Accelerating Materials Innovation with Modern Separation Technologies
Good morning and good afternoon. Thank you for joining us. It is my pleasure to welcome you to the materials session of the LC symposium. My name is Brian Katzenmeyer and I will be today's chair session. After a brief introduction, I'll be joined by Amandaa Brewer, Miro Janco, and Bastiaan Staal at different points during today's session. They will each present their topics, which will be followed by a live Q&A. Then
I will share a few more comments before we conclude today's session. Before we begin with our first presentation, I would like to take a brief moment and remind why we care about characterizing advanced materials. Polymer chemistry and materials provide the building blocks across diverse end markets and applications that form a critical part of many products from the world's most important and well-known companies.
In order to deliver fit for purpose materials, scientists need to have a thorough understanding of both the chemical structure and physical properties of their products. This is impossible without effective characterization. Especially in the world of increasing polymer complexicity, calling for increasing levels of information, from our analysis techniques and the combination of several analytical techniques, to fully understand a polymer material structure and subsequent behavior.
Size exclusion chromatography is one such technique, and is an essential part of polymer design and synthesis. The challenges in separation of polymers are compounded by the development seen in polymer R&D today. With the rise of high-performance polymer research, biotechnology, and more green chemistry routes, scientists must take into account further considerations when analyzing polymers.
Green, or sustainable, chemistry has been a buzzword in the chemical community, generally, over the last few years and brings a number of things to think about in terms of analysis. Similarly, if we think and talk about modern polymer chemistry, so those areas I referenced just a moment ago, for example biotechnology, there are again new challenges to think about, in terms of analyzing these materials. Today we have three talks addressing some of these considerations. And I hope you're looking forward to them-- will enjoy them as much as myself. Our first talk is entitled Mult-Detector Size Exclusion Chromatography, Accelerating Innovation Through Greater Insight Into Polymer Performance Properties.
And this illustrates how multi-detection can be leveraged to determine differences in polymer samples. I would like to introduce Amandaa Brewer. Amandaa received her PhD in analytical chemistry in 2010 from Florida State University and her bachelor's degree in chemistry from Roanoke College in 2005. She studied under Andre Streigel for her doctorate, where she utilized multi-detection SEC to characterize ultra-high molar mass polymers. Amandaa has worked at Tosoh Biosciences, where she was a GPC technical specialist and later GPC sales support leader. In 2015 Amandaa joined Arkema at their North American research and development site in King of Prussia.
Pennsylvania. She is currently a senior research scientist in Analytical and Systems Research Group, and heads the gel permeation chromatography laboratory. Her research interests are in the area of macro-molecular characterization, in particular using structure property relationships, obtained by multi-detector size separation techniques, in the aid to understanding end-product performance and applications, as well as the characterization of ultra-high molar mass polymers and applying multi-detector hydrodynamic chromatography in industrial applications. Amanda is the author of peer reviewed publications focusing on the area of size-based separation methods such as size exclusion and hydrodynamic chromatography, coupled to multi-detection methods, namely multi angle light scattering, quasi electric light scattering, differential viscometry, and differential refractrometry. She has published an encyclopedia article on gel permeation column packing material and lectured in the area of multi-detector size exclusion chromatography. I will now turn it over to Amanda.
Good morning and thank you for the introduction Brian. And as Brian mentioned, I'm going to be talking about how we can use multi-detector size exclusion chromatography as a tool to get insight into polymer performance properties. If we look around us, polymers are everywhere. They're found in things like water filtration, transportation, packaging, paper, electronics, health, and cosmetics.
And as technology progresses every single day, so does the complexity of those polymers. And it's really important for us to be able to understand how those polymers work and what they look like and use that as a tool to understand what properties it will have in it's in use application, and what applications, in fact, we will be able to use these polymers for. When we look at polymer characterization using a technique such as size exclusion chromatography, we're looking at solution-based polymer characteristics.
These things include molar mass, molar mass distribution, polymeric size, branching, and architecture. And those effect in use properties such as adhesion, packing, mixing, shear strength, haze, crystallinity. But these are all properties that are influenced by the solution properties of a polymer that we can determine.
And we can determine those solution properties by multi-detector SEC, and then we can go back and use our understanding of structure property relationships and in use properties to determine performance, and how we can use this polymer in the real world. Now I mentioned, we're going to look at these solution properties. And we're going to look at them using size exclusion chromatography. For those of you who are not familiar with size exclusion chromatography, or SEC, it's a solution based chromatographic technique that sorts the molecules according to their size or hydrodynamic volume in solution. So we have an SEC column and it's typically longer than a HPLC column, about 30 centimeters, and thicker, about 7.8 ID. And we have a sample.
And what we do is, we inject our sample into this column, which is packed with a porous material, and we separate our analyte based on size, where the larger analytes can't sample as much of the pore volume as the smaller analytes. So if we migrate our polymer down the column, we'll see that the larger analytes take a shorter path, then the medium ones, then the smaller ones. Until finally, they elute from the column and they elute based on size. And we end up with a chromatogram, depending on what detector response or detector we have at the end of our columns, where we have the blue analytes, which are our larger ones, elute first, followed by our medium sized ones, and then our smaller ones, the green ones.
So what's the most important thing to remember here is that we are separating our polymer based on size, or hydrodynamic volume in solution, and not molar mass. A polymer can be high in molar mass and small in size, or vise versa. Really more than just molar mass plays into that role. And that's why it's important to remember that it's the hydrodynamic volume or the polymeric size in solution. So what we're going to do is we're going to walk through a case study today.
This case study involves two polymers. They are very high in molar mass, greater than 2 million. They're acrylic. And what their goal is, is to enhance the processability of high output extrusions.
So they go into the application, and it's important for them to have a fairly high melt strength, or aid in improving the melt strength of the final product, and you want to do this efficiently, of course. So when we talk about size exclusion chromatography of high molar mass polymers, there's a lot of things we have to consider, and there's a lot of challenges that are traditionally faced. Some of them include the availability of columns and standards, the concentration limitations, as we get high in molar mass, we have to look at working with lower concentrations. We also run into solubility issues.
It can take a long time for something in the millions or 10 millions in molar mass to dissolve. We also have this possibility of degradation, as it goes in and out of those pores of the SEC columns. The most basic type of size exclusion chromatography experiment is to take the size exclusion chromatography columns and decouple them to a concentration sensitive detector, typically that's a refractive index detector, and to use that to determine the molar mass by peak position calibration. And that's, in fact, what we have done here, in this particular case study. As a reminder, we have two polymers, polymer A and polymer B, with the same chemical composition. We're looking at the molar mass of these polymers, relative to polystyrene.
If you look here at the table, you'll see that the molar mass of polymer A is greater than the molar mass of polymer B. But polymer A and polymer B have the same polydispersity index. Here's the issue though, when we looked at the performance comparison, polymer B always had superior performance to polymer A in the in use application. And the polystyrene calibration curve show that B was significantly higher in molar mass than A, and that's about all. So what happened was we tried to make a polymer that was the same molar mass as polymer B, a research sample, and couldn't get it to improve the application.
No matter how much we tried, if polymer A and polymer had the same molar mass, their in use performance was not the same. So there had to be something more to the picture than just simply the more mass of the polymer. So this is where multi detector SEC came in.
That we needed to take this single detector SEC experiment that was only giving us a relative molar mass and change it into a multi detector setup. And again, we still have those challenges we have to address, but hopefully we can use this to our advantage, and use our multiple detectors set up to gain a more detailed picture of the macromoleculer properties of our polymer. So how do we develop this multi detector SEC setup from our single detector. So we need to look at a few things. So the first of those is going to be the dissolution time. How long is it going to take for these samples, that are 10 million in molar mass plus, to dissolve? Overnight? A weekend? So we need to see at what point do we reach en equilibrium.
We need to make sure we're at a concentration below this critical overlap concentration in order to use our light scattering detectors. Column selection, we need to make sure that we have a particle and pore size that is appropriate, as well as packing material for these types of polymers. From there we want to do some offline measurements. And what are these? The first is we need to determine the dn/dc, the differential and refractive index increment, using our RI detector in batch mode. That enables us to see how the concentration changes with detector response.
And we need this value for our light scattering calculations, and to obtain the molar mass. From there we want to look at Zimm plot. A Zimm plot is also an offline measurement that uses our MALS detector.
And it provides us with a bulk property determination of the weight average molar mass, the second variable coefficient, and the radius of gyration. You may wonder, why is this important. Well, this is important because we can use this to determine if are polymers are degrading during the separation. So a Zimm plot looks something like this.
Like I said, we get the bulk property of mw and the radius of gyration, the z average. And ideally, in the real world, everything is all good. We want to have the weight average molar mass of the online measurement, so that involving the separation, and that of the offline measurement to be the same. As well as the ratings of gyration to be the same by both methods. And this will enable us to know what is going on with our sample during separation, if anything. Our next step is to determine what detectors we're interested in.
Are we interested in a single detector experiment? Or like we discussed, a multi detector experiment? Do we want to look at molar mass by calibration, absolute molar mass, the radius of gyration, the hydrodynamic radius, the viscometric radius, the intrinsic viscosity, the molecular architecture? All of these are things that we must consider as we determine what detectors to put in our multiple detectors set up. From there, once we have our detectors selected, we want to look at flow rate of our samples. At what flow do we want to run our system? This is because these samples are so high in molar mass, we need to consider something known as on column flow induced degradation, which occurs for high molar mass polymers, as they go in and out of the pores of an SEC column. So our ultimate goal is for the molar mass, and the radius of gyration of the samples, to remain constant as both a function of flow rate, and that of on and off line experiment.
So we want to run our experiments at various flow rates to determine if we see a change in molar mass and radius of gyration. The first flow rate that we looked at is the most popular flow rate used inside exclusion chromatography experiments, 1.0 mL/minute. And what we saw was we saw that at 1.0 mL/minute, the weight average molar mass and the radius of gyration of these acrylics was about 30% less of that obtained by offline MALS.
So this tells us that our polymers are experiencing what we call on column flow induced degradation during our size exclusion experiment. So the next step from there is to look at an even lower flow rate. Here we moved from 1.0 mL/minute to 0.5 mL/minute.
So if you look at our chromatagram here, with the retention volume on the x-axis, the RI detector response on the y. In red we have 1.0 mL/minute. In black we have 0.5 mL/minute. And as you can see a shift in the retention time of our elution profile towards a shorter retention time, indicating that there's a difference in the polymeric size between those two flow rates.
And in fact, if we looked at 0.5 mL/minute, both the weight averaged molar mass and the z average radius of gyration, for this high molar mass acrylic, it's only about 15% lower than that that was obtained by the offline MALS experiment. Next was to lower the flow rate even more, to 0.25 mL/minute. And again, if you look at the SEC elution profile here, the blue trace has shifted to an even shorter retention time than the black trace, indicating that again, the molar mass and the radius of gyration is greater than that of the previous flow rate.
And in fact, here, we're only about 5% lower than that obtained by offline MALS. So we've definitely minimized the sample degradation. And you might ask at this point, why don't we go any lower. And this is the time that you have to ask yourself, is it worth it. Do you need to run at a lower flow rate to be 5% closer in molar mass, and how much time does that take? These types of experiments with a differential viscometer at the low flow rate can take up to three hours per injection.
So here we decided, OK, we're OK being about 5% off, because we need to hit play time needs to play a role in our method development. So from there, you look at data collection. How long do your runs take? Do you run everything in triplicate? Do you run it more than that? And then your data analysis. What parameters do you want to actually study and extract from your chromatograms. Are you interested in the molar mass? The polymeric size? The architecture? So these are all considerations that are going into the development of these experiments Now that you've decided that these are the experiments you want to do and the ideal conditions for these experiments, we took them into this multi-detector setup. So we have the SEC columns coupled to four detectors in this case, a multi-angle light scattering detector, which is going to be used to determine the absolute molar mass of our sample, and a sizing parameter called the radius of gyration.
Within our moles unit. We have a quasi elastic light scattering detector or dynamic light scattering detector. And that's going to be used to determine a different sizing primer known as the hydrodynamic radius. From there, we have a differential viscometer, which can be used to actually get both a sizing parameter, the viscometric radius, as well as the intrinsic viscosity of the sample. And you'll see later on the architecture is also obtainable, once you have a viscometer.
Then the final detector is a differential refractometer. And that's our concentration sensitive detector. And we need that to determine information from the other detectors. Now I mentioned that there's three different polymeric sizes that we're going to obtain, and I should note that these are just three different ways of looking at the polymer. And you'll notice they differ within the polymer and that's expected also. So as a reminder, we have two polymers in this case study that we're looking at, polymer A and polymer B, and they have the exact same chemical composition.
Previously, we looked at these two polymers using polystyrene calibration, and we saw that polymer B was significantly higher and more mass than polymer A, and it was always outperforming polymer A also. Except if we made a polymer, the exact same chemical composition, that was the same molar mass as polymer B, we still saw that the polymer B was outperforming polymer A. So we needed a way or see if there was a reason why there is such a discrepancy in the performances. So our multi-detectors setup, we first determine the absolute molar mass of our polymer. So we do that using our multi angle light scattering detector. And as you see here, we have obtained the molar mass from polymer A and polymer B, and polymer B still has a higher molar mass than polymer A, but they no longer have equal polydispersities.
The polydispersity of polymer B is less than that of polymer A. If we look at the SEC elution profile, which we have done here, where we have the retention volume on the x-axis, the 90 degree light scattering detector on the y, And we've plotted the molar mass across the elution profile, we see that polymer A does extend further in the low polymeric size, or the lower molar mass in the chromatogram. Second thing that we obtain from our multi-detector SEC experiment is a polymeric size, known as the radius of gyration. This is obtained by the MALS detector. The radius of gyration is the root main square distance of an array of atoms from their common center of mass. Again, it is a sizing parameter and a way to look at our polymer.
If we compare the radius of gyration in the radius of gyration distribution for polymer A and polymer B, we can see that polymer B is greater than that of polymer A, as well as has a lower polydispersity than polymer A. If again we look at the SEC elution profile, with the retention volume on the x-axis, the 90 discrete light scattering signal on the y, and we plot the radius of gyration across the elution profile, we see that, when we get towards the lower size portion of the profile, that polymer A is significantly lower than that of polymer B. The second sizing perimeter we obtain from our multi-detector setup is known as the hydrodynamic radius. We determine this using our quasi elastic like scattering detector. By definition, the hydrodynamic radius is the radius of an equivalent hard sphere that has the same translational diffusion coefficient as a macromolecule. As you can see here, the hydrodynamic radius for polymer A and polymer B are pretty similar to one another.
And if we plot them across the SEC elution profile, as we did for the radius of gyration, we see very little difference in the two plots. Our third sizing parameter is known as the viscometric radius. We determine this using our differential viscometer. And by definition, this is the radius of a solid sphere that increases the fluid viscosity by the same amount as does a macromolecule. And for polymer A and polymer B, we can see that polymer B has a higher viscometric radius, especially in the z average range. But we can see A is little more polydisperse than that of B.
And if we plot this across the SEC elution profile, just as we did with the other radii, we'll see that polymer A extends further out in the low intrinsic viscosity region, compared to polymer B. We can also determine something known as the intrinsic viscosity from the differential viscometer. The intrinsic viscosity can be thought of as inverse density.
The more extended a polymer is the solution, the higher the intrinsic viscocity is going to be. And we determine the intrinsic viscosity by looking at the ratio of the viscometer to the refractive index detector. And you can see here, the intrinsic viscosity for polymer B is significantly higher than that of polymer A, as well as the polydispersity is more narrow.
And if we plot this across the SEC elution profile, just as we did with the other solution properties, we can see that polymer A continues to extend further in the low end of the distribution, compared to polymer B. Now we can take the information obtained from the differential viscometer, as well as our molar mass obtained from our light scattering detector, and combine those to look at the molecular architecture of our polymer, in something known as a Mark-Houwink plot. A Mark-Houwink plot is the relationship between the intrinsic viscosity, the molar mass of the sample, and the fractal dimension.
So what we do is we have a log log plot, with the molar mass in the x-axis, and the intrinsic viscosity on the y, and the slope of that line is what can be used to determine our molecular architecture, or our fractal dimension of the polymer. For real world polymers, the Mark-Houwink plot tends to look something like this, where we still have the molar mass and the intrinsic viscosity, but we don't have a linear relationship between the two. As the molar mass of a polymer grows, the Mark-Houwink plot changes in slope and tends to go from something that indicates a linear random coil, at lower molar masses, to a branch molecule, at higher molar masses. So here's the Mark-Houwink plot for polymer A. Again the molar mass, as a function of in-transit viscosity, so as the molar mass grows, we look at the slope of the line. And you see at the lower molar mass in and around 10 to 6, we have a fractal dimension that corresponds to that of a linear random coil.
As we grow and grow in molar mass, that alpha, or a value, becomes smaller and we're forming a molecule that has some random branching. So by the time this polymer gets high in molar mass, it has changed its architecture into something that is branch. Now if we compare this to polymer B-- remember polymer B is the one that had the superior performance-- we see that polymer B, across the entire molar mass distribution, is branch. Then we have this huge difference in polymeric architecture between the various samples.
So if we want to summarize our multi-detector SEC observations, we can see that polymer B has a narrow polydispersity than that of polymer A. We know that polymer B is branch across the majority of the molar mass distribution, where polymer A has a linear architecture in the low molar mass region and a branch architecture and the higher. We want to take this information about molecular architecture and see how it influences the in use properties. And we know in general, that the melt strength and elasticity are effected by branching The more branching you have, the higher the melt strength. And we also know that the cell structure is affected by branching, so the more branching you have, the better the cell structure, the lower use level and higher compressive strength you need, of this product. Like I just mentioned, we want to be able to compare the multi-detector SEC data with rheology and microscopy.
So here's the rheology data of polymer A and polymer B. We want to look at the melt strength and elasticity relationship. And we know that more branching something has, the higher the melt strength it should be. So if you look at polymer A here and polymer B, polymer B has a greater melt strength than polymer A, which is exactly what we would expect based on the molar mass, branching, and the polydispersity between A and B. Next, we looked at microscopy data and the cell structure. And we know that a polymer is going to have better cell structure and higher compressive strength when its branched.
So if you look at the microscopy images here, of polymer A and polymer B, you'll see that the images in B are much more uniform than that of polymer A. And in fact the compressive strength of polymer B is also greater than that a polymer A. Remember earlier I was saying that we were able to make a polymer of same molar mass of B in the same chemical composition, but the compressive strength was always off.
And that was basically because of the fact that we didn't have the branching, like I mentioned before. The cell structure, no matter how high the molar mass is, without the branching is just not as good as that of a molecule that is branched. So, to summarize, we took a single detector SEC experiment, that was based on looking at acrylic polymer using polystyrene calibration curves, and converted it to a multi-M detector setup, that involved MALS, QELS, viscometry and RI. And we used that to characterize these high molar mass polymers. Through the multi-detector SEC work, we were able to get a detailed characterization of our polymer, that included the absolute molar mass, the polymeric size, the intrinsic viscosity, and the polymer architecture, through those Mark-Houwink plots. This combination of absolute molar mass and polymer architecture, provides a key component in understanding the performance properties, specifically melt strength and compression strength, of these high molar mass polymers.
So by combining the multi-detector SEC data from the various detection methods and solution properties, we took those and combined them with rheology and microscopy to get this picture of the polymer that will help us understand the polymer performance properties. I'd like to thank Sara and Gunter, who helped with the rheology and microscopy portion of this comparison, as well as Mark, who helped the three of us combine our information to understand the in use properties of this particular product line. And thank you, the audience, for listening. And at this point, I'd be happy to answer any questions you may have. Any questions, you may have.
All right, thank you very much Amanda. That was a great talk. If you have any questions, please take a moment and submit them in the Q&A box on your Zoom screen and we'll get to them.
We do have a few questions already that have come in, and I'd like to address those. So Amandaa, if you don't mind, could you come on screen and unmute, and we can get to the first question, please. Yes, Brian, thank you. So the first question is regarding the degradation of high molar mass polymers on the column through the pores, and if there's any other parameters that might affect this.
So as I mentioned in my talk of the flow rate, we see that analytes that are large in polymeric size tend to degrade, especially if they're linear going in and out of the pores of the column. Concentration is going to affect this in some regards. There's a term known as viscous fingering. If something is too high and concentration, that's going to be an issue. We also have an issue with concentration for our light scattering detector, apparent that would be below the critical overlap concentration or C-star.
So the concentration, while it doesn't necessarily play as big of a role in going in and out of the pores and degrading the polymer, it does play a role in the actual separation, as well as the detection methods. Great. Thank you. Thank you Amandaa. That was great. We also have a second question from Rachel, who asked about Mark-Huewink plots, about how they provide great information on fractal dimension of a polymer.
And they wanted to know if you could explain a little bit more on how fractal dimension of a polymer affects the behavior and solutions, such as stability. So when you have a polymer, whether it's linear or branched, the solubility of it is going to vary a little bit, as well as how it behaves in that solution, especially if you're doing RI or polystyrene calibration relative RI measurements. And the fact is that something that's linear is going to have a larger hydrodynamic volume, than something, let's say, that's branched to the same molar mass. So we really have to take that into consideration when we don't know anything about the polymer's structure or architecture.
And when we're applying the calibration curves, we have to remember that it's relative to something linear, but maybe my polymer isn't linear and the number that I'm getting is not necessarily the most accurate number for that polymer. OK great. We also have another question coming in from Peter.
Do you do quasi electrostatic light scattering in line as well, and is there a reason for that somewhat scattered plot? The quasi elastic light scattering detector, or the QELS detector, is in line. It's actually housed within the MALS detector and, basically, one of the photo diodes is replaced with an avalanche photodiode to do the QELS measurements. That plot is a little more scattered because we collect data at a lower frequency there.
It takes a lot of computer power collecting those, so we do decrease the frequency of the QELS to actually, just prevent, really, at the low flow rates, crashing of the computers. Yeah, makes sense. Got to have a good computer. We do have another question from Denise here. She asks, polymer B performs better than polymer A, due to consistent branching throughout the whole molar mass distribution.
Is there an example in polymer A, where the molecular weight changes throughout the distribution, and it would be advantageous to a product or process. So, yes. If you just need a range of molar masses and a range of polymer architecture, this would be a perfect polymer to have for that. As far as this specific application that I talked about here, which obviously I can't disclose exactly what it was, that polymer wasn't performing as well as something that was completely branched in there. So you find kind of ways around it when the polymer may not be performing exactly, whether you add more of this, you add a different additive to it, or things like that. So a lot of times it's working with the end use user to determine what we can do to their so-called formulation that puts power in there, to get to the right product properties.
Great. We also have another question from another Denise here. And she asks, can you use multi-detector system for accurate co-polymer characterization.
So yes. But there are some caveats within that. And the co-polymers, when they have different monomers in there, that vary a lot in the DNDC value, or chemical structure, or architecture, things like that, where you're going to have to, kind of, do some additional work in your light scattering detectors. I know that there are a lot of things being done currently in the literature, where you're not necessarily correcting for the co-polymer, but they're characterizing it as best they can without spending a ton of time. Because, for those of us that are in industry, we know that sometimes we have to weigh the option of time versus how accurate we are in our measurements. OK.
All right. And I think we have time for one more question. And this also comes from Latin America here, so we had to get a quick interpretation. But the question asks, what type of column would be ideal for a polymer weight around 8,000 daltons. So that's going to depend on a few things, right? You need to look at the chemical composition of your polymer, and then obviously the 8,000 daltons is important, so you want to find a column that's-- basically all of the column vendors will tell you, based on your chemistry of your polymer, the various columns that they would recommend. And then you're going to want one that has a very low separation range.
If it's only in the 8,000's, then it's going to be kind of the lower end of the spectrum, and one that's going to separate, and have an ideal separation in that range. Right. Awesome. Well thank you very much Amandaa. I think we'll have to move on, but if there are still questions for Amandaa, please continue to submit them in chat and she will get to them here over the next 10, 15 minutes.
But I would like to thank you for coming today and putting on a great talk and I appreciate your support in this event. Thank you, Brian. Thank you.
So our next presentation is entitled Expanding Detector Capabilities in SEC and APC on-line SEC-ICP MS and APC-ICP MS Hyphenation. Quite the mouthful. This talk combines SEC and APC with advanced detection, in order to characterize polymers containing heteroatoms.
So I'm happy to introduce Dr. Miroslav Janco. Miro is a senior research scientist and a polymer separations leader within the analytical science team at the core R&D organization at Dow Chemical. He holds a master's degree with honors in chemical engineering from the Institute of Fine Chemical Technology in Moscow, at Russia. And a doctorate in macromolecular chemistry from the Polymer Institute of the Slovak Academy of Sciences in Bratislava. His extensive career in the field of separation science and analytical chemistry was built and within the renowned laboratories around the world, including Osaka University, University of California, Berkeley, Hahn-Meitner Institute, and the University of Torino, where he developed a unique LC cap NMR hyphenated technique for the separation of polymers by tacticity, fast polymer separations by HPLC on monolithic columns, and applied LC cap for separations for block copolymers respectively.
Currently, his research is focused on separation characterizations of polymers on both particle and molecular levels by HDC and LC, including SEC, HPLC, LCCAP modes, with conventional and enhanced detectors, as well as coupled hyphenated techniques. Doctor Janco has authored and co-authored 26 original papers in his peer reviewed journals, and presented his research at more than 60 International conferences. His latest contribution to the field of polymer separation characterization is the development of ultra high pressure size exclusion chromatography, which is commercialized as the advanced polymers chromatography system by water or APC. Miro I will turn it over to you. Thank you, Brian, for the kind introduction and for the opportunity to present at this forum.
Good morning, ladies and gentlemen and thank you for joining this talk. The topic of my presentation today is enhancing detection capabilities in size exclusion chromatography and advanced polymer chromatography, shortly SEC and APC, and more specifically, I will cover on-line hyphenation of this size based separation technique with inductively coupled mass spectrometry, shortly ICP MS. Before I start, I would like to acknowledge my colleagues Bertha Snow and Patrick Fryfogle for their contribution to this work. In the course of this presentation, after a short introduction of polymer heterogeneity and liquid chromatography molds, I will cover APC as a way to address the continuous demand for short analysis times, better resolution, and improved precision. I will also demonstrate the increase in obtaining information about analyze sample when multiple conventional detectors are employed.
Next, I will present the example of both SEC-ICP MS and APC-ICP MS hyphenations and demonstrate that these are effective approaches to determine, not only molecular weight, molecular weight distribution, and dispersal of analyzed polymers, but also to determine the amount of heteroatom containing monomer and is distribution within molecular weight of analyzed polymers, thus increasing information averageness about analyzed samples. Let me start with the statement that only mother nature can make uniform polymers. Man-made polymers are dispersed materials distributed in more than one direction.
In addition to a distribution in molecular weight, they show distribution in chemical composition, and group's toxicity and molecular architecture or topology. To process all these distribution with a single analytical technique is a very challenging task, if not an impossible one. It's clear that the simple analytical method cannot provide all the required information. Methods which do not involve any separation, such as osmometry, light scattering, viscometry, or even NMR spectroscopy, generally yield only an average of the property to be determined, such as molecular weight, chemical composition, and functionality. When information about the distribution of these properties is required, a separation step has to be included. Liquid chromatography and its differential chromatography modes depicted in this image are often used.
SEC and APC are chromatographic techniques that are most often used to determine molecular weight distribution. While HPLC and UPLC are most often used to determine chemical composition distribution. Liquid chromatography at critical absorption point is used to process end functionality and toxicity distributions, as well as molecular weight distribution of one block in block copolymers.
To further enhance the richness of obtained information on analyzed samples, still within one chromatographic graph, a coupling of two different chromatography models is required. We refer to that as a two DLC. This topic is out of the scope of this presentation, but is covered by one of my colleague presenters in this section. In my presentation, I will adhere to a single chromatographic mode, specifically SEC or APC, and the richness of the obtained information about analyzed sample will be achieved through the coupling of multiple conventional detectors or employment of enhanced detectors.
The representative list of detectors used in liquid chromatography that I divided into two arbitrary groups, labeled as conventional and advanced, is presented on this slide. Enhancing the richness of information about analyzed sample can be easily achieved by combining several conventional detector, such as RI, UV, light scattering, and viscometric detector, by using detector from advanced group, or by proper combination of detectors from both groups, as it will be demonstrated in the course of this presentation. It has already been several years since SEC scientific community is taking advantage of APC technology. APC is application technique for size based separation of polymers, using columns packed with substrate micron, rigid, high-pore volume, hybrid particles, combined with fully optimized low dispersion ACQUITY LC system. APC is superior to SEC in speed, resolution, precision, and, last but not least, sustainability.
Typical SEC run can take from half hour to one hour, while APC analysis can be finished as short as in several minutes. In terms of resolution, I believe that presented figures speak for themselves, demonstrating much higher resolution power of the APC. In terms of precision, RSD below 1% can be achieved using APC technology, while RSD of 5% to 10% is generally acceptable for SEC. Let me just say that, at the early stages of APC development, with only UV and LSD detector option available, APC technique deliver only a limited amount of information about analyzed samples. However, today the detection capabilities in APC are equal to those in SEC and the information revealed about analyzed sample is significantly improved, as it is demonstrated on the next slide.
The informational richness improves when we start to couple two or more conventional LC detectors together, as depicted in these setups on the left, where APC is combined with the set of four detectors. RI, UV, light scattering, and viscometer. Delivered result is not only a relative molecular weight, but also absolute one, plus information about the hydrodynamic radius, radius of gyration, and the branching, or confirmation of information. And in some instances, also chemical composition distribution within the molecular weight, as is shown in those figures on the right. All that, with the significantly shorter analysis times, due to the speed delivered by APC technology. All the obtained information about analyzed sample is quite impressive.
In some instances, it still might not be sufficient. So in the next part of presentation, detector from enhanced group of detector, specifically ICP MS, will be employed with attempt to further increase informational richness of our analyzed samples. Let me also make clear that ICP MS detection is feasible for substances that contain elements, I will refer to them as a heteroatom in the course of a presentation, that are detectable by ICP MS. These heteroatoms are highlighted in this table.
There is a vast number of small molecules, monomers, and corresponding polymers and copolymers, several of these structures are also shown on this slide, that contained the heteroatoms detectable by ICP MS. And that is also great interest to know the fate of these heteroatoms containing moieties when incorporated into polymer chain and in fully formulated products. When the information about the total amount of the heteroatom is only required, spectroscopic techniques, such as SRF and ICP MS, alone are most often used.
However, the result is only average value. So when the distribution of the corresponding heteroatom containing moiety is needed, a separations step has to be included. Our current state of the art on-line SEC and APC-ICP MS setup is depicted in this slide. We utilize water's ACQUITY APC separation model to achieve desired separations. And the key component is the ICP MS instrument as a on-line detector. In this particular case, Agilent 7700 series ICP MS instrument is shown and utilized as a very sensitive and highly selective online detector, with a high linear dynamic range, capable of detecting all elements from lithium to uranium, as it was demonstrated in the table on previous slide.
So far we have demonstrated successful utilization of ICP MS as on-line detector in HDC, SEC, and HPLC separation modes. And most recently, also in APC mode. In terms of compatible solvent, both volatile aqueous buffers and a wide range of organic solvents were confirmed to be compatible with ICP MS detection. On this slide, let me briefly describe ICP MS detection. Column effluent, actually our sample, is continuously pumped into nebulizer, where it is converted into fine aerosol with argon gas. The fine droplets of the aerosol are separated from the larger one in the spray chamber.
The fine aerosol is transported into the plasma torch. In plasma, at high temperature, atomisation and ionization of the material take place. Subsequently, the ions are extracted into mass spectrometer, where the elemental composition of the material is determined. Now, I would like to present a couple of examples, where a combination of the resolving power and speed of sizably separation techniques, both SEC and APC, with the high selectivity and sensitivity of ICP MS made significant contribution to the discovery of unknown information that helped more customer project forward. First example is a separation and characterisation of heteroatom containing polymer by SEC RI ICP MS hyphenation. Effluent from SEC column is split and monitored by refractive index, shown on this slide and ICP MS detector on the next slide.
Using SEC charge of analyzed samples and the proper calibration curve, a relative molecular weight data and fraction below 1,500 can be determined. While the first file lot of analyzed materials are consistent in terms of molecular weight, Lot number six, depicted here in cyan color, show lower molecular weight and is identified as an outlier. SEC charts of the analyzed samples, as detected by ICP MS detector, are shown here.
There are three distinct peaks on the chromatogram. Peak number one is assigned as heteroatom response in the polymer. Peak number two is label as a residual, or unincorporated heteroatom containing monomer. And peak number three is unintended heteroatom containing byproduct.
Using calibration curl of heteroatom containing monomer on the upper right, the total amount and the distribution of the heteroatom containing monomer can be determined within polymer molecular weight and is summarized in table below. It's clear that twice as much of that item containing monomer was charged in the sample five. By combining information from both reflective index detector on previous slide and ICP MS detector on this slide, two of six analyzed batches were identified as outliers. Next example is even more challenging one. There is a need to characterize heteroatom containing component in fully formulated system when component of interest, shown here in red trace correlates with other formulation components, shown here as a black trace, and is the level in the formulation is low in the range couple of percent, as shown by the blue trace.
In other words, this is an example and we are looking for a needle in the haystack. Shown here is that overlay of the SEC charts of the neat heteroatom containing component, blue chart, and the same heteroatom containing component in the fully formulated product, red chart, as detected by ICP MS detector. Due to high sensitivity and selectivity of ICP MS detector, a challenging task become quite a straightforward one. Broadening of molecular weight distribution, a heterotopia containing component is revealed with the high sensitivity.
It is hypothesized that the heteroatom containing component is undergoing chain scission and chained at a combination during formulation process, resulting in the presence of a low molecular weight and high molecular weight species. A summary of our analysis of several lots of fully formulated products by SEC RI ICP MS hyphenation is shown on this slide. While no difference can be revealed by our refractive index detector, see insert in the top right corner.
The ICP MS detector reveals significant differences in the level, summarized in the table on the left, in molecular weight and molecular weight distribution of the heteroatom containing components, which correlated well process variables. The disadvantage of the SEC ICP MS approach is a long analysis time, usually 30 to 60 minutes. That become an issue when using solvents that char or carbonize a lot, causing a drift in ICP MS detector signal. ICP MS detector drift can be minimized and even completely eliminated by utilizing APC technology, which allows to achieve desired separation in the significantly shorter time.
On this slide, four batches of heteroatom containing polymer were analyzed and compared using APC, with the analysis time as short as 12 minutes. Another advantage of using APC is that the range of the applicable solvent is greatly increased, due to elastic column and liquid chromatographic system back pressure limitations. Due to short analysis time, the ICP MS detector response stays linear over a wide range of concentration, as demonstrated by calibration curve on the left, and the amount and the distribution of the heteroatom containing monomer can be easily determined and compare. Three batches from different vendors are quite consistent in the amount of heteroatom containing monomer, but one batch shows significantly lower amount. APC ICP MS results were validated by independent technique, XRF, and good agreement between APC ICP MS result and the result obtained by XRF is demonstrated. On this slide, I would like to address a potential concern of peak broadening, caused by ICP MS detector.
Overlay of SEC charts of heteroatom containing standard, covering a broad range of molecular weight, as detected by refractive index and ICP MS detector, is presented. Detector responses are high normalized and appear as a single chart despite the fact that there are two lines. One smooth one is for refractive index detector and second, more spiky one, is for the ICP MS, suggesting no or minimal peak broadening caused by ICP MS detector. With that, I would like to conclude that SEC and APC using conventional detectors, including RI, SUV, MALS, and VIS are very robust and cost effective analytical tools, but provides somewhat limited information on analyzed samples. We have demonstrated that SEC and APC hyphenated to ICP MS detector deliver significantly richer information on analyzed samples, but add to the complexity of experimental setup and data processing, resulting in a longer turnaround times and higher cost. Further improvements in both instrument and software technologies are required to minimize the complexity of the experiment using advanced detectors, so that SEC and APC hyphenated technique become routine analytical approaches for polymeric characterisation in both academic and industrial settings.
Last but not least, I would like to acknowledge the support from both Dow and Waters and thank you for your attention. OK. Thank you very much, Miro, for that great talk on hyphenating ICP MS to SEC.
So I would ask you to come on camera and unmute your mic and we have a couple of questions coming in already, but for everyone else, if you have a question for Miro, please use the Q&A box to submit your question. The first question for you Miro is, can you use multi-detector system-- sorry that's the wrong question. Are there any solvent restrictions for the ICP MS detector? OK. Good afternoon, Brian. Can you just acknowledge quickly-- can you hear me? Yes. We can hear you just fine.
Thanks. OK. Perfect. So yes, as to your question, through the utilization of the technology and when the people will be using it, you will find out that there will be certain solvent restriction.
Let's assume that you set up your separation correctly and so on, but you might find out that, we can see flowing that solvent to the ICP MS, plasma might stay not on, actually will extinguish plasma. So of course, that would not be, then, what I would describe as an ICP MS compatible solvent. So to give you the guidance, or rule of thumb, you would be more successful in utilizing high boiling point solvent and infusing them for the plasma. There would be more trouble with that a low boiling solvent types. But also, your success and utilizing the technology will depend on the instrumentation.
There are definitely always windows pushing the instrumentation quality up. And most recent, a series of the ICP MS instrument-- stability of plasma significantly improved. So that level of the number solvent, which can be infused to the plasma and plasma stays on, broaden So this is what would be my answer to you, without going to the more specific. OK.
Great. Thank you, Miro. There's also a question from Denise. There's actually two questions. I'll go ahead with the first one and then we'll let you answer that.
And then do the second one immediately following. Her first question is, since APC is a low dispersion system, compared to conventional SEC, is there a low dispersion ICP MS detector available? Or is the one that you used just a universal, one detector fits all model? OK so let's answer it in probably two parts. First part would be, at least to my knowledge, I am not aware that there will be specific where we would describe as low dispersion, or let's call it APC type of the ICP MS detector. And I would appreciate, since we have a broad audience here, if we kind of got that feedback here that I'm actually wrong, then I would love to learn that there is that such instrument.
And now the second part. Yes, for now, I am using the currently available instrumentation in both those modes. And I would agree that you would see probably effect of dispersion in ICP MS, especially when you are going to that very short dimension of your separation column especially, in APC. If people remember the photo of the column setup up, 3,150 millimeter long one times 4.6. It's dispersion comparing to the water's dispersion-- that column will be negligible.
Once you start to push, and you are going to 3.5 centimeters long, which would be equal to the one, what I would call a regular APC column. There could be instances that you might observe the effect of the dispersity. So you just have to be mindful of that and to keep that in mind. Great.
Thank you, Miro. I think we'll have time for you to answer this last question and then for anyone else who has additional questions for Miro, just please address them in the Q&A box and then we'll type out those answers. So thank you. And the last question, Miro, is, since you mentioned that, with advanced detection, you can determine Rg and Rh, can you give an application examples for which you would either determine Rg or Rh? Well let me just start with the statement that both Rg and hydrodynamic ratios are most commonly used parameters to determine the size. So both will tell that one of your molecule is bigger or that another one is smaller. When you will be actually having that choice which you are using, it would be if your molecule is actually anisotropic scatter.
So then, it would be probably your choice which you choose to determine and use, hydronic volume or [INAUDIBLE] saturation. In the cases-- and for that one it would be like that once more, a rule of thumb, 10 nanometer, in some instances, people will tell you, even as big as 15 nanometers. But let's say your molecules are smaller in size, what I would describe as isotropic scatterer, then probably your choice would be hydronamic volume. So this is what I would give as more as a guidance, what your options would be, versus hydronic volume. Great. Thank you, Miro.
And thank you for your presentation today. Just to keep pace with everything, we're going to move on to Bastiaan's talk, but I do appreciate you coming on today and delivering an excellent talk. So thank you very much, Miro. Thank you. Our last presentation, entitled Recycling gradients in polymer chromatography, a new approach on HPLC on polymers. It will focus on interactive liquid chromatography and how it can determine the chemical composition distribution of a polymer.
It is my pleasure to introduce Dr. Bastiaan Staal. Bastiaan is a polymer chemist, who finished his doctorate in 2005, at the University of Eindhoven on copolymer characterization by MALDI TOF MS. Since 2006, he has been the lab leader responsible for GPC and HPLC of polymers for the center of research at BASF. His main interest include developing alternative methods and data processing for both HPLC and GPC, searching for stable conditions for performing proper GPC, unraveling MS data of copolymers, and hyphenation techniques like GPC IR, and pushing the boundaries of 2D chromatography to its limits. His main drive is forming the bridge between theory and an industrial practice. Bastiaan, the floor is yours.
Good morning or afternoon. My name is Bastiaan Staal and today I would like to present you something about recycling gradients in polymer chromatography. Is this a new approach or is this just another curiosity? Before we go on to dive into all kind of details, I would like to take you on a journey from a perspective like, what is it, why we are all doing this? And one of the main goals of characterization is understanding one of those fundamental relationships, which we always like to picture. The chemical, the physical, and of course, the application, and how do they interact with each other.
Forget about a picture on the left. Unfortunately things are more complicated if we want to understand the overall application properties. It is not just linear combination of chemical and physical properties. But first of all, we need to characterize our samples, then with the characterization methods we use for polymers, we are basically close to the chemical properties which we can describe. What are it's NMR or HPLC or GPC?.
If you want to know more about the physical properties, we have to use other techniques. Nevertheless, in many cases, understanding why application is performance, good or bad, it is simply defined by the difference between good and bad performance examples. And that is the main task of characterization, to see if we can distinguish between good and bad performers.
So why is it so difficult and why are we talking so much about the characterization. So what type do we have then? Well there are many types of different characterization techniques we can use for characterizing polymers SEC is just one of them, or GPC, but there are many more. The problem is that there is not a single technique who gives the full picture of a polymer and characterized it at once. So what is it then that it makes so hard to characterize polymers. The problem with polymers is that they are never single distributions. They are distributions in distributions.
Because think about the length of polymers. They can be long or short. It can be different in chemical compositions, that is already a challenge. And they can be different in topology, for instance branched or not branched.
And all those properties together form our final product, or our final polymer with its characteristic properties. So, as you can imagine, it is impossible to separate all those distributions simultaneously and especially for industrial problems, which normally consist of many more different components. Today, I will only focus on the characterization methods where we separate polymers. Because only by separating, we might have a chance to understand those underlaying distributions. So what I'd understand on matters for polymer separation, because that's where we need to go to. Well, GPC separation to size is definitely one of them.
Interactive chromatography is another very important. And there are many more, but those two are the main ones, if you're talking about polymer separation. And all those methods require that polymers can be dissolved, so I will not talk about polymers in a solid state. But we need to dissolve polymers otherwise, if they can't be dissolved, all those methods cannot be applied. So that narrows down our methods which we can use, to some extent. And one of them already addressed, the GPC, or SEC, to determine the molar mass distribution, or actually, it's the size which we measure.
And the other one, interactive chromatography, is one of the systems where we can determine chemical composition or endgroups for other kinds of things. And today, I would like to focus on the chemical composition distribution. So before we can continue, I need to make sure that you understand the principle of interactive liquid chromatography. I tried to put a scheme here, where green is the nonsolvent and red is the solvent, then at t0, we inject on top of our column our sample, with precipitate or absorbs there, and doesn't move. And while the time increases, we move from a nonsolvent to a solvent.
Then our peak starts to move and elute in a particular order, until everything is out of the column. It is based on the fact that we are weakening the interaction with the column, as we are moving to a stronger solvent. In this sheet, we see an example of how separation according to chemical composition could work. In this case, we use a gradient from acetronitrile to THF where we separate a copolymer, SAN, called styrene acrylonitrile. With the time, we increased the solvent composition from acetronitrile to more THF. And we see that peaks eluting from our system, it contains more acrylonitrile.
This was measured offline and used as a collaboration curve. That's why you see those four red peaks in there. The green sample was a commercial sample, which was used for comparison to see to what content this acrylonitrile was in there. So this shows the idea of how we can use interactive chromatography for separating according to chemical composition, but remember this only works if there is no molar mass dependency. And to prove that, it's what we see at the next slide. By doing two dimensional chromatography, if we combine our size exclusion methods with our interactive chromatography methods, we can easily see that this separation is not completely molar mass dependent.
Especially in the low molar mass range, we see there is a strong molar mass dependency in the HPLC separation, which we are doing. On the right, we still see the same picture as we see on the slide before. And remember, the part which is marked in red, where the green curve goes down, that is actually based to the molar mass dependency. You can also see that directly in a two dimensional picture, because if the molar mass should be independent on the chemical composition, it should not have a banana shape, but it should become horizontal. And they'll bend it down. So that means, if we go back to another example here for polyethylene glycols, which are lower molar mass we see this effective and much more pronounced.
What we see here is an HPLC separation, but from homo polymer. On the left, we see the very low oligomer for polyethylene glycol and the three ones to the right are standards, with an increasing molar mass. So this is all nice and good because the separation looks very nice, but this is actually not what we want, if you want to separate for chemical composition. We want to diminish this molar mass effect. Actually we want none to have any molar mass effect at all, because it only jeopardize our way, how we have to look to our distribution. So congratulations to the people who made it so far, because now we finally can define our aim.
Can we indeed diminish those molar mass effects of an HPLC separation for polymers? That is a very good question. And it took me a long while to think what we can do about it. So here comes the idea. If we use an infinitely long column, in such a way that each polymer or molecule have enough time to reach its equilibrium state within the gradient, we might receive-- or reach our target, that we do not have a molar mass dependency anymore. Wow that's really a mouthful.
So, how does this work then in practice, if it works at all. Let's first have a look to the experimental setup. What we see here is that we use two columns, column one and column two, like ordinary GPC columns.
And they are switched in such a way that, if we put a gradient in, we can recycle it. And I will show you on the next slide how that works, because it's a bit difficult to read it from such kind of pictures. So let's start at the beginning, where we are just doing an ordinary chromatography, in this case interactive liquid chromatography. On the left picture, we see column one. It's going to be filled with a gradient from our pump. Remember, green is an nonsolvent and red is the solvent.
Starting completely filled with ACN, and we see that the color to the right is red, it's coming in our column. So our column one is going up in solvent composition. If you follow all those arrows, you will see that column two still remains green and it's completely full with acetronitrile. So that is what happens in the beginning at t0. Then we move to the right and then we see something that, at the end of a certain time, nearly all the gradient, in which I put in the first column at the beginning, it simply flushed through the first column and parked, or moved, to the second column.
You can also say you simply moved the gradient from the first column, with time, to the second column. That is correct. And we notice that our pump has moved back from the gradient from a good solvent to our nonsolvent, very rapidly. And that's why we see this gradually changing of color, very rapidly from red to green in our first column. S