Analysis of the State of the Art: Liquid Chromatography Column Technology

In honor of LCGC's celebration of 30 years covering the latest developments in separation science, we asked a panel of experts (listed in the sidebar) to assess the current state of the art of liquid chromatography (LC) column technology, and to try to predict how the technology will develop in the future. This article is part of a special group of articles covering the state of the art in sample preparation, gas chromatography (GC) columns, GC instrumentation, LC columns, and LC instrumentation.

The Size of Superficially Porous Particles

The first question we asked our expert panel was whether they foresee any limits to the decrease in size of superficially porous particles (SPPs).

There should be no limits in terms of synthesizing the particles themselves, points out Ron Majors, a senior scientist at Agilent Technologies and the longtime editor of LCGC's "Column Watch" and "Sample Prep Perspectives" columns. It has already been demonstrated, he notes, that both smaller and larger SPP particles can be prepared with various shell thicknesses and pore sizes, and that they are robust, can be derivatized with the popular bonded phases, and can be successfully packed into most column configurations (albeit with a bit of work).

The question, then, is whether smaller particle sizes for SPPs are desirable, given the higher pressures generated, the real possibility of frictional heating, and the instrumental considerations of extracolumn effects. "Perhaps a new generation of low dispersion instruments may be required for smaller SPPs," he considers.

And of course, as several experts point out, making these particles smaller would eliminate the advantage of being able to use them on regular LC equipment to achieve efficiencies close to those of ultrahigh-pressure liquid chromatography (UHPLC). Thus, the panel almost unanimously agreed that the current size of SPPs, with most commercial SPPs at 2.7 µm and experimental SPPs between 1 and 2 µm, is likely to remain where it is.

And as Jack Kirkland of Advanced Materials Technology notes, particles smaller than 1 µm likely are not practical, and particles with diameters of <2 µm already are problematic, requiring very high pressures that tax instruments, decrease reliability, and increase operating costs. "Such small particles are only of value for fast separations of samples with limited components," he says. "How many users require <10-s separations rather than 30 s? Complex samples require larger particles, longer columns."

Joe Glajch of Momenta Pharmaceuticals also notes, "Additional gain in efficiency is probably not the major direction for column particles in the future; selectivity will become more important."

And there is indeed a lower limit to the size of SPPs, says Georges Guiochon of the University of Tennessee. The pressure required to achieve the optimum flow rate, he says, increases as the square of the inverse of the particle size at constant efficiency, for a given compound on a given system.

This, he says, becomes practically impossible for the analysis of low-molecular-weight compounds (such as pharmaceuticals and peptides) on columns packed with particles finer than 1 µm. However, Guiochon continues, the analysis of large biochemicals on columns packed with 0.5-µm particles would be possible if particles that size (but not smaller), with large mesopores could be manufactured and packed in efficient columns.

David Hage, of the University of Nebraska, agrees. "Pore sizes of 100 Å or less work well for small molecules, but pore sizes of 300–500 Å or even larger are needed to provide suitable access if the same kinds of supports are to be used for biomacromolecules," he says.

"The limit might also depend on the column diameter desired," adds Guiochon, "The efficiencies of wide-bore columns (>4.6 mm) and of capillary columns (<0.1 mm) tend to exceed that of narrow-bore (0.5–2.1 mm) ones."

Operations at Very High Pressure

In a related question, we then asked our panelists if they foresee a move to high performance liquid chromatography (HPLC) instruments operating at even higher pressure in the future, to increase speed or take advantage of ever-smaller packings.

LC Column Technology Panel of Experts

A few like the pursuit of higher pressure, at least to some extent, noting that higher pressures have provided excellent results so far.

"There are many applications in which UHPLC has been used to achieve better separations," says Hage. "One example has been to combine UHPLC with mass spectrometry (MS) to look at mycotoxins in food."

But most are not convinced that going to much higher pressures makes sense, both in terms of need as well as the costs and complications involved. "I think the current operating pressures are more than sufficient for the vast majority of problems in HPLC today and in the future," says Glajch.

Chris Welch, of the pharmaceutical company Merck, also points out that a lot of the recent improvements in HPLC performance can be attributed to simple adjustments like tuning up internal plumbing connections and reducing the volume devoted to mixing. "If higher pressure means higher complexity and cost, I would rather see simple solutions that improve performance while decreasing cost," he says. "I think vendors should give some thought to selling at the low end of the market, expanding the overall base of HPLC users to include things like microfluidic devices operating in medical and dentist offices, perhaps even in the home."

Stationary-Phase Chemistries

A number of our experts see the need for various types of new stationary-phase chemistries.

"When it comes to regular LC columns, I think that there must be more than reversed-phase-LC, hydrophilic interaction chromatography (HILIC) and the various ion exchangers," says Rainer Bischoff of the University of Groningen. "Orthogonality of separations is a critical attribute to getting a more comprehensive view of very complex samples. There was a time when mixed-mode stationary phases as used in solid-phase extraction were popular. Maybe there should be a revival of such specialty columns for high-efficiency LC."

Other areas for development that were suggested include new phases for chiral and biochemical applications, achiral phases that improve separations for closely related isomers and analogs, improved columns for oligonucleotide separations, and selective adsorbents for on-line extraction.

For supercritical fluid chromatography (SFC), says Welch, the "holy grail" is a column that would give comparable generality to C18 reversed-phase separations. "This would enable a shift to SFC for carrying out the routine mass-directed, small-scale purifications that support high-throughput synthetic chemistry," he says. "Until generality is comparable with reversed phase, it doesn't make sense to switch to SFC."

Majors predicts that for both HILIC and chiral separations, new phases will be developed that are more universal. For intact biological molecules, he feels improvements in biocompatible phases, especially for faster separations, would be of interest. "Non-silica-based superficially porous packings could fill this need," he concludes.

Nevertheless, he doesn't foresee a significant change in the landscape of stationary phases. "For the last three decades, for nonchiral small molecules, reversed-phase chromatography has been dominant as the preferred HPLC mode and I don't see this changing," he says. "Neutral, ionic, and ionizable compounds can all be separated using this mode."

Glajch sees the overall trend moving toward fewer stationary phases, not more, as stationary phases with overlapping selectivity are weeded out. "Some new phases are warranted, to fill in the gaps in current offerings, but many existing ones will likely be removed from use due to duplication," he says.

Kirkland believes that separation scientists should focus more on mobile-phase selection to obtain selectivity differences. "This technology has been exhaustively researched but largely ignored, even though it is much more powerful than stationary-phase selectivity," he says.

Monoliths

Monolithic columns offer a number of advantages, such as reduced back pressure and a lower C term at high linear velocities. Nevertheless, they have faced a number of challenges, and have yet to gain wide acceptance.

Guiochon outlined three major challenges faced by monoliths:

First, the maximum inlet pressure (currently, ~200 bar) is limited by the need to clad them with a plastic material. Improvement would require that the rod be prepared directly in the mold, which seems impossible. Second, radial mixing in these columns is poor, for fundamental reasons; the through-pores are more or less parallel to the column axis. This makes their efficiency most sensitive to radial heterogeneity, itself resulting from the presence of the column wall. And third, increasing their efficiency requires making their domain size smaller, which is inconsistent with high permeability.

Another problem, notes Hernan Cortes of H.J. Cortes Consulting and the University of Tasmania, has been difficulty constructing highly reproducible columns.

One of the reasons for the slow development and adoption of monoliths, both Majors and Kirkland point out, is that the technology of silica-based monoliths is tied up in patent protection, which precludes additional companies from pursuing it.

Nevertheless, say several panelists, recent research into monoliths shows some promise. For example, Majors points out that new-generation silica monoliths show reduced tailing, and that monoliths in stainless steel housings, which may enable longer columns, have become available.

Research on polymeric monoliths, furthermore, has shown that they can be useful not only for macromolecules but for smaller molecules too. "The kinetic plots made by Gert Desmet of the Free University of Brussels have shown that monolithic columns may be best for difficult separations requiring extremely high theoretical plates but at the expense of long separation times," Majors notes.

Yet no one envisions that monoliths will ever compete with packed beds for "standard" separations. Instead, they agree, monoliths will likely be valuable only in niche applications such as the separation of macromolecules, in high-throughput separations as a lower-pressure alternative in UHPLC, and in capillaries and chips, where there are advantages in cost and practice.

"I suspect that 'polymerized in place' or 'assembled in place' columns within microfluidic devices will become increasingly important over the next few years," adds Welch.

Alternatives to Silica Gel Supports?

The characteristics of non-silica gel supports like zirconia, titania, and polymers — such as pH resistance and temperature stability — continue to make these stationary phases attractive, notes Cortes.

Welch agrees that materials with mechanical properties similar to silica but with improved chemical inertness would be highly desirable. "I'm not sure how far we are from such materials, but I suspect that we will see an increased offering of ordered porous materials with very attractive physical and mechanical properties in the not-too-distant future," he says.

But the performance of silica gel is well demonstrated. It can be manufactured with high reproducibility at an industrial scale. It shows good flow properties, can be obtained in narrow particle size distributions, is rugged, is reasonably priced, and can be easily modified. As a result, the panelists say, it's difficult to develop another material that can compete with silica.

"Polymer-based materials surely have their place in preparative chromatography, but I do not believe that they will be a serious competition to silica for analytical LC," says Bischoff. "Metal oxides such as zirconium oxide may fill specialty niches, but will not replace silica to any extent."

Guiochon agrees. "Many smart and ingenious people have tried many things, but they do not have much to show for their pains," he says. "Only the inorganic–organic hybrid materials prepared by Waters and Phenomenex are widely accepted."

Packings for Preparative-Scale LC

We also asked our panelists for their views on the differences between preparative and analytical packings. A number of differences were noted.

Preparative chromatography has the primary goal of maximizing relatively pure sample amounts and recovery, notes Cortes. As such, operation under overload conditions is a common practice and requires attention to separation conditions that are different than those used in analytical sizes. "Efficiency tends to be less important than capacity," he sums up.

Bischoff sees promise for polymer-based packings in preparative chromatography, where their ruggedness is helpful for the cleaning-in-place procedures often needed in preparative applications.

Pressure limitations are key in preparative columns, notes Guiochon: the larger the columns, the smaller the pressure that can be used. As a result, he says, particle sizes will remain larger than 5 or 10 µm in most preparative applications.

Preparative columns will always be dominated by production cost constraints, comments Welch. Guiochon agrees, noting that costs will keep sophisticated stationary phases out of the preparative market.

Welch notes, however, that until recently, analytical columns were also limited by the cost of materials going into column production. "For example, expensive precursors or prolonged synthetic sequences are simply not cost effective for 250 mm × 4.6 mm columns," he says. But as microcolumn technologies have gained favor, the reduced packing material requirement has made expensive stationary phases more cost effective and has enabled access to new materials for specialty analytical applications. "This also suggests that prices should drop for microcolumns containing standard stationary phases," he concludes.

The Impact of MS Detection

The advent of MS as the preferred detection method for HPLC certainly has affected LC separations. MS detection has been a factor in the drive to faster and more efficient separations. It also has been the principal driver for nanoLC, and for the trend to use shorter columns (and high-speed separations).

The latter, says Welch, is likely to continue, and more development is needed. "We like to use the MISER (multiple injections in a single experimental run) format with very short (1-cm) columns operating at high linear velocity, and suspect that the development of well packed, short columns will become increasingly important," he says. "Faster autosamplers are also needed."

Another impact of MS detection, points out Majors, is that column bleed of bonded phase has come into the forefront, especially with some of the fluorinated reversed phases. (With UV detection, these bleeding phases could not be seen). "As a result, manufacturers are developing low-bleed MS-compatible phases," he says.

The phenomenal resolving power of MS also decreases the burden of chromatography to provide complete resolution of all components, notes Welch. "Chromatographic separation is still required to remove matrix components that would otherwise interfere with accurate quantitation, but in some cases, minimal chromatographic separation coupled with MS detection can provide the very fast analysis times required for high-throughput analysis," he says.

Guiochon, however, regrets the fact that because of the power of MS, many scientists pay little attention to the importance of good separations or how to achieve them. "Unfortunately, MS people use modern high performance columns the way my graduate students used columns in the 1960s: They connect high efficiency columns to ion sources through long, broad tubes where bands broaden and the column efficiency is just destroyed," comments Guiochon. "They have got to learn that they should listen to us."

On the other hand, HPLC has also revealed limitations in MS: The peaks coming off high-performance columns are too narrow and fast to be properly handled by most MS systems.

"MS will continue to drive column development, and UHPLC is also driving MS when it comes to faster scan rates and shorter dwell times," summarizes Bischoff. "LC and MS will continue to develop together and drive each other."

The Impact of Fast and Multiplexed Screening

We also asked the panel how fast and multiplexed screening has changed the demands for different column sizes, geometries, and packing materials. Most agree that these techniques are driving the development of short (for faster elution), narrow (for lower eluent consumption) columns packed with fine particles (for higher efficiency and shorter elution time).

"The application of 'ballistics' chromatography for extremely fast separations at modest pressures has required core–shell particles in short (2–3 cm) columns of narrow internal diameter (such as 1 mm) so that very high mobile-phase velocities can be used," says Kirkland. "Larger core–shell particles are best for this practice."

Majors is curious to see how these methods develop in the future. "It will be of interest to see if very fast separations of very complex samples such as protein hydrolysates or natural products will require the use of 2D instruments for full comprehensive analysis and how column systems will cope with these complex samples," he says. "Perhaps a new breed of truly orthogonal columns will be required."

Welch, in the meantime, is exploring what happens when these methods are pushed to the limit, and foresees many sub-1-min separation times in the near future. "We recently studied a group of standard test analytes for chiral chromatography, and just for the fun of it, we tried to push to the fastest possible separation," he says. "We were able to get all separations in less than 1.5 min, in each case tying or breaking the previous speed records. Why? The principal reason is that previous researchers weren't really that motivated to develop fast separations."

New Particle Technology?

We also asked our panelists if they see any new particle technology on the horizon. A number of ideas emerged.

Welch suspects that self-assembled porous coatings within microfluidic capillaries may become important.

Cortes, in turn, is intrigued by various recent developments, particularly shell-on-shell technology being developed by Peter Myers at the University of Liverpool, and monolith modifications, such as addition of particles, pioneered by Emily Hilder at the University of Tasmania.

Majors foresees that comprehensive multidimensional LC — particularly once the method is more widely used outside of academia — may require newer types of column configurations and truly orthogonal stationary phases.

Bischoff believes that the sub-1-µm particles being explored by Mary Wirth at Purdue University could bring about a new era of efficiency in LC. "However, the technical challenges are significant on all levels," he acknowledges. "We'll have to wait and see whether they can be tackled."

New ways of making LC packings may generate improvements in efficiency that were unheard of just 10 years ago, Majors notes. "Don't count out silica-, silica-hybrid-, and polymer-based monolithic columns as academics and industry work to further improve these interesting separation media."

In Guiochon's view, the only progress needed is to develop particles with a higher thermal conductivity, to reduce the heat effects that plague analyses performed with long, wide-bore columns packed with fine core–shell particles. The best candidate would be to use a solid alumina core instead of silica, he believes.

But, he says, the real problem with modern fine particles is their packing procedure. "It has proved to be very difficult for all manufacturers to develop a suitable procedure," he says, and wonders if it could still be improved.

Guiochon would also like to see instrument changes that reduce band broadening. "I suggest that manufacturers integrate the whole LC system — the sample injection device, the column, and the detector — into one subunit," he says. "Or if that is impossible, I suggest they design it in a way that minimizes the extracolumn volumes and reduces the length and diameters of connecting tubes."

Acknowledgment

I would like to extend a special thank you to Daniel W. Armstrong for serving as the chair of this LC Column Technology section of our special coverage for the 30th anniversary issue of LCGC. Armstrong is the Robert A. Welch Professor of Chemistry and Biochemistry at the University of Texas at Arlington, and a member of the Editorial Advisory Board of LCGC.

Laura Bush

Editorial Director, LCGC North America