In this lecture by Prof. Oliver Kühn, he discusses the importance of transition metal complexes, such as ruthenium and iridium, in various applications. These complexes are valuable, but their use is limited due to high costs and rapid deactivation of charge-transfer states. Prof. Kühn explores the potential of Density Functional Theory and optimal tuning of functionals for characterizing these complexes, especially iron complexes, highlighting the complexity of the design process.
So welcome everybody I’m delighted to get off a new session as a part of our Arab molecular simulation community and I hope we can achieve something great together so it is an absolute pleasure to introduce our distinguished speaker for today profor Oliver ton he of course like U I would introduce as wellknown
Scientist in the field of theory iCal and computational physics and chemistry with a particular focus on Photo physics and chemistry so we can say like light matter interaction so oler started his academic career or his academic Journey at hble University in Berlin where he studied physics and in 1990 he obtained his
Diploma in physics and 1995 he he obtained his PhD in theoretical physics from the same Institution uh later he did like two boss dogs in the United States and Sweden and in 2000 uh in 2000 he obtained his habilitation in theoretical chemistry at for University Berlin under the guidance of
Professor y mans in 2008 he he was appointed as a professor at the of theoretical physics at the University of rosto which I’m working of course under um his like guidance like a group leader with him um Oliver Kon visited like different institutions in United States Japan Russia and he made like
Significant contributions in Egypt particularly at K University today he is going to talk about photophysics of iron complexes and give like a theoretical challenge so please join me to welcome Oliver K and please so the floor is yours okay thanks a lot and yeah thanks for this nice introduction and for setting up
This uh this meeting I think that’s going to be a great opportunity to exchange ideas and communicate scientific results so today I will talk about uh essentially density functional Theory and uh with a particular focus on challenges which you’re appear uh if you look at iron complexes and the photophysics and also
For the chemistry of iron complexes okay so before I I start let me just thank those people who ESS the most of the work which is s boar who also will give a seminar in this series Anda barara who did her habilitation with this topic or will do she’s not finished yet uh
Financing comes from different sources most notably at these stage is um the German Science Foundation and here are our um experimental collaborators from Roso pabor mines and our theoretical collaborator from Vienna letisia Gonzalez and her group most notably Patrick T all right so here it started
So sometime ago we um uh had a grand to study the photocatalytic water splitting or essentially a half reaction and based on a photosensitizer which is see down here so it has a as a metal uh idium all right so then we studed basically the whole series of complexes separately like the
Exitation basically the electron transfer from the py ficial reductant and then the fate of the reduced species which basically donates an electron to iron based Catalyst which eventually like set free hydrogen we also studied what happens if you increase or or or modify the absorption Properties by just adding
Like a small silver clusters and so on so this was in a way rather uh successful and uh well also in terms of of experiments as proof of principle so there’s one thing in terms of artical challenges which is related to the fact that you have different
Charge transist so let’s see we we plot here are the potential curve well it could be any coordinate it doesn’t really matter at this point and uh you see what you start from the electronic ground state to one of the excited States so rious processes happen in
General you have kind of heavy metals you have inter system Crossing from the Singler to the triplet manifold and you have internal conversion due to nonadiabatic coupling that is the breakdown of the born up M approximation and uh so this all brings you back sayfe for instance from the
Initially excited state to some say metal to lean charge transfer state with this charged separated States you could either have photoluminescence or you do as I mentioned before like this electron transfer with this sacrificial reducted and well with this then you can go like do further reactions as I mentioned before like the
Electron transfer to the iron based cataly all right so what’s difficult now if you want to describe this so there as I said there are several charge transfer processes intramolecular so you have a charge separation from the metal to the liant and intermolecular from like one uh species uh like the
Photosensitizer to uh the uh sacrificial reduct so this is um um um this is uh basically the challenge we we Face here and of course you have to kind of make some predictions uh uh for these processes you have to be very accurate because it gives you some sort
Of constraints of ther dynamic or kinetic uh nature uh which tell you how to effectively make use of these charge subverted States okay so then you think of um calculations basically to get some mechanistic understanding of the properties or or the the what Happ happens after excitation by a photon uh
You would think of okay let’s just solve the electronic Shredder equation well which you all learn in physical chemistry course and well what you have here you have the kinetic energy you have the electron electron interaction and you have the electron interacting with the nuclear which are of course
Positively charged right so if you solve this then you well basically get all information uh you you ever could could have and to predict properties predict experiments the problem is that you have here uh the electronic coordinates and you have n electrons so you have three n coordinates just imagine you do this
Basically by discretization on a numerical um or discretization of the um of the coordinate say you take 10 points per coordinate and then you have 3 n Dimensions so you have 10 to the 3 N Point points you would have to calculate or you have to consider in order to
Solve this equation well this becomes very rapidly totally uh in conson okay so what’s the alternative and the alternative is density functional Theory there are these famous homeb corn theorems from the 1960s which in a nutshell Give the recipe how you can equivalent um or describe the properties
Of the molecule in terms of the electron density right so that’s the electron density at position X and here you kind of integrated it over all the other coordinates so it’s much simpler you have just three coordinates so that will be Dimension 10 to the three if you go
And do it on calculation on on on the letter well that that’s easily doable with noway noway computers uh in practice what you do is you don’t do this on in that way uh but you uh can show that there’s something called konam equation which give you the conam orbitals FJ and
Like their energies which is Mo can be seen as molecular orbitals and which are taken together then uh give you representation of the electron density right you some here over all occup uh the tricky point this is an exact prescription the tricky point is this kocham potential here which
Contains the electr nuclear repulsion which comes from over here and which contains the electron density I mean that’s basically coolum interaction which comes in here and then comes the most tricky part which is this exchange correlation or potential which deres from a um exchange correlation energy depending on the density or and that’s
Kind of functional derivatives never mind the point is uh in this derivation one has kind of put everything one does not know and whenever an obstacle occurred it was put into this exchange correlation part right so provided one would know this exchange correlation part one would be in game and solve the
Problem of this three and dimensional wave function or in this case three dimensional electron density EXA of course that’s would be too good to be true and there are only a approximate functional so the exact form of this functional is not known and some of the approximate functionals you might of
Course know this PBE or B3 lip is very popular and so on there eventually hundreds if not thousands of different functions all right um so what’s uh then the next step so that gives you the ground state the electronic ground state like the density are of the electronic
System for in the electronic ground seat of course if we want to study photochemistry photo physics uh we also have to look at excitations so uh there are different types of excitations uh say here you have like different parts say that’s a metal center that’s the liant orbitals
And uh here you would have local transition which would be in that context a metal center transition but you also have charge transfer transition right so from the donor to acceptor like metal to Legal charge trans Transitions and that’s a problem uh because the way you calculate now the optical or the ex
Electronic exceptation is what’s known as linear response time dependent DFT and uh well when you derive this there’s something it’s called CA equations which look like so basically that’s a generalized I value problem which you have to solve about to get the excitations right so um I don’t want to go into
Detail here just mention the point because you have here this two electron integrals like here and here and you see here okay I is a um occupied aital a is a virtual orbital J is occupied B is is virtual and you see here now you have i
A okay this basically at site R1 or or in at the r one and you have J here and B at R2 right and if J and um or so if you have a charge transfer transition like these two orbitals will be kind of centered very far from each other so basically that
The overlap are um the overlap whoops pass to my computer so the overlap is zero and so basically these terms don’t don’t not do not contribute and you get the excitation energy just as the difference in the single particle K energies they could be erally wrong so
You have a different issues which come along with this you have wrong Astic behavior of the density so it falls off uh not with the proper dependence you have uh so-called self interaction error you have um uh in a fact that the homo lumo Gap is not the difference between
The ionization potential and the electron infinity and then if you look at how charge separated states are well their energy depends on a distance between say donor acceptor you would expect from the kums law something like one over the distance and that’s not fil all because like this basically these terms are
Cancel each other of course there’s a a remedy for that and that’s uh like these so-called hybrid functionals like this famous B3 so what one does here is one artificially um adds some uh term which is coming here with a factor which is called C HF HF stands for hard to
and essentially um what one has then when one adds here something which is uh has the form of the exact Xchange integral for H Theory it’s completely if you want to like just to cure a shortcoming of this functional because if you uh of these functionals if you
Now look at this integral you see here you have i j being at the same well having the same electron um fornate and ab as well so what happens then is that uh for these charge separated States so these terms still vanish for for from the equation but you have then basically
This exact exchange right so basically you have you get the right uh in a way ASM totic or right dependence of charge separated States say as a function of the donor acceptor distance but but that’s really a little bit a talk and uh well now well as I
Said like B3 lip is one variant of this and um essentially um these different functionals so that’s basically like the two no uh four different um um like two singlet and two triplet excitations of charge transfer corrector here for this idium complex and that’s the transition energy and that’s basically different different
Methods and you see like depending on like all these different like Bop B3 Loop m6 and so on they um are differ in the amount of exact exchange you have artificially added to this in a way the C factor which I gave in the previous
Slide and you see okay here on that side are more high level calculations like CCF and cbd2 but you see essentially when you when you take them as a reference you can get well aary results right so basically the energies shift widely back and forth so that’s that’s
Very complicated and not very nice for to be predicted all right um okay so um when you then look at the absorption spectrum of these species for these different functionals you see okay it’s not only mean this is here for one state but you see also they have rather
Different absorption Spectre and that’s of course making life very difficult okay um what you can do to improve on that you can use the so-called optimally tuned rage separated hybrid functions so in a way U what you do for these sort of um well you have local Transitions and long range charge
Transer transitions are you add up um or for the long range part or you you you add up this kind of hard exchange for the long range part and for the short range part say for close for small electron electron distances you use your whatever you had as a functional R
Before all right but how how you decide on long and short range right and you do this in the following way that you say okay well this is uh the electron electron interaction here and then you add to basically you add a zero here right you see this cancel but you say
Okay this part defines short range this part defines long range and this is the error function and you see how this bit depends now on this parameter Omega which basically can be used oops okay can be used uh to T tune like what’s long range and what’s short range right
So Tunes the fraction of this exact f f exchange or F Exchange and uh this the fraction towards between this one and uh the short range say GGA function okay but U so what uh if I now take like different values of that so that’s this Omega which is basically like
The steepness of the switching function you see that for for this system I get like different Spectra so apparrently now I shifted the problem to finding optimal Army okay what’s the rational way of doing so of course like some of the um parameters like the 0.0 0.33 and 47 they have been obtained
By just like in a statistical Way by looking at many many different complexes and comparing with more accurate data and uh finding a compromise for this parameter okay but where’s the 0.18 from and in fact you can there a nonempirical way of tuning this range separation
Parameter so if you it works as follows so you have a energy of charge transfer state so it has the ionization potential of the donor minus electron Infinity of the acceptor minus this Kum in Direction one over um our distance between D and except right so you have donor
Acceptor um in exact K sham Theory the energy of the homo right so this conam orbital of the N electron system is should be equal to the um ionization potential um that’s an interesting result right so that’s that’s exact should be exactly true what’s the anization potential is basically the difference between the
Energy of the N minus1 electron system minus an electron so then can you you can just add one electron then you get the energy of the homo is like the electron affinity and this is IP of n + one so it’s e of n minus E of n +
One why this important because you now can say okay this should be true for for any for the exact functional why why don’t we enforce this right so why don’t we say okay I take a calculation and compare the ionization potential which I get just from this difference
Here so I do a calculation for the N minus one electron system and for the energy of the N electron system just take the difference compare with the homo energy and get this expression and I do the same for all this electron affinity term right and this gives me a
Function can kind of arrage this in a way uh it gives me a function called J of Omega of this range separation parameter and then I uh just argue that okay if I minimize this function I get the best possible Omega which in the best possible sense fulfills this exact con Theory
Constraint so basically I construct the system specific functional which fulfills certain exact properties well remember we don’t know the exact functional but here I can just say okay well look I I I um I uh just choose this functional to at least fulfill one known exact
Property all right so how does this work when you like this is here for this rum photo sensitizer you see again so that’s basically this J function is a function of this Omega and you see here okay well if you’re lucky then this function has has a minimum and uh
Which then gives the best possible values value of for of Omega for your particular system okay this is these are this one the blue one and the red one are the standard values which are in whatever gaussian and other codes fixed and you see you you get quite some some
Differences here for this particular system in particular you get like your optimal value here is 280 all right uh you can also work let me just skip over this you can also do this with salvation uh which has some interesting points and um we did a lot of work on
That and I just mention some of the papers we we wrote in this respect and I just want to show some results uh for this photocalytic water splitting uh half cycle for this idium photosensitizer so so it’s about the quenching of the excited of the fluoresence of the excited idium
Photosensitizer due to the reaction uh with the uh um Iron um with the sacrificial productor right and you see okay what what what you can see from this this value so that’s kind of the photo lience intensity T is a function of time and uh you see here are two
Decays say if you have um basically the bare photos sensitizer where the no TAA is in it has a very long Decay time and if you add a little bit of this uh sacrificial reduction which then leads to electron transfer and therefore quenches the photo luminesence you see a
Much more rapid uh decay uh actually one can estimate I mean it’s a bimolecular reaction and one can estimate what should be the uh kind of diff diffusion limited rate and it turns out that the diffusion rate is is much much faster than the uh rate of the
Reaction so the quenching efficiency is very very low why this is the case I mean we looked at this and in a way that okay so you you study the electron transfer uh between these two species with this range optimal tuned range separated functionals in a way to um how this has
Been done is basically you have the uh um you have the iron elium photosensitizer here and you move this ta around this and then check where basically it’s energetically favorable to do the reaction and uh when you do this you you find here that okay that’s basically like the
Uh solid angles here and what you see here is essentially the uh colorcoded The Binding energy and these dots here is like those energies where it’s energetically or those angles where it’s energetically favorable to do the reaction and turns out that um it doesn’t really come together so
Basically there are only few sides for good binding and uh and uh for those sites it’s not optimal for charge transfer so the systems I mean in a way not are when once it forms in counter complex it’s not really optimal for charge transfer which um uh kind of explains the low
Yield okay but I wanted I mean the title of of the talk is about iron comp and um uh so I mean idium is nice so everything works also you can go to ranium and so on so it’s all very nice but of course one would like to have
Just in terms of su ability to have these reactions working or this this complexes working for Earth abundant metals like like iron and this whole priority program like controlled reactivity of metal complexes so what’s the the issue why why iron is so complicated and some of you or I’m sure you know this
Tavano diagrams um are for D6 system and turns out that uh well in um our iron complexes right so you have like ranium here you have the position of the MLT states you have iron here so that’s basically the where it’s crosses it’s the metal centered stat and
You see when you take ranium and compared to U um to compared to to iron you see that the following happens that the metal to cented states are below energetically below the mlct states so what then happens is you have like a Cascade of events which lead to a very rapid
Deactivation and a short lifetime of this mlct States and that’s of course well hinders if you say like this you use this uh mlct state to well as a kind of reservoir kind of state which which interacts with with other species to have electron transfer
And so on so for that purpose you would like to have a long lift mlc Okay but well that’s basically in in in this idium like this MC States is way up in energy so it’s really everything can be stored forever I mean for NCS in this mlct States but in Iron complexes
That’s uh not the case so there’s a quite some active research going on for to find Long lived mlc States and quite some progress has been made um here we are on also made a contribution to to investigate a particular system and um or used what is called two parameter tuning remember I
Used this parameter Omega to tune our exact the contribution of exact exchange and uh um approximate exchange for long and for short distances there’s a next I mean higher level of this when you not only use one peret tuning ver two parameter tuning okay say it’s probably better to
Look at this lower part of the slide so you have like B3 lip that’s the distance between electrons and uh B3 liop it’s kind of hardcoded right so you’ve always 20% exact ex sh in this uh uh LC be which I talked about before uh well it’s with increasing distance
The heart to F or exact exchange the amount of it increases up to 100% And now with this two parameter tuning you can in addition for short distances mix in some except exchange from which of course gives you more flexibility okay let’s see how this this
Works um I mean first of all I have to to tell you where where do I get the other parameters from right so I I give more flexibility but I also need more constraints so you could of course look at the fundamental homo Gap uh could have look at Ground stability of ground
Seed Solutions look at what’s called delocalization error so um and of course also the experiment right so in terms of this delocalization error there’s another exact um uh exactor um point of quum theory which is called yanak theorem that basically if you allow for a fractional electron occupations then uh between the N
Electron energy and N minus one electron energy there should be linear um or dependence on the fractional electron uh number and so if you have a deviation from this linearity it tells you okay well this it’s can be related to all wave function being either too localized or too
Delocalized okay in short we have another exact property where we can uh tune uh the parameter in this case Alpha all right um let’s just look for this different I mean this was one set of different Iron complexes I just show here so it’s basically all this cine with this cine
Lians here and you see you have like different substitutions uh so you have like a series of compounds which have slightly different properties which can be um uh which can be compared uh what you see here is then first of all like this and uh for this range separation
Parameter you see it’s always kind of like in the same range which just tells you okay well this range separation parameter I mean it’s about the electron electron distance and uh say if you have like systems which have a kind of similar size all these typical distances
Will be kind of similar so you wouldn’t expect vastly different omegas here okay that’s fine um and uh uh what you then see here basically in this plot for this ic1 complex you see how this deviation from linearity behaves as a as a function of the fraction um of
The Delta so the fraction of of electron fraction fractional occupation or fractional electron charge and you see okay well they all like okay so it always looks like so so it’s it’s not really huge differences so basically if you take like like different pairs of of Alpha and Omega they have rather rather
Similar so it’s not really very strict uh um a very strict confirmation of of any or preference of any of these values right if you look at this for like two dimensional plot Alpha and Omega you see this here in color scale and you see there’s kind of
Valy of optimal values of or optimal Omega Alpha Pairs and um yeah so it’s it’s really there’s some flexibility so it’s not really that one alpha omega will stick out there will behave kind of similar all right um but in general one can argue that these parameters are uh comparable for class of
Mark okay W do they all give say now we have these optimal values which are given here like Alpha Omega pairs um or Omega Alpha pairs uh and uh or do they give good spectr well you see here the experimental spectrum and the theoretical Spectrum in in blue and you
See well it’s not really uniformly improving right so basically you you might argue well there’s no systematic Trend uh that’s a bit uh um a difficult Story I mean in a way you see already here there there are really many many many many Transitions and uh what you did is by
Tuning this Omega in a way you kind of worked on the homo lumo Gap why all these other states should have be have I mean why this Omega Alpha pair you get should be optimal for all these these transitions it’s not that that obvious okay so uh in order to to uh
Cope with this we said okay well let’s set out for one particular system and study this with kind of all different methods and and information we we had we had in hand it’s basically it’s called just cpmp it’s it’s basically the same type of system same type of liance as we
Had before and we it has been just synthesized and published last year and we also did this Omega Alpha Omega tuning for this system right you see here okay that’s Alpha Omega and to see okay why I get like this valy of Alpha Omega values which which are more or
Less optimal in particular in this lower part here for this compound if you ask yourself I mean how do the um properties depend on this Alpha Omega that’s kind of interesting you see here the position of the lowest mlct singular transition energetically right in wave numbers in
In wavelengths in nanometers and uh you see here Omega Alpha and you see basically being on that range you can get anything from what 700 nomer to down to 450 Nom right so basically this um MLT states are really really very dependent on these Alpha Omega all right how about the Spectrum
And and in fact this system was nice to us because it gave a very good very fair comparison with the experimental data so Bic you see here this Alpha Omega optimization that’s basically the electron difference density for the lowest charge trans for transition it’s looks very good definitely looks better
As uh as compared to this B3 results although it’s also not it’s also quite Fair okay but that’s not not all um in particular the question is uh well okay well we work pretty much on this getting a good values for these mlct transitions how about the other
Transitions This MC transitions right so I told you that uh the whole point of Designing um this or synthesizing these systems with different lians is to change the uh a relative position of metal centered and mlct States Al you could also ask okay what what happens here if I want to optimize
For like MC states do I get other parameters and so we we kind of looked at different ways of optimizing transitions um I don’t want to go into detail here just focus on this plot here right so this white line was U um like the the one which which I discussed
Before and these different other lines are like different other variants of of tuning and you see it’s all like in this range of of Alpha and Omega so it’s really not which are in this that value so it doesn’t really give to much U uh news here so well
Presumably you don’t get vastly different vs for the tuning parameters if you go from a mlc to an MC state all right okay so then what what we can do we could also say this these complexes are very big right but but when so far we compare to experiments
And like so the absorption Spectrum I told you these are like many many many states many transitions so it might not be and it’s also not very highly resolved so it might not be the best thing to compare with so let’s compare with exact calculation or presumably
Exact calculation which would be as a standard discuss CF cpd2 calculation of course this has a problem that it’s the the system is far too big to go and really make a one to one comparison between many transitions right so therefore we did this only for
Lowest MC and mlct states and to try to design this um CF calculation just to get good vales for these two types of transition okay so let’s look at this what you see here is the following you you get here you move along this Valley of optimal
Alpha Omega it’s only Alpha shown here to make the figure not too complicated but for each Alpha there’s the optimal Omega there these are the signals and um so uh you see how the red curve is uh the the mlc position of the mlct states the lowest one and uh the
Blue curve is the position of the MC state right and you see okay well look if I now go from along this Valley I even change the order of these two states okay and um I can also go for the triplets and also I see that’s Bic here
The right figure and also see okay I change the order of the states here at some point now look at the other points right so basic here the these straight lines are the C pt2 vales and like the other points is like different tuning recipes um the point is that and that’s
Very important message is that these are not pure States it’s not really like 100% MC or mlct states for most of the time and what you see here down and up there okay I mean mlcd is pretty good but the MC States are mostly mixed so
They always have a uh kind of mixture of mlc key character and mlc so this might be the reason why the tuning parameters are kind of similar but also this makes makes life rather kind of complicated because you you don’t get really clear results
And wouldn’t be able to um I mean as you see if you compare with this C pd2 here you never really get a gap between MC and mlct which is exactly um I mean at least not for those points here exactly the one for for you have this tuning so
You have to do a compromise in particular mean you want to have like the signals and the triplets right so they they it doesn’t come together so you can do a compromise and here we took like those points here at 0.2 uh which um Al 0.2 so it’s basically
Here where the order of the singlet and triplet MC and MCT transitions is at least correct okay so then we have like two uh sets of parameters one is here from this comparison of this range separated functional with cp2 and the other one is like from this
Optimal tuning right so this was 0 0.14 okay does this make any difference though we want to compare these two sets of parameter in respect to the performance okay so you can go well of course you know this this person and you can go for ask the experiment and do pump prop
Spectroscopy so which is essentially in the nutshell you have a pump PA and you have a probe PA you focus it on your sample and you have a time delay between pump and probe and this uh then you use to first of all the pump triggers reaction or an excitation and
The probe or then the certain time delay after the pump see looks essentially what what happens what happened in between right and um so all you do this in terms of spectral changes so also called Transit absorption so what what you then get typically is this type of
Signals so okay here is it’s like a just a chromophor doesn’t really matter in details we have like a um or the absorption Spectrum the fluorescent Spectrum here it’s exitation at a certain wavelengths and then you see okay well three contributions to the signal so you have like a first of all
You bleach the ground state just because you you excited some of the species you might have stimulated the mission range of where the fluorescence spectrum is so this two are shown here and then you have excited state absorption which has a different sign right because it’s xer
Absortion okay and this can be used now as if you just record this Spectra as a function of the delay time between pump and probe you can learn something about what’s going on in the molecule okay so here’s like one of these complexes which I showed you
Initially I see3 so that’s typically how these Spectra look like and has to get used to it to read them but just tell you that okay well if you have this part here which comes from our ground seed bleaching stimulated emission and then you have excited state absorption
And this excited state absorption is actually telling you okay there’s are you have populated MCT States right so that’s that’s known that has been investigated so I just mentioned this here and then you record all the spectrum is a function of time of delay time you see basically at very
Long delay time you back to like almost back to to well that you have no speal changes basically the system is returned to the ground then you can do some analy analysis of these traces and get some time skills right so you get this soall or Decay Associated spectr and then you
Get time skilles for the pro and you have then to work on get the processes for these time skilles okay well if you look at now to our cpmp liant you see okay well it has looks well compared to that one looks kind of not so interesting because like
All the changes are just in the uh um uh ground bleaching simulated emission Peak and there seems to be no excited state absorption that spectal region so then you would say okay there should be no long lift mlct okay um now what what how how to relate this
Now to our calculation so um let’s first see what the absorption spectr well we already looked at this so um absortion spectr are not really very telling here so it’s basically looking the same way depending I mean independent on what sort of parameters use okay that’s I will want to focus on
Dynamics right the photo I mean dynamics of the photo exitation and for this we need potential surfaces right so basically this is a potential surface cut along a breathing mode of the Fe nitrogen distance it’s about 200 centimeter in frequency all right so you see here uh the triplet
States and the um are the signal States and blue and red and blue respectively and uh then you also see these um uh triplet MC state which is or you see okay that’s the MC State that’s that’s the parameter set for the ground seat calculation well it looks like put
It like this it looks like spaghetti in first place uh but if you look at this closer you see for instance that in this uh cpt2 um uh derived parameter set the MC states are uh below the mlc set right en so it’s basically like this triplet MC
Here and this triplet MC is over there which would mean that there’s no I mean you would expect there’s no long lift m c states in the right hand side case and here they could be because the triplet states are higher all right um now well why there
There transition at all well mean you have non-tic copies basically if you look at this this curve well it’s it looks nice but but well it’s complicated but it even gets more complicated because you have not only spin orbit coupling between the single state blue one and the red triplet States uh but
You also have the breakdown of the B approximation right so basically you have when this is just a one-dimensional cut of this 3 nus 6 dimensional potential surface and if I just take another mode say well what you see this cut is So-Cal tuning mode right so if you look at go
Here okay you this state it’s kind of crossing here or that could be any of these pairs of states which you see here over here in the left hand side but then you have coupling modes which normally uh lead to um well if you move along these coupling modes reduce the symmetry
Of the molecule and uh then uh or if you take into this account this coupling mode you see you get well let’s zoom in here you get this photochemical funnels these con intersections uh which just tell you that you have very rapid uh Dynamics or can have very rapid Dynamics
Uh say starting on this higher state and then ending up in this Lower State here so that’s that’s a punch line This picture is much too too simple uh just gives you an idea about the relative position and an overal picture but in order to to really appreciate what’s
Going on are you need to take into account other coordinates uh notably these coupling coordinates and then this will give you very rapid processes that’s well known in order to to um I mean to investigate this now we teamed up with the group of latisa Gonzalez from Vienna and they did
Surface shopping simulation which is essentially um like a non-tic dynamic simulation to find out whether the two parameter sets one derived from the ground state calculation and one derived from the cp2 calculation lead to different Dynamics okay I don’t I mean my time is running out so I don’t want
To go too much into detail uh but so what you see here is always comparison population of the ground SE set that’s a and the population of the casp2 set that’s B and uh all that’s adiabetic populations that’s the diabetic population that’s probably easier to understand which is see here is
Essentially that you have a large population of and triplet mlct St um for this first parameter set from the ground right so that would mean uh that you trap the system in a triplet MCT State and um well which comes along with little change in in geometry and whereas for this other
Parameters set you have very rapid depopulation and you go back to to the ground so you no trans to Triplet population for I mean trapping in the triplet ABS population and that’s essentially in accord with experiment right which did not remember we didn’t find excited datee absorption so there’s
No long lift MCT uh population and therefore like we would be in favor by comp person with this simulations with experiment uh it would be it would be in favor to use this caspt2 derive parameters one can go like more into detail analyze different Bond lengths
And so on so you really have like large Bond elongations here for this cp2 set or which is a blue curve with respect to the equilibrium or the frankon geometry value and you you don’t have so much geometry change if you look at um the uh our ground
Derived perameter set so well that’s all in the court essentially ni try to summarize the what happens in this kind of cartoon it’s really cartoon because you might have noticed that we did not take into account the qued states which is kind of tricky computationally uh so
What happens is that you make an exitation and then really the the MC state is below energetically the mlct state and uh so the system goes over here to this blue State might be trapped and that’s also evidence of experiments in This qued MC State and then returns back to the to the
Ground ah yeah to be continued put it like this so to sum up let me just make a few points um what what I tried to convince you is that this functional tuning is an efficient way to obtaining a system specific density function um and uh which is system
Specific in that for this particular systems system fulfills certain conditions which should be true for exact con so it’s not like one an arbitary functional which from the shelf so it’s a functional just for your particular system that’s good um there are many many applications and
I think it’s mostly standard to use this functional tuning nowadays and while including small molecules even to solid state materials and yeah also our work on rium photosensitizer where it shows really considerable Improvement um with respect to Performance so if you compare to of the Shelf standard functions however
Um going to this Earth abundant elements like the iron is is kind of tricky because there you have a you have like this mixture of transitions of different characters like MC and M which I mean are at the heart of the photophysics of these complexes and which cannot really I mean addressed in
In in a simple way in that sense that you can just use this OT optimal tuned range separated typ functional as a black box and to just get some parameters and then you done why we have seen that okay you can do very well for absorption Spectrum but then for the
Dynamics which is not visible in the absorption Spectrum you you might be very wrong right so therefore um are you need some additional input from experiment or higher level calculation to really find Optimal parameters okay and with this I’m I’m finished and just thank my co-workers and which
Actually did the work a in particular particular a also sge and our experimental co-workers and uh yeah thank you for your attention and I look forward to your questions