 Okay, so we have talked about molecules, we have started discussion of nanoparticles, we have talked about plasmonic nanoparticles. Now the question that we want to address is, is there something in between? Something between molecules and nanoparticles or maybe atoms and nanoparticles rather and what, of course when I ask like this the answer is yes, yes there is something in between and what is in between is right now projected nanoclusters. Nanoclusters are of course as the name suggests they are a conglomeration of not molecules but atoms and we will talk about what their dimensions are and all, they are actually you can think that if you can, if you go on making smaller and smaller and smaller and smaller nanoparticles there will come a time when there will be a transition between nanoparticles to this cluster of few atoms which you can estimate the number of atoms, again we are not going to go into preparation, we are not going to go into how people know that there are so many atoms and so on and so forth that I leave for us to read by ourselves, we will only talk about the ultrafast dynamics but these nanoclusters have been, have drawn a lot of interest over the last few years because they are the newest kinds of fluorophores to be honest. And what people have found is that this cluster of atoms that are not covalently bonded or anything same kind of atom and that too noble metal, gold, silver, copper, well coinage metal. So these, these clusters, conglomerates of atoms sort of behave like molecules as far as spectroscopy is concerned. And you can use them to make sensors, you can use them in fate and you can do a lot of that. So there have been very intense activity especially from labs of fluorescent spectroscopy in the area of nanoclusters. And of course like anything new this also comes with its share of disbelief and non-acceptance. There have been lot of people who do not believe in nanoclusters but people who believe far out numbers people who do not. So what we will do is out of this vast body of literature we want to present two papers. First is this review which is exhaustive but a little old 2013, 6 years ago. So at least until 2013 the state of the art of the ultrafast dynamics in gold nanoclusters is summarized nicely in this review. Once again we are not even going to talk about silver or copper or anything that is self study but if you understand gold nanoclusters I think it will not be difficult to follow the rest. So this is a paper by Professor Theodore Goodson III the very eminent ultrafast dynamics researcher. So of course it is a review accounts of chemical research it does have papers from his group mostly but from other places as well. And for whatever reason in this paper in this module I have not written the references as such at least in the first part. I have written them as reference 9, reference 10, reference 12 those reference 9, 10, 12 are with respect to this paper accounts of chemical research volume 7 year 2013 page 1506. Now the thing is this this is where one usually draws a line between nanoparticles and nanoclusters. In nanoparticles as we have discussed already you have prominent SPR and as the size increases there is a progressive red shift and the color changes as well. In fact this change in color of gold nanoparticles is there even in CBSE class 12 textbook now it has become that common place. Now when you keep decreasing the nanoparticle size what happens you see a blue shift progressively and also the peak becomes less and less and less prominent because you are getting into scattering regime that scattering regime. That about 2 to 3 nanometer this peak vanishes. So what you see here is for this 2 to 3 nanometer this is what you expect to see from ME theory. If you actually have particles and nanoparticles ME theory is so there are two theories of light scattering one is Rayleigh and one is ME theory they work in different regimes of particle size. So ME theory still predicts a sharp band and what you see here is that if you look at this clusters I am not going to tell you how these numbers were found you can read up yourself for gold 2400 or something you do see still some surface plus 1 band it is almost gone in the next one and when you go down to things like gold 25 and all there is no band you do not see anything in absorption but you do see emission as we are going to show you later. Now it has been worked out that nanoclusters can be formed of certain discrete number of atoms you cannot have a cluster of 5 atoms then 6 atoms and 7 atoms 20 20 or no there are certain magic numbers and the reason why magic numbers arise is that the basic structure of nanoclusters is icosahedral. So of course when you go down to this clusters of atoms symmetry becomes very important in fact even do not have to go down to atom if you talk about assemblies of nanoparticles. So Gillard Haran's group has done significant work on spectroscopy of assembly of nanoparticle and there symmetry plays a very important role there symmetry and symmetry breaking here also symmetry has an important role to play. So they are found to be icosahedral mostly by theoretical calculations of course it is not very easy to see this individual atoms and what holds here once again is free electron model very much like the nanoparticles but as we will see nanoparticles are very different animals compared to nanoclusters. Nanoclusters are different the major difference between nanocluster and nanoparticle is prominent emission the previous module we stopped with the comment that gold nanoparticles do have some emission but then that emission is feeble and you need to work hard to see that emission it is easier to miss that emission than to see it in nanoclusters you cannot miss the emission no absorption emission is there. So what you see here is two kinds of emission one is visible and the other is NIR so in different regions you have two different kinds of emission and if you look in the time domain this comes up very nicely what you see here really is up conversion normalized counts so you see when you have a nanoparticle then the laser pulse and whatever light comes out from nanoparticle they are completely superimposed can you see that these two however when you go down to nanocluster regime then you see the decay here is different from the instrument function okay not very different but different nevertheless and the lifetime comes out to be 250 femtosecond once again not very easy to measure unless we are careful. So 250 femtosecond is the lifetime that you observe and what theorist group has done very carefully is that they have looked at the lifetime for the different size nanoclusters and what they establish is lifetime is around 250 femtosecond not so easy to differentiate between the clusters but the moment you go beyond say 2 nanometer or 3 nanometer there is a drop in lifetime so this is scattering all the way so this is the nanocluster nanoparticle divide as far as temporal properties of luminescence is concerned. And the mechanism of emission that has been proposed for golden nanoparticles is this but first of all you excite and if you excite to higher level then it comes down to a lower level pretty much like a molecule so this is basically a Gibbonsky diagram and you get visible emission from B as well as A so that would remind you perhaps of 7 azine dose and then the NIR emission has been attributed to surface states okay so this is the explanation why you have one emission invisible why you have another kind of emission in NIR. Now the other thing that has been done very nicely is 2 photon excited emission it turns out that these nanoclusters have a large 2 photon absorption cross section and that is what makes them very interesting and useful from the point of view of say microscopy because in microscopy you do not want to excite at 400 nanometer even 400 nanometer will be scattered if you want to excite at 350 nanometer then you need a different microscope which costs much more because all optics have to be made of quartz. So 2 photon excitation is the way to go and in 2 photon excitation you see this is the absorption and this is the 2 photon excited emission and this is for gold 25 as you see excitation wavelength is 1290 nanometer and 12 exciting by 1290 nanometer roughly this state is excited and you can see a nice emission coming from there and if you plot fluorescence counts against power well log log plot gives you a slope of 2 which confirms that it is really emission arising out of 2 photon absorption 2 photon excitation. So this is the visible emission with lambda ex at 800 nanometer once again the slope is about 2 and this is visible emission of another kind of nanocluster again with excitation wavelength of 800 nanometer once again you get a slope of 1.94 which is roughly 2. So from all these experiments it is established that 2 photon absorption is a reality and that can give rise to very well that can give rise to fluorescence that you can see. And once again by looking at this 2 photon absorption cross section itself you get a nice distinction between nanoparticles and nanoclusters and by the way the unit of 2 photon absorption cross section is gm 10 to the power 6 and this gm is for Gippert Mayer Maria Gippert Mayer was the Nobel Prize winner in physics for her work on non-linear optics in 1963 this unit is named after her. So you see just look at this this is 2 photon cross section this you get a linear variation with increase in number of volts in nanoclusters and then there is a break and then you get another line with little different slope for nanoparticles okay so this is another evidence that nanoclusters are really different from nanoparticles. And the thing becomes even more interesting when you work out delta per gold atom in gm then you see actually there is no increase this increase that you see was really a brute force increase an extensive quantity right because the number of gold atoms is increasing but what you see really is that delta per gold atom decreases as the nanoclusters size increases and then when you go to this nanoparticle regime then it is really very small per gold atom because do not forget nanoparticles are many more gold atoms than nanoclusters. So if you work out the intrinsic quantity then the separation between nanoclusters and nanoparticles is really really very good with that background let me show you transient absorption data this is what was so this is a little old data we will talk about new data as well. So here you see there is a ground state pleats and when you go from gold 25 to gold 55 to gold 140 how does it change gold 25 where is 55 what is the big one 55 or 140 140 so this size increases this delta A here the transient absorption becomes larger and you see look at the decay homo-lumo charge transfer in core that is how it was assigned at that time that was assigned at time constant of 1 picosecond and core to shell charge transfer what is this core shell all these things nanoclusters cannot exist by themselves you need some kind of a core some polymer or protein using HSA or BSA to make nanoparticles is very popular so some core is there and then like what happened in gold nanoparticles the energy does get dissipated so this core shell charge transfer is you can think in some way reminiscent of the electron phonon coupling right not really but more or less so it is basically active species to the environment kind of charge transfer so this is how it was assigned initially that there is a homo-lumo charge transfer in core and that is associated with 1 picosecond time constant and there is a core to shell charge transfer that gives rise to a longer than 1 nanosecond kind of lifetime. Now one problem that we have and you might remember what we said in the beginning one problem we have here is that there is no absorption spectrum so who has said that this is really emission from gold nanoclusters and not some impurity because you can record excitation no problem what will we compare the excitation with how do we know that it is not an impurity well transient absorption comes handy there because if you look at 1 picosecond well the earlier transient absorption I showed you was in of 550 femtosecond okay but then there is something that goes on for a long time so if you look in the 1 picosecond to 10 picosecond regime you see these ground state bleaches that are coming up okay what does that say that says that at that position spectral position there is something that is absorbing right so using that the absorption spectrum has been reconstructed and that can be compared conveniently with excitation spectrum okay so this is an elegant experiment where transient absorption helps you establish that you are really looking at the thing that you think you are looking at it is not some garbage right so this was the state of the art until that time 2013 now let me present this paper at least part of it that was published as recently as 2 years ago so this talks about electron localization in dot shaped try icosahedral gold nanocluster you might remember that we had said that icosahedral geometry is reported from DFT calculations right so this is actually the result so and this is the origin of the magic numbers also so what is proposed from theoretical studies is that the monomer so you need at least 13 gold atoms to get a nanocluster so gold 13 is the monomer and then this monomer units join up in different ways to give you other nanoclusters which are of different shapes for example you can have two monomer units coming together to make a rod or they can come together to make a timer the difference between this and this is of one gold atom yeah this one in the center is missing for the rod but is there for the timer similarly you can have gold trimer where you will have 37 gold atoms what those red dots are we will talk about a little later and this absorption spectrum can be conveniently discussed like conjugated molecules using particle in a box model because they are not very large okay so see these are the absorption spectra you can see a red shift and then when you go to the trimer you can see something new in the nir region so with this background what has been done is the transient absorption studies have been carried out on the monomer as well as the dimer and trimer and this is what you see for the monomer these are the transient are usually we do not show transient absorption like this but they actually are better looking more colorful and since both time and spectral information is there if you know how to read it then it can be more useful and it is the same thing you can take sections and compare but if you are used to it you can just look at this matrix and get spectral as well as temporal information alright so let us look at this one what they have done is they have used two kinds of pumps two wavelengths C 16 nanometer and 560 nanometer this first column is for C 60 nanometer pump this is for 580 nanometer pump what do you see here you see an excited state absorption in that 650 kind of region and here you see ground state bleach here of course you cannot look at this region of the spectrum that is why it is white and then you see another excited state absorption in the higher energy site okay so that feature is there here also but the excited state absorption in the 600 to 700 nanometer region is absent when you excite at 560 nanometer what does that mean well what this means is that when you use 360 nanometer pump and when you use 560 nanometer pump you are exciting to different energies right and again this should remind us of the 7 as a end old problem when they use the higher excitation wavelength they ended up exciting S 2 and then carefully they tried to decrease the wavelength so that they excite S 1 and not S 2 and remember what happened there when we excitation wavelength was varied what happened was that there was initially when excited at higher energy there was a fast component point 2 because again which vanished when excitation was done at sufficiently long wavelength so here what this tells us is this so this one 600 to 700 nanometer excited state absorption this is excited state absorption remember a positive signal it shows up only for 360 nanometer pump that means this excited state absorption is most likely from S 2 it is not accessed when you pump by 560 nanometer and what about this the higher energy excited state absorption which is observed in both that must be due to S 1 yeah and the evidence is that S 1 is accessed by the lower energy pump as well S 2 is not secondly what is the energy gap between S 1 and some higher energy state and S 2 and the same energy state the energy gap is higher for S 1 right so this absorption here is assigned to S 2 to S n smaller energy gap longer wavelength smaller energy gap this one is assigned to S 1 to S n S 1 to some higher energy singlet state and you see here we have started talking about singlet triplet once again because since it is a cluster you can actually talk about this spin states okay it is like a molecule so here you see you get a 400 femtosecond decay and you get a very long lifetime 1.72 microsecond long-lived state at 630 nanometer as well as at 750 nanometer when you pump at 560 nanometer you do get that 1.72 microsecond decay but there is no first component right what is that first component due to then 400 femtosecond must be from S 2 to S 1 that kind of thing yeah and this is corroborated by the long time pump probe and TC species studies and this here is the emission spectrum which is centered at 750 nanometer okay and then they did this experiment on monomer and already there was literature from 2011 I have forgotten who did this work there was literature report on dynamics in dimers and they turn out to be the same okay so monomer dimer not much of difference is there then they moved on to AU 37 trimer and in case you have forgotten what AU 37 looks like this is what it looks like now what do they see it is better that we look here at 560 nanometer probe you see a fast decay at 70 nanometer you see a rise and that is associated with 100 this decay and this rise look at the full scale 2000 picosecond the time constant associated with that is 115 picosecond so what they concluded from here is that this is due to excitation localization because they are done some calculations right excitation localization means this when you excite initially excitation is delocalized over the entire nano crystal but then they said that there is another state which a little lower in energy which you cannot access by direct excitation in which the excitation is localized that is what this means and also they compared the spectra pump probe spectra at long times 1 nanosecond and dimer and trimer turn out to have more or less similar features albeit at different wavelengths now dimer of course has no option it is known that the excitation is localized so what they said is that this long-lived species that is there is excitation localized species so when you excite excitation is delocalized in 115 picosecond post that ultrafast dynamics you get a localization of the excitation and then you get the emission also they made an important observation here the long-lived emissive excited state is has a lifetime of 28 nanosecond so much lesser than what you get for the monomers and they did some more studies of what they did is they oxygenated the sample and there was no change so they inferred that it is not a triplet state the second state that we are going to by this 115 picosecond time constant pathway is not a triplet state they also established by changing the medium dilatatory constant and all that it is not a charge transfer state hence the energetic that come out is something like this the post excitation initially this is the state that is produced where the excitation is delocalized then in 100 picosecond it has to go from here to here so if you pump at 1200 nanometer as we had shown earlier then you reach S1 by the way this was AU 37 has to be pumped at 1200 nanometer because that is what the band gap is so this is a probe and here this is S0 if you excite to S2 is from S2 there can be two pathways within one or two picosecond it can come down to S1 or what they propose was this another state S1 star which is not available by direct excitation and that is ascribed to this localized excitation excited state but a singlet excited state nevertheless so that has a little lower energy than this S1 you might ask why is it not being called S1 then and why is this not being called S2 the reason is notionally you can call this S1 that you can call this S2 the reason why they prefer to call it S1 star is that you cannot access it by direct excitation there has to be this relaxation and localization of excitation before you can get here so that time is 100 picosecond. So right now this is the state of the art this is how the excited state dynamics of this dimers and trimmers and the monomer have been discussed but I am sure there is ample opportunity to work in this and it is not as if you have one kind of nanocluster you will have the same thing the environment is also dependent so it is a good reason that there is so much of activity in this area and then as we said it is also useful for microscopy it is being used as novel 404 in many cases and it is not very difficult to say denatured the protein and then your nano crystal is completely gone so it can actually be very sensitive to inputs that one may provide so that is what we wanted to talk about in nano crystals in nano clusters in the next module we will go on to semiconductor nano to ultrafast dynamics in semiconductor nano crystals.