Understanding Capillary LC - Part 3: Choosing the Right Column

The column is the heart of any liquid chromatography system, and at the capillary scale, the right choice can make or break your separation. In the third installment of our Understanding Capillary LC series, Dr. Sam Foster is joined by Dr. Dwight Stoll of Gustavus Adolphus College to take a deep dive into capillary column selection. We’ll explore the different formats available, how packing differences impact performance, and what commercial options exist today. You’ll also learn practical approaches for when your ideal column isn’t commercially available, including custom and in-house solutions. Finally, we’ll discuss how Axcend’s Heated Column Cartridge (HCC) design allows the Focus LC to work with virtually any capillary column—giving you the flexibility to tailor your method without being locked into one supplier.

Whether you’re just starting with capillary LC or refining an existing workflow, this session will help you navigate the many factors that go into choosing the right column for your application.

Key Takeaways:

• Understand the range of capillary column formats and how they differ
• Learn how packing materials influence separation efficiency
• Explore strategies for sourcing or creating non-commercial columns
• See how Axcend’s HCC design maximizes flexibility in method development

Webinar

Speakers

Samuel Foster, Ph.D.
Application Scientist
Axcend

Samuel Foster completed his Ph.D. in Pharmaceutical Chemistry from Rowan University in 2025. His research has focused on the development and application of capillary scale liquid chromatography instrumentation. He currently works at Axcend as an application scientist focusing on the development of chromatographic workflows for a variety of analyte classes including oligonucleotides, monoclonal antibodies, and drugs of abuse.

Dwight R. Stoll, Ph.D.
Professor of Chemistry 
Gustavus Adolphus College

Dwight Stoll is Professor of Chemistry at Gustavus Adolphus College in St. Peter, MN. He has authored or co-authored more than 100 peer-reviewed publications and five book chapters in separation science, speaks internationally on the topic, and is co-editor of the recent book: Multi-Dimensional Liquid Chromatography: Principles, practice, and applications. He has also written the monthly “LC Troubleshooting” column for LCGC Magazine since 2017, and co-hosts the Analytically Speaking podcast with Prof. Jim Grinias. His primary research focus is on the development of two-dimensional liquid chromatography (2D-LC) for both targeted and untargeted analyses. Within this area he has made contributions on many aspects of the technique including stationary phase characterization, biopharmaceutical analysis, new 2D-LC methodologies and instrumentation, and fundamental aspects including re-equilibration in gradient elution liquid chromatography and analyte focusing. He has taught several short courses on 2D-LC in recent years at venues including Pittcon and the international HPLC20XX series, and hosts multi-day hands-on trainings for multi-dimensional chromatography in his laboratory at Gustavus.

Transcript

Samuel Foster

All right. I think we've had everyone come in that's going to make it in. Thank you, everyone, for coming. I hope we can all hear me. For those of you who are just joining us, this is a webinar series where we talk about various aspects of capillary scale liquid chromatography. I'm Sam Foster. I'm an application scientist here, and I'm joined today by Dr. Dwight Stoll who will be talking a little bit more about how to properly select the column using the Column selectivity database he has put together. So, to start off, in the past, couple of talks, we've discussed various aspects of the system in terms of pumping as well as just various differences, with capillary scale but today I want to go more specifically into capillary scale columns.

So to start off with I always like to define what capillary scale is just to make sure everyone is on the same page. We all are kind of aware of, what we're talking about. Then a lot of the questions that we get asked and what I want to start to go over is analytical scale columns or larger scale columns in general, versus capillary scale columns.

Then I want to go into a number of nontraditional formats that, can often lend their, their, properties well to the capillary scale flow rates, go over some of the existing commercial options and then I'm going to hand it over to Doctor Stoll, who is going to, discuss ways of selecting an alternative column if your specific column isn't available, through commercial vendors.

So first and foremost, what is capillary scale? Chromatographic scale in general is typically determined based on the inner diameter of the column you're using, 4.6 to 2.1 millimeters in diameter. That's typically considered analytical scale or the 2.1 is sort of narrow bore. And capillary scale is really defined as point three millimeters in diameter or below.

Some places have it is 0.5. It depends on the source. But really what's key about capillary scale columns is the flow rates needed to achieve similar linear velocities. So these three separate flow rates are one mil a minute or 0.2 mils a minute. And our .004 mils a minute or four microliters a minute. All give comparable linear velocities or move through the column at the same rate.

And so that dramatic change in scale is useful for solvent savings. But it also comes with differences in chromatography that we've been discussing. So the real question that I think we're here to answer today is are analytical and capillary scale columns the same. And in fact, I want to put a poll out to see what the crowd thinks.

So we should be seeing a poll pop up. Do we think analytical scale and capillary scale columns are the same? We'll give it a couple of seconds here for everyone to answer.

I'm seeing a lot of no's. A few yeses, but the no’s take it 75 to 25. All right, well, thank you all for answering. The answer is yes. They are the same, but they also aren't. And that's really what we're going to be discussing. So in terms of practical usage, the efficiencies you're going to get out of them, for the most part, capillary scale is going to perform the same as analytical scale if you get a commercially packed column from a vendor that's been doing it for a while, typically you'll see very similar efficiencies. But how we get to those efficiencies is a little bit different. So, we want to pop this up. We often think of the column and sort of the packing in general as this magic separation device. Right.

If you're an industry chemist, you typically don't care about, you know, your packing methods or the composition of your column. But really, there's a lot that goes on in there. And this could be an entire college course so I'm going to be making a lot of sweeping generalizations today. So, that being said, really there are differences between these columns based on the particles you use, the inner diameters,how you pack them.

In this case, we're going to be looking primarily at slurry packed columns, which is a common method of packing both analytical and capillary scale, and really discovering why we see similarities, but also differences between these columns. And a lot of it has to do with something known as wall effects. So when you're packing these columns, you will start to see something that is a wall effect or a highly dense region, in this case, the red region of particles directly around the column wall.

And these tends to be a very ordered, high density bed, because all of the other particles are pushing out and can typically make a very dense packing layer. From there, though, we actually see two more distinct regions. We see an intermediate zone. This is several particle diameters thick. And then we see a bulk zone. And the interesting part is that intermediate zone often has a slightly higher packing density than your bulk region.

And what that leads to is differences in velocity as you push through the column. So you actually are seeing slightly slower flow on the edges of the column when compared to, your central bulk region. That's what we see here, is that we see, very low flow and then picking up and as we hit the bulk central zone, we really start to get to, a fairly stable, velocity profile.

Now, what does that do for our chromatography? Well, it's going to cause some band broadening as we flow through that column our center region is going to flow slightly faster than our outer regions. And again, these are, you know, small percentages. We're not talking about very, very considerable amount, but it is enough to cause loss of efficiency or resolution and Poorer peak shape. So what does all of this matter. Right. If we know these exist and we really can't do much to stop them, and there are ways to stop them and different packing methods that can start to mitigate them. Why do we see similar performance from both capillary and analytical scale columns? And the answer comes down to the ratio of that intermediate region to its bulk region.

So I have a wonderful figure here from a paper by Fabrice Gritty. And what this is, is, a theoretical prediction as, of, of column efficiencies based on the various inner diameters. And so what we start to see is that our traditional 4.6mm column or our 2.1mm columns, the bulk region takes a majority of that column space.

And so because of that, the wall effects are pretty mitigated and we see that similarly on our graph over here, our 2.1 is in red and our 4.6 is in blue. These two are offering high efficiencies and aren't seeing many impacts from that wall region. On the other end we see the same thing but inverted on the capillary scale.

So we see almost all of the column at 400 micron is that intermediate region with a very small bulk region. And so because of that the wall effects actually are because they're the majority of your column are causing less, loss of sample than they are or less loss of efficiency than they otherwise would have. And so we see that that 0.4mm column has comparable predicted efficiency compared to the 4.6mm column.

And what we really see is that there's this sweet spot of, loss of performance at about one millimeter. And again, that that's not to say all one millimeter columns are bad or all of them suffer from this. There are methods of mitigating these wall effects. But what we do see is that the theoretical prediction is that these one millimeter and 800 micron offer very poor performances because the the wall region in that intermediate zone have an outsized impact, but don't make up the majority of the column.

And so we get very similar efficiencies in two very different ways. And that I want to, you know, touch on is that, capillary scale for all intensive purposes is the same but when you start to get very, very deep in and far deeper than we'll go today, there are significant differences. Additionally, I wanted to touch on heat generation and dissipation.

This can really be a factor if you're running, you know, sub two or in fact, sub one micron particles at extremely high pressures, what we start to see is similar to that velocity difference. We'll see a temperature difference, caused by the friction of that mobile phase flowing through the particle bed. We tend to see, heat generation across the column.

And there's really only two ways of dissipating heat once it builds up in that column. The first way is simple. You flow the hot liquid out the end of the column. That one happens pretty quickly. The other way is, heat dissipation through the walls of the column. And that one typically will only diffuse through the edges because heat had to dissipate.

And so what you see is as you push at these very high pressures, at these very high, flow rates, we have extremely high-power production. In fact, in some of the examples given, the power was high enough to boil the liquid when it moved from the start to the end of the column, it would go from, you know, room temperature to boiling before it ever exited the column.

And what that causes is differences in viscosity across the column, where the center region, because the heat isn't properly dissipated, will have lower viscosity and slightly faster flow because it's warmer than that of the edges of the column, which again, is going to hurt your efficiency. Because we have less stationary phase, we have less energy or heat production.

And so because of that, these temperature effects are felt less when you go down into the capillary scale. Now, a lot of this happens when you're working at 7000 bar with sub one micron particles. So for practical everyday use at, you know, not ultra-high pressures. These these are necessarily the most important. But again, when we start to get down into, the finer details, they do tend to, play a role in capillary columns start to act a little differently than the analytical scale.

Additionally, I wanted to go over some of the lesser known but still relevant columns. And in fact, there are a couple of column types here that we're going to talk about that really lend, their benefits specifically to capillary for a number of reasons. So we'll get into that.

First and foremost we have monolith columns. So rather than packing a number of smaller particles into the column, we tend to instead make this porous membrane, this can either be a silica backbone or an organic polymer. And the idea of this is that this or its structure can serve similar to a packed column bed and providing a large amount of surface area, while at the same time requiring less packing and can be, you know, more easily fabricated.

This is just a couple of images of that and again, we can see that very porous structure. So a very high surface area, inside of this model with the columns. Now there are a couple of useful benefits to model columns. I think first and foremost, they produce a lot of lower back pressures. This can be great if you're trying to flow extremely quickly.

This also lends credence to very long columns as we'll get into some of these can reach meters in length, which is very different from, traditional packed capillaries or packed columns in general. Additionally, there is less mass or reduced mass transfer resistance, which allows for very high-speed separation, especially when it comes to, large molecules. You don't necessarily have to worry about, putting a lot of strain or stress on them. Now, the one downside or the drawback is that typically there is a broader dispersion of these poor sizes. So you'll see slightly lowered efficiency when compared to packed capillaries. That's not to say they can't meet efficiencies. In fact, there are plenty of papers where they do that.

Sort of the the rule of thumb, though, is traditionally we like to think of these as having slightly lower efficiencies than packed, bed capillaries. Our next mode is open tubular columns. And so these really came about, alongside that of GC columns or gas chromatography columns, with the idea of having a thin capillary coated with some functional group in order to perform these separations.

And in fact, it was actually Milton Lees group, who did some of the, the initial work with these back in the 70s. The difference really is the inner diameter needed. So with gas chromatography you have much higher diffusivity because gas can move around a lot more than liquid. Because of that. We really need to go to much smaller inner diameters for these columns to produce comparable efficiencies.

In fact, most of what I've seen recommends columns with 20 micron inner diameters or less. And so because of that, you really have to go slow in terms of flow rates to, to not clog it or overpressure it. They are very high efficiency. We don't necessarily have, the same, you know, packing concerns when we work with these, we, we also don't necessarily have to worry, about, you know, particle size distribution and things like that.

We really just have, you know, a capillary with a coated layer. They do suffer from drawbacks in terms of load ability. They aren't, you know, necessarily the same surface area, and therefore you can't load as much sample onto them as you could, with a packed capillary. But they do offer very high efficiencies in different mode.

Another mode that I wanted to talk about is pillar array columns. So these are sort of in the same idea as open tubular in that if we reduce the amount of heterogeneity in the column bed, we can produce higher efficiency separations rather than having a bunch of particles packed in, you know, fairly randomly, if we were to make a perfectly homogenous bed of these pillars and then coated them with a functional group, we'd be able to perform these separations with significant reductions to band broadening.

However, it does come with the stipulation that you actually have to make these beds, and so making very large beds or very high pressure beds of all of these different pillars can take a considerable amount of time. It often is fairly costly. And they certainly don't necessarily have, high enough pressure limits that you'd want to flow through at, you know, milliliters per minute.

So instead capillary scale becomes very useful with these columns. In terms of hyphenating LC to them. Additionally these and I thought this was a really cool graph is, they're able to perform these separations and much like, a packed capillary, the particle size or in this case, the pillar size and pillar distribution plays a role and plays an impact very similar to that of particle size where the smaller you have or pardon me, the smaller the pillar and the closer together they're packed, the higher efficiencies we tend to see. And so it acts very much like a packed, column, but it's made of these homogenous pillars. And so I thought that was, pretty interesting to show. This graph here just shows the differences of a Gen one versus Gen two column, showing just that, whereas we reduced the pillar diameter and put them closer together, we again see the expected, efficiency gains that we would.

So this is another method where it requires low, you know, often microliter a minute flow rates, which is great for capillary scale, and offers just a new way of doing chromatography that's fairly promising.

So looking at the commercial offerings of capillary columns, there really isn't a phase or, technique you can't find. Really, the only one I found so far is Protein A columns, and I'm on the hunt for that. There's some interest there. This was part of a, LCGC article that was recently published, going over, to my knowledge, every manufacturer, if you're a manufacturer watching this and you aren’t on the list, please let me know. I'd be happy to update that. But this is most of, if not all of the commercial offerings of capillary columns. And we start to see a couple of key geometries that, that stand out, primarily that of the .3 millimeters inner diameter or the 0.15mm inner diameter. These are sort of the two most common, inner diameters that we find, the standard column lengths we tend to see anywhere from, 5cm to 15cm.

We can get really crazy lengths. In fact, if we look at some of these monolithic columns like those offered by, GL Sciences, we can get all the way out to 2000mm, which, is pretty ridiculous for a column. But as we can see, there are a number of different phases. C18 is the most common, but we have, cation exchange, you know, HILIC, size exclusion, you know, various numbers of them.

And what we see, because this list is going to go on for a while, is a lot of sort of the, the standard column vendors that you look at, Waters, Thermo, YMC, Phenomenex, AMT, we, we also see a lot of distributors. So MAC-MOD they tend to have their own line but they also stock a lot of these.

So if you're looking for, distributors, you'll often find that one distributor can carry many of these capillary brands. We also see that the pressure limits tend to fall in line with that of HPLC columns. They're often either, you know, 4 to 600 or a thousand bar or if you're going UHPLC, really the only thing that you can't find with these columns is large particles.

And, that's for a great reason is typically the highest particle size we'll see is about five. That's not to say you can't pack capillary columns with larger particles. I packed some with 30 micron into ten micron. But really what you see is a significant loss of efficiency. I mean, if you're at a, you know, 100 micron inner diameter and you're using 30 micron particles, your bed is only three particles across.

It tends not to, offer incredibly high efficiencies. And so, really, you know, basically the sky is the limit with commercial offerings. That being said, you know, we'll get to Dwight in one more slide here. You know, there are certainly phases and columns where we miss now to, to talk about, extend a little bit.

Our system was designed with the intention of connecting to these commercially available columns. So, we have a flow rate range of 0.5 to 10 microliters a minute, which is sort of the standard operating range for the most common, column inner diameters, the 0.15, the 0.2, in the .3 millimeter inner diameters, we're able to accommodate columns all the way up to 15cm in length.

So unfortunately, we don't have a cartridge two meters long to accommodate some of those monoliths. But we do have, you know, the ability to use the majority of commercially available columns. And we're also able to heat them up to about 80 degrees C, which enables high temperature workflows if you're, you know, doing, you know, high temperature or biopharmaceutical compounds that require temperatures to be that high.

Now, the big question is, what if I can't find a column on that list? And I think there's no one better to answer that question than Dwight Sowell.

Dwight Stoll

All right. Thanks, Sam. So, I think, there are a number of different ways one could approach this question of how do I find, a suitable column of similar selectivity, that meet the requirements that I have for specific dimensions.

But, what I'm the approach that I'm going to talk about here briefly is, using the so-called hydrophobic subtraction model of reverse phase selectivity. And the reason this is really powerful here is because we we have today a database of selectivity, parameters for columns that stands at about 800 columns. That's freely available. So that's, a really a, freely available tool that anybody can use, sort of in this space.

So, what I find is that, the community is kind of split into two groups. Either there are people who really use this model a lot and know a lot about it. And there's a large population of people who've never heard of it before. So before we get into actually looking at the data, I just wanted to go over some of the basic ideas here so that we're all on the same page.

So this hydrophobic subtraction model, really started life about 25 years ago. These three gentlemen here were some of the key players early on in the early 2000’s Following some what Lloyd Snyder called trolling of retention data. He thought, oh, there really could be a chance here to build a selectivity model that could be used for, finding columns of similar selectivity.

And so the development phase, really persisted through the early 2000s there. And it was a highly collaborative effort, involving USP, PQRI column manufacturers and some others stakeholders in the industry. Then around 2010, Lloyd and John said, yeah, we're going to retire soon. So we'd like somebody to sort of take over this project and so at that time, we transitioned the characterization of new columns using this model to my lab here at Gustavus Adolphus College in Minnesota.

And we've been, doing that for the past 15 years. So as I said, the database currently stands at about 800 columns. Actually, we just have some new data today. We're going to flip over 800 here shortly, and it's growing pretty consistently at about one new column every two weeks. So next slide. Yeah. So, this is going to be a really brief, introduction and discussion of it here.

But for people that are interested in knowing more or just pointing to two resources here, if you have a short time and you, you want to know just a little bit more than what I'm saying, that I point people to this LCGC article on the right, published about ten years ago. It's a ten or so page read, that, like I said, gives a little more depth and kind of gives an update on what had been going on with the model at that time.

If you have if you want to know a lot more and you have a lot more time to spend that, I then I'd refer people to the book chapter indicated on the left, which is just, tremendously rich resource, about selectivity, reverse selectivity in general, but also specifically this model.

So just a little bit about how the characterization works. So the model was initially developed, using about 90 different, small molecule compounds and, ten type B, pure or pure silica, C18 columns, and the, the conditions really since day one pretty much have been, acetonitrile and phosphate buffer.

So 50/50 mix 35 degrees. And we do work at both pH 2.8 and 7. Since that initial development, the site set has been reduced to 16. So that's the current characterizations or protocols speak about in just a moment here. But the hydrophobic subtraction model HSM for short. What it, what it attempts to do is, is quantify selectivity where selectivity is defined.

Here is the retention factor of any compound relative to the retention of ethyl benzene. That's the that's what's in the log term here as a function of five solute column parameter pairs. So and these are kind of cartoons that try to reflect the chemistry that we try to capture in these terms. So the η’H term is about hydrophobicity.

The σ’S term is about solute size more or less. The β’A and α’B terms are about hydrogen bonding between the analyte in the stationary phase and the κ’C term is about electrostatic. Or you could say, ion exchange interactions between the analyte and some charge functional group or property of the stationary phase. Next slide. So the process, what we actually do to characterize columns that we receive is first, at pH 2.8, we measure the retention times of 16 probe compounds.

You know, we have three bases like nortriptyline, amitriptyline, berberine. We have, few, carboxylic acids, like butylbenzoic acid. We have a handful of, uncharged molecules like toluene, ethylbenzene, and things like that. Those are the probes. And, then we switch to pH 7 and we only measure the retention of berberine there. We then take those, retention values and we do a multiple linear regression against, the solute parameters for the 16 probe compounds that have been established a long time ago. And then finally we, calculate the C parameter, ph7 using, the retention of berberine. And so what we compile into the database are the resulting H/S/A/B/C parameters that reflect the contributions to retention shown in the cartoons on the previous slide.

Next. So when it comes to actually comparing columns in terms of, similarities or differences in their selectivities, we use this so-called similarity factor, F_S. Here is the symbol we use. And what it is, is essentially a distance in, in five dimensional space between the two columns. So we can’t represent five dimensions here. So what this, picture shows is three dimensional space and calculating the distance between two columns in this three dimensional space.

So the green dot and the green square here represent two different columns. And so their H, S and A parameters define the point in space where that column lives. And the F_S value is the square root of the sum of the squares of the differences between each parameter pair. So H₂ minus H₁, square that, so on and so on.

For S, A, B and C then we take the square root of the whole thing. So the math is really simple, but conceptually this is, this is what it is. And there are some weighting factors that, that attempt to, let's say give equal weight to hydrophobicity and ionic electrostatic interactions, hydrogen bonding and things like that.

Because if we didn't do that, the hydrophobic term would really dominate things. And from a selectivity point of view, that's not so interesting. And you can see more on the website I’m going to show you in a moment that you can adjust these weighting factors. If you want to do that. The big takeaway from this particular slide is that we have this threshold of three.

Anytime we do a column comparison and we calculate that the F_S factor is less than three, then we say we declare that for functional practical purposes the two columns are equivalent. Of course, they will never be exactly equivalent. There are going to be small differences. But from the point of view of practical separations, this is kind of the benchmark, the threshold that we use, anything less than three equivalent and bigger than three, some degree of difference.

Okay. Excellent. So, you can access the database that I'm referring to here. Data for 800 columns, H/S/A/B/C parameters. Two different ways. One is through the USP website. The URL is here. And, I think we if you can advance one more Sam. So, what this what is shown in this slide is a comparison of the top column here, the ACE 5 C18 to everything else in the database, sorted according to the F_S factor.

So what this table tells us is that out of the other 700 columns, this Develosil ODS-UG-5 column exhibits the most similar selectivity to this ACE 5 C18 column. And you can see, kind of rank ordering there. So there are looks like six, six columns that have SF factors less than three here. And, and should exhibit, similar selectivities to our target column here.

Next slide. So if we look at this in chromatograms these are the first three entries in the table, we just look at these chromatograms and you know, qualitatively they look very similar. Yes, there are small shifts in where the peaks are. But in terms of actually resolving this mixture, any one of these columns would do fine. Even for the critical pair.

There are peaks three and four. Next slide. So the other way to get at the database is through this website that, that my group maintains www.hplccolumns.org. This is just a screenshot of the landing page. And if you go to the next slide, now we bring it to the question we're really trying to address here, which is this scenario at the bottom. I'm running a separation on a Kinetex C18.

I want to go to a 75 micron column. But the table and Sam's paper tells me that that particular option is not available. So what alternatives, are there that will give me the same selectivity, but in a different column dimension. Let's go to the next slide. So I look again at Sam's table and I see, oh, Agilent and some other vendors make, 75 micron i.d. columns.

You can pick any of those. And so now I can go to the database and say I'm looking to replace this Kinetex C18. So I select that column in the top there. And then I go to the table and I see oh about number eight or so in the list there with an F_S factor of 2.5 is the Zorbax Eclipse Plus C18 column and all of the columns in the table here are sortable.

And you can also filter on anything. So I could say I could have made my life easier here by filtering on Agilent columns if that's what I was looking for in particular. And if I did that, it looks like that would have been the number one hit in the list, because that's the top Agilent column in the table there.

So that's the basic idea. I think HSM is a really powerful, tool for making these kind of comparisons and we've just given you a little demonstration here of how you would actually, use it as a tool to answer the question. That's relevant here today.

Samuel Foster

Awesome. Thank you. That was excellent.

So we'll wrap up with our conclusions. So first and foremost, our capillary scale columns offer very comparable sensitivity, pardon me, comparable chromatographic efficiencies when compared to, analytical scale. There are a number of different column formats that also can be used at the capillary scale, including monolithic open tubular and pillar array. There are considerable commercial offerings, for a number of different phases of capillary scale columns and nearly all chemistries and geometries.

When a phase is unavailable, feel free to check the column selectivity databases that will provide great insights into, you know, equivalence or near equivalent columns that will point you on the right track, and that the hydrophobic subtraction model as a whole offers great insights into column selectivity across, a wide variety of different chemical properties.

For those who have been following the series and want to get a little bit more involved, we have started a LinkedIn group. The Capillary Innovation Forum. Feel free to join, take a look, ask questions. We have a number of experts in the field there who, are certainly interested in talking with you and helping to spread the word of capillary scale.

Here is our references slide. Finally, I'd like to thank Doctor Dwight Stoll for taking his time today. And I'd like to thank the audience for coming out and listening. And with that, I think we'll open the floor for questions.

Alright. Questions are starting to filter in here. I have a good one for those who've worked mostly with packed columns. What trade-offs should we expect when trying monolithic or open tubular formats? Yeah. So monolithic formats, traditionally you're going to see, reduced back pressures, by allowing you to, to move much faster, operate at higher linear velocities.

For open tubular formats, you'll typically see higher efficiencies, but much, lower, you know, load ability. And so those are definitely things to be aware of. This is a great question. What effect do nonspecific bindings have on capillary columns? Yeah, it's a very great question. So the answer is they have, you know, similar impacts.

I haven't necessarily gone about and quantified them. That being said, when you're working with something like, a, you know, protein, you will see, you know, binding and may have to use passivated columns or, may have to do a number of runs to actually coat the column in any surfaces. But after that, you'll see very similar performance again to, to analytical scale.

This is one for you, I think. Dwight. Can the column selectivity, but pardon me, can the column selectivity database be used for identifying orthogonal columns in 2D-LC applications?

Dwight Stoll

Yeah. So the as I said, the table, when you use the column comparison tool, the table is sortable. So what we were talking about today is trying to identify columns of similar selectivity to, trying to identify a kind of drop-in replacement.

But you can, just with a click of a button, do the inverse sort on that F_S factor and, reorganize the table so that the columns at the top of the list are the most different, from the target column, at least as measured by this, similarity factor. And, you know, that's going to return, a lot of possibilities.

And so the filtering tools, or filtering capability in the table is really effective for certain narrowing things down to some possibilities. But yeah, that's, it's a good question. And, another good use of the data.

Samuel Foster

Awesome. Thank you. We have another one here. What are the compatibility requirements for capillary columns from various manufacturers with the Axcend LC column cartridge?

Yeah, there's very few. So I think first and foremost it has to be less than 15cm. You can have the column poke out the back of the cartridge. Absolutely. So if you did want to use that two meter long column, go for it. Just made not, you know, look, it's pretty, that being said, basically all you really need is the proper fittings in which we have, you know, a number of different fittings that allow for connection with the standard, you know, 1/16, 1/32 and 360 OD formats.

And beyond that, you know, purchasing a column that is 15cm or less and, hooking it in. It's a very simple cartridge that was designed to work just the same way it works with analytical scale.

All right. I see another question here. When scaling a method down to capillary dimensions, what's the most common challenge people run into? And how do you usually overcome it? Yeah. No, that's a great question. I think probably the most common, challenge is finding the proper column, from there, it's just math to make sure you have, you know, a comparable flow rate and a comparable gradient.

That being said, the other major factor is the limiting of extra column band broadening. And in fact, that's what we'll be covering next time is, is methods of reducing extra column effects to your efficiency. In order to, you know, maintain high quality and high efficiency separations.

Oh. This is a good one. So Dwight, has the theory of the hydrophobic subtraction model seen any changes since inception?

Dwight Stoll

Yeah. So, that's a good question as well. So we, my group, in collaboration with Sarah Rutan, from Virginia Commonwealth University, have been, looking at this in a few iterations. I would say for the first, maybe 20 years, there wasn't a whole lot of 20 years of life in the model.

There wasn't a whole lot of, changing of anything. But what we did around five years ago was basically ask the question. Okay. Originally the HSM then was developed using C18s pretty much. What if we think about refining the model, using all of the data that had been acquired since the original development, which included a lot of data on, you know, PFP, phenol, various polar embedded phases and things like that.

So, we did that. We we referred to that model as HSM 2, which, involved, adding a couple of parameters. And, we think that, kind of, let's say tightened up the model in terms of its ability to describe selectivity for things that are not simple alkyl silica phases. More recently we also talked about an HSM 3, which was based on, fairly large, much larger data set than we've been working with historically with just the 16 Pro molecules.

And, we were able to do that by accelerating the retention measurement. And that allowed us to sort of open up the scope, to about 90 molecules. And I, you know, we're continuing to work on that and I'm, I think the future will hold some more developments in that space.

Samuel Foster

Perfect. All right. I think probably we'll have time for 1 or 2 more.

I see one here. Have I worked with MEMS columns? I have not, actually. And you mentioned taking it offline. Please shoot me an email. I'd be curious to see what you have working on that. That would be something interesting. I see another one here. This one should be for you. How can vendors get new columns added to the column selectivity database?

Dwight Stoll

Yeah. Good question. Just, shoot me an email. Typically we, historically, the column dimensions we used for the characterization have been 150 millimeter length and 4.6mm diameter. But some of the newer, columns, typically are made in those dimensions. So we can really work with anything that's just been, the conventional dimensions.

But go ahead and shoot me an email. And, we can tell you a little bit more about our timeline and things like that. And, we always, kick the data back to vendors before we publish it to make sure, everything, you know, there's no unexpected surprises there. But, yeah, just get in touch with me and we can work it out.

Samuel Foster

All right. Awesome. All right. Thank you very much, everyone, for attending. Thank you again, Doctor Stoll, for taking the time out of your day to come talk about this. If anyone has any questions, feel free to visit us. And, anything else you want to add before we go Dwight?

Dwight Stoll

Nope thanks for. Thanks for having me. It's been fun.

Samuel Foster

Perfect. Thank you so much again. All right, everyone, bye.