Supercritical Fluid Chromatography

Volume 1

Gérard Rossé

• Publisher: De Gruyter
• Year: 2018
• eBook ISBN: 9783110498127
• Edition: 1
• Topic: Naturwissenschaften
• Subtopic: Umweltwissenschaft

Supercritical Fluid Chromatography (SFC) provides a timely overview of SFC application areas which were unimaginable just a decade ago. This two-volume series opens with an overview of the history and expectant future of SFC and continues with recent applications in the pharmaceutical industry and other fascinating areas of science. SFC has found its place in the pharmaceutical industry with an increasing body of applications for chiral and achiral molecules in both the research and development phases of the drug discovery process. As illustrated in this two-volume series, the current interest in SFC extends well beyond the pharmaceutical industry. Chapters encompassing applications for polar and non-polar mixtures of importance are covering widely disparate areas in substance abuse, natural products including cannabinoids, bioactive lipids, flavor and fragrance. With its broad balance and coverage, this two-volume book constitutes a unique educational platform to students and scientists for many years to come. The major objective of this book editions is to inspire and stimulate readers to continue exploring the possibilities of exploiting supercritical fluids as a particular media for analysis, purifications and synthesis

Terry A. Berger

1A history of supercritical fluid chromatography (SFC)

Terry A. Berger, SFC Solutions, Inc.,9435 Downing St., Englewood, FL 34224

Abstract: Supercritical fluid chromatography (SFC) has slowly developed over more than 60 years, and has only fairly recently become mainstream. Some of the reasons for such slow development were a few shaky concepts from the nineteenth century and a lack of adequate technology. When SFC was first proposed in 1957, high-performance liquid chromatography (HPLC) did not exist. Gas chromatography (GC) was in its infancy but growing rapidly. SFC was seen as a possible means to extend GC to heavier, low-volatility compounds. The details regarding fluids were not completely understood. Pumping highly compressible fluids at high pressures was fraught with difficulties. All equipment used up until the early 1980s were homemade and surprisingly crude. Most of the work of the first 20 years pointed in a direction different from where we wound up. Many of the scientists involved were some of the biggest names in the field of chromatography. The stature of these scientists made it difficult to change directions to pursue a different path later on. Consequently, the 1980s were filled with controversies. The viewpoint expressed in this chapter is somewhat biased since the author was directly involved during the years of the greatest controversies.

Nowadays, instrumentation for SFC is as robust as HPLC. Most applications use small particle columns and binary mobile phases, often with gradient elution. There is accurate independent dynamic control of flow, composition, temperature, and outlet pressure. The physical and chemical nature of the mobile phases is reasonably well understood. None of this was true for the first 25+ years of SFC.

Keywords: pre-1980 confusion, 1980’s controversies, Modern SFC, mainstream chiral analysis, 2nd and 3rd generations

1.1The beginnings up to 1980

The existence of a critical point was first discovered by Charles Cagniard de la Tour in 1822 [1]. He showed that CO2 could be liquefied at temperatures up to 31 °C by increasing pressure up to 73 bar, but above 31 °C, it could not be liquefied even at 3,000 bar. Dmitri Mendeleev may have named it the “critical point” in 1860. In 1869, Thomas Andrews [2] also described the end of the liquid/ vapor equilibrium line as the “critical point” and coined the terms “critical temperature” and “critical pressure” using CO2.

Ten years later, in a short communication, Hannay and Hogarth [3] described the solubility of a number of inorganic salts and chlorophyll in several fluids above their critical point. Most of the work was performed with supercritical methanol or ethanol. According to Krukonis [4], this report was greeted with disbelief, which was only overcome with substantially more work. The criticisms were quite severe, and the response remains the same. In a following paper, Hannay stoutly defended his previous work and provided substantial further evidence. He claimed that “The gas must have a certain density before it will act as a solvent.” This was later called “threshold density” by Giddings. Hannay [5] has been widely quoted as having said: “The liquid condition of fluids has very little to do with their solvent power, but only indicates molecular closeness. Should this closeness be attained by external pressure instead of internal attraction, the result is the same or even greater solvent power....” He ignored or was unaware of the differences in intermolecular interactions by different types of solvents and suggesting it is only density that is important. Such a claim is, of course, unrealistic. Today, beside dispersive interactions, we also look at proton donor, proton acceptor, and dipole and electrostatic interactions.

There is no indication that anyone up to this point knew what the density of the various fluids was above their critical points. Perhaps more importantly, there also appears to have been no attempt to try similar high-pressure experiments, but below the critical temperature, using fluids such as CO2, which are gasses under atmospheric conditions. Typical of developments with supercritical fluids, it was only in 1906 (27 years later) that Buchner [6] studied the solubility of naphthalene in supercritical CO2. Naphthalene became a model compound for various studies.

After another 50+ years, Jim Lovelock, the designer of the gas chromatography (GC) that analyzed the Marian atmosphere, inventor of the helium and argon ionization detectors, and the electron capture detector, suggested, at a 1956 international meeting, that some gases compressed above their critical point might be usable as a mobile phase in GC, to elute larger compounds with little or no volatility. It is unclear if he knew of Hannay’s work. He typed a one-page letter he had notarized in 1957, copies of which still exist. He suggested the name “critical state chromatography.” Lovelock never did any lab work to test his supposition.

Lovelock suggested that some inorganic gases such as SO2 under high pressure were likely to solvate more polar molecules that were not volatile in GC. He also suggested that columns “like Desty used” would be useful. Golay [7] had presented his theory for open tubular columns for GC, at the spring meeting of the American Chemical Society in 1956, and subsequently published in 1957. Dennis Desty [8] was an early leader in their application using very small bore glass and metal columns. Years later this letter resulted in an argument about the patentability of capillary SFC, as proposed by Novotny et al. [9], but we are getting ahead of the story.

1.1.1Klesper and Corwin

It took another 5 years after Lovelock’s suggestion, before Klesper et al. [10] published the first report of what we now call supercritical fluid chromatography in 1962. Of all the places, they chose to report the results as a brief (approximately two pages) preliminary “Communication to the Editor” in the Journal of Organic Chemistry. They used dichlorodifluoromethane and chlorodifluoromethane (refrigerants) as the mobile phases to separate metal porphyrins of the types found in crude oil. These porphyrins were thought to offer a means of understanding the origins of petroleum but were too thermally unstable for regular GC.

Details were sketchy. There was no pump. In fact there were no pumps that could accurately control the flow rate of a compressible fluid at high pressures at that time. The fluid was placed in a closed reservoir, which was then heated in an oven to generate higher pressure, and then passed through a packed GC column containing large particles. A downstream restriction throttled flow. The solutes were colored and were observed visually as they migrated down the column. After the separation, the fluid was bubbled through liquid solvent to collect fractions that were analyzed off-line. There was no in-line detector. This extremely crude approach is indicative of the state of the art. The results were preliminary, but were never completed by Klesper. Klesper was at the time a research assistant under Prof. Corwin, at Johns-Hopkins, in Maryland. Nevertheless, Klesper is generally considered the “father” of SFC. Klesper did not publish further on SFC until the mid-1970s.

The details on the separation of porphyrins were clarified somewhat in several later reports by Karayannis and Corwin [11], which included a number of chromatograms. Perkin Elmer had produced a “hyperpressure gas chromatograph” which replaced the previous apparatus. It was apparently extensively modified by Corwin and included a spectrophotometer as detector. The instrument does not appear to have been developed commercially as there are only a few references to its use.

Columns were typically 43 in. long (≈110 cm) with one-eighth inch ID (3.2 mm), packed with supports of Chromosorb 60–80 mesh, or 80–100 mesh. These are equivalent to 177–250 μm and 149–177 μm, respectively. These supports were coated with various stationary phases commonly used in GC. These columns probably had very low pressure drops.

The same chlorofluorocarbons were used as the mobile phases. Pressures were ≈75–200 bar. Temperatures were 140–170 °C, but at unknown densities. There appears to have been no independent means to control pressure and flow. Pressures were reported with an associated flow rate. Flow rates were between 324 mL/ min at 75 bar and 756 mL/ min at 180 bar. These appear to be the expanded flows measured at atmospheric pressure. The range suggests the use of a fixed restrictor to provide a pressure drop. Peaks were minutes wide.

1.1.2Sie and Rijnders

Sie and Rijnders [12] at Shell, in Amsterdam, published a number of reports between 1966 and 1969, and they were the first to call the technique “supercritical fluid chromatography,” although almost all their CO2 data were collected under subcritical conditions. They do appear to have been the first to use a flame ionization detector (FID) in SFC which was invented the same year Golay produced his theory on capillary columns (1956). Initially, they used mostly CO2 as mobile phase, and typically 50/70 mesh (210–297 μm) Sil-O-Cel C22 firebrick as stationary phase in a 1 m × 6 mm ID column. At that time, broken-up firebrick was a common support in packed column GC. In one report they used uncoated 120–140 mesh (105–125 μm) alumina. All these columns would have negligible pressure drops, except at high flow rates.

They immersed a cylinder of CO2 in a large water bath and heated the water to generate a higher pressure. Today, this would be illegal in many jurisdictions for safety reasons, although some still use heating blankets, where legal. The heated reservoir acted as a pressure source. Flow was “controlled” by a manual metering valve mounted downstream of the column. Most of the pressure drop in the system probably occurred across this metering valve. All the early papers can be characterized as providing an amazing lack of detail.

Maximum pressures were ≈80 bar. Temperatures range between 30 and 40 °C. Plots of log k versus pressure, between 10 and 80 bar (40 °C), produced continuously decreasing retention that were nearly linear until they approached Pc, where the decrease accelerated. The only data point on these plots that was supercritical was at 80 bar. They generated the data as a bridge between GC and supercritical fluids. This work was really high-pressure GC, showing that CO2 acts as a solvent (at least for low polarity solutes) even at quite low pressures (like 10 bar) and low temperatures.

At 40 °C, the density at 10 bar is 0.0177 g/cm3. At 70 bar, still below the critical pressure, the density increases more than 10-fold to 0.198 g/cm3. At 80 bar (supercritical), the density only reaches 0.281 g/cm3. These have very low densities and explain the use of light hydrocarbons as solutes. Solutes were paraffins and light hydrocarbons, which are quite volatile and easily separated by GC.

Later they switched to pentane and isopropanol (IPA) [13, 14] since these fluids were much easier to work with, being liquids at room temperature. They used the same columns in an air bath oven, but at much higher temperatures (200–250 °C). The liquid mobile phase was placed in a reservoir and pressurized with nitrogen gas at up to 50 bar. This led to some nitrogen being dissolved into the mobile phase. The densities used were...

Table of contents

• Cover
• Title Page
• Copyright
• Preface
• Contents
• List of Contributors
• 1 A history of supercritical fluid chromatography (SFC)
• 2 Achiral preparative supercritical fluid chromatography
• 3 Addressing robust operation and reproducibility in SFC
• 4 SFC within early drug discovery research – a user’s perspective
• 5 What is the potential of SFC-MS for doping control analysis?
• 6 SFC as a novel approach to assess polarity and identify intramolecular hydrogen bonding
• 7 Applications of supercritical fluid chromatography: Natural products in pharmaceutical, cosmetic, and food applications
• 8 Analytical chiral supercritical fluid chromatography
• Index