June 1, 2013 (Vol. 33, No. 11)

In 1964, University of Utah Chemistry Professor J. Calvin Giddings enunciated a theoretical platform that would provide liquid chromatography (LC) with resolving power equivalent to that of gas chromatography (GC). “Unified separation science” became a buzzword soon after.

What Giddings envisioned, in the words of Waters scientists Christopher Hudalla and Patrick McDonald, was the “higher mobile phase diffusion and efficiency” of GC, and the “higher selectivity via orthogonal modes of separation” from LC, which could be achieved by combining supercritical carbon dioxide as the major mobile-phase component.

Supercritical carbon dioxide is an inexpensive, low-viscosity, highly compressible, environmentally benign solvent that improves chromatographic efficiency for a given stationary-phase particle size and linear velocity.

By contrast, conventional high-performance liquid chromatography (HPLC) solvents are toxic and expensive, in both their acquisition and disposal. Aside from water, acetonitrile is arguably the most used solvent in HPLC. It has become prohibitively expensive in recent years; users pay “both ways” for the luxury of using the overwhelming majority of solvents. A four-liter bottle of HPLC-grade acetonitrile costs between $300 and $400. Then, because acetonitrile is generally not recycled due to its high boiling point, disposing of spent mobile-phase diluted with buffers and other organic solvents easily doubles the cost of using the solvent. The carbon footprint for manufacturing acetonitrile is substantial as well, both in terms of atom economy, purification, and shipping.

Supercritical carbon dioxide is the greenest HPLC solvent, hands down. The carbon dioxide is extracted from the atmosphere, to where it is vented after use. Greenhouse gas contribution is zero.

Waters Acquity UPC2 System simplifies laboratory workflow, separates structurally similar compounds, and provides an orthogonal mode of separation to HPLC/UHPLC for full characterization of samples, the company reports.

Convergence Chromatography

Waters has named its new-generation supercritical fluid chromatography (SFC) system UltraPerformance Convergence Chromatography, or UPC2 ®.

“It’s the combination, or convergence, of GC and LC,” says Ken Fountain, director of chemistry applied technology and global UPC2 applications.

Convergence chromatography exploits the advantage of carbon dioxide-based mobile phases in either supercritical or subcritical mode. UPC2 combines the unique properties of carbon dioxide mobile phases with the ability to run gradients with common organic solvents, like methanol or acetonitrile, on stationary phases that have both normal- and reverse-phase characteristics, to cover a wide selectivity range.

Like other LC systems built on Waters Acquity® platform, UPC2 employs sub-two-micron stationary phases, especially for nonchiral analysis. It is possible to use five-micron particle technology as well, as with UHPLC, but also like conventional LC, efficiency increases as particle size decreases. Efficiency improvements from smaller particles are in fact comparable to those obtained in HPLC/UHPLC. UPC2’s additional benefit is much higher linear velocity and throughput.

“Mass transfer kinetics are greatly improved. Diffusivity is closer to that of GC’s, between 10 and 100 times faster than that of LC,” Fountain says.

Waters’ original target market was chiral analysis, where SFC already enjoyed a healthy reputation. The company has demonstrated chiral SFC separations that require one-thirtieth the time compared with standard normal-phase HPLC. Solvent consumption and lower cost are additional benefits. In a typical example, the SFC method consumed 135 microliters of methanol, compared with 10 mL of hexane/ethanol for HPLC. Overall cost savings are dramatic, falling from about $6 for a conventional HPLC run to $0.05 for UPC2.

“When you consider this savings on the scale of hundreds to thousands of injections, the financial impact to an organization can be quite exceptional,” says Waters’ John van Antwerp.

Since its selectivity overlaps significantly with normal-phase chromatography, SFC is orthogonal to reverse-phase LC. The technique is applicable to a diverse range of compounds, including most organic-soluble compounds, most salts of organic acids and bases, strong organic acids and bases, small lipophilic peptides, and nonpolar solutes (e.g., waxes and oils). In addition to being the go-to method for chiral separations, SFC also separates positional isomers and diastereomers, and is compatible with most popular detection modes.


Waters’ 2004 introduction of UPLC (the company’s branded version of what later came to be generally referred to as UHPLC) was met with some pushback. Users were concerned about the extraordinarily high operating pressures, the cost of new HPLC systems, and method transfer. Eventually, customers recognized UHPLC as an extension of HPLC—a way to exploit the benefits of sub-2 micron stationary phases. The pushback with UPC2 was based on the perception that SFC lacked robustness, was not easy to use, and was less-than ideal outside of chiral separations.

Adapting supercritical chromatography to everyday analytics required more than simply pumping supercritical carbon dioxide through an existing instrument. Fountain makes the analogy to the early development of UPLC. Making minor adjustments to existing columns or instruments while employing 1.7-micron particles with large diameter columns would have been easy. “But we knew from practice and theory that the advantages of sub-two-micron particles would disappear. That was the rationale for redesigning columns, hardware, end fittings, and the instrument itself. That is how the Acquity UPC2 System came to be.”

Waters has gone out of its way to differentiate UPC2 from conventional SFC. Both use carbon dioxide-based mobile phases, and not without justification. “Convergence chromatography brings the SFC practice into mainstream analytical techniques,” Fountain explains. The most formidable hurdle to “instrumenting” SFC in the form of UPC2 was to make the new platform as robust and reliable as traditional analytical chromatography.

“Laboratory scientists know deep down that GC and LC are robust. They trust those technologies. We wanted to bring that same level of trust to SFC,” he says. “We’re not trying to replace GC or LC. But we are finding successful application areas we did not anticipate during development, for example in the areas of semiconductors and organic light-emitting diodes.”

Origins of an Instrument

Years ago, while managing a unit at Merck & Co.’s New Technologies Review and Licensing Committee, it occurred to Christopher Welch, Ph.D., that SFC had not kept pace with HPLC, either technologically or in terms of features. A fully up-to-date SFC system was therefore added to the group’s “wish list” for analytical instruments.

“We were already using analytical SFC,” Welch explained. At that point, SFC had reached the proof-of-principle stage, and was becoming the go-to method for the analysis of chiral compounds. Thus its utility with chiral pharmaceutical intermediates and active ingredients. “But the current standard just wasn’t up to snuff,” Welch adds.

He observed the activity in HPLC to improve interfaces, streamline methods, and run separations faster and with higher sensitivity and robustness. “HPLC was getting better all the time, and we knew those same improvements could be applied to SFC, but nobody was interested in implementing them. It was time to remove the bandages,” Welch says.

Partly through Welch’s suggestions, Waters took on SFC technology through its 2009 acquisition of privately held Thar Instruments, which was then the world’s most advanced manufacturer of SFC systems.

Merck eventually became a beta test site for the Waters instruments that later evolved into UPC2. “We got an early version of the system, ran it through its paces, and provided early feedback. Waters has a mechanism for reaching out to customers for input, which is smart,” Welch says.

SFC’s main claim to fame is its handling of stereoisomers, particularly enantiomers. Welch is something of an expert on chiral separations, having studied in the lab of Prof. William Pirkle at the University of Illinois. Pirkle invented the first commercial chiral stationary phases that interact differently with enantiomers (mirror-image isomers).

Chiral interactions with columns tend to be weak, which is why baseline chiral separations are difficult to achieve. “The early columns provided enantioselectivity, but their efficiency was poor, making the separations difficult to see,” Welch explains. “With the needle-sharp peaks provided by modern columns and instrumentation it’s possible to see even tiny differences.”

Under SFC, this translates to baseline complete separations where one might have observed shoulders, and much faster operation. “In the old days you’d be lucky to develop a chiral method for half of your samples. Now, with SFC, you can do 95 percent,” Welch says.

Difficult Applications

Metabolomic science involves the analysis of broadly diverse molecules, many unknown, with a concentration dynamic range as high as 1014. Both separation and identification pose tremendous challenges, says Prof. Vladimir Shulaev of the University of North Texas. “You can’t put a name on every peak, on every ion. Even mass spectrometry is incapable of de-convoluting some mixtures.”

Identifying compounds through mass peaks alone becomes difficult because masses lack uniqueness: Compounds form adducts, and many compounds break up within MS to identical fragment ions.

Because spectral libraries tend to be instrument- and platform-specific, even that avenue is not always fruitful, particularly for LC-MS. (GC-MS provides more predictable, reproducible fragmentation.) Positive identification is normally achieved by acquiring a reference standard for both chromatography and MS, but not all are commercially available, and isolating or synthesizing them takes time. Metabolomics researchers, Dr. Shulaev says, must often generate their own custom spectral libraries.

Dr. Shulaev’s specialty within the field is lipidomics for which SFC is especially advantageous compared with LC or GC. One advantage of SFC over more conventional separations is lower sample preparation requirements.

“We take the lipid extract and shoot it through directly,” Shulaev explains. SFC might provide similar benefits for other molecular classes, he adds.

One need only consider the structure of glycerin-conjugated lipids to understand the analytical tribulations of lipidomics researchers. Glycerin esterifies with three, two, or one fatty acids . Acids differ in ways that complicate analysis, for example in degree and location of unsaturation. Reconstructing even a simple triglyceride—including the position on the glycerin blackbone, carbon number, and location/orientation of double bonds—tests current analytical instrumentation.

Considering SFC’s position within the GC-LC continuum, its potential to provide analytical orthogonality—a type of missing link in the analyst’s tool chest— becomes palpable. For example, SFC handles both non-volatile compounds (typically LC’s domain), and volatile species (GC’s forte). Molecules requiring derivatization to pass through a GC column elute under SFC easily.

“SFC provides greater bandwidth to our arsenal of analytical methods,” says Gerard Rosse, Ph.D., associate director at Dart Neuroscience. “It expands our ability to purify compounds and to simplify method development to enable purifications.”

Having two orthogonal methods facilitates rapid method development, Rosse adds, particularly for separating complex mixtures.

“SFC represents the future, and I’m expecting it will gain broad adoption over the next two or three years.” Theoretically, SFC is expected to be up to 10-times faster than HPLC, an improvement not yet reached but quite possible given recent developments in “instrumentalizing” the technology.

As we’ve seen before, in analytical laboratories speed is king. “That’s why I believe SFC will be revolutionary, but to get there will require further instrumentation development,” Rosse says.

Previous articleAxe Falls at Addex, Two Employees Remain
Next articleBioproduction Approach Could Build Better Oligonucleotides for Less