand it and we have no super fluidity it's a generic effect of collision hydrodynamics and we can understand it and this is the result of the Monte Carlo simulation for the same conditions so just collision hydrodynamics and we see this p-desperate ok, we can analyze this more in a more quantitative way looking into how many atoms are in the central core and then we can plot it the experimental result I think are the black ones the field dots and the open red ones are from the simulations so we understand what is going on in collision hydrodynamics with extremely large collision weight but nevertheless it's nice to see in this hydrodynamic strong hydrodynamic behavior there's always a pre-corrosion of other things happening collision is stability I meant three body decay and we did a lot of loss measurements and we see the typical lifetimes are at least a few hundred milliseconds and from these measurements we can extract the recombination coefficients now if you work with three body recombination maybe you can interpret these values 10 to the minus 25 cubic centimeter to the power of 6 over seconds and I can compare this result for example on all those experiments here the potassium rubidium and all these experiments on resonance are 10 to the minus 23 or 10 to the minus 22 10 to the minus 21 centimeter to the 6 power over second whereas we are in the Fermi-Fermi mixture a few orders of magnitude lower so we have I would claim something like 2 to 4 orders of magnitude we see this polysupersion effect it is definitely there otherwise something would decay within less than a millisecond ok, now a superfluidity in reach actually we have a very nice collaboration with John Carlos Trinati an expert on these superfluids and he modeled with his group exactly the situation that we have the trap situation the trapping depth and the mass imbalance and so on in a safe consistent team metric approach which is very known to be very good for the mass balance systems and I just want to show this diagram it is a phase diagram for the trap system so here is the majority of this posion what we usually have in the experiments majority of potassium temperatures in terms of the search of this posion you see the lift shift point here and here you should get a superfluid below that line and in experiments we are typically doing so we are not there but it is not a long way to go I think with some improvement for resonant superfluid we should end up here to enter these regimes ok, so this is the outlook I mean we have shown that we can cool in a deeply controlled regime, we can tune the reaction by a broad resonance we have power suppression effect looks all very nice and very promising good reasons to be very optimistic but there is still one problem this deep cooling here that was done at low field a few hundred milligounds and these experiments are done at high field, 220 gauss and if you ramp up in your field then you cross many resonances you have problems for the experimentalists with eddy currents and the apparatus and so on this makes it a bit difficult it makes it really very hard particularly in view of the complex behavior of disposing background losses but disposing background losses this is just a loss spectrum loss in temperature for disposing health few hundred milligounds in a trap as a function of the B field this is not experimental noise these are resonance structures and you see there are some funny points where you have little losses we can address them these narrow loss minima and you can jump on these points but you have to control the field then within few milligounds you can use them but it's all kind of a complication and so therefore faster precise control of the energy field is very important but here if you look at low field this now is arranged up to one gauss and then you see we can nicely resolve loss features of disposing background loss and so we were should we look maybe more into the low field region we can control it a little bit better and we had a pleasant surprise recently when the new team started on the experiment we started to check the low field region again and we found resonances at low field not as wide as the one before not a few 10 gauss but there are resonances and what I show here are just binding energy measurements just by the standard magnetic field modulation method binding energy measurements a resonance at one gauss or a little bit less at 970 milligauss interspecies resonance a triplet near 3 gauss and a nice resonance at 7 gauss 3 gauss and we see here is a normalization measurement again which tells us there is a resonance which has a delta width of one gauss roughly which allows us to control the interactions and this curvature in the binding energy for the expert tells us there is a universal range and so this looks very promising and now we explored these resonance in more detail the parameters I don't want to go into detail here but this is a nice resonance for the experiments not as broad as the one at high field but for the experimenters it is so much easier to control the field at 7 gauss than it is at 230 gauss it has a lot of advantages in particular avoiding these bad positions for disposal and so on so we are working now on these resonance just to show you this one this is a repetition of the hydrodynamic expansion experiment with the 7 gauss resonance so we can do on the 7 gauss resonance what we did before on the resonance but with much better the prospects of controlling the field and we can do also something more we can produce molecules this is a recent example where we start with the mixture we ramp across the resonance then we apply a Stangala separation and then we see the dispersion the potassium atoms going up the magnetic field gradients the potassium atoms going down the gradients chosen in such a way that the molecules stay levitated very similar to what we have seen in David Jean's work 20 years ago on the molecule formation one stayed here, one stayed there in the center of the molecules very similar to what we see here then what was next if you look back these almost 20 years next step was formation Stangala and Stangala and we looked into the properties of this molecular cloud it's super cold it's nearly degeneracy and we hope with some minor improvements we can bring it into degeneracy so I'm quite optimistic about the BAC of heteronuclear fashion of molecules ok so ok this is some outlaw coming we can create normal superfluids hopefully look into full but few body states and look into Fermi-Potter ones and the medium of heavy particles many things to do but I think the general conclusion is on the fermions other called fermion mixtures are a great playground for physics, of stormy and active many body systems and there are many opportunities and challenges for experiment and theory ahead of us thank you for your attention