Modern Supercritical Fluid Chromatography

Skip to Navigation

Education Article

  • Published: Jul 29, 2008
  • Author: Larry T. Taylor
  • Channels: HPLC / Gas Chromatography
thumbnail image: Modern Supercritical Fluid Chromatography

Modern Supercritical Fluid Chromatography

Larry T. Taylor
Department of Chemistry
Virginia Tech
Blacksburg, VA 24061-0212, USA

Historical Development of SFC

Supercritical fluids (SF) have densities and dissolving capacities similar to those of certain liquids, but lower viscosities and better diffusion properties. Accordingly, SF used as mobile phases in chromatography should act both as substance carriers like the mobile phases in gas chromatography (GC) and substance solubilizers like the solvents in liquid chromatography (HPLC). This chromatographic variant has been known as Supercritical Fluid Chromatography (SFC) for over 40 years.

The first renaissance of SFC is generally recognized to have come in 1981-82 with Hewlett-Packard's introduction of instrumentation for packed column SFC at the Pittsburgh Conference and with numerous subsequent studies by both Terry Berger [1] and Dennis Gere who were employed at the time by HP [2]. Concurrent with this event was the first report on the use of open tubular wall-coated columns in SFC by Novotny, Springston, Peaden, Fjeldsted, and Lee [3]. Capillary SFC, as popularized in the 1980's to almost the exclusion of packed column SFC, was practiced during this time using (1) open tubular columns (50µm i.d.), (2) a GC-like oven, (3) pure carbon dioxide, (4) a pump used as a pressure source to perform either pressure or density programming, (5) a fixed restrictor to maintain pressure in the column and to serve as an interface between the column outlet and the laboratory atmosphere, and (6) a flame ionization detector [4]. Historically, capillaries tended to be operated at temperatures well above the critical temperature of the fluid. Thus, this type of SFC was viewed as an extension of GC (but with a greater sample base) where some of the thermal energy required for mobilizing solutes was replaced with solvation energy.

The conviction that open tubular columns were preferred over packed columns was popularly held during the 1980's even though linear velocities 10-20 times the optimum were required to achieve reasonable analysis times. Column efficiency was noted to markedly decrease with carbon dioxide density programming because flow increased with pressure and temperature using fixed restrictors which continues to be the norm even today. Nevertheless, many fantastic separations of low molecular weight polymers and surfactants were reported employing open tubular columns. The initial publicity talked of SFC having all the advantages of GC and HPLC but none of the disadvantages [5]. Unfortunately, a disregard of the physical properties of the fluids and the resulting problems associated with them were rapidly discovered when attempts were made to apply the method. Furthermore, it was observed that the polarity and the solvating power of carbon dioxide are low and many analytes of interest were simply not soluble although earlier reports had postulated that dense carbon dioxide should exhibit polarity similar to isopropyl alcohol [6]! Most pioneers from the pharmaceutical industry who tested the available instrumentation in the 1980's found the technology was very limited, if not almost useless because of its poor reproducibility and limited application range. SFC thus became known as a separations technique that was considered revolutionary when first introduced but whose reputation had slowly ebbed over the years.

The other form of SFC uses packed columns, usually binary or ternary fluids, composition programming, and a UV detector. Stationary phases have much higher surface area to void volume ratios than capillaries and are thus much more retentive. Polar modifiers (which are usually incompatible with flame ionization detection) mixed with the main fluid (CO2) increase the solvating tendency and decrease the retention time of solutes. Once modifiers are added, mobile phase composition becomes much more important than carbon dioxide pressure or density in determining retention unlike capillary column SFC. Polar modifiers may also alter the surface chemistry of the stationary phase such as forming hydrogen-bonds with uncapped silanol sites on the solid support. Packed columns are usually operated near the critical temperature of the fluid with flow control pumps and electronically controlled back pressure regulators mounted downstream of the column to obtain both accurate flow rates and mobile phase composition. In other words, the combination of upstream flow control and downstream pressure control allowed volumetric mixing of the main fluid and modifier and also gradient elution. Most of the components of packed column SFC including columns are similar to HPLC. Thus, packed column SFC can be viewed as an extension or subset of HPLC.

SFC Unifies Chromatography

Berger has noted that the solvent characteristics of the fluids used in SFC are not unique to supercritical fluid conditions [1]. The characteristics are present irrespective of whether the fluid is defined as a liquid, a dense gas, or supercritical fluid. In some instances, the initial pressure used in SFC is actually below the critical pressure. Under such conditions, the technique could be defined as GC. Therefore, the operator cannot tell when the definition (or name) of the mobile phase changes from gas to supercritical fluid to sub-critical fluid to an enhanced fluidity liquid to a normal liquid during a typical separation. As a result, the nomenclature has often been loosely applied in this area, but what was generally called SFC then (and even today) employs carbon dioxide above or near its critical temperature of 31 oC and critical pressure of 73 bars, combined with an organic modifier such as methanol or ethanol. The differences in SFC, subcritical fluid chromatography, enhanced fluidity chromatography and high performance liquid chromatography (HPLC) have been overstated in the past. Diffusion coefficients, viscosity, density, and mobile phase solvent strength show minimal change when the fluid changes from super- to sub-critical. Chester has noted that it is no wonder that many people seeing the apparent complexity of these chromatographies are slow to embrace change [7]. When outlet pressure is elevated and pressure and temperature are controlled, the resulting techniques are similar and the behaviors of conventional GC and HPLC are completely and seamlessly bridged. Berger has stated that fluids slightly above the critical temperature are almost identical to the same fluids slightly below their critical temperature [8]. Each chromatography represents a part of a continuum of increasing mobile phase solvating power coupled with increasing mobile phase viscosity and decreasing mobile phase diffusivity as depicted in Figure 1.

Figure 1
Figure 1: An ordering of variously named chromatographies
in terms of both variable mobile phase diffusivity and solvating power

SFC Re-Focuses on Packed Columns

Sophisticated commercial instrumentation allowing independent flow control under both pressure and composition gradient conditions came out in 1992 boosting the development of applications in every major field of the industry. By 1997 it was clear that the future of SFC would focus more on separation of moderately polar analytes with a combination of polar bonded silica-based packed columns, modified carbon dioxide, and spectroscopic detectors. Many of these early developments were inspired by the work of Dr. Terry Berger who spent more than a decade systematically undoing many of the misconceptions about packed column SFC. In 2004 Berger was awarded the Martin Gold Medal by The Chromatographic Society for work in SFC [9]. Some of the Berger activity included: (1) showed that very long columns with large pressure drops were feasible, (2) de-convoluted density and solvent strengths, (3) introduced the use of mobile phase additives and studied their effects on peak shape and retention, (4) demonstrated that packed column SFC was broadly applicable to small drug-like molecules, and (5) led a team in the development of separator technology which allowed quantitative recovery of solutes without cyclone separators and aerosol generation.

Nowadays, packed column SFC is widely accepted. It uses the same injector and packed column configurations as in HPLC. It is more robust and more adaptable to a broader spectrum of compound classes than just low molecular weight polymeric compounds and nonionic surfactants. It thus is more useful for routine separation of pharmaceuticals for example than open tubular column SFC. Difficulties with back-pressure regulation, consistent flow rates, modifier addition, sample injection, automation, inadequate stationary phases, etc. have been resolved. Pure fluids like CO2 are limited to use with modestly polar solutes, but binary fluids will elute a wider range of polar molecules. Strong or multifunctional acids and bases often need tertiary mobile phases. Many drugs that are salts can be eluted. SFC is, however, not a good candidate for biological molecules such as high molecular weight hydrophilic polypeptides and proteins. Molecules containing both strongly acidic and strongly basic functional groups also tend to be difficult to separate. When working with water-soluble compounds reversed phase HPLC remains the first choice unless water can be removed via a pre-or guard column.

SFC Equals Normal Phase HPLC

Chromatographers readily agree that SFC is normal phase chromatography without most of the problems usually associated with normal phase HPLC. As such, its orthogonality to reversed phase liquid chromatography is an attractive asset for purity assessment. Equilibration is very fast, usually requiring flow turnover of little more than three to five column volumes. Composition, pressure, and temperature can all be programmed with rapid recovery to initial conditions between analyses. Traces of water do not cause the variations in retention often encountered in normal phase HPLC.

The advantages of packed column SFC relative to more seemingly mature HPLC methodologies are clear: (a) lower viscosity and higher diffusivity of supercritical mobile phases relative to liquids which lead to both faster, more efficient separations per unit time and shorter turn-around time between injections, (b) an inert, environmentally "green", more volatile carbon dioxide-based mobile phase for large scale separations and energy efficient isolation of the desired product, (c) longer, stacked columns with the same or multiple phases with total theoretical plates in excess of 100,000 (d) selectivity that matches reversed phase HPLC, but it is more easily adjustable [10], and (e) HPLC applications can be run on SFC instrumentation. These advantages are especially significant in the development of methods for the separation of enantiomers since the stationary phase controls selectivity; therefore numerous columns should be readily screened as shown in Figure 2.

Figure 2
Figure 2: A typical experimental setup for screening various stationary phases
and polar modifiers in the development of a chiral separation.
Figure courtesy of Rodger W. Stringham, Chiral Technologies, Inc.

Furthermore, selectivity is governed also by CO2 polar modifier and predictions of the most effective modifier must be determined by mostly trial and error as illustrated in Figure 3 where one of three low molecular weight alcohols has a unique dramatic effect on resolution of an enantiomeric pair.

Figure 3
Figure 3: Separation of racemic epinephrine on a Chiralcel AD-H column
with three modifiers and 0.1% ethane sulfonic acid (ESA) as the additive.
Figure courtesy of Rodger W. Stringham, Chiral Technologies, Inc.

SFC Equals Preparative Chromatography

The current instrumentation of preparative SFC is in general similar to that of HPLC. Most hardware is directly adaptable from HPLC instrumentation with minimum modifications. The main problem encountered in the development of a preparative SFC prototype was the technology available for fraction collection at the time. Dave Berger (not related to Terry) has stated [11] that the existing cyclone separation technology did not allow for the building of a self-cleaning system which would be able to process up to 12 samples per hour. Phase separation technology was thus developed by Berger Instruments Inc. (then Mettler Toledo, Inc. and now Thar Instruments Inc.) which controlled the gentle decompression of CO2 to the gas phase just beyond the electronic backpressure regulator without aerosol formation. A film of modifier is pushed along the capillary walls to the fraction selection valve. The eluted compounds remain soluble in the modifier liquid phase, and peak collection is driven by an electronic peak detector (i.e. UV/vis). The introduction of this feature has subsequently led to the production of larger amounts (i.e. from milligrams to kilograms) of pure product in a very reproducible manner. The faster SFC process makes the separation cycle time significantly shorter such that it is practically viable to make purification runs by "stacking" small injections onto smaller analytical scale columns in short time windows without compromising the throughput. In this way the utilization rate of expensive column material is much higher.

Alternatively, a preparative mode injection system onto a larger column can be utilized. This operation is accomplished by installation of the sample loop within the modifier stream before the modifier/CO2 mixing point which enables large injection volumes to be made without impact of sample solvent onto column equilibrium [11]. Nowadays, preparative instruments for industrial-scale runs can be subdivided. Semi-preparative instruments have flow rates from 20 to 200mL/min; while larger systems handle flow rates at liters per minute with columns half as big as an automobile. For comparison purposes analytical-scale instruments have liquid flow rates less than 20mL/min.

The use of SFC for preparative separations has received considerable attention during the last two years as a tool for supporting preclinical development in the pharmaceutical industry. The resulting decrease in solvent use and waste generation offers a green advantage with an economic bonus that makes preparative SFC especially attractive for providing purified materials on a kilogram scale [12]. The SFC product is recovered in a more concentrated form relative to HPLC, thereby greatly reducing the amount of solvent that must be evaporated and gives rise to considerable savings in labor, time, and energy costs [13]. The higher SFC flow rates also contribute to higher productivity relative to HPLC methods. Furthermore, a full scale up study from analytical to preparative is required for HPLC processes since it is not a straightforward exercise. This is part of the reason that the preparative chiral HPLC process is still laborious because the amount of time used for extra development counts as a significant portion of the overall process. SFC is in a perfect position to overcome this disadvantage since the lower viscosity of supercritical fluids makes it practical to run the same separation at 5 times higher flow rate without the pressure buildup issue. In addition, this means that the same type of high efficiency stationary phase particles can be used in both analytical and preparative processes; whereas, in preparative HPLC, larger, less efficient particles are often used to reduce the pressure effect [14].

Current Applications of SFC

Major emphasis in SFC today concerns packed columns although several studies using open tubular columns continue to appear each year. Packed column SFC applications include both analytical scale chiral and achiral pharmaceutical separations. For example, enantiomeric separation of several chiral sulfoxides belonging to the family of substituted benzimidazoles has been reported with a Chiralpak AD stationary phase and methanol modified CO2 mobile phase [15]. Resolutions higher than 2 and separation times shorter than 10min were observed. Another application concerns the coupling of SFC with mass spectrometric detection for support of drug discovery. Packed column SFC-MS/MS has been used for the separation of fifteen estrogen metabolites [16]. A gradient of methanol in CO2 with a cyanopropyl silica column connected in series with a diol column packed with 5µm spherical silica-based particles resulted in separation and quantification in less than 10 minutes. The limit of detection and limit of quantification was determined to be 0.5pg (S/N = 3) and 5pg respectively. As a third application, SFC via simulated moving beds continues to attract much interest. Relatively large peptides (at lest 40 mers) containing a variety of acidic and basic residues have been eluted via SFC as a fourth application to natural products [17]. Trifluoroacetic acid was used as an additive in a CO2/ethanol mobile phase to suppress deprotonation of peptide carboxylic acid groups and to protonate peptide amino groups. A fifth area of great interest concerns preparative SFC. The first direct multi-gram purification of all four isomers of the unnatural amino acid (beta-methylphenylalanine) using SFC with stacked injection was reported [18]. A Daicel Chiralpak AD-H column (20mm x 250 mm) using 50:50 methanol/ethanol as the organic modifier resulted in purification of over 3.4 g of material in 6.25 hour with >90% recovery.

In summary, SFC affords a continuum of chromatographic conditions that span GC and HPLC. Instrument makers have addressed many of the past experimental problems to the point that SFC is truly user friendly. Reported applications are diverse and significant separation and purification problems are being addressed.

[Back to Top]


  1. Berger, T. A., "Packed Column SFC", Royal Society of Chemistry, Letchworth, UK, 1995
  2. D. R. Gere, D. R., R. K. Houck, R. K., F. Pacholec, F., Rosselli, A. C. P. "Supercritical Fluid Chromatography: A Technology Update", Fresenius J. Anal. Chem., 1988, 330, 222-224.
  3. Novotny, M., Springston, S. R., Peaden, P. A., Fjeldsted, J. C., Lee, M. L., "Capillary Supercritical Fluid Chromatography", Anal. Chem., 1981, 53, 407A.
  4. Griffiths, P. R., "Contemporary SFC: Accomplishments and Limitations", Anal. Chem., 1988, 60, 593A-597A.
  5. Smith,R. M., "Supercritical Fluids in Separation Science - the Dreams, the Reality, and the Future", J. Chromatogr., 1999, 856, 83-115.
  6. Giddings, J. C., Meyers, M. N., McLaren, L. M., Keller, R. A., "High Pressure Gas Chromatography of Nonvolatile Species", Science, 1968, 162, 67.
  7. Chester, T. L., "Chromatography from the Mobile-Phase Perspective", Anal.Chem., 1997, 69, 156A-169A.
  8. Berger, T. A., "Practical Advantages of Packed Column Supercritical Fluid Chromatography in Supporting Combinatorial Chemistry", in "Unified Chromatography", ed by J. F. Parcher and T. L. Chester, ACS Sym. Series, 2000, 748, 203.
  10. Sandra, P., Medvedovici, A., Kot, A., David, F., "Selectivity Tuning in Packed Column Supercritical Fluid Chromatography", in "Supercritical Fluid Chromatography with Packed Columns", ed. by K. Anton and C. Berger, Marcel Dekker, Inc., New York, 1998, pp. 161-194.
  11. Berger, D., "Supercritical Fluid Chromatography: A Review of Technical Developments", LCGC Europe, March 1, 2007.
  12. Wu, D-R.; Leith, L.; Balasubramanian, B.; Palcic, T.; Wang-Iverson, D., "The Impact of Chiral Supercritical Fluid Chromatography in Drug Discovery: From Analytical to Multi-gram Scale", Amer. Lab., 2006, June/July, 24-26.
  13. Yan, T. Q.; Orihuela, C.,"Rapid and High Throughput Separation Technologies - Steady State Recycling and Supercritical Fluid Chromatography for Chiral Resolution of Pharmaceutical Intermediates" J. Chromatogr. A, 2007, 1156, 220-227.
  14. Wang, Z., "Development of Supercritical Fluid Chromatography for Chiral Separation in the Pharmaceutical Industry", Amer. Pharm. Rev., 2007, 10, 96-100.
  15. Toribio, L.; Alonso, C.; del Nozal, M. J.; Bernal, J. L.; Jimenez, J. J., "Enantiomeric Separation of Chiral Sulfoxides by Supercritical Fluid Chromatography", J. Sep. Sci., 2006, 29, 1363-1372.
  16. Xu, X.; Roman, J. M.; Veenstra, T. D.; Anda, J. V.; Ziegler, R. G.; Issaq, H. J., "Analysis of Fifteen Estrogen Metabolites Using Packed Column Supercritical Fluid Chromatography-Mass Spectrometry", Anal. Chem., 2006, 78, 1553-1558.
  17. Zheng, J.; Pinkston, J. D.; Zoutendam, P. H.; Taylor, L. T., "Feasibility of Supercritical Fluid Chromatography/Mass Spectrometry of Polypeptides with up to 40-Mers", Anal. Chem., 2006, 78, 1535-1545.
  18. Nogle, L. M.; Mann, C. W.; Watts, W. L.; Zhang, Y., "Preparative Separation and Identification of Derivatized ß-Methylphenylalanine Enantiomers by Chiral SFC, HPLC, and NMR for Development of New Peptide Ligand Mimetics in Drug Discovery", J. Pharm. Biomed. Anal., 2006, 40, 901-909.

[Back to Top]

Social Links

Share This Links

Bookmark and Share


Suppliers Selection
Societies Selection

Banner Ad

Click here to see
all job opportunities

Copyright Information

Interested in spectroscopy? Visit our sister site

Copyright © 2019 John Wiley & Sons, Inc. All Rights Reserved