Webinar: Advances in Measuring Gas and Vapor Permeation Through Membranes and Barrier Polymers

So, yeah, thank you everyone. Those who are joining. I'm just going to give it a couple of minutes before we start the session and then I'll give a brief introduction of our panelists for today and then start off with my presentation to open proceedings. So just give it, give it a minute too. Well, we're at half two. I'll make a long start, if you will, and then hopefully the stragglers can join in when they can. Yeah. So, welcome to today's session. We are at Service Management Systems. We are a company that produces instruments traditionally to characterize porous materials by gravimetric methods. But recently we've expanded into the characterization of membranes and barrier polymers, looking at permeation, kinetics and equilibrium data using our new instrument, the Membrane Permeation Analyzer Horizon, a multi component system designed to look at permeation under simulated real world conditions. And we have a very eclectic panelist today. We have John Zelinsky joining us from Intertek, we have Rafael Oris and Dan Zhu joining us from Michigan State University. And we have Sunshine Iguodala joining us from Imperial College as well. And I want to thank them for joining us today. And they will be presenting. Following my initial presentation, there will be a Q and A after each speaker's session. We will briefly be taking questions after each presentation is over. You can use the webinar control panel to submit your questions during the talk and then we'll read them out to the presenters or the presenters perhaps might be able to find them themselves. But nevertheless, we'll endeavor to make sure your questions are answered during each Q and A session, after each presentation, and potentially at the end, there'll be time for an open discussion, but that depends on the timings available at the end of the webinar. So, yeah, this is my presentation, starting off with I want to thank you for joining today's webinar and I'm going to be discussing open proceedings measuring gas and vapor permeation through membranes and barrier polymers. It's going to be an introduction and I think a lot of the speakers today are going to be doing a deeper dive into specific aspects of this presentation. It's my privilege today to be introducing our new instrument, the MPA Horizon, designed around tackling multiple component permeation challenges. It is a very flexible instrument designed for complex research applications. And yeah, I'll be going over some of the specifications and a few case studies that we've done to show off its capabilities. So I want to start by introducing the issues. One of those issues is that polymers can be a very expensive contributor for an individual process. You know, depending on the number of cycles that the polymer can do, depending on the lifetime of each polymer in a particular process, those costs can build up. If you have to keep replacing that polymer, say it's a carbon capture plant and there's a lot of fouling going on. That fouling and repurchasing those membranes can add cost to the process. So one way around that is by characterizing the degradation pathways and fouling that occurs in your polymer. And to do that, really you need to have knowledge of how the polymer behaves under the process conditions which you're going to be using. And to do that, you need an instrument that has the capability to reproduce those conditions. So polymers can be costly. And to reduce those costs, we need to increase the lifetime of these polymers. One other challenge that we're facing is that a lot of new polymers are entering the industry with very different properties to those traditional petrochemical polymers that we're used to using. I'm talking of course, about bio based polymers, which tend to be less stable when interacting with water. They tend to have a lot of degradation pathways linked to water uptake, and therefore the interactions with water can be very significant. There's also a lot of CCS polymers coming onto the market around polymers of intrinsic microporosity. And those with amine functionalization designed to have stronger interactions with CO2, also have stronger interactions with humidity. And water is a very active foul. And that foul can change the properties of the polymer. It can disrupt the linkages between polymer chains, and that can affect the permeation of gases and vapors through the polymer, affecting its usefulness as a polymer in that particular process. So watering directs very strongly with bio based enamine functionalized polymers. And one of the challenges is defining realistic conditions. I want to show now just the CCS industry. So I want to look at the impurities coming out of different waste streams from processes that produce CO2 and the impurities that are present. And you can see the concentrations of each impurity differs very significantly from process to process. There are a number of poisonous species, your knots and socks, carbon monoxide, CO2 concentration is fairly stable, but you've got changes in concentration of humidity, oxygen and nitrogen. Those need to be taken into account when introducing membranes for these particular applications. Now, the traditional way of characterizing membranes is to use a constant volume, variable pressure method. In this, you're vacuuming one side of the membrane and then introducing a specific gas in the feed to the other side of the membrane, tracking the change in pressure as gases move from one side of the membrane to another. This is a very powerful technique for determining of gas permeation under different process conditions. Where you've got changes in temperature, changes in pressure, concentration, polymer aging, different types of membranes, this process works very well. But if you are checking pressure change and pressure change is not a selective method of determining which components are permeating through your membrane at each time. And I'm talking about the presence of contaminants, the humidity present, the presence of competing gases and fouling agents. And to do those studies you really need a system that has specific sensors designed to pick up and track the concentrations of those individual components under real time. So we need to characterize the effects of individual filants and the effects of those on permeation. And to do that, we produce the APA Horizon. This is our new permeation unit. As I said and keeps saying, it's designed for multi component permeation, but also more importantly, it's designed for real time measurements of those individual components going through your membrane. That allows for the determination of permeation kinetics, that allows you to give diffusion coefficients, allows you to get partition coefficients for individual species. So the real time measurement of individual species through the membrane is an important parameter which we provide with the MPA Horizon. We have in the system specific sensors for humidity, CO2, organics, oxygen. We have a TCD held in the system as well for the tracking of inert gases and active species. And everything is held inside of a temperature controlled incubator which allows us to reach high levels of humidity up to higher temperature conditions as well. So we really can reproduce those process relevant conditions to look at how the membrane is behaving when we have these contaminants present. So I want to go over the two main contributors to permeation. The first is pore flow pressure driven process typically found in nanofiltration. And the other one is solution diffusion, a concentration driven process where first the permeating species must first absorb into the polymer, diffuse through and then dissolve. In reality it's the two processes that contribute to a particular pollen earth. But in our MPA horizon, it's solution diffusion which is the main driving force. So we start off with concentration generation. We're generating biobubbler concentrations of organics and humidity, then introducing select gases at concentrations applicable for the process and that is introduced over the top of the membrane. And then we have a sweep gas shown in green, which takes the gases permeating through the membrane to our sensor suite. And that from that we can get the permeation curve. And I say it's solution diffusion. And that's because this process is all done under atmospheric pressure, in which partial pressure, the concentration difference is the driving force which moves molecules through the membrane. You can generate a slight overpressure on the feed side, but it's not a significant overpressure. Okay, so this is a very brief schematic of the MPA horizon as shown in the top left. You can see the concentration generation with regards to humidity or organics, as well as the dilution of dry gases and gases of interest such as CO2. It then passes through an inlet sensor suite where you have relative humidity and optional CO2 sensor which then passes in the feed over the top of the membrane. And then as I said before, we have a sweep gas which takes those species which permeate through the membrane to an outlet sensor suite. This is what the control software looks like. So you get real time data of individual components going through your membrane. In this case, I was first conditioning a membrane at select humidity and then introducing a concentration of CO2, tracking the outlet CO2 concentration as well. And you can obviously change temperatures, monitor pressure drop across your membrane. All of these very significant contributors to a mathematical formula which you need to calculate certain parameters from your membrane. Okay, I think Sunshine is going to go into this a little bit more, but I just want to go through some of the data that you can get from a permeation curve. The first is equilibrium permeation data. So just how much of a particular gas permeates through when you introduce a feed gas? In this case, the equilibrium permeation data shows, you know, we're achieving about 4.5% relative humidity on the outlet. You can look at the integral of the curve and from that you can essentially get the mass of a penetrant over time. And that will give you metrics such as the water transmission rate, oxygen transmission rate, CO2 transmission rate, which of course is an important variable in packaging industry where you want to be able to track how much of a particular gas is moving through your barrier over time. Okay. And that is also possible on the MPA horizon. And finally there's time lag measurements as well that allow you to get diffusion constants from steady state portion of the transient increase in relative humidity. Oops. So as I said at start, single component permeate, multiple component permeation, determination of permeation kinetics, measurement of transmission rates for individual components. And on this system as well, we have methods and sequences for easy use, which allows for a lot of flexibility, but also a lot of usability in terms of putting a membrane into the system and then walking away and collecting data. So let's go through a few case studies. Now. I'm going to start by going over humidity permeation through barrier films. Then I'm going to move on to multiple component permeation of humidity and vapors and how those two can interact and affect the permeation of each other. And then I'm going to move on to measurement of gas permeability under humid conditions. In this case, how the concentrations of CO2 permeating through membranes can be affected by the humidity of the process. So let's start off by looking at humidity through barrier polymers. And we're going to be looking at two different polymers. Pet, a typical packaging polymer. You'll find it in your plastic bottles for water and Coca Cola. You know, it's used in a lot of food packaging. It has a very low glass transition temperature, about 70 degrees C, but it has very effective water barrier properties. And then we've got Kapton, which is a common polymer electrical insulation, has a very high glass transition temperature above 300 degrees C. And it acts as a much more rigid barrier at those higher temperatures. And we're going to compare the two, starting off with how the permeation of humidity varies through these polymers as we increase the temperature in this case up to 130 degrees. Here we're measuring, introducing absolute humidity. So about 22 grams per meter cubed of absolute humidity to the surface of the membrane. And as you can see, at this temperature, PET is showing a higher equilibrium permeation concentration of water, but also faster permeation kinetics. Okay, it's past its glass transition temperature. And that permeation is a lot faster than Kapton, which remains below its blast transition temperature and much more stable. In this case, we've also used humidity as a probe, in this case, monitoring the water transmission rate as we're changing temperature. And we've done this again with PET and Kapton. And as you can see, below 70 degrees C, the water vapour transmission rate remains around 11 grams per meter cubed squared per day. But as we approach the glass printed temperature range of pet, that permeability increases quite significantly. Okay, so it increases from 11 to about 15, so about by 50%. And that allows us to observe when the polymer is increasing its free volume, when those chains are becoming more flexible. And that is characterized by an increase in permeability for Kapton because of its higher Permeability, sorry, glass transition temperature. That permeability remains fairly consistent. We've also looked at composite membranes. So the metallization of polymers is fairly common in the packaging industry. An example of that would be a packet of crisps. They tend to have a very small layer of metal. In this case, we looked at adding a very small layer of aluminium to our pet. We saw that in this composite material, the water vapor transmission rate is decreased significantly and we no longer observe the change in permeation that happens when we go over the glass transition temperature. And this is because in this case, aluminium is acting as a barrier to humidity and yeah, it's just stopping all the humidity permeation that might have occurred through the traditional PET at those higher temperatures. Let's go over now some more complex studies looking at how some binary mixtures of gases and vapors change their properties in those multi component mixtures. Okay. And we're going to be looking at pet, again, a robust polymer under humidity. And we're going to be looking a little bit as well at pla, but I think Rafael and Diane are going to go into that a little bit more. But I just wanted to briefly introduce that work as well. So pet, very robust under humidity. PLA polymer, which can change under humidity depending on the conditions in which it's being introduced. Let's start off by looking at pet, looking at mixtures of water vapor and toluene. Okay, so when we have a very small concentrations of toluene, how does the introduction of water vapor change? Now PET is an interesting polymer, especially when toluene is introduced, because toluene can induce crystallization of the polymer. And in this case it looks like that's what we're seeing. So we're seeing a decrease in humidity permeation as we're increasing the Ptolemy concentration, both in terms of absolute values at the outlet, but also in terms of the kinetics of permeation of humidity as well as now we did expect to see some plasticization of the polymer, but in this case what we saw was the opposite. We saw a decrease in the permeation of humidity as we increase that toluene concentration. And I think that demonstrates the necessity of characterizing these complex mixtures permeating through your membranes. We can of course also do the reverse. So we looked at how that toluene concentration was increasing, permeation was changing when increasing relative humidity. And we also saw the same story. So we saw that increasing the relative humidity produces a decrease in toluene permeation. And again, no plasticization occurs. Those two species act as competing species in the membrane for that free volume and therefore decrease the permeation through the membrane. Finally, we're going to be looking at PLA with water and ethanol. And I didn't want to go into too much data, but I just wanted to show a different case study where we have the addition of ethanol increasing that water permeability and that is consistent with temperature. This is of course different to the behavior of PET and toluene. It's just another case study which shows the necessity to characterize these multi component mixtures permeating through your membranes. I think Rafael and Diane are going to go into the complexity of ethanol water mixtures and how those interactions PLA in their presentation later on in Proceedings. So finally I'm going to go over some work I did in collaboration with University of Manchester with Professor Peter Bird's group going over polymers of intrinsic microporosity designed for applications in carbon capture and sequestration. We're going to be looking at Tim1, which is a slightly hydrophobic polymer. It has a very narrow pore distribution around 2 nm, designed to sieve CO2 from mixtures of CO2 nitrogen in Flue gas. We're going to be looking and comparing that to its carboxylated analog CPIM1, which is slightly hydrophilic. And we've done some DBS studies to show the water uptake in these two polymers. And as you can see with CPIM1, that mass change occurs over the whole range of humidities. So it's very hydroscopic. It takes up humidity up to 20% of its mass at the highest humidity level when compared to PIM1, which has a very small humidity uptake, up to about 70% relative humidity, and then it increases sharply up to about 6% of its mass. Okay, so these two polymers, very different hydrophilicities, CPIM1 much more hydrophilic. And that'll become relevant when we look at the interactions with humidity and how that affects CO2 permeation. So these are the two structures of PIM1 and CPIM1. PIM1 is shown above with those cyano groups. And CPIM1 has these carboxylic acid groups, which can form hydrogen bonding between polymer chains inside of the polymer. And this results in a decrease in the free volume polymer, therefore lowering CO2 permeation but increasing CO2 nitrogen selectivity. Okay, so the voids in the polymer are much smaller and therefore the size selection for CO2 is much greater. So we have an increase in CO2 nitrogen selectivity. So in these typical experiments, we first start by introducing our humidity and we introduce that target humidity shown in blue there for about three hours. And then we introduce our CO2, in this case a 10% CO2 in nitrogen for about an hour, tracking the inlet and outlet concentration whilst maintaining the selected humidity. And we're using helium as the sweep gas to determine the concentrations of nitrogen permeating through the membrane and and then how that moves to outlet sensors. Okay, so let's have a look. First, humidity permeation through PIM1. It starts off initially high, and then as humidity increases, that permeation of water decreases. Okay, so there's a buildup of humidity inside the polymer and that is causing a decrease in overall humidity permeation. The opposite is what we observe for CPIM1. That humidity permeation starts off initially low. Again, it has a much smaller free volume, much narrower pores, and therefore humidity permeation is a lot lower than PIM1. But that permeation appears to increase as we increase humidity. And we think what's happening is the water is disrupting the hydrogen bonding between the chains in the polymer and that is causing an increase in free volume. It's causing some swelling to occur, and that is resulting in an increase in water permeation as we increase the relative humidity. Let's look now at how that CO2 permeability changes in relative humidity. So in this case for PIM1, increasing that relative humidity results in a decrease in CO2 permeability from about 3,000 bara down to 2,500 barre, which is about a 20% decrease, which is somewhat significant in terms of how applicable the membrane is to carbon capture processes. Again, this is caused by a buildup of humidity in the channels, which reduces CO2 permeability through the membrane. And we also see this occur for the more hydrophilic polymer, but this happens much more significantly. So humidity increases, the channels get blocked and CO2 permeation decreases. But for CPIM1, we see a decrease of about 40%, whereas for PIM1 we only see a decrease of about 20%. Now, in terms of absolute values, this is a decrease of about 600 for PIM1 and a decrease of about 900 BARY units for CPIM1. This has implications in the selectivity of the polymer. And this is the CO2 nitrogen selectivity for CPIM1, which shows, okay, you know, our CO2 permeation is decreasing with relative humidity. The nitrogen permeation only starts to decrease at much higher relative humidities. So when Those channels are really getting blocked. That's when there's nitrogen permeation starts to decrease and this results in a sort of tailing off of the selectivity. But we still see a decrease of about 20% in the CO2 nitrogen selectivity of the polymer, which decreases the necessity, if you will, of using CPIM1 when it has a much lower permeability. But now the benefit of the highest selectivity is also being lost because of its interactions with water. So sometimes the modifications we make to improve polymer performance, in this case the selectivity of the polymer can have a detrimental impact when introducing multiple species into that feed stream. So when humidity is present, that water is interacting with those carboxylic acid groups, which causes a buildup of humidity of the polymer and results in a decrease of CO2 permeation and of course, the selectivity. So I want to finish by just going over again the MPA Horizon and its capabilities. So we are tackling multi component permeation of gases and vapors. It allows for the complex determination of permeation kinetics, but also partition coefficients and diffusion constants. It has the measurements of transmission rates of multiple components with high selectivity sensitivity and selectivity. And it's a very powerful, but also more importantly, very flexible permeation instrument. So it can tackle a variety of different permeation challenges. And I want to finish up by thanking you for your attention. I want to thank the other speakers for joining us today, my colleagues at Central Event Systems and our collaborators, both at University of Manchester, Michigan State University, who are also in the panel today. And I invite any questions you might have on my talk. If you have questions, please add them into the question tab and I will try and answer them best I can. So what is the maximum temperature and humidity of the MDA Horizon? That's a good question. So the maximum temperature in which you can go to in the incubator is between 10 and 60 degrees, and that is the temperature at which you can achieve 0 to 95% relative humidity. Now, we do have an individual local heater at the sample which allows you to go from 10 to 150 degrees. Okay. So different parts of the instrument can go to different temperatures. And at your sample you can go to 150. But for the generation of humidities and VOCs, we only go from 0 to 60 in terms of generation. Okay. Is it possible to test hollow fibers? We are working on producing a sample holder for hollow fibers, but currently it's just flat sheets of membranes because the hollow fiber bundle introduction adds A little bit of complexity into the flow. So we're working on it. But currently this version only does flat sheet membranes. So yeah, it's my privilege to introduce our next speaker, John Zielinski from Intertek over in the usa. And he's going to be talking about diffusion and solubility and how insights into those particular metrics affect gas and vapor permeation through barrier polymers. Please take it away, John.

Well, first of all, thank you very much for the opportunity of presenting today and for setting up this webinar. The focus of my presentation really is going to be on extracting diffusion and and solubility information from permeation experiments. I'll talk a little bit about when you do a permeation measurement, what is actually being measured. Relate that to discussions of flux diffusion, permeation coefficients, solubility coefficients. I'll talk a little bit about trends that I've seen and how to go from solubility coefficients to actually isotherm data and finish off with some special considerations, maybe for vapor systems. As a brief introduction, I work at a company called Intertek which is actually just down the road from SMS in Allentown in the USA. Intertek is a quite large company, about 40,000 people worldwide and we do auditing, inspections and materials testing. So our lab in Allentown, we were the corporate analytical labs for air products and chemicals where I spent majority of my career. But we do materials characterization. So we have the nmrs, lcms, SEMs, those sorts of things. So if there's any material characterization needs that you have, by all means please reach out to us. But we also have adsorption lab and I'll tell you a little bit about the types of studies that we do there. I like always starting off this talk with a question to the group and asking If I had two balloons, one filled with helium and one filled with CO2, which one would you expect would deflate faster? Most people would say that the helium balloon would deflate faster because helium is a smaller molecule. That's what we use for leak checking systems. CO2 is molecular weight 44 versus helium. That's molecular weight 4. Most people are surprised to find out that actually CO2 deflates a whole lot faster than the helium balloon does. This really illustrates the difference between the mechanism in that the balloons deflating because of permeation as opposed to diffusion. And we'll talk about that a little bit more. But permeation really is a combination of both those quantities. If you think about it, if this is the inside of the balloon and this is in the case that Shawn mentioned earlier about solution diffusion, what happens is the molecule has to come to the surface of the balloon, dissolve in the membrane, migrate across to the other side and pop out. And it's true that helium, being the smaller molecule, diffuses a little bit faster across the membrane. But CO2 is much more interactive with the rubber, and therefore more molecules dissolve in that membrane and maybe they migrate a little bit more slowly. But at the end of the day, more molecules of CO2 are migrating across the membrane in a given period of time, and that's why the balloon deflates faster with CO2. This transmission rate coming across the membrane I'll refer to as flux is related to the gradient and concentration on both sides of that membrane, the thickness of that membrane, and also that diffusion coefficient. And if you assume a linear isotherm, you can convert this over to coefficients where this permeation coefficient, which is more of an intrinsic property of the polymer and the penetrant, is related to both the diffusion coefficient and solubility coefficients. And we'll talk a whole lot more about that. Quite often I'll get asked is, can you measure a transmission rate? For me, because I'm concerned about my nitrile glove. How long is it going to keep my hands protected from exposure to chemicals? I'm like, well, sure, I can measure a transmission rate or a flux from a steady state measurement, but really what you want to know is how long is my skin protected? When is the breakthrough of the molecule going to be upon and giving me the first exposure? So it's really the diffusion a lot of times that people are interested in. Same way when you have electronic components, if you encapsulate something in a polymer and you want to find out how long are my electronics protected against high humidities or, you know, when you have a spill, it's the diffusion rate that you're interested in terms of, you know, how much, how long it takes to break through as opposed to how much is transmitting over a long period of time. Sean mentioned earlier that, you know that there are units that measure transmission rate by pressurized. And that is one of the units that I have in my lab, having worked for air products and chemicals. When we were using membranes for gas separations and looking for things like oxygen nitrogen separation or methane, CO2, hydrogen separations or oxygen argon, it was a lot easier to have a sensor, which is effectively my pressure transducer, so I could have any gas flow over the top surface with a vacuum initially on the bottom surface. And when I close that Vacuum off the gas would migrate across the membrane and cause a pressure rise. And so, effectively, what we're measuring is how does that pressure rise over time? And once that vacuum is closed off, I have a finite volume, therefore I know the volume, I know the pressure, I know the temperature by ideal gas law. I know the number of moles of gas then that are transmitting over a certain area as a function of time. And that's nothing more than flux. Right? And so when you're doing an experiment like that, what you're truly measuring is the flux, the transmission rate. If you're dealing with oxygen, you'd call that otr. If you're dealing with carbon dioxide flowing, it'd be the cdtr. But it's nice in this system that I could swap out the gases, but it's true that it's limited only to a single component gas. I can't do multi components in this case. But once you have that steady state flux, if you multiply that by the thickness of the membrane and divide that by the driving force pressure across the membrane, then you can calculate a permeation coefficient. Now, the nice thing about having a permeation coefficient is if I wanted to change what my driving force pressures are on both sides of the membrane, I could calculate what the effect of that would be on the transmission rate. Same way, if I changed the thickness of my membrane, I could calculate directly how that would affect my transmission rate. The main takeaway from this slide is that you get the flux, the transmission rate, from steady state behavior. If we do the experiment in a slightly different way and first pull vacuum on both sides of the membrane and then shut the vacuum off and initiate the gas flow, well, my pressure doesn't change for a period of time. It's flat. And it takes a finite amount of time for a gas molecule to migrate across the membrane. But when it pops through, you start seeing the pressure rise. And again, at long term behavior, you reach the steady state flux. And as Shawn pointed out, if you extrapolate that back, you can get this time lag from which you can calculate a diffusion coefficient. So again, the things that you're measuring in this case are really the time lag from which you can calculate a diffusion coefficient, the steady state flux, and from there you can calculate the permeation coefficient and solubility coefficients, but you're not measuring them directly, you're kind of inferring them. So from the transient experiment, you can get both sets of information, whereas from the steady state, you really only get the transmission rate, the flux now, again, with our system, I showed you we could swap out gases. So I did that and a couple of systems where we varied molecular weight of gas from helium to xenon, krypton. And you could see that the diffusion coefficient decreases as molecular size increases. That's not really surprising. The bigger the molecule, the more effort it takes to move around. But you can see in the glassy system of Teflon, which has a glass transition of about 125 degrees C, there's a stronger dependence on molecular weight than in the more rubbery system of silicon rubber, which has a glass transition temperature of about -120 C. But in general, diffusion is decreasing with increasing molecular size. The solubility coefficient, on the other hand, is a bit more of a scatter plot. But if I had to say anything, I would say that in general the trend is that solubility increases with molecular size. So you have really two competing effects, which one dominates. And if you look at the permeation coefficient, it's interesting that it has opposite trends. For the more glassy system, it seems that diffusion is winning over solubility. And in the rubbery system, it seems that solubility is the predominant mechanism that's governing permeation. And it really got me interested in solubility measurements. And, you know, when I think of solubility running an absorption lab, I typically think of gas isotherms. And you know, you usually take a porous media or a polymer, expose it to a gas either by a gravimetric or manometric type of experiment, and you're measuring directly the, the weight uptake at a fixed pressure of the gas. Here I have a system of silicon rubber around room temperature. This is the isotherm for CO2 measured with a low pressure micromerx asap unit and a high pressure differential pressure unit that we have developed in our lab. You can see the data are in quite good agreement, but you can see that CO2 absorbs much more so than oxygen does. But when you're measuring these isotherms, usually you're reporting the data in terms of an amount of uptake, grams of gas, and you're dealing with the grams of dry material that you typically put into the, into the system. So it's the weight percent is typically reported grams of gas per grams of polymer. I'll show why I'm highlighting that here. When you're doing the membrane experiment, we're evaluating the concentration of gas really at this surface of the membrane. If you go through the math, this concentration is really grams of gas per cubic centimeter of solution. Now, in a very dilute system, if you're dealing with low uptakes of gas or very low uptakes of vapor, you can assume that this is cubic centimeters of polymer of pure polymer. And then if you know the density of the polymer, you can convert that easily to grams of gas per gram of polymer. But when you're dealing with vapor systems, in particular, when there's a lot of absorption taking place, that assumption that this is really species of polymer breaks down. So I understand that this is a busy slide, but try to follow me through on this. But what we're assuming is really Fick's first law here that I mentioned, that we could measure the transmission rate. From the breakthrough, we can get the diffusion coefficient. If gas is being swept on this side of the membrane, or if the pressure is maintained low, then the concentration on this surface of the membrane is effectively zero. So really, the only unknown in this equation is what the concentration is at this surface. Now, typically, to convert this over to a weight percent, you need to assume some sort of an equation of state or mixing rule. Normally what I do is to look at essentially weight fraction additivity of volume. So if this is the weight, let's assume that this is water transmission. So this would be the weight fraction of water divided by the density of liquid water plus the weight fraction of polymer divided by the density of the polymer. And that gives you the mass concentration in a solution. So really, the only unknown in this equation is the weight fraction. And from there it's easy to calculate the weight percent, which is again, grams of gas per gram of polymer leading you through all of this. If we go back to our silicon rubber system, this is what the data looked like for breakthrough. So CO2 takes longer to break through, so the diffusion coefficient is slower, it's lower than for oxygen, but you can see the transmission rate. Ultimately, the permeation rate is going to be a lot faster. If we do the analysis as I suggested, you measure the flux, you have the diffusion rate, you know, the thickness of the membrane, you can calculate what that concentration is at that surface. And when you do that analysis, you can see that the various techniques line up quite nicely. So, yeah, this is the point for CO2, that's the point for oxygen. Like I said for the Dow cell system, I unfortunately can only do this at a single point because the gas flowing over my membrane is at one atmosphere of pressure. So you're measuring the isotherm point, they're loading at 1:1 atmosphere of gas pressure. For vapor systems like the one that Sean talked about, their horizon unit, you can actually vary the concentration along the top surface. And in principle, you should be able to measure the isotherm as a function of the relative humidity. Unfortunately, I don't have one of those units. I'm trying to get my management team to talk me into, or talk them into buying one of those units for me. So I'd love to be able to do this experiment directly. But in the meantime, I did a poor man's approach to this where I looked at a system of toluene with my silicon rubber, and I filled the vial with toluene, sealed a membrane across the top surface, and put that system on an analytical balance. And effectively, what I'm measuring is weight loss, right? So there's nothing at first, and then you see the breakthrough, and then at long time, we're measuring the steady state weight loss of toluene through this membrane. If you look at this a little bit more carefully, the point at which we're measuring the concentration is at this interface where the vapor is in equilibrium with the liquid. So this is really 100% relative humidity or 100% relative pressure in the case of toluene. So when you do a comparison to an isotherm that we'd measure gravimetrically, you could see that at the 100% relativity point, it lines up quite nicely. But you have to know where you're measuring the concentration. Right? And even doing a system like this, it seems like it works out quite nicely. So I'd like to conclude that. That the isotherm. You can measure isotherms with vapors by doing permeation measurements as well. But there's a caveat, and I want to talk about that a little bit. We'll talk more about that later on in the presentation. But the diffusion coefficient in this system is relatively independent of concentration. So when we're making that Fick's law, first law approximation that there's a linear concentration gradient through the membrane, it's valid because the diffusion coefficient is weakly. Is really weakly concentration dependent. And I'll show you an example later on where that's where that may be more problematic. But let's talk a little bit about how you do a permeation experiment. Normally, as Shawn describes, usually what you're doing is you're humidifying a gas and flowing it over the top surface and watching the response. Now, if you start off with a flat concentration profile and then do a step change in the humidity on the top surface, you form an equilibrium point immediately on that surface, and it takes a while for the system to break through. So that's what you're ultimately following the break, breakthrough, and then the rise to a steady state transmission rate. And then normally what's done is that you'd take the next step, just immediately increase the concentration to the next level and follow the transmission until you reach that steady state. And then you can kind of work your way up to higher and higher relative humidities. And that's a perfectly good way of doing a vapor transmission experiment. And here I'm showing some data for water vapor transmission as a function of relative humidity, function of temperature as well. And the data line up, as you'd expect from Fick's first law, where, you know, if we're keeping this concentration low, the relative humidity increases the concentration of the uphill side of the membrane. So if you're increasing this driving force, well, your transmission rate's increasing. And it's also no surprise is that when you increase the temperature, your diffusion coefficient increases. So your transmission rate goes according to expectation, right? It's increasing with increasing temperature. It's increasing with increasing driving force. But if I take this data and I convert it over to a permeation coefficient, like I calculate it knowing that transmission rate that I measured the membrane thickness and the driving force pressure, and I make an Arrhenius plot, I was surprised to see that the behavior, you know, it's linear. But usually when you're making Arrhenius plot, you have slopes going the opposite direction. When you have diffusion, when you go to higher temperatures, your diffusion coefficient increases. In this system, the permeation coefficient is decreasing with increasing temperature. Certainly not, because the diffusion rate is decreasing with increasing. The diffusion coefficient is not decreasing with increasing temperature. So it must be true that the solubility coefficient is. And in this case, the solubility is seen to have a stronger temperature dependence than the diffusion process. But let's talk just for a couple of minutes on what happens if the concentration, if the diffusion process is concentration dependent. And here I'm showing simply modeling results. So normally we're assuming that the diffusion constant is flat, right? And that leads ultimately to a linear constant concentration gradient across your membrane. If I force it to have a dependence of, say, an exponential dependence, where a can be positive 20 or negative 20. And I chose that. So it gives me about an order of magnitude change in the diffusion coefficient on either direction. What you'd see is that if A is positive and D is increasing with increase in concentration on the Upstream side of the membrane, that's where you have your fastest diffusion point. And as the molecules migrate through the membrane, they're actually slowing down the diffusion process. On the other hand, if it's, you know, if A is negative and it slows with increase in concentration, it's actually slowest at the highest point of the membrane where every molecule has to transmit across is going through this slow, very slow diffusion region. So what we see experimentally are these breakthrough curves where this is your increase in concentration D. This one's where D is decreasing with increase in concentration. But experimentally, you'd have no idea. It just looks like a normal breakthrough curve. And if I looked at this region extrapolated back, got my time lag calculated, my diffusion coefficients, the values that I get are intermediate to the diffusion coefficients at the two faces of the my membrane, which is really no surprise. But the thing that bothered me about this most is that in my simulation, I set the upstream concentration at point 1. What I evaluated from this breakthrough analysis, as we did with the other systems, are concentrations that show about a 50% error. So clearly that's unsatisfactory. And you just can't get a concentration or measure an isotherm in a system that's like that. So my proposal. Well, and again, the reason that this happens is that, you know, when you have an independent diffusion coefficient that's constant, no concentration dependence, you wind up having a linear gradient through the membrane, whereas when there is dependence, you wind up having curved profiles. And so that impacts how you need to do that calculation. My proposal would be that as you do with, if you're trying to measure a diffusion coefficient in a gravimetry experiment, it's really try to take small, small steps in your relative humidity. So you'd start off the same way. By having a flat profile, you take a much smaller step in the concentration than we did in the previous example. And then what I propose is that you flow humidity on both sides of the membrane so you can flatten the profile. And then when you take the next step again, you'd watch the breakthrough, your transmission rate wouldn't be increasing to a much higher level. Your driving force would maintain about the same, but your profile would be flatter and there wouldn't be as much concentration dependence that you could see throughout the membrane. Again, you do that for the following steps. I think that's the better way of doing the experiment. If you're trying to extract out both diffusion and concentration, isotherm points. To end on a positive note though, in most membrane Systems of interest, if you have cross linked polymers and you're dealing with gases, your solubility is going to be very low and so you really won't have very strong diffusion effects even in membrane systems where there is uptake. If you're at very high glasses, it could be that your diffusion coefficient won't be very strongly concentration dependent. But I think it's important to keep in mind that we're, when going through the analysis, we're implicitly making the assumption that Fick's law holds that it's a linear gradient. And that might not be the case. Performing the experiment in this manner may be a better way of doing it. But I'd love to be able to borrow one of the SMS systems and actually run through the testing or maybe we can get together and do that at some point. But in conclusion, I just wanted to indicate that when you do steady state experiments, you measure the transmission rate, transmission rate you can then convert over to a permeation coefficient. If you're doing a transient experiment, you get both the steady state transmission rate, but you can also back out to diffusion information. And once you have both of those, you can get information about permeation coefficients that then allow you to calculate, calculate transmission rates at different driving force pressures, different membrane thicknesses and it allows you to measure solubility. And I've shown you how to do that for isotherm points. And lastly, you know, if you are interested in measuring diffusion and isotherm points from permeation experiments, I think it's important to be able to do that with small diffusion or small relative humidity steps. So that concludes my presentation. I'll be happy to answer any questions that you might have.

Thank you John for the very interesting talk. Yeah, if you have any questions for John, please post them in the chat and we'll try and get them answered for you. I think while we're waiting. John, I had a question myself which was around the interesting proposal of flattening the permeation curve before each humidity step. I wonder, would you achieve the same diffusion constants gravimetrically, let's say in a DBS where you have essentially equilibrium around the system and you're monitoring the mass change, would that replicate those experiments?

Absolutely. But again it's important to, if you had a system where the diffusion coefficient is say strongly concentration dependent, you really want to measure, minimize that step change. And so there's a work that I can send you that the actually diffusion constant that you measure is related to a concentration that's somewhere in between the equilibrium level. So if the diffusion is increasing with concentration, it's at about 0.7 or 70% of the equilibrium level. If the diffusion coefficient is decreasing with concentration, it's at about 56% of that interval. But in principle, yes, if you do those steps, what you measure via the membrane experiment or gravimetrically or by pressure decay, they all should agree.

Okay, brilliant. Definitely look into doing some of those experiments. We'll move to our next speaker, which is Jewel speaking with Rafael Oros and Dan Zhu of Michigan State University. And they're going to be presenting their work on ethanol interactions with polylactic acid and how that induces crystallization behavior with temperature. So please take it away. Rafael and Dan. Perfect.

Can you hear me okay?

Yes, we can hear you.

Okay, perfect. Okay, thank you very much. Thank you for the opportunity to present here. As Sean said, this is a dual presentation. Very nice presentation from John. It would be nice that all my students taking our classes had the opportunity to be in the seminar because I think that was a very good walkthrough, the measurement of permeation and diffusion and what happened when it's concentration dependent. So the work that we present today is little more into the liquid phase of perniation. And you will see why we're interested in that. So I'm from the School of Packaging. My name is Rafael Auras. I'm a professor there. I've been there for many years. And Diane Shu, the one that will present the second part, is a PhD student, actually PhD candidate already. And she will present a little more of the research that she's working on. So next slide, Dian. So for the ones that actually don't know where Michigan State University is, we are in Michigan, in the middle of the meeting. We are the Green University, so please don't confuse with the blue one. And next slide, Diane. And we are in islancing.

Next slide, please.

And in the middle of campus there is a school of packaging. So yes, we do have a school of packaging. And the School of Packaging, next slide, Diane, please. Was founded in 1952. And since then we have a bachelor program in packaging and a master program that was funded two years after. And in 1996, we realized that in order to conduct research in the area of packaging and for packaging, we need to have a PhD. So we create the PhD in packaging. And students at the School of packaging will work or will do research in many different areas. So mass transfer is one of the areas that we do. It mostly, mostly concentrated for the area of packaging. Application for medical or for food and also some for the military packaging. Next slide, Diane. So in the US we're ranking number one in the packaging program. Maybe because we're the oldest and the largest. But anyway, so that we're ranking number one. 40% of the students actually the packaging professionals in the US are coming from the School of Packaging at MSU. So as I said, we have a PhD in packaging. So if you want to have a super, super, super degree in packaging, you can come and have your bachelor, your master and you can do your PhD all in packaging. So we have around 600 students, 520 are undergraduate students and around 80 graduate students doing the master's and the PhD. And you will see that many of our students around the world, we have trained many of those students. And since we are very active in the research and layering, as packaging become more concerned, that is scaling a lot. And we have offered comfort more than 10,000 degrees since the 1952. Next slide. Ian. So as I say, this is the team on the screw of packaging. I'm there in one side. We are recruiting new people and we have people working in many different areas in food packaging, material science, very heavily in plastics and paper based work. We have distribution packaging and makeup packaging. As I say, next slide. And I'm particularly working in three different areas and today we'll actually discuss one of them. I work in developing a packaging structure with tailored property and that tailor property tend to be a lot in the area of a mass transfer for different type of packaging application in medical or in food where you need to have particular oxygen barrier, water barrier, or you want to create some functional barrier, or you want to create some particular positive migrations and compounds. Work a lot in the area of inner life and since in the last 10 years a lot also in the environmental footprint.

So today.

Next slide. Jan. So we would like to discuss about the part that we're working for a couple of years already in polylactic acid as a polymer derived from bio based resources. And the end will explain more on that. But we are as one of the newest polymer coming for packaging application derived from Biobase. We still don't have a lot of information of this polymer and how this polymer interact with different organic compounds. And Diane will explain a little more why that is important and why we are concentrating in looking at the polylactic acid properties. Okay, Diane, over to you.

Okay. Can you hear me?

Yes.

Okay, thank you for the introduction, Dr. Auras. And the research topic I am present today is temperature dependent crystallization behavior of polylactic acid films induced by adeno. So there are five sections in today's presentation which are introduction study goals and materials and method results and discussion and final remarks. So let's start with introduction and motivation. We know that Lake PLA is biodegradable and bio based polymer. And it is widely applied in many fields such as agriculture, medical and packaging field. And it is one of the lowest environmental footprints per kilogram of resin. And it can be completely industrial composted to carbon dioxide and water by the action of microorganisms in the environment. And we can see in the life cycle of PLA here the corn can be processed to the starch and we use bacteria to ferment the carbohydrates to the lactic acid. And by using different methods, the lactic acid can be polymerized to the plaque. And then the PLA resin is used to produce the product such as food packaging like recycled bottles or caps and so on. After that it may go to the incineration and landfill. And we know that in the past decade research on PLA films has been growing and a large part of the research has been published in the fields of polymer science, chemistry and engineering, which also reflects this wide range of applications. So I'm going to show you some of the applications of the PLA right now. And there are four main application fields which are medical, textile, environmental and packaging or consumer goods. So there are many chances that PLA films can be contacted with the aqua or in this case ethanol and water during their service life. So this leads to our motivation for this study. For instance, in the field of food packaging, some studies have shown that organic compounds can interact with food packaging and change their properties and performance. So commonly used food stimulants specified by the US FDA are ethanol solution with different concentrations. And in the agriculture field PLA can contact with water in growth stimulation. And by testing different properties of PLA immersing ethanol, we can understand the changes in the PLA during the application. Already know that PLA films have many chances to contact with the ethanol like or other organic compounds during its use. And there will be mass transfer during the contact that showed here. And we know that the diffusion it's followed fixed low and the permeation is followed Henry's law. So due to the diffuse of adenoid PLA fumes, there might be some plasticization, swelling and crystallization inside the PLA films. And the crystallinity may related to the different fraction in PLA spirulin. So we can see here, there are three fractions in PLA's fabric which are rough MOF and crystal crystalline part. So the RAWF and MOF is rigid and mobile amorphous fractions. So we know that the absorption of ethanol can only happen in those amorphous fractions. And the plasticization can increase the chain mobility so that reduce the TG and swelling and crystallization may affect the sorption amount of ethanol in PLA films. So right now we have some basic understanding of whole mechanism. And let's see our study goals. The purpose of our study is to evaluate the mass transfer mechanism of ethanol in PLA films using experiments and also the prediction models. So we want to determine the diffusion behavior and quantify the sorption amount of ethanol in amorphous field of films. To reach that goal, we test our PLA films to make it amorphous at 160 Celsius degree and heat for 10 minutes and use the dry ice to cool it very fast. So we can get the crystallinity of the hot press film around 0%. And then we put five or six sample in the 20ml well and put it immersed into the ethanol. And the immersion time is from half hour to 144 hours. And we do have like two different conditions, one is $0.25 degree and another one is different temperature gaps Delta T that I will explain later. And then we use dsc, NMR and analytical balance to measure the crystallinity soption amount and also the swelling degree of PLA immersed in the ethanol. So the NMR we use the quantitative NMR and using the internal standard we to compare the concentration of ethanol absorbed into the PLA fumes so we can get absorption amount. And we also use the hermetic pen in DSC to avoid the liquid evaporate during the experiment. And getting the creativity of PLA films after immersion. And we use the analytical balance for for measuring the degree of swelling. And we also use the RAMI equation to feeding the curve which is the crystallinity changing during different immersion time. And we use the XRD to see what is the crystal type of the PLA films after immersion. So let's see the results. First of all this is the dry amorphous PLA films thermal properties we know for the dry PLA after hot press the glass transition temperature is around 60 Celsius degree and the cold crystallization temperature is around 126 south degree. And the melting temperature is around 152 south degree. And I mentioned before that we have Delta T that one means the gap between the in situ TG and also the experimental temperature. So what is in pseudo tg? That means the TG of PLA immerse in the ethanol. So during the immersion we measure the TG by DMA and we got the TG of PLA immersed in methanol is around 24 cells degree, which is much lower than the dry mode PLA TG which is around 60 cells south degree. So we have four different temperatures has the different delta D which is temperature gap. The four different experimental temperature is 10, 20, 30 and 40 Celsius degree. We want to know like what is the temperature effect of the crystallinity changes when we immerse the PLA into ethanol. So the PLA's poisoninity emerged in ethanol. From the figure A we can see the yellow part is 40 Celsius degree and green is 30 and the dark blue and the light blue part is 20 Celsius degree and 10 south degree. So we can see the crystallinity changes is very fast. When it is one hour immersion, the crystallinity of PLA at 30 and 40 Celsius degree is already reached around like 25% and it's almost reached a steady state. And we see that there is a large yellow shadow here that is for 95% confidence interval for the remedy feed. So we also do the bootstrap to narrow it down. And we can see the plaque stability immersed in Adeno under The in situ TG which is the blue part was below 10%. And from the figure B we can see this part, the red part is the temperature lower than the in situ TG and the the left part is the temperature higher than the in pseudo TG. So we can see a temperature gap of 10 Kelvin was insufficient to provide enough mobility and induce significant crystallinity changes in the PLA films under the TG and at the temperature below tg. We know that chain segment activity is limited and the free volume in PLA films does not change that much. And because of that, PLA films absorb less ethanol. And from the degree of swelling we can also see that. So after around like 20 after around 12 hours the swelling degree reaches the steady state and we can see no more crystals can be formed during that time. See results. We can see there is no obvious reduction of the TG at the 10 cell screen. But we can see there is obvious reduction of TG when we immerse into ethanol at 20 Celsius degree from 0 hour to 48 hours there is a chance that we can see the TGS is reducing. And from SRD profile we can find that the alpha form crystal peaks started to appear after four hours which is the green line here. And after that we did an Mr. Experiment. We can see from different chemical shapes. This is for PLA spectrum. We know that maple group is around 1.5 and the OCH group is around 5. By doing that we know we can measure how much how much ethanol can be absorbed into the nmr. So we know the integral of different group peaks in the PLA films which is 1.56 and 5.14. And for the ethanol peak is around 1.20, 1.24 and 3.72. So according to the equation here we can measure how much is the sorption amount of ethanol into the PLA fumes. And from 12 hours to 144 hours immersion we know that we have different integral of the ethno methyl group and we can calculate how much is this option amount. So this is the substance amount that we measured from NMR at 25 centigrade which is a little bit higher than the institute TG of PLA immersing ethno and that is 24 centigrade. So we can see there is a trend. And first of all the disruption amount is increasing and at around 4 south degree. At around 4 our immersion the suction amount reached the maximum amount which is about 8%. And after that it is dropping and then reach the steady state at around 24 hours immersion and which is around 5%. And for the crystallinity measured by DSC which can see there is a chance of increasing between 0 to 8 hours of immersion and reach to the steady state around 24 hours. And we also got the swelling degree which is similar to the chance of the suction amount because we measured the weight of the film after immersion so it increased from zero to four hours and go down after four hours and reach the steady state around 24 hours. So while we have a drop in here between 4 hours and 24 hours we think it is because at the beginning so the crystallinity of PLA is still very low. And after four hours immersion the ethanol can induce crystalline crystal in the PLA film. So less amorphous fraction in PLA films to absorb ethanol. So the sorption amount is dropping. And then after we got like the stabilizer amount of the crystallinity in pla, the suction amount is also stabilized because no more crystal are induced. And in conclusion for immersion at different temperature gap, we know that below the ing which is 24 Celsius degree. Here lysiums show lower crystallinity that is below 10% in adenol. And it is maybe due to the chain segment freezing and low free volume. So it can reduce the amount of ethanol soption by PLA films. And above the instituteg we have the experimental temperature 30 and 40 Celsius degree. The chain segment starts moving, the free volume increases and the PLA films can absorb more ethanol. So that increasing the crystallization rate and maximum crystallinity and the swelling degree of PLA films immersed in ethanol it shows a positive correlation chant. And for immersion at temperature 25 degrees, it takes 24 hours. For PLA Films immersing ethanol to reach the steady state, the maximum absorption amount is around 8%. And we do have some implications from the US FDA. So testing PLA at 40 Celsius degree is an extraction and it does simulate its use at even temperature patients with real food. But we also need to cautious for another aliphatic and aromatic polyester as they are certified as food contact. So thank you for your listening and please let us know if you have any questions.

Thank you, Diane. Thank you for presenting the data. Just to wrap up Diane's presentation and to discuss little implications as we have more polymers that are being certified for using package application. And we're looking for these a biodegradable composable polymer. Most of those polymers are polyesters. And so I think that there are two things that would be important to get from this from this presentation is one, most of these polymers will be in contact with some type of product when they are actually in the packaging application. So they will be in the liquid phase. But also you imagine the bottle, it would be in the vapor phase. And you can see there when you have that, that the issue that can happen is that if you are measuring that, what we want to try to show is like this in situ tg, what is the polymer? C is much, much lower than that. We generally determined by DSC. No, it's almost 20 degree. And the penta alcohol will change. So that's very important for applications and for certification purpose. Because for example, in the FDA and the EFSA we still certifying this material as food contact with food stimulants. And those food stimulants are mostly based on water, 10% ethanol, water 50% ethanol and 95% ethanol. You can see it's really easily that we're doing an extraction there. And it's not actually a footprint application. But if we do have those Products with the application. So you can see that the indication for those type of issues. Thank you and happy to answer any questions. If there is any question from the audience for Diana for me.

Thank you, Rafael. Thank you, Dan. Yeah, really great presentation. Really great insights into those important interactions that occur with these new materials and packaging. Yeah. If you have any questions, please post them in chat. I'm noticing a hand up in the audience from Gabra Hamadin. Apologies if I'm pronouncing your name wrong, but perhaps we can unmute your mic and you can ask a question. I'm just waiting for your microphone to be unmuted self muted. Okay, a question for me perhaps while we're waiting. I just wonder what implications. I mean you talked briefly about the implications around other material. I'm just wondering what implications the interactions of ethanol have on the permeation of other components through the material in terms of changing its crystallinity.

That's the thing that we want to answer with some of your new equipment and machine. Now try to understand now when we have more than one compound because as you can imagine, these simulants are very ideal situation.

But.

But in a real system, when you actually have ethanol, water and also other type of compounds, I think that the new equipment that you are developing is allowing us to get some of that information. When you have a mixed compound and it will depend. No, it depends. What if ethanol or ethanol, methanol, butanol, all those compounds will swell the polymer, they will crystallize a different degree and the permeation of the co Permian there will be different. And these are all the things that we don't know yet. So these are the things that we need to measure and understand. And actually this work came long time ago. Just basing on trying to develop a bottle application for this polymer and understanding that you want to put the wine 14% ethanol, that polymer will swell and the properties that you needed there will not be there.

Absolutely. Thank you. Okay, well, if there are no other questions, perhaps we can move on to Sunshine Nugudala. He's going to be talking about the best methods around and practices around the measurement of permeation through polysulfide membranes. So please go ahead.

Let me just go and share my screen. Can you see my screen at all?

Yep, we can see it. Yes.

Okay, prep it. Hi everyone. So my name is Sunshine. I'm a PhD student in the Department of Chemical Engineering at Imperial College. I came to talk a bit slightly from what the title is and focusing slightly more on PIMs. Which is what we work on also, so it doesn't overlap too much with what we've kind of seen already. Just a bit about what we do in our lab and why these sort of measurements are so relevant to what we do as well. And especially my background's actually in kind of organic chemistry. So I'll talk about why these measurements are so relevant to me as a chemist when I'm developing new polymers, as is quite well known. I'll just close this if I can.

Yeah.

As is quite well known, as population grows and CO2, atmospheric CO2 starts to increase, there are going to need to be kind of changes in the way we deal with this to reduce kind of atmospheric CO2 emissions. And of course, a very kind of mature technology is amine swabbing, which uses atrious amine solution to remove CO2 from gases and has been used since the 1930s. It's quite very well known technology. However, it has many disadvantages, such as the higher energy input. It requires a very large physical footprint. And so since then we have seen the emergence of polymer membranes being used for this high volume CO2 removal. And as with many advantages it has over kind of aiming scrubbing, such as the lower operating energy, lower maintenance cost and this kind of smaller physical footprint. This technology is, there's kind of a lot of work to do before it can be kind of widely adopted globally for CO2 arrival. So there are a lot of membrane materials. I will probably be focusing on kind of what we do in our lab. And what's quite important is that we have specifically for CO2 separation, a membrane has to meet several requirements, whether that be kind of chemical stability, thermal stability. If we look at kind of inorganic membranes, they do have this, but they have kind of a lot of poor reproducibility, they're quite brittle and generally more expensive than organic polymers. And this is what we really focus on. And this is because of course they're solution processable and a lot cheaper compared to inorganic materials. And if we look at some of the conventional gas separation membranes and those that kind of predominate in the industry, such as cellulose, acetate, polysulfone polyimides and so on, we can see that of course with most gas separation polymers, the performance is limited by this permeability selectivity trade off. But the polymers that are kind of leading this technology, I'd say they have relatively low CO2 permeability and kind of moderate CO2 and nitrogen selectivity. So there is a lot of room to kind of improve this Performance both on kind of the chemistry side and materials development and also kind of moving on to those working more in sort of an engineering perspective. One thing that's kind of really highlighted on is the need to improve CO2 permeance. And I guess with reading any literature on gas separation membranes, there's a lot of focus on Polaris developed by mtr and one of their. You'd see a lot with their kind of trends in what they've developed so far from first generation membranes to kind of their most recent. I think it's third generation. Although the selectivity is roughly similar. You see that there's an increase in permeance. And one thing that they really highlighted is the need to improve permeance so that we can. Well, in terms of reducing capture costs, the main way to do that is to improve the permanence of the material. So this is kind of why we work on polymers of intrinsic microporosity. So just wanted to. This is kind of a schematic we always use in kind of undergrad. But I think it's probably the best way to show it if and kind of when understanding what PIMs are and how they differ from some other polymers. So if we look, think about a lot of kind of polyethylene oxide based polymers, they're very flexible. So these chains can rearrange and they very efficiently. And that's similar to those with linear or rigid polymer chains. They have this kind of same effect where they do pack together quite efficiently, but with PIMs they are have a contorted backbone. And this means that there's. They don't pack efficiently and you have this space in between the polymer chains. And you will kind of say since 2004, when they were first established by Professor Neil McKeown and Peter Buddhims, they kind of follow this general structure where they have this site of contortion and this is what causes the intrinsic microprocessity and then leads on to kind of this excess free volume in the structure and therefore high permeability. And in general, and I'd say the kind of a great thing about PIMs is that they can be kind of, I'd say following these kind of design rules with a structural unit with a site of contortion and a linking group that kind of follow these rules. And it means that there is kind of a lot of structural diversity between different PIMs from canal PIMs to those with benzotrypticine units, BI units, ethanoanthracene units. You know, there's a lot kind of you can do with a pim. And there's I guess a slight relaxation in how we define what a PIM is. But if we actually look at the most recent upper bounds for various gas pairs, these PIMs do actually dominate kind of the upper bound itself, specifically the benzotrypsine pims. And this is because they have this very rigid kind of structural unit that I spoke about before. And because it's increasingly more rigid than some of the previously reported kind of previously developed PIMs, you get this excess free volume to kind of significantly higher. And then you get these ultra permeable polymers which as I said, kind of state of the art and have CO2 permeability as high as like 40,000 barrel. So I want to just kind of talk a bit, a little bit about what our research group does. So this, I'd say kind of a wide range of skills. So kind of focusing on polymer design, whether that's through modeling and kind of working with a few of us who are chemists and really focus on the synthesis of polymers. Looking at membrane fabrication and then the performance evaluation, which I'll kind of go into and just wanted to highlight that these are kind of things we go back through all the time. Maybe kind of the synthesis works, but then it doesn't form a membrane. So you're going back to the synthesis and kind of going between each of these points to speak. So one thing I wanted to highlight first, which is kind of the first thing on kind of that process is the polymer design and synthesis, which is what I kind of work on. And one really important thing as a chemist when we're deciding what polymers to kind of further develop, is looking at synthesis itself, reaction kinetics, the scale up and the general conditions that are required. So kind of avoiding things with multiple steps or that require columns, very expensive catalysts, for example. So these are all things that we need to consider when scaling up these sort of polymers. And now kind of initially we always start with like a 50 gram, 50ml flask. But now we're kind of reaching the kilogram scale in a lot of our polymers. So it's kind of also understanding how these reaction kinetics change as you go from grams to kilograms. And again, why this is so important to kind of think have a kind of intentional polymer design or have that in your mind when designing a lot of these polymers. If we look at a lot of the state of the art pims at the moment, as you can see, the synthesis is quite, it's multi step, it's quite Complex. And this does limit its scalability. So yeah, this was just to kind of highlight why it's so important to kind of think about if we do want to start to commercialize PIMs, thinking about the synthesis of them itself. So just to have a quick look at some of the sort of synthesis that we do, it's usually one step or one to two steps, and we still follow the same design rules. We have a structural unit with a site of contortion. We have a fluorinated benzaldehyde as our linking group. And then we get these PIMs that still kind of reach similar kind of beauty surface areas as traditional pims. And what's great is that they're synthesized from a lot of commercially available monomers. So it kind of, we don't need to do those multiple steps. And in that case it's relatively easy to scale up. And in general the synthesis is a lot cheaper. Unfortunately I can't show any structures, but kind of hopefully it makes makes sense. And as I said, another thing we do in our group is various different techniques for membrane fabrication. So usually we do start off with just solution casting using a petri dish. But once we've kind of been able to scale up the synthesis, we can move on to using a casting knife to develop these large area membranes collofibers as well that we've done in our group. But what I specifically focus on is mostly thin films, using both KISS coating methods and spin coating methods as well. And kind of looking at why that is so important. So not only from the thin films that we work on, as I said, they have multiple layers. So we've got our support layer, usually kind of a polyacrylonitrile support layer. We are developing some gutter layers, but the selective layers themselves are the PIMs that we are developing. And in general from again a cost perspective, because that selected layer is so thin, if you're going from like 200nm for the selective layers, whereas dense membranes are like maybe 100 micron, you are obviously using a lot less material compared to some traditional membrane configurations. So this is why we kind of want to focus on thin films specifically. And one thing that you might see in a lot of PIM literature, there are a lot of materials that although kind of have some non thickness, say some limitations in terms of their performance that aren't thickness dependent. And so what the aim is to do is to kind of transform them into thin films. And this kind of allows polymers with excellent theoretical intrinsical intrinsic performances and making them into Kind of a practical thin, ultra thin film. And one thing that I also want to highlight, of course there are multiple layers. So the optimization of each layer is really important. So for example, we are also working on this gutter layer that I said. And the reason for this, it provides kind of a flat uniform surface to allow us to then coat a ultra thin structure selective pin layer. And this goes on to boost permeance, which is, I guess the aim in, in all of our research is to boost this permeance. And it does this by kind of stopping the selective layer from penetrating the pores of the porous support. So it kind of sits quite flat on, on this gutter layer. And for this we use slight different materials. So PDMs, we do a cross linking reaction between acyl chloride groups and the amino propyl terminating end group on the pdms. So again, just wanted to highlight kind of the importance of choosing the right materials for each layer. And this is kind of what I was previously going to present a bit more on, but is what the testing method that we do in our lab, which is the constant volume variable pressure. And a few years ago we were involved in a multi lab study with 10 independent labs kind of collecting pure gas permeation data using polysulfon. So this is kind of what we still kind of use this method. It's kind of an in house built permeating rig and it kind of helps us to standard, kind of set a standard protocol and a list of best practices for sample preparation, equipment setup and in general permeation testing. So this is kind of still what we follow to this day. And I just wanted to show how we prepare our membranes. One of the kind of things that we that's highlighted in this work is the importance of the sample preparation. So we mask all our membranes and reduce the kind of testing area down and this reduces the chances of any defects. And the masking also helps to protect the film itself. And this is quite important, especially for thin films. We saw that when we were doing this masking method, our films would break during testing and five minutes would shoot right up to testing pressure. So yeah, we realized that kind of this is the quite a good way for us to test our polymers. And it seems to work quite well. And similarly again, we use the time lag method. And this is incredibly important for, let's say on the polymer design side because it helps us to determine diffusion permeation and solubility coefficients from a single dynamic gas permeation experiment. So going back to some of the polymers I Think Sean showed, which is this CPIM1 and the performance itself, there is a significant, significant selectivity improvement for various gas pairs. And this is driven by this diffusion selectivity. And this is kind of why it's so important for us to see what drives this increase in performance. And we're able to see this using the time lag method and kind of then looking at why this is so. This is because this polymer then starts to go tightening of the structure when it goes, undergoes this acid hydrolysis and these carboxyl groups are added on and you can, I think in this paper they're able to see with like PALS that a lot of these large free volume elements kind of collapse and then form these smaller free volume elements. So yeah, this is quite like so important to have these diffusion coefficients. Similarly, a different way for us to kind of enhance performance by increasing the diffusivity selectivity is also tailoring the backbone architecture of these polymers. So for this polymer itself, it was about increasing the molar content of tryptacine, which is one of the rigid sites of contortion. As you can see, the solubility coefficients are pretty stable throughout as you increase the trypsine molar content. But the diffusivity selectivity does increase. And this kind of verifies that there's a molecular sieving effect that is dominant and responsible for the improved gas separation performance in the tryptophane based films. And again, while this time lag method is so useful for us, it also helps us to kind of look at this solubility selectivity, which is a project that we're doing with Sinha Cell and kind of both collaborators. So this is kind of the enhancement of this solubility selectivity in PIMs, whether that's behind transport or gas separations. And this is by introducing various substituent functional groups. And yeah, we're able to kind of first see this through modeling the synthesis itself and then also performance, looking at the performance and comparing it to some traditional Pims. And so we see kind of compared to Pim 1, the polymers that we have developed, they fall into this target area for post combustion CO2 capture. And this is due to a solubility selectivity enhancement as a result of kind of some various functional groups that we've added into the polymer side chain. And I guess for one thing that can't be ignored for the future, and if we want to kind of commercialize a lot of these PIM membranes is some of the kind of the primary obstacles to their practical use for Example physical aging. And this is just the relaxation of the polymer chain over time. As we know, PIMs have this excess free volume. And so over time it reaches an equilibrium state and then this fraction free volume reduces significantly. And we can see this through kind of when we do permeation measurements and we see that this results in a pretty significant decrease in gas permeability and a very slight increase in selectivity. So this is very important for us to kind of track with pims. Of course, if they age, then the likelihood of them being used for kind of the likelihood of them being commercialized is pretty slim. And there are several ways to improve and kind of mitigate the effects of physical aging, whether that's blending with non porous materials or other microorganism materials. Post modification, there's lots that have been published and everyone seen it within kind of this, let's say industry is trying to kind of tackle. One of the most recent ones is the slight change in the topology of these polymers. And this is, and also the network content. And this can be kind of easily tailored through some of the kind of reaction parameters. And what the authors found was that one of the branched polymer topologies actually showed very little aging. So there is kind of promise to improving PIMs and kind of helping them to kind of say overcome this significant barrier to commercialization. Similarly, we are also working on some post synthetic modification of polymer structures, specifically PIM1. So we have made some range in between 150-200nm in films and we have been kind of tested them against PIM1. So over kind of the space of one month, PIM1, the CO2 permits of PIM1 decreases significantly from around almost 5000 GPU and it decreases well below 1000 GPU just over the span of 28 days. Compared to our modified PIM structure, we see that it's pretty stable so far and there is relatively no change in the permeance. And again this, this improvement in both the selectivity and is as a result of an improvement in the solubility selectivity. Again that we have very kind of useful to be able to see through our herniation measurements. And in terms of some future perspectives and from like a chemistry point of view, it's really important that we do these sort of measurements so that we can address the kind of main issues with pims and we can go back to the synthesis and see what we can do to improve the performance. And it's like a constant and forth looking at what we can do on our side to achieve certain targets. And I think one of the main things for on our side is exposing these membranes and polymers to additional contaminants. As I said in our lab we currently only do kind of pure gas measurements which give you very ideal.

Performance.

We also do kind of just water vapor sorption on its own, which is useful kind of for preliminary results. But the most important thing is that we start to move on to multi component analysis. So for example, with Semester we've been able to have get some preliminary results and looking at the CO2 permeability under the effect of humidity. And some of our chlorinated polymers do seem to be quite stable even at high relative humidities compared to CPIM1. And so yeah, just wanted to highlight why these sort of measurements are kind of so important for us as when designing new polymers. And I'd just like to acknowledge various collaborators and thank you for letting me present. If there are any questions, feel free to ask away. Thank you.

Thank you Sunshine for your talk. It was very nice to see the sort of synergy between your talk and John Zelinsky's talk earlier in Proceedings, especially around the measurement of diffusion and solubility and how the interactions of different species in the polymer affected the permeation. Yeah. If you do have any questions for Sunshine, please put them in the chat and we'll try and answer them as much as we can. Let's have a look. Okay, we've got a hand raised from Jerry King. Jerry, do you have a question you want to pose? I guess while we're waiting Sunshine, I have a question myself around fouling. So have you done any, I guess studies to look at changing the structure of these PIM materials to try and increase their anti fouling properties?

So that's actually something that were doing with somebody because initially a lot of these PIM materials were to be used for like marine based CO2 sensors. So they're having a look at some kind of effects of kind of biofouling and looking at how water affects these, the kind of sensing capabilities of these films. So waiting to get the results from that. But yeah, we are working on looking at the founding of these polymer thin films specifically.

Great to hear. Great to hear. Yeah. If there's any other questions, please post away and we'll get them answered for you. Okay. Jerry, you need to unmute yourself if you have a question that you want to ask Sunshine. Oh yeah, there's a general question from Jerry. So can turfing migration through polymer films be measured by the instrumentation you currently have One of the concerns is flavor scalping and vice versa.

Probably not with what we kind of use in our lab. I'd say not with kind of what we have currently at the moment. It's just kind of just pure single gas measurements and quite low pressure at the moment as well. So I'd say we would need to. Yeah, not with what we have at the moment. I guess there is scope to do that.

Yeah.

And another question from the audience. Are PIM membranes in general susceptible to plugging by humidity? Yes.

So this based on kind of what we're starting to see, it depends on the PIM itself and what kind of the PIM itself. So for example, the fluorinated PIMs, it seems that they don't suffer so much from this, but we do are working on.

A lot of.

So for example afm1c seen pim1, they do suffer from this. And yeah, we both see this in the results even as thin films, thick films. When we do expose the kind of few measurements we've done, they do suffer from this kind of plug in in humidity based off kind of a few, some. A few measurements that we've done so far. But yeah, it depends on the polymer itself because there's so much structural diversity, it's hard to kind of bunch them into one group. It really depends on the functional groups and also the backbone.

Okay, if there's any more questions, post away. If not, then of course you'll be able to contact Sunshine directly by email. And of course that applies to the rest of our speakers who have presented today. And I just want to thank them again for giving their talks and their time and of course for you in the audience for giving your time to listen and for asking your questions. Please join us again for our future webinars. If you have any questions, you can contact me at Social Management Systems. Our email address is on here as well for the general information. But yeah, we'll be presenting more upcoming webinars in the future. And yeah, please do join us again. I hope to see you soon.