HPLC Guide

Departments R450, R452, R45R

Authors: Seble Wagaw, Jason Tedrow, Tim Grieme, Lalit Bavda, Weifeng Wang,
Shekhar Viswanath, David Barnes, Maureen McLaughlin
Table of Contents

Table of Contents………………………………………………………………………...2

Introduction................................................................................................................... 3

Role of Assays in Chemical Development..................................................................... 3

Method development..................................................................................................... 4

HPLC Columns ........................................................................................................... 4

Mobile phase............................................................................................................... 7

Gradients..................................................................................................................... 8

Assays ............................................................................................................................ 9

Sample preparation...................................................................................................... 9

Data collection and analysis ........................................................................................ 9

Standards and assays ................................................................................................. 10

Appendix 1: Calculating a Correlation Coefficient………………………………….13

Appendix 2: Theory……………………………………………………………………14

Appendix 3: Quick Reference Guide to Setting up an HPLC Assay ……………... 16

2
 Introduction
The purpose of this guide is to provide information on the effective use of HPLC
analysis in chemical development for departments R450 (Process Chemistry), R452
(Process Engineering), and R45R (Hazards Analysis Lab). Topics included are basic
method development, assay techniques, data analysis, and applications of HPLC assays in
chemical development. This manual will only cover the use of reverse phase
chromatography.

Role of Assays in Chemical Development

HPLC assays are powerful tools to gather qualitative and quantitative data to
understand and monitor a chemical reaction and subsequent product isolation. Since
process chemists and engineers are typically assigned to projects before analytical
chemists, it is up to them to develop suitable HPLC methods and assays for development
work. The degree of rigor with which HPLC assays are used increases throughout route
development, process optimization, and scale-up work. It is important, especially during
early route development when relatively unproven HPLC methods are being used, to
employ more than one method of analysis (e.g. NMR or TLC). In preparation for GMP
deliveries, information gained by engineers and chemists about critical assays is shared
with analytical chemists, who then do further method development and appropriate
method validation.
In early route development, HPLC assays are primarily used to quickly assess
conversion, relative reaction purity profiles, and purity of the isolated product. The
emphasis at this point is on speed more than rigor. The use of peak area percents (PA%)
to describe reaction mixtures is acceptable to quickly rank proposed reaction schemes.
However, the use of a working standard should be incorporated as early as possible, even
before a standard of high purity is available. Once a sufficient amount of product is
isolated, the cleanest lot is used as a working standard and replaced with cleaner material
as it becomes available (vide infra). Subsequent working standards may then be
referenced back to earlier standards to accurately compare data. Standards are typically
not available for impurities at this stage, therefore impurity levels are reported as relative
peak area percents at a specified wavelength. Preliminary mass balance calculations
based on analyzable species, typically the limiting reagent and product, are carried out.
Once a synthetic route is identified, HPLC assays are used to demonstrate control of
the process, to track and minimize impurities, and to maximize the yield and purity of the
isolated product. Characterized working standards are utilized for quantification. Mass
balance is monitored to enable full understanding of the process. Product purity profiles
from various experiments are used to establish product specifications for upcoming
campaigns. When the route advances to a stage at which high purity products are
produced, peak area percent (PA%) of impurities may be a more useful indication of
purity than potency vs a standard. For example, potencies of 98 –99 % may be within
experimental error, but an increase in peak area percent of an impurity from 1 PA% to 2
PA% may be relevant.
In scale-up work, well-characterized standards of high quality are preferred. Process
chemists and engineers use HPLC assays to ensure that the scale-up is proceeding as
planned and to help design further refinements to the process. Analytical chemists use

3
validated methods for critical in-process assays and isolated intermediates to confirm that
the process and product conform to previously established specifications.

Method development
The determination of the suitability of a HPLC method is based upon the level of
development. However, at a minimum a HPLC assay should provide baseline separation
of starting materials, desired products, known impurities, and expected by-products. The
chromatographic conditions should also be chemically compatible with the analytes.
This section will cover the role of the HPLC column, mobile phase and gradient
development. For a brief discussion of the basic theories of chromatography, including
capacity factor (k), selectivity (α), system efficiency (N, also known as theoretical plates)
and resolution (R) see appendix 2.

HPLC Columns
The heart of a HPLC system is the column. Changing a column will have the
greatest effect on the resolution of analytes during method development. The three main
components of an HPLC column are the hardware (column housing), the matrix, and the
stationary phase. Column hardware will not be discussed; handling guidelines are
included with each column and should be followed for optimal results and column
lifetime. Generally, modern reverse phase HPLC columns are made by packing the
column housing with spherical silica gel beads which are coated with the hydrophobic
stationary phase. The stationary phase is introduced to the matrix by reacting a
chlorosilane with the hydroxyl groups present on the silica gel surface. In general, the
nature of stationary phase has the greatest effect on capacity factor, selectivity, efficiency
and elution1
There are several types of matrices for support of the stationary phase, including
silica, polymers, alumina, and zirconia. Silica is the most common matrix for HPLC
columns.2 Silica matrices are robust, easily derivatized, manufactured to consistent
sphere size, and do not tend to compress under pressure. Silica is chemically stable to
most organic solvents and to low pH systems. One shortcoming of a silica solid support
is that it will dissolve above pH 7. In recent years, silica supported columns have been
developed for use at high pH (vide infra).3
The nature, shape and particle size of the silica support effects separation.
Smaller particle results in a greater number of theoretical plates, or increased separation
efficiency. However, the use of smaller particles also results in increased backpressure
during chromatography and the column more easily becomes plugged. For this reason 5
Å columns are more frequently used than 3 Å columns in development work. Narrower
particle size distribution of the silica particles also results in better resolution. Hence,
similar phase columns from different manufacturers, or different lots of columns from the

1
For a discussion of capacity factor, selectivity, system efficiency (theoretical plates) and resolution see
appendix 2.
2
For more information on polymer, hybrid and alumina columns see
http://www.waters.com/WatersDivision/pdfs/lc3AC.pdf.
3
The column will come with specifications as to solvent and mobile phase compatibilities and tolerated pH
ranges. For maximum efficiency and lifetime, follow these guidelines

4
same manufacture may have very different separation properties due to differing methods
of matrix preparation.
The nature of the stationary phase will determine whether a column can be used
for normal phase or reverse phase chromatography. Normal phase chromatography
utilizes a polar stationary phase and a non-polar mobile phase. Generally, more polar
compounds elute later than non-polar compounds. Types of columns suitable for normal-
phase phase chromatography include underivatized silica, nitrile (vide infra), amino (or
aminopropyl), glycerol and nitro columns. Chiral separation is usually performed under
normal phase conditions.4 Since highly polar and ionic compounds are retained on
normal phase columns, a guard column or silica gel sample purification should be used to
extend the column life.5
In reverse phase chromatography the stationary phase is non-polar and the mobile
phase is polar, causing polar peaks to generally elute earlier than non-polar peaks. To
create a stationary phase for reverse phase chromatography on silica support, the free
silanols are reacted with a chlorosilane with hydrophobic functionality to introduce the
non-polar surface. Due to steric constraints, only about 1/3 of the surface silanols are
derivatized. The remaining free silanols can interact with analytes, causing peak tailing.
Typically, after the derivitization of a column with the desired stationary phase, the
column is further reacted with chlorotrimethylsilane to end cap the remaining free
silanols and improve the column efficiency. Common stationary phases are C4 (butyl),
C8 (MOS), C18 (ODS), nitrile (cyanopropyl), and phenyl (phenyl propyl) columns. In
general, longer alkyl chains, higher phase loading, and higher carbon loads provide
greater retention of non-polar analytes. Selectivity is most influenced by the amount of
accessible surface area of the derivatized silica gel particles and the carbon load. Thus it
is often a benefit to not only have columns with different stationary phases, but columns
with the same phase from different manufacturers. Commonly used reverse phase
columns and their uses are listed below.

Propyl (C3), Butyl (C4), and Pentyl (C5) phases are useful for ion-pairing
chromatography (C4) (vide infra) and peptides with hydrophobic residues, and other large
molecules. C3–5 columns generally retain non-polar solutes more poorly when compared
to C8 or C18 phases. Examples include Zorbax SB-C3, YMC-Pack C4, and Luna C5.6
These columns are generally less stable to hydrolysis than columns with longer alkyl
chains.

Octyl (C8, MOS) phases have wide applicability. This phase is less retentive than the C18
phases, but is still quite useful for pharmaceuticals, nucleosides, and steroids. Octyl
columns are also useful for peptides, peptide mapping and small hydrophilic proteins
when bonded to 300 Å silica particles. Examples include (Zorbax SB-C8, Luna C8,
YMC-Pack-MOS).

4
For commonly used chiral columns and applications see: www.chiraltech.com.
5
For more information about normal phase columns, please visit the various manufacturers’ websites (Agilent,
Phenomenex, Waters and Supleco).

6
More information about different columns may be found at the specific company website. Zorbax =
www.agilent.com YMC and Xterra = www.waters.com, Luna and Synergi = www.phenomenex.com.

5
Octadecyl (C18, ODS) columns are the most widely used and tend to be the most retentive
for non-polar analytes. This phase is useful in ion-pairing chromatography and has wide
applicability (same as C8 in addition to vitamins, fatty acids, environmental compounds).
Examples include Zorbax SB-C18, YMC-Pack ODS and Luna C18.

Xterra RP-C18 and Zorbax Extend-C18 columns have been formulated to tolerate high pH
systems (pH >7, normally up to pH 11). Varying the pH can dramatically affect
selectivity and resolution of polar analytes, especially for ionizable compounds (vide
infra).

Phenyl (Ph) columns offer unique selectivity from the alkyl phases and are generally less
retentive than C8 or C18 phases. Phenyl columns are commonly used to resolve aromatic
compounds. Examples include Zorbax SB-Phenyl, YMC-Pack Phenyl and Luna Phenyl-
Hexyl.

Nitrile (CN or cyano) columns are polar and can be used for both reverse and normal
phase applications. This phase is often used to increase retention of polar analytes. The
nitrile derivatization allows for rapid column equilibration. Examples include Zorbax
SB-CN, Luna-CN, and YMC-Pack CN.

Column compatibility with various mobile phases is determined by the stationary

phase. In reverse phase chromatography the mobile phase is composed of an aqueous
buffer and a water miscible organic solvent that has little or no absorption above 200 nm.
The common organic solvents in reverse phase chromatography are methanol,
acetonitrile and tetrahydrofuran (vide infra). The operational pH range of a column is
dependent on the stability of the silica gel support. Special columns with high phase
loading, extensive end-capping, or cross-linked stationary phases have been developed
for use at pH < 2, or at pH > 7. It is important to check the literature provided with the
column for mobile phase compatibility.
Standard C18 Columns and similar stationary phases will undergo phase collapse
at highly aqueous mobile phases, typically at less than 5-10% organic composition; this
will decrease analyte-stationary phase interaction. Collapsed phases are also difficult to
re-equilibrate. To prevent phase collapse, C18 columns with a polar group embedded in
the alkyl chain have been developed to help solvate the hydrophobic chain in >90%
aqueous mobile phases.7 Examples include Zorbax SB-Aq, Synergi Hydro-RP and
YMC-Pack ODS-Aq.
Further versatility of C18 and C8 columns can be achieved by utilizing ion-pairing
chromatography. In ion-pairing chromatography, ionic modifiers such as
tetrabutylammonium or octylsulphonate salts are added to the mobile phase to increase
capacity factors for ionic or ionizable compounds1. The alkyl chain of the ion-pairing
modifier will partition into the hydrophobic stationary phase, producing an ionic surface
and leading to increased retention of charged analytes (figure 1).8 Capacity factors of
uncharged analytes will also change slightly due to the change in stationary phase.

7
Such modifications will impart selectivity differences from normal C18 columns.
8
For more information on ion-pairing chromatography see http://www.registech.com/ionpair/.

6
 ion paring reagent

stationary phase
silica support
figure 1. ion pairing chromatography

Column temperature control is important for long-term method reproducibility as

temperature can affect selectivity.1 A target temperature in the range of 30–40 °C is
normally sufficient for good reproducibility. Use of elevated temperature can be
advantageous for several reasons. First, operating at a temperature higher than ambient
reduces the viscosity of the mobile phase and thus the overall backpressure on the
column. Lower system pressures allow for faster flow rates and thus faster analyses. The
temperature may also affect selectivity patterns because analytes will respond
dissimilarly to different temperatures. Finally, use of a column oven eliminates
variability due to normal fluctuations in the air temperature surrounding the column.

Mobile phase
Though choice of column has the greatest effect on resolution, the mobile phase
also effects selectivity and efficiency and is the aspect of chromatography over which we
have the most control.1 In reverse phase chromatography, the mobile phase consists of an
aqueous buffer and a non-UV active water miscible organic solvent. The effect of the
organic and aqueous phase, and the proportions in which they are mixed will be
discussed in this section.
The aqueous buffer serves several purposes. At low pH, the mobile phase
protonates free silanols on the column and reduces peak tailing. At sufficiently low pH
basic analytes are protonated; when ionized the analyte will elute more quickly but with
improved peak shape. Acidic analytes in buffers of sufficiently low pH will remain
uncharged, increasing retention. Conversely, at higher pH neutral basic compounds will
be more retained, and ionized acidic compounds will elute earlier. Peak splitting may be
observed if the pKa of a compound is similar to the pKa of the buffer, and the analyte
elutes as both a charged and uncharged species. The pH of a buffer will not greatly affect
the retention of non-ionizable sample components.
Typically a 10 – 50 mM solution of an aqueous buffer is used. The most
commonly used aqueous phase is a 17 mM solution of H3PO4 in water (0.085% v/v
H3PO4). The pH of a phosphate buffer is easily adjusted by using mono-, di-, or tribasic
phosphate salts. However, when phosphate salts are used the solution should be filtered
to remove insoluble particles. Other non-UV active acids and bases may also be used to
effect differences in peak shape and retention. Table 1 gives a selection of buffers
covering a range of pKa’s.

7
Table 1.
Buffer pKa PH range UV cutoff
HClO4 – 10
H3PO4/KH2PO4 2.12 1.1 – 3.1 <200 nm
KOAc/AcOH 4.8 3.8 – 5.8 210 nm
KH2PO4/ K2HPO4 7.21 6.2 – 8.2 <200 nm
NH4OH 9.2 8.2 – 10.2 200 nm
.
Et3NH/HCl Et3NH 11.0 10.0 – 12 <200 nm

Since the degree of solubility of the components of a sample will vary

independently in different solvents, the choice of organic solvent can affect selectivity
and therefore resolution. In some cases the order of elution of sample components of
similar polarity will change with different organic mobile phases. Organic solvents
typically used in reverse phase chromatography are methanol, acetonitrile, or THF in
order of increasing strength; the strength of a solvent refers to its ability to retain analytes
in the mobile phase. Acetonitrile is a highly polar aprotic solvent, providing adequate
resolution for many compounds. Due to its ability to form hydrogen bonds, the use of
methanol either as an additive or as the organic phase can provide significantly different
selectivities. Tetrahydrofuran is the most hydrophobic of the three organic solvents
mentioned. Care should be exercised when using tetrahydrofuran in the mobile phase as
it is difficult to fully remove from a column. Selectivity is also greatly affected by
amount of aqueous solution in the mobile phase, with higher percentages of aqueous
phase leading to increased retention and frequently to improved selectivity.

Gradients
Solvent gradients are commonly used in reverse phase chromatography, whereas
isocratic methods are commonly used in normal and chiral phase chromatography.
Gradient methods permit faster separation with samples containing multiple components
of varying degrees of polarity in comparison to isocratic methods, and tend to result in
better peak shape for late eluting compounds. Gradient methods typically run from a
mobile phase of high aqueous composition to high organic composition, eluting polar
compounds before non-polar compounds. When gradient methods are used the column
has to be equilibrated to the initial mobile phase composition prior to the next run. When
using a standard length C18 column, this can be achieved by programming a 5 – 10
minute post time in the method.
A common gradient method in chemical development runs from 10% to 90%
organic over 15 minutes, followed by a 5 minute 90% organic isocratic period. While
this gradient is ideal for samples containing a range of polar to non-polar compounds
with sufficiently different relative retention times, optimization is needed when this
method fails to provide adequate resolution. 1
In general, a slower ramp rate will give better peak separation but will result in
longer run times and may cause peak broadening. Changes made to the beginning of a
method will have a greater effect on resolution than those made to the end of method.
Therefore, introducing an isocratic period in the early portion of a run, or using slow
gradient followed by a steeper gradient is often sufficient to improve resolution while
maintaining reasonably short analysis times. Alternately, when a region of poor

8
resolution is identified with respect to solvent composition at time of elution, resolution
may be improved by introducing an isocratic region or slower gradient at the appropriate
solvent composition. However, in instances where resolution is poor due to poor peak
shape resolution may be improved by increasing the ramp rate. Regardless of the method
chosen, care should be taken, especially when using isocratic methods, to ensure that all
compounds of interest are eluting from the column during the method run time. This may
be tested by initially injecting the sample on several different methods or by extending
the method run time.

Assays
After the appropriate column and method conditions are established, attention
shifts to getting useful data from an HPLC assay. Consistent practices in assay
techniques should be used to produce accurate and reproducible data.

Sample preparation
Assay solutions are prepared by adding a known weight or volume of the analyte
to a volumetric flask and diluting the sample to an appropriate concentration.9 Typically,
target concentrations of 0.1-1.0 mg/mL are used with 5-10 µL injection volumes (vide
infra). When assaying reaction mixtures, the concentration of the sample should be
estimated to achieve an appropriate final concentration. The diluent should completely
dissolve and be chemically compatible with the analyte. Reaction mixtures that contain
insoluble inorganic solids should be filtered prior to dissolution. The solvent of choice
for dissolution is usually the method mobile phase (ie. acetonitrile, acetonitrile\buffer, or
water). If the sample does not have sufficient solubility in these solvents, then alternate
solvents that are miscible with the mobile phase and are chemically compatible with the
sample are used. Keep in mind that UV active solvents may obscure sample peaks.
When assaying highly reactive reaction mixtures, samples should be quenched to ensure
accurate results are obtained.

Data collection and analysis

Data collection also affects the quality of the analysis. Optimally when analyzing a
single component, the analysis should be performed at the UV max of the compound of
interest to maximize sensitivity. The UV spectrum of a peak can be obtained with a
diode array detector. Diode array detectors are available on some HPLC’s or on the
walk-up LC/MS systems. When assaying a multiple component mixture, for example a
reaction mixture containing starting material, product and impurities, a wavelength
should be chosen which allows for detection of all relevant components. In some cases,
this may involve assaying the sample at several wavelengths. In a chromatogram, the
peak height of the major component should not exceed 1500 mAu (target 800-1000 mAu)
to obtain proper integration. Sample concentrations should be adjusted as needed to
achieve the targeted range. Linearity, or system suitability, may also be used to more
rigorously determine an acceptable range of sample concentrations. A linearity study is

9
When it is necessary to achieve an accurate and quantitative result (e.g. assay yields), it is desirable to
measure samples by weight rather than volume. If a volume assay is required, care should be taken to
avoid the use of pipettors and measure the volumes using syringes and volumetric flasks using proper
dilution techniques.

9
performed to determine a range of sample concentrations that is linearly proportional to
the UV response at a given wavelength. To determine the linearity, start with a sample
concentration that gives a peak height of 800-1000 mAU. Prepare two samples of higher
concentration and two samples of lower concentration, for example 80%, 90%, 110% and
120% of the original concentration.10 Plot the UV response (mAu*s) versus sample
concentration (mg/mL). The linear range of the plot indicates the range of concentrations
in which wt% assays will be accurate. The linear range may be determined visually, or
more rigorously by calculating the correlation coefficient. [See appendix 1 for
information on calculating a correlation coefficient]

In order to accurately follow reactions and track purity profiles it is important to

make all UV active components of the sample visible when displaying and printing
chromatograms. Using appropriate sample dilution insures that the main component is
within scale and therefore accurately integrated (vide supra). In order to emphasize
smaller peaks, the y-scale of printed chromatograms should not exceed 400 mAu
(suggested range is –20 to 180 mAu). Integration methods should be set so that the
minimum integration threshold is low enough to integrate minor components of the
chromatogram. When printing chromatograms, it is also beneficial to include the
method on the printed plot.

Standards and assays

As previously mentioned, product standards are used throughout all stages of
chemical development to obtain quantitative results from HPLC assays. In addition to
performing weight percent (wt%) assays, standards are used to obtain information on
relative response factors between reagents, products, and when standards are available of
by-products. With a relative response factor, wt% and therefore mol% conversion of
reactions can be determined without weighing assay samples.
A standard is the highest purity lot of isolated compound, preferably purified by
crystallization or silica gel chromatography. Minimally all standards should be
characterized by 1H NMR and MS, and corrected for wt% purity. As the process
advances further characterization of the reference standard by GC, TGA, IC, EA, or KF
may be warranted. Ideally, a fresh standard should be prepared and injected immediately
prior to sample injection for quantitative assays. The standard should be run at least
weekly; attention should be paid to the stability of a standard when stored in solution for
extended periods of time.
Weight percent assays are calculated by relating the response of the reference
standard (mAu/mg) at a given wavelength to a known weight of the sample to be
analyzed. Weight percents are calculated using the following formula.

10
For cases in which samples of very low concentration may be injected (such as in a solubility study), it
may be necessary to extend this range to low concentration.

10
 areasoln* volsoln/weightsample
wt% = * 100
area*mL
mg
standard

areasoln = area count (mAU*s) of the peak in question of the diluted assay sample
volsoln = volume of dilution for the assay sample (size of volumetric flask)
weightsample = weight of sample in volumetric flask
note: use same mass units throughout the calculation

equation 5. weight percent calculation

Weight percent assays of solutions are used to quantify crude assay yields,
product loss to extracts and crystallization liquors. These wt% assays are obtained by
multiplying the assayed wt% by the weight of solution in question. Determination of
crude assay yields allows the user to quantitatively differentiate between reaction
conditions, offering more information than is obtained from PA% analysis. Similarly,
wt% assays are used to quantify product loss to extractions, which allows for the
optimization of work up procedures.
Similar to wt% assays, external standards can be used to determine relative
response ratios. A relative response ratio is a comparison of the response
(absorbance/mg) of two sample components at a given wavelength. Once this ratio is
determined, it can be used to calculate relative weight percents and therefore mole
percents of two components in a sample. This becomes particularly important in
instances where reactants and products have significantly different UV spectra, and PA
percentages do not accurately reflect the sample composition.

response ratio (n) = (area/mg)component A

(area/mg)component B

wt% component A = areaA/n

(areaA/n) + areaB

equation 6. weight percent conversion calculation

Assays are also used for the measurement of supernatant concentrations in

crystallization development to maximize purity and recovery. These measurements are
used in solubility studies to identify optimal crystallization solvents, and to follow the
desaturation of crystallizations. Supernatant concentrations are determined by diluting a
known volume of the supernatant solution in a volumetric flask. Supernatant
concentrations are calculated following equation 1, but replacing weightsample with the
volume of the supernatant solution (mL).

11
 Finally the isolated product is assayed against the reference standard to obtain a
weight percent purity using equation 1. All isolated yields should be adjusted for the
weight percent purity, and reported as the purity adjusted yield. The purity adjusted yield
is the true reaction yield and is the quantity to be used in calculating reaction mass
balance. It should be noted, when the weight percent purity of the isolated product is
>100% the purity adjusted yield cannot be adjusted higher than the actual amount of
product isolated. This new lot of product will now become the new standard.
The sum of the purity adjusted yield and product loss to extractions and liquors
should add up to the calculated assay yield of the crude reaction mixture. The reaction
mass balance is the sum of the residual starting material, isolated product, and all possible
sources of product loss including extractions, washes, and crystallization liquors. The
reaction mass balance is a measure of reaction efficiency, where a low mass balance
reflects the formation of significant amounts of by-products.

12
 Appendix 1

Calculating a Correlation Coefficient

Linearity is indicated when the coefficient is equal to 1. The correlation

coefficient can be calculated from Microsoft Excel by the following procedure.

1) Enter your data in columns as shown below.

2) Click on a separate cell then click = which is to the left of the formula bar. From the
drop down menu to the left find CORREL; CORREL can be found under formulas, more
functions, then statistical.
3) You will be asked to add the array 1 (highlight the cell range of values for the
concentration) and array 2 (highlight the cell range of values for the Area).
4) Click OK, and the correlation coefficient will be calculated.

Conc (mg/mL) Area (mAu*s)

0.0830 7137893
0.0934 8071364
0.1038 8997555
0.1142 9856508
0.1246 10682971

Correlation Coefficient =
0.9996

Figure 2. sample linearity plot

13
 Appendix 2

Theory
As with any type of chromatography, there are four basic concepts in basic HPLC
method development: capacity factor (k), selectivity (α), system efficiency (N, also
known as theoretical plates) and resolution (R). Capacity factor (k) is the ratio of a
compound’s elution volume, or retention time, relative to the time of elution of an
unretained compound.11 It is a measure of how much time a compound spends in the
stationary phase versus the mobile phase. Peaks of interest should have a k>2 for
optimum resolution. The capacity factor is generally independent of the equipment and is
a function of the choice of column and mobile phase.

(RTanalyte - RTunretained)
capacity factor (k) =
(RTunretained)

equation 1. capacity factor

Selectivity (α) is the ratio of the capacity factors of two peaks. It is a measurement of the
difference in interactions of two analytes with the mobile and stationary phases, and
therefore the difference in retention times. As with capacity, the selectivity of a column
is an effect of the column packing material and the eluents used.

k1
selectivity (!) =
k2

equation 2. selectivity

Separation efficiency (N) is a measure of the sharpness of peaks eluting from a specific
column, and is described as the number of theoretical plates in a column.12 It is
calculated from the retention time and the width of the peak at ½ peak height. In general
shorter columns decrease analysis time, but do not increase the system efficiency (N).
Often, increasing column temperature will increase the efficiency of separation.

RT
Separation efficiency (N) = 5.54 *
Widthhalf height

11
Void-volume, or dead-volume, may be determined by the retention time of an unretained analyte such as
uracil, which is not retained on common reverse phase columns. It also can be estimated as the time of
refractive disturbance at the beginning of an HPLC run.
12
Technically N is a function of the entire HPLC system, but in the average HPLC system the effect of the
instrument is minimal. With decreasing diameters and lengths, instrument contribution to this parameter
becomes significant.

14
 equation 3. separation efficiency

Arguably the most important concept in HPLC chromatography is resolution (R), which
is dependent on all the factors listed above. R is calculated as the difference in retention
time of two analytes divided by the average width of the two peaks at the baseline.

(RTpeak 1 - RTpeak 2)
resolution (R) =
1/2 (widthpeak 1 + widthpeak 2)

equation 4. resolution

As shown by the equations described above, developing a method with good resolution
and a reasonable run time is a balancing act between selectivity and efficiency. The most
effective way to improve resolution is by changing the stationary and mobile phases.

15
 Appendix 3

Quick Reference Guide to Setting up an HPLC Assay

Below are guidelines to assist in “getting started” on an HPLC assay. A more detailed
description of the following parameters can be found on pp.10-12 (Assays).

• Sample prep: Samples should typically be diluted to a concentration of 0.1 – 1.0

mg / mL to ensure the sample lies within the linearity range of the assay. The
solvent used for dissolution is usually mobile phase, but may be any compatible
non-UV active solvent. Samples containing inorganic solids should be filtered.
• Injection: Injection volume should be 5-10 µL to ensure reproducibility.
• Column Temperature: The column oven should be set above ambient
temperature to eliminate error introduced by fluctuations in room temperature.
35oC is a good starting point.
• Column Choice: The choice of column will be dependent upon the nature of the
sample to be analyzed. A standard column such as C8 or C18 is a good starting
point.
• Peak height: The height of the peak of interest should not exceed 1000 mAu to
avoid saturation of the detector.
• Reference Standards: All quantitative assays are carried out relative to a
reference standard of assigned weight percent purity. That standard will usually
be a sample of the product, preferably of high chemical purity. All reference
samples should be analyzed by at least two methods, typically NMR and either
GC or HPLC. It is best to inject the standard immediately prior to the sample for
quantitative assays and at least weekly to minimize day-to-day variance in analyte
response.
• Linearity: For the most reliable numbers, one should run a linearity study of the
reference standard. The importance of doing this is related to the stage of
development. To do this, one prepares solutions over a wide range of
concentrations, and plots the response vs. the concentration. Future analyte
samples should be run in the range in which the resulting plot is linear.
• Detector: If possible, assays should be performed at the UV max of the
compound of interest. Standards of all available products, impurities, starting
materials, etc., should be injected to determine their relative response factors.
• Plotting results: It is important to plot results such that the baseline is blown up
to reveal small impurities. In general, one should be able to see ~0.l % peaks off
the baseline. A range of –20 – 180 mAu is recommended.

16