R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.

com

A Guide to Validation in HPLC

Based on the work of G. M. Hearn PERKIN ELMER

1. Introduction

Sometimes you may wonder who was the first to make an experimental procedure for a specific
analysis. The route from an idea to an actual standard operating procedure (SOP) is not easy. In
the early days it was a long period of trial and error to establish a document where common
users could benefit from. Nowadays the use of computer programmes like Drylab or HIPAC
makes it easier to optimize the intended procedure to a practical analysis for other laboratories
and schools. This document will try to explain some of the basic jargon and glossary concerning
HPLC. At the end you may understand the long route to a method that has been validated as
have many of the experiments and analytical procedures you will use within StandardBase.

2 METHOD VALIDATION

"Validation of an analytical method is the process by which it is established by laboratory

studies, that the performance characteristics of the method meet the requirements for the
intended analytical application - "(6)

Validation is required for any new or amended method to ensure that it is capable of giving
reproducible and reliable results, when used by different operators employing the same equipment
in the same or different laboratories. The type of validation programme required depends entirely on
the particular method and its proposed applications.

Typical analytical parameters used in assay validation include:

1. Precision
2. Accuracy
3. Linearity
4. Range
5. Ruggedness
6. Limit of detection
7. Limit of quantitation
8. Selectivity
9. Specificity
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com

2.1 Precision

"The precision of an analytical method is the degree of agreement among individual

test results obtained when the method is applied to multiple sampling of a
homogenous sample -"(6)

Precision is a measure of the reproducibility of the whole analytical method (including

sampling, sample preparation and analysis) under normal operating circumstances.
Precision is determined by using the method to assay a sample for a sufficient number of
times to obtain statistically valid results (ie between 6 - 1 0). The precision is then
expressed as the relative standard deviation
std dev x 100%
%RSD = ──────────
mean

2.2 Accuracy

"Accuracy is a measure of the closeness of test results obtained by a method to the

true value. " (6)

Accuracy indicates the deviation between the mean value found and the true value. lt is
determined by applying the method to samples to which known amounts of analyte have
been added. These should be analysed against standard and blank solutions to ensure
that no interference exists. The accuracy is then calculated from the test results as a
percentage of the analyte recovered by the assay.

Accuracy and precision are not the same, as the diagram below indicates. A method can
have good precision and yet not be accurate.

Errors in measurement can be divided into two general categories: systematic errors and
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
random errors.

Systematic errors result from sources that can be traced to the methodology, the
instrument or the operator, and affect both the accuracy and the precision of the
measurement.

Random errors only affect the precision, and are difficult to eliminate, because they are the
result of random fluctuations in the measured signal, due to noise and other factors.

Whilst systematic errors are proportional to the sum of individual contributions, random
errors are proportional to the root of the sum of the squares of the individual contributions.
Thus, the imprecision of the entire procedure is often dominated by the random errors of
the most imprecise step.

2.3 Linearity

This is the method's ability to obtain results which are either directly, or after mathematical
transformation proportional to the concentration of the analyte within a given range.
Linearity is determined by calculating the regression line using a mathematical treatment of
the results (ie least mean squares) vs analyte concentration.

2.4 Range

The range of the method is the interval between the upper and lower levels of an analyte
that have been determined with acceptable precision, accuracy and linearity. It is
determined on either a linear or nonlinear response curve (ie where more than one range is
involved, as shown below) and is normally expressed in the same units as the test results.
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
2.5 Ruggedness

Ruggedness is the degree of reproducibility of results obtained by the analysis of the same
sample under a variety of normal test conditions ie different analysts, laboratories,
instruments, reagents, assay temperatures, small variations in mobile phase, different days
etc. (ie from laboratory to laboratory, from analyst to analyst.)

2.6 Limit of Detection

This is the lowest concentration in a sample that can be detected, but not necessarily
quantitated, under the stated experimental conditions. The limit of detection is important
for impurity tests and the assays of dosages containing low drug levels and placebos.
The limit of detection is generally quoted as the concentration yielding a signal-to-noise
ratio of 2:1 and is confirmed by analyzing a number of samples near this value (6) using
the following equation. The signal-to-noise ratio (5) is determined by:

s = H/h

Where H = height of the peak corresponding to the component

h = absolute value of the largest noise fluctuation from the baseline of the
chromatogram of a blank solution.

Since the limit of detection is dependant on the signal-to-noise ratio, it

can be improved by enhancing the analyte signal and reducing the detector noise.

The signal (ie peak height) can be increased by selecting the optimum monitoring
wavelength, increasing the injection volume or mass (below signal or column saturation),
increasing the peak sharpness with high efficiency columns and by optimizing the mobile
phase. For absorbance detectors, longer path lengths in the flow cell enhances sensitivity
though often to the detriment of post column dispersion.

Noise can be reduced by using high sensitivity detectors with low noise and drift
characteristics, slower detector response time, mobile phases with low absorbance and
pumps with low pulsation.

2.7 Limit of Quantitation

This is the lowest concentration of analyte in a sample that can be determined with
acceptable precision and accuracy.

lt is quoted as the concentration yielding a signal-to-noise ratio of 1 0: 1 and is confirmed

by analyzing a number of samples near this value (5).

2.8 Selectivity and Specificity

R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com

Selectivity is the ability to measure accurately and specifically the analyte in the presence
of components that may be expected to be present in the sample matrix.

Specificity for an assay ensures that the signal measured comes from the substance of
interest, and that there is no interference from excipient and/or degradation products
and/or impurities.

Determination of this can be carried out by assessing the peak identity and purity.

Diode array detectors can facilitate the development and validation of HPLC assays.
Spectra] data obtained from diode array detectors, effectively supplement the retention
time data for peak identification, also spectral manipulation often provides information
about the peak purity. The table below lists several of the techniques available for
assessing peak identity and purity.

The purity index is a measure of the peak's relative purity, measured using a full
comparison of spectral data for the leading and training edge of the peak Figure 4. A value
of 1.5 is commonly accepted to indicate a pure peak but >1.5 would indicate the presence
of an impurity, (9) as shown in Figure 5.
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com

3. SYSTEM SUITABILITY TESTS (SST)

Once a method or system has been validated the task becomes one of routinely checking
the suitability of the system to perform within the validated limits.
The simplest form of an HPLC system suitability test involves a comparison of the
chromatogram trace with a standard trace (as shown below). This allows a comparison of
the peak shape, peak width, baseline resolution.
Alternatively these parameters can be calculated experimentally to provide a quantitative
system suitability test report:

Number of theoretical plates (efficiency)

Capacity factor,

Separation (relative retention)

Resolution,

Tailing factor

Relative Standard Deviation (Precision)

These are measured on a peak or peaks of known retention time and peak width.
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
3.1 Plate number or number of theoretical plates (n)

This a measure of the sharpness of the peaks and therefore the efficiency of the
column. This can be calculated in various ways, for example the USP uses the peak
width at the base and the BP at half the
height.

where
Wh = peak width at 1/2 peak height
Wb = peak width at base
t = retention time of peak

Therefore the higher the plate number the more efficient the column.

The plate number depends on column length - ie the longer the column the larger the
plate number. Therefore the column's efficiency can also be quoted as:

Either- as the plate height (h), or the height equivalent to one theoretical plate (HETP).

h= L where L = length of column n

n
Or- as plates/meter.
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com

3.2 Capacity factor (capacity ratio) k

This value gives an indication of how long each component is retained on the column
(ie how many times longer the component is retarded by the stationary phase than it
spends in the mobile phase).
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
k' is used in preference to retention time because it is less sensitive to fluctuations in
chromatographic conditions (ie flow rate) and therefore ensures greater reproducibility
from run to run.

In practice the k value for the first peak of interest should be >l to assure that it is
separated from the solvent.

3.3 Separation Factor (relative retention)

This describes the relative position of two adjacent peaks. ldeally, it is calculated using
the capacity factor because the peaks' separation depends on the components'
interaction with the stationary phase.

Therefore considering peaks A and B

Figure 1 1: Separation factor calculation

k for the later peak is always placed in the numerator to assure a value >l.

lf the capacity factor is used then the separation factor should be consistent for a given
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
column, mobile phase, composition and specified temperature, regardless of the
instrument used.

NB. The separation factor gives no indication of the efficiency of the column.
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
3.4 Peak Resolution R
This is not only a measure of the separation between two peaks, but also the efficiency
of the column. It is expressed as the ratio of the distance between the two peak
maxima. (At) to the mean value of the peak width at base (Wb).

Figure 12: Peak resolution calculation

3.5 Tailing factor T

This is a measure for the asymmetry of the peak.
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
3.6 Relative Standard Deviation or precision

For an HPLC system this would involve the reproducibility of a number of replicate
injections (ie 6) of an analytical solution.

The USP requires that unless otherwise specified by a method:

- if a relative standard deviation of <2% is required then five replicate injections should be
used

- if a relative standard deviation of >2% is required then six replicate injections should be
used

Tabie 5 below lists factors w,hich could affect the precision of an HPLC system.

Tabie 5: Factors affecting LC system precision

Precision Controlling Factors

Retention time Pump f low and composition precision Column temperature

Mobile phase composition

Peak area Autosampler: inj mode, inj volume

Pump: flow, pulsation
Detector: noise and drift, response
Data system: sampling rate,
integration parameters

In most cases the system's Relative Standard Deviation is required; deciding which of the
other tests are required is not straightforward. To assist with the decision it has been
suggested that those parameters which have an affect on the system precision should be
used. For instance the resolution of two peaks with similar retention times should be quoted,
because, if it is below a critical value, the precision will be affected.

In addition "diode array" detectors allow for the determination of the relative purity factor
typically called: Peak Purity.

The retention time precision is important, because not only is retention time the primary
method for peak identification, but also variations can indicate problems within the LC
system (ie with the piston seals, check valves etc). Use of a column oven can overcome
laboratory temperature variations, which is the most common cause of retention time drift.
R.A. van Iterson Drenthe College Emmen Holland for www.standardbase.com
The most dominant factor controlling the repeatability of peak area is the autosampler's
precision, though the affect of noise and integration parameters will become more
significant with small peaks.

CONCLUSION

Reproducible quality HPLC results can only be obtained if attention has been paid to the
method development, validation and the system's suitability to carry out the analysis.

REFERENCES

1. GLP - The United Kingdom Compliance Programme (Department of Health) 1989.

2. Code of Fed: Reg 21 Part 211.160 (Government Printing Office Washington DC (1
978).
3. BS 7501 EN 45001 General Criteria for the operation of Testing Laboratories 1989.
4. Practical Liquid Chromatography An lntroduction R W Yost, L S Ettre, R D Conlan
Perkin-Eimer 1980.
5. United States Pharmacopeia XXI I (United States Pharmacopeial Convention,
Rockvilie, MD, 1990).
6. Validation of compedial Assays-Guidelines'Pharmacopeial Convention, Rockvilie,
MD, 1985".
7. Perkin-Eimer Technical Bulletin LCTB 19 "Liquid Chromatographic Considerations for
High Sensitivity Impurity and Stability Testing of Pharmaceuticals", Michael W Dong, et
al.
8. British Pharmacopeia 1988. Appendix A84 fit D.
9. Poile A E & Conion R D J. Chromatogr. 204 149 -152.
10 http://www.standardbase.com for more experiments.