Dynamic Vapor Sorption (DVS) Applications for Amorphous Content Calculations & in situ Raman studies

Hi everybody, My name is Damiano Cattaneo, I'm a product manager at Surface Measurement System for the product of DVS Adventure Resolution Endeavor and today I would like to guide you through a presentation following the application of DVS for amorphous content, BAT surface area and Raman studies. The presentation will be divided into two main part. The initial section will cover general introduction to DVS and DVS studies, part of the system and how we collect data. The second section of the presentation will follow up on a study of amorphous content calculation where we are going to show you two methods for amorphous content calculation, then BT surface area measurement and in the end we are going to show you a full study of in situ RAMA on hydrate solvate compound. So starting from the initial part of the presentation, whenever we talk about analysis of property of sample, we always refer to obviously sample in potentially three different phases, vapor, liquid and solid. In a dvs we usually use vapor phase molecule, which we refer to as probe, to study the surface or the bulk of solid sample, powder or membrane or polymers or other type of formation. Now, if we have a look at a standard method of analysis, e.g. spectroscopy or calorimetry, what we usually do is provide a form of energy to the surface or the bulk of the material, and then the energy is transferred or absorbed by the sample and then release and measured by a detector. Now this is a very similar of what happen when we use a dbs. We provide probes as matter. The probe interact with the surface or the bulk of the sample and then they can be released. Now this schematic basically shows whatever happened. Typically in powder, XRD or NMR or ftir, we have a source that generates energy. The energy is transferred through the sample, then released in a different way, and then we have a detector that measures it in a very similar way. We can use heat, for example, in TGA or DSC to do the same type of process. We provide heat. The heat transfers through the sample is released out or it changes the structure of the sample itself. Following the same type of schematic, in a dvs, we have molecules that are provided to the sample, for example water molecules in a gas phase. These molecules interact with the surface of the sample by adsorbing or interacts with the bulk of the sample by absorbing through the sample itself. Now, this process can be irreversible or irreversible. So what it means is that when we lower the humidity present in the chamber where we have our sample, the molecule can be released and the mass of the sample can reduce progressively. Now, what I'm going to do in the next couple of slides is focus specifically of the definition and the concept of adsorption and absorption, because these two main concepts are going to be used throughout the presentation to analyze the structure of different types of material. Let's start from adsorption. When we talk about adsorption, we usually refer to two potential phenomena, which are called physisorption and chemisorption. In physisorption, we have a very weak interaction between the molecule, the probe, the gas that we are introducing, and the surface of the sample. Most of the time, the process is reversible, which means that when we lower the humidity or the partial pressure, the sample release those molecules and loses mass. So why is it important to study physisorption? Physisorption gives information in regard to surface area, pore size, surface roughness, heat of absorption, and free energy. So when we are going to talk about surface area, for example, we are going to refer to a physisorption process rather than a chemisorption or a bulk sorption. Especially now, when we talk about CAM insertion, the interaction always happens on the surface of the sample, but we are usually referring to as a strong interaction. The molecule can be bounded to the surface of the sample reversibly or irreversibly, but the energy of the interaction is usually quite high because we are talking about covalent bonding. Now, typical of this type of interaction is active surface area of catalyst dispersion. So molecules that are specifically designed to interact with the probe that we are using for our experiment. Now, when we talk about adsorption, we usually refer to as two types of interaction. Bulk sorption or an adsorption into the lattice structure. Now, when we talk about bulk sorption, we have to think about this process is usually slowed by a diffusion parameter. So what we have is that molecules reach the surface of the sample, pile up on the surface, and then they diffuse throughout the experiment in the bulk of the sample itself. That is automatically limiting, a limiting factor on our process because it slows down the kinetic of the experiment itself. But it also gives information about the type of interaction that the sample have with that molecule. So what type of information do we get? We obviously have diffusion and permeability coefficient drying kinetics. Those are very simple experiments that we can definitely do in a dbs. But we also have phenomena like transition. When the molecules go through the bulk of the material, they can cause a change in the structure of the material itself. So we can have glass transition Recrystallization, for example, of amorphous material. That's what we are going to talk about in the next phase of our presentation. The last example that we are going to talk about actually refer to absorption into the lattice structure. What this means is that the molecule penetrates through the bulk of the material and change the crystal structure of our material from an anhydrous form to a hydrate or a solvent. Now, that process can be again, reversible or irreversible. But we can also define an extra parameter, which is, is the hydrate stoichiometric or non stoichiometric. What that means is then can we exactly count how many molecules of our solvent goes into the structure of the material? So in a stoichiometric hydrate, we will have an extra hydrate or a tetrahydrate. So we can exactly say how many molecules are going into this bulk of the material. In a non stoichiometric hydrate, for example, we cannot define a number of molecules that form that new structure. Okay, so these are the type of information, the type of interaction that we can have with the structure of the material when we introduce a molecule. How do we study this type of interaction? So, to explain you how advs work, we are going to look at the origin of, of the design of a dvs. So to do that, we are going to focus specifically on how was the IFDVs introduced, starting from the previous method that was used historically. Historically we have a method called JAR method. In this method we have a series of desiccators that contain a saturated solution of salt at the bottom of the desiccator. Each desiccator contains a different salt. And each salt will be able to generate a different level of partial pressure in specifically relative humidity for that specific type of desiccator. So, for example, we will have a desiccator that generates 11% relative humidity by containing lithium chloride. We will have a 75% relative humidity desiccator that will have sodium chloride inside, and so on and so forth. And what you will do is record an initial mass of your sample, put it inside the desiccator, wait for a couple of days, then take it out and measure the mass again. And you will do this process over and over and over until you record a final mass of the sample that is at equilibrium, so the mass stopped changing. That will give you an equilibrium point for that relative humidity value for each one of the desiccator. And then you can build up an isotherm in that way. Now, a DVS basically does exactly the same process but in continuous. What it means is that we have an in situ balance that constantly records the mass of the sample and we generate relative humidity by controlling it step by step. So automatically we are able to measure and control the relative humidity. So it's a process that is completely done in C. So what are the difference the advantages of using a DVS compared to a desiccator? Now when you use a desiccator obviously you need a lot of mass of the sample, you need grams of sample. In a DVS we use a micro balance. So consequentially we don't need a lot of sample to make an experiment accurate and repeatable. The second parameter that we need to focus on is equilibrium. When do we reach equilibrium? Now, in a jar method, the only force that push towards the equilibrium, the mass equilibrium, is diffusion inside the desiccator itself. In a standard devs we have a gas flow that pushes molecules on the surface of the sample. This allows us to reach equilibrium much quicker and much weaker means saving of time but also higher accuracy of the experiment. Because you have to think about the fact that whenever you measure something in a desiccator you need to take it out of the chamber of the desiccator itself, you need to put it on the balance and record it. This creates an issue of potential cross contamination whenever you expose it to the outside condition of the lab temperature and the lab humidity and put it back in. Now different user, different scientists can take longer or shorter time. Whenever you do this type of experiment, consequentially introducing an extra error In a dvs the entire experiment happens in situ on the balance. So automatically we don't have to touch the sample, we don't have to open the chamber. Everything is recorded constantly so there is no issue of contamination, cross contamination or error in the handling of the sample itself. Now here I'm showing you a plot of kinetic of time over partial pressure and it gives you an idea of a full sorption dissorption experiment which takes about 6,000 minutes. Now this same sample has been analyzed using a desiccator and a desiccator takes about 300 days. In a, in a DVS it takes about 4 days, 4.2 days. So it's a massive save in time. The other parameter that you have to think about is the fact that in a desiccator on the left side you can see the data are only showing absorption, so the mass going up, it doesn't show data for Desorption So we do not have a full study of your isothermic. In a DVS we have the data for absorption and the data for desorption. So when we are going to plot our hysteresis and isoturn plot you can see that the JAR method only gives you the data of absorption while a DVS will give you the data on absorption and desorption and it also gives you the information in regards of the gap of hysteresis that you are recording during your experiment. So some more complete set of details, a more accurate set of data and also is quicker to obtain. So here is a schematic of our DBS and I'm going to highlight how the process of generation work. I'm going to start from the bottom right side where we have the generation here is divided into two reservoir. Even though your system will probably be provided for water only with one reservoir. One of the reservoir is water which is a container, a glass container for water itself which will represent your 100% wet, your saturated vapor pressure water. And then we will have what we call dry which is your reservoir containing your carrier gas, nitrogen or air in a cylinder. The two reservoir are controlled by valve called mass flow controller that allows you to control the generation. Let's say you're doing a dry stages so you want the Kerry gas to be completely dry. So what you do is the system will automatically open your dry completely and keep your wet, your reservoir of water completely sealed. If we are generating for example let's say 80% relative humidity then 80% of the flow will be open on the wet side right here and only 20% will be open on the dry side. The gas that is generated by this two reservoir is then pushed in front of a sensor that allows us to check if the generation is correct. In this case for water we have a relative humidity sensor. For organic solvent we will use a speed of sound sensor that we're going to talk about in a later stage of the presentation. The gases are then pushed towards the chamber that contains the sample and the reference side. So here on the left you have your sample and on the right you have your reference. Those are two pans that are connected with a hand and wire to a two plated balance. Picture this balance as an old pharmacy's balance with two plates. Now what is the difference between the sample and the reference size? There is no difference between the sample and the reference size. The sample and reference size are basically pans with end and wire. The only difference between the two sides is due to the presence of the sample itself. So why is it designed in this way? We are only interested in measuring the mass change caused by the absorption of molecules of water or organic solvent on the surface and the bulk of the circle. But we are doing it very inaccurately. What we are just doing is pushing molecule of water into a chamber. Water will absorb everywhere, will absorb on the hand and wire, will absorb on the pan, will absorb on the side of the chamber. So the only way to make it out accurate is to make it symmetrical right here. So what we are doing here is providing molecule on both sides. And if there is no sample, the system is perfect or symmelectrical. So molecule will absorb in the left side in the same way they will absorb on the right side. So the mass will not change, the balance will not read any mass change. However, when you load the sample, you will have a mass recording on the left side that is not present on the right side. And that in the case is the measurement that we are interested into collecting. That's the reason why the system is designed to be symmetrical, to be accurate. Now, how does the standard experiment work? Is very straightforward. We start from loading the sample into the chamber. We record an initial mass, a dry mass of the sample, which is usually used for reference for the calculation for the percentage mass change. And then we do a kinetic experiment in which we introduce a constant level of relative humidity, let's say 10%. We wait for the mass to reach equilibrium and then we move to the next stage, going from zero to 90% back to zero. We record the mass change at each stages and then we plot an isotherm of absorption and desorption by recording the mass equilibrium at each stage against the relative humidity plot. To give you an example of a standard set of data, here we have a sample of rice collected in a DVS full study of absorption and desorption. The data that we are showing you here are kinetic data. So we have a plot of time over percentage mass change, which is highlighted in blue, sorry, red. And in blue we have the relative humidity or partial pressure. So what we have here is a full study of absorption and desorption with the initial drying stage. So what information do we get from this type of data? We can see first that the material is perfectly reversible. We start from a dry mass here and after a first cycle we go back to the initial mass again and we can repeat it in a second cycle again. So the process is perfectly reversible. We can wet the sample and dry as many times as we Want. The second thing is that the mass uptake is exactly the same at the end of this period. Again pushing towards the concept of reversibility of this sample. The third interesting thing is the mass change. There is a lot of mass change in our sample. The material absorbs over 20% mass change. So obviously this sample absorbs a lot of water over time. So if we take every single point of equilibrium over kinetic plot and, and plot it against the relative humidity, we get our isotone plot. And you can see here, there is the full study of absorption and desorption. So what does it highlight? This plot? It highlights first that is reversible, that has a lot of uptake, but also that there is a gap of hysteresis between the absorption and desorption. Now the very interesting thing is the fact that this is telling us that this material have a bulk absorption. It means that the molecule not only pile up on the surface of our right sample, but also go through the bulk of the material. The reason is, if you think about it, when molecules migrate through the bulk of the material, when I'm drying the sample, when I'm going through the desorption stages, then they will require more energy to leave the sample to migrate throughout the sample. And that is causing a gap of hysteresis because I have to go to a lower level of relativity to get to the same level of mass during the desorption stage compared to the absorption stage. However, the other important thing is the process is completely reversible. Now, thinking about bulk absorption, we always think about thermodynamic equilibrium. Now this is a significant gap of hysteresis. And this experiment was performed at 25 degrees Celsius. The desorption process, specifically on bulk, is usually determined by amount of energy provided to the system. What I would expect, for example, in this type of experiment is that if I increase the temperature of the experiment, let's say 30, 35 degrees, which we can do in our system, I would probably expect that this gap of hysteresis right here to close progressively because we are providing more energy to the system, consequentially reducing the energy requirement for the desertion stage. Okay, so this gives you a brief introduction of our technique and the type of information that we get. What I'm going to do now is a more focused study on our amorphous content and then going over to the other section of our present stage. So let's start from amorphous content. Obviously, whenever a company is studying a material that contain a certain percentage of Amorphous content is never looking to a purely amorphous material or a purely crystalline material. Most of the time what we are looking at is a partially amorphous material. And what we are interested in is determining what's the percentage or the content of partially amorphous material that we want to study. Different type of material, behave in different way and have different crystal structure related to the crystalline form and the amorphous form. I'm not going to focus specifically on any type of material, but what I'm going to give you is the methodology that can be used to study the amorphous content in a dbs. And it all starts from the analysis of a standard DBS plot for a material that is amorphous and it moves to crystalline. And this is the plot that I'm referring to. What you see here is amorphous lactose exposed to water over a two cycle experiment. What we are looking at here is an initial stage of absorption. You can see here our mass increases progressively, but when we reach a certain level of relative humidity, we have a mass drop. Now, whenever you see a sample experiencing a mass drop, so the mass rather than going up, it goes down. Even though you are providing more molecule to the system, it usually tells you that something very particular is happening to the sample. You expect that logically a sample will always obtain more mass whenever you are providing more molecule. What is really happening here is that the molecules that are absorbing through the bulk of the material are triggering a crystallization process in my amorphous sample. And what is accidentally happening is also the fact that because the material is moving from amorphous to crystalline, the surface area of the sample is reducing progressively through. Think about in this way a perfectly ordered material, a crystalline material does not have a lot of space in the structure itself to absorb water. Water will have to go in between the structure to allow for the packing of the system. However, an amorphous material, due to this irregularity on the structure, have a lot of space in which you can occupy with water. So whenever a material move from an amorphous form to a crystalline form, the surface area available for solvent reduces. And consequentially what happens is that when it reorder, it releases molecule of water or organic solvent that has been absorbed, consequentially losing mass. And here is very evident that there is a significant mass loss caused by the crystallization stage. The interesting thing is that this second cycle that is highlighted over here does not contain any mass drop, which means that it does not contain any recrystallization stage. So we have a crystallization step, probably a second crystallization step over here. But we do not have any crystallization stage in the second cycle of our spelling. So everything that was amorphous in the first stage has moved to crystalline in the second cycle. Now, if we plot, the isotherm of this data is even more evident. You can see here, this is my mass drop during my first cycle. And the second cycle shows perfectly repeatable data, which means that a single cycle is able to transform my entire material from fully amorphous to fully crystalline. Now, how do we analyze this data? How do we get information of amorphous cone? I'm going to show you two examples on how to do this calculation. The first example is based on using solvent that does not cause the crystallization, so does not cause a mass drop. And the method that we are using here is based on identifying a material that is purely amorphous, a material that is purely crystalline, and then creating a mixture between amorphous and crystalline, and creating a calibration in between these two. I want to specifically focus on a single detail in regard to every methodology that we use. I will refer to a material that is purely amorphous and a material that is purely crystalline. The DVS cannot define what is 100% amorphous. It does not have the ability to do that. You pick a material that you define as purely amorphous and that will be your reference towards the calculation. It could be the material as purchased, it could be the material as synthesized, and then you will calculate all the other value of a marker's content compared to your reference that you pick. It could be the material stored after six months or any other process or packaging or post processing of synthesis. So you will pick a material, you refer as that material as your reference amorphous and the other reference crystalline, and then you will do the calculation on any other product referring back to your reference that you pick. So here is for example a study done on sorption. In this case is water relative humidity over amorphous lactone, crystalline lactose and then 10% amorphous lactose at 25 degrees Celsius. As you can see here, your amorphous material sits on the top very, very high uptake because it has a higher surface area. Your crystalline lactose is basically sitting at zero. So in this case what you see is the assumption that the crystalline form almost has no uptake compared to the amorphous form. And then basically here you can create a calibration that is these two points. Your zero your 100. And then you assume that everything in between will be a ratio between the crystalline and the amorphous section. You can create extra point by mixing the amorphous and the crystalline and creating that mixture artificially, or you can just dial up the calculation by using the two value of 100% amorphous and 100% crystalline. So what are the advantages of this method and the potential disadvantages? So the big advantage of using this specific method of calculation for the amorphous content is the fact that the method is extremely accurate. As I showed you before, we can use this method to basically measure small mass changes between the amorphous form and the crystalline form. And the accuracy of the measurement is within 0.25%. So which is much higher compared to what you get from example from powder xrd, which you can see over here, but which is usually about 5%. It does require an amorphous counter standard that you define in between your sample. And it has an assumption, and the assumption is quite evident by the data presented in this slide before. You are assuming that the crystalline material has no uptake. And that's a big assumption. The majority of the material that you're going to study, the crystalline form, has such a small uptake compared to the amorphous form that is insignificant. So this assumption does not hold up in those type of situation. However, if the crystalline form has a little bit of uptake, then it becomes a more important factor. So that is something that needs to be considered. However, the big advantage of this method is highlighted at the bottom here. We do not have to find a solvent that caused the crystallization. The data that I showed you in the slide before had no mass drop. So consequentially you can just use your standard solvent. You don't need to figure out a solvent that allows you to have a recrystallization state. You just have to measure the absorption stage. That's all you have to do. So that's a big advantage because you can just pick whatever solvent you want. There is no mass drop in that specific method. Now, if we want to have a look at a second method, a second method will not require the assumption of a crystalline form because it's only based on transformation between an amorphous and a crystalline. The important factor that we are focusing on in this case is the fact that we are going to identify a solvent that causes a crystallization and then we are going to study our material before and after that crystallization stage. So how do we do that? So first we are going to have to identify a solvent that causes a crystallization and that's the first stage. So we look at this set of data and obviously the solvent that we pick for this specific study is not good enough because you can see here, my amorphous does not have any mass drop. So obviously for this specific sample, water is not a good solvent. It's not very useful for our type of study. So we move to a different type of solvent. In this case we move to acetone. And in this case acetone has a very significant mass drop at about 60%. Now that is perfect because now we have the solvent that caused the crystallization. So how do we gonna complete this study? We're gonna do two stages and we are gonna identify specific partial pressure that allows to do the calculation. The first partial pressure that we're gonna pick is 30% right here. At 30% partial pressure, our material is amorphous. There is no crystallization, the mass drop happen at a higher value. And the second partial pressure that we're going to pick is 90%. At 90% that you see over here, our material is crystalline because it's after the master perfectly. So now we have our 2 target, 30 and 90. So, so what we do in the next stage is take a fresh new sample and repeat the study that we had done in this way. We start from 30% relative humidity, partial pressure. Sorry, right here. And we measure the mass change at 30% of the amorphous because there is no crystallization yet. Then we move to 90% right here. And you can see at 90% the material moves from amorphous to crystalline. Perfect. Now we go back to 30% in the third stage and you can see the mass uptake is much lower compared to what we have in the first stage. The difference between the initial value and the last value is dependent only on the amorphous content of the material that we want to study. So what we do is we do the experiment of our reference, what we call 100%, that is our standard and then we repeat the experiment on any other material that we want to calculate the amorphous concept towards. So what are the advantages of this method? This method is very well recognized, is reported in the pharmacopeia and is extremely reliable, is accurate up to 0.2%. It does not require a calibration process, it does not require a calibration curve because you just have to define a standard of amorphous. It's very reliable, is much more reliable than powder xrd. However, it has two requirements that you Define an amorphous standard in your sample batch that you will call 100%. And second, you have to find a solvent that caused the crystallization. In our case was acid. Now what about your sample? Which solvent should I use? Well, the question is more or less, how did you prepare your amorphous material? Your amorphous material is probably prepared by synthesis. So probably with a CO solvent and an antisolvent. So the CO solvent and antisolvent will be my first two guesses that you are going to use to study the recrystallization process. Same thing for a spray dry material. Spray dry material are prepared using a mixture of solvent. Those two solvents are probably the one that can cause a recrystallization stage during your experiment. So the choice of solvent is usually defined by the method of preparation of your sample. Okay, so the next application that I want to talk about is BT surface area. Now I know that there are techniques that use nitrogen absorption as a volumetric method to measure BET surface area. We use gravimetric method. If the question is, are the data going to be different? The answer is no. We are analyzing a property of the material, this surface area of this material. Consequentially, the data provided by different machines has to be the same because we are referring to the same property. So that will answer that question. How do we measure it? So the measurement of BET surface area is a fitting applied to the isotherm plot independently of what type of material you are studying and what type of solvent you are using. The isotherm will have a shape that corresponds to one of these six plots reported over here. Now, to measure BT surface area, we have to generate a plot. There is a type 2 right here or a type 4 right here. Isotherm. Why do we need a type 2 type 4 isotherm? A type 2 type 4 isotherm are the only two plots that have a linear section in a lower monolayer region between 5 and 30% partial pressure of relative humidity. We need that linear plot because the BET surface area calculation is a fitting of a line over the isoton plot. And what I mean by that is the fact that we are trying to fit this equation to the isoton plot in a monolayer region. And the GRIS equation, if you look at it, is the equation of a line you see here, y =mx q. So what we are really doing is trying to fit a line in a section of your isotherm and try to match it up with this equation. And the only type of isotherm that give you a Linear plot are type 2 and type 4. Once we selected the area that we want in the isotherm, we can easily do this calculation and we can easily do the fitting. The better is the fitting between the point of your isotherm between 5 and 30, 35%. And the higher is the R squared value and the more accurate is the monolayer calculation. So more accurate is the surface area measurement. So how do we define this interaction? How do we define a linear plot? A linear plot is usually caused by a monolayer formation, a monolayer distribution, and that is defined by the interaction between the molecule of solvent that we're using and the surface of the sample itself is not dependent on the sample, is not dependent on the solvent, is dependent on the interaction between the solvent and the sample. What I'm referring to is basically how these two components interact with each other. Is it coating the sample uniformly or is it spreading in droplets that do not cover the entire sample uniformly? And this is usually basic interaction on thermodynamic and also energy of interaction. So if I think about water on the surface of a pharmaceutical drug, usually water does not spread uniformly like this. It forms droplets like this. This is because the delta H of adsorption is much higher than the delta H of condensation. However, if I get a situation in which I have a solvent that really interacts very well with the surface on the sample, it spreads as a monolayer, so it will give me a linear clock so I can do the PT calculation. The easy example that I show you here is for example, lactose and this isotherm plot, the blue line up here shows the interaction with water. And you can clearly see that the plot is not linear, it is exponential. So I have a type 3 isotherm, which is not exactly what I'm looking for. What I want is a linear plot. Now if I pick a solvent that is less pore under water. So right here I have methanol, ethanol, isopropanol and butanol. You see that the isoton drops immediately and it becomes more linear. In the section that I'm interested. And all this plot right here can be used to calculate the BT surface area. So again, it's not due to the sample, it's not due to the solvent, it's due to the interaction between the solvent and Henderson. So I've been talking about organic solvent for a while. All the data that I showed you so far are collected using our DVS resolution. How do we measure organic solvent in a DVS resolution? Of course, for water we use Water sensor, humidity sensor that are fairly available on the market. However, to measure organic solvent we have to develop our own sensor which we define as a speed of sound sensor. A speed of sound sensor measures time of flight of a carrier gas plus solvent combination and measure the density of that gas basically and it relates back to the relative humidity or partial pressure of that specific solvent. We have to build up an entire library of solvent plus carrier gas combination to measure accurately the concentration of gases during our generation. The important thing is the fact that by using this combination you can get a very reliable response, very quick response, and ultimately a live feed measurement of the organic solvent generation throughout the experiment. This is our proprietary technology. We provide the system with about 30 different type of solvent calibration. Your most. This is not a complete list. We have a fully complete list of solvent. If you're interested in what type of solvent we have already calibration for, you can just contact us. However, the most common solvent that you are going to use in your lab, so ethanol, isopropanol, cyclohexane, octane, acetone, those are already provided. As I said, we have about 30, 30 different solvent already provided. What if you are doing a specific study that requires a specific type of solvent that is not in that list, you can easily contact us and we can provide you with an extra calibration if that is necessary for your specific study. That is not a problem. You can contact us and we can work with you towards this type of calibration. Now the last type of application that I want to show you today is an in situ study using Raman on hydrate and sulfate. The sample that we are focusing on is carbamazepine. And the study that we perform is using a DVS in situ with a Raman. What you see here is basically the Raman probe mounted on our DVS chamber. And here we have our sample pan. And what we do is we load our sample in here and we we measure the Raman spectra at the end of each partial pressure stage and we record automatically all the spectra together. So we can measure the difference between each spectra at the end of each stage and identify if there is a transition in the spectrum. So what study did we do? We started from a high anedrous carbamazepine and we did a study in three different sections. We started for water absorption, then we moved to acetamsorption by itself. We did dvs with Raman for water, DVS with Raman for acetone and then we did a full study on first an absorption in Acetone and then water on the same sample together to see which one of the two solvent is more stable at the end on the structure of the material. So let's start from water absorption. Here is the kinetic plot of water. As you can see, the material form hydrate is a very slow process. It's not super defined because the mass uptake is slightly small. However, one of the things that you might notice is that the reading of the mass change always contains small humps during each stage. That is due to the fact that obviously we are collecting Raman data. So whenever we are triggering the Raman, we are turning on a laser and the laser slightly changes the temperature of the sample causing a small mass loss. It becomes very significant only if the mass change of the sample is very very small. However, it's something to notice when you're collecting DDS data. Now if we plot the isoton plot, you can see that the material ever basically constant uptake, absorption and then desorption. There is a section in which the hydrate is more stable right here. But the mass change is relatively small for this type of sample. We've seen in published data that carbamazepine has a slightly higher uptake. But this is just a small reference from a hydrate formation from the sample that we purchased from Sigma Outreach at the moment. Now what happened when we look at the Raman data? We start obviously from our dry sample which is referred right here. This is the as prepared as purchased material and then we introduce water throughout the experiment. Now unfortunately what we notice is that for water the Raman spectra does not change. We don't have a formation of new peaks, we don't have the shift of existing peak. The only thing that happen is that the intensity of the peak becomes slightly higher. Even if we look at the 90% from 0 to 90% partial pressure and we zoom in into the data. The only thing that we notice is that there is a higher intensity of the peak. No changes, no shift. So water does not cause a significant change in the shift in the Raman spectrum. For caramelzepine it only increase the intensity of the peak. Assess signal to nodes. Okay, so this is our first step of our study. The second stage is to look at the acetone absorption. So we start from exactly the same sample and we absorb and desorb acetone. Now the interesting thing here is the fact that the mass change is very, very high. We are Talking about over 20% mass change and there's a significant jump of absorption and desorption. Again, the sample is perfectly reversible, which is very good for our type of study. Which means that we can repeat this process as many times as we want to. If we look at the isotherm data, there is something very intriguing because what you see is that the sample basically has no uptake whatsoever up to 70% over 70% relative humidity. Exactly. At 80 we have the transition to the solvate form. And the solvate form is perfectly stable until we go back to below 10% relative humidity. That is when the solvate destabilizes and it goes back to anhydrous. So we have this fascinating structure that basically has a full stability up to 80% partial pressure. It transition immediately to solvate over 80%, it's stable up to 10% and then it goes back to the initial condition. So very interesting data from the dvs. What we want to see is a correspondence in the Raman data. So let's remember that everything happened when we do the assortion over 80. So we start from our dry sample. Again perfectly relatable to the data shown before. Now from 0 to 70%. There is not a significant change in the Raman data which match exactly what we've seen for the DBS data. However, when we move to higher partial pressure, we see that the transition happened. If we zoom in on this data right here, you can see that there are the formation of new peaks. When we move over 70% there are shift of existing peak. The intensity of the peak can change. This is in the lower region. So we are talking about up to 1600 centimeters to the minus one. But you can see here there are a few peaks that are disappearing, a few peaks that are shifting in position. Some peaks are doubling. But if we look in the higher section of the spectra over around the 3,000 centimeter to the minus one is very evident that there are changes in the Raman spectra corresponding to the change in structure and the DVS data. So this is very good because the acetone data has no significant change up to 70 and there is the change starting over 80% partial pressure. The DVS data and the Raman data support exactly this type of pressure transition. Okay, so the last stage of our study will focus on creating the solvate first stage and then pushing water into the system to see what happened to our solvent. So again the first stage is we dry the sample, we push acetone into the system and we want to see the formation of the solvent. That is exactly what's happening in our first step. Here is our solved form and it forms exactly over 80% as shown by the data before. So everything is in agreement with what we've shown you so far. Again, the assets on data right here shows the transition step up to 80 and then over 80% exactly as the single component experiment. Now, as the second stage, what we do is stop the generation of acetone and start to introduce water adsorption and desorption. We have to remember the fact that water does not cause a lot of mass change compared to acetone. So there is a little bit of difference in the mass change right here, but is marginally smaller compared to what we've seen so far for the data from acetone. Now, how does our DVS and Raman data compare to. So here is the plot of the isotherm. So there is definitely the significant mass loss indicating that acetone is coming out from the structure. So the material is not stable in presence of water. As soon as we put water in it, acetone disappears from the structure and moves back to the initial form. Does the Raman data agree with the DVS data? And what you can see here is the fact that when we start pushing water in, the structure goes back to the initial form. So in the second stage of the experiment, we are able to reverse back the structure of the solvate acetone back to the initial form of our sample. Now there could be a question right here of so this is the form with the acetone and this is as soon as we put water in it. So reverting back. Now there could be a question right here related to are we going back to the hydroform or are we actually just drying the sample? That's an interesting question because the main problem we have to answer this question is the fact that if you remember, water does not cause a Raman shift and the mass change caused by the presence of water is relatively small compared to the acetone data. So it's a little bit hard to define if we are going back to the initial condition of the sample of or if we are forming back the hydro, we could do powder XRD and see the difference in structure potentially there. But from the DVS data and from the Raman data, at the moment we are not able to provide a definitive answer. It will be much easier if the water dehydrate form had a different Raman spectra. If that was the case compared to the anidrous form, that would be extremely simple because by looking at the Raman data, we could identify immediately which part is generating which structure. We are ending up as the final four. Okay, so this is all I wanted to show you today for this type of application. Obviously you might have some question in regard to these type of application or different type of application that we can do for DBs. So what I would suggest if you have this type of question, is to get in contact with your sale manager or directly to us at Science and Surface Measurement System and we can answer as many questions that you want. Thank you very much.