 after the lecture, I will be here available with this. Thank you, sorry, Florian. OK, so thanks, Albi, for that. So let's make a start then. So the London of Franco is going to carry on a second part, photo-thermal microscopy and spectroscopy. OK, thank you very much. For the beginning, we will refresh a little bit the material that we covered yesterday. And this was about thermal lens microscopy. And I have promised to show you the last three slides in five minutes and discuss the question of, let's say, expanding the applicability of thermal lens spectrometry in the sense of or by enabling additional wavelengths or excitation. And of course, because of the limited number of available lasers, one of the ways of doing it is, of course, using incoherent light sources, such as a xenon lamp, for example. And here you have a scheme of a thermal lens microscope, as we have seen it yesterday. But instead of using a laser source, we use optical fiber to bring in the entire spectrum of the xenon lamp. Of course, the power of such a lamp is relatively low. So in, let's say, at around 550 nanometers, which would correspond to one of the argon laser lines, within one nanometer bandwidth, we can get about one milliwatt power. And if you remember the enhancement factors that we spoke about, this would be less sensitive if we do measurements in water samples. This would be less sensitive than spectrophotometric determination. In addition, we have to take into account that in the thermal lens microscopy, we don't have one centimeter optical interaction length, like in conventional spectrophotometry. But we reduce it by a factor of 100 to the distance of about 100 microns. But however, when we go from the macroscopic system to a microscopic system, we have to take into consideration some phenomena which were neglected in the case of macroscopic thermal lens. Particularly, we have neglected the conduction of heat into the windows of our detection cell. Of course, when you consider these few microns, maybe some 10 microns, in comparison to one centimeter, this is, of course, negligible. But when you do it in comparison to 100 micrometers, this does not become negligible anymore. And we can see that the temperature rise due to the photothermal heating appears also in the surrounding of our sample. Of course, this can generate thermal lens in this material, not only in our sample. So here, we got the idea that eventually, by exploiting this temperature change, we can enhance our thermal lens signal. So we have proposed, let's say, a so-called three-lens system. Here is our sample, generating a thermal lens. And then we have additional thermal lenses in the front and at the rear of our sample. Now, by using appropriate materials with appropriate photothermal properties, high change in the index of refraction with temperature and low thermal conductivity, we can generate thermal lenses comparable to the thermal lens that is generated in our sample. And here we have some examples for different materials octane as a representative of organic solvents. We have seen yesterday that they have extremely high enhancement factor. And you can see that compared to the adiabatic, adiabatic means that in the vicinity, we have a signal, we have a material with exactly the same thermal properties as our sample. So in this case, we would get such a signal. If we put octane, we increase considerably our signal. Polystyrene is somewhere in between. But in fused silica, we get a decrease in the signal. Why? Somebody has listened yesterday, remembers what we discussed yesterday. Silica has a positive dn over dt, while n octane has strongly negative dn over dt. And this is that the sign of the thermal lens signal changes. So here we have what we call a positive thermal lens signal. In the vicinity, we have a negative thermal lens signal. So we have something we don't want in this kind of system. But using organic materials, we can significantly improve the sensitivity of our system. And we have designed such a similar detection cell where we can control the thickness of our sample. And we have two unmixable solvents, one on the bottom, one on the top. And here we have the results showing just a single layer, just a sample in this system. Very flat calibration curve. When we add one layer on the bottom, we already get an improvement with two layers. This is now with the UV lamp excited already. And with three layers, so we have our sample. And the organic solvents on the bottom, which do not mix, we get quite a strong enhancement of almost one order of magnitude. So by using light sources such as Xenon lamp, which has an overlap in the emission spectrum with the absorption spectrum of our analyte. In this case, it was iron complex, iron two with 110 phenanthroline or ferroene short. This is the absorption spectrum. So we were measuring somewhere in this range around 510 nanometers. So this is a suggestion for improving sensitivity in thermal lens microscopy when using incoherent light sources or low power light sources, like lead emitting diodes, which gives you more versatility for your applications. And we were able, still, in a 100 micrometer cell, to detect absorbances on the order of 10 to the minus 5 absorption units, which is, I believe, still better than commercial transmission mode spectrophotometers. OK, our yesterday's lecture is over. I would only like to point out some of the literature, which is in the PDF also. I think as of half an hour ago, both of my lectures are available online to you. And you can find some suggestions for readings, including the Bible of Analytical Thermal Lens Spectrometry by Stephen Bielkowski and some other references which were published recently. Maybe something is still missing. No, this is the last one from mainly microscopic thermal lens spectrometry. And of course, if you have additional questions, you can ask me anytime, today, till Friday, or just send an email if you don't have questions ready today. So now I would like to move on to the applications of thermal lens spectrometry and microscopy. I have added here for bio-imaging, and particularly for bio-analysis. Bio-imaging by thermal lens spectrometry is not very common, and I think only one paper appeared which can be really classified as bio-imaging. Microscopic analysis of a single cell was reported. Of course, there are many, many other variations of photothermal techniques that are used for imaging, detection of nano-objects. So labeling proteins and biomolecules with nano-silver, nano-gold particularly can be used. And you can refer to works of Brachim Lounis from Bordeaux. He did a lot of work in this area, but we cannot classify that as a clear thermal lens spectrometry or microscopy. So here you can see the paper that I was referring to. It was done by the group of Kitamori in Japan who was basically the pioneer in thermal lens microscopy. And here we can see a single cell cultured in a microchip. This is a classical fluorescence image based on the, of course, staining of the material in the cell. And here we see the thermal lens scan, two-dimensional thermal lens scan with one micrometer resolution, 30 micrometers in both directions. And we can see quite nicely the cytochrome distribution within the cell that corresponds very well to the fluorescence image. And after the cell apoptosis, of course, the cytochrome C is everywhere in this microchip. So this is the presentation of a capability of thermal lens microscopy to do imaging of single cells. But we will now focus more on to the detection of the fluids inside the cell or other biological fluids. But first, before going there, let's refresh a little bit what we were saying about the Vanguard methods. So detecting a signal which will tell us that something is present in our sample. And now we are going basically from satellite imaging down to bio-imaging at the cellular level, so to say. But what we see from the satellite is actually the algal bloom in the northern sea, or my geography is bad. But this is part of Great Britain here, so Atlantic. And we can see the algal blooms due to the different phenomena that happen in our environment. Too much nutrients in the water, the change in temperature. But of course, biologists are not interested in phenomena as they are seen from the satellite, but they would like to know what is going on here, why this is happening, what is the consequence of such phenomena. And a simple way of doing it, which also shows the superior sensitivity of thermal lens spectrometry again, is by looking at the pigments that are present in microorganisms in the phytoplankton, so to say. When the algal blooms happen, the phytoplankton cells also die, and the contents of the cell is released into the water. And of course, pigments absorb indivisible, and we can very nicely detect them by a sensitive technique such as thermal lens spectrometry. So here we see, we have caused the decay of the microorganisms by inducing some cytotoxin. This is some, what does it say? Oh, addition of poly-APS. Poly-APS is the cytotoxin that comes from C-spungy, and it causes the breaking apart of the cellular wall. So the pigments, in this case, particularly carotenoids, are released, and we can observe very nicely an increase in thermal lens signal, while spectrophotometry gave us actually no response because of insufficient sensitivity. So this is a Van Gogh method which detects carotenoids or pigments in general and doesn't tell us more about the process. But if we are interested in the beginning of such a process, this is a good technique because we can use it to observe that something is changing in this water. After we see this, then we can use, as we described yesterday, the retroguard methods to look more in detail what is going on. Here we see another variation of a thermal lens spectrometer, which is interesting for the fact that we actually have two thermal lens spectrometers. Here we use a prism to split the two beams, and we have one system here, and the other one is here. Why do we need this? Because we need to control with the control population of microorganisms what is going on because this is a growing, it's a living sample. So we must have a reference. And basically what I want to point out is that we can construct two thermal lens spectrometers by using only one excitation laser. OK, now what are the species in such? You see here we have only mainly diatoms, very nice organisms. And what biologists like to do is not to use a microscope to identify them, but they like to use chemistry. And what they call this is a chemotaxonomy because each species has a characteristic fingerprint of pigments. And these are both carotenoids, fucoxantin and diadenoxantin, and they have a dual function. They also transform one into the organism transforms one into the other. One is used to harvest the solar energy. The other one is used to protect the organism against solar light when there is too much light. So during the daily cycle and night cycle, the ratio of these two changes a lot. But first of all, if in such a chromatogram a biologist would see fucoxantin and diadenoxantin, they would say, OK, this is we have diatoms here. And here we have two diatom species. One is skeletonemacostatum, and the other one is phyldaptilum trichornutum. I'm bad in Latin, but here you see this is phyldaptilum criodutum, and skeletonemacostatum looks like this. So by looking at the chemical composition of pigments, you can predict which of the species you have in there. But still, you see, between these two, you cannot distinguish. So we were thinking, why don't we do something? And perhaps we will have a clue how to distinguish between diatoms in between diatoms. So to do this, we have to be able to analyze a broad spectrum of different pigments. I will not go into many, into identification of them. Here you can recognize something that we described already yesterday, alpha carotene, beta carotene, lycopene here. What you can also observe is the increase in the baseline. This is not the drift in the signal, but the reason is change in thermo-optical properties. What we use here to get all these pigments out of our chromatographic system in a reasonable time. Still, this is a very long time. Not something we like, but for such systems, it's acceptable. To get all these pigments that are highly retained on the column washed out, we are applying what we call a gradient illusion. So the gradient is in the fact that the mobile phase contains more and more organic solvent when we go towards the end of the chromatogram. And organic solvents means better thermo-optical properties, higher signal. This is what we have seen yesterday by the enhancement factors. So this is not a drift. This is actually the change in the composition of our solvents. So you see, we are able to separate a lot of such pigments. Now, the problem that arises in gradient illusion, and we will see this later on in some other applications, the problem is the signal noise. The mixing of two solvents generates inhomogeneity in our eluent in terms of thermal conductivity and dn over dt. And this mixing, which is not complete, causes increase in the signal noise, which, as you can see here, with the gradient going from eluent a to eluent b with higher degree of organic phase, we have the highest noise here at the end of the gradient. And then, with time, this decreases again. So here, in a part of chromatogram, we have a very significantly deteriorated limits of detection, so to say. And this doesn't change even if we change the rate of the gradient. You see here the two simil, actually the same flow rates. But one gradient is faster than the other. It even seems that the faster gradient works a little bit better. Here, we have a slower gradient, but the noise is more or less the same. So this is a drawback. That's why we prefer what we call isocratic elution in chromatography with the constant composition of the eluent. But there are ways of improving it. And particularly, we can improve it by using longer separation columns. These are packed columns. Longer column means better mixing of the solvents, so less noise. And here, you can see on the longer column, this is the ratio between the spectrophotometric detection to thermal lens LOD. The lower thermal lens LOD, the higher is this ratio. So we can see that in the part of the chromatogram, when we apply the gradient, this ratio is the lowest. But then it reaches, again, relatively high values, about 20 times improvement for the most sensitive detection of carotenoids with thermal lens spectrometry. When we go to shorter columns, of course, the LOD of thermal lens increases. That's why this ratio is lower, but still showing deterioration in the part of the applied gradient. This also reflects in the limits of detection for different carotenoids or different pigments here, even chlorophyll. Gradient is always worse than isocratic. Of course, with isocratic, we cannot separate all. Why the chlorophyll is much, much higher LOD? What's wrong with chlorophyll? It's fluorescing, yeah? So the heat yield with chlorophyll is much, much lower. Chlorophyll is known as highly fluorescing. Pigment, so we are losing energy, that's why. So chlorophyll, it's better to detect it with fluorescence. And still, despite the problems with the gradient, here is the chromatogram detected with the diolta ray detector. You see very low signal to noise ratio and very noisy and unstable baseline. So this shows the advantages of using thermal lens spectrometer, despite the problems with the gradient illusion. Now, this is again the same chromatogram on which I have marked different sections, A, B, C, which show completely flat with the diolta ray detection. Now we will look, what do we see with thermal lens detection? Here, for comparison, is the signal from the UVVs, diolta ray detector on the bottom. Again, both species, skeletonema and phyodartilum. But in all the three areas, we can see additional pigments, which are not present in the original chromatogram because of insufficient sensitivity. So you see several pigments appearing. And we can say that several of them are characteristic for one or the other species. So based on the analysis of what we call minor and trace carotenoids in diatoms, we can, in principle, distinguish between productilum skeletonema, which according to the known chemotaxonomy cannot be distinguished because of insufficient detection system. OK, from here on, now it's up to biologists to apply this and to demonstrate that this is useful. We will now go on to the biomolecules in body fluids and also in intracellular solutions. One of such interesting molecules is bilirubin. Bilirubin is one of the endogenous antioxidants in our body, which is formed from the hemoglobin. The porphyrin ring of the hemoglobin breaks, and this then forms bilirubin. Bilirubin is, as I said, endogenous antioxidant, but it's also neurotoxic compound, especially with newborns who have this Junis effect I don't know exactly how the medical doctors call it, but they become yellow. And this is because of bilirubin, so they have to be treated, usually with photodegradation of bilirubin and other treatments to prevent damage of the neural system. Therefore, it is very important to understand the behavior of bilirubin in our body and our colleagues here from University of Trieste were interested in the transport of bilirubin across the cellular membrane. And the problem, of course, is detecting free bilirubin. So until this work, all the experiments were done on the model molecules, which cannot be exactly considered the same as bilirubin. The problem is that the solubility of bilirubin is on the order of 50 to 100 nanomolar. And this cannot be detected by spectrophotometry. With thermal lens, we have pushed the limit of detection in this system below 1 nanomolar. So we were actually able to monitor the process in the real physiological situation. And we can see that the concentration of bilirubin in the substrate did not change very much if the ratio of NADH to NAD+, was low. This is a normal situation in the cell. And there was no transport into the cell because there was no decrease in the bilirubin concentration. But when we have stressed the cells with either high lactate content, and this is an interesting observation, when we work hard, we work physically, we produce lactate. And during the work, also radicals are produced. So our body, interestingly, opens the channel to transport bilirubin into the cell to protect it. Supposedly, this was our explanation. But also, drinking alcohol, so be careful. Drinking alcohol also causes radicals. But our body has a system. It increases the intake of bilirubin. This is in both cases. Now, when we have used a specific antibody which blocked the active site of the bilitranslocase, bilitranslocase is the protein in the cellular membrane that actually carries bilirubin from the outside to the inside of the cell. And we have shown when this active site, which was known before, was blocked, there was no transport of bilirubin into the cell. So this confirmed for the first time the active transport mechanism. The mechanism is active. It needs protein that takes the substrate and brings it into the cell. It's not like different ions that go through ion channels. This is nonactive transport across the cellular membrane. So this was one of the first interesting works in this respect. And we have continued. Oh, this is an extra slide, which will not hurt. But what I want to show is that we have improved the detection system by coupling in the previous cases. This was just the model solution. We have a physiological solution in which the cells were grown. And we have just added the bilirubin. So there was no question of selectivity or possible interferences. Now, we were interested in the levels of free bilirubin in the blood serum, which for human blood was not reported yet. And there, of course, we have to separate bilirubin from all the other possibly absorbing compounds in our blood serum. So we have performed HPLC separation. And here you can see the bilirubin, the peak corresponding to bilirubin. And this is dioderae detection. You see the big difference in the signal to noise ratio. What is also interesting is that in the standard sample, this is a synthetic standard of bilirubin. We can see three different isomers of bilirubin. And this clearly confirms that there is no cheating here, because in nature, only one isomer is always synthesized. Our body doesn't synthesize three different isomers, which cannot be avoided when you do this chemically. So this is clearly standard because it has three isomers. Our sample contains only the pure peak. And secondly, with this very sensitive method, we are able to monitor degradation products. Bilirubin is very sensitive to light and also to other oxidative species. So when we see some degradation products, here we know that the sample preparation was not performed as it should be. You can see extremely low limits of detection. And with this, we have, for the first time, determined the free bilirubin in the blood serum for three volunteers here. And it's close to some animal samples that were analyzed before. Remember the values here because we will come to another important discovery. But first of all, I would like to show you another application, which is basically the termination of bilirubin and its counterpart, biliverdin. Because bilirubin is transformed through biliverdin and by biliverdin reductase back to bilirubin. So this is an important cycle in human body to understand the antioxidative potential or activity of our cells. So now, again, we are facing the problem of the gradient illusion. We are changing the methanol content in our mobile phase from 67% to 97%. Of course, again, we see the increase in the thermal length signal, the baseline, and here a big, big noise, which of course doesn't hurt because we are interested in biliverdin, which is alluded here. Then we have this noisy part of the chromatogram. And then at the end, we get illusion of bilirubin. These are the concentrations. So you see that we can work well below the levels in normal cells. So we can do both pigments at the same time. And finally, this is what I consider the most important contribution of thermal length spectrometry to scientific discoveries. Most of the people were mainly reporting extremely high limits of detection. They have shown some applications, but it was never put into the routine analysis or applied for some important scientific work. In this case, we have actually confirmed that bilirubin does not exist only in the blood plasma, but also in the endothelial cells, vascular endothelial cells. These are cells that are in the inner walls of our veins. And they also represent about 60% of our heart. 60% of our heart is endothelial cells. So it's very important that we have antioxidants to prevent different vascular diseases, which are among the most important diseases of humankind, I would say. So here, again, you can see the demonstration of standard in the red and free bilirubin in our sample confirmed by UAVs spectrometry, also mass spectrometry. If you want to publish something, you always need to do mass spectrometry. Now, you will ask me why you bother with thermal length if you can do mass spectrometry. With mass spectrometry, you can also analyze the pigs which you know they exist. So before using thermal length spectrometry and analysis endothelial vascular cells, there was no indication at all that bilirubin exists in there. And of course, once you know, and you know the chromatographic condition, you can collect sufficiently large fractions to do mass spectrometry. Because mass spectrometry, you can hardly do on 10 microliters of non-concentrated sample. So this molecular peak and the fragmentation, which is also supported by the fragmentation of bilirubin standard, confirms that bilirubin is present in endothelial cells. And even more, we were able to show that we can modulate. We can increase the concentration of bilirubin by either activating hemoxygenase. Hemoxygenase will transform hemoglobin partly into bilirubin, or by adding additional bilirubin. So in both cases, the levels of bilirubin in endothelian cells were increased compared to the control. And what we can see here from the plot which shows the cellular antioxidant activity units, the EC concentration 50, the effective concentration 50, is about 10 nanomolar. And this is what we have measured in the three volunteers that, if you recall, that I have shown for the first measurements in the blood plasma. So this corresponds very well and shows that it gives additional credibility to the results. And if you want to read more, this is the reference. OK, now I will finish with applications of thermal and spectrometry in liquid chromatography, so to say. And the ultimate result in this is the so-called liquid chromatography in the extended nanospace. In October of last year, this was still without the numbers. It was published online. Probably now this is, you can get the full reference. What is the extended nanospace? We have nanospace in two, we have nanoscaling in two dimensions. X, Y, you see that this is 910 nanometers wide micro channel and 220 nanometers deep. And it extends over one millimeter in distance. That's why this is not entirely considered as a nanospace but extended nanospace. What is the advantage here? We can do separation of molecules. And this was done by using amino acids. Amino acids actually, why amino acids and thermal and spectrometry? You can find a lot of papers on detection of amino acids because about 30 years ago or even more, there was a big quest for the human genome. So the amino acid sequencing and detection of amino acids was very important. That's why thermal and spectrometry was considered as one of potential techniques to detect amino acids at the very low level. And since then, this is used quite a lot as a model system. So here we have this chromatographic system, which requires very high pressure pumps. And these are the nano channels shown on this microchip. Basically, by controlling the flows, we can control the injection. So sending our sample in this direction and stopping the flow and then starting the flow in this direction will inject this part of the sample into the separation channel. And of course, we need a very, very sensitive detection technique because we are talking about 220 nanometers, optical interaction length. Now, thermal lens spectrometry cannot be so sensitive. So they had to invent a variation of thermal lens spectrometry. This is what we call a thermal lens microscope. But it has a beam splitter that separates the probe beam into two branches. And only one branch is excited by the excitation beam. And measuring or observing the phase shift in the signal, which would result in the defocusing of the probe beam in such a short distance is not possible. That's why the signal arises from the interference because due to the photothermal effect, there is a phase shift in this signal, in this beam or the branch of this beam, which is later on combined with the unperturbed probe beam. And from this combination, we can observe an interference pattern, which is the measure of the absorption at this point. And this is how then you get signals for different amino acids. And look at the time scale here. You see, in 25 seconds, not 1,200 seconds, as we have seen before for carotenoids in diatoms, here in 20 seconds, you can separate different amino acids and detect them at relatively low concentrations, 10 to 100 ppb, or actually 370 molecules in about 350 nanometer deep microchannel. So this was done in a different system. So this would conclude now the combination of thermal length spectrometry in liquid chromatography, which we can give interesting and important results. But we could not really call it a Vanguard approach. Now I will return to this fast screening Vanguard techniques. And one of the first applications that we have done was the termination of pesticides, organophosphorous pesticides in different food samples. And what we know about organophosphorous pesticides is that they are inhibitors of acetylcholinesterase, which is in every living organisms. It's responsible for proper nerve functioning or transmission of signals through the nervous system of different living organisms. So in principle, organophosphorous pesticides act as a mild nerve agents. We have nerve gases used for military purposes. This is a milder variety of those. But they act in a similar way. They inhibit the acetylcholinesterase. Now to have this system faster, we have immobilized acetylcholinesterase in such a column. And by a flowing system, we inject a sample, larger volume of a sample through one valve. This interacts or inhibits binds irreversibly to acetylcholinesterase. And by injecting the substrate for acetylcholinesterase, we can obtain the initial activity of acetylcholinesterase as shown here and the remaining activity. And the difference here shows the inhibition. So the higher inhibition, the higher the content of pesticides in the sample. This was some sample of anions treated with pesticide. And the other one was iceberg lettuce. We can see clearly the decrease in the activity, which confirms the presence of pesticides. Now, why do we want to do this? Because we are not interested in looking at a particular pesticide. This will take too much time. We cannot analyze every single pesticide present, but we are just interested, is there a pesticide or not? And once we see that there is a pesticide, we can then apply, as we said yesterday, the rear guard methods, HPLC, liquid chromatography, MS, which takes at least three hours for sample preparation, and so on and so on. So a very, very time consuming and costly analytical techniques. But this has shown as a very reliable yes, no response to analyze different food samples. From here on, we have inverted the philosophy a little bit or tried to do other applications. If we can detect the activity of acetylcholine esterase and the effect of organophosphorous pesticides, we can also detect the activity of choline esterase in the body fluids, particularly blood plasma, because acetylcholine esterase is known to be an indicator of our liver function. But lately, it's also associated or considered a potential biomarker of Alzheimer's disease, because acetylcholine esterase regulates the nerve signals in our body. So if we are able to measure, in a very fast way, activity of acetylcholine esterase in blood plasma, we could have a good method for diagnosis. So we have tried. And basically, instead of using the column with immobilized acetylcholine esterase, we have simply put the substrate for acetylcholine esterase into the carrier buffer. Here we have a pump which carries this buffer to the thermal lens detection cell. And with the injector, we are injecting about 20 microliters, in this case, of a sample to which we have added a reagent. This reagent, DT and B, reacts with the product of reaction between the acetylcholine esterase, which is contained in this sample, and the substrate. So acetylcholine esterase plus substrate gives a compound A that reacts with DT and B and produces this yellow-colored compound B, so to say, to make it simple for you. So this is an online reaction. And you see here, we get very nice peaks, which correspond to the activity of acetylcholine esterase. We can calibrate this by knowing the concentration. This is a calibration curve, basically. So one times diluted means 164 units of acetylcholine esterase per milliliter of solution. Here we were able to dilute it 10,000 times, and we still see the signal, which at present corresponds closely to the limit of detection, which is about, let's say, 10 million units per milliliter. The blood levels in human plasma from different studies were at the level between 2 to 8 units per milliliter. So we are about 500 times, 1,000 times, no, 1,000 times, and even more lower than that. So we can detect in a very sensitive way, and what is most important, we can do this in a very short time. Here you can see, we still have some margin. You see, my student was probably drinking coffee in between two injections. But in principle, this is about 300 seconds. We could do three injections in about two minutes time very easily. So two minutes for one analysis. If you take a commercial assay from APCAM, this is between 10 to 30 minutes, at least. So we can go much, much faster, and providing also higher sensitivity. So this is one of the starting points for our future work in this regard. Another combination with bioanalytical methods, now we have to go back to yesterday's lecture and remember, why do we need bioanalytical methods with thermal lens spectrometry? Why did we need a setil cholinesterase? One of the problems is selectivity. The other one is limited excitation wavelengths. With a setil cholinesterase, we have converted, so to say, our pesticide, which does not absorb in the visible, into something measurable in the visible spectrum range. Here, we are after food allergens. If you go now to the restaurant, you see all kinds of different markings, labels by different food offers, even pizza, gluten, beta-lactoglobulin, ovalbumin, all these are allergens. Then the protein is coming from peanuts and so on. So we want to detect them. Producing allergen food free is very important. Such food is very expensive. That's why very easily people are trying to cheat on this. So you must have reliable methods to determine. Now, of course, we cannot detect food allergens as such. So we have to use other techniques, other methods, and in this case is so-called ELISA test. And I will just briefly explain what is the secret of this test. We use what we call a primary antibodies, which are immobilized to some solid surface. And this primary antibodies bind selectively to anti-gene, as it's shown here, which is actually our analyte. This would be the food allergen in this case. Now, of course, food allergens could also bind to the support. So after the first step, we have to block all the active sites by adding some proteins. These are these yellow dots here that block other active sites. So the binding is possible only to the active sites of the primary antibody. Then we use secondary antibody, which again binds specifically only to this particular anti-gene. So we have secondary antibodies for every particular allergen. We have antibody for beta-lactoglobulin, antibody for ovalbumin, and so on. And this secondary antibody is then labeled, usually with some enzyme. It can also be labeled with the fluorophore. Now, using horseradish peroxidase as a label, then we need a substrate, which is converted by the action of this enzyme. And this forms a colored product, which is then detected either spectrophotometrically or by thermal and spectrometry. Now, consider all these steps here. And the fact that proteins are large molecules, so they have very low diffusion constant. So it takes a lot of time for such a protein to diffuse to the active site if we have relatively large sample or large distances. This makes such analysis very, very long. Sometimes it requires even 24 hours to be completed the entire sequence. Now, here for the same system, we have used a little trick by pre-incubating our sample with the secondary antibodies, which were already labeled. So we went from this step directly to this one. We have skipped two steps here. So we pre-incubated our sample. The antibodies bind to the allergen, and this allergen then bind to the primary antibody. So here you can see the injection of our sample, which was this part here. And then we were injecting the substrate. This is this OPD. And this has resulted in high signals, which were proportional to the concentration of our analyte. And you can see that we were able to do it in a relatively short time. It doesn't require 24 hours, but these tests are usually much, much faster. But what I want to show is much lower limits of detection. This one is a little bit optimistic. I would trust this here comparison. This is a commercial transmission mode measurement by Microtiter plate reader. This one 1,000 times, I think either my student exaggerated a little bit, or the producer was a little bit conservative in specifying this limit of detection. Because this is not what we tested. We just took it as a specification from the producer. Usually from what we know now, from the enhancement factors and from the laser powers that we get, you can easily determine that for about 100 milliwatts power you can expect between 1 to 2 orders of magnitude improvement in limit of detection compared to spectrophotometry. So here now we are going from the microscopic. This would be a microtiter well on the microtiter plate. Usually we have immobilized antibodies. And we put a few millimeters of a solution. If we go to the microfluidic systems, you see this is now 100 micrometer channel with antibodies immobilized on some microscopic bits or even nanoparticles as we will see later. We significantly reduce the distances. And the molecular diffusion time goes with the square of the distance. So going from, let's say, 1 millimeter to 100 micrometers, we reduce the molecular diffusion time by a factor of 100. And that's why you see SA for immunoglobulin A was reduced from 24 hours to 20 minutes by using such a system with the thermal lens microscopic detection. Why thermal lens microscopic detection? Because we have only 100 micrometers interaction length. And there in the transmission mode measurement it is very difficult to get any reasonable signal. OK, in addition to this we can do all different processes from extraction as shown here for this determination of cobalt by introducing the ligand that makes complexes with cobalt, but also with other metal ions, which are then extracted into this microfluidic system into m-xylene. And then by adding acid, some metal ions metal complexes are destroyed. Some are extracted into this basic medium while cobalt complex remains and is detected. Very impressive limit of detection considering the small detection volume. As I mentioned yesterday in terms of concentrations, this is not so impressive because the interaction lens are extremely short. We have learned about thermal lens microscope and its operation yesterday, so we will not repeat it today. Here I would like to go on with another approach in detection of non-absorbing species, I mean non-absorbing at the wavelengths of lasers which are available. So if you want to make them absorbing, we have to do some colorimetric reactions. And just to demonstrate the principle of this analysis, I will show you the case for the determination of chromium-6, chromate. This is a cancerogenic species. That's why we don't like it in the drinking water. Another problem is that it is very, very soluble as compared to chromium-3, which is, on the other hand, another species of chromium but is essential for our body. So we have an essential and we have a highly toxic species. So it's important to determine not the total concentration of chromium but what we call the speciation of chromium. So here we have a carrier fluid. This is a microfluidic chip. Here you can see our home-built thermal lens microscope with the probe beam coming up. And here is this position of the microchip here. The pump beam comes from the top down. And here we see the detector. So we have a carrier fluid which runs at 5 to 50 microliters per minute. And then we do a very fast injection at about five times higher in flow rates of our sample which contains chromium-6. The carrier contains the specific reagent that forms a colored complex with the chromium. This is the one that we have demonstrated yesterday as an example of photosensitive analyte. So we don't want to have too much laser power or power density. Again, the considerations of the flow which we described yesterday, so we have to keep the pump and the probe beam slightly displaced in terms of their axis. Now, here we have a demonstration of different injection volumes. And we see that after we go too close to one microliter injection, the peaks split. We get a double peaks. The reason is that this sample inside the microchannel is very, very long. It extends over several centimeters. So the diffusion time for the reagent to come from front edge and from the rear edge of our sample to the center is too short. Only when we have a shorter sample, which means shorter volume of about 0.7 microliters, this reaction is complete at the center. Even this is surprising for such a long distances. But in another work, we have shown that not just the molecular diffusion but convective mixing in the microfluidic system is a very important way of transporting reagents. So in microfluidics, we have a lot of convective mixing which assists to the speed of those reactions. Then, going with the optimal sample size at the different carrier flow rates, you see at the low flow rate, we have a lot of diffusion again. That's why it peaks are much broader. Going with the faster flow rates, we lose a little bit of signal because of the known effect, loss of heat with the flow. We lose a little bit of signal. But still, the diffusion is much less. Diffusional broadening is much less. And we can do up to 20 injections in one minute. 20 samples. OK, if we do in duplicate, 10 samples in one minute can be analyzed. So this is now already the question of mechanical engineers. No automatic sampler is so fast as we can do it. So we don't want to push it further before we resolve the question of autosampler and injections. But this shows you how fast we can do such analysis. And now let's go back to the bio. Even though Chromium-6 is related to cancer, so it cannot be excluded from bioanalysis. But I would like to show you an experiment with, again, linking the algal blooms. This is related to cyanobacteria. When cyanobacteria are blooming, there is a lot of dead cyanobacteria which form microcystine. And microcystine is a known neurotoxin. That's why there is a very low regulatory limit for drinking and recreational waters. And the big interest of fast screening techniques to detect microcystine, which is this macromolecule. The interesting property or important property of microcystine is that it inhibits the phosphatase enzyme. So similarly to the Acetylcholinesterase case, we can observe the activity of phosphatase by the conversion of this para-nitrofenyl phosphate into para-nitrofenol, which is then in equilibrium with para-nitrofenolate, which is colored species that we can detect. So we have a carrier flow with our reagent. And we are injecting our sample, as we have done with the Chromium 6 in the previous case. And further down, after this reaction happens, we have a thermal lens microscope that detects the product. So the higher the concentration of the product, higher the activity of the enzyme, less microcystine. Lower the activity of the enzyme, lower the signal, higher the concentration of microcystine. And this is the calibration curve that we have done. And we could reach a limit of detection of about 80 nanograms per liter, which is 12 times below the WHO limit for drinking water. What is more important, we can do it eight times faster than batch mode, as I say, due to the shorter distance or due to direct mixing, because this commercial assay does this on the macroscopic microtitter plate. Another, we still have about 20 minutes, another important application in biomedical diagnostics. For example, detection of biomarkers of acute kidney injury. Where does it come from? And the reason why we have undertaken this research, medical doctors, when they do imaging of soft tissues, they usually use different contrast agents. And one of the MRI contrast agents, it's a group of iodinated benzene, iodinated aromatic compounds, causes such kidney failure. So of course, medical doctors are interested not only of measuring this biomarker, which is N-gal. This stands for neutrophil gelatinase associated lipokalin, if it tells you something. It's basically a protein that is released when your kidney is under inflammation process, basically. Not only that they want to measure the status of your kidney, but they would like to do it during the operation. Because once they have the patient there, they would like to know whether they can do another injection of the contrast agent for imaging, or they better don't do it because the patient is already bed shaped due to the kidney. So in the past, they were using creatinine as the indicator of kidney failure, but creatinine responds in about two days time. So there was no way that they could wait with the patient there open heart and do the imaging. The commercial kit for N-gal takes four hours. So this was the idea, do it faster than four hours if we can. And of course, we have tried. This is a commercial microtiter plate, as you can see it, with different standard solutions and different samples. And this is then read on what we call the microtiter reader in a transmission mode, in this case, because there is no fluorescence engaged. We have taken such samples and analyzed with thermal and spectrometry. And you can see that in this microfluidic system, we can get a very, very fast response of fast injection of several samples. With in triplicate, for example, we could analyze two samples in one minute, with still 500 times diluted samples with better LODs than in the case of transmission mode. But this was still a macroscopic system and the reaction took four hours, as in the commercial kit. So we want to transfer it into the microfluidic system. And for this, we use the magnetic nanobits. These are magnetic nanobits with iron oxide core and the silica shell that makes them insoluble or inert. And we can immobilize the primary antibodies for Engal on such nanobits. And we can insert them into the microchannels and keep them there by magnetic field. One of the important obstacles in such systems is filling and replacing the biomolecule once it's either inhibited or used, consumed, degraded. It's quite complicated. If you use magnetic nanobits, you have a microchannel. You trap them with the magnetic field when the antibodies are used. You remove the magnetic field, flush it out, and new injection of magnetic nanobits again. And here, you can see probably a little bit of blue coloration. Or here, which is already the product of this horseradish peroxidase-labeled antibodies, secondary antibodies which were also reacting in this small microchannel in a relatively faster way. This is now actually the image of, let's say, our prototype. This is a thermal lens microscope. Here, you can see two lasers, one for excitation. The other one is probe. Here is the microfluidic chip with represented magnetic field that traps the nanobits. These are the microfluidic pumps. And we have reduced, by performing all these steps, all the six steps that I have shown initially for allergens, we have performed all these steps in the microchannel. And by doing this, we have reduced the time of analysis of the entire process from four hours, which is a commercial analyzer down to 35 minutes already. So we can, in principle, analysis with a half an hour time resolution, so to say, even better if we take samples at a much shorter intervals. But what is interesting is that by using this on different patients, which were subject to this diagnostics, we have also shown the comparison of thermal lens microscopy to ELISA, which gives relatively good correlation. But we have observed a significant increase in n-gal concentration only in one patient. And this was relatively slow. It's a two-hour time response. So again, showing that n-gal responds too slowly to meet the requirements of medical doctors. So we need to find another biomarker. But the analytical tools are here. Of course, we can immobilize any available antibody on such magnetic nanobits. And by having appropriate antibodies, we can then selectively detect different biomolecules. One of them could be, for example, viruses. One such virus is human papilloma virus, which is particularly known to our female participants because it can cause cervical cancer. And we want to avoid this. So by using what we call pseudovirions, human papilloma virus pseudovirions can be synthesized, basically, in laboratory. We can start with a protein, which forms such five protein structure. And these 72 such structures then form non-infectious human papilloma virus. It has, on the surface, it has all the properties of human papilloma virus, but it's not infections because it doesn't have any content, which would be the mechanism of infection. So we can use such pseudovirions, as we call them, immobilize them. And then they will attract antibodies, which can be found in the infected person. So the antibody that is formed in our body to fight this infection would bind to such pseudovirion in this microfluidic system. And we will be able to detect whether a person was infected or not. Here I am showing some different calibration curves performed on such a system. And here we have a microscopic flow injection analysis with thermal mass detection on magnetic nanobits with the best limit of detection compared to the commercial ELISA or ELISA combined with the thermal mass microscopy. So we can do it in a sufficiently sensitive way. And the comparison of ELISA and microteater microfluid at LM with nanobits shows good agreement. And in some cases where ELISA does not provide sufficient sensitivity, we are still able to detect the antibodies. And the persons that were under this investigation, they were known by the medical doctors that they were infected. So this was no question. Is this true or not? Yes, they were infected. The level of this antibodies to human papilloma virus, of course, it depends on the stage of the infection. So if a person was infected, I don't know, two weeks ago or one year ago, of course, the level of antibodies decreases. Also the magnitude of the infection itself. But important to demonstrate always the reliability of the technique. Conclusions probably we can save some time for questions. But something that I forgot my glasses, that's first. But what I wanted to point out is actually the time of analysis, which unfortunately I don't have here. But the total time of analysis is, again, 35 minutes with thermal lens in the microfluidic system. I bet that the commercial ELISA for human papilloma virus is on the order of 10 hours, something like that. So this is the time difference which we can get by performing the steps of the ELISA test in a micro on micro scale by reducing the distances, which reduces the molecular diffusion time. That's the main secret of this approach. So with this, I think I have exhausted all my slides. And of course, I will be happy to answer any question you might have. Hope I will be able to. Thank you.