Trends in Analytical Chemistry, Vol. 29, No.

7, 2010 Trends

Greening analytical
chromatography
Christopher J. Welch, Naijun Wu, Mirlinda Biba, Robert Hartman,
Tanja Brkovic, Xiaoyi Gong, Roy Helmy, Wes Schafer, James Cuff,
Zainab Pirzada, Lili Zhou

We review recent approaches for greening analytical separation technolo- The nascent field of green analytical
gies. Whereas conventional analytical separation technologies produce as chemistry, recently reviewed by de la
much as 50 mL of waste per analysis, greener options considerably reduce Guardia [5], is concerned with the reduc-
the amount of waste generated per analysis, ranging from simply adoptable tion or elimination of the use or genera-
small-particle stationary-phase technologies and reduced column diameters tion of reagents or byproducts that are
to more specialized technologies [e.g., ultra-high-pressure liquid chroma- harmful to human health or the environ-
tography (UHPLC), supercritical-fluid chromatography (SFC) and microscale ment. Green analytical chemistry can take
or nanoscale HPLC]. In addition, greener eluents and chromatography con- many forms, but most segments of this
ditions can also be used, including substitution of ethanol for acetonitrile in emerging field minimize the amount of
reversed-phase HPLC and the use of water alone as an eluent at elevated waste associated with either the sample
temperatures. We describe the merits of these emerging technologies, and preparation or analysis. The greening of
discuss the factors influencing the decision to move from conventional an- sample preparation is an exciting field that
alytical separation methodologies to greener alternatives. has recently been reviewed by Namiesnik
ª 2010 Elsevier Ltd. All rights reserved. and co-workers [6]. In this review, we
survey recent advances in the greening of
Keywords: Acetonitrile; Analytical separation; Ethanol; Fast chromatography; Green
analytical separation technologies.
analytical chemistry; Microflow chromatography; Solvent saving; Supercritical-fluid
chromatography; Superheated water chromatography; Waste
2. Greening analytical separation
1. Introduction technologies
Christopher J. Welch*,
Naijun Wu*,
Mirlinda Biba, In recent years, the subject of green High-performance liquid chromatography
Robert Hartman, chemistry has gained increasing attention (HPLC) and related chromatographic
Tanja Brkovic, and acceptance from researchers in both techniques are the most widely utilized
Xiaoyi Gong, industry and academia [1–3]. Under- analytical tools in chemical research. In its
Roy Helmy,
standably, early efforts targeted improving most common current incarnation, ana-
Wes Schafer,
James Cuff, process economy, waste reduction, and lytical HPLC is performed using a station-
Zainab Pirzada, elimination of risks and hazards in those ary-phase column of 4.6 mm internal
Lili Zhou large-scale industrial processes where the diameter (i.d.) and 25 cm length at a
Merck & Co., Inc., greatest environmental impact could be mobile-phase flow rate of about 1–1.5 mL/
Separation & Purification
realized. There has been recent growing min [7]. Under these conditions, each
Center of Excellence, RY 801,
A100 Rahway, NJ 07065, USA interest in the application of green-chem- operating HPLC instrument generates on a
istry principles to all aspects of laboratory daily basis well over 1 L of effluent that
work, where more modest savings, multi- must be disposed of as chemical waste.
plied by a greater number of individual While 1 L of waste solvent may be incon-
researchers, can have a significant sequential in the industrial manufacturing
cumulative impact [4]. Consequently, the realm, the pervasiveness of analytical
previous focus on greening chemical pro- HPLC instrumentation makes the cumu-
duction and purification technologies has lative effect substantial. For example, it is
*
recently expanded to include other labo- not uncommon for a single pharmaceuti-
Corresponding authors.
ratory technologies as well. cal company to have well over 1000
E-mail: christopher_welch@
merck.com

0165-9936/$ - see front matter ª 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.trac.2010.03.008 667
Trends
operating HPLC instruments. In addition, HPLC is
becoming more intensively used in recent years, with
automated technologies gaining favor to allow round-
the-clock operation of instruments.
HPLC is a workhorse analytical tool that delivers tre-
mendous benefit to the chemical-research community.
For many applications, the value produced by an HPLC
instrument in a day of operation easily exceeds the com-
bined cost of the purchase of 1 L of solvent, and the asso-
ciated waste disposal. However, a fundamental question
remains: ‘‘Could these same results be obtained with de-
creased waste generation?’’. Recent technological ad-
vances have shown that the answer to this question is a
resounding, unequivocal ‘‘Yes!’’. A variety of new, re-
cently available, analytical separation technologies can
deliver the required analytical results while significantly
reducing solvent consumption, waste generation, or both.
3. The three Rs of green analytical chemistry
In addition to the well-known 12 principles of green
chemistry [1], three Rs (Reduce, Replace, Recycle)
(Fig. 1) are commonly mentioned in connection with
green analytical chemistry. To date, most of the efforts at
greening analytical chromatography have focused on
either replacement of existing solvents with greener
alternatives or on overall reduction in the amount of
solvent used and waste generated. Solvent recycling via
distillation [8] or other methods [9] has received only
limited investigation in analytical chromatography, al-
though the technique is widely used in industrial-scale
chromatography.
4. Solvent reduction: fast chromatography is green
4.1. Fast separation using conventional HPLC with
elevated pressures
The most straightforward and easiest way to reduce the
solvent consumption of an HPLC separation is to reduce
the run time. The mobile-phase consumption (V) for a
given separation can be expressed as:
V ¼Ft ð1Þ
where F is the volumetric flow rate and t is the run time
of the separation. A fast, efficient separation can easily be
achieved by using a short column packed with smaller
particles, since column efficiency per unit length in-
creases with decreasing particle size. The column length,
L, required for a given separation is proportional to the
particle size [10]:
L ¼ 2N req  d p ð2Þ
where Nreq is the required column efficiency for a given
separation and dp is the particle size of the packing
668 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
Figure 1. The three Rs of green separation technologies. Most
green separation technologies focus on reduction and replacement,
with recycling being important primarily in larger scale preparative
separation technologies.
material. For example, a 100 · 4.6-mm column packed
with 1.8-lm particles can provide column efficiency
similar to a 250 · 4.6-mm column packed with 5-lm
particles (as illustrated in Fig. 2). It can be seen that,
while both columns provide similar separation, the sep-
aration speed for the 100 · 4.6-mm column (1.8-lm
particles) is approximately 2.5 times faster than for the
250 · 4.6-mm column (5-lm particles) simply because
the former is 2.5 times shorter. As a result, solvent
savings of about 2.5-fold can be achieved by reducing
the column length and the particle size. However, it
should be noted, that the column pressure for the
100 mm, 1.8 lm column is much higher even though
the same flow rate was used, since the pressure drop for
a packed column increases faster with decreasing par-
ticle size than it decreases by reducing column length
[11]. Nevertheless, the pressure is still within allowable
pressure ranges for a conventional instrument.
4.2. Fast separation using ultra-high performance
liquid chromatography (UHPLC) systems
Recent years have seen a veritable explosion in new and
improved methods for speeding up chromatographic
analysis, mostly based upon increased use of smaller
particle (sub-2 lm) stationary phases operating at
higher back-pressures [11–13]. Theoretically, much
higher flow rates can be used for sub-2 lm particles to
achieve faster separation, since the optimum linear
velocity is inversely proportional to the particle size [12].
For example, a flow rate as high as 3.3 mL/min can be
used for the 100 · 4.6-mm column packed with 1.8-lm
particles from the previous example. Obviously, no waste
reduction can be achieved by this approach, since the
solvent saving through reducing the separation time are
compensated for by the solvent increase through using a
higher volumetric flow rate, as shown in the Equation
(1). Furthermore, such high volumetric flow rates on a
4.6-mm column can result in a frictional heating be-
tween the mobile phase and the stationary phase, which
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010 Trends

30
(a) 60 mL waste
20 per analysis
Response

10

0
0 5 10 15 20 25 30 35 40
Retention Time (min)

(b) 23 mL waste
20
per analysis
Response

10

0
0 2 4 6 8 10 12 14
Retention Time (min)
Figure 2. A shorter column packed with smaller particles and operating at higher pressure affords comparable resolution of a drug candidate and
process impurities in shorter time, leading to a 2.5· reduction in solvent use and waste generation. Conditions: (a) 250 · 4.6 mm, 5 lm, Zorbax
SB C18, 210 nm, 10 lL injection, 40C, 1.2 mL/min, 1730 psi, 75/25 to 5/95 0.1%H3PO4/MeCN in 40 min, 10 min post run, Agilent1100. (b)
100 · 4.6 mm, 1.8 lm, Zorbax SB C18, 210 nm, 5 lL injection, 40C, 1.2 mL/min, 5600 psi, 5/25 to 5/95 0.1%H3PO4/MeCN, in 15 min,
5 min post run, Agilent1100.

can negatively impact column performance [14]. To
reduce solvent consumption and minimize the frictional
heating effect, a smaller i.d. column can be used with
sub-2 lm particles to achieve higher optimum linear
velocity. Reducing the i.d. of the column can dramati-
cally reduce the generation of waste solvent, as described
in a modified form of Equation (1):
p mobile phase was used per analysis by UHPLC with a
V ¼ d 2c  u  t ð3Þ
4
where dc is the column i.d. and u is the linear velocity. A
flow rate of 0.25 mL/min can therefore be used for a
100 · 2.1-mm column packed with 1.7-lm particles to
achieve the same linear velocity, separation speed and
efficiency (as shown in Fig. 3a), affording about five-fold
additional solvent savings. For a given instrument,
extra-column volume effects become more significant as
the column diameter is reduced and as the peak width
decreases, so, for optimal performance, extra-column
volume should decrease commensurate with decreases
in column volume.
In addition, the separation speed can be further in-
creased by using higher flow rates, which require much
higher back-pressures. Recently, several commercial
UHPLC systems have been introduced to overcome the
high back-pressures generated by small particles while
minimizing extra-column volume effects for smaller
dimensional columns. Fig. 3b shows a separation that
employed a Waters UPLC system using a 100 · 2.1-mm
column (1.7-lm particles) operating at 0.65 mL/min.
The separation is three times faster than in Fig. 2A, due
to the use of a higher flow rate. By comparing Fig. 2a
and Fig. 3b, it can be seen that 60-mL mobile phase was
consumed per analysis by a conventional LC system with
a 250 · 4.6-mm, 5-lm column, while only 4.9-mL
100 · 2.1-mm, 1.7-lm column – an overall 92%
reduction in solvent use and waste generation, with only
minor differences in selectivity observed between the two
separations.
In general, these advantages can be obtained irre-
spective of separation method or analyte structure, and,
while some variation in the savings may be observed, it
is evident the UHPLC fast analysis technologies using
smaller particles, higher pressures and shorter columns
with smaller diameters offer a greener alternative to
conventional HPLC. Chen and Kord recently reported
that a UPLC system with a 2.1-mm i.d. column could
afford solvent savings overall of 80% compared to
conventional HPLC [15].
4.3. Conventional HPLC systems with minor
modifications for greener separation
It is important to note that the purchase of new UHPLC
instruments is not essential for implementing greener

http://www.elsevier.com/locate/trac 669
Trends Trends in Analytical Chemistry, Vol. 29, No. 7, 2010

30
(a) 5.0 mL waste
20
per analysis
Response

10

0
0 2 4 6 8 10 12 14
Retention Time (min)

(b) 4.9 mL waste

200
per analysis
Response

100

0 1 2 3 4 5
Retention Time (min)
Figure 3. Fast separation of a drug candidate with its in-process impurities using ultra-high-performance liquid chromatography (UHPLC). Oper-
ation at higher flow rate and pressure affords faster chromatography, although solvent use remains unchanged. Conditions: (a) 100 · 2.1 mm,
1.7 lm Shield C18, 210 nm, 2 lL injection, 40C, 0.25 mL/min, 6020 psi, 75/25 to 5/95, 0.1%H3PO4/MeCN, in 5 min, 5 min post run, Agilent
1200. (b) 100 · 2.1 mm, 1.7 lm Shield C18, 210 nm, 2 lL injection, 40C, 0.65 mL/min, 14,500 psi, 75/25 to 5/95 0.1%H3PO4/MeCN in 5 min,
2 min post run, Acquity UPLC.

analytical HPLC in the laboratory. Simple approaches
(e.g., use of new column technologies or minor plumbing
modifications of existing HPLC instrumentation) can
often enable significant reductions in generation of waste
solvent.
Significant solvent savings can be obtained simply by
changing column dimensions. In order to maintain the
separation speed or the linear velocity, while reducing
column i.d., the flow rate must be reduced proportionally
to the square of the ratio of the two radii, as described by
Equation (4):
 2
r2
F2 ¼ F1 ð4Þ
r1
where F2 and r2 are the flow rate and the radius for the
smaller i.d. column, and F1 and r1 are the flow rate and
the radius for the larger i.d. column. A reduction in
column i.d. from 4.6 mm to 3.0 mm with constant
length requires the flow rate to be reduced by a factor of
0.43 [i.e. (3.0/4.6)2], affording a 57% reduction in sol-
vent consumption and waste generation for any isocratic
separation.
In Fig. 4, we present an example of a greener sepa-
ration using a conventional HPLC system. The gradient
separation of six process impurities from a drug-
substance intermediate was initially achieved on a C18,
100 · 4.6-mm, 2.7-lm column at a flow rate of 1.5 mL/
min, with a combined separation and equilibration time
of 27.5 min, resulting in consumption of 41.3 mL of
mobile phase and generation of an equivalent volume of
waste. By simply changing the i.d. of the column from
4.6 mm to 3.0 mm and scaling the flow rate with respect
to the column and the system dwell volume, the solvent
consumption and subsequent waste stream were re-
duced by 47% with virtually no change in the quality of
the separation.
In combination with the use of smaller particle col-
umns and operation at slightly elevated pressure, sig-
nificant solvent savings can be realized. As discussed
previously, re-plumbing such systems is often required
for optimal performance, as decreasing dwell volume and
extra-column volume to the absolute minimum is
essential for optimal performance at these lower flow
rates to avoid overly long re-equilibration times and
extra-column band broadening, respectively. While
these types of instrument modification have long been
performed by expert users, conventional HPLC instru-
ments with optimized and improved dwell volumes have
begun to appear in the marketplace in recent years (e.g.,
AgilentÕs model 1200 RRHT). Presumably, this trend
will continue, leading to a continuing decrease in col-
umn diameter and solvent utilization in LC.

670 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
Figure 4. Significant reduction in solvent consumption and waste generation can be achieved with conventional HPLC instrumentation by
simply reducing column diameter from 4.6 mm to 3.0 mm. a) 100 · 4.6-mm column operating at 1.5 mL/min, total solvent consumption =
41.3 mL. b) 100 · 3-mm column operating at 0.8 mL/min, total solvent consumption = 22.0 mL. Both columns are packed with C18 stationary
phase, 2.7 lm, and have a total run time per assay of 27.5 min.
4.4. Newer HPLC column technologies for greener
separation
In addition to the use of HPLC columns packed with sub-
2 lm stationary-phase particles, there has also been an
increased interest in employing superficially porous
packings, also known as Fused-Core particle columns,
which provide the advantage of small-particle columns
without the need for ultra-high pressures. These col-
umns employ a solid core surrounded by a 0.5–0.34 lm
porous silica layer.
Cunliffe and Maloney recently compared the perfor-
mance of columns packed with Fused-Core particles to
those packed with sub-2 lm particles under isocratic and
gradient conditions [16]. The Fused-Core columns of-
fered lower back-pressure compared to the sub-2 lm
columns typically employed in UHPLC with only a slight
sacrifice in peak efficiency. In addition, these columns
can provide higher column efficiency per unit length
compared to 3-lm or 5-lm-particle packed columns.
Consequently, more efficient, fast separations and thus
solvent savings can often be achieved using relatively
shorter columns on a conventional HPLC system.
It is important to note that not all fast chromato-
graphic techniques are sparing in solvent use. For
example, monolithic columns have a long history of
usage for speeding chromatographic analysis, the
advantage of the technique being that the columns can
be operated at elevated flow rates without generating
excessive back-pressure [17–20]. Faster chromato-
graphic analysis is often possible using this approach,
although the amount of waste generated per data point
is typically the same as or even greater than conven-
tional HPLC, when columns of comparable dimensions
Trends
are used. Unfortunately, the availability of commercial
narrow-bore monolithic columns is limited.
5. Solvent reduction: microflow and capillary
HPLC
Another approach for green analytical chemistry focuses
on dramatically reducing the overall scale of the ana-
lytical chromatographic experiment. HPLC technology
was first commercialized around 1970, and subsequent
industry standardization on 4.6-mm i.d. columns with
flow rates in the neighborhood of 1 mL/min is something
of a historical relic driven by equipment limitations from
that time period, rather than by performance consider-
ations [21]. We have seen in Section 4 how reducing
column diameter can reduce solvent consumption and
waste generation. The use of microscale columns and
lower flow rates for HPLC has been known for many
years, and a number of early researchers, particularly
Novotny, have long pointed out the important solvent
savings that can be gained with such columns when
flow rates are reduced to only a few mL/min [22].
Microflow chromatography has gained renewed
interest in recent years, driven largely by considerations
having little to do with waste reduction and environ-
mental responsibility. However, as these instruments
become more intensively utilized, the green-chemistry
advantages are beginning to have an impact in labora-
tories [23]. As a case in point, our entry into the use
of microflow HPLC technologies for supporting
pharmaceutical-process research stems from our interest
in creating new capabilities for carrying out high-
http://www.elsevier.com/locate/trac 671
Trends
throughput analysis to support high-throughput exper-
imentation workflows. As part of this effort, we collab-
orated with Eksigent Technologies on the development of
a multi-parallel microflow HPLC instrument utilizing
eight columns of 300-lm i.d., each under independent
binary pump flow control [24,25]. In this instance, the
choice of lL rather than conventional mL HPLC flow
rates stemmed from a need to co-locate the required 16
high-pressure pumps for the device in a reasonably-sized
instrument console. By using miniaturized pumps
capable of accurately delivering flows of a few lL/min, it
was possible to develop an instrument with a reasonable
laboratory footprint. Subsequent studies have shown the
utility of this instrument for not only high-throughput
analysis, but also multi-parallel method-development
studies, where different columns and mobile phases are
simultaneously investigated [26,27]. With increased
usage of this equipment, it soon became evident that not
only time, but also substantial amounts of solvent can be
saved compared with conventional HPLC equipment.
Not only can the instrument complete analysis work-
flows more than eight times faster than conventional
chromatography, but it also does so with at least a
99.5% reduction in solvent consumption and waste
generation. The resulting advantage can be quite sig-
Figure 5. Separation of an alkyl benzene mixture on a) conventional HPLC using 4.6 mm i.d. column at 1.5 mL/min requiring 12 mL of solvent
per analysis; and, b) microscale HPLC using 0.3 mm i.d. column at 6 lL/min requiring only 39 lL of solvent per analysis. Conditions: Agilent SB-
18, 150 mm, 3.5-lm particles, UV detection @ 220 nm, Gradient elution 50–95% acetonitrile in water over 5 min, then 95% in 1 min.
672 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
nificant for laboratories carrying out the analysis of
hundreds or thousands of microplates each year. Fig. 5
shows a representative chromatogram illustrating how
microflow chromatography can be used to afford fast
separations with decreased solvent usage. In this
example, an analyte mixture is purified on a 300-lm i.d.
column using only 39 lL of solvent, as opposed to 12 mL
of solvent required per analysis using a conventional
column. The somewhat faster elution observed with the
microcolumn stems from the much smaller (about seven
times) proportional gradient dwell volume of the
microflow HPLC system, a property that enables fast
gradient separations [24].
We have investigated microflow chromatography for
other analytical applications outside the high-through-
put analysis area, and have generally been pleased with
the technique. We have collaborated with Eksigent
Technologies on production of a small, mobile, single-
channel microflow HPLC instrument that can perform
on-line analysis of organic reactions [28]. The instru-
ment can withdraw reaction aliquots at specified times,
perform a precise dilution, and carry out HPLC analysis
with graphical representation of the results. We have
recently applied this tool to on-line analysis of flowing
streams coming from flow-chemistry reactions, with
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
Microcolumn (300 µ i.d)
120
Abs. UV320 nm 100
80 Flow rate = 6 µL/min
60
Inject 43 µg
40
20
0
0 2 4 6
2000
Preparative Column (30 cm i.d)
Abs. UV290 nm
Flow rate = 6 L/min
1000 Inject 43 g
0
0 2 4 6
Figure 6. Preparative chromatography-modeling experiments carried out on microcolumns accurately predict performance at million-fold scale,
thereby significantly reducing the amount of material and solvent required for a modeling study.
good results [29]. This small, mobile instrument operates
in an automated, unattended fashion, and generates less
than 10 mL of waste during 24 h of analysis.
An additional green-chemistry advantage of micro-
flow HPLC is that only a small fraction of the
chromatographic stationary phase needed for a con-
ventional column is required to pack a microcolumn,
as pointed out by Novotny and others more than 25
years ago [22]. Microflow chromatography can there-
fore obtain the same results using only a fraction of
valuable stationary-phase material. While this can
clearly lead to dramatic cost savings for analyses using
expensive chromatographic stationary phases (e.g.,
chiral stationary phases, or affinity or other biomole-
cule-containing stationary phases), the impact can be
even more dramatic, as commercialization of stationary
phases in the microcolumn format is possible for exotic
materials that would simply be cost prohibitive in the
conventional column format [30]. Similarly, as illus-
trated in Fig. 6, preparative chromatographic loading
studies carried out using microcolumns can afford re-
sults with <1% of the material typically required for
such studies, meaning that the necessary information
for purification scale-up can be obtained faster and
without the need for re-synthesis that is typically re-
quired today [31].
Our experience with microflow HPLC has thus far
been restricted to the somewhat specialized areas of
high-throughput analysis, high-throughput method
development, and on-line analysis. We have, of course,
Trends
8 10 12 14 16 18 20
1,000,000 x
scale up
8 10 12 14 16 18 20
considered the broader implementation of the technol-
ogy to replace the more routine HPLC analysis that
forms the bulk of the HPLC utilization within the phar-
maceutical industry. While solvent savings are undeni-
able, there are also some negatives, including the basic
unfamiliarity of traditional HPLC users with microflow
HPLC equipment and control software. In addition,
microflow HPLC currently has a slightly lower limit of
detection arising from UV-detector limitations. This is
typically not a problem for workflows where plenty of
sample is available, but is more problematic for the low-
level quantitation required in metabolite and trace-
impurity studies.
Nevertheless, the adoption of microflow HPLC is
certain to increase in coming years. As noted previ-
ously, recently introduced HPLC instruments are well
suited to the use of not only conventional columns,
but also smaller columns in the 2-mm i.d. range. This
new capability is largely due to minimization of
plumbing connections that affords the low extra-col-
umn volumes required for good chromatographic per-
formance with microcolumns. In addition, current
issues with UV-detector sensitivity are likely to go
away as further technical advances are made. Further,
as mass spectrometry (MS) continues to grow as a
preferred approach to chromatographic detection,
microcolumn HPLC becomes even more attractive, as
evidenced by the growing widespread use of nano-
HPLC approaches in the field of proteomics analysis
[32,33].
http://www.elsevier.com/locate/trac 673
Trends
injector
solvent A solvent B
Differing scenarios for greening analytical chromatography
(a)
Non-green solvents and waste
(most normal phase chromatography)
(c)
Green solvents, green analytes and waste
(ethanol-based HPLC or SFC, hot water HPLC)
Figure 7. Different scenarios for greening analytical separation science.
6. Solvent replacement
6.1. General considerations
In addition to solvent-reduction strategies, other green
analytical chemistry efforts focus on replacing toxic or
flammable solvents with greener alternatives. A variety
of options exist, as summarized in Fig. 7. The worst
possible case, from a green analytical chemistry per-
spective, is illustrated in Fig. 7a, where two non-green
solvents are used to carry out the chromatographic
separation, with the waste generated in the experiment
also being non-green. Fig. 7b shows a preferred option,
as one of the solvents used in the chromatographic
experiment has been replaced with a greener alternative
(e.g., water, ethanol, or carbon dioxide). Fig. 7c shows
the ideal, as all solvents used are green, as is the waste
generated in the experiment. In principle, such a tech-
nology would not generate any waste at all, as the
effluent could be directly sewered. However, we must
point out that, when dealing with analytes that are toxic
or radioactive, the best-case scenario would be Fig. 7d,
where green eluents are used, but the chromatographic
eluent becomes contaminated as a result of exposure to
the analytes, and thus must be disposed of as laboratory
waste.
6.2. Replacement of acetonitrile with ethanol in
RP-HPLC
The most common form of analytical separation tech-
nology carried out in laboratories is reversed-phase (RP)
HPLC using a hydrophobic stationary phase and a
mobile phase comprising acetonitrile with water con-
674 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
waste
solvent
column
(b)
Green and non-green solvents, non-green waste
(SFC, water-based RP HPLC)
(d)
Green solvents, non green analytes and waste
(best case for analysis of toxins, radiochemicals)
taining additives to adjust pH and ionic strength (Fig. 7a
or 7b, depending on the nature of the water additives).
The use of acetonitrile as an organic modifier in RP-
HPLC has long been preferred by HPLC users, based on
the remarkable combination of physical properties that
this solvent possesses, including low viscosity, complete
miscibility with water, and very low UV absorbance at
wavelengths of interest to the HPLC analyst, especially at
lower wavelengths in the 200-nm range, where other
solvents often have strong background absorbance. De-
spite these remarkable properties, acetonitrile possesses
other properties that make it less preferred from a green
analytical chemistry standpoint, namely, acetonitrile is
flammable, volatile and toxic. Consequently, HPLC waste
streams containing acetonitrile must be disposed as
chemical waste, which is costly and adds to the envi-
ronmental waste-disposal burden of the laboratory. In
addition, acetonitrile is an expensive solvent, as brought
to the fore during a worldwide shortage in 2009, where
limited availability caused prices to skyrocket [34].
In a recent study, we investigated the potential for
replacing acetonitrile in RP-HPLC with the greener sol-
vent, ethanol [35]. While it is well known that ethanol is
inferior to acetonitrile as an RP-HPLC eluent, we were
interested, as other investigators had been previously
[36–38], in investigating whether, in combination with
state-of-the-art high-performance column technologies,
ethanol could deliver satisfactory results. In other words,
we were interested in exploring how bad ethanol would
be as an RP-HPLC eluent, and whether satisfactory
performance might be possible. The question is especially
important when one considers that chromatography
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010 Trends

carried out with the two green solvents, ethanol and
water, could afford the best-case scenario, as depicted in
Fig. 7c and 7d.
The study clearly showed that ethanol performs rea-
sonably well as an RP-HPLC solvent, and may be suitable
for replacing acetonitrile in some instances. Fig. 8 depicts
a representative example, where reversed-phase HPLC
was carried out under limiting pressure conditions for the
separation of a test mixture of alkylbenzene. While the
results with acetonitrile (Fig. 8a) were clearly superior,
the results obtained with ethanol were surprisingly good
(Fig. 8b), particularly when modern column and instru-
ment technologies and elevated temperatures were used
(Fig. 8c).
Use of the greener solvent ethanol may therefore be an
option for some analytical laboratories, particularly
those where the poorer elutropic strength and UV cut-off
would not be a significant problem (e.g., within labora-
tories that specialize in the analysis of compounds with
strong UV chromophores, such as aromatic hydrocar-
bons, pigments, dyes and most drugs). In addition, as MS
continues to supplant UV as the preferred detection
technique for HPLC, the higher UV cut-off of the alcohol
solvents may become less important over time.
One of the most exciting opportunities relating to the
use of ethanol as a modifier for RP-HPLC is the possibility
for the ideal green analytical chemistry scenario depicted
in Fig. 7c, where all incoming solvents and outgoing
waste are environmentally compatible, opening up the
possibility of using HPLC instrumentation outside the
normal boundaries of the chemical laboratory. The
widespread availability of ethanol in all areas of the
globe suggests that this solvent could be ideal for
instruments deployed in environmentally-sensitive areas
(e.g., laboratories within ecological preserves, or schools
and universities in developing countries), where chemi-

(a) 1000
ACN
UV 220 nm (mAU)

800
5 mL/min
600

400

200

0
0 0.2 0.4 0.6 0.8 1.0
time (min)

(b) 1000
800 EtOH (25 °C)
UV 220 nm (mAU)

1.75 mL/min
600

400

200

0
0 0.2 0.4 0.6 0.8 1 1.2
time (min)

(c) 1000
800 EtOH (45 °C)
2.5 mL/min
UV 220 nm (mAU)

600
400

200

0
0 0.2 0.4 0.6 0.8 1
time (min)
Figure 8. Comparison of ethanol and acetonitrile in the reversed-phase HPLC separation of an alkylbenzene test mixture demonstrates that,
while inferior to acetonitrile as an HPLC eluent, ethanol can sometimes obtain surprisingly good results, particularly when modern column
and instrument technologies and elevated temperatures are used. Conditions: column = Ascentis Express C18; flow rates as noted in chromato-
grams; temperature = 25C, except as noted; aqueous 0.1% H3PO4; UV detection at 220 nm; gradient = 80–100% organic in 0.5 min.

http://www.elsevier.com/locate/trac 675
Trends Trends in Analytical Chemistry, Vol. 29, No. 7, 2010

Vodka baseline Ethanol baseline

DAD1 A, Sig=220,16 Ref=450,100 (2009-10-21\NS_LINDNER_0040.D) DAD1 A, Sig=220,16 Ref=450,100 (2009-10-21\NS_LINDNER_0039.D)

300 1.8

1.6
-1

1.4

250 1.2 -1.2

-1.4
0.8

200 0.6
-1.6
0.4

0.2

150 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 3.4 3.6 3.8 4 4.2 4.4

100

50
Vodka
0

-50 Ethanol
-100
0 2 4 6 8 10 12 14
Time (min)
Figure 9. HPLC separation of alkylbenzene test mixture using inexpensive vodka as a mobile phase. Conditions: column = Ascentis C18
4.6 mm · 50 mm, 2.7 micron; mobile phase = either 40% EtOH/water or vodka; flow rate = 3 mL/min; column temperature = 60C; UV
detection at 220 nm.

cal-waste disposal is not yet established or is prohibi- High-temperature water (>100C) possesses reactive
tively expensive. Fig. 9 illustrates this point, with an properties, so it is incompatible with most conventional
example of HPLC separation using inexpensive vodka as silica-based stationary phases (e.g., ODS silica). More
the eluent. One could imagine that, in coming years, temperature-resistant packing materials are usually re-
such benign analysis technology could be deployed in quired for SHWLC, including polymeric phases [e.g.,
the offices of doctors, dentists, farmers, and many other polystyrene-divinylbenzene (PS-DVB)] or zirconia-based
industries that do not currently utilize HPLC-analysis materials [e.g., zirconia particles with polybutadiene
tools. In combination with current trends in miniaturi- (PBD), polystyrene or a carbon coating)] [48]. SHW was
zation and decreased cost, HPLC using benign solvents investigated as a mobile phase for LC using not only UV
could even find applications in the home. and FID, but also various other detectors [e.g., refractive
index (RI), nuclear magnetic resonance (NMR), MS, and
6.3. Solvent replacement: superheated water infrared (IR)] [49–54]. The applications of SHWLC in-
chromatography clude natural products, pharmaceuticals, and environ-
In recent years, there has been a considerable interest in mental samples [48]. The limitation of SHWLC includes
the use of neat water, especially superheated water potential on-column degradation of thermally-labile
(SHW) or pressurized water, as a mobile phase in liquid compounds and insolubility of hydrophobic compounds
chromatography (LC) [39–45]. SHWLC can be consid- in neat water.
ered a green alternative for conventional RP-LC sepa-
rations where water-methanol and water-acetonitrile 6.4. Solvent replacement: carbon dioxide-based
are generally used as eluents, since water is readily chromatography
available, inexpensive, non-flammable and certainly Supercritical fluid chromatography (SFC) utilizes pres-
environmentally acceptable. In addition, water has no surized carbon dioxide in the subcritical or supercritical
significant flame-ionization-detector (FID) response, so it state as a chromatographic eluent [55–59]. Somewhat
is potentially promising as a mobile phase for LC when a surprisingly, pressurized carbon dioxide has solvent
‘‘universal’’ FID is desirable for UV-transparent com- properties similar to petrochemical-derived hydrocar-
pounds. However, water often cannot elute many or- bons, so it affords a greener replacement for commonly-
ganic analytes at room temperature in LC [46,47], so used normal-phase chromatography solvents (e.g., hex-
high temperature is usually needed to increase elution ane or heptanes). Carbon dioxide is typically used with a
strength instead of organic composition in the mobile polar modifier solvent, usually methanol, although the
phase. Recently, Hartonen and Riekkola provided a greener alternative, ethanol, is often used. In addition,
comprehensive review on SHWLC in terms of column very small amounts of acid or base additives are often
technologies, detection, and applications [48]. added to improve the peak shape of acidic or basic

676 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
analytes. Pressure must be maintained on an operating
SFC system through the use of a back-pressure regulator
in order to keep the carbon dioxide in the liquid or
supercritical form; however, upon depressurization,
carbon dioxide again returns to gas form, allowing
convenient recovery of all non-volatile components. It
should be noted that this addition of carbon dioxide to
the atmosphere is not a net addition to the environment,
as industrial carbon dioxide is typically derived as a
byproduct from refining natural gas or alcohol fermen-
tation.
The critical parameters for pure carbon dioxide are
31.1C temperature and 73.8 bar pressure, meaning
that, at temperatures and pressures above these values,
carbon dioxide exists as a supercritical fluid. Carbon
dioxide is environmentally friendly, and, in the liquid or
supercritical fluid state, has a low viscosity that allows
high flow rates and faster separations, and a high dif-
fusivity that allows efficient separations. The optimal
linear velocity for SFC is approximately three times
greater than for HPLC, and SFC also generally loses less
efficiency at high velocities than HPLC [60]. For a gra-
dient separation, the faster diffusion of SFC means faster
re-equilibration and thus a shorter cycle time [61]. Be-
cause of these speed advantages in analysis, savings in
organic solvent consumption can be 88–98% compared
to normal-phase HPLC, even when 5–35% organic sol-
vent additives are used. Due to these advantages in SFC,
very fast separations are possible and very impressive
separations of the order of a few seconds have been re-
ported (Fig. 10) [62,63].
Analytical SFC has made significant inroads in dis-
placing normal-phase HPLC as the preferred technique
for chromatographic separation of enantiomers on chiral
stationary phases. The combined advantages of speed,
convenience and environmental friendliness have made
SFC the preferred instrumental approach for stereo-
mAU
300 Fast SFC
Chiralcel OF
(50 x 4.6 mm)
200
100
0.72 mL waste
per analysis
0
0 0.2
Figure 10. Fast chiral supercritical fluid chromatography (SFC) separation of alcohol enantiomers. SFC conditions: Chiralcel OF column
(50 · 4.6 mm); isocratic 30% methanol/70% CO2, 4 mL/min; 35C; 200 bar; UV detection at 215 nm.
Trends
chemical analysis in many laboratories [64,65], and a
variety of approaches have been described for fast chiral
method development and analysis [66–69]. Despite the
continued success in the past decade, analytical SFC is
far from the robust, reliable technology that HPLC has
evolved into over four decades. The problem is com-
pounded by instrument companies focusing primarily on
the preparative SFC market segment, rather than ana-
lytical SFC, as the considerable advantages in purifying
compounds have driven recent instrument sales.
In addition to its use in chiral analysis, SFC has proven
useful for achiral separations. Recent interest in devel-
oping a universal purification platform to replace mass-
directed preparative RP-HPLC as the preferred tool for
high-throughput purification has focused on the promise
of SFC [70]. In addition to green-chemistry advantages,
this emerging focus on high-throughput mass-directed
SFC has been driven by the considerable speed advan-
tage for sample recovery that comes from avoiding the
need to evaporate aqueous fractions. The results to date
are promising, but, in order for this approach to match
the universality of RP-HPLC, the development of more
general stationary phases and chromatographic condi-
tions will be needed – an area of considerable interest in
current research [71].
SFC represents one of the true success stories of green
chemistry, where the replacing technology is both
greener and better. The improved performance of
the technique has driven a widespread implementa-
tion in chiral chromatography, for both analysis and
smaller-scale preparative separation, where considerable
advantages in speed and performance can be obtained.
In a recent article, SFC was criticized for not being
truly green, or at least not as green as generally depicted
[72]. While the authors of this study raised some
important points about the need to consider total energy
costs in the production and the distribution of carbon
OH
Ar
0.6 min
0.4 0.6 0.8 min
http://www.elsevier.com/locate/trac 677
Trends
dioxide, they focused their analysis on a single pre-
parative chromatography example where productivity is
much higher with normal-phase solvents than with SFC.
While occasional examples of this type can be found, SFC
typically has better chromatographic productivity than
HPLC [73]. The authors also based much of their anal-
ysis on the energy costs of recycling organic solvents by
distillation vs. recycling of carbon dioxide by chilling.
While these considerations are indeed important for
operation at manufacturing scale, almost all discovery
and development-scale preparative chromatography is
carried out without recycling solvent.
As the popularity of SFC has grown in recent years,
there has been an increasing interest in bringing ana-
lytical SFC instruments to market. First commercialized
around 20 years ago, analytical SFC instrumentation is
largely based upon older HPLC technology, and has been
slow to keep pace with instrument improvements that
have taken place in the HPLC field. For example, one of
the greatest promises of SFC is the ability to carry out fast
chiral separations, with many compounds being resolv-
able in 1 min or less. Consequently, the technique should
be amenable to high-throughput analysis of chirality,
but many instrument limitations have made it difficult to
realize this potential. For example, analytical SFC
instruments were, until recently, unable to inject from
microplates, and current capabilities in this area are far
from what is standard in the HPLC realm. Analytical SFC
instruments are generally unable to carry out multiple
injections within a single experimental run. A flow-
injection analysis technique commonly used in HPLC is
very important for fast analysis. In addition, existing SFC
instruments have historically had extremely long auto-
sampler delays, ranging up to several minutes between
samples. Clearly, a 30-s SFC method with a 5-min
autosampler delay would be sub-optimal for high-
throughput chiral analysis.
Poor detection sensitivity in SFC has historically pre-
vented the technique from being embraced by analysts
carrying out regulatory testing, where low-level impu-
rity quantitation is important. We showed in a recent
study that this limitation can easily be overcome by
simple instrument modifications and plumbing changes,
allowing SFC to achieve the level of sensitivity required
for analytical release testing [74].
Finally, control software for analytical SFC instru-
mentations has been problematic for a number of years,
with none of the available choices being comparable in
ease of use with what has become standard in the HPLC
realm.
On the positive side, all of these limitations of analyt-
ical SFC technology can be overcome using solutions
that have already proved to work in the HPLC arena. In
addition, other current trends in HPLC-instrument
design (e.g., minimization of extra-column volume,
reduction of column diameter, use of higher back-pres-
678 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
sures and use of smaller stationary-phase particles) could
also be applied with benefit to SFC-instrument design.
Consequently, there are new product entries that address
some of the historical limitations of analytical SFC
instrumentation.
We recently carried out a comparative evaluation of
several new commercial analytical SFC instruments. In
typical Good Manufacturing Practice (GMP) tests [e.g.,
detection linearity, injection precision, retention repro-
ducibility, low level of detection (LOD) and low level of
quantitation (LOQ)], we found some products to be
substantially improved relative to the previous genera-
tion of SFC instruments, while others were worse.
However, we found that most new models had signifi-
cant detection-sensitivity improvement compared with
previous instruments, with signal-to-noise (S/N) ratios of
3–6 achieved with 0.02–0.05% of target 1 mg/mL-
sample concentration, nearly 10 times better than pre-
vious analytical SFC instruments. This significant
improvement in LOD allows use of new-generation
analytical SFC instruments for GMP release testing.
In an evaluation of features [e.g., back-pressure reg-
ulator (BPR) stability, gradient-delay volume, system
robustness and mobile-phase-pumping stability at both
low and high modifier percentages], we found several of
the new product offerings to be superior to older models.
The Jasco analytical SFC instrument, for example, dem-
onstrated excellent pumping stability with the ability to
run as low as 2% modifier (with typical 40C tempera-
ture and 200 bar outlet pressure), representing a sig-
nificant improvement over previous instruments with a
low-level modifier cut-off of 4% or 5%. When considering
software and instrument ease of use, as well as the
performance criteria described above, the Aurora SFC
Fusion A5 system showed a distinct advantage. This
system comprises a carbon-dioxide handling module that
is ‘‘piggy-backed’’ onto an existing HPLC instrument,
allowing SFC operations using robust and familiar HPLC
software, which is a considerable advantage, especially
for newer SFC users. While none of the new offerings
possesses all of the attributes that would be desired in an
ideal analytical SFC instrument, recent progress is
encouraging, suggesting that analytical SFC instru-
mentation may soon show the degree of robustness and
performance required for broader implementation of this
important green analytical chemistry technology.
7. Implementation of green analytical chemistry
technologies
We have shown in the preceding sections that green
analytical chemistry technologies can afford greener
analytical separations. But, under what conditions, do
adoption and implementation become compelling? Just
because a new technology has been shown to consume
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
less solvent and generate less waste than traditional
HPLC, a switch to the greener alternative may or may
not be justified. For example, a new HPLC instrument
that consumes 90% less solvent and generates 90% less
waste over a projected 10-year lifetime, but costs 10
times more than a conventional HPLC instrument,
would probably not be cost effective from the perspective
of ‘‘total cost of ownership’’. Similarly, a change to a
like-priced, but greener alternative instrument that lacks
robustness, dependability, and user friendliness may not
be warranted, as the costs for repairs and down-time
must be factored into the total cost of ownership.
In general, while it may be reasonable to pay a small
premium for a greener technology, the best green-
chemistry solutions are often characterized by being
both cheaper and greener. Finally, even when the
replacement technology is clearly superior from both a
cost and performance perspective, there can still be
additional impediments to change. For example, while
instruments that have reached the end of their useful life
are obvious candidates for greener replacements, should
we replace well-functioning instruments projected to
remain useful for 5 or even 10 years? Clearly, replacing
such equipment is a luxury that few can afford, and
throwing away such equipment is certainly not envi-
ronmentally responsible. In addition, there can some-
times be additional costs involved in changing over to a
new, greener technology alternative, especially where
substantial changes in operating software and user
accessibility are issues, or in a heavily regulated envi-
ronment, where revalidation of a separation method
may be required.
As a result of all these considerations, we can expect
to see the uptake of green analytical chemistry
technologies lag behind what has already been dem-
onstrated to work. However, the future for this field is
clear. We can project that, in the next few years, there
will be continuing decline in the amount of solvent
usage and waste generation associated with collecting a
single analytical measurement, and we can look for-
ward to a time in the not-too-distant future when the
idea of generating 50 mL of chemical waste for every
analytical data point will be viewed as a careless
anachronism from a bygone era.
8. Conclusion
In this review, we have shown that the amount of waste
generated per data point for analytical separation tech-
nologies can be decreased dramatically by a number of
green analytical chemistry technology alternatives. As
these newer, greener analytical separation technologies
become more widely utilized, they are having a dra-
matic, beneficial impact in the workplace, allowing more
work to be performed with fewer instruments and
Trends
greatly reducing the amount of waste solvent associated
with the collection of highly-valued analytical HPLC
data.
Acknowledgement
We are grateful to the New Technologies Review &
Licensing Committee (NT-RLC) for providing funding for
the acquisition and evaluation of many of the technol-
ogies described in this review.
References
[1] P.T. Anastas, J.C. Warner, Green Chemistry: Theory and Practice,
Oxford University Press, Oxford, UK, 2000.
[2] M. Lancaster, Green Chemistry, Royal Society of Chemistry,
Cambridge, UK, 2002.
[3] C. Alfonso, J. Crespo (Editors), Green Separation Processes:
Fundamentals and Applications, Wiley, New York, USA, 2005.
[4] C. Welch, Drug Discov. World (Fall 2007) 71.
[5] S. Armenta, S. Garrigues, M. de la Guardia, Trends Anal. Chem.
27 (2008) 497.
[6] M. Tobiszewski, A. Mechlinska, B. Zygmunt, J. Namiesnik, Trends
Anal. Chem. 28 (2008) 943.
[7] R. Majors Trends in HPLC Column Usage, LCGC N. Am. (1
November 2009).
[8] R.M. Katusz, L. Bellew, J.A. Mangravite, R.F. Foery, J. Chromatogr.
213 (1981) 331.
[9] A. Welch, Am. Lab. 38 (2006) 44.
[10] J.C. Giddings, Unified Separation Science, John Wiley, New York,
USA, 1991.
[11] N. Wu, A.M. Clausen, J. Sep. Sci. 30 (2007) 1167.
[12] H. Chen, C. Horváth, J. Chromatogr., A 705 (1995) 3.
[13] J.R. Mazzeo, U.D. Neue, M. Kele, R.S. Plumb, Anal. Chem. 77
(2005) 460A.
[14] F. Gritti, M. Martin, G. Guiochon, Anal. Chem. 81 (2009) 3365.
[15] S. Chen, A. Kord, J. Chromatogr., A 1216 (2009) 6204.
[16] J.M. Cunliffe, T.D. Maloney, J. Sep. Sci. 30 (2007) 3104.
[17] N. Tanaka, H. Kobayashi, N. Ishizuka, H. Minakuchi, K. Nakani-
shi, K. Hosoya, T. Ikegami, J. Chromatogr., A 965 (2002) 35.
[18] K.K. Unger, R. Skudas, M.M. Schulte, J. Chromatogr., A 1184
(2008) 393.
[19] T. Ikegami, N. Tanaka, Curr. Opin. Chem. Biol. 8 (2004) 527.
[20] F. Svec, Chem. Listy 98 (2004) 232.
[21] L.S. Ettre, J. Waters, LCGC 23 (2005) 752.
[22] M. Novotny, Clin. Chem. 26 (1980) 1474.
[23] C.J. Welch, Chirality 21 (2009) 114.
[24] C.J. Welch, P. Sajonz, M. Biba, J. Gouker, J. Fairchild, J. Liq.
Chromatogr. 29 (2006) 2185.
[25] P. Sajonz, W. Schafer, X. Gong, S. Shultz, T. Rosner, C.J. Welch, J.
Chromatogr. 1147 (2007) 149.
[26] P. Peter, X. Gong, W.R. Leonard, M. Biba, C.J. Welch, Chirality 18
(2006) 803.
[27] P. Sajonz, W. Schafer, W.R. Leonard, X. Gong, S. Shultz, T.
Rosner, Y. Sun, C.J. Welch, J. Liq. Chromatogr. 31 (2008)
2296.
[28] W.A. Schafer, S. Hobbs, J. Rehm, D.A. Rakestraw, C. Orella,
M. McLaughlin, Z. Ge, C.J. Welch, Org. Proc. R&D 11 (2007)
870.
[29] C.J. Welch, X. Gong, J. Cuff, S. Dolman, J. Nyrop, F. Lin, H. Rogers,
Org. Proc. R&D 13 (2009) 1022.
[30] C.J. Welch, M.H. Hyun, T. Kubota, W.A. Schafer, F. Bernardoni,
H.J. Choi, X. Gong, B. Lipshutz, Chirality 20 (2008) 815.
[31] C.J. Welch, P. Sajonz, G.A. Spencer, W.R. Leonard, W.A. Schafer,
F. Bernardoni, Org. Proc. R&D 12 (2008) 674.
http://www.elsevier.com/locate/trac 679
Trends
[32] J. Hernandez-Borges, Z. Aturki, A. Rocco, S. Fanali, J. Sep. Sci. 30
(2007) 1589.
[33] T. Wehr, LCGC 24 (2006) 286.
[34] R.E. Majors, LCGC 27 (2009) 458.
[35] C.J. Welch, T. Brkovic, W. Schafer, X. Gong, Green Chem. 11
(2009) 1232.
[36] R.L.V. Ribeiro, C.B. Bottoli, K.E. Collins, C.H. Collins, J. Brazil.
Chem. Soc. 15 (2004) 300.
[37] C. Capello, U. Fischer, K. Hungerbühler, Green Chem. 9 (2007)
927.
[38] A. Balaguer, A. Chisvert, A. Salvador, J. Sep. Sci. 31 (2008) 229.
[39] T. Greibrokk, T. Andersen, J. Chromatogr., A 1000 (2003) 743.
[40] D. Guillarme, S. Heinisch, J.L. Rocca, J. Chromatogr., A 1052
(2004) 39.
[41] D. Guillarme, S. Heinisch, Sep. Purif. Rev. 34 (2005) 181.
[42] C.A. Bruckner, S.T. Ecker, R.E. Synovec, Anal. Chem. 69 (1997)
3465.
[43] Y. Yang, A.D. Jones, C.D. Eaton, Anal. Chem. 71 (1999) 3808.
[44] M.D. Foster, R.E. Synovec, Anal. Chem. 68 (1996) 2838.
[45] R.M. Smith, Anal. Bioanal. Chem. 385 (2006) 419.
[46] R.M. Smith, R.J. Burgess, O. Chienthavorn, J.R. Bone, LCGC 17
(1999) 939.
[47] N. Wu, Q. Tang, J.A. Lippert, M.L. Lee, J. Microcol. Sep. 13 (2001)
41.
[48] K. Hartonen, M. Riekkola, Trends Anal. Chem. 27 (2008) 1.
[49] C.L. Guillemin, J.L. Millet, J. Dubois, HRC & CC, J. High Resolut.
Chromatogr. Chromatogr. Commun. 4 (1981) 280.
[50] R.M. Smith, O. Chienthavorn, I.D. Wilson, B. Wright, Anal.
Commun. 35 (1998) 261.
[51] D. Louden, A. Handley, R. Lafont, S. Taylor, I. Sinclair, E. Lenz, T.
Orton, I.D. Wilson, Anal. Chem. 74 (2002) 288.
[52] S. Saha, R.M. Smith, E. Lenz, I.D. Wilson, J. Chromatogr., A 991
(2003) 143.
[53] D. Louden, A. Handley, S. Taylor, I. Sinclair, E. Lenz, I.D. Wilson,
Analyst (Cambridge, UK) 126 (2001) 1625.
[54] R.M. Smith, O. Chienthavorn, I.D. Wilson, B. Wright, S.D. Taylor,
Anal. Chem. 71 (1999) 4493.
[55] T.L. Chester, J.D. Pinkston, Anal. Chem. 74 (2002) 2801.
680 http://www.elsevier.com/locate/trac
Trends in Analytical Chemistry, Vol. 29, No. 7, 2010
[56] P. Raveendran, Y. Ikushima, S.L. Wallen, Acc. Chem. Res. 38
(2005) 478.
[57] E. Lesellier, J. Chromatogr., A 1216 (2009) 1881.
[58] P. Mourier, P. Sassiat, M. Caude, R. Rosset, J. Chromatogr. 353
(1986) 61.
[59] L.T. Taylor, J. Supercrit. Fluids 47 (2009) 566.
[60] N. Wu, Q. Tang, Y. Shen, M.L. Lee, Anal. Chem. 71 (1999) 5084.
[61] D.R. Gere, Science (Washington, DC) 222 (1983) 253.
[62] C.J. Welch, M. Biba, P. Sajonz, Chirality 19 (2007) 34.
[63] B. Bolanos, M. Greig, M. Ventura, W. Farrell, C.M. Aurigemma, H.
Li, T.L. Quenzer, K. Tivel, J.M.R. Bylund, P. Tran, C. Pham, D.
Phillipson, Int. J. Mass Spectrom. 238 (2004) 85.
[64] C.J. Welch, W.R. Leonard Jr., J.O. Da Silva, M. Biba, J. Albaneze-
Walker, D.W. Henderson, B. Laing, D.J. Mathre, LCGC 23 (2004)
16.
[65] J.O. Da Silva, H. Yip, V. Hegde, Am. Pharm. Rev. 12 (2009) 98.
[66] M.S. Villeneuve, R.J. Anderegg, J. Chromatogr., A 826 (1998)
217.
[67] Y. Zhao, G. Woo, S. Thomas, D. Semin, P. Sandra, J. Chromatogr.,
A 1003 (2003) 157.
[68] Y. Zhang, W. Watts, L. Nogle, O. McConnell, J. Chromatogr., A
1049 (2004) 75.
[69] M. Biba, C.J. Welch, J.O. Da Silva, D.J. Mathre, Am. Pharm. Rev. 8
(2005) 68.
[70] R.T. McClain, A. Dudkina, J. Barrow, G. Hartman, C.J. Welch, J.
Liq. Chromatogr. 32 (2009) 483.
[71] C. West, E. Lesellier, J. Chromatogr., A 1191 (2008) 21.
[72] G. Van der Vorst, H. Van Langenhove, F. DePape, W. Aelterman,
J. Dingenen, J. Dewulf, Green Chem. 11 (2009) 1007.
[73] C.J. Welch, in: G. Cox (Editor), Preparative Enantioselective
Chromatography, Blackwell, London, UK, 2005, pp. 1–18.
[74] R. Helmy, M. Biba, J. Zang, B. Mao, K. Fogelman, V. Vlachos, P.
Hosek, C.J. Welch, Chirality 19 (2007) 787.