History of supercritical fluid chromatog

Journal of Bioscience and Bioengineering
VOL. 115 No. 6, 590e599, 2013
www.elsevier.com/locate/jbiosc

REVIEW

History of supercritical fluid chromatography: Instrumental development
Muneo Saito*
JASCO Corporation, 2967-5 Ishikawa-cho, Hachioji, Tokyo 192-8537, Japan
Received 12 September 2012; accepted 5 December 2012
Available online 11 January 2013

In the early days of supercritical fluid chromatography (SFC), it was categorized as high-pressure or dense gas chromatography (HPGC or DGC) and low boiling point hydrocarbons were used as supercritical mobile phase. Various liquids
and gases were examined, however, by the late 1970s, carbon dioxide (CO2) became the most preferred fluid because it
has low critical temperature (31.1 C) and relatively low critical pressure (7.38 MPa); in addition, it is non-toxic, nonflammable and inexpensive. A prototype of a modern packed-column SFC instrument appeared in the late 1970s.
However, in the 1980s, as open tubular capillary columns appeared and there was keen competition with packed
columns. And packed-column SFC at once became less popular, but it regained popularity in the early 1990s. The history
of SFC was of “the rise and fall.” Advances in chiral stationary phase took place in the early 1990s made packed-column
SFC truly useful chiral separation method and SFC is now regarded as an inevitable separation tool both in analytical and
preparative separation.
Ó 2012, The Society for Biotechnology, Japan. All rights reserved.

[Key words: Supercritical fluid; Chromatography; Open tubular capillary column; Packed column; Preparative supercritical fluid chromatography;
Chiral separation]

What is a supercritical fluid? It is a highly compressed gas
that has a density similar to that of a liquid. Fig. 1 shows the phase
diagram of a pure substance.
At the triple point (TP) the three phases (gas, liquid, and solid) of
the substance coexist in thermodynamic equilibrium. As the
temperature goes high, the substance coexists in two phases, i.e.,
gas and liquid. At the critical point (CP) and in the region the
temperature and the pressure are above the critical temperature
and pressure, the substance exists in a single gaseous phase. The
substance in this region is defined as in the supercritical state
where the density is liquid-like while the viscosity is gas-like, and
the diffusivity is in between those of a liquid and a gas as shown in
Table 1 (1). It is expected that a supercritical fluid, which has the
higher diffusivity and the lower viscosity than a liquid solvent, will
function much better than a liquid solvent as an extractant and
a mobile phase in extraction and chromatography.
Chromatography that uses a supercritical fluid as the mobile

phase, i.e., supercritical fluid chromatography (SFC), was first reported by Klesper et al. (2) as high-pressure gas chromatography
(HPGC) in 1962, a little before the advent of high-performance
liquid chromatography (HPLC). Although SFC has a history as long
as or even a little longer than that of HPLC, it was not so long ago,
probably in the latter half of the 1990s, when SFC was recognized as
a truly useful separation method. There are a few reasons why it
took such a long time. The greatest reason is the advent of HPLC. At
the time of the first report on SFC, gas chromatography (GC) was
* Tel.: þ81 42 646 4111x204; fax: þ81 42 643 0053.
E-mail address: muneo.saito@jasco.co.jp.

already a well established method and its instrumentation was
readily available from several commercial sources and researchers’
interest was shifted to an analytical method that could analyze
thermally labile, non-volatile or polar compounds that could not be
separated by GC. Liquid chromatography (LC) has the potential to
realize these requirements, however, stationary phases and
instrumentation available at the time did not allow high-speed and
high-efficiency analysis comparable to that of GC. And tremendous
efforts were paid to the development of HPLC. Although SFC is not

as versatile as LC, it too has good potential. The development of SFC
was unfortunately shaded by the rapid development of HPLC that
took place in the latter half of the 1960s and the 1970s.
There are two important review articles on SFC, one published in
2009 by Taylor (3) and the other in 2011 by Guiochon and Tarafder
(4). Taylor (3) overviewed various techniques in SFC and put them
into a compact and comprehensive article. Guiochon and Tarafder
(4) covered every aspect of SFC including theoretical and empirical
treatments of physicochemical properties of high temperature and
high-pressure fluids, even including thermodynamics and the
equations of state. Their article is roughly 77 pages and could be
a small monograph. The author recommends the readers to read
these review articles for an in-depth understanding of SFC. In this
article, the author covers mainly a history of instrumental development that was partially discussed in the above reviews.
Fig. 2 shows the number of articles on SFC published each year
from 1962 to 2012. In the 1960s, the number is very small and does
not chart as well relative to the later years. In the 1970s, the number
varies but not more than 30 per year. However, in the 1980s, the
number exponentially increases from 15 to over 400 in a decade. In

1389-1723/$ e see front matter Ó 2012, The Society for Biotechnology, Japan. All rights reserved.
http://dx.doi.org/10.1016/j.jbiosc.2012.12.008

VOL. 115, 2013

HISTORY OF SUPERCRITICAL FLUID CHROMATOGRAPHY

591

900
800
700

Supercritical
Fluid

Melting curve

600

Pressure

Pc

500
400

CP

Liquid

300
200
100

rv
e

n
io

at
or
p
a
Ev

cu

0
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982

1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012

Solid

Gas

TP
FIG. 2. Numbers of publications on supercritical fluid chromatography. Based on
a Google Scholar search performed on July 2, 2012.

Sublimation
curve

Tc
Temperature
FIG. 1. Phase diagram of pure substance. TP: triple point; CP: critical point; Pc: critical
pressure; and Tc: critical temperature. These parameters are specific to each substance.
Reproduced from Saito et al. (47) with permission of John Wiley & Sons, Inc.

the 1990s, the number is somehow saturated at around 500e600.
Then, it rapidly increased again in the first decade of the 21st
century.
EARLY WORKS IN HIGH-PRESSURE GAS CHROMATOGRAPHY
(HPGC) OR DENSE GAS CHROMATOGRAPHY (DGC)
Works of early pioneers In the early days, presently
accepted terminology “supercritical fluid chromatography (SFC)”

for the method was not established, and various terms were used
such as (ultra) HPGC and dense gas chromatography (DGC).
The first report on SFC by Klesper et al. (2) appeared in the
Communications to the editor section of Journal of Organic
Chemistry in 1962. It is a brief report, less than 2 pages. They
indicated thermally labile porphyrin mixtures were separated on
polyethylene glycol stationary phase with two mobile phase gases
such as dichlorodifluoromethane (Tc ¼ 112 C) and monochlorodifluoromethane (Tc ¼ 96 C) at temperatures of 150e170 C. They also
stated “above 1000 psi (7 MPa) with the first and 1400 psi (9.8 MPa)
with the second, vaporization increased with increasing gas pressure”, indicating that porphyrins were dissolved in the supercritical
mobile phase. In addition, they foresaw the possibility of preparative SFC by stating “the porphyrins could be recovered at the outlet
valve.”
Sie et al. (5e8) published a series of articles on HPGC in 1966 and
1967. They used supercritical carbon dioxide as the mobile phase
and discussed fluidesolid and fluideliquid separation modes. It
should be noted that they developed a sophisticated pneumatically

operated injector in order to inject a sample under high-pressure
and high-temperature conditions. It is also unique that they used
a UV absorption detector with a quartz cell that was equipped with

a gaseliquid separator and detection was carried out under atmospheric pressure. Both devices did not survive, however, the author
thinks it is worth mentioning these pioneering works.
Sophisticated instrumentation developed in the late
1960s In 1968, Klesper’s group reported a new SFC system (9).
The system was equipped with a mechanical backpressure
regulator that could control the pressure independent of the flow
rate. The detector was a filter photometer with a high-pressure
flow cell. It should be noted that the detector used in HPLC at
that time was a simple single-wavelength photometer with
a low-pressure Hg discharge lamp as the light source which emits
UV light at 254 nm (10). In 1968, vacuum tubes were still in use
and the transition to transistors had just begun. Therefore, the
author believes that the system consisted of a sophisticated
analog servo system with vacuum tubes. It is remarkable that
a prototype of a modern packed-column SFC appeared more than
40 years ago.
In 1969, Giddings et al. (11) reported DGC. In the article they
stated “One of the most interesting features of ultra-high-pressure
gas chromatography would be its convergence with classical liquid
chromatography. A liquid is ordinarily about 1000 times denser

than a gas; at 1000 atm, however, gas molecules crowd together
with a liquid-like density. At such densities intermolecular forces
become very large, and are undoubtedly capable of extracting big
molecules from the stationary phase. Thus in effect, non-volatile
components are made volatile.”
The first sentence seems to imply the unified chromatography
that was later realized and reported by Ishii et al. (12) in 1988 and
more recently by Chester and Pinkston (13). The latter part of the
paragraph describes solvation of a solute in a supercritical fluid.
This phenomenon was investigated and well elucidated by Kim and
Johnston (14) in 1987 and Kajimoto et al. (15) in 1988. Their works
will be explained later.

TABLE 1. Properties of gas, liquid and supercritical fluid.a
Property

Units

Gas 1 atm, 25 C

Liquid 1 atm, 25 C

Supercritical fluid
Tc, Pc

Density
Diffusivity
Viscosity
a

r (g cm 3)
Dm (cm2 s

h (g cm

1

After Takishima and Masuoka (1).

1

)

s

1

)

0.6e2  10
1e4  10 1
1e3  10 4

3

0.6e1.6
0.2e2  10
0.2e3  10

5
2

0.2e0.5
0.5e4  10
1e3  10 4

Tc, 4Pc
3

0.4e0.9
0.1e1  10
3e9  10 4

3

SAITO

Giddings et al. proposed an extension of the Hildebrand solubility parameter to a supercritical fluid (11). They used various
gases including He, N2, CO2 and NH3, and examined retention
behavior of various substances such as purines, nucleosides and
nucleotides, steroids, sugars, terpenes, amino acids, proteins, carbowaxes, etc. However, CO2 later became the most preferred fluid
because it has low critical temperature (31.1 C) and relatively low
critical pressure (7.38 MPa), in addition, it is non-toxic, non-flammable and inexpensive. Today, an SFC mobile phase automatically
assumes a pure or mixture of CO2 and organic modifiers.
SFC instrumentation with pressure programming and
fractionation capability
In 1970, Jentoft and Gouw (16)
reported a new SFC system that allows pressure programming.
They separated polycyclic aromatic hydrocarbons and styrene
oligomers utilizing pressure programming. In the case of
a supercritical mobile phase, the higher the pressure, the stronger
the solvating power becomes. Thus, it functions similar to
gradient elution in LC. They claimed that their new SFC system
offered comparable performance to high resolution LC (note that
the terminology “high-performance liquid chromatography” was
not yet generally accepted at that time). It is remarkable that they
already developed such a sophisticated SFC system as early as in
1970. At that time, HPLC was still in an early developmental stage
and there was an argument as to which type of the pump (i.e.,
a syringe, a reciprocating or a constant-pressure pump) was best
suited for the HPLC pump. In 1972, they developed a highpressure fraction collector to fractionate components eluted from
an SFC system, proving that SFC can be used for preparative
applications (17).
In 1977, Klesper and Hartman (18,19) reported a sophisticated
preparative SFC (prep-SFC) and fractionation of styrene oligomers.
Fractions were analyzed with mass spectrometry, thus, constituting
offline SFC-mass spectrometry (MS). In 1978, Randall and
Wahrhaftig (20) described a dense gas chromatography (DGC)/
mass spectrometer (MS) Interface.
In 1980, a monograph edited by Schneider et al. (21) was published. This book covers developments that took place mainly in
Europe in the 1960s and 70s such as plant scale supercritical fluid
extraction (SFE) and their applications, physicochemical studies of
supercritical fluid including phase equilibria, analytical scale SFE
combined with thin layer chromatography (TLC), and SFC.
OPEN TUBULAR CAPILLARY COLUMN VERSUS PACKED COLUMN,
AND OTHER DEVELOPMENTS

J. BIOSCI. BIOENG.,
a schematic diagram of a typical GC-like open tubular column SFC
system. The SFC system sold well in the beginning mainly in the US,
however, within a few years, it was revealed that it had intrinsic
technical difficulties.
Technical advantages of SFC over other modes of chromatography such as GC and LC are summarized as in Table 2. In SFC, all
three parameters (i.e., pressure, temperature and modifier content)
can independently or cooperatively control retention, or even
a gradient method can be applied to all the parameters. These
advantages were too heavily emphasized at that time. Therefore, it
often misled chromatographers to think that SFC was a type of
super chromatography. However, in the case of open tubular
capillary SFC, pressure (or density that is inversely proportional to
the pressure under the certain conditions) which is the most
important operating parameter, could only be varied by changing
the flow velocity due to the limitation of the constant restrictor.
This means that the pressure could not be changed independently
of the flow velocity. Therefore, one could never obtain the optimum
flow velocity at the optimum pressure. In addition, the standard FID
detector could not be used with an organic modifier because even
a small amount of organic solvent produced too high a background
on the baseline, limiting application range. There were a few
attempts to use premixed CO2 with an organic solvent in the
cylinder to add some polarity to the CO2-based mobile phase in
order to elute polar compounds using a UV absorption detector
equipped with a micro flow cell. However, this technique did not
attract chromatographers because they could not change the
modifier content as required or run modifier gradient elutions.
Thus, it could not extend the application areas, and open tubular
capillary column SFC rapidly diminished in the early 1990’s. Taylor
intensively described the history of open tubular capillary column
SFC (3).
Packed column
Before the advent of open tubular capillary
column SFC, all SFC research works were performed using packed
columns. While open tubular capillary column SFC was GC-like
instrument, packed-column SFC was more like LC. In 1982, Gere
et al. (24) modified a HewlettePackard (HP) HPLC system to operate
as an SFC system by adding a backpressure regulator and other
devices. They showed that SFC gave higher efficiency with 3, 5
and 10 mm packing materials especially in high flow velocity
region. Packed-column SFC was developed almost independently
of open tubular capillary column SFC. Packed-column SFC at once
became less popular, especially in the US due to the marketing
strategy of open tubular column SFC in the middle of 1980s.

Research activities on column technology and instrumentation
were very active and diverse in the 1980s, and this led to the
commercialization of SFC instruments.
Open tubular capillary column In 1981, Novotny and Lee’s
group introduced open tubular capillary column SFC (22). A typical
open tubular capillary column was a 50 mm inner diameter fused
silica capillary tube and the internal wall was coated with
a polymer such as dimethyl polysiloxane that functioned as the
stationary phase.
Novotny et al. (23) previously studied retention behavior of
packed columns under various conditions, and stated that a packed
column could not give high-efficiency at high-linear velocity
because of the pressure drop along the column that functions as
a negative density gradient. They emphasized that a small pressure
drop across an open tubular capillary column would give higher
efficiency than a packed column. Later open tubular capillary SFC
was patented and exclusively marketed by Lee Scientific. The
system consisted of a syringe pump, an injection valve with a split
mechanism, a GC-like oven, a wall-coated open tubular column,
a fixed restrictor, and a flame ionization detector (FID). Fig. 3 shows

Injection Valve

Preheat Coil

CO2
Cylinder

Drive
Mechanism

Column

FID

Restrictor

592

Column Oven

Syringe pump
FIG. 3. Schematic diagram of typical GC-like open tubular column SFC system. Since the
flow rate is very low, a screw-driven syringe pump is used. Backpressure is applied by
a restrictor that has a certain flow resistance to keep the system pressure above the
critical pressure of the fluid. Pressure was controlled by changing the mobile phase
flow rate. Reproduced from Saito et al. (47) with permission of John Wiley & Sons, Inc.

VOL. 115, 2013

HISTORY OF SUPERCRITICAL FLUID CHROMATOGRAPHY

TABLE 2. Various modes of chromatography and available control parameters.
Parameter/Mode

GC

SFC

LC

Pressure
Temperature
Modifier

No
Yes
No

Yes
Yes
Yes

No
Yes
Yes

However, it regained popularity when packed columns were found
to have a wider application range than open tubular columns
(25,26). Fig. 4 shows a schematic diagram of a typical LC-like
packed-column SFC system with automated backpressure
regulator. It is very similar to an HPLC system. However,
a backpressure regulator that keeps the fluid pressure above the
critical pressure and an oven that keeps the fluid temperature
above the critical temperature are vital devices specific to SFC.
Photodiode array UV detector and electronic backpressure
regulator In 1985, Sugiyama et al. (27) developed a packedcolumn SFEeSFC hyphenated system, and demonstrated the
extraction and chromatography of caffeine from ground coffee
beans. The SFE directly coupled to an SFC system allowed an
online introduction to an SFC column and the signal was

Sampling syringe

RF

Sample
INJ

PU1

Column

Preheat Coil

PU2

CO2 Modifier
OVEN

PDA

PT

BR

Collection tube

FIG. 4. Schematic diagram of typical LC-like packed-column SFC system with automated backpressure regulator. PU1: liquefied CO2 delivery reciprocating pump with
chilled pump heads; PU2: modifier solvent delivery pump; RF: safety relief valve that
prevents over pressure; INJ: injection valve; PDA: photodiode array UV detector; PT:
pressure transducer; and BR: backpressure regulator. The pressure transducer monitors the pressure real time and the backpressure regulator compares the set pressure
and actual pressure and control the flow resistance of the regulator so that the actual
pressure becomes equal to the set pressure.

593

monitored with a photodiode array UV detector (PDA). PDA has
soon become the standard detector in packed-column SFC.
In a modern SFC system, the most important device may be the
backpressure regulator which allows pressure control independent
of mobile phase flow rate. Saito et al. (28) developed an electronically controlled backpressure regulator that had a very small
internal volume and allowed efficient fractionation without cross
contamination between fractions. This type of the backpressure
regulator has become the standard device in packed-column SFC.
Chiral separation
In the 1980s, Okamoto et al. (29,30)
developed highly efficient and versatile chiral stationary phases
(CSP) and published a series of articles. Later, these CSPs were
commercialized by Daicel Corporation, Osaka, Japan, and rapidly
spread throughout the world; first used with LC and then
extended to SFC.
In 1985, Mourier et al. (31) demonstrated a chiral separation of
phosphine oxides with supercritical and subcritical carbon dioxide
mobile phase. In 1986, Hara et al. (32) demonstrated an SFC chiral
separation of dl-amino acid derivatives. In the same year, Perrut
and Jusforgues (33) developed a prep-SFC system with a 60-mm i.d.
column with carbon dioxide recirculation. They mentioned that
a preparative SFC is a sophisticated high-pressure gas equipment
and thus expensive (33). Therefore, preparative SFC is suitable for
fractionation of high-valued compounds such as chiral drugs,
essential oils, etc.
Later in the 1990s and 2000s, advances in chiral stationary
phase and instrumental development made chiral separations one
of the most important and preferred applications in both analytical
and preparative SFC (34,35).
Cluster theory For the utilization of a supercritical fluid as an
extraction solvent or a mobile phase for chromatography, the fluid
must have a solvating power. Kajimoto (36) illustrated the behavior
of molecules in gas, liquid, and supercritical state in view of the
intermolecular potential and the average molecular energy
as shown in Fig. 5. At lower temperatures, an energetically
lower state is strongly favored as is known in statistical
thermodynamics. In the liquid state at low temperatures,
therefore, each molecule feels the attractive intermolecular
potential and most molecules are trapped in the potential well,
the depth of which is usually larger than the average kinetic
energy per molecule, moving around only a small region
surrounded by adjacent molecules. This is a rough picture of the
liquid state. On the other hand, in the gas state, most molecules
at high temperatures, where the average kinetic energy is large,
can move freely over the attractive potential well to expand the
free volume of the system.
In the supercritical fluid region near the critical temperature,
some molecules may move freely and some may be trapped to form
so-called weak clusters since kinetic energies of each molecule are
fluctuating around the average value clusters formed when the
molecular kinetic energy is smaller than the attractive energy
between adjacent molecules. In addition, these clusters are rapidly
changing in size and constitution due to molecular collisions. When
a solute molecule is thrown into the supercritical fluid, and if the
soluteesolvent attractive integration is larger than the
solventesolvent interaction, the solute molecule may be surrounded by the solvent molecules which form cluster because
attractive potential energy around the solute molecule is larger
than the average kinetic energy of the solvent (supercritical)
molecules. Clustering around a solute molecule is now considered
a major cause of enhanced solubility in supercritical fluids.
In 1987, Kim and Johnston (14) experimentally showed that the
local concentration of the fluid solvent molecules around a solute
molecule, phenol blue, is higher than the bulk concentration by
measuring the UV absorption wavelength shift of phenol blue in

594

SAITO

J. BIOSCI. BIOENG.,

FIG. 5. Behavior of molecules in gas, liquid, and supercritical state. Reproduced from Kajimoto (36) with permission of Kagakudojin.

various supercritical fluids such as CO2, CF3Cl, and CHF3. In 1988,
Kajimoto et al. (15) obtained the first experimental evidence of
solvation by cluster formation via measurements of the UV
absorption wavelength shift of 4-(N,N-dimethyl amino)benzonitrile
(DMABN) both in a supersonic jet and supercritical CHF3 with
various densities. They calculated the change in the number of the
fluid molecules around the solute molecule by employing a clustering model based on the Sutherland potential and Langmuir type
adsorption. Comparison of the experimental data agreed well with
the calculated values.
Symposia and workshops held in the 1980s In 1987, Smith
organized the first international workshop on SFC at Loughborough
University of Technology. Many research groups; Bartle’s, Smith’s,
Leyendecker’s, Lee’s, Sandra’s, Game’s, Lane’s and Saito’s groups,
gathered from Europe, US and Japan, and had an intensive discussion. Commercial instruments from Lee Scientific (open tubular
capillary column SFC) and JASCO (packed-column SFC and analytical SFE) were demonstrated. Contents of the discussion were
published as a monograph edited by Smith from Royal Society of
Chemistry (37).
In 1988, Perrut organized the first International Symposium on
Supercritical Fluids (ISSF) that covered a very wide range of
research on supercritical fluids, including industrial scale extraction, chromatography, phase equilibria, equations of state, etc., in
Nice, France. This symposium was very successful and gathered
many researchers in various fields from various countries. The ISSF
has been held in every 3 years and the latest one was held in May
2012 in San Francisco.
In 1988, Lee and Markides organized the 1988 Workshop on
Supercritical Fluid Chromatography in Park City, Utah. They have
organized the Workshop/Symposium in Utah a couple of times. The
4th workshop was held in Cincinnati, Ohio, and the 5th one was
held in Baltimore, Maryland. Lee and Markides (38) edited
a monograph that collected works presented in the series of
symposia and workshops in 1990.

DEVELOPMENT OF SFC AS A PRACTICAL TOOL IN ANALYTICAL
AND PREPARATIVE CHROMATOGRAPHY
Analytical SFC
The solvating power of a supercritical fluid
mobile phase depends on the density of the fluid. This means that
under the isobaric condition, the lower the temperature, the higher
the solvating power becomes. Thus, the lower the temperature, the
retention becomes shorter which is contrary to normal retention
behavior in GC and HPLC. In GC, the higher the vapor pressure, the
shorter the retention. This means that the higher the temperature,
the shorter the retention. Therefore, if the column temperature is
significantly higher than the critical temperature of the mobile
phase fluid, the fluid’s solvating power competes with the solute’s
vapor pressure. Fig. 6 shows the relationship between the
logarithm of capacity factor k0 and the reciprocal of column
temperature T (K) (39). At the temperatures of 2.4 (144 C) or

FIG. 6. Relationship between the logarithm of capacity factor k0 and the reciprocal of
column temperature T (K). Conditions: column, Capcell Pak CN, 5 mm; mobile
phase, CO2, 4 mL/min as liquid; pressure, constant at 20 MPa. Reproduced from Saito
et al. (39) with permission of John Wiley & Sons, Inc.

VOL. 115, 2013
lower, the retention (k0 ) decreases roughly linearly to the reciprocal
of the temperature; according to SFC theory. However, at the
temperature of 2.4 (144 C) or higher, the retention (k0 ) decreases
as well according to GC theory. At 2.4 (144 C) there are maxima
that are generated by the competition between the changes of
solvating power and the vapor pressure effect.
In the case of open tubular capillary SFC, it is often operated in
this temperature region, and makes it difficult to predict the
retention. Controlling the pressure by changing the flow velocity
further complicates the prediction. In open tubular capillary
column SFC, the mobile phase is often pure CO2 and a pressure
(density) gradient is used. While in the case of packed-column SFC
as stated before, it is more LC-like from view points of instrumentation, and it is common practice to add polar modifier to CO2 and
perform LC-like modifier gradient.
In packed-column SFC, chromatographers started to use organic
modifiers in higher percentage; a few to several 10s%. In such cases,
both critical temperature and pressure are rapidly elevated as
shown in Fig. 7 (40). For example, 5% (30%) methanol in CO2 gives
the critical temperature 51 C (135 C) and the critical pressure of
105 bar (168 bar) as shown in the gray box. Therefore, under
commonly used chromatographic conditions such as 100e120 bar
pressure and 40 C temperature, the mobile phase fluid is not in
a supercritical state. Cui and Olesik (41) started to use highconcentration modifiers in liquefied CO2 as early as in 1991. They
recognized that their mobile phase was not a supercritical fluid and
they called it “enhanced-fluidity mobile phase”. However, this term
was not generally accepted by SFC chromatographers and the term
supercritical fluid chromatography remains as is regardless of the
actual state of the fluid used. It should be noted that in such
conditions, the solvating power or retention can hardly be
controlled by changing the pressure because the temperature and
the pressure are well below the critical values of the binary mixture
fluid and the densities do not change much by the pressure. In
short, such a mobile phase is a simple mixture of liquefied CO2 gas
and an organic solvent, though when the fluid temperature and
pressure are a little under the critical values it may be called
a subcritical fluid. Advantages of this type of mobile phase are
lower viscosity than a liquid mobile phase and easy recovery of the

HISTORY OF SUPERCRITICAL FLUID CHROMATOGRAPHY

595

sample solute by decompression which is very useful when it is
used in preparative separation. An LC-like SFC system together
with the use of high-concentration modifier and modifier
gradient offered great flexibility in analytical work and chromatographers have finally found it as non-experimental ordinary
chromatograph.
Preparative SFC As discussed previously, Klesper foresaw the
possibility of preparative SFC (2), in their pioneering work in the
1960s (9) and 1970s (18,19). Saito and Yamauchi’s group
demonstrated the enrichment of tocopherol from wheat germ in
1989 (42) and the fractionation of lemon peel oil in 1990 (43) by
semi-preparative SFC using a 20-mm i.d. column. Berger and
Perrut (44) reviewed preparative SFC works in 1970s and in the 80s.
In 1992, Ute et al. (45) demonstrated isolation of methyl
methacrylate (MMA) oligomer, according to the degree of polymerization employing a negative temperature gradient. In 1995,
Saito and Yamauchi (46) separated flavanone enantiomers on a 20mm i.d. column with a stacked injection technique using a photodiode array UV/Vis detector (PDA). These works proved that SFC is
suitable for analytical and preparative separations; and that the
same SFC system could be used for both analytical and semipreparative applications. Chiral separation is now the most
successful application in SFC including analytical and preparative
separations.
Text books and commercialization of packed-column SFC
systems in the 1990s
Saito et al. (47) published a monograph
that describe the practice of SFE and packed-column SFC
including preparative SFC in 1994. T.A. Berger published
a monograph on packed-column SFC in 1995 (48). In 1998, Klaus
and C. Berger (not related to T.A. Berger) (49) published a book on
SFC with packed columns. These books encouraged
chromatographers to use SFC, thus, packed-column SFC became
the main stream of SFC by the late 1990s, and packed-column SFC
instrumentation became commercially available from several
sources; Hewlett Packard (later Berger Instrument), JASCO, Gilson,
Novasep, etc.
Standard
configuration
of
a
packed-column
SFC
system The standard configuration of a packed-column SFC

FIG. 7. Relationship between the calculated critical temperature, pressure and mass % of a CO2-methanol mixture. Recalculated using the program by Saito and Nitta (40) with
permission of John Wiley & Sons, Inc.

596

SAITO

J. BIOSCI. BIOENG.,
TABLE 3. Overview of history of development of SFC.

Publication year

Authors (ref. no.)

Application/Event

Mobile phase

Stationary phase

Apparatus/Detector

1962

Klesper et al. (2)

Porphyrin mixtures

1966, 1967

Sie et al. (5e8)

Paraphins

Diclorodifluoroethane
monochlorodifluoromethane
CO2

1968

Karayannis et al. (9)

Porphyrin mixtures

1969

Giddings et al. (11)

1970

Jentoft and Gouw (16)

Purines, nucleosides and
nucleotides, steroids,
sugars, terpenes, amino
acids, proteins,
carbowaxes, etc.
PAHs, styrene oligomer

1972

Jentoft and Gouw (17)

Preparative separation of
above solutes and fraction
collection

CO2

1977

Hartman and
Klesper (18,19)

Fractionation of styrene
oligomer

n-pentane þ methanol

1981

Novotny et al. (22)

CO2

Open tubular capillary
column/dimethyl
polysiloxan

1982

Gere et al. (24)

Various chemicals such as
drugs, natural products,
etc.; separation of styrene
oligomers were often
demonstrated to show its
high resolution
PAHs

CO2

ODS

1985

Sugiyama et al. (27)

CO2

Silica gel

1985

Mourier et al. (31)

CO2

Pirckle type CSP

1986

Hara et al. (32)

CO2 þ methanol

Homemade CSP

1986

Perrut and
Jusforgues (33)

CO2

NA

1987

Saito et al. (28)

CO2

Silica gel

LC-like commercial SFC
system/PDA

1987

Kim and Johnston (14)

CO2, CF3Cl, and CHF3

NA

NA

1987
1988

Smith (37)
Perrut

NA
NA

NA
NA

NA
NA

1988
1990

NA
CO2 þ ethanol

NA
Silica gel

1991

Lee and Markides
Yamauchi and
Saito (43)
Cui and Olesik (41)

CO2 þ methanol

Hypercarb PGC

NA
LC-like commercial SFC
system(JASCO)/PDA
HP GC/Isco syringe pump

1992

Ute et al. (45)

CO2

Silica gel

Commercial SFC system/HP
GC oven/negative
temperature gradient

1995

Saito and
Yamauchi (46)

CO2 þ ethanol

Silica gel (20 mm i.d.
column)

Commercial semi-prep-SFC
system

2001

Wang et al. (61)

CO2 þ methanol

NA

2006

Zheng et al. (80)

Caffeine extraction and
separation
Chiral separation of
phosphine oxides
Chiral separation of d-l amino
acid derivatives
Large scale preparative SFC
(60 mm i.d. column) with
CO2 recirculation
PAHs, experimentally
showed outlet mass flow
reduction and elucidated
the phenomenon
theoretically
Experimentally showed
clustering in supercritical
fluids
First Workshop on SFC
First international
symposium on
supercritical fluids
Workshop on analytical SFC
Fractionation of lemon peel
oil
High-concentration modifier
enhanced-fluidity mobile
phase
Isolation of methyl
methacrylate (MMA)
oligomer, according to the
degree of polymerization
Chiral fractionation of
flavanone enantiomers on
a 20-mm i.d. column with
stacked injections
Mass-directed fractionation
for drug discovery
SFC/MS of polypeptides

2005
2006

Xu et al. (81,82)

Estrogen metabolites

CO2 þ (methanol þ
trifluoroacetic acid)
CO2 þ methanol

2012

Bamba et al. (85)

Metabolite analysis, review

CO2 þ modifier

2-Ethylpyridine bonded
silica column
Cyanopropyl silica column
connected in series with
a diol column
NA

Homemade semi-prepSFCeMS
Commercial SFCeMS

Polyethylene glycol

GC-like very simple
homemade system/FID

Diclorodifluoroethane

Silica gel coated with
glycerol, squalene
Chromosorb/cabowax 20 M

CO2, NH3, etc.

Chromosorb/silicone oil

GC-like relatively simple
homemade system/FID
GC-like sophisticated
homemade system/
photometer
GC-like sophisticated
homemade system/FID
Ultra-high-pressure
system, details unknown

CO2

Woelm basic alumina.
Porosil/n-pentane
polystyrene
divinylbenzene
Woelm basic alumina.
Porosil/n-pentane
polystyrene
divinylbenzene
Porosil

LC-like very sophisticated
homemade system/
UV photometer/pressure
programming
Automated fraction collector
added to the above
system
LC-like homemade very
sophisticated system/UV
photometer/pressure
programming
GC-like system with syringe
pump/FID

Modified Commercial HPLC
system/backpressure
regulator
LC-like sophisticated
system/PDA
Modified Varian HPLC
system
LC-like commercial SFC
system(JASCO)/PDA
Semi-pilot plan scale
preparative SFC

Commercial SFCeMS/MS

Commercial SFCeMS/MS

VOL. 115, 2013
system established in the 1990s comprises of a reciprocating CO2
pump with chilled pump heads, a reciprocating modifier solvent
pump, a manual or an automated injection device, a column oven,
a UV absorption detector (typically a PDA detector), a backpressure
regulator that allows the pressure control independent of the flow
rate, and a chromatography data system (CDS) as shown in Fig. 4.
Other types of detectors have also been used.
Various detection systems Randall wrote a long article in
1982, on dense (supercritical) gas chromatographyemass
spectrometry (MS), trying to stimulate research in this area (50).
There are many articles on capillary column SFCeMS appeared in
the 1980s (51e53). Crowther and Henion (54) reported packedcolumn SFCeMS for polar drug analysis. Those works in the 1980s
were reviewed by Kalinoski et al. (55) in 1986 and Sheeley and
Reinhold (56) in 1989. However, practical application of packedcolumn SFCeMS started in the 1990s after successful interfacing
with atmospheric pressure ionization, i.e., APCI and ESI (57). These
works were reviewed by several researchers (58e60).
In 2001, Wang et al. (61) reported mass-directed fractionation
and isolation by packed-column SFC/MS, and proposed a fractionation method utilizing the matching of mass spectra of the sample
compounds and those stored in the library. Zhang et al. (62)
developed a similar mass-directed preparative SFC purification
system in 2006. Recently, Li and Hsieh (63) reviewed SFCeMS. MS
detection is now a very powerful and indispensable method to
accurately identify the target compound especially in the pharmaceutical industry (64).
In chiral separation, a chiral detector plays an important role as
no other detector can differentiate chiral compounds. A chiral
detector is an optical detector by nature. There are two types of
detectors based on different optical property. One is based on
optical rotation (OR) and the other on circular dichroism (CD). An
OR detector measures the difference in refractive index of enantiomers, whereas a CD detector measures the difference in optical
absorption. A refractive index is subject to the change of temperature and density, thus, it is extremely difficult to obtain a stable
baseline in SFC. On the other hand, an absorption-based CD signal
is very stable as a UV absorption detector, even in SFC. In addition,
the g-factor (CD/UV signal), indicates enantiopurity independent of
the peak concentration. Kanomata et al. (65) reported advantages
of CD detection in SFC especially when it is employed in preparative SFC.
FID has been the standard detector in open tubular capillary SFC.
In packed-column SFC, the stationary phase is much stronger than
that in open tubular capillary SFC, therefore, the addition of a polar
modifier is necessary to elute a sample solute. As discussed before,
even a small amount of organic modifier interferes with an FID
however, it is still used with packed columns for specific analyses
such as analyses of petroleum fuels using pure CO2 as a mobile
phase. These methods are published by ASTM as D5186 (66) and
D6550 (67).
Evaporative light scattering detection (ESLD) is regarded as
a pseudo-universal detector in HPLC and SFC. Since it does not
require analytes to have UV absorption, it is a preferred detector
in SFC in place of a refractive index detector that is not
compatible with the high-backpressure required by SFC. Early
attempts in using an ESLD in SFC were carried out by Carraud
et al. (68), by Nizery et al. (69), and by Hoffmann and Greibrokk
(70) in the late 1980s. In 1996, Strode and Taylor (71) reviewed
previous works of SFC-ESLD and investigated optimum conditions under various elution modes such as pressure gradient and
modifier gradient.
To facilitate the readers to overview the half-a-century long
history of the development of SFC, publications and events related
to SFC are listed in chronological order in Table 3.

HISTORY OF SUPERCRITICAL FLUID CHROMATOGRAPHY

597

RECENT TRENDS IN SUPERCRITICAL FLUID CHROMATOGRAPHY
As discussed in the previous section, standardized SFC systems
finally became commercially available from several sources by the
late 1990s and were recognized as a truly useful separation
instruments. However, the history of SFC was of “the rise and fall”,
as Smith (72) wrote in his review article in 1999. Harris (73) wrote
a very interesting story in her review article based on interviews
with Karen Phinney of NIST and other SFC experts. “When Phinney
joined NIST, she said to one of her colleagues, ‘I’m doing SFC’, ‘Oh,
science fiction chromatography’ replied the colleague.” This
happened in the mid 1990s.
As we are in the 21st century and the resources are growing
scarce, demands for sustainable chemistry or green chemistry have
been getting stronger. CO2-based SFC could significantly reduce the
use of organic solvents in the separation process, and also energy in
a fractionation and evaporation processes. Therefore, SFC has been
expected to contribute greatly to green chemistry. In fact, the Green
Chemistry Group (Oakmont, PA, USA), has been promoting the
International Conference on Packed-Column SFC every year since
2008.
The author’s intention, as indicated by the title, is to review
history of supercritical fluid chromatography in view of instrumental developments and he will not go into the details about SFC
applications. Nevertheless, the author would like to touch on some
of important research works. Since the late 1990s, SFC research has
been focused on expansion of application areas associated with
development of column technology.
CO2-based SFC is normal phase chromatography and as such,
polar analytes are difficult to separate by SFC. In order to expand
the application areas, there are many attempts to separate such
analytes (25,26,74e77). Works done in the 1980s and 1990s were
intensively reviewed by Berger (78) in 1997.
In the 21st century, advances in column and mobile phase
chemistry finally allowed the analysis of biomolecules that were
difficult to separate by SFC (79). In addition, advances in tandem
mass spectrometry enabled the analysis of a very small amount of
biomolecules and extended SFCeMS application range to metabolism research. Taylor’s group successfully analyzed a polypeptide
with up to 40-mers using trifluoroacetic acid as additive in a
CO2/methanol mobile phase to suppress the deprotonation of the
peptide carboxylic acid groups and to protonate the peptide
amino groups on a 2-ethylpyridine bonded silica column, which
was specifically developed for SFC (80). They Xu et al. (81,82)
separated 15 estrogen metabolites by using SFCeMS/MS with
a mobile phase of CO2-methanol mixture in gradient mode on
a cyanopropyl silica column connected in series with a diol column
in 10 min, whereas HPLCeMS/MS required 70 min. Bamba’s group
(83e85) performed extensive study on metabolic profiling of
various natural products including lipids and polar lipids using
SFCeMS/MS.
To conclude this article, the author would like to mention that
there are a huge number of articles published since the advent of
modern SFC, however, he had to mention less than 1% of such
articles on a sampling basis. The readers may find that this review
article examines many older works from the 1960se1980s.
However, it is the intention of author to introduce these
pioneering works to the readers which are often omitted in more
recent review articles.

References
1. Takishima, S. and Masuoka, H.: Kagakukogaku bussei josu (Physical properties of substances related to chemical engineering), 11, 31 (1989) (in
Japanese).

598

SAITO

2. Klesper, K., Corwin, A. H., and Turner, D. A.: High pressure gas chromatography above critical temperatures, J. Org. Chem., 27, 700e701 (1962).
3. Taylor, L. T.: Supercritical fluid chromatography for the 21st century,
J. Supercrit. Fluids, 47, 566e573 (2009).
4. Guiochon, G. and Tarafder, A.: Review Fundamental challenges and opportunities for preparative supercritical fluid chromatography, J. Chromatogr. A,
1218, 1037e1114 (2011).
5. Sie, S. T., Van Beersum, W., and Rjinders, G. W. A.: High-pressure gas chromatography and chromatography with supercritical fluids. I. The effect of
pressure on partition coefficients in gas-liquid chromatography with carbon
dioxide as a carrier gas, Sep. Sci., 1, 459e490 (1966).
6. Sie, S. T. and Rjinders, G. W. A.: High-pressure gas chromatography and
chromatography with supercritical fluids. II. Permeability and efficiency of
packed columns with high-pressure gases as mobile fluids under conditions of
incipient turbulence, Sep. Sci., 2, 699e727 (1967).
7. Sie, S. T. and Rjinders, G. W. A.: High-pressure gas chromatography and
chromatography with supercritical fluids. III. Fluid-liquid chromatography, Sep.
Sci., 2, 729e753 (1967).
8. Sie, S. T. and Rjinders, G. W. A.: High-pressure gas chromatography and
chromatography with supercritical fluids, IV. Fluid-solid chromatography, Sep.
Sci., 2, 755e777 (1967).
9. Karayannis, N. K., Corwin, A. H., Baker, E. W., Klesper, E., and Walter, J. A.:
Apparatus and materials for hyperpressure gas chromatography of nonvolatile
compounds, Anal. Chem., 40, 1736e1739 (1968).
10. Snyder, L. R. and Kirkland, J. J. (Eds.): Introduction to modern liquid chromatography, 1st ed., p.130e131. John Wiley & Sons, New York (1979).
11. Giddings, C. G., Myers, M. N., and King, J. W.: Dense gas chromatography at
pressures to 2000 atmospheres, J. Chromatogr. Sci., 7, 276e283 (1969).
12. Ishii, D., Niwa, T., Ohta, K., and Takeuchi, T.: Unified capillary chromatography, J. High Resol. Chromatogr., 11, 800e801 (1988).
13. Chester, T. L. and Pinkston, J. D.: Supercritical fluid and unified chromatography, Anal. Chem., 76, 4606e4643 (2004).
14. Kim, S. and Johnston, K. P.: Molecular interactions in dilute supercritical fluid
solutions, Ind. Eng. Chem. Res., 26, 1206e1213 (1987).
15. Kajimoto, O., Futakami, M., Kobayashi, T., and Yamasaki, K.: Charge-transfer-state formation in supercritical fluid: (N, N-dimethylamino)benzonitrile in
trifluoromethane, J. Phys. Chem., 92, 1347e1352 (1988).
16. Jentoft, R. E. and Gouw, T. H.: Pressure-programmed supercritical fluid
chromatography of wide molecular weight range mixtures, J. Chromatogr. Sci.,
8, 138e142 (1970).
17. Jentoft, R. E. and Gouw, T. H.: Apparatus for supercritical fluid chromatography
with carbon dioxide as the mobile phase, Anal. Chem., 44, 681e686 (1972).
18. Klesper, E. and Hartmann, W.: Supercritical fluid chromatography of styrene
oligomers, J. Polym. Lett. Ed., 15, 9e16 (1977).
19. Hartmann, W. and Klesper, E.: Preparative supercritical fluid chromatography
of styrene oligomers, J. Polym. Lett. Ed., 15, 713e719 (1977).
20. Randall, L. G. and Wahrhaftig, A. L.: Dense gas chromatograph/mass spectrometer interface, Anal. Chem., 50, 1703e1705 (1978).
21. Schneider, G. M., Stahl, E., and Wilke, G. (Eds.): Extraction with supercritical
fluid, Verlag Chemie, Weinheim (1980).
22. Novotny, M., Springton, S. R., Peaden, P. A., Fjeldsted, J., and Lee, M. L.:
Capillary supercritical fluid chromatography, Anal. Chem., 53, 407Ae414A
(1981).
23. Novotny, M., Bertsch, W., and Zlatkis, A.: Temperature and pressure effects in
supercritical-fluid chromatography, J. Chromatogr., 61, 17e28 (1971).
24. Gere, D. R., Board, R., and McManigill, D.: Supercritical fluid chromatography
with small particle diameter packed columns, Anal. Chem., 54, 736e740 (1982).
25. Taylor, L. T. and Change, H.-C.: Packed column development in supercritical
fluid chromatography, J. Chromatogr. Sci., 28, 357e366 (1990).
26. Smith, R. M. and Sanagi, M. M.: Retention and selectivity in supercritical fluid
chromatography on an octadecylsilyl-silica column, J. Chromatogr. A, 505,
147e159 (1990).
27. Sugiyama, K., Saito, M., Hondo, T., and Senda, M.: New double-stage separation analysis method: directly coupled laboratory-scale supercritical fluid
extractiondsupercritical fluid chromatography, monitored with a multiwavelength ultraviolet detector, J. Chromatogr., 332, 107e116 (1985).
28. Saito, M., Yamauchi, Y., Kashiwazaki, H., and Sugawara, M.: New pressure
regulating system for constant mass flow supercritical-fluid chromatography
and physico-chemical analysis of mass-flow reduction in pressure
programming by analogous circuit model, Chromatographia, 25, 801e805
(1988).
29. Okamoto, Y., Kawashima, M., and Hatada, K.: Useful chiral packing materials
for high-performance liquid chromatographic resolution of enantiomers:
phenylcarbamates of polysaccharides coated on silica gel, J. Am. Chem. Soc.,
106, 5357e5359 (1984).
30. Okamoto, Y., Kawashima, M., and Hatada, K.: Chromatographic resolution:
XI. Controlled chiral recognition of cellulose triphenylcarbamate derivatives
supported on silica gel, J. Chromatogr., 363, 173e186 (1986).
31. Mourier, P. A., Eliot, E., Caude, M. H., Rosset, R. H., and Tambute, A. G.:
Supercritical and subcritical fluid chromatography on a chiral stationary phase

J. BIOSCI. BIOENG.,

32.

33.
34.

35.
36.
37.
38.
39.

40.

41.
42.

43.

44.
45.
46.

47.
48.
49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

for the resolution of phosphine oxide enantiomers, Anal. Chem., 57,
2819e2823 (1985).
Hara, S., Dobashi, A., Kinoshita, K., Hondo, T., Saito, M., and Senda, M.:
Carbon dioxide supercritical fluid chromatography on a chiral diamide
stationary phase for the resolution of D- and L-amino acid derivatives,
J. Chromatogr. A, 371, 153e158 (1986).
Perrut, M. and Jusforgues, P.: Un nouveau procédé de fractionnement: la chromatographie préparative avec éluant supercritique, Entropie, 132, 3e9 (1986).
Phinney, K. W.: Review enantioselective separations by packed column
subcritical and supercritical fluid chromatography, Anal. Bioanal. Chem., 382,
639e645 (2005).
Mangelings, D. and Heyden, Y. V.: Review Chiral separations in sub- and
supercritical fluid chromatography, J. Sep. Sci., 31, 1252e1273 (2008).
Kajimoto, O.: Supercritical fluid: visual observation of its characteristics,
Kagaku, 47, 444e450 (1992) (in Japanese).
Smith, R. M. (Ed.): Superc