PCL Hybrid Workshop: Recent Advances in the Characterisation of Pharmaceutical Powders

Okay, so a very good morning to all of you. Thank you for taking on time to join us today. We'll cover some topics on dbs, IGC and Paulo Neurology today. We've also got the Question Answer session for you at the end of the session. This webinar will be hosted by Professor Darren Williams. Professor Darren Williams is the Founder and Managing Director of Surface Measurement Systems Particle Characterization Lab and the professor of Particle Science in the Department of Chemical Engineering at Imperial College. He has published over 125 papers in referred journals and been granted international 5 patents. His research interests include the surface and bulk characterization of complex organic solids, especially biopharmaceutics, foods, pharmaceuticals and polymers, as well as their manufacturing using spray drying, print crystallization, freeze drying, milling and granulation. In 2019 he was awarded the Geldard Medal by the Institution of Chemical engineers and in 2020 the Royal Society of Chemistry named him as Chemistry World Entrepreneur of the Year. His current work in research includes investigating complex materials in range of industrial applications, including manufacture, performance and stability. I would like to hand over this mic to Professor Dariel and let him start the first session that is on the topic Surface Area and Surface Energy by igcse.

Good morning everybody and thank you for attending today's webinar. I'm sorry I'm not able to be with you personally, some change of plans, but I hope you'll nevertheless enjoy the presentations this morning. Let me share my screen and let the work begin. So the first technique I'm going to talk to you about today is a technique called inverse gas chromatography. So firstly, don't worry if you haven't heard about it. It's still a technique that not everybody knows about. But if you're interested in surface properties of materials, particularly pharmaceutical materials, this can be a very valuable technique. It can give you information around the behavior of particular materials and particularly the surface energy, which is something we don't often hear talked about in many areas of particle science and powder engineering. I'll explain all that during today's presentation. So the plan for today's lecture is to explain about this technique called inverse gas chromatography. As I said, many of you won't know anything about it. We'll talk a bit about some quite basic measurements you can do around surface area, and then we're going to spend most of the discussion today discussing the area of surface energy. And first we will have an introduction to what is surface energy, some examples as to why it's important in allowing us to understand powder Behavior, particularly for pharmaceutical materials, in terms of milling, crystallization, and also in terms of powder flow. So that's the agenda for today's presentation. So, firstly, what I want to tell you about is that inverse gas chromatography is related to normal analytical chromatography, which we'll talk about in a moment. It's a technique for characterizing surface properties of solids. It's not an analytical method. It's a method which is used to characterize the surface properties of solids. It's not a new technique. It's been around for a very long time, been around since the 50s. It basically focuses on determining kinetic thermodynamic information for a solid powder, a pharmaceutical powder, in many cases today, based on the absorption of selected probe molecules. It can be used for catalysis, materials, pharmaceuticals, many materials. And we'll see that it's been around for probably 50 years, actually, as a research technique in universities and being commercialized by surface measurement systems over the last 20 years. So what's the idea here? Well, the idea is to use gas phase molecules as probes. So what I mean by that is we're going to use gases or vapors from molecules such as hexane or octane or ethanol to absorb on the surface of the powder. We want to understand this is an adsorption technique. Some of you may be familiar with doing nitrogen absorption for BET surface areas. This technique is a cousin. The difference here is that we don't use nitrogen. We use other molecules that are more interactive, and they can give us a lot more chemical information. They can provide chemical information about the surface and also chemical information around the bulk properties. And as well as looking at surface properties, we can look at the effects of temperature, humidity on these materials. And essentially, as we'll see today, we can measure thermodynamic properties. So from the thermodynamic properties, we'll see that these can tell us a lot about how materials behave in terms of flowing, milling, dissolution, and other critical properties for performance of formulated pharmaceuticals. So I think it would be useful to start. Let me just find a little pointer here. No, the pointer. Is the pointer there. Is there a pointer? Can you see my mouse? Okay, well, I'll assume you can see my mouse. Can you?

Yes, we can see your mouse.

Thank you. Great. So here's a schematic of an inverse gas chromatographic instrument. And the first thing to note, it has a column holder in here. And this column will contain the powder of interest. So this is an important distinction between analytical GC in igc, we study a powder excipient formulation, API, inhaled drug, it does not matter. We study that material and what we're going to do is we're going to introduce a carrier gas, like a normal gas chromatography, typically helium. It's going to pass through the device here, through the column, and the end and FID detector will see in the end. And we're going to introduce different probe molecules, ethanol, hexane, decane. Those molecules will elute through the column, and we'll detect them at the end as we would normally in a gc. So what happens in the experiment is some ways quite simple. As these molecules pass through the column, they adsorb on the surface of the powder. As they absorb, their passage through the device is delayed. The stronger the molecules absorb in the column, the longer it takes from the parts of the column. From these experiments, we can determine adsorption isotherms. And from those we calculate physiochemical properties which are very helpful to understand the material we want to characterize. Now, there are a very large range of properties we can measure. We are not going to discuss many of these today. Probably three. We're going to talk about the surface energy. I'll explain what that is in a moment. We're going to talk about surface area, and I'll remind you about that in a moment. And we might at some stage, and I think just today, we'll do BT surface area and we'll do surface energy. Just two things you can measure using this approach. Oh, okay. We will, of course, look at this heterogeneity maps, which is another way of looking at the energy of a material. So this is an important technique for characterizing pharmaceuticals. It's been used now extensively. I think probably all the world's largest pharma companies using IGC for characterizing their materials. And typically, surface energy is the property that they are most interested in. And we'll see why in a moment. So before we get into surface energy, we're just going to be a little simple. We're going to do surface area. Now, at the moment, you probably measure surface area using nitrogen bet. That's a very good technique. It's very sensitive for large surface areas or high surface areas. Nitrogen bet was particularly designed for characterizing catalysts and porous materials. And for that it is very good. But the truth is, when we're dealing with pharmaceutical materials, typically the surface areas are much lower. And for materials with low surface areas, nitrogen bet is often not accurate enough. Let me explain a bit more so here is a scale of surface area of a material. This is meter squared per gram and we've got one scale up to say 400 meters squared per gram, which is typically for a catalyst or for a porous material. And for that, nitrogen is very accurate and a go to method. But once we get into pharmaceuticals, we very rarely have 400 meters squared per gram. Indeed, we're more likely to have 1 meter squared per gram. When you start to use nitrogen absorption at that level of 1 meter squared per gram, you start to get very large errors, typically about a 10% error. And 10% error is not that good. And what we'll see today, briefly is that if we use this inverse gas chromatographic technique, not using nitrogen, but rather organic vapors such as octane or decane, we can measure surface areas down to about 0.1 meters squared per gram and typically errors about 1%. So much smaller areas and much higher accuracies. And I think for drugs, fibres, excipients, hair, typically 1, 2, 3 would be often the highest surface areas you would come across. So let's have a look at an example of this, just to remind people. The determination of surface areas using BET is covered by international standard ISO9277. Although people think about this in terms of nitrogen, the standard allows for a wide range of molecules and you can see here, molecules such as heptane and octane are also within that standard. So although we tend to think about BET as being a nitrogen or an argon or a krypton only experiment, that's not entirely true. You can do the experiments using octane, heptane and many other molecules at room temperature and at room pressure. And of course they're much more relevant conditions for your material. If you're making a pharmaceutical material, it's very rarely going to be used at 77 Kelvin, which is a temperature you do nitrogen absorption at. So doing experiments at room temperature is also a bit more realistic. So let's look at some data. This is some work done on a reference material. This is CRM170. This is a material that you would use to quantify or validate a nitrogen BET analyzer. But of course we can use the same material to validate our inverse gas chromatography determination of surface area. So using a 500 milligram sample running at 30 degrees, we can determine adsorption isother material here. So what you can see is over here. This is the adsorption isotherm for octane. So It's a hydrocarbon C8. This experiment is done at a basically at room temperature. And you can see here, we have done the isotherm on IGC and we construct the adsorption data using the BET model here. And this reference material gives us a surface area of 1.01 meters square to gram within the experimental uncertainty for a reference standard. So this is something quite simple. But again, if you're dealing with low surface area pharmaceuticals, and if you've got materials that are 10 or 20 microns, they will be low surface area, then this IGC technique can give you very accurate surface area determinations. And just to give you some idea around the accuracy of this technique, here we looked at some paracetamol powder and these were quite large particles. And you can see here the surface area is actually 0.3 meters squared per gram. Very low indeed. Too low to do nitrogen bet with. You would be wasting your time doing the experiment. You can see here we're able to measure surface areas of 0.727 meter square per gram. An error typically 1 or 2%. That's probably an order of magnitude or more more accurate to using nitrogen. So that's something to be aware of. If you are characterizing particular materials that have low surface areas, this IGC method is very, very accurate. One thing you can do with the IGC method, which is slightly unusual but sometimes helpful, is you can determine the surface area of materials at different humidity. You cannot do this using nitrogen, of course. Course. And you can see here, for this cotton fibres, the surface area is different when the material is dry to when it's wet, due to the cotton fibers being swollen by humidity. So surface area as a function of humidity is something you can also do using your IGC method. But the main focus of today's presentation is around surface energy. That's the main measure we're going to discuss today. So let's talk about what is surface energy? The easy way to think about surface energy. It's very much like the surface tension of a liquid. Liquid. If you wish to expand the surface, you must put some energy in. That's called the surface tension. To create new surface of a liquid, you need to put a small amount of energy. This is the same for solids. And technically it's the excess energy at the surface of a material compared to the bulk. And it's a direct measure of the thermodynamic properties of the surface. And really what it measures are the intermolecular forces that are at the surface of your particle. We will talk about two different Sources of surface energy components Today there are some surface energies related to long range physical interactions, the sort that are related to nitrogen absorption. We call these dispersive interactions London van de Veldes forces. They involve non specific interactions between the molecules. Also for some molecules that have active functional groups, hydroxyl groups, amine groups will also have acid base interactions. These are short range chemical interactions. So when we look at the surface property of a material, such as a pharmaceutical solid or a catalyst, we will consider the surface energy as having two components, a short chemical direction and a long chemical direction property. So I'm measuring the surface energy solids. There are lots of ways to measure surface energy solids. It depends on the form of the solids you've got. So here I summarize the sort of techniques that are out there that you may already be aware of. Atomic force microscopy, contact angles, weighting balances. They are very good for measuring the service energy of a flat material or even for fibres. But for particular materials they are virtually impossible to use. So if you want to surface energy, if you want to serve the energy of a powder, inverse gas chromatography is the only technique in place. Inverse it as we'll see soon, it's excellent for particulates, it gives repeatable data. There's no hysteresis or roughness effects. We can use a wide range of solids and we can also determine the heterogeneity, the variations, energies. So for particular materials, IGC is the go to method for doing those sorts of determinations. So I think the question we might ask is why do we care? Why do we want to measure these properties? What's the point? Well, there are a number of things you can measure and learn about by measuring the surface energy. If we want to understand how particles interact with each other, and that is powder cohesion, we need the surface energy. If we want to understand how vapors interact with solids, which we call sorption, the surface energy helps us with that. If we understand how liquids will wet a powders wetting behavior is also dependent on surface energy. And if we want to know will our powders adhere to the walls of vessels or to other excipients or active particles in a formulation, then we want to have adhesion. So any of those phenomena involve intermolecular forces, and those intermolecular forces are directly probed using the inverse gas chromatographic method. That's why we want to know about it. So when we think about pharmaceuticals, why do we care? Well, if for instance, drug excipient Interactions, adhesive and bonding interactions, powder mixing, segregation, powder flow, static charge. Surface energy is a factor in understanding those properties. Not the only factor. It's a factor. Things like particle size will be important. Static grouping, static charging is important. Surface energy is one of the properties that defines some of those interactions. If we're interested in process induced disorder milling, grinding, microsation is a very common unit operation for manufacturing of drugs. But it changes surface properties and it changes product stability. And that's critical. It can affect the ability of a material to become crystalline or amorphous, it can affect polymorphism, it can do a lot of things to complicate your material that you want to know about. Batch to batch variability probably the most common problem we come across using inverse gas chronography is the following. I've got two batches of a pharmaceutical. One was made here, one was made in our other plant. I can't get them to mix, I can't get them to granulate, I can't do something with the manufacturing wise, even though in theory they are identical. So when you get problems which are outside analytical chemistry, sometimes it turns out surface properties are the critical differences. And of course there are also fundamental material properties which we may want from time to time. Surface properties, surface acidity based entities, surface modifications, coatings, stability of surface, functional groups, these are all factors that will be affected by the surface energy we measure. So we can measure surface energy, we have a handle on those sorts of measurements. So here are just some examples of how surface energy affects certain sorts of behavior. So wetting, you can see here a series of droplets on a surface. The surface at the top, this, this surface here has the highest energy. Droplets want to wet that surface. But these surface here that has the lowest energy drops do not want to wet. So if I'm trying to disperse my powder particles into a liquid media, this is really, really important. Also powder cohesion, as the surface energy goes up, powders will stick together more, there's more cohesion, that might be good, that might be bad, but cohesion will directly be affected by surface energy. And also surface energy is a measure of disorder at a surface. So if we start milling a surface that tends to give us high disorder and high energy sites. So the surface energy is a way of understanding disorder at a surface. Particle cohesion and the wettability of particles, all important problems for certain classes of research or process development. And finally, adhesion. If we're developing respiratory drug systems where the adhesion between our active API particles and the carrier particles is important, IGC is probably One of the most important techniques for understanding how the particles adhere to each other, how they adhere to the carrier particle. And there are very few ways of measuring those surface properties except igc. One of the things we're going to talk about today is heterogeneity. And let me talk what I mean by that. For the sort of materials we have here, we tend to think that our pharmaceutical materials or our excipients are all homogeneous. They're all the same. They have the same surface properties, same surface chemistries, and everything is uniform. It turns out in the real world, this is rarely the case. And actually, we tend to get a lot of heterogeneity in our complex APIs and the excipients as well, and formulations. And why there's heterogeneity? Well, sometimes the reasons are really quite simple. Just the presence of contaminants and impurities, more fundamental reasons. Different crystal planes can exhibit different chemistries. There are different crystallinities or morphicities. There are a range of morphological states, polymorphs, hydrate, solvates. We sometimes get different surface groups or different surface modifications. So these can all affect the uniformity of the surface properties and indeed will make them heterogeneous. And also they were called structural heterogeneities. Crystals have defects, steps, edges, pores. There are irregularities of crystal lattices, and there are pore sizes of different sizes and shapes. So these heterogeneities cause differences in adsorption behavior. And therefore the IGC technique, from looking at different strength absorption, can understand these heterogeneities. Indeed, it's the only way you can see them. And these can be one of the reasons we get batch, batch variabilities, which we'll talk a little bit about later on. So a little bit of adsorption science here. In adsorption, we are measuring the amount of a species adsorbed. So it could be hexane, could be ethanol. We're measuring how much is absorbed as a function of the concentration of that molecule in the gas phase. If we're doing bet surface area, as in nitrogen, we do that at quite high concentrations of molecules. But a lot of the work we'll talk about today is done at lower concentrations, which we call the Henry Law region. Sometimes that's also called infinite dilution. And these tell us a bit around the surface properties at low surface coverages. Here are low surface coverages, and here are high surface coverages. So here a lot of the surface is covered by my adsorbing molecules. And then here only A small number are covered by absorbing molecules. And studying down here is a good way of me finding the hot spots, finding those sites that are particularly energetic, which might have particularly important contributions to the properties of my overall material. So I want to talk to you now about. Let's talk to you now a bit about surface energy heterogeneity. So if I talk to you about particle size, you and I said to you the particles were all 10 microns in size, you wouldn't believe me. You'd say, I want to see the particle size division, and rightly so. So it is the same for surface energy. The particles rarely have the same surface energy. So we have here what's called a surface energy distribution. So let's start by looking at the. Right here, it's plotting surface energy. And this is the population shown here on the Y axis. And what you can see here is that just like we get particle size distributions, we can get surface to energy distribution. So here the blue curve says that there's a very broad range of distributions and the average is quite high for the blue material. And it says here for the red material shown in the red curve. The red curve has a much lower energy and a much more uniform distribution. If we replot that in a different fashion, where we plot serves energy as a function of coverage, you can see here that the red curve is much more level, okay? And it has a lower number, whereas the blue one, which has got more defects, it has a higher average and a broader distribution. So this is not unlike part and size distribution, but it's about energy now. And energy is important. And we can also measure the Average service energy, gamma, 50 new materials highlighting the different surface energy for 50% of the sites in that population of particles. So now I'm going to take you through a little project where we decided to investigate this technique and its approach in a lot more detail. And we used a material that you're quite familiar with, which is mannitol. And what we wanted to know is whether or not the data we measured using inverse gas chromatography would agree with the data we would get using wettability measurements. And just to help you a little bit here, Conant angle or wettability measures are the standard method for determining the surface energy of flat solids. Of course, what you know here is wearance and powders. We did a special project where, as you'll see very shortly, we grew some very large crystals. And those crystals allowed us to understand the surface properties macroscopically and to compare those with the microscopic surface Properties. We're going to have two materials in our study. We have D Mannitol which is an AS received, the same material you'd buy for manufacture, and another one which is cyanide. In other words, we are going to put methyl groups all over it and make it very hydrophobic. We'll do two things then we'll use contact angle measurements using sessile drops to look at the surface properties of these large crystals. And then we're going to look at the characteristics of powders with the same surface chemistry. Let's look a bit more detail now just to remind you. This is the structure of mannitol. As you can see it has a lot of hydroxyl groups. So potentially it's a very hydrophilic material. And here is a photograph of some of the materials that we crystallized in the lab. As you can see, they're big crystals. You can see there 120010 and 011 are the primary crystal faces for those. And we'll hear more about those crystal faces in a moment. But I can tell you that the surface chemistry of each crystal face is different. That's why the crystal has a long aspect ratio because the crystals have different surface chemistry on each of those facets. So the first thing we did is content angle measurement and I'm not sure if you're familiar with that, but let's talk about it briefly. We put a droplet of a liquid on the surface when we typically use two different liquids. For that we use a liquid that does the van der Waals interaction. I only called diodomethane. It's called summarize here dimm. It just measures long range van der Waals forces. And the angle that the droplet makes in this case it's going to be on the crystal surfaces allows us to determine the surface energy using contact angle measurements. And if we want. So this tells us about the long range forces. Water is a very hydrophilic polar molecule. It tells us around the short range forces. What I'd like you to notice here firstly is 011 notice here the very low Conant angle. In other words, the water molecules when we do this experiment are very strongly attached to this crystal facet. Reason we'll see in a moment. There are lots of hydroxyl groups on the end and there are very few on the sides of the crystal. So in other words, the crystal has different properties depending which facet you look at. That's why it grows in solution the way it does. We can determine the surface energy of the materials. And what you can see here is that this is gamma D, this is the dispersive van der Waals forces. They're similar. I'll just get my mouse back here. They, they're similar for all three faces. Gamma P is polar, which is due to the acid base forces. And you can see here a big difference between the O11 facet and the other facets. Where's the mouse gone? There it is. You can see here it's twice as high. And it's twice as high because there are lots of hydroxyl groups on the end of that face, the 011 face. And we'll see in a moment. That's important. So this is a very hydrophilic face and the other ones aren't. If we do X ray photoelectron spectroscopy, which is a way of determining the population of groups on the different crystal facets, we'll see here the 011, it has a large concentration of our H groups, probably about four times to three times higher than on zero. So on the 1 2, 0 facets, in other words, there are lots of hydroxyl groups here and not so many there. So let's just think about that. What's that mean? Well, just imagine I took that crystal and I milled it up. And in the. Excuse me. And in the milling process, I created lots of 011 facets that would have very different surface properties to the material at the start because I would have created a lot more of these high energy.011 sites. So just briefly, we cyanized these materials, we made them hydrophobic or some of them. And we did that work on both the single crystals and on the powders. And what you can see here is that when we cyanized them, you can see the surface energies are a bit lower here, all about 35 on each facet. So a very little difference. And they're a little bit higher on the untreated as received mannitol. But when we cyanized them, they became very uniform in terms of their dispersive energy. So now what do we want to do? Well, what we're going to do now is we want to compare the IGC data, which is measured the surface energy of a powder. We now want to compare that with the wettability data for the same material. And the question is, does IGC give me the same sort of data that I get with wetability, which is the classic method? And what we can see here is that for these materials, the untreated D, mannitol Cyanide, mannitol. And also we did a racemic ibuprofen. The data is in very good agreement. So that just is a way of us validating the IGC method gives us data that's consistent with other approaches in the literature. So now we're going to do some, start looking at some applications. This is all very good in theory of what's it mean in the real world. So the first thing we did is a study on powder flow. So I created two different batches of mannitol, about 50 grams of regular mannitol which got from the container, and we created 50 grams of cyanide mannitol. And of course you saw before the cyanized mannitol has lower energy. So from our igc, I can compute the work of cohesion between the powders. And you see here that the blue one, the cyanized mannitol is much lower, it's got lower energy. Therefore we'd expect that powder to stick less. Okay, and we're now going to see how does the material behave when I go into a flow. And also to be clear here, the particle sizes are identical in these materials. The only difference is one is hydrophobic, it's solidized, and one is hydrophilic. The question is, does it affect powder flow? This just gives me the equation that I use to calculate the surface energy and the cohesions. So here is powder cohesion taken from an FT4 study. And what you see is the total flow energy here. Here is the solenoid data. Here is the data for the astroceived mannitol in red. And what we can see here at low air velocities, where there's very little aeration, there's a very big difference in the flow energies. In other words, the cyanized mannitol, which is very hydrophobic and very uniform, flows very, very easily. But the addressive mannitol, because it has much higher energies, it actually flows much more difficult. It takes a lot more energy in the system for it to flow. So basically the solenization affects the behavior of the material flow wise. And again, the particle size is identical. The only difference we're looking at here is just a difference in surface chemistry of the particles. So let's look at another example. This is some work on aspirin. So what we have here are two different aspirin materials. The particle sizes are the same. Okay? Particle size is the same. One has been milled down, the other has been crystallized to the different size fractions here. And there is also no amorphicity. These materials are 100% crystalline, but one is milled and one is unmilled. So you might expect within that there are differences in particle shape and there will be different particle shapes. But you can see here, we're looking at the dissolution kinetics here. This is given by the UV vis absorption. And what you can see here is that the milled material here, which is blue, it dissolves much faster. Okay. Compared to the unmilled material as a factor of probably 20 or 30% difference. And it doesn't matter what the particle size is, the milled material dissolves faster. So why is that? Well, if we look here at the inverse gas chromatography, we can see here that the milled materials have much higher surface energies, and those surface energies facilitate faster dissolution rate. So we can see here that for materials that are normally identical, the only difference is surface energy. The surface energy has a direct link to the ability of the material to dissolve. That's an important property for us when we're developing our materials. And we estimated the milling change between 20 and 25% of the surface groups and has a measurable effect on the dissolution of the final product. So that's important. Now we change into a different material. This one, I should say, I think this is ibuprofen. And what we're seeing here is as we change, we've milled the materials, they're different shapes of particles. And as we do that, the dominant facet in the materials changes. So when you start milling materials, the facets which have the lowest attachment energy, they start to dominate. And that change reflects in the actual surface energy we see. So this is IGC data for different sieve fractions. And what we're seeing here is a change in the population of facets as we mill the material. Here's another study on mannitol, same idea. Here we're milling the mannitol. We have different aspect ratios of the material. And you can see here the surface energy is changing as we mill the material and as we change the population of facets. So milling affects the different sorts of facets that are present in the powder. IGC can see that because you can see different facets. It's probably the only technique that can give you facet relevant information. So let me start just to give you some conclusions. We've seen today that IGC can measure the surface area of low surface area materials. So materials that are only 1 or 2 meters square per gram, these materials can be measured using igc. IGC can be used to map surface energy variations of powders using statistics known as an IISSD based approach. The IGC data on powders agrees well with single crystal wetting studies. We can see the energy variations for milled APIs which probably can't be seen any other way except using igc. And even something simple like powder flow can be related to surface energy which is an important. So do excuse me for the odd cough during my presentation. I thank you for your time and attention and I'll be happy to have any questions on the chat. I think it's probably the best way of doing that. Thank you. No questions.

What I'll suggest is let's give it a minute or two for the audience online and if not then we go ahead and we proceed. I think there is a tea break, 10 minutes.

Tea break? Yep.

Let's just give it, give it two minutes.

That's fine. No problems.

Now we've got one question here. I'll read it out loud for you. Can you see me okay?

Yep, go for it.

Okay, so we have how to calculate dispersive surface energy and polar surface energy from contact EndoVascience.

There's a couple of well established, quite simple theories to calculate gamma D and gamma P from con measurements. The one we use is technique reported by folks fowks and essentially from the wettability data for di dyomethane and from the wettability data of water. It's a fairly simple equation which allows you to calculate gamma D and gamma P using that. And that's our preferred method to do that. Look up folks theory, gamma D, gamma P that you'll find that very easily online. Thank you.

Darc, there is another question here. What is the correlation between surface energy and mass flow?

Mass flow. So we think powder flow.

Yeah.

Well, powder flow will scale with a number of factors. It will scale with particle size. Probably most significantly, it will probably scale with distribution the uniformity and shape of the powders a lot. Surface energy is one of the factors in the mix. So for example, if you had two powders that were identical shapes, identical sizes, then the surface energy would be coming the determinant factor. But particle size, particle uniformity, particle shape would still dominate. So differences in particle shape and size would be probably most important. Surface energy would be a secondary factor, I would say. Thank you.

Daryl, There is another question here. I think you covered this in the beginning of the slide, so maybe if you could repeat this part. It says what are the inputs required for surface energy determination?

Sure. So if you're doing it with Contact angles. You need the advancing contact angle for water and di iodomethane on that substrate, that flat surface. So you need the wettability data for the two liquids on the flat surface. And then you can calculate the surface energy using wetting. If you're doing igc, you need to do a number of injections of different probe molecules across a range of different surface coverages to extract from that. Right. So next question. I can see in front of me here, effective surface area. So when you mill an API, you're doing three or four things at the same time. This is where it's complicated. Yes, you're changing the surface area for sure. Yes, you're changing the server energy for sure. And yes, you're changing the amorphicity as well. So when you do milling, you are doing three things at the same time and they have in a complex way. So definitely you're going to change the surface area, you're going to change the surface energy and you're going to change the surface of all three things happen when you do that milling and teasing out which of those is more important is tricky.

Can you see the questions from your slide?

Yep, yeah, I can see them. Static charging, which I assume is what you're referring to here, is it is a difficult problem, certainly I would expect using igc, but I mean static charging of powders typically relates to polar groups on the material. Polar groups on the surface of the material. Those groups will typically be acid base groups. And we would expect igc, you would expect IGC to be sensitive to many of those surface groups. So I would say that the IGC method would tell you about the polar and non polar acid base groups and that might help you a little bit in terms of understanding static charging, but it is a difficult problem. Well, IGC gives you. IGC does give you information about adsorption which we then use to give us surface area and surface energy. So it's adsorption information which we then turn into surface properties. Does that make sense?

Thank you, Daryl. We have one more question here from the audience. I'll ask him to ask the question directly just so there's a camera.

Yeah.

Hi, Professor Darrell, this is Manu.

Hi there.

My question is specifically for the APIs which have been obtained from reactive crystallization. So sometimes we face that even after achieving a similar physical characteristics like particle size PETB or those shape of the crystal, we still face sometime variation in the resolution. So do you have any certain experience like particularly for hydrochloride salts or the API which have been obtained from reactive crystallization. So that, so this is your so far presentation. I could refer that there may be some change in the surface energy, but can you put more light to that?

Yeah, I caught most of that. I think there are lots of examples where you, you make formulations or products which based on all the measures you've got should be identical. I mean they got the same particle size, the same particle shape. Everybody says these are the same. But when you go to do dissolution or other measurements, they turn out not to behave the same. Now I think it's fair to say that we've seen over the years quite a few examples where the only technique that could spot a difference was inverse gas chronometer. Let me give you one example that was quite an old example. Quantum interesting one. We worked with a major pharma company in the USA quite a while ago and they were making a new age drug. And the age drug, age drug involved a wet granulation process. And the problem was that half the batches were faced were failing release because they wouldn't granulate properly. And they raman, they did everything and they, for them these materials were all the same and they were throwing out half the batches. So we ended up having the only technique, IGC which could spot the batch prior to granulation. They would then change the formulation for the batch and it would pass. So the IGC was very sensitive to surface energy and that turned out for granulation to be critical. So I can't say to you IGC is going to solve all your problems. But what I can tell you is from time to time there'll be problems where everything else seems the same and you don't have an answer. From time to time IGC will spot the difference.

Darren, I think we have time for one more question.

One or two.

One of them is how reliable would you say IGC data would be when you're dealing with highly porous materials?

Let's define highly porous. Give me a bit more information. Do you mean high surface area or high pore volume? Just tell me a bit more.

Okay, we can ask a member of the audience, rephrase that question.

Yeah, we can do porous materials, but it's not ideal for. If you've got a catalyst material of 400 meters squared per gram, IGC is probably not the best method. There are better methods to do it if you want to do that. So but if you know, if you, if it's 50 meters squared per gram, IGC is pretty good. But for very, very high porosities, I would probably not use igc. I'd use dynamic vapor absorption, which I'll talk about in my next lecture.

I can read on the last question I've got here before we can go for the break. And it says when you pass a gas, chances are that it might be absorbed as well.

And the surface.

Sure. So let's talk about it. So in our experiments, there are typically two species present. There's helium, which is the carrier gas, and there's the absorbing molecule, say, octane. Our assumption, which is generally pretty good, is that helium does not adsorb on the substrate. So we are basically studying the absorption of heptane, for example. So in gas phase phenomena, we can pretty well ignore the adsorption of helium on the surface. Helium does not want to absorb. It's not going to at room temperature is not going to happen. So we're very much only looking at the organic vapors that we inject. We do not have to worry about the actual carrier gas in this case. Okay, yes.

Thank you so much. So what I'll do is we'll collect the questions and we'll answer there directly or in the chat Directly to.

Yep, no problem.

So being a hybrid workshop, what I'll say is I'll say thank you. Let's take this 10 minutes. We'll reconvene exactly in 10 minutes. So we'll do the same here online. I'll turn the camera off, but for our colleagues online, please stay there. We'll be right back in just 10 minutes.

Thank you.

Bye. Bye. Everybody's in the room, so we're ready to start whenever you're ready.

Okay, so can share my. Okay, thank you. Okay, share my screen. Is that clear to everybody?

We can hear you. We can hear you.

We can see you.

Okay, here.

Great. Fantastic. Okay, well, welcome to installment two on today's workshop on advanced characterization of particulate materials. This presentation is onto a different topic. We're going to talk really around materials now that have been freeze dried materials. So before we talked about the characterization or surface properties of pharmaceuticals and APIs, we're now going to talk about the behavior of freeze dried materials. And we're going to use a different technique called dynamic vapor absorption, which I'll explain to you during this presentation. Some of you, I'm sure, involved in manufacture of freeze dried materials. An important part of freeze dried materials is that you end up with a material that's effectively amorphous. If it turns out that material has been freeze dried from water, it means in reality that it's going to be very sensitive to water in the environment. So in this particular project we're going to look at the way moisture can affect the collapse of freeze dried cakes. And this is a part of a collaboration with NIBS and also psrc. So we're going to look at the relationship between stability and particularly moisture content, cake structure and other physiochemical properties of real world freeze dried products. We're going to be particularly interested in how water vapor interacts with the materials and how can we look at a range of different freeze dried materials, including model materials, as well as reference freeze dried biologics provided by National Laboratory. We are particularly into understanding in the formulation space what can give us best product stability and particularly how stability relates to the glass transition temperature of the freeze dried solid. That's important part of the presentation. So of course we need to recall that water, and particularly protein hydration, is very important in the structure and activity of many biologic materials. And if you take too much water away from a protein, it'll lose all of its biological activity. So the structuring of water around proteins in the solid state is very important. And you're typically looking at a couple of nanometer surface layer for many proteins in solid state systems. If we remove too much water, we will denature the protein and will lose the structure that gives us the biological activity. So water has an important role in stability of materials. The question is having too little or having too much. So just reminding you about freeze drying as an important manufacturing process for biologics. There's monoclonal antibodies and peptides and other biological materials. We take the material we want to stabilize. And just a bit of background here. If we have a biological material in liquid that is a solution, it has a relatively short shelf life. If we can get that material in a solid state formulation, it has a much longer shelf life. And a very common way of turning a material into a solid is known as freeze drying. So we describe the process broadly. Firstly, we create a solution of the material used to freeze dry. We then freeze the material below room temperature in a freeze dryer, typically at minus 20 to minus 50. We then, as the diagram shows on the right, we then use a vacuum to remove the water molecules. So we sublime away the water molecules. We do that, we end up with a porous cake which contains only the active product we want, typically in an amorphous form. So if the product originally had 5% in solution, we'll end up with a solid that is 95% air and solid. And this is described the product is described as a porous, normally dry cake. Now what you can see here is that these cakes, and you can see one on the right looks nice and stable, but the one on the left has had a bad experience. Some water vapors got it into that vial and the materials collapsed. So that looks like there's nothing in there, but there's a small amount of material in there. There we are, which is the cake that's collapsed. So it's really important in freeze dried materials to keep water vapor out, keep humidity out. And the question is how much humidity or how much water can you have in the material and not compromise its performance. If there's too much water in the freeze dried cake, it will collapse, we'll get degradation and loss of biological patency. And the moisture levels in vials can change over time. And what we'll see in a moment is that over time as water can get in, I mean these stoppers here, they will let moisture in. And if you're unlucky, the moisture gets in, it can cause the cake to collapse. So the exercise in today is to try and understand what are the critical conditions by which we can keep this cake stable. So we're going to introduce a new technique now called dynamic vapor sorption. This is the method we're going to use to characterize how much water is in our freeze dried products. This is a gravimetric approach, so its precocious cousins are tga. Now what we're going to do is take a small piece of the freeze dried cake, we're going to weigh it and then we're going to look at how much water it absorbs as a function of time for a particular temperature and a particular humidity. And that's going to tell us essentially allow us to understand the stability of this particular solid state amorphous freeze dried material. This just gives you a simple schematic of the instrument. So our sample goes into here and typically 5 or 10 milligrams we're going to flow humidified air over this. So we have air coming into the instrument. It's humidified, it passes over the sample. Some of the water molecules go into the sample, some don't. And we will see a change in sample mass. The more the mass increases, the more water has gone into my sample. So that's the idea of the experiment here. And just to take you back to a bit of basics here, what we're going to be looking at is determining how much mass here in percent goes into our sample. As we increase the humidity Higher humidities, more water goes in. And we're going to be wondering within this research, where are the actual water molecules? Are they in the area of multi layers? Are they, where are they? So I think there's a question here that we can answer. Are the water molecules on the surface? Are they in pores multilayer? Are they dissolved inside the material? Have they reacted to form a hydrate? And we'll see today that 4 and 5 are the main places water molecules go in freeze dried solids. They either dissolve into the amorphous domains or sometimes they can form hydrates. We'll see an example of both of those today. But basically it's the mass versus humidity is our primary output from these experiments. So just to give you some example of the sort of data that comes from dvs, so here we have a rice starch, so just a powder, 10 milligrams or so of rice pouch rice starch powder. And the blue represents the humidity changes. So we started at 0% humidity and we've gone up to 90% humidity. So we stepped the humidity up in the blue and stepped it back. And the sample mass, the sample mass responds. On the red line here is change in mass. And you can see that at a high humidity here of 90%, rye starch has about 22% water present. Rice starch is quite amorphous. It has a capacity for absorbing a lot of water and it can absorb essentially about 22%. And the reason we do these experiments in two cycles is we're looking for any changes in stability. Sometimes when moisture goes into a material, the material changes and when you repeat it, you get a different result. And if you look very carefully, there's a very small decrease. My mouse, this result here just very slightly lower than this one, but they are broadly the same. So this is the kinetic, this is the mass versus time. And from each of these points that you can see where I'm stopping with my mouse, you are now going to get an isotherm point. So where's the. Okay, fine. You now get, you can use each of these points to get you the isotherm. So here is some data for some sprinkles, dried alpha amylose materials. And you can see that these materials have taken up about 15% by a mass of water vapor. And this is done at different temperatures and there's some differences in temperatures of uptake. But basically this material is taking up quite a lot of water, which is quite common, quite common for freeze dried materials. So one extra thing we're going to do in our experiments today is not only are we going to measure the change in mass, but we're going to take some pictures. We're going to look and see what is our material doing. So in the instrument we've got a small high resolution camera and it allows us to take pictures of the sample as we do the experiment. And remember, we're looking for collapse of a freeze dried cake. So we hope that we might see that. So if you look down bottom here, this is the sample pan and in that pan is where the sample goes, the powder. In this case we're going to put in some freeze dried material, a small sliver taken from the cake. And this gives you some example of what you can see using this simple microscopy technique. Here is a spray dried lactose powder. It starts in image A down here in the bottom left down here this is a freeze dried powder at, sorry, spray dried lactose at 0% humidity. By the time I've got to 50% humidity it's gone from a powder and it's formed a gel, same material. A couple of hours later after 50% humidity then the material has taken on so much water. Now it starts to form a liquid like material and finally it crystallizes and you can see the actual change in opacity as crystals are formed. So amorphous powder, gel, liquid, crystalline salt. We can see all these in the instrument as we change the humidity and as we observe the sample and that can be very, very informative. But in this case we're not doing spray drying. In this particular study we are going to freeze drying. So here is the study we're going to do. We're going to look at some freeze dried samples, typically 5, 10 milligrams at 3 temperatures, 15, 25 and 40 degrees Celsius. And the samples are listed over here. 5% sucrose, 5% trecherolose in aqueous solution, 5% dextrin. And we're also going to study some commercial products which are some actual influenza antigens which is shown listed down here, right down there, four different samples. We're going to test at three different temperatures and we're going to ramp the humidity up from 0 to 90% and take pictures and see what happens. So it's a visual study. We got the mass data as well, but the visual pictures will tell us a lot. So this just shows you a schematic. So we can see the sample pan is here. There is my sample. Just a piece, a piece of the freeze dried cake. We are going to ramp up the Humidity and we're going to look for any mass changes which can sometimes tell us about the TG or crystallization. So that's the experiment. It's a video experiment on a small sample. It's shown there. So let's start with some simple experiments. So the first experiment we're going to see is a ramp for 5% Trekulose. So that's the ramp up and down, up and down. The next thing I'm going to show you is the mass change. The mass goes up to here and actually it gets to about 11 and a half percent and it drops down a little bit and it stops at about 10%. Well it turns out, it turns out that Trekalose forms a dihydrate about 10.5% by mass. So in the case of the Trekalose cakes they simply have formed a hydrate and not much else has happened. Interesting. But more data to follow. So here is a picture and you can see the humidity here going up. Here is my trichalose sample right there. I'm weighing it and I'm measuring it, looking up increasing humidity. And the question sees does that material stay stable or does it collapse? And you just keep your eye on that and see what we see over the next minute. Ah, here we go. So what that says is at about 52% that material collapses. So you're going from being a freeze dried cake to actually having a drop of liquid solution. And we're just seeing it again. We'll just watch, watch that once more while we're here. And again that's the piece of cake. You can see where I'm tracing with my mouse. Right there, that's the cake. I take it with a razor. We cut a piece of the freeze dried cake and put it in there and it's quite stable. I'll put my mouse just there for a moment. You see that my mouse there, it's stable to about just under 50% and then it goes. So we can see that Treeclose does two things. The cake collapses a and during the process it also wants to form a actual dihydrate. So two things happening there. So let's press on now. Okay, so we can see here, this is the mass trace in red and the first sign there is a collapse which is here. And the second event is the formation of the actual dihydrate. So we can see both those events by picture and by video for Trekalose. So now we're looking at this is a commercial influenza antigen. Again freeze stride. And you see there that that collapses quite early on at quite a low humidity. Not very stable material. Okay, and we press on again. This is dextran. Dextran was very stable and between 0 or 5, 0 and 90% there was no collapse observed at all. It was stable at all humidity. So it's a very, very stable simple freeze dry system. This is obviously data here for sucrose and for tricholose. So you can see here there are two events associated with the trekalose and the sugar does a number of other things involving recrystallization and other changes of the material. So what we decided to do was to standardize a method for determining at what humidity the material collapsed. And we took as an arbitrary metric, when the cake area had changed by 50%, we called that the collapse point. So in other words, when the initial area you see in B goes down to be half the area, we call that collapse. Okay, so what we did was we looked at measuring the collapse humidity for these materials at different temperatures, the different formulations we have. So here are some of the data we have. This is sucrose, 5% by mass. And you can see here at the different temperatures there are different humidities for stability. So at the highest temperature we have a lower humidity of stability. So we can actually map, so above this line here the material will collapse and below here it will be stable. And we can do that for trichalose and for the flu antigen and we can plot them all up together and surprisingly they're somewhat similar really. So we can now put a general plot as to when our freeze dried materials are actually stable. Now of course, you know, we do of course recall that we haven't put a dextran because the dextran was actually stable at all humidities. But for the ones that were unstable there, we could map out a stability place for the trichulose, sucrose and the flu antigen. Stable down here. Oops, sorry. So we're stable basically down the bottom and up here we're at unstable. What I mean by unstable is above that red line we are at risk collapse. If the humidity and temperature is above that line, there is a risk that this material will collapse. But below that line the material will generally be stable. So we're able to use the DVS with its video accessory to basically do a stability map for freeze dried cakes. So conclusions here are that humidity induced collapse of freeze dried cakes can be successfully observed in real time images. Using the dvs we can record eventual change in phases induced by humidity variations for quite a wide range of freeze dried biologics. And we can look at the collapsed behavior of any cake routinely and determine some fairly accurate guidelines for the still region stability of that particular formulation. And the person did the work is Arnold. And Arnold over there on the right. Okay, so Arnold did this for his PhD. So I thank you for your time and attention and happy to move into a mode to answer any questions you might have.

Daryl, we have a question coming through and it might be a moment, so I'll just read it aloud. So what are the possible causes, the different causes that for powder caking.

Well, can I make a suggestion? I'm going to talk about powder caking a bit later on when we talk about powder flow. But I mean in short, you're looking at either deliquescence events on the surface of particles or looking at plasticization which softens amorphous particles. They're probably the two single biggest factors. The change in mass on the Y axis is actually dry. The actual mass. Dry mass. Yeah. The Y axis will be a change in mass of the sample relative to the initial starting mass. That's the question. That is correct. So we will call the starting mass probably 100%. I just want to see what I had in the slide packing on. Yes. So typically we refer to the starting mass sometimes as 100%. Okay. So the mass will go from 100% up if it absorbs water. Sometimes it will go down if water is released that's been captured or froze into the structure. So although 100% does seem strange number, it is sometimes useful for any events that involve the loss of water.

We have one question from the audience, Daryl, so they'll just speak directly to you.

Sure. Go.

Professor Darrell, how can we determine the percentage crystallinity or percent percentage of opacity? With the help of.

Yeah, sure. The question is how can we determine the morphicity or percent of morphicity or crystallinity using dvs? Maybe that could be my next workshop seminar in the next time. But basically the way it can be done, and it's very well established in literature, you choose a molecule, could be octane, could be ethanol, that ideally dissolves in the amorphous phase, only does not dissolve in the crystalline phase. So you can use the absorption of a specific molecule on your amorphous materials to work out the percent amorphous content. Now to do that measurement, ideally need a 100% crystalline sample. And ideally you need a 100% amorphous reference sample to do that. But many, many people do that and devis is probably the Easiest technique for the determining amorphous content for formulations where it's not 100% or not zero. And there are many, many papers that we can send you references to where devious is an established technique for determining amorphous cones down to as low as 0.1 to 0.2% up to very high concentrations of amorphous solids. If you email us offline, we can send you up. There's a couple, a couple of quite good review papers out there on this topic. We can share that with you in the literature.

Okay. I don't think we have any more questions coming at the moment. What I would stress is if there are more questions that come through, then please do email them to us@infoaticlaboratories.com or directly Daryl at the email that you see on these slides and we'll just get to you. So we are right on time for another break. This one is a very short, quick break just for a quick glass of water and be back. So another 10 minutes and the next session will be exactly at 12.

Well, I don't know what time is.

Anybody in exactly 10 minutes. And it's water vapor in compens organic materials. So we reconvene exactly in 10 minutes.

Thank you.

Yes, Dar, we are ready now whenever you want to start.

Okay. Hi there. Yeah, right. Good afternoon to everybody for joining the third part of our little workshop today. We're going to be spending some more time in this session discussing watersorption in complex materials. Some of the materials we'll see today are natural materials. They're cellulose fibers and some are made main materials. And we'll be spending most of our time in this session looking at a technique called dynamic vapor absorption. So the background to this is the fact that there are many organic materials that are hydrophilic and they absorb a lot of water and for example, cellulose materials. A piece of paper. Piece of paper typically have 10 to 15% water in it. And if you were to measure an absorption of them for that piece of paper, you would see an isosome which we'll see later on. That is a classic sigmoidal shaped indeed. It looks very much like a gas absorption isomer for bet. And part of the background here, which is a little bit confusing, I'm sorry to say, is a lot of the people in the food industry particularly talk about water monolayers forming on the surface of amorphous solids. I'm sorry to say these sorts of discussions are very misleading. So in terms of answering some of These questions, what we'd really like to know is where is the water? We'd like to know what defines the shape of these isotherms. Do these isotherms that look like BET ones, is this just coincidence? And are there water surface monolays forming on these particles that are biological materials? So we're going to answer those questions during the next half an hour. So we saw this before. This is our standard adsorption isotherm. I've just changed the axis, the units. So rather than humidity over P over P0, which is another way of writing humidity, volume adsorbed or mass absorbed depending on your measurement. But here is the adsorption isotherm. Adsorption here desorption been in the middle is called hysteresis. And we're very interested in knowing, we're very interested in knowing where are the water molecules? Could they be here? Or could they even be forming a hydrate? We'll discuss that during our presentation today. But as a general observation, unless you have very porous materials, the water molecules will typically only be in two or three places. They'll be on the surface, they'll be in the bulk. Maybe some multi layer absorption, occasionally maybe some pores. But for a lot of materials, actually this is where they live, in the bulk, they dissolve into amorphous domains. If we look in the literature, we'll find there are a lot of theories out there for modeling adsorption isotherms. The classic BET one for nitrogen, the fewer newer ones. Some of these models have been developed by protein formulation people and have come from the food scientists. The food people are concerned about water in food because it affects the stability of your food products and that's clearly important. So some of these models are fairly empirical. And more recently we've had some developments in more theoretical models. But part of these models are to try and help us understand at a fundamental level where are the water molecules? What are they actually doing? I think one other thing which is sort of confused literature a little bit, comes from some work published a long time ago by Ernest Pauling. He wrote a paper talking about the adsorption by water onto proteins. And clearly yes, water do absorb on the surface of proteins in different application spaces. But it sort of slightly confused in my view the literature around where the water molecules really are. And I think that sort of complicated the adsorption science a little bit. So in terms of the experiments, it's a DBS product. Again, we can typically do up to 10 gram samples. Sorry, we typically do up to 10 grams of samples. But reality is most samples are about 10 milligrams high resolution. We can take video pictures, we can take ramen, we can do all sorts of things. We want temperature range from 5 to 85. And if we need to, we can heat the sample up to very high temperature. Should there be a need to do that? Just to remind you a little bit about the method, comparing it with tga, which is probably the closest technique you might already be aware about. So here's a comparison. Is TGA gravimetric? A few milligrams, we're changing the temperature, normally a ramp. It's really around the thermal perturbation runtime, a few hours. And that's a temperature range. The DVS, it's gravimetric, about same sample size, 10 milligrams. We normally have humidity as a variable. That is they're done isothermally. We can step or ramp the humidity. Looking at the moisture partitioning, how much water dissolves in the material. Typically 24 hours to an experiment, sometimes a bit longer, typically 0 to 98%. Just to help you get a comparison between those two different related techniques. Both use mass change to understand material behavior. So I thought it might be worth having a little more detail to understand how these technologies work. So here you can see that the sample goes onto a small pan. It's measured on a balance. There's a reference pan. It's normally a blank reference, so it's not that important. We're going to flow our humidified air. So we take a dry and a wet gas stream, mix it together, pass that through the sample and out again. And some experiments we might even use a second vapor phase. So we can present two vapor phases if we want to for certain sorts of experiments. There's a preheater here if we wish to heat the sample up to 200, 300 degrees. And we generally purge the balance head with nitrogen to keep it just not affected by the actual humidity of the experiment. And the entire system is kept in a temperature controlled incubator. So things are isothermal, which gives us best performance, best stability. And it's worked out very well over the years as a way of maintaining high quality test environment. So I just thought a little cartoon to help you remind you what goes on here. The water molecules go in. They are attracted to the sample because water molecules like to interact with materials, including solids. The mass changes, we measure the mass change as a function of time as we step up the humidity. The humidity results in a mass change, which is shown in red. We go up and down and we get an adsorption and a desorption isotherm. So that's the basic methodology of this experiment. We flow a humidified K gas over a small sample and that allows us to actually get the mass change versus time, which then is computed into very simply into an absorption isotherm. It's quite a simple process really. So here is behavior. Reichstag, we saw it before, I just remind you. So 0%, 10, 20, all the way up to 90% humidity. And the red data shows the mass change. So to start the experiment, the mass decreases by about 7%. We're removing residual moisture from the rice powder and eventually goes up to be about 22%. Once we've got that, we now compute the adsorption isotherm. Here's the humidity on the X axis, change in mass in the Y axis, it picked up about 22%. So that was a standard uptake for that material. So that's a standard isotherm for that material. So we're going to do here now is look at some different materials. Now these are three different cellulose based materials. So we've got a micro crystalline cellulose, an alpha cellulose and a cotton material. You can see here, this is the moisture content versus timing already obvious. The kinetics and the uptake levels are different. And if we look more carefully, we can see here that these isotherms are similar, but are different. And one of the things we're seeing in this isotherm, and we'll talk about it in a moment, is we're seeing the water molecules dissolve inside the amorphous domains of the cellulose. Cellulose materials are partially amorphous. Those amorphous materials have a very strong affinity to water. The water goes into the amorphous domains, not the crystalline domains, the amorphous domain. And what you can see here, those of these materials are very similar. You can see, you know, some subtle but significant differences in the kinetics and the amount of water going in. For example, you see over here that the cotton linter takes up about 14%, whereas the alpha cellulose has taken up about 21%. So there's significant differences in the amount of water dissolved in those materials. So this takes us back to some of the core scientific questions, which is rearound the glass transition temperature. This is the fundamental property we're interested in. It's a property of amorphous solids. And if you've got a material that's partially amorphous, from polymers to biomaterials, including hair, Then the amorphous domains will exhibit atg. And to remind you, if we have a crystalline materials, it's a well defined molecular structure with long range molecular order. The amorphous material, not well defined molecular structure and only has some short range order. So they're different molecular structures. And if we have a tg, only the amorphous part will see a tg. Amorphous materials exhibit very different properties above and below the tg. Above the tg they're soft, rubbery material, whereas below the TG they're glassy solids, which with much lower. So they're rubbery mechanically above and glassy solids below. Typically, if we want to determine the tg, and maybe you measure that in the moment using dsc, you typically heat up material. You look for a change in thermal properties in a dsc, for example. However, it turns out that if a material absorbs something like water, that will also cause the TG to be apparent. Absorption of water will result in the TG decreasing. So for example, if you have something like polystyrene, which has TG of 100 degrees, if you expose that to toluene vapor, that will decrease the TG by 20 or 30 degrees. So the solvent acts as a plasticizing agent that allows you to see changes in tg. This cartoon here tells you a bit more about the different things that happen between amorphous and non amorphous. So we can see that water vapor is a strong plasticizer for hydrophilic materials. The glass transition behavior depends on the temperature and the humidity and of course the molecular structure of the solid. So if it's amorphous, glassy solid, you get mainly surface adsorption of molecules. Fast kinetics, but very low uptake. If it's rubbery, then of course you have faster, larger uptake and slow kinetics because the materials are diffusing all the way into the bulk. If it's crystalline, there's no bulk absorption, it's just surface absorption. Very low uptake levels. Just to remind you what happens above and below the tg. And the list down here is probably the most important. Properties vary enormously above and below the tg. And sometimes the best way of seeing a TG is mechanical properties. So for example, above and below the tg, the heat capacity might only vary by a factor of 2. So sometimes DSC is quite tricky. Spot TGS. Mechanical techniques can be a bit more difficult to employ, but are very, very sensitive. It turns out also that you can actually measure TGs using DVs, particularly for materials that are hygroscopic anamorphous. And we'll see that coming up. So I'm going to take a moment or two to talk you through. Hang on, slide there. I take a moment or two to talk you through, talk you through this slide here. So anybody who's used DSC will be, will be quite familiar with the process of ramping the temperature and looking for a transition event. We're going to describe a similar thing here using the dynamic vapor sorption method. So we're going to do is we're going to ramp the humidity, okay. And when we ramp the humidity, we're going to be looking to see if there are any changes in the slope of the adsorption isother, which is not unlike what you do in tdsc. DSC look for a change in the heat, the slope of the heat capacity as a function temperature. And we do a similar thing here with dvs. So let me describe what's going on here. The first thing to realize is that the humidity for this sample is being ramped up at a fixed rate, okay? Up to 90% humidity. The red line is the sample mass. So at the start there's a little bit of water which is probably surface absorption. It turns over, it becomes a fairly straight line for a certain ramp of humidity. But then a particular humidity, it changes and it goes to a different slope. It's not unlike what you see in DSC for measuring tg. So actually, whereas in DSC you think about the temperature, in our experiments here we think about the temperature and the humidity. This experiment would have been done at 25 degrees, but there's a critical humidity which we can pick off the graph here at which we go from being a glassy to a rubbery solid. So we talk about sometimes the rhg, not a tg, an rhg. And these ramping experiments for a fixed temperature can allow us to determine the RHG for all sorts of amorphous materials. And this just gives you an example. And this is for 100% amorphous lactose. It's a 6% per hour ramp experiment at 25 degrees. And lactose, as we saw before, can be quite complicated. So again, the blue line is the ramp and the red line is the actual mass change. So it goes up here, glassy solid here, rubbery solid here. And of course we saw in those pictures a bit early on today that actually it will start to recrystallize and we can see some other events. So in these ramp experiments you can see two or three different events. But the one we're particularly focused on today is the tg, or rather the rhg, the glassy humidity. Just to remind you this. So what we're looking at is the temperature, the humidity at which we go from here to here, what to critical humidity. We go from being a glassy powder to a rubbery material. And that's what we can determine using the DBS method. In some of our earlier work we would do these ramp experiments at different ramp rates. So here is an estimation of the critical glassy humidity for lactose. And in the early methods we did different ramp rates. 2, 4, 6, 8. And we can see here that if we want to extrapolate, if we want extrapolate to a very, an infinitely slow ramping rate, we get about 30% as the critical humidity. That is above 30% humidity, amorphous lactose will become a soft rubbery solid, whereas below 30% humidity it will maintain a glassy powder. And that's important for us to know if you want to engineer a formulation where we've got, let's say a spray dried lactose. So this allows us to determine the critical humidity where we're stable or unstable with respect to a glass transition temperature. This is for a lactose. So we've done other materials. So here's a complex material. This is hair. The ramp is shown here in blue and you can see the mass change. You can see in fluid point in here, which we're going to construct. Again, same idea. We can construct this point here and determine the RHG, which is about 70% for hair. This is the amorphous domains in human hair. Same data here, different ramp rates or higher for human hair. And we're able to extrapolate the conditions to estimate what the humidity is, where the material transitions from being glassy to being rubbery, and again for different temperatures as well, which is relevant. This is the data published here and references some other data published literature. Let me be clear. Measuring the TG of some materials is very tricky, difficult. Hair is particularly difficult because it's got a lot of crystalline domains and the amorphous domains are heavily cross linked. So the relaxation processes are constrained. It's a tricky measurement to do, but the data we got was quite encouraging. So we just moved on to do really just a series of different cottons and other cellulose materials. It's the same approach. We ramp up the humidity and we fit the data and we determine the critical humidity. And for cotton it was 76% twice. Okay, fine. And we've got some Nanocellulose. So this is materials being ground up to provide very small particles that are more crystalline. What you can see is that the RHD is much lower. So yes, there are small particles, but the amorphous particles that are also present are much more susceptible to moisture uptake. In other words, the amorphous domains attached to nanocellulose are much more easy to plasticize. So that's an important observation. Here's another standard fiber material, viscose. Same measurement and calculated a 72% RHG again for 25 degrees Celsius for this material. Here's a material that you might use in your lab now. Micro crystalline cellulose, probably one of the most popular excipients in use. Same approach. You can see there. We can interrogate that data set. Sorry, we can interrogate that data and determine a very clear humidity for the TG at 25 degrees of 65% h, below 65% glassy, above 65% humidity. Rubbery material. And this is whatman filter paper which you use in the lab. It has a RH is quite High, about 76% at room temperature. And we took some of this material and we ball milled it. Typically when you do ball milling, you change particle size and you tend to change amorphicity. Okay. And you can see here, when we bore milled the material, a fairly substantial decrease by about 20% the RHG. So in other words, not only did we change the surface the particle size, but we changed the amorphousity. So in other words, the ball milling induces a lot of disorder in the material. That disorder is reflected in the RHG for that particular material and substantially decreased. Here we changed effectively the molecular structure of the amorphous domains in this bormeeld power. So what can we say in conclusion? Well, for partially amorphous solids, the water isotherm is basically reflect the TG in the material. So the shape of the isotherm is very much dependent on the amorphicity and particularly the tg. So the amorphous content is one thing, but also the TG of the amorphous material is also reflected in the adsorption isotope. We can determine the glass transition humidity from water salt at different temperatures using the dynamic vaporsorption method. And it is possible to determine these humidity induced glass glass conditions as a function temperature relatively easily. And we can see that the RHG varies for different silage materials and also varies on processing history. So if we do things to process our material, we can change the amorphous structures and that is reflected in the isotherms and also reflected in our ramp rate experiments. So I thank you for your time and attention and I think.

Thank you very much, Dara. We'll just give it a couple of minutes then to just wait for the questions to come in.

Thank you. Sure. Thank you.

Daryl, there's one question that might be interesting for you. Maybe you want to spend a couple of minutes in here, which is.

Good question. I got a good question. Hey. Yeah, yeah.

So it says in 100% amorphous lactose, why is the mass increasing and then decreasing as you continue to increase the relative humidity?

Yep, that's a. That's an excellent question. Thank you for asking that question. So actually, what's happening towards the end? And maybe if I would get the slide up, it might help a little bit here. Yes, right. So. Oh, okay. Let me just do that. So you're asking about this part of the curve here, which is interesting. And what's happening here is the lactose is forming a monohydrate. Okay. So what's happening now is we've got a lactose monohydrate forming here that's ejecting the water out of the sample. So this is loss of water due to monohydrate formation. So basically crystallization is occurring when the crystals form. There is no longer amorphous structures and the water that's absorbed is now being ejected. And the only water left is the water of hydration. That's a good question. Thank you.

Thank you.

Sorry, there's a question.

Yeah, I'll just read it out loud. Sometimes the substance cycle ends with negative values. So what is the reason for that?

Yeah, I think in some cases you may have water that's trapped inside your structure. And when you go to high humidities, that's released. And then when you go back to low humidities, you end up having less mass than you started with. So sometimes water is trapped in amorphous structures. You go up into the rubbery state, some of that trap water is released. Then when you go down, it's not there anymore. So you'll sometimes get negative masses. So that's one reason. So can we determine transition temperature? Ready for lipids? Yep. Yep. If you're. If your lipids have a tg, for example, we can do that. I think the only question to think a little bit about is what the probe molecule is. If the material is hydrophilic, then I'll use water molecule. But if the materials hydrophobic, I might change the molecule to measure the tg. I Might use octane, for example. We should be able to do that. Well, in terms of information, in terms of lipids, we can do an isotherm that tells us the solubility of organic solvents or water molecules. Just get the questions one at a time here. There's a few coming in there. Just get those up here a bit. So you get the solubility of your molecules into your lipids and that tells you a bit about the hydro diversity or hydrophilicity of the materials. You get that from dbs, the kinetics you might also be able to measure, if you do the experiments carefully, might be to measure diffusion constants, so molecules into lipid layers, which might also be interesting to you. So there's a few things you can do potentially. Next question, hysteresis. So hysteresis broadly has two or three main causes. Let me talk you through them. If it's surface adsorption, then the only reasons you will get hysteresis is if there are pores present. So in other words, if in the process of adsorption, molecules are dissolved, doing poor filling, there will be some hysteresis associated with the desorption from pores. So the wetting of a pore desorption from a pore will cause hysteresis. If you have a hydrate reaction that will cause hysteresis. If it's amorphous solid, you'll get some hysteresis due to the differences in the absorption processes. So if you've got bulk absorption, that will also result in hysteresis as well. So hydrate reaction, pore filling or water going into an amorphous domain, any of those three will give you hysteresis. So just following on that question, you asked the question, does hysteresis occur in all materials? Answer is no. If you've got a, you know, a fine powder, if you've got a powder and you've only got surface adsorption, you'll probably get no hysteresis at all. The big so question, the question we'll go to tga. What does DVS bring you? Well, the big difference is that DVS tells you humidity's impact on tg. The TGA only tells you the TGA only gives you the TG for a dry material. And the truth is materials are very rarely dry. They've always got some moisture in them. So the DVS tells you what the TG is at different temperatures and different communities. If you're only interested in dry powders, then TGA is Fine. But actually in the real world, powders are very rarely dry. So measuring the TG at different humidities and different temperatures is practically sensible.

There's another question that came in. It says, in your experience, what other techniques complement well with dbs?

Coefficient titration is sometimes useful. Complementary information to have DSC can be helpful as well. So the question here can community profiles give us information about specific surface area? If we do adsorption isotherms of the correct molecules, we can tell you what the specific surface area is. But I think what we have to be careful about. So for example, we saw some cottons before. Now, if I want to do the surface area of a cotton, I have to choose a molecule that does not dissolve in cotton. So I can't do water. Water will tell me about the tg, not the surface area. But if I have a molecule, say ethanol, that doesn't dissolve in cotton, then I can use the ethanol data to give me the surface area of my materials. So you. So if you're doing surface area determination using dvs, for example, you've got to use a molecule that does not dissolve. So if it's a crystalline solid, you can use octane, no trouble. So if you choose the right molecule, you can use DVS to give you the surface area. And same sort of thing for IGC as well. They'll both give you surface area and low surface coverages and low surface areas as well. Crystalline samples lost during water absorption. So this is about something losing some sample mass, by the looks, crystalline samples key. I think I'd do a few. I think I would do a couple of blank experiments just to check what's going on here. The report here is someone's running crystalline samples on a dvs. They seem to be losing some sample. Now, what's that mean? Well, there are two or three possibilities. One, if you're unlucky, maybe some of the samples being blown away, not common, but have seen it happen occasionally, particularly for a material that's very fluffy, electrostatically charged. Two, if the sample is electrostatically charged, then mass readings can be a bit tricky and you might have misleading data. You could have some trapped moisture in the material that's being released which will give you an apparent negative mass loss. Maybe that person might want to email us offline to get a bit more detail about what they've observed. We can give them some advice. Yep, good question here about differentiation between adsorption and absorption. So there are three or four things that help us answer the question. Right. The first thing is the kinetics, sometimes in DVs, the most important thing you measure is the mass as a function of time. I know the isotherm that everyone wants to show you, but the actual kinetics are sometimes most important. So if it's surface absorption, it will be fast, like 10 minutes. If it's bulk absorption, it will be slow, many hours. That's the quickest way of spotting if it's surface or bulk is the kinetics. The other thing which can help you is the amount of uptake. If it's surface adsorption for pharmaceutical material, for example, you're probably looking at more than 1 or 2% maximum uptake. If it's absorption, you could pick up 30, 40, 50%. So the amount water you take up and the speed will tell you if it's bulk or surface. So hysteresis can relate to water in the bulk or it can be water in a pore. Okay. And although if you're talking about pharmaceutical materials, I don't think pores are very common. So I think for most pharmaceutical materials that are amorphous, for example, it's probably bulk sorption, I think. So if you've got, if you've got a amorphous or partially amorphous solid, hysteresis almost certainly going to be due to bulk watersorption. If you've got a crystalline material and you've got hysteresis, maybe there's some pores in it. Possibly.

Okay, I think we're pretty much done with the question. There's one more pass and that raises hands. So maybe they like to ask the question themselves. I can see that you have your hand raised. Would you like to ask the question or would you like to just put.

It on the chat? Y is equal to MX +C.

This is the slope. Slope means kitna angle, inclined head. Slope is always tan theta sin theta by cos theta. And this is the point. This is the Y intercept.

Now.

Where the mass change is representing 100.

Can someone help you with that question? Around the slope of the lines, I.

Believe that person might have unintentionally unmute themselves because they were having a conversation.

Oh, right. Okay.

But if the question does come in, then please send us the questions that. Okay, we are a little bit early, which is great. So what we'll do is we'll take another 10 minutes break from here from now on and we'll be back the next session which will be called Para Rheology for the characterization of pattern.

Okay, ten minutes.

Hello, Val, can you hear me?

Okay, loud and clear.

Yeah, we are pretty good here. So whenever you're ready, we can start.

Great. Okay, let's go. Yeah, that's good. I'll turn that. Oh, actually turn that off. Turn that off. Is that clear? Yes. Great. Thank you. So I'm now going to talk about a different area of research which is not around DVIs or IGC. This is not a technique that was developed by Surface Management Systems. I'm going to talk about a technology which has been developed by Freeman Technology for characterizing powder flow. And we'll talk about the FT4 rheometer in this presentation. Right, okay, that's good. So this is a part of a research collaboration between Imperial College. Roquet is the manufacturer of excipients in France and BASF in Germany, who are one of the world's largest chemical manufacturing companies. And it was really around powder caking. The person who did all the work is Alexandra, a PhD student of mine at Imperial College. And she's been working for the last three years on developing a new method of determining the conditions under which powders caked, with particular emphasis on humidity induced caking. So someone asked a question earlier today. Why does caking occur? There are three or four mechanisms by which caking occurs, but I'll give you some of them just to help. If a crystalline solid absorbs adsorbs a lot of water, then actually you can dissolve the surface of the particle in question. And then later on when that particle dries out, you get a solid crystalline bridge form amongst a powder bed. That is one reason you can get caking. If you have amorphous solids that absorb a lot of water, then those materials become very soft and they become so soft they become sticky. That is another primary mechanism for determining caking, which can cause caking. And of course you can get caking in some quite large scales, particularly if you're an industrial chemical manufacturer. So here are the factors that are at play if you're an industrial company and you're moving particulate materials in bags around the world. The temperature, humidity, the load on the bag.

We lost you there for about 10, 15 seconds.

Yeah, that was okay, me having a small break. So, packaging material, the load on the bag, temperature, time, humidity, all affect the probability that a material will cake in an undesirable way at scale. So the traditional way of measuring this is to take a packet containing a kilogram of powder or thereabouts, putting a weight on it, compressing it, then doing some indentation test to see whether or not a cake had been produced. These tests are typically set up on a three month basis. So the mission in this project was how could we determine the propensity of material to cake and not take three months to give you the answer? And ideally we want to give the answer that day. That's the overall mission of this project. In terms of the molecular properties, the properties of the powder that are critical for caking then crystallinity, solubility, lasicity, plasticity, desorption, surface roughness, particle size are all factors that are going to contribute to the probability of caking occurring. But the traditional approach used by many manufacturers including basf, is a very slow and meticulous process which involves the samples being put in ovens and a lot of time being spent waiting to find out if the materials formed a hard cake or not, and then measuring the hardness of the cake using an indentation approach. So in this project we decided to adopt a very different strategy and we're going to talk about the FT4 power. This approach is a standard method used by many companies days to understand the amount of energy requires to get powder to flow. The technique is quite simple at some levels. It involves an impeller being driven through the powder bed and the amount of energy required to push that impeller through the bed is determined. So basically you look, it's not unlike rheology of a liquid, but it's rheology of a powder. You're working out how much energy do you require to put into the blade to displace the powder to allow the blade to advance in the system. Now there are a few important subtleties to how this measurement's done, which I need to explain to you to help you understand the project here. When the blade goes into the powder bed, there are three distinct things it does. Most commonly people talk about the downward cycle. So it's pushing through the powder to the bottom and as it gets to the bottom it slows down because it doesn't want to crush the pattern. So there's a downward cycle followed by an auto tune which is just simply a slowing down. And finally it comes up again. So it's down, slow down, up again. And this is in when you do this, every time you're doing this measurement, the machine's reporting the amount of energy required for the blade to push through the powder during this experiment. And this will be one of the techniques that have been the PCL laboratory, which is one of the reasons for today's presentation on this topic. So let's press on. So here's some data you would get from an FT4R and on the Y axis, this there's the energy and this is the height as the blade pushes through the bed. When it gets to the bottom, there's nothing to resist it and there's resistance at the top. So the black is a downward cycle. So we're compressing the powder in the downward cycle and the UC is the upward cycle. We are pulling up through the powder. And traditional wisdom when you're doing powderymetry is to focus on the downward cycle data. But we'll see in today's study that actually the data was most informative is actually the upward cycle data because we're interested in caking. Instant understanding about caking. And it turns out, as we'll see in the presentation today, that pulling the rheometer blade up through the samples much more sensitive to aggregates and agglomerates forming. So the novelty in the project that I report today is that if you have an FT4 in your lab, and some of you may have one, they have no humidity control. And it's slightly curious really, because considering the humidity is the single biggest factor controlling the formation of caking, having a powerometer that doesn't have humidity seems slightly strange. So in our project, we modified the machine to incorporate a humidity control system. And that's what's shown on the left. Also, the way we did this was slightly novel. There's a car turn up here, and it shows you the way you can condition a sample for a rheometry experiment. So mark four here, model number four is what we did. And we continuously pump in Mach 4. We continuously pump humidified air through the sample base, humidify the bed during the actual experiment. That's a novel approach. Here we see more detail how this is achieved. We take dry air, we humidify the stream, and we pass it through the base of the unit here and it comes up that way. So essentially we're able to humidify the sample very easily and very effectively. If you don't do this, you really don't know the moisture content of your sample. It's entirely random what you get. But for a process in which the humidity is the critical variable, having this sort of control is actually essential. So let me describe the experiment we're doing here. It has a few bits to it, so just take a moment to think through this. So the standard method FD4 Riometer has is it typically when you load the sample in the bed, it goes does a stabilization run. So it goes up and down 20 times to essentially stabilize and unify the pad A bed. That's the first part of the experiment. And all FD4 measurements do that. The experiment I'm going to describe to you takes about two and a half hours to run. But and it's an automated experiment and to compare with the three months it takes to do it manually in the lab, then once you've done this preparation, which we call stabilization, then we introduce humidity into the powder bed. Now we've chosen three different communities, but you can choose any communities you like. We use 33, 58 and 84, but it doesn't matter which humidity, those are the ones we chose for our test. So once we've aerated the sample for this test phase within rest the sample at 33%. We aerate the sample, but we do not put the mechanical blade through the sample. And what we're looking to see is have solid, solid bridges formed, have cakes formed. And then after we've rested for 20 minutes, we then do the test again at 33% to probe whether or not solid bridges agglomerate caking has occurred. So at the start we're doing this experiment at 33%, then we humidify at 33, 50, 84, then we go back to 33, then back to 33. So still bit to think about there. Now just to compare with the other technique, we put the powder into a container, we store it at 33, 58 or 84%. And then we do an indentation experiment after a few days or weeks and there's a small indentation. So these are petri dishes, small indentation sample. That's another way of confirming. So what we're looking to do is to show that our new test gives us the same insight as this mock up of the old mechanical test, but of course to do it much faster. So the first thing we're going to use is used as a reference material, sodium chloride, salt. It's a well known material that will cake. But of course the critical caking humidity is 75% humidity. So in our experiments to accelerate the process slightly, we did the caking at 80, we induced humidity at 84%. And just to show you what sodium chlorate does, this is some watersorption data. So up to 70% it picks up no water. And then as we go above there, it would have picked up there and got very fast. And once you get above 75%, the sample, the powder, the granules, picks up water very fast and forms cake. And here are some pictures showing you the sort of bridges, these solid solar bridges that can form Once you've gone to 75% humidity and gone down again. Okay, so basically deliquescens is one mechanism by which caking will form via solid state bridges. Right, now let's talk about the actual experiments. So we've got two different experiments here, but I want to today focus on just one of those. I want you to focus on this experiment here on the right says upward cycle. Okay? So we don't report the first 20 cycles because that's just stabilization, remember? But we report on the next set, okay, and in the next set here we are putting through and I think we go back here, we'll see it. We're humidifying for 30 tests. Okay, so this is an experiment where we have humidified air passing through the sample and we're measuring here the flow energy, how much energy does it take for the blade to go through? And in this case the blade is coming up. So all we can see is at 33% and 58%. The energy does not change for my 50 cycles at all, but at 84% when I'm above the deliquescence point. So in other words, right now the sodium chloride particles are starting to stick together. You can automatically see a substantial change in the flow energy. It's gone up by a factor of well over 2. So in other words, this material that deliquesces, we can see this very, very quickly. Now when we turn a humidity off again, that goes away. In other words, we're able to remove the humidity before substantial solid bridge formation occurs. So we're seeing some deliquescence in here that's affecting the flow behavior. We probably haven't got solid bridges forming quite yet. We haven't given enough time for that. So on the right hand side here we have the established method or similar to run. So what you can see is 33, 58 and 84%. So what we can see in the traditional test that after one day we don't see anything more than baseline. It takes us three days or seven days to get enough caking at 84% for us to see a difference in the penetration hardness of the material. So using a traditional approach, it took us a week, let's say, and our new approach took us two hours. So in that sense it's a successful way of us defining the caking potentially this material. So let's press on. We can also see during these processes one thing you can see in the FT4, as you will see some density changes. This is the energy and that's the bed height. And we zoom in, we can See here that there's been some small changes in the overall density at 84% compared to 33 and 58. In other words, the powder density has become hot. The powder has gone down as the material starts to deliquest and similar over here. So we can see some small changes in powder bed density during our actual experiments as well, which is associated with the caking behavior. So let's move on to a different material. We're now going to look at a mannitol and we've got two different mannitols. What's interesting about these Matanols, they're commercial grades manufactured by Rokette, is that there's one material which is a spray dried material, it takes up three times more water than the 160 grade, which is a crystalline material. You can see that 0.38% uptake versus 0.1% uptake. So we might have thought possibly that the SD material might cake more because it takes up more water. More water, more caking. Well, that might be a good place to think about. So does it. So this is the caking strength based on our penetration test. Okay. And we can see here that actually the after 14 days the caking strength was only 0.18 neutron newton for the SD grade which had more water. But this grade here that had less water had even more caking. It has 0.71 Newton. So the cakes are much stronger for this C grade material, which is slightly surprising. So the question is, can we see the same sort of behavior in our FT4 data? The FT4 data give us the same sort of behavior. So on the left you can see the actual spray dried material. It's a downward cycle to start with. The downward cycles are all the same, but this material is definitely much higher. This is the crystalline material that has stronger indentation caking. That's the downward cycle and the upward cycle even clearer. So we can see here in the upward, in the upward cycle for the material which is sd, there's some sign of a little bit of caking here. Sorry about that. So even the SD material shows a small amount of caking propensity. But very clearly the C material, the 160C shows way more caking behavior, which is quite consistent with the indentation experiments. So though we'd like to say that moisture content correlates to caking propensity, isn't true for this material. But we can see that the FT4 test is able to spot caking very quickly in a couple hours, making it an attractive test. So this is A slightly more complex data set and we won't go through all of it. I will just focus on this one down the bottom here. Okay. Now in the FD4 experiments you are aerating the powder bed. You're flowing air through the bed and therefore within that there's a risk that you'll aerate the bed, it'll become fluidized. So in this experiment we change the velocity of air through the bed just to see where that affected the results. And this is for sodium chloride. And what you can see here for sodium chloride there is a small change as you change the velocity of gas at 33 and also 50% humidity. But for 80% where there is caking, the trend is very unambiguous. So in other words, even the effective velocity of the aeration media is not very significant. It is there. It's real, but it's not significant. So if you use the same aeration velocity for your experiments, the data is going to be very comparable. And again, very good sensitivity for this upward cycle, which is our preferred method. And these just give some idea of the errors. It's typically under about 7 or 8%. Seven or 8%, which is quite good for these pathology experiments. They tend to have big, bigger errors than you might expect. But these were not bad, quite credible. I think if you look at the machine after the experiment is sometimes interesting. This is the machine at 10% and what you can see here is a little bit of powder stuck to the actual impeller. That's probably a bit of electrostatics. At low humidities things get charged and you see a little bit of electrostatics. Even though you don't want to, you do. At 50% there's nothing at all attached to the sample impeller. However, after 80% there's a lot of cake material attached to the impeller. Just giving you further evidence of the caking behavior that you see. The high humidity for sodium chloride, this is after 40 cycle experiment. So we'll start to move into our sort of conclusion phase here a little bit now. We now have a new method by which we'll be able to introduce humidity into FT4 experiments. It's a modular approach which is quite easy to reproduce and gives very reliable data. We've developed some new FD4 methods particularly focused on the upward cycle. We can interpret the downward and upward cycle flow energies quite well and we're able to detect deliquescence very easily as well as other caking amidst which not deliquescence as in the case of mannitol. And therefore this Is a potential research tool for looking at powder caking, wet granulation powder, electrostatics and the powder drying. I haven't talked about it today, but we also have some evidence to show that when you put humidity through an FT4, you get more reliable data, better quality data. And the effect of electrostatics, which sometimes can complicate the interpretation of FT4 data is not the case with humidity control because essentially neutralizes surface charges, which is sometimes rather helpful. So we've got time resolved measurement. We can vary the humidity less time, less product. And we can reduce substantially the number of conventional storage tests by the targeted planning of this approach. I said the young lady who did all the work is hidden behind that face mask there, which is Alexander. She led all the students doing the work and we many collaborators in BASF and other companies. I thank you for your time and attention and happy to have any questions you might have.

Thank you for that presentation. So we'll just open the floor for questions now.

I have questions.

Sure.

So Mary, you can. We've got a question here in the room. So I'll ask him to directly direct. It's about slide number eight in particular.

Yes. Okay, hang on, let me get that up. Yep. Can you see slide 8 now?

Second chart. Any particular reason why the energy has reduced in case of 58% RH as compared to 33% RH?

Sure. Yeah. So if we look at the data there, it's decreased. Our general view of that is probably some charge neutralization. We think there's a bit of charge in there, but electrostatics. And at 50%, we're better at neutralizing the humidity. The surface charge is better neutralized at that high, that slightly high of humidity. So I think that drop is due to the humidity neutralizing some surface charge on the particles because it is sodium chloride. So it is an insulator and it does get. It can get charged. Okay, thank you.

A question that just came on the chat, so I'll just read that out loud for you. It says how can you reduce the caking effect in powders?

Sure. Well, the first thing you do is understand what causes it. Okay. If you don't understand what causes it, you can't get rid of it. So my first suggestion was you need to understand what is causing it and or which component is causing it. So if I was doing a study, I would be looking at the powder rheology of my component in my formulation to identify which component is causing the caking. And that's where I'd start.

You mentioned that the pattern rheometer has an atom for humidity. Will that be available in pcl?

Yes. Yep, it will be available at pcl. PCL would be any place in the world that has a. Beside the Imperial College.

Professor Darrell I just have a follow up question to what has been asked in the chat box. So in case of any powder which is having electrostatic charge during, you know, maybe powder processing or during the particle size reduction, some static charge is generated on that and that is causing agglomeration kind of tendency to a powder in bulk. So could you share some industrial level solutions like how can we dissipate that electrostatic charge during the packing or during the powder processing?

I can't say I have a lot of industrial experience with this problem, but I can certainly give you a couple of thoughts. I think there's an opportunity. I think one of the challenges is can you use. I mean humidity can be your friend, it can be your enemy. Too much humidity can cause problems, but a little bit might help. So sometimes you might be able to induce some humidity into the system to help neutralize that. Okay, that might be. That might help a little bit. But we haven't had a lot of experience of solving the problem mainly and identifying and quantifying the formulations that are at risk. But I guess it also goes onto your packaging solution. I mean, are you keeping moisture out of your final package product? This moisture tends to be the more common thing that you can control. That is a factor at play. Okay.

Okay, Darl, we don't have any more questions at the moment, but we'll continue monitoring the time.

Sure, no problems.

If it's okay with you, I think we can proceed to the last session of today.

Sure.

Measuring the specific surface area of freeze dry biologics using invert gastronomatography.

No problem. Let me just get that presentation up. Bear with me a second. The last presentation today is another project from Arnold's PhD who worked on biological material, particularly freeze dried materials. When it comes really to a topic I introduced a bit earlier on today is that there are lots of pharmaceutical and biopharmaceutical materials that you really want to know what their surface area is. And although traditionally you might think about doing nitrogen absorption, the truth is for probably a majority of biopharmaceutical and certainly a lot of pharmaceuticals, the nitrogen BET method is just not accurate or sensitive enough to do a very good job. And freeze dried materials are a particular example. As we'll see today, they have low surface areas and doing it with nitrogen is quite challenging. Right, let's get going here. So, okay, so today's presentation, which is part of a general project we were doing on freeze dry products, is really looking at the surface area determinations on freeze, freeze dried products. You saw earlier on, we discussed their moisture stability, which is an important issue for their performance. But the final presentation is really talking about looking at their surface area and how you measure it. We heard about water before, I don't need to tell you about that again, just remind you about freeze drying if you didn't join us before. We're going to take our active biopharmaceutical, which is we dissolve it in solution with excipient and other additives. We put that in a freezer, cool it down to minus 50, apply a vacuum and we sublime off the water. We do that, we end up forming a dry porous cake which consists of predominantly our active material in a solid state form. It is going to be in an amorphous form, but it is in a solid state form and that's quite useful because it has better stability as a solid than in solution as a liquid. So liquid solutions, not very stable amorphous solids, more stable crystalline solids, most stable. So these freeze dry materials have some intermediate stability which is good for these sorts of biopharmaceutical materials. So if we've got a, if we freeze dry solution containing active species, we almost always get a porous keck which is amorphous. And the surface area is quite important because one of the things we need to be aware of is that this material in a freeze dried form here we're going to add liquid to that and reconstitute that for the final product at the hospital for the patient. And the surface area is a factor that defines how fast it will dissolve and how quickly it will dissolve. You do not want to have to take half an hour to redissolve your material. Having a high surface area allows you to reconstitute the formulation much more, much more quickly. And that's what you want. Now we've heard about inverse gastroenterology a bit early on. I don't tend to spend too much time on that. But what we're going to do though in today's experiments is we're going to pack our freeze dried cake into an IGC column, we're going to chop some up and pack into a column and we're going to pass helium through here. In the helium we'll add small amounts of octane for example, or ethanol and they will be used to determine an absorption isotherm which would in this case be used to determine the BET surface area. This is a measurement that we've done at room temperature and room pressure. Unlike nitrogen absorption which is done under vacuum at 77 Kelvin, it's much more a real world determination. So as we saw before and we should see again, we started by doing our validation work using CRM170. It's a standard reference material. So we do bet adsorption using octane. I should say. Yeah, octane is the adsorption isotherm. And from that we determine the BET surface area and it comes out to be about 1.01. So which is very close to the reference result from the standard of 1.05. So we're quite comfortable that our IGC method is working well and is working well for low surface area solids. And 1 meter square per gram would be a very common surface area for many pharmaceutical materials. And in this case it turns out to be relevant to our freeze dried materials as well. So here is, here's another reference material. This is CRM171. This is, has a slightly higher area. It has an area of 2.95 meters squared per gram. Using IGC 30 degrees, we've got 2.995 meters squared per gram. What we can do with IGC, we can't do it Nitrogen BT. We can also measure surface area at different humidities. So if we do the surface area at 50% humidity, it drops slightly. That's because the water molecules occupying some of the surface. That's an experiment we can do on igc, we can't do in nitrogen. Just remind you about the traditional technique. So nitrogen absorption is going to work very well for high surface area materials up to 1500 square meters per gram. But when we get under 10, the errors start becoming substantial. And what we can tell you is that the IGC method will take you down to about 0.1 and the error is about 10 times less. So that's, you know, quite important. If quality, if area or surface area is a metric that you're using for product release or an important research parameter, having a method that does accurate determinations is something to be noted. So why do we want to measure the surface area of these cakes? Here are some reasons we may want to do. Might correlate with the biological stability of the material. It could assist in predicting the reconstitution time for the cakes. It's a potential quality attribute which might track batch batch variation in final products. It directly relates to the average pore size distribution in the cake. And it could also predict variations in the mechanical forms of the cakes. And these cakes have to be shipped, they're not very mechanically strong, they break. And people don't like to buy freeze dried cakes that are broken. Some countries will automatically reject batches of product that have got cakes that are broken in transportation. So here is the experimental matrix for the materials we look at today. 5% sucrose, 5% trek lows, 5% IgG, an influenza antigen and two different influenza antigens with sucrose. The freezing temperatures, the rates and the conditions it took and these typically would take two days to three days to freeze dry the product. It's not a fast process, freeze drying, it's quite slow. So the method we would take about 100 to 60 milligrams of the sample very lightly crushed and packed into some Columns at a 4mil ID with the ends play with some glasswool to stop it falling out. We use helium as a carrier gas and methane for dead volume determinations. Experiments on a 10 cc at 0% humidity peak maximum method was used. And we looked at, looked to analyze the linear region of the BET isotherm. We got the monolayer capacity from the slope and intercept using the BET method. And I think we'll see that most of the studies were done with octane, but we'll see that in a moment. So this just gives you some idea on the repeatability. These are three of the samples here and they vary from about 2 to about 1 meter squared per gram. This is the same column repeated. Okay. So you put in the machine three times. Typically errors between 1 and 2% in terms that's a repeatability measurements. Okay. Which is of course only one measurement of relevance but certainly a useful measurement. And this is a similar thing done using volumetric notching. Okay. Again, the same sample tube was repeated three times and what you can see there, the errors are typically between about three and about 5%. So they're about two times higher, maybe three times higher, something like that. So here is just a general comparison of the methods of the measurement. Approaches nitrogen versus octane absorption using igc. The amount of sample is fairly similar. To be honest, the time to run the RSD errors maybe about a factor 2 lower for the IGC. The runtime is a little bit faster, you know, a lot faster depending on the preconditioning arrangements, but it's a little bit faster. The temperature is quite different because we're doing our temperatures at room temperature, not at 70 per kelvin. And the experiments are not done at the vacuum, they're all done at atmospheric pressure. So there's a few small differences there. So here we've got some data sets. Okay, this is three different sets of vials and first comment I make is that the vials don't have necessarily the same surface area. You'd like to think that if I've got a thousand vials, they all have the same surface area. Well, I'm sorry to say that isn't quite the case. They do vary a bit. So you can see here that sucrose which is the top line, see these are three different three sets of vials that give them the same batch. And you can see there's some quite substantial variations in the sucrose surface areas for what are normally the same material. Treklos a bit better behaved. The IgG varied a bit. The influenza had quite a high surface area, but again the variability was a bit lower. So one thing you can see here is that there is some definite non uniformity across the vials in a batch. This is data obtained using the igc. But we've got the error levels here. They do vary a little bit. This is the mean surface area and these are the errors here. But these are certainly better than you've achieved for nitrogen. This is also some similar data. Three different vials using nitrogen. The variations in the vials just depend literally on where the vials come from in the batch. And of course in an ideal world you might want to analyze all the vials in a batch to work out is there a pattern to the surface area depending where you have a vial in the batch physically inside the machine because the freeze dries will have a couple hundred samples in them. Maybe the batches near the front or the back or the side are different to the batches in the middle. So the vials might vary in terms of their physical and chemical properties during the freezing process and that might affect the surface area. Don't know. But you do see quite a bit of variation within a batch, which is interesting observation. I'm not sure that's well understood or well reported, but it does vary quite a bit within the actual batchelor but nominally the identical vials. Here's some data looking at one of the materials where we can see we measure the surface area as a function of humidity and you can see here this is a androgen and the surface area decreases by quite a lot due to humidity changing the process properties of the particles in the freeze dried material. I think we've seen this before, so I don't think we need to see this again. No, no, no. You've said this before now, so it's a fairly short presentation, this one. So that means your lunch is even closer than you thought. So we can look at the freeze drying of collapsed cakes if we want. We can also determine the surface area. The surface area using IGC is a bit more accurate, little bit easier to do than using nitrogen. And we also can see that there's some variations in the surface areas of vials, which is in trinity to the process. But igc, operating at room temperature and room pressure, using octane, for example, can measure surface areas down to as low as 0.1 meters squared per gram. So the ones we saw today were about 1 meter squared per gram. Nice. You can do those slightly high errors, but when they get smaller than that, they the advantage of the IGC method become more substantial. I thank you for your time and attention and happy to answer any questions you might have.

Okay, so I'll just go ahead and close the session. Nothing forward to say. Big thank you to D and everybody for joining us. Thank you, D, for taking the time for doing this. In the next webinar this session is being recorded and we'll go ahead and we'll try to send it by email to everybody who attended and people here in the room. So thanks again and see you later.

See you later. Bye Bye.