Shimadzu - Introduction To HPLC
Shimadzu - Introduction To HPLC
Shimadzu - Introduction To HPLC
HPLC
Shi madzu
LC Wor l d Tal k Speci al I ssue Vol ume1
C190-E093
2
3
Contents Page
1. Reversed-phase Chromatography
4
2. Reversed-phase Ion-pair Chromatography
5
3. Derivatization in HPLC
6
4. Reversed-phase Chromatography and Hydrophobic Interaction Chromatography
7
5. Silica Gel Based Packing
8
6. Absorptiometric Detection
9
7. Affinity Chromatography
10
8. Normal-phase Chromatography
11
9. Ligand-exchange Chromatography
12
10. Explaining GLP/GMP Terminology: RSD (C.V.)
13
11. Explaining GLP/GMP Terminology: Tailing Factor, Resolution
14
12. Detection Limit
16
13. LC-MS Discussion (part 1)
18
14. LC-MS Discussion (part 2)
19
15. LC-MS Discussion (part 3)
20
16. Separation of Sugars
21
17. Detection of Sugars
22
18. Detection of Sugars - Continued
23
19. Evaporative Light Scattering Detector
24
20. Methods of Amino Acid Analysis
25
21. Analysis of Organic Acids
26
22. Size Exclusion Chromatography
27
This is an compilation of introductory section of past issues of the "LCtalk",
Shimadzu's newsletter for HPLC users, Japanese version.
4
Reversed-phase chromatography is the most commonly used HPLC
separation mode. It is far superior to the other modes in the variety of
target compounds it can handle. The dominant phenomenon retaining
the sample in the column in reversed-phase chromatography is the
hydrophobic interaction between the solid phase and sample. Two
types of reversed-phase chromatography column packing are used: one
type is a silica gel matrix with chemically bonded alkyl chains and the
other is resin-based packing. Except in special circumstances, the silica
gel matrix type is used due to its high number of theoretical plates. A
resin-based packing must be used if the pH of the mobile phase used
for separation is set outside the range that can be used with silica gel
or if unreacted silanol groups remaining on the silica gel surface have
a detrimental effect on separation and this problem cannot be resolved
by changing the composition of the mobile phase. However, these are
comparatively rare cases. Typical alkyl groups that are chemically
bonded to the silica gel include the octadecyl group, the octyl group, and
the trimethyl group (see Fig. 1). The longer the alkyl chain, the greater
the retaining force. The column is selected according to the hydrophobic
retaining force on the target compounds. Use a column with weaker
retaining force if the target compounds do not elute within an appropriate
time under mobile phase conditions
providing the strongest elution force
(100% organic solvent). Conversely,
use a column with stronger retaining
force if target compounds are not
retained for an appropriate time
under mobi l e phase condi t i ons
providing the weakest elution force
(100% buffer).
The mobile phase composition is
determined after the column has
been selected. It is basically a mixed
solvent comprising water or buffer
mixed with an organic solvent. The
three major factors that influence
separation are (1) type of organic
solvent, (2) proportion of organic solvent, and (3) pH of the buffer
solution. Acetonitrile and methanol are the most commonly used organic
solvents (point (1)). Acetonitrile is the most desirable organic solvent for
HPLC analysis. Acetonitrile offers the dual benefits of low noise effects
due to its low UV absorbance and extended column life, as it permits
analysis at low pressure. However, in exceptional cases, methanol may
be chosen due to its high selectivity. Methanol is selected if peaks A and
B can be separated using methanol but not using acetonitrile, or if peaks
A and B separate easily in both solvents but the a value (k' value ratio) of
A and B is larger for acetonitrile, such that analysis can be completed in
a shorter time using methanol.
The proportion of organic solvent (point (2)) is selected as the proportion
that achieves the required separation and most rapidly achieves elution.
(Increasing the proportion of organic solvent achieves more rapid
elution. Decreasing the proportion results in slower elution but better
separation.) When only neutral substances are analyzed, the retaining
time is not affected by the mobile phase pH, so that analysis is normally
conducted using water and organic solvents. In this case, only conditions
(1) and (2) need to be set. However, condition (3) must also be set when
analyzing acidic or basic substances.
For example, benzoic acid exists in the solution in the state of
equilibrium shown in Fig. 2. Dropping the pH of the solution pushes the
equilibrium toward the right, increasing the proportion of the uncharged
state. Conversely, increasing the solution pH increases the dissociated
state (the left side). The uncharged state is better retained during
reversed-phase chromatography: lowering the pH increases the retaining
force; increasing the pH achieves faster elution. Conversely, for basic
substances such as amines, increasing the pH increases the retaining
force. The method of setting the pH based on awareness of this pH
dependency to increase the retaining force while suppressing dissociation
is called the "ion suppression method." As acids and bases are affected
by the pH of the mobile phase, a buffer must be used instead of water to
ensure analysis repeatability. To control separation, the mobile phase pH
can be changed to achieve selective separation of acids and bases from
other substances.
The buffer is normally prepared by dissolving a weakly acidic or weakly
basic salt in water. Commonly used buffers are phosphoric acid, acetic
acid, boric acid, citric acid, and ammonium. The buffer is selected
according to its pKa, as a buffer exhibits the strongest buffer capacity at
the same pH as the weak acid pKa used (or pKa of the conjugate acid for
a weak base group). Assume that the target buffer pH is 4.8, for example.
As the 4.8 pKa of acetic acid is extremely close to the target buffer pH,
an acetic acid buffer is desirable. However, an acetic acid or citric acid
buffer is not suitable for measurements with a UV absorbance detector
at short wavelengths near 210 nm, as the background increases due to
the absorbance of the carboxyl group if acetic acid or citric acid is used.
Setting condition (3) requires information on the properties of chemical
compounds and attention to the points described above to set a pH that
achieves separation from other substances in the minimum possible time.
Reversed-phase Chromatography
Fig. 1 Types of Solid Phase
Fig. 2 Benzoic Acid
5
In this article, we discuss reversed-phase ion-pair chromatography
(RP-IPC), which extends the application range of RPC.
IPC was originally developed from a solvent extraction method called
ion-pair extraction. The ion-pair extraction method involves extraction
to an organic solvent layer after applying counter ions with the opposite
charge to an ionic substance in aqueous solution to form ion pairs and
neutralize the overall charge. Silica gel was used in the initial application
of IPC to HPLC but now most IPC is conducted using RPC. The
basic feature of RPC is that retention increases as the sample polarity
decreases. As discussed in the previous issue, when analyzing an ionic
substance using RPC, the retention depends on the dissociation state:
retention is enhanced in the undissociated state.
Consequently, adding counter ions to an ionic substance to form ion
pairs reduces the polarity and increases the retaining force. Consider the
example of the analysis of thiamine (vitamin B1). Thiamine is a strong
basic substance that contains quaternary nitrogen in its molecule and
normally carries a positive charge in a mobile phase. Therefore, it passes
directly through the hydrocarbon solid phase of the RPC packing without
undergoing a hydrophobic mutual interaction. Negatively charged
octanesulfonic acid ions added to the mobile phase form ion pairs with
the thiamine, thereby reducing its polarity such that it is retained by the
solid phase. That is, the ion-pair reagent (counter ions) mediates between
the nonpolar solid phase and ionic samples.
The actual RP-IPC retention mechanism cannot in practice be fully
explained using the simple model above. Three models are available,
the first of which is the ion-pair model that was explained above.
Next is the ion-replacement model, in which the nonpolar parts of the
ion-pair reagent are initially adsorbed onto the packing to form an ion-
replacement surface that retains the sample by ion replacement. Finally,
we have the ion interaction model, which has a broader range of
application than the ion-pair model and ion-replacement model. This
model considers the dynamic equilibrium; including the effects of not
only electrostatic forces but also mobile phase attractive and repulsive
forces and solid phase attractive and repulsive forces. Although some
phenomena are observed that cannot be explained by these models and
other models are also available for consideration, the models will not be
discussed further here.
An alkyl sulfonate salt such as sodium 1-pentane sulfonate or sodium
octane sulfonate is generally used as the ion-pair reagent for basic
substances, as described in the thiamine analysis example. Tetraalkyl
ammonium ions such as tetrabutyl ammonium hydroxide are often used
as the ion-pair reagent for acidic substances. All of these reagents are
readily obtainable from reagent manufacturers. Recently, highly refined
high-purity ion-pair reagents for HPLC applications have appeared on
the market. As with normal RPC, RP-IPC sample retention is affected
by the amount of organic solvent in the mobile phase, the concentration
of the buffer, and the pH. However, RP-IPC sample retention is also
affected by the type and concentration of the ion-pair reagent. Generally,
the longer the alkyl chains of the ion-pair reagent and the higher the
concentration (up to a certain level), the greater the sample retaining
force. A reduction in retaining force (foldover) is observed when the ion-
pair reagent concentration reaches a certain limit.
Two other points must be considered when using RP-IPC. Firstly, it takes
some time for the column to completely stabilize from the time that the
ion-pair reagent starts to flow into the column. Sometimes, the retention
time will not stabilize if a sample is injected after the time required to
achieve equilibrium with a normal RPC mobile phase. It is important to
wait patiently for the column to fully stabilize. Secondly, water should
not be used immediately to flush the ion-pair reagent out of the column
(especially a basic ion-pair reagent such as tetraalkyl ammonium ions). If
the column is flushed with water instead of buffer solution, the tetrabutyl
ammonium ions, which have hydrophobic regions, can remain adsorbed
on the packing and cause deterioration of the silica gel. In this case, it
is important to flush out the ion-pair reagent using a mixture of acidic
buffer and organic solvent with a proportion of between 1:1 and 1:2.
RP-IPC is a powerful technique that greatly expands the range of
applications of RPC and, consequently, HPLC. We expect that new
applications for it will be developed in the future.
Reversed-phase Ion-pair
Chromatography
6
Advances in HPLC and its widespread application bring demands for the
separation and quantitation of ever-smaller traces of target components
in increasingly complex samples. However, special methods are required
for this type of analysis, as it can be difficult to conduct using a normal
HPLC detector. One approach is derivatization, which enhances the
detection sensitivity and selectivity for the target component. Many
special reagents and techniques have been developed to improve
derivatization. In this article, we will discuss the general derivatization
methods used in HPLC.
Two met hods of derivat i zat i on are used i n HPLC: pre-col umn
derivatization and post-column derivatization. The difference between
these methods is that the derivatization reaction occurs before the column
with the pre-column method and after the column with the post-column
method. The features of each method are discussed below.
a) Pre-column derivatization
Pre-column derivatization is often conducted off-line. As the
derivatization reaction takes place outside the chromatography system,
the pre-column method offers the following advantages over the post-
column method:
1) simple equipment configuration;
2) no restrictions on the reaction conditions: reaction time, reaction
temperature, number of reagents, etc.;
3) small volumes of derivatization reagents used, permitting the use of
expensive reagents;
4) the derivatization reagent can be separated from the derivatization
product and have no influence on the quantitation accuracy or
detection limits in cases where the derivatization regent could be
detected (for example, the reaction reagent has photoabsorbance
at 400 nm and the reaction product also has photoabsorbance at
400 nm); and
5) derivatization often permits easier selection of the separation
conditions than the original components.
While the pre-column method offers significant freedom in selecting
the reaction, the following conditions must be met to achieve accurate
quantitation:
1) excellent reaction reproducibility;
2) reaction products are stable over time; and
3) byproduct peaks do not interfere with the target peaks.
Reagent manufacturers market numerous reagents for pre-column
derivatization that react with groups such as the carboxyl group, amino
group, and hydroxyl group.
b) Post-column derivatization
As the post-column method conducts derivatization online, it offers
the following advantages over the pre-column method:
1) automated reaction process permits unattended operation;
2) handles time-unstable reaction products, as the reaction time can
be accurately controlled by the flowrates of the mobile phase pump
and reaction reagent pump; and
3) permits partial reaction.
However, the post-column method must satisfy the following conditions:
1) to avoid increased background, the reaction reagent itself must
be undetectable - it changes to a detectable substance only after
reacting with the target component;
2) separation must not be sacrificed due to the restrictions on mobile
phase factors (pH, organic solvent concentration, etc.) required to
improve the reaction conditions; and
3) spreading of the component bands must not impair separation (in
other words, short reaction time).
Examples of the post-column method include the analysis of amino
acids using o-phthalaldehyde (OPA) and reducing sugar analysis using
arginine.
The derivatization method must be selected according to the goal of the
analysis, based on a good understanding of the features outlined above.
Demands on HPLC are likely to become more extensive and more severe
in the future. Meeting these demands will require the development of
novel HPLC systems that incorporate derivatization.
Derivatization in HPLC
7
Reversed-phase Chromatography and
Hydrophobic Interaction Chromatography
The separation of proteins and nucleic acids was conventionally handled
by gel filtration chromatography or ion-exchange chromatography.
Recently, research has been conducted into separation using reversed-
phase chromatography and hydrophobic interaction chromatography.
Thanks to the many new hard gels based on chemically bonded silica
gels that have been developed recently to dramatically accelerate the
separation of bio-macromolecules, the role of HPLC in this field will
become increasingly important.
Separation methods exploiting the hydrophobic interaction between
proteins and the solid phase include reversed-phase chromatography
(RPC) and hydrophobic interaction chromatography (HIC). The features
of these two methods are described below.
RPC has conventionally been the main stream of HPLC that was widely
used for the analysis of low molecular weight substances and has
been recently applied to the analysis of nucleic acids and proteins. A
chemically bonded silica gel with a large pore size is used as the column
packing when analyzing proteins. The gradient elution method is used
for the mobile phase conditions, with the amount of organic solvent
increasing at pH 2 to 3 or near neutral pH. As the protein sample is often
denatured during this process, this method should more correctly be
known as polypeptide analysis.
Conversely, HIC employs packing that is less hydrophobic than RPC
packing and uses gradient elution in which the salt concentrations
gradually decreases. HIC often permits the analysis of proteins without
destroying their higher-order structure, as no organic solvent is used
in the mobile phase, unlike RPC. Consequently, this method should be
applicable to preparative separation.
Let us now compare the two packings used in these methods. The main
packing type for RPC is chemically bonded silica gel with a large pore
size of at least 30 nm (compared to the conventional 8 to 10 nm small
pore size). This packing that has an embedded alkyl group is said to have
a good sample recovery rate. The residual silanol group in the chemically
bonded silica gel affects the recovery of protein samples and the
amount of residual silanol group differs according to the manufacturer
and production lot, such that quantitative considerations are currently
difficult to make.
Silica generally contains approximately 8mol/m
2
silanol groups. During
synthesis of normal RPC packing, a silanol-coupling agent containing
alkyl groups reacts with the silanol groups. The amount of the silanol
group that reacts with the agent is 3 to 4mol/m
2
, although it depends on
the length of the alkyl-group chain.
To reduce the effects of residual silanol groups during chromatography,
secondary silylation is generally conducted using a reagent such as
trimethylchlorosilane. Problems with denaturing and recovery rate
occur when such RPC packing is used for protein analysis because
an organic solvent must be used as the mobile phase, as described
above. Consequently, RPC is mainly used to analyze polypeptides with
comparatively low molecular weights. A known example is the analysis
of polypeptides obtained by the digestion of proteins by an enzyme such
as trypsin.
HIC generally uses packing created by embedding hydrophobic groups,
such as alkyl groups and phenol groups, onto a hydrophilic gel used
for gel filtration chromatography. However, the amount of embedded
hydrophobic group is significantly lower than for RPC. Silica-based and
polymer-based commercial HIC packings are currently available. The
silica-based packings are designed by the manufacturers to eliminate the
effects of residual silanol groups. The RPC and HIC packings exhibit
a large difference in the degree of hydrophobia, which results in the
significant difference in elution conditions described above.
RPC is superior to HIC in sample separation, whereas HIC results in
lower sample denaturing. As both HIC and RPC are expected to play an
important role for the analysis of proteins and nucleic acids in the future,
the careful selection of the appropriate technique will be essential.
8
Silica Gel Based Packing
Octadecylsilylated (ODS) and other chemically bonded, porous,
spherical silica gels are currently widely used as the packing in HPLC
columns. Many people who have conducted HPLC analysis have
probably experienced problems with differences between column
production lots or have questions about the differences with ODS
columns produced by different manufacturers. This article is intended to
clarify some of these queries about silica gel-based packings.
General-purpose ODS silica gel is typically produced by reacting
spherical, porous silica gel that has 5m average particle size and 6 to
10 nm average pore size with a silanol-coupling agent such as dimethyl-
octadecylchlorosilane. Approximately 8mol/m
2
silanol groups remain
on the silica gel surface but the reaction with the silanol-coupling
agent embeds octadecyl groups via siloxane bonds. However, when
using a bulky substituted group such as the dimethyl-octadecylsilyl
group, unreacted silanol groups always remain due to steric hindrance.
The amount of the unreacted silanol groups depends on the reaction
conditions, but is around 5mol/m
2
for normal ODS silica. Consequently,
attempts were made to conduct silylation of the residual silanol groups
using a silanol-coupling agent with a substituted group with lower steric
hindrance. This is the so-called secondary silylation, which uses
trimethylchlorosilane (TMS-CI) or other substances. Modern commercial
ODS silica gels are treated with TMS in this way (Fig. 1). The residual
silanol groups affect the adsorption of basic samples but the amount
of residual silanol groups differs from manufacturer to manufacturer.
Therefore, special attention is required during the analysis of basic
samples.
Let us now take a closer look at ODS silica gels. When spherical silica
gel is observed under a microscope, broken particles of silica gel are
often observed mixed in with it. Also, despite the nominal 5m particle
size, many fine particles below 2m and large particles of about 10m
are also included. This disparity in particle size varies according to the
manufacturer, and it has an adverse effect on the state of the column
packing. Most customers probably purchase packed columns. If a gap
occurs at the column inlet after a short period of use, there are probably
problems with the column packing due to particle size discrepancies.
A silica gel with a nominal 10 nm average pore size has a certain pore-
size distribution and the size of the pore-size distribution varies from
manufacturer to manufacturer. A bulky silanol coupling agent, such as
ODS, does not disperse in a fine-pored silica gel, such that the amount of
embedded ODS decreases and in some cases remains as residual silanol
groups without undergoing secondary silylation. Consequently, pore
size distribution control is essential to eliminate differences between
production lots in packing manufacture. In general, as the average pore
size increases, the silica gel specific surface area decreases, resulting in a
decrease in the apparent amount of embedded ODS and weaker retention
of low molecular weight substances. So-called wide-pore silica gels
with a pore size of 30 nm or more have recently been applied to the
analysis of macromolecules, especially proteins and nucleic acids. This
is because larger packing pore sizes result in more rapid dispersion of a
macromolecular sample in and around the pores.
Using different types of silylation agent permits the manufacture of
octyl, phenyl, cyanopropyl (CN), and aminopropyl (NH2) types of
chemically bonded silica-gel packing, in addition to ODS. The packing
must be selected according to the aim of the analysis. These packings
can be used in a pH range from 2 to 8; the silica dissolves in the alkaline
range and the chemically bonded solid phase separates at a pH below
2 due to cleavage of the Si-C bonds. Columns with polymer-coated
silica gel and columns with octadecyl groups embedded onto a synthetic
polymer surface are used to improve alkali resistance. However,
chemically bonded silica gel columns are these days widely used as
cheap, durable, high-performance packing.
Fig. 1 Silylation Reaction on Silica Surface
9
Absorptiometric Detection
Absorptiometry is a detection method that measures the absorbance
based on excitation of the valence electrons in the molecules. It is
the most widely used detection method for HPLC as it offers great
generality, superb selectivity, and good ease-of-handling.
The absorbance (A) is defined as the negative logarithm of the ratio of
the transmitted light intensity (I) to the incident light intensity (Io) when
monochromatic light passes through a light path length r in a solvent. It
bears the following relationship with the analyte concentration C:
A = Cr
Where, is the molar absorptivity of the analyte.
This relationship, where absorbance is proportional to the concentration,
is known as Beers Law. It applies when the interactions between
analytes can be ignored and the analyte chemical status is not dependent
on the concentration. With HPLC, the analysis target components that
elute from the column have a dilute concentration, such that Beers Law
applies and is used as a linear function for component concentration
during quantitation.
So, let us consider what substances can be analyzed by absorptiometry
and what the measurement conditions are. When the electrons,
electrons, and n electrons (which are non-bonding) in a molecule are
excited by light, they undergo transition from bonding or nonbonding
orbitals to antibonding orbitals. (An antibonding orbital is denoted by
an asterisk *.) As the -> * transition requires significantly higher
energy than the other transitions, compounds with only single bonds,
such as saturated hydrocarbons, cannot be detected in the 190 to 700 nm
wavelength range by this detection method. Compounds with n electrons
and isolated bonds exhibit absorbance in the ultraviolet range due
to n -> *, n -> *, or -> * transition but the molar absorptivity is
generally small. Consequently, in practice, target substances are limited
to conjugated systems. However, as many substances handled by HPLC
are conjugated systems, absorption photometry detection is commonly
used.
The measurement wavelength setting is an important point that
determines the detection sensitivity and selectivity. If many impurity
components exist, the wavelength can be selected to reduce their
influence; but the measurement wavelength is normally set to the
maximum absorbance wavelength of the analysis target component. This
is not just to obtain the highest sensitivity but also to eliminate effects of
wavelength-setting errors on the sensitivity. The maximum absorbance
wavelength is determined by measuring the absorbance spectrum with a
spectrophotometer. For a simple polyene structure or enone structure, the
maximum absorbance wavelength can be calculated using the Woodward
and Fieser rules. This calculation involves adding values applied to the
basic skeleton and the bonded functional groups. An example of this
calculation for a steroid enone, such as a mineral steroid, is shown below.
This method is always described in organic chemistry textbooks that
deal with spectroscopy. Refer to such a textbook for more detailed
information.
The absorbance spectrum and molar absorptivity are governed by the
chemical structure but in solution they change due to the solvation,
interactions with the third substance, and dissociated state. These
changes are particularly large when the dissociated groups are in
conj ugat ed syst ems. As an exampl e, t he maxi mum absorbance
wavelength of aniline and the molar absorptivity at this wavelength
change as follows.
Consequently, the optimal mobile phase conditions must be investigated
for detection as well as for separation. For aniline, setting the mobile
phase pH higher than the aniline pKa is superior from the point of
view of both sensitivity and selectivity. For this type of substance that
exhibits significant changes in absorbance due to the dissociated state,
good repeatability and linearity can be obtained only by using a buffer to
maintain a constant degree of dissociation.
The mobile phase is an important factor for this type of detection, as
it affects the chemical status of the target component. In addition, it
is important to consider the absorbance of the mobile phase itself. For
example, the noise increases and system peaks may appear if the mobile
phase contains a substance with high absorbance. This may also cause
baseline fluctuations during gradient elution. Ideally, a mobile phase
with low absorbance at the measurement wavelength should be selected.
Wi t h t odays dr amat i c i ncr ease i n i nst r ument per f or mance,
absorptiometry can detect some components at pmol levels if the optimal
conditions are set. It is a powerful tool for trace element analysis if the
absorbance of the substance is known.
10
Affinity Chromatography
Biological substances, such as enzymes and substrates or antigens
and antibodies, have the property of selectively recognizing each
other to form complexes. Affinity chromatography is a separation and
purification method that exploits this biological affinity. For example, a
specific enzyme can be extracted from a matrix containing an extremely
high number of components by using a substance as the ligand that
exhibits specific affinity for the target enzyme, such as its corresponding
substrate or inhibitor.
Procedure for Affinity Chromatography
The procedure comprises the following four steps:
(1) Column equilibration
The column is filled with a packing onto which a suitable ligand is
immobilized. The column is then equilibrated with the solvent that
will be used for sample adsorption.
(2) Sample addition and adsorption of target substance (Fig. a)
The sample is introduced into the column and the target component
adsorbed onto the packing.
(3) Column washing (Fig. b)
Washing is conducted to eliminate components not adsorbed onto the
packing from the column.
(4) Elution of target components (Fig. c)
A solvent is introduced into the column to elute the target component
adsorbed on the packing to recover the target component. The
procedure then reverts to step (1).
In principle, affinity chromatography can effectively extract a target
component from a complex matrix that contains multiple components
just using simple operations.
Ligand Selection and Immobilization on the Packing
The selection of the ligand and packing matrix and method to immobilize
the ligand must be carefully considered before conducting actual affinity
chromatography.
Selecting the Ligand
Elution from the column may be difficult if the affinity between the
substance used as the ligand and the target component is too strong. If
multiple ligands can be considered, select the substance which offers the
greatest selectivity for the target compound and which permits elution
from the column under the most moderate conditions.
Selecting the Matrix
Commonly used matrices are agarose and hydrophilic synthesized
macromolecules (polyvinyl alcohol, polyacrylate, etc.) beads. In
principle, a matrix is selected that is inactive with respect to the sample
and is physically and chemically stable under the conditions used. It
must also be noted that if the beads are porous, their pore distribution
and specific surface area affect the amount of immobilized ligand and the
effective amount of ligand on the packing surface.
Immobilizing the Ligand
The CNBr method and many other techniques to immobilize the ligand
have been reported. A ligand immobilization method must be selected
that permits the effective function of the regions exhibiting affinity
for the target component. The length of the spacer between the ligand
and matrix is also important, especially if the target component is a
macromolecule such as a protein. Many types of active matrix that
permit easy ligand immobilization are now commercially available.
Application Examples
Many cases of separation and purification by affinity chromatography
have been reported. Separation of lectin with sugar as the ligand
and separation of hormone receptors using a hormone-bound matrix
have been reported as examples of the separation of proteins with
an immobilized low molecular weight ligand. Examples using a
macromolecular substance as the ligand include the separation of
calmodulin-bound proteins using calmodulin immobilization matrix and
the separation of fibronectin using collagen as the ligand.
Many cases of enzyme purification have been reported. However, if
the substrate is used as a ligand, careful control of the chromatography
conditions is required to prevent the enzyme breaking down the ligand.
Antigens or antibodies are widely used as the ligand to isolate a specific
antibody or antigen. In this case, the extremely strong affinity between
the antigen and antibody requires measures such as lowering the eluate
pH or the use of a denaturing agent or chaotropic ions.
Affinity chromatography often presents the disadvantage that the user
has to make the adsorbent. Even so, this technique is likely to continue to
be used as a convenient and effective way to isolate target components.
13
Explaining GLP/GMP
Terminology: RSD (C.V.)
The importance of quality control is growing internationally, as typified
by the ISO-9000 Series and GLP/GMP for pharmaceuticals. The
validation of instruments and analytical measures is now required as a
method to objectively verify analytical instruments and data reliability
in a variety of industries. A check of the precision, i.e., the degree of
discrepancy in the analysis results, is one item in the method validation
and system suitability tests that are conducted during validation. This
article explores RSD (C.V.), which is used to express precision.
When recording or displaying analysis data, if only one data point
is collected, that value is used directly; if multiple data points are
collected, the mean value or median value is often used as statistical
characteristics. However, when discrepancies do exist in the data,
the exact status of the data cannot be accurately expressed by these
statistical characteristics alone. In practice, data never matches
perfectly from analysis to analysis and a discrepancy results. RSD
(C.V.) is often used as an objective indicator in the statistical analysis
of such discrepancies. In Table 1 below, the mean value is 100 for both
Results (1) and Results (2), but the discrepancies in Results (2) are
clearly larger. The RSD calculation yields RSD = 2.92% for Results (1)
and RSD = 29.2% for Results (2).
Table 1 Mean Values and RSD
Data Results (1) Results (2)
x1 101 110
x2 96 60
x3 103 130
x4 102 120
x5 98 80
Mean (x
) 100 100
RSD (C.V.) 2.92% 29.2%
Relative Standard Deviation (RSD) and Coefficient of Variation (C.V.)
are synonymous. These values are used to objectively express the data
discrepancy (precision). It is defined as the standard deviation (SD)
divided by the mean (x
x
1 46.204 2567643 501.91
(
x
1
2
) 426.962162 1318559569475 50382.7849
x
Where,
a0.05 is the width of the front of the peak (start point to apex) at 5% peak-
height position; and W0.05h is the peak width at 5% peak-height position.
Generally, if Tf<1, the peak is described as leading or fronting; if
Tf>1, the peak is described as tailing. A range of values from 0.5 to
1.5 is normally appropriate. If Tf lies outside this range, the cause should
be investigated. By way of reference, the System Suitability item in the
FDA Reviewer Guidance
3
) states that Tf<2 is desirable.
2. Resolution
Resolution (Rs) is an indicator of the degree of separation between two
components. It is defined as the difference in retention time between two
peaks, divided by the mean peak width. Like the number of theoretical
plates, resolution is defined slightly differently in the Japanese
Pharmacopoeia
1
) and U.S. Pharmacopoeia
2
), which adopt the peak width
at half height method and peak width tangent method, respectively.
1) Japanese Pharmacopoeia
1)
Where,
tR1 is the retention time of the front peak;
tR2 is the retention time of the rear peak;
W0.5h1 is the front peak width at 50% peak height position; and
W0.5h2 is the rear peak width at 50% peak height position.
2) USP
2)
Where,
W1 is the peak width of the front peak *;
W2 is the peak width of the rear peak *;
(* Note: The peak width is taken as the time amplitude between the two
points of intersection of the baseline and the tangents to the left
and right peak inflection points.)
Explaining GLP/GMP Terminology:
Tailing Factor, Resolution
15
Normally, Rs >= 1.5 indicates that separation has occurred (baseline
separation), thereby providing desirable conditions for reliable
quantitation. This value is particularly important for chromatography,
as one characteristic of chromatography is its ability to conduct the
simultaneous analysis of multiple components.
By way of reference, the System Suitability item in the FDA Reviewer
Guidance
3)
states that Rs >2 is desirable between the target peak and
the nearest adjacent peak (impurity, diluting agent, separated component,
internal standard, etc.).
In some cases, the separation coefficient (, relative retention) is also
used. The separation coefficient is the ratio of the capacity factors (k') of
two peaks. It is defined as follows:
Where,
k'1 is the k' value (capacity factor) of the front peak;
k'2 is the k' value (capacity factor) of the rear peak;
t0 is the unretained peak time (dead time);
tR1 is the retention time of the front peak; and
tR2 is the retention time of the rear peak.
The col umn performance val i dat i on funct i on of t he Shi madzu
LCsolution Workstation for HPLC can automatically calculate the
tailing factor, resolution, and separation coefficient using the Japanese
Pharmacopoeia, USP, and other calculation methods. If the calculation
results for these parameters do not meet the pass criteria, pass/fail
evaluation using QA/QC functions can take actions such as reinjection or
stopping analysis.
References
1. Japanese Pharmacopoeia 12
th
Edition, General Test Methods (Liquid
Chromatography)
2. USP (United States Pharmacopeia) XXIII-5, <621> Chromatography
3. Center for Drug Evaluation and Research (CDER, FDA), Reviewer
Guidance, Nov. (1994) Validation of Chromatographic Methods
4. LCtalk Vol. 34, TEC Calculation of Number of Theoretical Plates.
Peak
Japanese Pharmacopoeia, DB USP
Tailing factor
Number of
theoretical
Separation Tailing factor
Number of
theoretical
Separation
A
1 1.41 15649 1.41 12701
2 1.28 20444 11.34 1.28 17558 10.34
3 - 20389 1.65 - 17718 1.53
4 1.20 22245 8.47 1.20 20233 7.97
B
1 2.49 5972 2.49 5426
2 - 7917 7.02 - 7310 6.70
3 - - - - 5371 0.90
4 1.71 9957 - 1.71 9316 4.91
Sample Calculations of Tailing Factor and Resolution
16
Often, the word sensitivity is used to express the ability of an
analytical method or analytical instrument when discussing the detection
ability with respect to minute quantities of substances. Sensitivity is
actually a term indicating the magnitude of the detection response with
respect to an amount of a component in a sample, and to be correct,
detection limit or quantitation limit is an index of detection ability.
Detection limit is defined as the amount of a substance that generates
a signal which differs significantly from a blank. However, differ
significantly is not actually prescribed in ISO or in JIS, and the lack of a
uniform interpretation causes confusion in the handling of the detection
limit.
Here we wi l l expl ai n t he defi ni t i on of det ect i on l i mi t and i t s
measurement method as set forth by the ICH (International Conference
on Harmonization of Technical Requirements for Registration of
Pharmaceuticals for Human Use), organized for the purpose of
harmonizing interpretation among Japan, the United States and the
European Union.
According to ICH, detection limit is defined as the lowest amount of
analyte in a sample which can be detected but not necessarily quantified
as an exact value. The phrase not necessarily quantified as an exact
value means that there is no necessity for an accompanying permissible
degree of truth and accuracy, differing from quantitation limit in this
respect.
The method for determining the detection limit differs depending on
whether or not the analytical method is an instrument analytical method,
however, it can be categorized broadly into the following 3 methods.
(1) Method based on visual evaluation
This method is used when the analytical method is not dependent
on an instrument analytical method. Here, a sample of known
concentration is gradually diluted, and the greatest dilution stage at
which the sample can be distinguished from the blank is taken as the
detection limit.
(2) Method based on signal-to-noise
This method primarily used in chromatography. The analyte signal
and the baseline noise are measured, and the analyte concentration
(mass) at which the ratio is 2:1 or 3:1 is taken as the detection limit.
First the sample concentration is adjusted so that the peak height is
at least 10 times that of the baseline noise, and then measurement
is conducted to obtain the peak height h. (European Pharmacopeia-
1987)
For noise, the maximum fluctuation hN is measured over a time
period 20 times that of the peak width at half height, and 1/2 of that
value is taken as noise N (Fig. 1, European Pharmacopeia- 1987), or
the baseline recorded for 15 minutes is divided into 0.5 - 1 minute
segments (X1, X2, X3....), parallel lines are drawn in the horizontal
axial direction separated by the minimum width required to include
the noise within each segment, the vertical distance between the
parallel lines is calculated (Y1, Y2, Y3....) along the segment division
lines, and the average value is obtained and taken as noise N (Fig. 2,
JAIMA S 0005-1984, ASTM E 1657-1994).
Figure 2
Figure 1
Detection Limit
17
(3) Based on standard deviation of response and slope of
calibration curve
This is a method in which a blank sample is analyzed and its
standard deviation is obtained, or several samples containing an
analyte with concentrations (mass) close to that of the detection limit
are analyzed, the residual standard deviation of the regression line
or the standard deviation of the y-intercept is determined, and the
detection limit is calculated using the following equation.
DL = 3.3/a
( : standard deviation of response
a : slope of calibration curve)
This is called the Currie detection limit (L.A. Currie : IUPAC Provisional
Draft, 1994), in which the detection limit is the quantity of substance
corresponding to the signal at the position B + 3.29 B obtained from
the average B of the blank signal distribution (B, B). In other words,
signal distribution N (s, B) when the probability of an error of the
first kind occurring (probability of falsely deciding that a signal is
present when it is not) and the probability of an error of the second kind
occurring (probability of falsely deciding that a signal is not present
when it is) are both 5%.
In HPLC, the determination method normally used is either that based
on the signal-to-noise ratio or that based on the standard deviation of
response and slope of the calibration curve.
The determination method based on signal-to-noise is easy and is the
most commonly used, however, when there is baseline fluctuation
such as in gradient elution analysis, or when there are nearby impurity
peaks, this method is not suitable. Although the method based on
the standard deviation of response and slope of the calibration curve
requires abundant data for the calculation, it has the advantage of being
applicable to any case.
The method based on the standard deviation of response and slope of the
calibration curve is shown below. However, since the standard deviation
of the blank cannot be obtained in HPLC, it is obtained from the standard
deviation of the regression line (residual or y-intercept).
Theoretical Conc. 0.04% 0.06% 0.08% 0.10% 0.12%
1st run 3551 4446 6182 7963 9405
2nd run 3282 5089 6294 8154 9226
3rd run 3013 5050 6418 8078 9084
4th run 3635 4907 6793 7668 9780
5th run 3119 4686 6109 7525 9591
Table 1 Data of Samples Containing Analyte at Known Concentrations
Detection Limit : DL = 3.3sy /x/a = 0.011 (%)
(Residual standard deviation :
sy /x= {{yi - (axi+b)}
2
/(n-2)}
1/2
=252.857)
or
Detection Limit : DL = 3.3sy/a = 0.013 (%)
(y-intercept standard deviation :
sy=sy /x{1+1/n+ (xi/n)
2
/S (xi-xi/n)
2
}
1/2
=294.879)
(a = 76182 : regression line slope,
b = 267.36 : y-intercept
n = 25 : total analysis repetitions)
Note that the y-intercept standard deviation calculation formula
shown above represents the distribution of the regression-dependant
variable (extrapolation predicted value) when the independent variable
(theoretical concentration) is 0. There is a separate equation for
calculating the y-intercept standard deviation, however, it should be
pointed out that when that equation is used, a smaller value is obtained
compared to when using other methods (calculation by S/N ratio or
residual standard deviation). For that reason, it is preferable to use this
equation.
18
LC-MS Discussion (part 1)
Introduction
Currently, liquid chromatograph-mass spectrometry (LC-MS) is
spreading widely throughout the pharmaceutical, environmental, food,
industrial materials and other fields. From this issue forward we will
offer commentary on the topic of LC-MS over several issues.
What Are the Merits of LC-MS?
In general, in LC, various constituents of a sample are separated due to
differing affinity (retention force) for the stationary phase (column) and
mobile phase, and depending upon the properties of the components,
they are detected using UV, fluorescence, electrical conductivity, etc.
Using these detectors, qualitative analysis of the substances is conducted
primarily on the basis of retention time, and quantitation is conducted
using peak height and area. Chromatography can provide excellent
separation, however, reliable identification and quantitation are difficult
in cases where many constituents elute from the column at about the
same time, such as when conducting simultaneous analysis of many
components.
On the other hand, MS is a high sensitivity detection method in which
analyte species are first ionized using various techniques, then generated
ions are separated in a vacuum based on the ratio of their mass and
electrical charge (m/z), and finally each ion intensity is measured. The
obtained mass spectrum shows the extent to which an ion with a given
mass number is present, thereby greatly assisting in qualitative analysis.
The mass number is information that is specific to the molecule, and
this information is obtained directly from the MS. However, this is the
situation in which constituents are analyzed individually. If multiple
analytes are injected simultaneously, spectral analysis becomes
extremely difficult.
LC-MS is an instrument system that combines the excellent separation
ability of LC with the excellent qualitative ability of the MS. A mass
spectrum obtained using scan analysis can provide molecular weight and
structural information, supplementing the retention time given by other
LC detectors for conducting qualitative analysis (Fig. 1). Moreover, in
SIM (Selected Ion Monitoring) analysis, detection is conducted using the
mass number, a parameter providing high selectivity. Even in the event
of inadequate LC separation, quantitative analysis can be conducted to
circumvent the influence of impurities. The combination of wide analyte
accommodation and selectivity afforded by the MS make it a powerful
LC detector.
Comparison with GC-MS
Mass spectrometers were used as detectors in gas chromatography (GC)
relatively early, and their benefits are widely recognized. From the
viewpoint of separation and qualitative analysis of substances, GCMS
is an effective method. However, the samples that can be analyzed are
limited to relatively low molecular weight gases or volatile compounds,
and compounds that are very thermally stable. In LC, on the other
hand, if the sample can just be dissolved in the mobile phase (liquid),
substances that are not suitable for GCMS analysis, such as those that are
not easily volatilized or are thermally unstable, can be analyzed. In other
words, LC-MS has the advantage of being suitable for analysis of a wide
range of samples.
LC-MS Instrument Composition
The mass spectrometer anaytical system consists of the sample
introduction component (HPLC, GC, etc.), the interface between the
sample injection and the MS sections, the ion source which ionizes the
sample, the electrostatic lenses which efficiently guide the generated
ions, the mass spectrometer which separates the ions according to m/z,
and the detector which detects the separated ions.
There are various types of mass spectrometers depending on the ion
separation method. Shown here (Fig. 2) is the composition of an
atmospheric pressure ionization type quadrupole MS which is generally
used as an LCMS detector. Atmospheric pressure ionizers include the
electrospray ionizer (ESI) and the atmospheric pressure chemical ionizer
(APCI), etc., and they serve as an interface with the HPLC and as an ion
source. Following desolvation, ions which are generated here are directed
by the octapole, to the quadrupole mass analyzer. In the quadrupole mass
analyzer, combined direct current and RF alternating current potentials
are impressed on the rods so that only ions with the selected m/z can pass
through. The number of ions that reach the detector are converted to a
signal, which is acquired at the PC.
Fig. 2 LC-MS System Composition
Fig. 2 Atmospheric Pressure Chemical Ionization (APCI): Ion-Molecule Reaction
Fig. 1 Vitamin B2 Mass Spectra
Fig. 2 Erythromycin Collision Induced Dissociation Spectrum
24
Evaporative Light Scattering Detector
After the evaporative light scattering detector (ELSD) was introduced,
we received many letters requesting more detailed information about
ELSD. This article features ELSD.
What is ELSD?
The ELSD (Evaporative Light Scattering Detector) is an instrument
which nebulizes the column eluent to remove the mobile phase solvent
through evaporation, then directs light at the remaining non-volatile
constituents and detects the scattered light. Since any substances that
do not evaporate with the mobile phase are detected, this instrument is
sometimes referred to as a universal detector. Actually, ELSD has been
on the market for more than 20 years, but because of its relatively poor
sensitivity and stability, it has not been very popular. However, especially
in response to increased requirements in the pharmaceutical field, many
improvements have been made in recent years, greatly enhancing the
ELSD performance and thereby increasing its popularity.
Principles of ELSD
Fig. 1 illustrates the detection principle of the Shimadzu ELSD-LT.
The ELSD detection process is divided into 3 steps: (1) nebulizing the
column eluent, (2) evaporating the mobile phase, and (3) detecting the
scattered light from non-volatile sample constituents. The column eluent
is nebulized by the flow of nitrogen or air from the nebulizer. As it
proceeds through the temperature-controlled drift tube, the mobile phase
is removed through evaporation, and only the non-volatile components
are sent to the detector. In the detector, light is irradiated onto these
constituent droplets, and the resultant scattered light is detected by the
photomultiplier. Fig. 2 illustrates the structure of the detection unit.
ELSD vs. RID
RIDs (Refractive Index Detectors) are also universal detectors that
detect any substance even without UV absorption. The RID detects the
difference in the refractive index between the sample constituents and
mobile phase. Since there is normally some difference between these
refractive indices, fundamentally, anything can be detected by RID. Let
us compare the ELSD and RID, which are both universal but based on
completely different detection principles.
Merits of ELSD
The primary advantage of the ELSD is its sensitivity. The S/N in sugar
analysis using the ELSD is 5 - 10 times higher than that of the RID.
Secondly, the RID is associated with baseline drift. With the RID, an
extremely small difference in refractive index between the sample
constituents and mobile phase is detected, and therefore it is necessary
to maintain the mobile phase refractive index stable at all times.
However, as the refractive index is affected by slight fluctuations in
temperature and flow rate, it is difficult to obtain a straight baseline in
high sensitivity analysis even with the newest technology. Naturally
gradient elution is impossible when using the RID. The ELSD, on
the other hand, is free from this problem. Especially, the capability
of gradient elution provides a powerful merit for analyzing multiple
substances without UV absorption. Furthermore, with the RID, if the
sample solvent is different from the mobile phase, the peak of the
solvent may interfere with the analyte peak. The ELSD is also free
from this problem.
Cautions with ELSD
With the RID, there is no restriction with the mobile phase. With
the ELSD, however, the mobile phase must be volatile. This means
phosphate buffers cannot be used. Also, the ELSD detects not only
target substances but also dirt and dust. Care is required for impurities
in the mobile phase and soluble column packing materials, because
they cause noise in detection. Another thing to point out is linearity.
The RID provides linearity over a wide range. However, with the
ELSD, substance quantity and the intensity of scattering light are
exponentially related, and the calibration curve needs to be linearized
using the double logarithm.
Application Fields of ELSD
The ELSD has been used to analyze sugars and lipids, which cannot be
detected by UV absorption. In addition to this, the ELSD is currently
attracting increasing attention as an instrument for impurity verification
in the pharmaceutical field. (Refer to the issue No. 51.) It is also being
used in analysis of herbal medicines and other natural substances. It
is expected that the ELSD will be more popular as the range of its
Fig. 2 Detector Structure (ELST-LT)
Fig. 1 Detector Principle (ELSD-LT)
25
Methods of Amino Acid Analysis
Amino acids are attracting increasing attention with the heightening of health
consciousness. Amino acids are most commonly analyzed by HPLC. Here we
discuss the methods of amino acid analysis, focusing primarily on the methods
of detection.
Types of Detection Methods
The only means available to detect amino acids with UV absorption is to utilize
the absorption of the carboxyl group (-COOH) at 200 - 210 nm. Although
amino acids with benzene rings can be detected at 250 - 280 nm, it is generally
difficult to directly analyze amino acids with high selectivity and sensitivity.
Accordingly, the derivatization method has been used. Most amino acids have
amino groups (-NH2, -NHR) in their structure, and a derivatization reagent is
used to selectively react with these groups.
Pre-column Derivatization
With this method, amino acids are derivatized before injection, and the
resultant substances are separated and detected using HPLC. Fig. 1 illustrates
the outline of this process.
The merits of this method are as follows:
Reagent consumption can be minimized by setting up a small reaction
system.
As a result of the merit above, expensive reagents can be used, making
possible high sensitivity analysis with low background.
Even if a portion of the derivatization reagent remains unreacted, it is
separated in the column and therefore does not affect the analysis results.
This method also has a disadvantage. Since the sample is mixed directly with
the derivatization reagent, the reaction efficiency is largely influenced by the
sample matrix (co-existing constituents and solvents, etc.).
Considering the advantages and disadvantages, it is said that pre-column
derivatization is suitable for high sensitivity analysis of some specific samples.
Representative pre-column derivatization reagents for amino acid analysis
include o-phthalaldehyde, phenyl isothiocyanate, fluorescamine, and dansyl
chloride. The procedures to cause reaction differ for each reagent. Some
reagents cause reaction simply by being mixed with the sample at room
temperature, and others require heating. There are also reagents that require
cleanup after the reaction.
Reversed-phase chromatography (RPC) is often used for the separation of the
reaction product. Since RPC is capable of quick analysis with high resolution,
high-throughput analysis is possible if the conditions are set optimally.
Post-column Derivatization
With the post-column derivatization method, the amino acids are first separated
in the column, then the derivatization reagent is introduced into the eluent to
cause reaction, and finally, the resultant compounds are guided to the detector.
Fig. 2 illustrates a typical flow line diagram of this process.
The merits of this method are as follows:
The reaction can be automated, improving quantitation performance and
reproducibility.
Sample constituents are separated before derivatization, minimizing the
influence of the matrix. As a result the method is applicable to a wide
range of samples.
The disadvantages are the difficulty in enhancing sensitivity, and the large
consumption of reagent (reagents need to be continuously delivered).
Considering the advantages and disadvantages, post-column derivatization is an
excellent technique for quantitation in routine analyses that can be applied to a
wide range of samples once the reaction system is optimized.
Since the reaction reagent continuously flows into the detector, only specific
reagents can be used in post-column derivatization, to prevent unreacted
reagent from being detected. Currently only two reagents are available for
analysis of amino acids: ninhydrin for UV-VIS detection and o-phthalaldehyde
for fluorescence detection.
The most common separation method is cation-exchange chromatography.
Since amino acids are ampholite ions that have both amino and carboxyl
groups in their structure, in cation exchange, constituents with higher acidity
(constituents that easily become anions) elute faster, and constituents with
higher basicity (constituents that easily become cations) elute more slowly.
By the way, why is the cation-exchange method is used even though the
mainstream of HPLC is reversed-phase chromatography? This is because
the separation of amino acids and other substances including amino groups
(amines), which is difficult with the reversed-phase chromatography, can be
conducted easily and with good efficiency using cation-exchange.
Since the derivatization reagents used in amino acid analysis possess reactivity
for amino groups, they also react with amines, and those may be detected as
peaks. Amines generally do not have an anion type functional group like a
carboxyl group, showing a stronger basicity than amino acids. Therefore they
elute more slowly than amino acids in cation-exchange. This is how the cation-
exchange method prevents amines from interfering with the quantitation of
amino acids.
As described above, cation-exchange chromatography, which can separate
amino acids and amines, and post-column derivatization, which realizes
selective reaction with amino acids, provide a perfect combination.
Fig. 3 shows an example of a chromatogram obtained by analyzing an
amino acid beverage by the post-column derivatization method using
o-phthalaldehyde as the reagent.
Fig. 1 Pre-column Derivatization
26
Analysis of Organic Acids
Organic acids, along with amino acids and sugars, are often analyzed for
deliciousness in food products. Here we will discuss the analysis of
organic acids using HPLC.
Separation
There are three major modes for the separation of organic acids: anion-
exchange, ion exclusion and reversed-phase. We briefly describe each mode
below.
Anion-exchange mode (Fig. 1)
In this mode, separation is realized by the negative ions and organic acid
ions (negative ions) in the mobile phase competing for the positive ions on
the column packing. The retention behavior on the column surface differs
depending on the organic acid radius and ion valence. Separation of many
organic acids requires gradient elution, and thereby a long analysis time.
Ion Exclusion Mode (Fig. 2)
This is the most common mode for the analysis of organic acids. H-type
ion exchange resin is used as the packing material, and the organic acids
are separated according to the degree of Donnan exclusion between the
stationary phase (H-type ion exchange group) and the mobile phase. In
this mode, strong acids are subjected to a large electrostatic exclusion by
the negative electrical charge of the stationary phase, and cannot enter
the pores of the packing material. With weak acids like organic acids,
however, the degree of infiltration into the pores is determined by the size
of the electrical charge (pKa), causing differences in elution time, namely
separation. According to this principle, organic acids elute in the sequence
of their pKa values (from small to large), with all of them having eluted
before the elution time of the neutral substances (the time when the pores
are completely infiltrated). In actual analysis, however, highly hydrophobic
organic acids may elute more slowly due to hydrophobic interaction with
the packing material matrix. This mode is widely used because of its
operational simplicity.
Reversed-phase Mode
Although this is the most versatile HPLC mode, it is not widely used
for the analysis of organic acids. This is because sufficient retention or
selectivity cannot be achieved due to the hydrophilic properties of organic
acids. Recently, reversed-phase columns that also retain highly hydrophilic
substances are appearing on the market.
Detection
The easiest detection method is the direct UV direct detection, in which
the carboxyl group absorption at 200 - 210 nm is detected. However, since
many organic substances show absorption in this wavelength region, there
is great interference due to impurities, making accurate quantitation difficult
unless the sample is extremely clean. Although differential refractive
index detectors can be used, they do not provide sufficient selectivity and
sensitivity. In direct detection using the electrical conductivity detector,
it is necessary to lower the electrical conductivity of the mobile phase
background.
Post-column methods that selectively detect organic acids with high
sensitivity include the visible absorption detection using pH indicators and
electrical conductivity detection using pH buffers. The pH-indicator method,
which utilizes the color change of pH indicators in response to organic acids,
is slightly poor in linearity and procedural simplicity. The pH buffer method
is excellent for its sensitivity, selectivity and accuracy.
pH-buffer Method
In the ion exclusion mode widely applied in analysis of organic acids, acidic
mobile phase is normally used and therefore, the background electrical
conductivity is kept high. The pH-buffer method continuously adds a
pH buffer reagent to the column eluent to bring it near a neutral pH and
dissociate the organic acids. Then the dissociated organic acids are detected
by an electrical conductivity detector with high sensitivity. This is a very
effective method that lowers the mobile phase background, and at the same
time enhances sensitivity by dissociating the organic acids.
The pH-buffer method enables highly sensitive and selective detection
of organic acids, as well as analysis of samples over a wide range of
concentrations, thanks to its wide range of calibration curve linearity. This
method is adopted in Shimadzus organic acid analysis system. Fig. 3 shows
the results of soy sauce analysis. Using this system, organic acid analysis
even in samples like soy sauce can be conducted with good accuracy,
without interference from impurities.
Fig. 1 Anion-exchange Mode
Fig. 2 Ion Exclusion Mode