Molecular features contributing to the non-ideal mixing behavior of deep eutectic solvents
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Title of Series | ||
| Number of Parts | 22 | |
| Author | ||
| License | CC Attribution - NonCommercial - ShareAlike 3.0 Germany: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal and non-commercial purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this | |
| Identifiers | 10.5446/57513 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
Leibniz MMS Days 20222 / 22
11
19
00:00
Solid geometryGrothendieck topologyMaxima and minimaMathematicsPhysical systemMultiplication signObservational studySet theoryDistribution (mathematics)Right angle2 (number)Atomic numberProcess (computing)Routing1 (number)LiquidCross-correlationFood energyWeightForcing (mathematics)DistanceModel theoryRule of inferenceEvent horizonSurfaceClassical physicsScaling (geometry)Algebraic structureGroup actionMusical ensembleAverageGoodness of fitEstimatorResultantExistencePoint (geometry)SupersymmetryHand fanMoment (mathematics)Degree (graph theory)Translation (relic)INTEGRALStudent's t-testOperator (mathematics)Gastropod shellPopulation densityCategory of beingDifferent (Kate Ryan album)Green's functionEuler anglesMaxwell's demonArithmetic meanMonster groupCoefficient of determinationRegulator geneMatching (graph theory)Line (geometry)MeasurementWater vaporOrder of magnitudePotential energyState of matterFreezingGoogolPlane (geometry)ComputabilityThermodynamisches SystemSocial classMaterialization (paranormal)Slide rulePosition operatorClosed setMereologyNichtlineares GleichungssystemMany-sorted logicCartesian coordinate systemModulformCurveVolume (thermodynamics)QuantumDivisorFluxGamma functionAdditionOcean currentMathematical optimizationElectric dipole momentFunctional (mathematics)CoefficientPower (physics)Condition numberDensity functional theoryVelocityPhysicalismDirected graphVapor pressureKörper <Algebra>Boundary value problemLinear regressionPrice indexMass flow rateElement (mathematics)SummierbarkeitConnectivity (graph theory)Mixture modelChemical equationProbability distributionFrequencyKinetic energyProduct (business)Numerical analysisRotationParameter (computer programming)Fluid staticsMechanism designTheoryAutocorrelationSchmelze <Betrieb>Generating set of a groupLocal ringTheory of relativityThermal conductivityAngleFraunhofer-Institut für Physikalische MeßtechnikDirection (geometry)Lecture/Conference
Transcript: English(auto-generated)
00:30
Okay, let's start. Okay, I want to shortly talk about the IOM and what I do especially in this debate tactics solve what I am. I see them in this room so shortly. We use ion beams plasma
00:46
photons or electron beams to generate surfaces functional surfaces. And this is mainly focused on experimental studies so far. And we're doing also some software development and modeling and
01:02
simulation. Stefan Gersch is doing some software development for our tools and which we need to make it structures on the nanometer scale. Martin Rudolf is doing some plasma modeling. I will talk later a little bit about what he will do. Then we have Stefan Meyer and maybe people
01:22
know him, because the states were 2018 in light take and he doing computational physics, for example, developing cosplay models or models to describe inorganic materials, which is biocompatible. And furthermore, myself, and I doing electron beam and photochemical
01:44
use reactions. I study these things by applying quantum chemistry approaches, including also post Hartley-Fock methods for static, for systems with strong static electron correlation, and main topic from my service also to fit in people tactic solvents. And this is in a moment
02:04
expertise at our institute a short overview. And in the perspective, we want to build up a machine learning expertise in close cooperation with the center of source scalloped data and artificial intelligence, so that our experts in there are integrated in the large team. And
02:21
furthermore, we want all the future indicate how we see results in the IT team into the collection sent from Leipzig, which will be integrated in 2026. But now I want to talk about some research activities of myself, and what are the protecting solvents you see here to example, or what are the taking solvents and example is a mixture some principle of a cheap
02:45
organic salt, which can you can combine this organic compounds and then you get a liquid at room temperature. And these are most popular because clean chloride is produced on the medicartan scale and fed to animals is very cheap and environmentally friendly. And it's
03:01
solid up to 300 degrees and you can mix it for example, which is also solid above 100 degree and then you get a liquid at room temperature. And as you can imagine, these liquids are very cheap, have a low vapor pressure and are environmentally friendly, even some bacteria can survive in these solvents. So and now you can ask what is a deeper technical science
03:24
principle initially when it started in 2003, when they were reported by effort the systems then they said, okay, it's a mixture of two compounds with significantly increased melting point compared to pure compounds. And you say, okay, it's also salt and water and that's right. So then the phrase appears what is a deep or a tactic solvent. And what you see here is
03:47
in principle, the next panel point of view. So as you see here, a typical or tactic system, which has an idiot behave have also a melting point can have a melting point decrease. And here you see the equation for it. And the most important part is the first one,
04:06
and which is important. So you see it depends on the difference of the melting point enter the melting entity and the melting point of the form of the pure compounds. And when you have idiot behavior here, this factor gamma is the activity coefficient is one. And if you have
04:27
value smaller than one, then you have a non-ideal behavior, which results in decreased melting point. So when you say now what's deep in principle, you should have a strong non-ideal behavior. This is a gamma, which is significantly wrong. Unfortunately, this data is not easily obtained.
04:45
And therefore, at the moment, they are still the phrase used, the phrase deeper technical or it might be better when you report for the first time, because you do not have all the physical chemical data to say if it's really an evil technique. So where fields of applications
05:03
I have selected for applications examples, you can use them to generate unique nanomaterials. Here you see golden star and the golden star is covered completely by high index planes, which results in high electrochemical activity. You can also get smaller nanoparticles of about 5 nanometer size in these liquids. You can use it as a solvent catalyst that the product
05:25
forms a separate phase. And you can also use it to generate high stretched and non-miotile ionogales, which is also an application you can use for pressure sensors and other things. And you can also use them for front-type quantumization at low temperature with full
05:42
conversion, because you have obtained a stable front. And this makes them interesting for applications, for example, for proteins. Okay, now you can ask, does it contribute to the low melting point? And when I started, the literature was suggested that the charged localization according to hydrogen bonding between
06:02
the halide and the anion and hydrogen bond donor majority is responsible for the decrease in the freezing point of the mixture relative to the melting points of the inner components. And this is something you can easily check by first principle, simulations. There we have seen that in case of ionic liquids, which are liquid salts at
06:22
room temperature, that about eight ion pairs are sufficient to converge dipole moments under periodic boundary conditions. And we have carried out some first principle molecular dynamic simulations with three systems and make subsequently a part zero charge analysis, this case the Hirschfeld part zero charge analysis, which is based on electron density.
06:42
And the most important thing is here, what you see, this is the organic compound. So when you look at some of the charges on the organic compound, and you see in the most popular example, this is this one, this is the charge on the organic compound is overall mechanical. So only in case of the oxalic acid, you see a significant negative charge located
07:04
there. And therefore, you can say, okay, it might be not the origin for the urea system. Okay, how and what you see here is this linear trend in the negative charge located on the organic compound. And what you can do now is you can look at the hydrogen
07:26
bonds in the systems and you have similar on the right side is this is the case from hydrogen oxygen, then you can just look at the distance between the hydrogen bond donor atom and hydrogen sector atom. And you see here the distribution function value of one means a statistical
07:45
distribution. And you see here the first solvation peak, and the blue one is between the anion and the oxygen atom of choline. And the second one, the green one is the one between the anion and the nitrogen atom of the hydrogen bond donor atom. In this case,
08:03
here's the nitrogen atom, here the oxygen atom, and so on. And what you see is in case of the urea system, the hydrogen bond between the cation and anion is stronger than between the anion and the organic compound. And therefore, the charge transfer might be mainly
08:21
between the cation and anion. However, when you look in the last example, then you see here, the hydrogen bond strength is stronger between the organic compound and the anion, and therefore, significant charge is transferred to the organic compound of the system, and you have a negative charge located. And this is also important data for force field flow elements.
08:43
So now, we have seen, okay, what doesn't contribute to the low melting point, and we can ask, okay, what might be something what we can use to discuss, this is a so-called energy landscape, so it's been, as I said, portion of the potential energy surface that represents the liquid or glassy region, as unlike the portion, as it's added to the crystalline
09:02
solid, a large number of minima of Orion gaps. So if you have a shallow energy surface, and you need low kinetic energy to push the atoms around, if you have a steep potential, then of course you need a high kinetic energy to push atoms around. And this was, for example, can be used to discuss properties, and the potential energy is a high complex compound,
09:24
and you must select, if you look on examples, here's one example where you see the potential energy of the surface of an ionic liquid, and what you see is when you replace this acidic hydrogen atom by a metal group, then you place an attractive interaction by the powers of form,
09:44
and the experiment groups have expected that this would decrease the melting point, but the opposite is significantly increased. And what you can look now is, first by static chemistry, look at the potential energy surface for the anion flipping around the cation, and you see you have an overall shallow potential energy first when you have this acidic
10:03
hydrogen atom. When you add the metal group, then you see that you have a high activation barrier in front, and therefore the anion is fixed above and below the plane. And you see this not only for the isolated ions in the gas phase, you see this also in molecular
10:22
dynamics simulation, and then you can understand, for example, why you have this property. And now we can apply this to study our system, and here we have different kinds of hydrogen bond donor atoms in our system. We have, for example, this hydrogen bony atom, and the experimental
10:43
group suggested that you have here a fast rotation of this chlorine oxygen atom, resulting of the side change, and this results in a fast hydrogen bond dynamics zone. And in this case, in our area, we don't, we're not considered, but we also investigated this, and what we
11:04
can do then is we can look at the lifetime of the hydrogen bond times to check where we have the fasted hydrogen bond dynamics, and this is just done by autocorrelation functions, here, so you just simply define a hydrogen bond criteria. So, and this is, for example,
11:20
you can take this from the distribution function, look at the first minima and say, okay, then you have the distance cut off, then you also make an angle cut off, and this then defines the hydrogen bond. If it's there or not, then you make an autocorrelation function, integrate it, and then you get the lifetime. And if you're doing so, then you see that the fasted hydrogen
11:41
bond dynamics is not here, and this chlorine oxygen hydrogen atom, in principle, you find that the fast hydrogen bond dynamics is in these two hydrogen atoms, and it's a very fast dynamics, and so you have a fast hopping of the animal. So, however, this only explains, okay,
12:00
and now you can look to the experiment, and the experiment values you see here, you have the mixture of urea, and you see when you introduce one metal group, you see a slight change in the melting point. If you add two metal groups, but you still have these two hydrogen atoms, you see it's also slightly increased, but you have a different strong
12:21
effects if you really remove one of the hydrogen transposition. So this now explains why you have a liquid at room temperature, but not why you have non-EDL mixing behavior. And this was then the next step. So we study homologous areas of de-protactic solvents, starting here, you see we have urea, we have thiourea, material urea, dematerial urea,
12:50
and another material urea. And for this, we also expand the values known, which allows us to force field. And what we can do now is we can employ theory to data mine the relative of the
13:05
chemical activity from the molecular dynamic simulation. And what we obtained, you see is this green curve, and you see a good correlation. And what we can also then do is we can
13:22
calculate the experimental activity coefficients, and these are the red values, and you see this is also a good correlation. And please, this looks now like it's strange you have an inverse behavior, but actually this relative is the relative of the chemical
13:41
activity with respect to the molarity, and therefore you have an inverse behavior in both because it's not the activity coefficient is a directive of the chemical activity. Okay, and then the next step, we develop a polarizing force field. This is done mainly
14:01
by my PhD student Omichai Stapor, and this force field is based on a truth approach. So we have a particle on a spring, which has some charge on the heavy atoms, which models and the polarizability of the systems. And what we see is that we need a screening factor
14:23
between the oxygen atom of choline and the anion. Otherwise, we do not get good results. As you see here, the light green color, this is without this damping factor, and the red one is for NASA RDF. And as soon as we apply this additional screening factor and parametrizers,
14:46
then we see that both of the arpeggios match this archer, which, as in our reference, is a DFT-based molecular simulations. So then we have the force field, which will produce very well the structure of a DFT simulation, and also the dipole moments and polarizabilities are
15:03
also fitted on chemistry data. And then you can also use it to look on some experimental properties if they are reproduced very well. And there we use diffusion coefficients, which were available in the literature. And this can be determined by velocity out of correlation
15:22
functions. And you see we have an overall good match for a force field. And then we can also use the in-coupled relation to calculate the conductivity and also the pressure-tension out of correlation function calculated with pure viscosity. And you see we have also good agreement so that we have now a force field, which reproduces very well the structural
15:44
data, but also dynamic properties, and now is now a good study. And then we can, for example, apply this to study some dynamic properties, which is not available by first Winston-Muller because the simulation of the system size is not so good. And here we have, you see,
16:05
the electric current out of correlation function of the ions. And what you see is this curve here, the cross-correlation of the electric current out of correlation function between cation and anion. And what you see if you have, on the short time scale,
16:21
joint migration of cation and anion, but on the long time scale, you do not see any joint migration. Okay, so let's sum up this result of the demo checking solvents. So we see that we have a fast high chamber dynamics at Rhea, which constitutes the melting conclosed room temperature of the lean, the incorporation of the anion into the
16:42
hydrogen bond network of Rhea is crucial for the observed melting point depression. And the ion correlation is only observed on short time scales, so there's no joint ion pairs moving together, and these polar isophores need additional damping function between choline and the anion. So now I want to show the talk about the results,
17:05
which was done by Martin Woodolf, and he's doing a plasma simulation as our institute's only short overview because it's not done by my group. And so, and this is focused on high power-imposed macron sputtering. So here you see the picture, how it looks like when the
17:25
when the plasma is visible, and this is in principle the setup of the device. So we see we have a cathode and the anode. On the cathode is the target, so the material you want to put on the substrate, on the anode, and you see behind the cathode, you have a
17:41
magnet. And in high power-imposed macron sputtering, you have only short pulse, which is active on the microsecond time scale, and again, it's some plasma, and then material is deposited towards the anode. And yeah, and this is the systems which Martin investigates, and for this
18:04
he uses global discharge models, so it's about, it's based on volume average plasma chemistry, and the advantage of his models is that he can run these simulations on the laptop. And for example, what he used now for this model is, what you want to understand is then
18:25
how is the flux of the material on the substrate, and how you can optimize it. And there he used the semi-empirical model, the ionization region model, and this is in principle fitted to some easily accepted properties like voltage and current. And with this, then he can use
18:45
this model, for example, to determine what is the optimal process conditions to have a large deposition of material on the anode material. What his IPM model uses makes also several
19:01
assumptions, and one assumption is that we have a Bmax volume distribution of the electron, and therefore he used a model which describes the electron, or calculates the electron energy distribution, and you see here the assumption from the IRM, these are these black lines in this
19:25
graph, and then with this other model, with this obelix, he calculated the electron distribution, and what you see that the colored lines, that's the mouse now here, the colored lines matches overall very good the assumption from the IRM, and so you can say,
19:45
okay, that's the IRM model is overall valid for these things. And what's important to know that the obelix model is about four orders of magnitude slower than the IRM model. Okay,
20:01
but when you want to know something about plasma simulation IRM, then best is to contact Martin, and yeah, so then finally I want to thank my co-workers, Charles Stiebel, the chief of funding, my cooperation partners, and also thank you for your attention.
20:35
Okay, so essentially you're just based on
20:42
molecular dynamics models. Yeah, molecular dynamics models. And the question is, so just, I'm asking from the context of my trans-institute, and what possibilities do we have to, let's say, extract parameters like
21:02
solution effects or something like this from your approach? So you need activity coefficients, for example. Yeah. Okay, so the activity coefficients can be
21:24
in principle calculated by these Kuiper-Puff theories and by integration. So this is in principle, what you do is you have here the, this is the distribution function. This is the molarity and this is the chemical activity. And what we have done, what you see, we have
21:46
plotted this one because when you want to get this, you must integrate over the composition. But this is in principle a possibility what you can get from the simulation activity coefficients.