Online Webinar: Recent Technological Advances in Adsorption Based Direct Air Capture

Thank you for joining us everyone. We will now be beginning with today's webinar session. So without further ado, I would like to hand over to today's host, Dr. Daniel J. Burnett, Vice President of Surface Measurement Systems North America. Over to you, Dan.

Thanks, John. Well, good morning, good afternoon, good evening, good day to everybody attending. Thank you for taking time out of your busy schedule to attend our webinar. Today we are pleased to have three speakers from the Chemical Engineering Department at the Imperial College London. Just to give you, the attendees, a little heads up, we're going to have three speakers in succession. I'll introduce all three speakers now. They will go through their presentations. This will be followed by a presentation by Dr. Paul Iacomi from Service Measurement Systems. And at the end, after all four speakers have presented, we will have a live Q and A roundtable open chat. However, in the meantime, you're welcome to submit questions through the panel. You can submit questions with the GoToMeeting box and we will go through those questions that have been submitted and then also open it up to open questions at the end. So if you want to ask that question at any time, feel free to do so. However, we will address and answer all those questions at the end of the four speakers today. So without further ado, the three speakers are going to be started by Ashlyn Lowe. She's in the final year a PhD student in the Chemical Engineering Department at London Imperial College. She's supervised by Professor Camille Petit and her research focuses on synthesis, characterization and evaluation of CO2 absorbance for direct air capture. Ashlyn is also a member of the International Adsorption Society Carbon Neutral Working Group and before joining Imperial College, she completed her Bachelor's Degree in Nanotechnology Engineering at the University of Waterloo in Canada. The second speaker will be Lucy Barton and a third year PhD student at the Department of Chemical Engineering at Imperial College London. She works in both groups of professor camille petit and Dr. Ronnie Pini. She completed her Bachelor's of Chemical Engineering with honors at Curtin University in Western Australia before working for four years in liquefied natural gas processing. Here she worked in facility life extension projects, maintenance engineering and then process control. Her current work relates to process modeling and design as well as experimental design of a unique process configuration for absorption based carbon capture applications. Our third and final speaker from Department of Chemo Engineering at Imperial College London is Hassan Azan. He joined the research groups of Dr. Ronnie Pini and Professor Camille Petit in 2019, having completed his MN degree in Chemical Engineering from Imperial College London. Also in 2019, his work relates to gas adsorption, equilibria kinetics and process evaluation for zeolytic materials for carbon capture applications, with a particular focus on pre combustion carbon capture. So with those introductions, we will turn the presentation to Ashlyn and she's going to start, followed by Lucy and Hassan, and then finally Dr. Paul Iacomi will finish with an SMS presentation. So I'll turn the attention over to you, Lucy, and go ahead and take the floor.

Yeah. So thank you everyone for attending. So today myself, Hassan and Lucy are going to be doing a joint presentation on the key technical, key technological considerations for the advancement of absorption based direct air capture. And we just introduced ourselves, so I'll skip that. Right, so just to give a bit of background. So most of us know that in 2015 there was the Paris Agreement that said we needed to limit global temperature rise to below 2 degrees Celsius and ideally below 1.5 degrees Celsius. And the IPCC has said that we need carbon dioxide removal technologies in order to achieve this goal. And, and so direct air capture using solid absorbance is one of these promising carbon dioxide removal technologies that is being researched today. So if.

Let me just share the pointer.

So if anyone's not really familiar with an absorption based act process, here is just a general schematic where we have air being fed into your vessel containing your solid absorbent that can selectively capture CO2 and remove it out from air. And then you can apply a purge, typically either heat or steam or vacuum or a combination of these to remove your CO2, which can then be stored or used further downstream. And your absorbent bed is regenerated and able to be used again for another cycle. So today there are several pilot scale absorption based DAC units that have already been commercialized. And in total these are capturing approximately 10 kilotons of CO2 per year. But, but the goal by 2050 is to capture and remove 2.7 gigatons of CO2 per year using engineered CDR technologies. So obviously there's still a lot of scale up that needs to happen and progress in a very short amount of time. So in terms of how we do this scaling. Okay, thanks. In terms of how we go about the scale up, before you go spending money to build all these new plants, you'd first want to do modeling and optimization in order to assess how different absorbents and process designs could work at scale. So if we look at the absorbent side of things, you need various absorbent properties in order to do process modeling. So if we want to describe the general process, this general process I showed previously, you need to know different information. So if we think about the, the vessel itself, well, you'd want to know how much solid, how much mass of absorbent is in your vessel. So that's knowing the bed density as well as the particle density of your material. And then as well you'd want to know how much gas is in your system. So that's like the total pore volume as well as the inter particle spacing. And for the adsorption process, well, you'd want to know how much of each gas is being absorbed. And not just CO2, but there's other gases that are in much higher concentrations in, in air, such as nitrogen, water, oxygen, etc. And so in order to implement the essential process models, you need the isother stem equations that describe the loading of each gas as a function of pressure. And then you'd also want to know how fast you know your, your absorption occurs because that, that could greatly affect impact stuff like your productivity of your process. So that's needing to know the mass transfer kinetics. And then lastly for the desorption process, well, you'd want to know how much heat is required to increase the temperature of your absorbent. So that's the specific heat capacity. You'd want to know how fast your absorbent can heat up. So that's its thermal conductivity. And lastly, how much energy does it take to desorb all the gases that has been captured. So that's the heats of absorption for your different components.

I'll just speak to this slide and then hand back to Ashlyn. So as part of scaling up, Ashlyn's just, just discussed the material properties that are important for process modeling. And a key aspect of scale up that has been mentioned by the International Energy Agency is providing plug in and modularized systems. So processes where these sorbents are actually capturing CH CO2, which are easy to install and have low capital intensity. And so as part of this, what we have seen in the literature is that there's little research on unique process configurations and contactors in carbon capture applications, including dac where these sorbents are actually used. And so as part of this, these type, you know, these process configurations also need to be investigated to improve the scale up or improve the path to scale up essentially. And so that's where the use of material properties in these process models, looking at these different configurations is very important to assess the feasibility of them. And so I'll pass back to Ashwin.

Now Yeah, so basically for the. So that was kind of our introduction to daq. And in terms of the rest of the presentation, we're basically breaking it down into three sections. So for myself, I'll be speaking on the material equilibrium absorption properties of two commercial resins for dac. Lucy will be discussing process design and modeling, and then lastly Hassan for kinetics absorption in shaped adsorbent. So yeah, I'll be starting. Yeah. Right, so in terms of again, back to adsorbance for DAC, since 2016, there's been over 200 absorbance reported in, in the literature. But the main issue is that a lot of the critical data, a lot of those key parameters that I discussed previously that's needed for process modeling is unavailable, not reported for most of these materials. So here we have a graph where on the Y axis is a list of the different parameters I discussed earlier. And then the X axis shows the number of adsorbents that these parameters are available for. And so what you can see is that there's a lot of missing information. And if you want to find an absorbent that has all of these properties available for it. Well, to the best of my knowledge, there's only two adsorbents so far that have all of these reported. So it's a big gap in the literature and yeah, a hurdle to trying to advance, trying to scale up and do process modeling for these materials. So in a recent study that we did that was just published last month, we basically wanted to measure all the material and equilibrium absorption properties that you need for process modeling for two adsorbents. And so the two materials that we looked at, one was Lutet Vpoc1065. So this is a commercially available amine functionalized polymeric resin. And it's basically regarded as one of the benchmarks in the literature for dac. It's one of the two adsorbens. I mentioned that all the currently has all the necessary properties needed for process modeling available, although there's still some missing parameters that would be interesting to investigate. And then the second material we looked at is Purelite A110. So this is another again commercially available amine functionalized polymeric resin, but had not yet been investigated for a deck and there was very little information about it available. And so we thought it would be an interesting comparison to see if we could potentially find another benchmark material through purelite. Now I did say that kinetics are also very important parameter to look at in terms of the our study and the journal we submitted to Kinetics was out of, kinetic measurements were out of the scope. But Hassan will be touching on that later in the presentation. And so in terms of our investigations for the material properties, we look into the chemical features of the materials, their textural properties and their densities and their thermal properties using these different techniques that I'll go through. And then for the equilibrium absorption side, we measured volumetric absorption isotherms not just for CO2 but also nitrogen, oxygen, argon and water. And also looked at doing isotherm fittings and calculating heats of absorption. So yeah, we'll be going through those results. So for the chemical features here we are basically just trying to confirm that the material is as we expected. So we confirmed through FTIR that indeed has the primary amine functionalization and the expected polystyrene backbone. And from elemental analysis we found that fearolite actually has a slightly higher nitrogen content compared to the benchmark glue it. And so it means that it potentially has more binding sites for CO2 compared to the benchmark. Because in these materials basically how it works is that two amine groups are needed to capture one CO2 molecule in dry conditions. So, so yeah, having a higher nitrogen content, potentially more CO2 loading in terms of the textural and density properties. So we did nitrogen absorption measurements at 77 Kelvin and we basically saw that these are macropore resins with expected low surface area. They have quite similar skeletal density. And the main difference is that Luitit has higher particle and bed density and correspondingly lower total pore volume and porosity compared to pure light. So what this means is that you in the same volume of your vessel you could fit more mass of luitic compared to pure light, which which again could affect how much you the how much CO2 you can absorb in your process. But I'll discuss that a bit further down and yeah, I'll just note that all of these measurements were done on the dry sorbent. So these materials are basically shipped with like 50% water content. But all of the densities we measured was from the gas material. And so then for the thermal properties we looked at its thermal stability. So again you can see that both materials lose quite a lot of water. But then it's, the material itself is relatively pretty stable up to almost at 120 degrees it's still stable in both nitrogen and air. And 120 degrees is already on the upper end of temperatures that would be used for desorption process. So it would be stable for those conditions. And then on the right we measured the specific heat capacity. And we see that they're both pretty similar, which is what we expect since their chemical features are basically, basically the same. And we just compared their results to that of polystyrene from the literature, just to, as a quick sanity check to make sure these results make sense, which they appeared to do. Yeah. So then in terms of the equilibrium absorption investigations, so here we, I'm showing CO2 isotherms for Lewitt and pure light as loading as a function of pressure shown in the log scale. And so with a, you can see here, I've highlighted 0.4 millibar, which is equivalent to the concentration of CO2 you'd find for direct air capture. And we've measured isotherms at seven different temperatures going from 15 degrees up to 120 degrees Celsius. And so if we start comparing the absorption capacities, theoretically, if we assume that CO2 capture is only happening by two nitrogens capturing one CO2 molecule, theoretically pure light should have a, should capture more CO2 just based on the nitrogen content that I showed earlier. And we see that that's also what also the case experimentally. So the gravimetric CO2 absorption capacity, so in other words, the amount of CO2 absorbed per gram of adsorbent purolite has a higher gravimetric capacity of 1.3 versus 1 millimole per gram compared to Lua tip. And then in terms of the volumetric absorption capacity, which is the amount of CO2 absorbed per volume of your bed, even though Lua, as I said before, had a higher bed density overall, when you account for the, when you, even when you account for that pure light still has a higher volumetric absorption capacity, which means, you know, you could use basically a smaller footprint of your facility to Capture more, more CO2. So that could have cost benefits there.

Yeah, right.

So now to actually implement all these isotherm information into, into your process models, they need to be accurately represented with isotherm equations. So from the literature in Luata has always been represented with the temperature dependent toss isotherm model which we shown here in the dashed lines in both graphs. But as you can see, it doesn't really represent our data well, especially at the lower temperatures. The reason for this, you know, why it was okay for other studies in literature, not ours, we're still not fully sure. You know, it could be because where we've measured quite a wide range of temperatures compared to other studies and also have measured a lot more points in the low pressure range. But the point is that we just need an equation that accurately represents our data so that it can be inputted into models. So we proposed a new isotherm model which we called the chemi physisorption model shown here in the solid lines. And you can see that the fitting is a lot better. And so basically here we've basically introduced a new model that accounts takes into account that it's not only just chem absorption that could be happening in your material, but also there could be a significant contribution from phys absorption because you need, you need, your CO2 molecules need enough activation energy to participate in chem absorption. And at the lower temperatures, not all the molecules might necessarily have enough energy to do so it seems especially at the lower temperatures there is a significant contribution from physisorption as well. And we think that's why the fitting does seem to be a lot better. So. So, yeah, all right. And so then we also measured nitrogen isotherms at three different temperatures shown here in the squares, as well as argon and oxygen isotherms shown in the diamonds and triangles. And you can, the main takeaway here is that the nitrogen, argon and oxygen isothermic loading is a lot, a lot lower compared to CO2 and I think almost on the level of, of negligible basically compared to CO2 which is ideal for both materials. And we also measured the water isis water loading. Yeah, again at three different temperatures. And we, we fit it to a known, a known isotherm model from the literature. And basically both residents display similar behavior of unrestricted multi layer water absorption. And then lastly we also calculated the heat of absorption for CO2, nitrogen and water. So you can see that purolite has slightly lower CO2 heat of absorption compared to Luitip. But it's pretty similar for, between both for nitrogen and water. Yeah. So with the lower CO2 heat of absorption it potentially could take less energy to remove the CO2 and again that could be beneficial on the cost side of running the process. Yeah. So basically to summarize what we've done, so this was the table I showed at the beginning with all the data available in the literature. And after our work the table can now look like this. So we basically filled in almost every gap for the different parameters except for thermal conductivity, which we haven't been able found the facilities to measure yet unfortunately. But yeah, we've been able to address everything else. So in summary, compared to lunatic, purolite has pretty similar physical chemical properties. It has both higher gravimetric and volumetric CO2 absorption capacity and a lower CO2 and similar nitrogen and water heat of absorption. So overall, based on these results, we think purelight could be a suitable absorbent for deck. And these parameters are now available to be implemented in process models to determine key performance indicators and assess how this material might do at scale. Yeah, so that's it for my section. I'll hand over to Liz.

Thank you, Hassan, Lucy and Ashlyn, for your presentations today. As mentioned earlier, we're going to hold off any questions and answers until the end. So I'm going to turn it over to Dr. Paul Iacomi from Surface Measurement Systems, and he's going to take the floor for the next several minutes and then we'll open it up for questions. If you have anything, please continue to enter questions. We'll ask them at the end if you can't hang on, but you can enter them into the question and answer box in the GoToWebinar panel. So go ahead, Paul.

All right, well, thank you everybody for joining me. So we've heard a lot of great details about directory capture from our three previous speakers, and that was good because it meant that I could remove most of my introductory slides. I think we all know here, we all know what we're here for, and we've heard a lot of details about measuring key performance indicators on different materials, various process configurations and modeling, particularly the importance of kinetics. But I'd like to focus on one of the other challenges which was identified, I think, pertinently by all our previous speakers. Humidity. Sir? Water. Why is water a challenge in direct air capture? So first it has really strong interactions. It's a liquid at room temperature, while CO2 condenses at minus ATC. The enthalpy of vaporization of water is high. So through a combination of dipole moment and hydrogen bonding, it stands at around 44 kilojoules per mole. What that means is that for absorption in porous materials, which rely Mostly on physisorption, for example, in the group of Baron Schmidt at EPFL coined the term of adsorberce, where CO2 adsorbs onto porous materials. It also is generally ideal for water. So the two molecules compete quite heavily for interactions in physisorption, although certain configurations of aromatic rings have shown to be more desirable in this case. Similarly, through hydrogen bonding or charged interaction, it can be very easy for water to interact with pretty much anything on the surfaces or functionalizations. When it comes to adsorption approaches, the combination of a hydrophobic material and idealized van der waal interactions for CO2 absorption is required. Chemistorbents are less impacted by the affinity as chemistry can be made a bit more specific. But nevertheless, water is a really good solvent. So in the first instance that means it has the ability to dissolve various parts of the sorbent itself. For example, here we can take the case of carbonation based CO2 capture processes, where an alkali or an alkali earth metal oxide reacts with CO2 to form carbonates, which is a basis of the process of calcium looping or enhanced weathering or cement curing. Water absorbed on the crystal surface and in the pores facilitates the carbonation process by dissolving some of the sorbet ions and some of the CO2. So allows a kind of liquid phase reaction at the interface and speeds up kinetics. But it is also the case that water can dissolve itself in the material. Many polymers, which are the backbones of amines, take up water and swell when water is present. And this plays a big role in not only the supports, but also in case of membrane based approaches where the membranes can swell or the pores can fill. Also, water is chemically active, it can act as a weak acid or weak base can also coordinate. And generally it's quite highly reactive. Maybe not as reactive as other materials, other compounds, but still particularly metal. Organic frameworks with open metal sites or like copper paddle wheels are particularly impacted by coordination with water, so it will destroy the base material. In the case of amines, there is a change in chemistry depending on the primary secondary, tertiary. I mean, the chemistry slightly changes, but it can act as an aid in CO2 capture. Even if there is no interaction between water and CO2 on the material used for CO2 capture, it can simply be a thermal sink. This is more relevant for temperature swing processes, but also a little bit for pressure swing. If it absorbs, that means it changes the required thermal load of the system. So, so you have a material which takes up 15% water, you need to put in the heat to dissolve that water and it also changes the heat capacity of the support. So that's something that needs to be taken into account. The thing is, these can be true for many other compounds, but water is something that's very common. Air is mainly oxygen and nitrogen, but the third most common would be water. So up to 4% by volume at 25C and up to much higher concentrations as temperature goes up. But also because it is process useful, it's also a challenge. Turns out water, or more specifically steam, is very commonly used as a process fluid. And one of the fastest ways of heating to temperatures to generate material is by using steam, sort of like a steam or moisture swing absorption or moisture vacuum swing absorption is one of the most commonly used methods. So not only is something that you might find in a directory capture process, but something that you might introduce yourself. So this is mainly why water is a challenge in directory capture. And there have been a number of perspectives out there and a number of research interests that highlight this as a potential interference or as a challenge. So why is this a current challenge? Partly it's also because of a disconnect between materials and process research. So what chemists might do in the lab. I'm a chemical engineer myself, but I worked with metal organic frameworks. So what our key performance indicators are when when synthesizing and what process the process actual requirements are can be a little bit disconnected. But I also think it's a challenge in instrumentation. So there are not many dedicated devices out there which you can get the kind of data easily. So water absorption data, co absorption data, regeneration data with humidity, etc. And water was at the heart of SMS since the company started 30 years ago. So we were eager to see how we can apply our DNA to this challenging problem. And we now have a number of gravimetric and chromatographic devices for sorbents in powders, monoliths, membranes, high throughput, et cetera. And what I'd like to go in the remainder of the presentation is just go into a couple of examples and case studies to see what happens when water and CO2 sorbents mix. Okay, so the first example is fairly straightforward. This is a metal organic framework, MIL101 chromium. In the literature it's specified as a material that's resistant to water and Indeed cycling at 25C can be cycled. The absorption of water can be cycled for many cycles at 25C. But what you might have in a regeneration process in a carbon capture would be using a humid stream for regeneration. This is the uptake of the material as it is being cycled at a constant water loading. So this is 5.6 Kpa of water vapor loaded in the air and then the material is cycled up to 140C. And then when it goes back to ambient temperatures, what we see is that the uptake starts decreasing. So materials which are considered to be resistant in terms of water might not. Be. So when they're exposed to both water and high temperatures. Another prototypical material, excuse me there a bit of technical trouble, are amine functionalized substrates. And in here, as was mentioned before, the. The Presence of water can act to change the chemistry of the amine, sorption, CO2 sorption, chemisorption, to essentially add more CO2 to allow the material to take up more CO2 into the process. So here is an experiment where we first expose the material to 50% relative humidity. It takes up about 3.3% in weight. We expose it to CO2 at DAC levels of 400 ppm and it takes around 1.1%. Fairly slow kinetics here, but otherwise good uptake. And when we expose it to both components together, what we see is that it takes up more than the individual sum of its parts. So this is how the enhancement effect can be quantified. Now, of course we cannot say for certain which one is absorbed onto the surface, whether it's water or CO2, but there has to be at least some CO2 in that extra uptake. Another example here is the carbonation of lime for calcium oxide. This is a main component of hydraulic and non hydraulic cements. And it has been considered as a basis for calcium looping process where it reacts with CO2 from the air and then it is subsequently regenerated at high temperatures to to get back to the calcium oxide. So in the presence of water it forms calcium hydroxide, in the presence of CO2 it forms calcium carbonate. These are the main reactions and it is quite a strongly exothermic process. So that means it is very selective for CO2, even at low concentrations, like in the air. But what we wanted to look at is how the presence of humidity changes this process. And what we have here is a single experiment where we started with calcium oxide and then went all the way to calcium carbonate with various cycling and various subsequent experiments to quantify things like sorption of water, kinetics of transformations, etc. I'm going to go through this step by step because it's quite a complicated one step experiment. So in the first process we have just calcium oxide finely divided. It is loaded in a gravimetric device and we start exposing it to different concentrations of humidity. The humidity absorbs quite quickly on the surface and it starts to convert the calcium oxide to calcium hydroxide. What we see is that the mass uptake changes, the rate of mass update changes as more humidity is present. Which makes sense because there is more humidity in the air, there is more to react. And we see the rate of reaction increases as humidity goes up. It then afterwards stabilizes. We can start heating the material and remove any excess humidity that is absorbed on the surface of the crystals and we end up with calcium hydroxide. And if we look at the stoichiometry of this reaction, it turns out that the conversion rate of calcium oxide to calcium hydroxide is pretty much complete. It's at 99%. We then run a simple water absorption isotherm. So we want to look at how much water does this material take up. And this is just looking at increasing humidity. We see the mass increase, but as we decrease the humidity, we see this is fully reversible. So there is no such thing as more reaction happening. It's only surface sorption. And we can quantify that. It's about 2.5% water uptake in the case of this transformed calcium hydroxide. And then we introduce humidity and CO2 at 400 ppm, this is multiplied by 10. So this is actually 400 ppm CO2. And we see what happens with the particles. We see uptake of water on the surface. And when CO2 is introduced, we see a start of a reaction. We see the increase here, and we know that this corresponds to a carbonation process. So first the water covers the particles and then CO2 starts carbonating the particles. And what we see in this particular case is that the kinetics of uptake don't really change with humidity, because it might be the case that this is a kind of surface process where it only depends on the surface area that is exposed from the particles. This was a commercially sourced lime sample, and other people have observed out there that the kinetics of sorption of carbonation do depend on the humidity. The transformation is also confirmed in situ using Raman. And we see here that at least from the surface particle, from the surface of the particle, we see a transformation of calcium oxides in characteristic bands here to calcium carbonate at the end. If we look at the stoichiometry, the carbonation extent was only about 8.5%, which also reinforces the idea that this was a surface process. It's a kind of formation of a core shell here, where we have a core of calcium carbonate, and still inside is a calcium hydroxide unreactive. So this particular particle might benefit from a particle size optimization. But just to summarize, the effect of water water absorbs on the. On the crystal faces and in the porosity, it does change the kinetics of the reaction when it comes to the transformation to calcium hydroxide, but in this case didn't have an effect when it comes to the transformation to calcium carbonate. And this is just only one experiment which I've shown here. There are other Parameters to investigate the temperature. This was only at 25, but director capture can happen anywhere from minus 20 to 40C. So these are also important variables to change. I mentioned optimizations of the particle size and morphology and the extent of reaction and carbonation could also be further investigated. Move on to another sort of prototypical material for carbon capture, zeolite 13x. This material was also shown in Hasan's study. It is very prototypical because there is a lot of data out there where we can compare pure component uptakes, kinetics, multi component uptakes, etc. It's a little bit of information about materials here. Essentially it's forgesite aluminosilicate zeolite with sodium counterions. It's already used in gas separation applications out there, which means it's commercial, very easy to obtain. And in this particular impact of. In this particular study of the impact of water we looked at most more post combustion conditions. So higher concentrations of CO2. Purely recording single component isotherms on 13x show that it has a high uptake for CO2. That's great, that's one of the things that we want. It has what seems to be a good working capacity from 0 to 15%. But looking at the water isotherm it seems that the interactions with water are also very high. So we can reasonably assume here that water would displace CO2 and would take preference in Co absorption conditions. If we record a second isotherm with a background of just 5% relative humidity, we can see based off the wet mass, we only see a 0.3% increase in uptake. So if we're looking at just pure, if we're looking at just the weight, the uptake we see, we think that while the uptake decreases by around 10 times, we think that the total uptake of CO2 is around 0.3. But if we look closely at the kinetics of this isotherm, so this is the same isotherm that was shown on the previous slide, but with a bit of kinetics information. So here is the prehumidification with 5% water and then we look into the dosing of CO2 and what we see is yes, there is an increase, but there is also a subsequent decrease. What this means is that there is some competition and desorption of water, replacement of adsorbed water with CO2. So while if we look at just equilibration conditions, we would think that the CO2 uptake is massively reduced, it might not be as massively reduced. The only way to look at this is to use to use techniques which can measure multi component absorption such as breakthrough. And here we look at a breakthrough of CO2 and humidity on 30. Next we see the breakthrough of CO2. First kinetics is quite fast and then as water makes its way through the column, we see that there is a displacement of CO2. The concentration at the outlet is more than the concentration at the inlet. That means that there is a displacement of CO2 by water as it progresses through the column. Once we stop the experiment here, we can integrate both curves and we can determine the amount absorbed of both CO2 and water which turn out to be around 0.5 by weight CO2 and 23.4% by weight of water. We also see that the data to Data variability in CO2 breakthrough is quite high. This is often a problem in adsorbance where one in multi component systems with Breakthrough where one adsorbent breaks through much faster than the other. Here there are actually three experiments overlaying one on top of the other. There are almost no visible differences. But when we do the integration it turns out that the variability exists purely because of the error of the sensor and the integration that's cumulative here as the experiment progresses over 300 minutes. So a nice way of obtaining accurate multi component data in these kind of systems which was initially suggested by the group of Arvind Projendran at University of Alberta, is to measure at the same time a graph metric experiment with the same conditions. We take the total from the gravimetric experiment. We take the water from the breakthrough experiment where we have no uncertainty or almost no uncertainty. And we get the CO2 uptake by subtracting the 2. And just as an overview of what actually happens when water and CO2 are competing for the same absorption sites in 13x we have CO2 single component uptake at 3.5%, around 7.6 by weight, 24.4 by weight when just water is absorbed. This is for single component data. And when we have both together we have about 10 times lower CO2 but still 3 times more than what was initially estimated by using just relimetry. So often a combination of techniques are needed to really assess multi component updates. I'd also like to go into a case study on MOF for post combustion capture provided by our industry partners novamov. This was a promising material for post combustion capture scenarios and wanted to investigate also its water stability and CO sorption performance. Looking at pure component data, this looks like great material. We have a large uptake we have large uptake also at post combustion conditions we have a high enthalpy of absorption. So it's an optimal range, around 40 to 50 kilojoules per mole. Not too high that they would introduce regeneration penalties and high enough that it will still be selected for nitrogen over for. For CO2 over nitrogen and oxygen. We also see one very nice thing is that the isotherms are completely irreversible. So in theory this material would not even need an activation, just a kind of pressure swing or mild activation conditions. Just switching one stream to another would completely clear CO2 from the absorption. Looking at the cycles of CO2 absorption, up to 10 cycles here, perfect regeneration, no degradation in time. Looking at the kinetics, 90% of sorption and desorption uptake takes place in less than five minutes. So very nice, very very fast uptake as well. So overall a really nice and promising material. We then wanted to also look at water absorption and we tested it at two temperatures. We see here two isotherms and the associated kinetics. We tested water absorption at 25C and water absorption at 40C. And what we saw in the 20C experiment was this large uptake commencing around about around 70% RH and then a subsequent loss in mass when it went up to 20 to 40C. We see the same thing happening a little bit earlier at around 50% RH. Investigating the sample after the experiment see it sort of bubbled here. So it looks like the material is starting to degrade at higher humidities. If we look at just purely the sorption characteristics, it is hydrophobic. The enthalpy absorption is lower than 44 kilojoules per mole. So it is still a nice material from the point of view of hydrophobicity. It's just when high humidity is introduced that it begins to collapse. So I wanted to also test this impact on CO2 sorption. So we devised experiments, sequential experiments, where we introduced first CO2 and then water at different humidity with 0% relative humidity. As we've seen also before, it's perfectly reversible in seven cycles here. We then introduce 25% humidity over here and we see almost no impact. Maybe a slight loss of uptake at the regeneration, but overall total uptake stays the same. Same thing at 50% RH. When we introduce 75% RH we see this increase in the kinetics and we also see a loss of uptake, maybe around 40% in CO2 and also we see an increase in the time constant. So it takes longer after exposure to 75% RH. And finally when we expose it to over 90% RH, we see this large kind of condensation uptake step. And then afterwards, when CO2 is being introduced, there seems to be Almost no more CO2 absorption. But can we still make use of this material? First we wanted to see what would happen if we have CO2 as a carrier gas, so we have CO2 as a carrier gas instead of nitrogen. And then we introduce humidity on top. So this experiment was devised like this. This is the kinetics, these are the resulting isotherms. So to go through each step we have first nitrogen. The material is then exposed to pure CO2 and we see this uptake here, 18% like before. And then humidity is introduced. And what we see is the isotherm that is being recorded still appears to include a condensation step later down here. So the material still seems to be decomposing at high humidity. On the other hand, the decomposition seems to happen at higher humidities compared to before. So instead of around 70% it looks around 85%. And one also wanted to see what happens if we keep the humidity below the material decomposition. Where we did a kind of step by step experiment to look at uptakes. We see here in the first step we, we see 40%. So this is below the threshold that was identified. We see this uptake and then we introduce CO2 and we see another uptake. If we look at the pure component uptakes at these levels, 40% and 10% CO2, 6.7% weight of CO2 if it's pure, 6.1% by weight of water at 40% relative humidity. And when two are combined, it almost looks like you add these two together. So the material seems to be very promising for taking up CO2 even in the presence of water. And there seems to be no interference of the two below the threshold for material decomposition. So to summarize this section, great CO2 uptake data, easy regeneration, fast kinetics, good overall key performance indicates from this point, including hydrophobicity. The only downside identified on this material is that collapses at high humidity, but it still affords near full CO2 uptake below this threshold. So can still be a nice choice for some processes where humidity is not as high. And finally, just before finalizing a very quick overview on a different kind of material on membrane permeability. So now we're moving away from equilibrium sorption, we're looking at membranes for CO2 capture. These are PIM1 membranes, so polymers of intrinsic mesoporosity. This Is work Spearheaded by Sean McIntyre from SMS in collaboration with Peter Budd's group at Manchester University. Essentially, these are polymers which are intrinsically microporous and have been shown to have great CO2 permeability and selectivity. In terms of CO2 nitrogen. If we look at the original material, PIM1, here we have high CO2 permeability and fairly nice CO2 nitrogen selectivity. If we look at also a similar material which has been slightly modified here, a carboxylated PIM1C PIM1, we have lower permeability but higher selectivity. So in terms of the material, this could be a nice material to increase the selectivity of CO2 nitrogen separations. But it was also observed from contact angle measurements that it is slightly more hydrophilic. The introduction of these carboxylic groups here increased slightly its hydrophilicity. So, in a real world process, because water is present, how does water affect the permeability of CO2? We tested it in a membrane permeability setup here. So it's sort of like a cross flow configuration. I won't go into detail in the interest of time, but I just want to quickly show some interesting results. In both materials, for pure CO2 permeability we have a negative permeability constant. So this follows a sort of anti arenas behavior for the activation energy, which means that CO2 permeability is favoured. The transport through the membrane is favored. When we look at pure water permeability as a function of relative humidity, we see a sort of different behavior in the case of the two polymers. In the case of regular PIM1, permeability of water decreases as humidity is increased, while in CPIM1 the permeability actually starts to increase as humidity is being introduced. Has some theory of why this is the case. In particular with the hydrophobicity and sort of like the swelling of the membrane in the case of the CPIM one. But probably the more interesting experiment is when both are being introduced. So this is a constant concentration of CO2 and increasing humidity. We're measuring the permeability of CO2 and it turns out that yes, water impacts the permeability of CO2 in both PIM1 and CPIM1. But while in the case of PIM1 the permeability loss is about 20%, in the CPIM1 it's about 40% when we go to around 60% relative humidity. So even though in theory this functionality that was introduced in the material improves membrane permeation and selectivity for CO2 nitrogen separations. Because in the Real world water is present. This kind of functionalization did not really improve the selectivity when water is being added. So testing in these kind of real world conditions is a must. Okay. And just to bring everything back in terms of conclusions of the various case studies that I talked about today, first, water is obviously one of the most important challenges in CO2 capture, and I hope I've shown that and I've demonstrated the kind of various behavior materials can have in the presence of water. So water can be a material's biggest enemy, but also its closest friend, depending on the configuration. And really a combination of techniques are required to fully assess a material. One CO2 isotherm and one water isotherm is really not enough. And I want to conclude this by a recommendation. So just a friendly reminder that you also need hydration. So the recommendation is that you drink about 2 liters of water a day, but besides your own water intake, you should consider the hydration of materials that you're working on. And with that, I'd like to thank everybody from the SMS team and our external collaborators, particularly the team here at sms. Thank you for listening and thanks again.

Thanks, Paul, and thank you to all of our speakers. We do have a few questions that have come in, if you don't mind. We'll go through those quickly. I mean, just there was a question, you know, just as a housekeeping item, that somebody missed the first part of the presentation. Their presentation is being recorded and I believe a link will be sent out to the attendees, you know, after the webinar is completed. So SMS will follow up with that information to you. So Yero Wang, although they've left the presentation, was asking, I believe this was during Lucy's presentation. Just a general question on what. What type of software was used to make all your graphs? So apparently they were impressed with the graph. So I think it was again, during.

Yeah, the modeling's in Matlab, but I use Origin for the graphs.

Okay, thank you.

Question for Ashlyn. This was by Audrey Bezrukov. They ask thanks for the great presentation. You showed only adsorption branches of CO2 isotherms for chemizorbants. Is there any significant hysteresis in those sorption isotherms and how would it affect DAC process?

Yeah, so we did measure the desorption isotherms as well, just to. So that the graphs didn't look too confusing. I just, I just showed the adsorption. Can I remember the graphs at the top of my head? So there was some hysteresis And I think they were.

Can't remember. It was more.

I think they were more significant at the lower temperatures. I'd have to tell. It's all in the paper that we put and I think we commented on the effect of the hysteresis there. Sorry, I haven't actually looked at it in a little while, so I won't say.

Was there a link to the reference in your paper in the presentation?

Yeah, there was. It's all in the bottom left corner.

Okay, so then when we send it out, they can find the. The details there.

Yeah, yeah, sorry about that.

Thank you. We have a question from Greg Sanko. They ask for the Liu Tit and Perlite. Do you see oxidation effects at 120 degrees F and above in air? So I guess it's a question for Ashlyn. Correct?

Yeah. So we didn't look too much at. So we didn't measure. The only oxygen isotherm we measured was at 25. I didn't point it out, but again, it was discussed in the paper. But for paralyte, if you look closely at the graph, you'll see that the oxygen isotherm doesn't. When we have the full adsorption, desorption, it doesn't come back to zero at the intersection. So we think that shows there's like a bit of oxidative reaction there. But for Lewis, it comes back through the intersection. The other thing we did was that for the paper, we measured a CO2 isotherm, basically repeated it after one year. So, like the same material, same batch from the same container, measured one year apart. And we saw for Purolite, the CO2 capacity actually decreased slightly compared to Liu. So in terms of like, storage wise, it seems like pure light might be a bit more susceptible to oxidative degradation. We didn't do any more thorough studies other than that, just because there's been quite a lot of work in the literature on oxidative degradation on amines, and some papers have even broken down, like the percentage of degradation associated to storage to high oxidation concentration and temperature, high oxygen concentration, ambient temperature. So, yeah, again, all referenced in the paper, but yeah, that's about the extent that we did for that. Thanks.

Ashton, Can I also step in here just for a quick second? In terms of the perlite material, do you know the structure of it? So what is the difference in terms of the structure and amine concentration backbone between that?

So on paper, like from the. The product specification sheets, it basically looks the exact same. So they're, they're both beads. It's both styrene divinyl benzene backbone with primary amines. We have a more thorough table in the paper that puts in all other differences. Basically, there's really not much of a difference other than when we check the elemental analysis, there was a slightly higher nitrogen content, so like 10.2 versus 8.3 or so. But yeah, and it also has, when it shipped, it seemed to have slightly higher water content. So it was about 60% loss in the TGA compared to about 50 for Lewathid. So those were the main differences we saw. But again, all our measurements we did were on dry conditions. So we desorbed all the water that was originally in the material just to have the same baseline, I guess.

Yeah.

Thank you. We have another question from Prabhu. I apologize, I'm sure I mispronounced that name horribly, but they ask which MOF is used from the novo for the CO2 adsorption.

That is proprietary material. So unfortunately I can't go into details.

Okay, I believe there's another one for you, Paul. This is from Huifang Yang. They asked what is the stability difference between PM1 and CPM1?

That's a good question. That's one thing that we're looking into at the moment. Essentially, it's quite well known that PIM1 undergoes some shrinkage and some sort of changes as time passes. The shrinkage itself is not in the physical term, very high, but it just means that the pore size changes by enough that the permeability of CO2 changes massively. So this is something that we're looking into and we hope to detail a little bit more further. But there is a lot of data of time changes of the polymer of PIM1 already in literature.

All right, thank you. We have another question for Paul and then I'll jump to Hassan. This is from Katia Pesina. They ask, when you looked at amine functionalized adsorption, did you notice if the kinetics change for the CO adsorption experiments?

That's a good question. The experiments that were done in the graphs that I showed were sequential, so not CO absorption. Introduction, as in both at the same time. So it's not something that I've noticed in that experiment, but it might change the kinetics due to the change in chemistry. So that's definitely something to look into. Thanks for the suggestion.

Thank you. Hassan. This is from Sean McIntyre. They ask, when looking at temperature effects on the micro and macro pore diffusion via zlc, how are you accounting for the heat effects from from absorption?

Yeah, so yeah, that's a good question. So one of the reasons that we reduce the size of the sample used, as opposed to using a whole pellet, is to reduce the heat effects in the experiment itself. So we use, we prior to these experiments, we calculate the. The dimensionless parameter that is typically used in ZLC experiments that determines whether or not you are in isothermal or non isothermal conditions for all of these materials. So it's an experimental check that we do that essentially tells you that under these conditions, with this sample size, flow rate and this material and concentration, you're going to be under isothermal operation. Other than that, in terms of monitoring, we basically have a thermocouple that is essentially within the adsorption cell and around a millimeter away from the adsorbent that measures the temperature very. It's Fairly sensitive to 0.001 Kelvin in terms of measurement accuracy, the temperature. So we are able to measure the temperature of the bulk gas during the course of adsorption and desorption. And we see that during the desorption experiments it does not change significantly to the point where it would affect the resource in any way.

All right, thank you. Question for Paul from Jan Steckel. They asked what is the type of detector that is used in the BTA instrument?

There are a number of detectors that are available in the Breakthrough Analyzer. There's obviously, of course, the capability of coupling it with a mass spec. But the design philosophy behind our particular brand of breakthrough analysis was that it would be equipped with sort of individual detectors that allow you to do 80% of the experiments that you might want to do. So there's an NDIR detector for CO2, there's a capacitive detector for relative humidity, and there's photoionization detectors for various volatile organic compounds at different ranges. We and to top it all off at the end, there's a thermal conductivity detector which sort of gives you an extra flexibility in gases or binary mixtures that are not detected by anything else in the tray.

Thanks, Paul. We have several questions here. We'll go through them one at a time. These are all submitted by Yabir Shabanian. The first couple of questions. First they. Thank you for the great presentations. First couple questions for Ashlyn, who was our first presenter. Question number one is which size of powders did you use?

So the Luita and Purolite are both manufactured as beads and they're about like one millimeter in diameter. Again, if you look up this product specification sheet, they give the range and what percentage falls in that range? Yeah, by about a millimeter.

Okay, second question for you. Isotherm. Under which CO2 concentration in the inlet gas, DAC or pure CO2?

So all our isotherm measurements, we're done with pure CO2, but, you know, we. It goes from zero to one bar. So we're going through the range of pressures. And yeah, we basically had like 20 points below 0.01 bar. So we really focus on the low pressure region. Yeah, thank you.

Another question for you, Ashlyn. With which unit did you measure the absorption capacity?

So all the isotherms were measured on a Micromeritics 3 Flux Porosity Analyzer.

Okay. They also asked, do you have an iga? No, that's from Haydn. Second. Thank you. Second presenter. So these are questions for Lucy. How do you prevent gas leakage between adsorption and regeneration sections in the rotary absorber?

Yeah, so there's different sealing mechanisms, so you might not be able to completely prevent any leakage. And there will be some mixing zones. So we've selected a special. A certain type of seal called brush seals, basically in our experimental setup, which we're still commissioning. But I don't know what seals like, span to use, for example, because, yeah, that would be proprietary for them. But, yeah, we use. We use brush seals where you can have motion against them, essentially. But I couldn't provide you information at this stage on how effective they are. And it might be that there would be some mixing, but like in a very large system, it might be that that mixing as a proportion of the total size of your absorber is. Okay, so that's something that still needs to be understood better at scale. Yeah.

Thanks.

Lucy, another question for you. In which form did you have your absorbance in the rotary unit? Beads.

Yeah. So in the simulations that I showed in this presentation, they were beads. However, the model was also able to do structured laminates, like sinusoidal laminates. So the modeling I've showed is beads. But it's. The model that we have is not limited to beads.

Thank you. Now we have a question from the same person for Hassan. Which of bound or binderless shaped absorbents work better based on your study? So which is a bound or boundless shaped absorbance work better based on your study?

Yeah, so that's a good question. So that's really the avenue that we want to explore moving forward. So the background for my project is basically pre combustion capture, which is why we wanted to go with Y Zolites. And for the work that we're doing. We wanted to completely focus on binderless pellets. But in addition to that we've also looked at bound and bundle as ZIF8 made in the lab and also commercially available ZIF8 and evaluated the kinetics on those materials. And for that case the binderless seemed to be faster in terms of kinetics and also had a higher absorption capacity obviously because you have find it that doesn't absorb. That should be hopefully published soon, so keep an eye out for that. That is on the Fed.

Thanks Hassan. Our final question so far. This is from Andre Bezrukov. This is for Paul. The behavior of novo MOF sample looks similar to deliquescent materials from the water vapor isotherm shape. Have you seen any liquid formation at high Rh during water absorption?

We haven't seen it per se, but we do have a camera accessory for our device which you can follow anything that happens on the sample as a function of the experiment. That I think would be a great thing to try on next. So thanks for the suggestion.

Well, that wraps up the submitted questions. I know there was, you know, we have a few more minutes if there's any general topics Paul or any of the other presenters want to throw out there for discussion or we can, we can wrap it up. But from my standpoint I just wanted to thank everybody for their time and maybe I'll see Paul if there's any other questions you wanted to throw out to our generous collaborators and presenters from the Imperial College.

If you don't mind me taking advantage of your expertise, I do have a couple of questions over here. So first to Lucy. It's very interesting that at director capture levels the rotary system seems to be primarily kinetics limited. I really enjoyed the graph as what sort of limitations as a function of the kinetics and then followed by thermal properties. It's more like an open discussion question, I guess. Does CO2 capacity even make an impact?

That's a really interesting question. So in the specific cycle that I showed you, I would say that the capacity of the sorbent didn't matter so much. However, keeping in mind that the linear driving force equation, the driving force is a bit driven by your equilibrium uptake of your material which does then impact your kinetic, your kinetics a bit. So it's not that it doesn't matter, but the capacity isn't as important. So actually your kinetics are a bit more important and the thermal properties. But then without having for example simulated the steam process yet, I can't say for sure that with Now a different heating fluid and also different driving forces in your desorption that the capacity won't matter anymore. So I think it's, it's really is good, a good question to highlight the importance of measuring properties of material and development. And obviously I still need to know the isotherm and the equilibrium uptake, right? I still need to know that information to do the process model. It's in the linear driving force. You still might want to utilize all your sorbent and see where you saturate it. But then a lot of material ranking is done based on uptake and you can see in this process specifically that what would be more important is fast kinetics and thermal properties. So yeah, but it does depend on your vessel, your cycle, all of those things.

But I think that that's a great insight. And like you mentioned, a lot of materials are often ranked by capacity or sometimes working capacity, which is what people should be looking at. But sometimes not even working capacity, just purely uptake at one bar. And that might not be the choice for the process at hand. And often I think it might, it isn't like you've just shown.

Thank you.

Okay, yeah, thanks. Thanks for going into that. I do have another question for Hassan and that sort of finalizes my questions here. You did show the two models, the near driving force model and the macropore press micropore linear driving force model. One inch is a bit more complicated than the other. I'm not a diffusion expert by any means, so please excuse my ignorance here, but when does the linear driving force model break down? So when would I want to pick the more complicated one over the other?

Yeah, yeah, that's a good question. And there's really two ways to approach this question. So the first way is really from a process design engineer's perspective where the question to be asked is how good is good enough? Essentially? Whereas where you're not really looking at a log scale in the concentration, if you're looking at the overall process, you don't really necessarily care that much about the slope of the tail of the desorption curve. So that is one thing. And the other thing is in terms of where the LDF breaks down, where the LDF breaks down is when you're more kinetically controlled in your operation of the diffusion process. So let's say you have a. Which is sort of determined by the choice of the LDF model. So there's LDF models for fast cycles and there's LDF models for slow cycles. What I've shown in my results is the LDF model proposed by Glukov. So it's the Glucoff equation where you have 15 diffusivity by r squared. And that works at conditions where your kinetics isn't really you're not really driven by the kinetics of the separation. So in which case it should be good enough for most separations. But then again, going into dac, where you're really driven by the mechanistic sort of reaction on the surface, that becomes a problem. And that's something that we really, really want to explore in more detail.

Yeah, very interesting, the Grim Reaper here. But unfortunately we do have some time constraints and limited on our availability for this webinar. So I want to wrap it up by thanking all of our speakers today. Thank you very much for your time. Your attention. This webinar will be available offline. It will be on our adsorption hub and will be available for those who could not catch the entire webinar. So please be on the lookout for details to get access to it. And I want to sincerely thank all of you for your time today. I know it's a busy time and we really appreciate your presentation and your expertise and your contribution. So with that, thank you guys and everybody. Have a good day.

Thank you everybody.

Cheers.

Thank you everyone. And a final note, for anyone who attended to hear the roundtable discussion, please be aware as we run out of time today, we will endeavor to schedule another session where we can get this in and have this discussion separately.

Thank you.