\t
`
`jl
`
`Q.
`
`1
`
`C I
`
`Thomas E. Beasley
`Advanced Separation Tecbnoiogies, Inc.
`Whippany, New Jersey
`
`i CD 3Jam Bughu
`
`Hoffmann—La Roche, Inc.
`Nutiey, New Jersey
`
`Raymond P. W. Scott
`Georgetown University
`Washington, 0.6; and
`Birkbeck (foliage, University of London
`London, Engiano‘
`
`MARCEL
`
` DEKKER
`
`MARCEL DEI<1<Ez2., INC.
`
`Nxsw YQRK «- BASEL
`
`FRESENIUS KABI 1021-OOO1
`
`

`
`Library of Congress Cataloging-in-Publication Data
`
`Beesley, Thomas E.
`Quantitative chromatographic analysis E Thomas E. Beesley, Benjamin
`Buglio, Raymond P.W. Scott.
`p. cm. -— (Chromatographic science series ; 85)
`includes bibliographical references and index.
`ISBN 0-8247-0503-3 (acid~free paper)
`1. Chromatographic analysis. 2. Chemistry, Analytic-«Quantitative.
`I. Buglio, Benjamin. H. Scott, Raymond P. W. (Raymond Peter
`William).
`Ill. Title. IV. Chromatographic science ; V. 85.
`QDl17.C5 B35 2001
`545'.89—-dr:2l
`
`00~i)4757S
`
`This book is printed on acid-free paper.
`
`Headquarters
`Marcel Dekker, Inc.
`270 Madison Avenue, New York, NY 10016
`tel: 212-6968000; fax: 212-685-4540
`
`
`
`Eastern Hemisphere Distribution
`Marcel Dekker AG
`
`Hutgasse 4, Postfach 812, CH-4001 Basel, Switzerland
`tel: 41-6l~26i-8482; fax: 41-6}-261-8896
`
`World Wide Web
`
`hitp:!fwww.del<:l<er.eoni
`
`The publisher offers discounts on this book when ordered in bulk quantities. For more infor-
`mation, write to Special Salesz‘Professional Marketing at the headquarters address above.
`
`Copyright © 2001 by Marcel Dekker, Inc. All Rights Reserved.
`
`Neither this book nor any part may be reproduced or transmitted in any form or by any
`means, electronic or mechanical, including photocopying, miomfilrning, and recording, or
`by any information storage and retrieval system, without permission in writing from the
`publisher.
`
`Current printing (last digit):
`10
`9
`8
`7 6 S 4 3 2
`
`l
`
`PRINTED IN THE UNITED STATES OF AMERICA
`
`FRESENIUS KABI 1021-OOO2
`
`

`
`Contents
`
`Preface
`
`iii
`
`PART 1 Intruductiun to Quantitative Chmmatographic Analysis
`
`Chapter 1 The Critical Factors that Govern a Successful
`Quantitative Chromatographic Analysis
`Historical Introduction
`The Importance of Chromatography as an Analytical Technique
`Critical Factors Involved in a Successful Chromatogluaphic:
`Analysis
`Sample T:'(m.s'p0rfa:io:z mm’ Storage
`Sample Pr‘eparm‘z'0:2
`Analyiicai Pz'oce(fm'es
`Data Processmg
`Ana!yfz'caZ Reports
`Synopsis
`References
`
`Chapter 2 Sample Collection, Transport and Storage
`
`Gas Samples
`Samplingfor She Major Conzpomenfs ofa Mixz‘z.:re of
`Permcmem Gases
`Sampfizzgfar Minor Components ofa Gas Mz',~:tzm3
`Tile Azmfysis 0fF!ower Fragrances
`Head Space Arzaiysis
`Scaiia’ Phase Micro-E.m'aciio:2
`Liquid Sampling
`Solid Sampling
`Powder, Gra2m1’eS 0:‘ Cry.s‘za!s
`Reductian oftize Sample to xfizafyficaf Size
`Sample Size
`Synapsis.
`Reconlnwnded Reading
`
`3
`
`9
`
`10
`12
`13
`13
`16
`16
`17
`18
`
`21
`
`21
`
`22
`25
`29
`31
`33
`36
`38
`39
`41
`43
`43
`44
`
`FRESENIUS KABI 1021-OOO3
`
`

`
`vi
`
`Chapter 3 Sample Preparation
`Introduction to Sample Preparation
`Extraction Techniques
`Pre—Extraetz'on Pmofices
`Sofvent Exzmctiorz
`Sofia’ Pizase Exrmcrion
`The Solvent E,r:zrczcii0n ofSo:’idS
`Saaper—Critical Fiuici Extracriorz
`Derivatization
`ESterz'fca2f2'on
`Acylariozz Reactions
`P0s2‘—Coz’unm Deriwziizatiorz
`Concentration Techniques
`Synopsis
`References
`
`Chapter 4 The Chromatography Detector
`Introduction
`The Dynamic Range of the Detector
`Detector Linearity
`The Determinmfz’orz oftfze Response Ir2a'e.r ofa Detector
`The Linear Dynamic Range of a Detector
`Detector Response
`Detector Noise
`Shore‘ Term Noise
`Long Term Noise
`Drift
`Measuremezzt offlerector Noise
`Detector Sensitivity or the Minimum Detectable Concentration
`The Mass Sensitivity of a Chromatographic System
`The Concentration Sensitivity of a Chromatographic System
`The Maximum Capacity Factor of an Eluted Peak
`Practical Considerations
`Synopsis
`References
`
`47
`4’?
`47
`48
`51
`55
`59
`62
`63
`64
`68
`72
`74
`T?
`80
`
`81
`81
`82
`83
`85
`98
`90
`91
`91
`92
`92
`93
`94
`96
`97
`98
`101
`101
`103
`
`FRESENIUS KABI 1021-OOO4
`
`

`
`Chapter 5 Processing Chromatographic Data
`
`Introduction
`Chromatographic Resolution
`The Efficiency Required to Achieve a Specific Resolution
`Peak Deconvolution
`The Detector Response
`Chromatographic Data Processing
`Manual Data Processing
`Compznfer Data Processing
`Data Processing
`Some Simple Computing Techniques
`Chromatographic Control
`Quantitative Analytical Methods for GC and LC
`Qzrmztittztive Amziysis Using Reference Sfcrizrfands
`Quantitative Analysis by TLC
`Conzparative Spot Assessnzezzt by I/isuai E.s'zz‘maz‘iozz
`Synopsis
`References
`
`PART 2 Quantitative Gas Chromatographic Analysis
`
`Chapter 6 Gas Chromatographic Apparatus for Quantitative
`Analysis
`
`Gas Supplies
`Gas Suppiiesfrom Tam/cs or Cylimfers
`Pure Air G€!Ié‘f‘(Z€Oi‘S
`Pure Nz‘f:'0gez2 Generators
`h{ydroge12 Ge1zeraz‘0rs
`Pressure Com‘ro!Iers
`Ffow Com‘ro!Zer:s
`New Progmxrzmers
`Injection Devices
`Gas Sampiiizg Systems
`Packed Coleman Iiyectiozz Sysierzzs‘
`Open Tz.'Z)zr?ar Cofzmm Injecfioiz Sjystems
`/iutonzafic Injectiozz S3?.S‘l‘€i?iS
`The Column Oven
`The Terrrzpe2‘atm‘e Progranmzez‘
`
`vii
`
`105
`
`1&5
`105
`10’?
`110
`113
`116
`116
`123
`I 31
`135
`139
`139
`14}
`149
`149
`150
`152
`
`155
`
`156
`159
`159
`159
`160
`I60
`161
`162
`170
`1’? 1
`173
`I74
`I’?9
`179
`180
`
`FRESENIUS KABI 1021-OOO5
`
`

`
`viii
`
`Detector Ovens
`Colum:nz’Detector Connecting Conduits
`Gas Chromatography Detectors
`Tire Ficzrme {orzizotion Detector
`The Nitrogen Piiospfzorus Detector‘ (NPD)
`The Electron Capture Detector
`The f<'at:‘zrrron:eIer' Detector
`Synopsis
`References
`
`Chapter 7 Gas Chromatography Applications
`The Extraction and Analysis of Carbonyl Compounds from
`Some Natural Products
`Determination of Insecticides in Drinking Water Using Dual
`Electron Capture and Nitrogen Phosphorus Detection
`Analysis of Chlorobenzenes in Soil by Headspace Solid-Phase
`Extraction and Ion~Trap Mass Spectrornetiy
`Determination of Fentanyl in Whole Blood at the Snbnanogram
`Level Using Nitrogen Sensitive and Mass Spectrornetric
`Detection
`The Measurement of Within—T1‘ee \»"ariation in Lignin
`Components by Pyrolysis Gas Chromatography
`Use of Capillary Columns for Sample Microextraction
`Extraction Technique for the Analysis ofProtective Clothing
`Breakthrough indicator Pads
`Analysis of Pethidine and Methadone in Human Urine by Solid
`Phase Microextraction and Gas Chromatography
`The Determination of Methyl Mercury and Inorganic Me1‘cury
`in Whole Blood by Head Space Ciyofocusing Gas
`Chromatography with Atomic Adsorption Detection
`Determination of}‘rzorgonz’c Mercmj2 in Whole Blood
`Dez‘err22ir:atio:z QfOrga.m'c Mercmy in Whole Blood
`The Determination of l5—Alaninediacetic Acid in Waste Waters
`and Aquatic Environments
`WaterSrmzp1’es
`Sediments
`
`186
`187
`18‘?
`188
`191
`194
`198
`201
`203
`
`205
`
`208
`
`211
`
`213
`
`215
`
`219
`222
`
`224
`
`226
`
`228
`229
`229
`
`239
`231
`231
`
`FRESENIUS KABI 1021-0006
`
`

`
`Method for Analyzing Organoclilorine Pesticides in Water Using
`Solid Phase Microextraction and Pulsed Discharge Electron
`Capture Detection
`Use of Supereritical Fluid Extraction for the {Determination of
`Steroids in Animal Tissue
`Synopsis
`References
`
`PART 3 Quantitative Liquid Chromatography Analysis
`
`Chapter 8 Liquid Chromatographic Apparatus for Quantita-
`tive Analysis
`
`The Basic Liquid Chromatograph
`Tire Sofvem.‘ Suppiy Slvszfem
`Liquid Ciiromczfograpliy Pumps
`Sample Valves
`Column Ovens
`General Comments on Detectors
`Data Acquisition and Processing
`The Modern Versatile Liquid Chromatograph
`Liquid Chromatography Detectors
`The UV Detectors
`The Fzlred Waveiezzgfiz UT/Absorptiozz Detector
`The Variable Wavelength Detector
`The Diode Amity Detector‘
`The Fluorescence Detector
`The Electrical Conductivity Detector
`The Refractive Index Detector
`Synopsis
`References
`
`Chapter 9 Liquid Chromatography Applications
`The Determination of Bixin and Norhixin in Human Blood
`Plasma
`Determination of the Enantiomers of Methamphetamine and its
`Metabolites in Urine
`Trace Enrichment of Alkylthio-s——triazine Herbicides by
`Supported Liquid Membrane Techniques
`
`ix
`
`233
`
`235
`239
`2240
`
`245
`
`245
`246
`248
`253
`255
`255
`256
`256
`261
`262
`262
`266
`268
`272
`277
`281
`284
`286
`
`287
`
`288
`
`292
`
`294
`
`FRESENIUS KABI 1021-OOO7
`
`

`
`The Determination of Trace Amounts of the Transition Metals
`in Parenteral Solutions
`
`Determination of Theophyiline and Caffeine in Blood Serum
`by Direct Injection.
`The Determination of Spectinomycin Residues in Various
`Tissue Types hem Husbandry Animals
`The Determination of Alternarioi in Tomato Paste by Solid
`Phase Extraction and Liquid Chromatography Using
`Fluorescence Detection
`
`Automated Determination of Amphetamine Eiiaiitioniers
`Employing Two—Diniensionai Column Switching
`Determination of N-Methyl Carbamate Pesticides in Foods
`Using Solvent Extraction at Elevated Temperatures with
`Minieolumn Cleanup
`Ascorbic Acid Determination in Foodstuffs by Microdialysis
`Sampling and Liquid Cht‘omatog1*aphy with Electrochemical
`Detection
`
`Synopsis
`References
`
`PART 4 Thin Layer Chromatography
`
`Chapter 10 Thin Layer Chromatography Apparatus
`
`Elution Development in Thin Layer Chromatography
`Thin Layer Cinematography Chambers
`Continuous Plate Development
`Forced-Flow Development
`Sample Application
`Detection Techniques and the Qiiaiititative Evaluation of TLC
`Spots
`The Iodine Reagent‘
`
`The Sm’fm‘ic Acid Spray
`Cfzrosiiic-Sai1fftm'c Acid Spray
`F1tzorescenee
`
`Scanning Densitometry
`Synopsis
`References
`
`297
`
`300
`
`302
`
`306
`
`309
`
`311
`
`314
`316
`318
`
`321
`
`322
`
`324
`
`329
`
`33C)
`
`332
`
`337
`338
`
`339
`
`339
`
`339
`340
`
`345
`347
`
`FRESENIUS KABI 1021-OOO8
`
`

`
`Chapter 11 Thin Layer Chromatography Applications
`
`Measurement Methods
`The Assay of Sulfamethazine in Pork Carcasses
`The Determination of Aflatoxins in Palm Kernels
`Determination of the Platelet-Activating Factor and Other
`Pliosphoiipids in Human Tears
`A Mass Spectometric Measurement of Peptide—Lil<e Materials
`Using a Hybrid Thin Layer Plate as the Transport Interface
`to a Matrix Assisted Laser Desorption Ionization System
`Determination of Caffeine Using Both Densitoinetry Measure-
`ments and an Image Analyzing System
`The Use of Digital Auto—Radiograpliy for the Analysis of
`Biological Samples and for Studying Drug Metabolism
`Purity Measurements of Phthaloyl-Ainioclipine Using Over-
`Pressurized Thin Layer Chromatography
`Synopsis
`References
`
`index
`
`xi
`
`349
`
`349
`352
`354
`
`356
`
`358
`
`362
`
`364
`
`366
`369
`3’?I
`
`373
`
`FRESENIUS KABI 1021-0009
`
`

`
`105
`
`Chapter 5
`
`Processing Chromatographic Data
`
`Introduction
`
`A quantitative analysis can only he successfully carried out if the
`substances of interest are adequately resolved. That is, the peaks must
`be separated sufficiently to allow the unique area or height of each
`peak to he accurately measured or calculated. This is necessary
`irrespective of whether the measurements are to be made manually on
`the chart paper or be calculated from acquired data by a computer. It
`follows that
`the degree of resolution that
`is necessary to achieve
`adequate quantitative accuracy needs to be addressed. In addition,
`despite chromatographic Optimization, some components may still be
`only poorly resolved and, as a consequence, some action must be taken
`to access the peak areas and heights by special techniques. First, the
`conditions necessary to achieve adequate resolution will be considered.
`
`Chromatographic Resolution
`
`The resolution of a pair of closely eluted solutes is usually defined as
`the distance between the peaks measured in units of the average
`standard deviation of the two peaks. In practice, the standard deviation
`(0) of two peaks that are eluted close together are virtually identical.
`However,
`it
`is necessary to decide the degree of separation that
`constitutes adequate resoiution for accurate quantitative analysis. In
`figure 2, five pairs of peaks are shown, separated by 20, 30, 40, 56
`
`FRESENIUS KABI 1021-0010
`
`

`
`106
`
`Quantitative Chromatographic Analysis
`
`and 66, the area of the smaller peak being half that of the larger peak.
`
`However, it should be noted that the relative size of the peaks will
`
`affect the minimum separation that will be suitable for accurate
`quantitative analysis and thus, must be taken into account.
`
`60
`
`5::
`
`45
`
`3:7
`
`2::
`
`Figure 1 Peaks Showing Different Degrees of Resolution
`
`It is clear that a separation of do would be ideal for accurate
`quantitative results. Unfortunately, such resolution will often demand
`very high efficiencies, which may also entail
`long columns and
`consequently, very long analysis times. Furthermore, a separation of
`60 is for more than necessary for accurate quantitative analysis and
`
`this would be true even if manual measurements of peak area and peak
`
`height are employed. In fact, accurate quantitative results can usually
`
`be obtained with a separation of only 46 and, as will be seen later, if
`
`peak heights are employed, then even less resolution may be needed.
`
`Duplicate measurements of peak area or peak height on peaks separated
`
`by 40 should not differ by more than 2%. If the chromatographic data
`
`are acquired and processed by a computer,
`
`then with modern
`
`software, even a separation of filo could be more than adequate. In the
`
`following mathematical argument it will be assumed that a resolution
`
`of 40 is sufficient, but should greater resolution be thought more
`
`appropriate, the equations below can be easily modified and the same
`
`arguments used. Defining the resolution required for accurate
`quantitative analysis as 469 it is now possible to calculate the number
`of theoretical plates necessary for the separation of a specific pair of
`solutes.
`
`FRESENIUS KABI 1021-0011 11
`
`

`
`Processing Chromatographic Data
`
`107
`
`The Efficiency Required to Achieve a Specific Resolution
`
`Consider the two peaks shown in figure 2. The various parameters of
`the chromatogram are labeled according to the expressions derived
`from the Plate Theory. A discussion of the Plate Theory is not
`germane to the subject of this book, but certain relationships will he
`used that are derived from it and so those readers interested in their
`origin are recommended to another book in this series, Introduction
`to Anaigtrzicai Gas C12r'0ma2‘og2'ap!2y [1].
`
`Vria) = n(Vm+K(3)"s}
`<3
`4-«-——— V.-(,g= n(vm+KWvs) --—————+
`
`
`
`—-———>
`
`{---
`
`2
`
`n(vm+K(Mvs)
`
`Figure 2 The Separation of Two Salutes
`
`The difference between the peaks is given by (.o\»’), where
`
`AV : “(Vin 1" KBVs)"' “(Vin + KAVQ : “(KB "' KA)Vv..-
`
`is the column efficiency in theoretical plates,
`where (11)
`is the volume of mobile phase per plate,
`(vm)
`is the volume of stationary phase per plate,
`(V3)
`and (KAMKB) are the distribution coefficients of solutes (A) and
`(B).
`
`FRESENIUS KABI 1021-0012
`
`

`
`108
`
`Quantitative Chrom atographic Analysis
`
`Bearing in mind that the separation ratio (ot) is defined
`
`= M“
`
`then
`
`t-AV 2 t1KA(0t - l)vs
`
`(2)
`
`Now, by definition for adequate separation, av 2 46
`
`or
`
`AV x Marv,“ + KAVS)
`
`(3)
`
`Equating equations (2) and (3)
`
`4«/lit»-m + Km) = nKA(ct — l)vS
`
`Dividing throughout by (um) and remembering that the separation
`K
`ratio 0:’) is given by
`k‘A = -5-if-1
`Vm
`
`Then
`
`4-\/E(1+i(‘A): nk‘A(ot—~1)
`
`Solving for (n),
`
`n =
`
`4 l
`
`k‘
`
`2
`
`kg (Ot -1)
`
`(4)
`
`Equation (4) was first developed by Purnell in 1959 {2} and is one of
`
`the more important equations used in column design and one that can
`be a great aid in column selection for the practicing analyst. in a given
`analysis, if the separation ratio of the ‘critical pair’ is measured (the
`critical pair comprise the two solutes that are eluted closest together),
`
`the column efficiency required (for a separation of 46) can be
`
`calculated from equation (4). If the actual efficiency is found to be
`
`inadequate, the flow rate can he reduced until sufficient efficiency is
`
`attained or until the maximum efficiency is realized at the optimum
`
`velocity.
`
`If, at
`
`the optimum flow rate, even more efficiency is
`
`necessary, then a longer column is required or a column packed with
`smaller particles. The latter is to be preferred, if possible, as the
`increased efficiency will be obtained at reduced analysis times
`(assuming adequate inlet pressure is available). As most modern LC
`
`FRESENIUS KABI 1021-0013
`
`

`
`Processing Chromatographic Data
`
`109
`
`in the case of the GC
`columns are homogeneously packed, oz‘,
`capillary column, homogeneously coated, then for capillary columns
`of the same diameter and columns packed with the same diameter
`particles, the column efficiency can be assumed to increase linearly
`with the column length. Employing a similar argument, an expression
`for the efficiency required from a TLC plate to effect a given
`separation can also be developed. The (Rf) value of a solute is given
`by
`
`R __ Distance Migrated by Solute
`Z S
`f
`Distance Migrated by Solvent - Zm
`
`Now the distance (D) between two solutes (A) and (B) will he
`
`D = 33(3) - Zs(A)
`
`= Zm(Rf{B) ~ Rf(A))
`
`Now if the diameter of a spot is (d) then, [3]
`
`Zsox) : f_ or d 2 42%)
`d
`4
`vii
`
`(5)
`
`(6)
`
`Now the diameter of the spot can be considered to be equivalent to
`(46) and thus to achieve resolution, the spots must be separated by a
`distance equivalent to their diameter, then from equations (5) and (6)
`
`42 (A)
`x/% 2 Zm(Rfto) "’RnA>)
`
`Rearranging, and beating in mind that Rf =2 -33-,
`
`II}
`
`\/I :
`
`4Z‘s(A}
`Zm(Rt«;B> "RttA>)
`
`.-;
`
`4
`I
`(“WA “‘ )
`
`where 0&5/A =2 Rf{B)!Rf(a)
`ratio used in GC and LC.
`
`is the TLC equivalent to the separation
`
`FRESENIUS KABI 1021-0014
`
`

`
`110
`
`Quantitative Chromatographic Analysis
`
`Thus
`
`l6
`n = ——-—--——-3
`(“BIA ‘U
`
`(7)
`
`Equation (7) gives the efficiency required from a TLC plate to
`separate a particular pair of substances employing a specific phase
`system. To increase the plate efficiency the particle size of the
`adsorbent must be reduced, which,
`in most cases,
`is difficult to
`
`the same plate adsorbent will probably be
`In practice,
`change.
`employed, but the phase system will be adjusted to increase the value
`of (man) so that the separation is achieved with the efficiency
`available.
`
`Peak Deconvolution
`
`Unfortunately there will be certain samples, usually complex in
`nature, where adequate resolution is not obtainable and, as a
`consequence, it is necessary to resort to other procedures to determine
`the respective areas of the merged peaks. This situation can arise when
`there is a large number of peaks in the chromatogram and attempts to
`improve the resolution of one pair results in the fusion of another
`pair. One such technique that may be used under such circumstances is
`peak deconvoittrion. Many data acquisition and processing systems
`include software that mathematically analyze convoluted (unresolved)
`peaks,
`identify the individual peaks that make up the composite
`envelope, and then determine the area of the individual peaks. The
`algorithms in the software must embody certain tentative assumptions
`in order to analyze the peak envelope, such as the peaks are Gaussian
`in form and are symmetric or that
`there are explicit equations
`available that describe the specific asymmetry of each peak.
`Furthermore,
`the software must assume that all the peaks can be
`
`described by the sani.e function (r’..=3., the efficiency of all the peaks are
`
`the same), which is also not always true. Nevertheless, providing the
`composite peak is not too complex, deconvolution can, under some
`
`circumstances, be reasonably successful.
`
`However it must be emphasized that the use of deconyoluting software
`has limited capabilities and must be used with considerable caution.
`
`FRESENIUS KABI 1021-0015
`
`

`
`Processing Chromatographic Data
`
`111
`
`Every effort should be made to separate all the components of a
`mixture by chromatography and only employ peak deconvolution as a
`last resort. In general, it should be remembered that,
`
`"Clever algorithms are not an oeeeptnbfe snbstitntefor good
`chromntogrzzpity. "
`
`Having made the point, if resolution is partial, and the components are
`present in equal quantities, then the deconvolution approach can be
`quite successful. An example of the convolution of two partially
`resolved peaks of equal size is shown in figure 3.
`
`1.25
`
`
`
`SoluteConcentration
`
`5:‘ xx '11
`
`1” UI
`
`0.25
`
`0
`
`4.?
`
`__.
`
`4.9
`
`5.1
`
`5.3
`
`5.5
`
`5.7
`
`Volume Flow of Mobile Phase (ml)
`
`The Deconvolution of Two Partially Resolved
`Figure 3
`Peaks Representing Solutes Present in Equal Quantities
`
`It is seen that two Gaussian shaped peaks can be easily extracted from
`the composite envelope and the software could also supply values for
`either the heights of the deeonvoiuteti peaks andfor their areas. It
`should he noted, however, that the two peaks are clearly discernible
`and the deconvoluting software can easily identify the approximate
`positions of the peak maxima and assess the peak widths. Such
`information allows the software to arrive at a valid analysis quickly
`
`FRESENIUS KABI 1021-0016
`
`

`
`112
`
`Quantitative Chromatographic Analysis
`
`and with reasonable accuracy. Nevertheless, most software would
`construct a perpendicular from the valley to the baseline,
`thus
`bisecting the combined peak envelope. The area of each half would
`then be taken as the respective area of each peak. If the peaks are of
`significantly different size the problem becomes more difficult but,
`nevertheless, solvable. An example of the analysis of a composite
`envelope for two peaks of different size is shown in figure 4. It is
`clear that the resolution will not permit peak skimming (a technique
`that will be discussed in detail later) with any hope of a reasonable
`degree of accuracy. It seen that the position of the peak maximum,
`and the peak width, of the major component is easily identifiable.
`
`1.25
`
`0.75
`
`0.5
`
`0.25
`
`
`
`SoluteConcentration
`
`0.‘_
`4.‘?
`
`4.9
`
`5.1
`
`5.3
`
`5.5
`
`5.?
`
`Volume Flow of Mobile Phase (ml)
`
`The Deconvolution of Two Partially Resolved
`Figure 4
`Peaks Representing Salutes Present in Unequal Quantities
`
`The software can accurately determine the Gaussian function of the
`major component and the reconstructed profile of the major
`component would then be subtracted from the total composite peak
`leaving the small peak as difference value. However, in the examples
`given, the solutes were at least partly resolved. Unfortunately, as the
`resolution becomes less and less, and the need for an accurate
`
`FRESENIUS KABI 1021-0017 to
`
`

`
`Processing Chromatographic Data
`
`113
`
`the value of the
`deconvolution technique becomes even greater,
`software presently available appears to become minimal. A typical
`example at the other extreme, where a deconvolution technique would
`be useless, is given in figure 5. It is seen that however sophisticated
`the software might be, it would be virtually impossible to deeonvolute
`the peak into the three components. The peaks shown in the diagram
`are discernible because the peaks themselves were assumed and the
`composite envelope calculated.
`
`1.5
`
` Solute
`Concentration
`
`0
`4.7
`4.9
`5.1
`5.3
`5.5
`5.’?
`
`13'” UI
`
`Volume Flow of Mobile Phase (ml)
`
`Figure 5 The Deconvolution of Three Unresolved Peaks
`
`The envelope, however, would provide no basic information; there is
`no hint of an approximate position for any peak maximum and
`absolutely no indication of the peak width of any of the components or
`even how many peaks are present.
`
`The Detector Response
`
`One further form of peak deconvolution needs to he mentioned and
`that is the practical use of the detector response index. The detector
`response index has been presented as a means of defining linearity,
`but, as mentioned in an earlier chapter, it can also be used to take into
`
`FRESENIUS KABI 1021-0018
`
`

`
`114
`
`Quantitative Chromatographic Analysis
`
`account any non—linearity that is present and appropriately modify the
`peak height or peak area calculations and thus improve quantitative
`accuracy. An example of two peaks constructed from Gaussian
`functions using response factors of 0.95 and L05 are shown in figure
`6. Such values would be considered outside these which would be
`acceptable for a detector to be defined as linear.
`
`
`
`Two Identical Gaussian Curves Drawn with
`Figure 6
`Response Indices of 0.95 and 1.05
`
`It is seen that there is a clear difference between the peaks, and the
`error involved is quite considerable, as discussed in detail in Chapter
`4. Thus,
`in order to obtain a value that truly reflects the solute
`concentration being measured, the detector output must not only be
`corrected for the linearity constant, but also the numerical value of the
`response factor, This is not a difficult calculation and should be
`automatically incorporated into any data processing software that is
`not dealing with the ideal detector. The equation given for detector
`I‘€Sp()IlS€
`in Chapter 4 is
`
`V =: Aer
`
`Where (V)
`(P3)
`
`The voltage output from the detector,
`is the linearity constant,
`
`FRESENIUS KABI 1021-0019
`
`

`
`Processing Chromatographic Data
`
`1 15
`
`is the numerical value of the response inclex.
`and (1')
`Consequently the output
`that would truly represent
`concentration would be
`
`the solute
`
`c = I‘-;:-
`
`(8)
`
`or
`
`C : ARV? where A’ = #-
`x/Lit‘
`
`If (r) is unity, then the expression reduces to the simple relationship
`that is normally used in quantitative analysis,
`
`c:-
`
`(9)
`
`Unfortunately, (r) is not usually unity and may differ from unity
`sufficiently to require equation (8) to be used as opposed to equation
`(9) in order to achieve the necessary accuracy. Equation (8) can be
`employed to calculate the effect of the response index on the level of
`the component as calculated assuming true linearity. The results are
`shown in table i. It is seen that depending on the relative peak heights
`of the standard and the solute of interest the result may be greater or
`less than the true values.
`
`Table 1 Effect of Response Index on Apparent Component
`Level
`
`8835
`
`89.18
`
`89.60
`
`78.64
`68.92
`
`79.10
`
`69.28
`
`90.76 9l.l3 9l.47
`
`80.87
`
`81.30
`
`81.72
`
`FRESENIUS KABI 1021-0020
`
`

`
`116
`
`Quantitative Chromatographic Analysis
`
`The data in table 1 can also be presented in the form of % error and
`these are shown in table 2.
`
`Table 2 Effect of Response Index on Component Error
`
`% Error of Comonent Assumin Linearit
`r:
`1’:
`1‘:
`{'2
`
`0.92 0.94 0.96 0.98
`
`It is seen that absolute errors of l.7% at the 30% level can be realized
`
`and in relative terms this would he a 17% error. It is clear that if the
`
`response index is outside the range 0.98.< r <l.02, then appropriate
`
`corrections should be applied to the results.
`
`Chromatographic Data Processing
`
`Chromatographic data can be processed either manually or by
`computer and, although virtually all modern chrornatographs are
`provided with data acquisition and processing,
`there are still a
`considerable number of instruments in use that require manual
`
`processing. Consequently, both computer and manual processing will
`he discussed. As the iogic of the computer software is based on manual
`measurement techniques, the manual procedure will be discussed first.
`
`Manual Data Processing
`
`The information required from the chromatogram is basically an
`accurate measure of, if the retention time, 2! the peak height, 3/ the
`
`peak width. The retention time is used for peak identification and the
`
`FRESENIUS KABI 1021-0021
`
`

`
`Processing Chromatographic Data
`
`117
`
`peak height and peak width for quantitative evaluation. In fact, it is the
`peak area that is often required, but manual measurement of peak
`area, although possible,
`is difficult and a compromise is usually
`reached by taking the peak area as proportional to the product of the
`peak height and the peak width measured at the points of inflexion.
`This assumption is theoretically sound provided the peak is Gaussian
`or near Gaussian in shape. A typical peak on which measurements
`might be made is shown in figure 7. Measurements, other than the
`peak width, are normally made directly on the chart using a good
`quality steel rule and should be estimated to the nearest 0.2 mm (if
`necessary using a lens or loupe). If capacity ratios ((k‘) values) are
`used for identification purposes, then the chart speed or flow rate
`should be adjusted so that the distance between the injection point and
`the dead point is at least 2 cm. This is necessary to ensure that the
`measurement error will be no greater than 1% (tie. 0.02/2 x 100).
`
`Corrected
`Dead
`
`Time Retention Time
`
`Retention Time
`
`Peak
`at Points
`of Infiexion
`
`0.6065h
`
`Pgak
`Height
`(h)
`
`
`
`
`
`
`Base Line
`
`Injection Dead
`Point
`point
`
`
`Phak
`Maximum
`
`,3‘igure 7 Peak Measurements
`
`The most precise measurements must be made on the peak width and,
`as the peak width is frequently only a fraction of a centimeter wide,
`
`FRESENIUS KABI 1021-0022
`
`

`
`118
`
`Quantitative Chromatographic Analysts
`
`such measurements can not he made sufficiently accurately by means
`of steel rule. In some cases measurements can be made more precise
`by increasing the chart speed.
`
`However, when capillary columns are employed the peak widths can Q
`be so narrow that even at high chart speed the peaks are still less than
`a centimeter in width. Under such circumstances it is advisable to use
`a comparator with a gratieule that is calibrated in at least 0.1 mm.
`Comparators are simple to use and give accurate results, but care must
`be taken over the measurement procedure, A diagram depicting the
`measurement of peak width is shown in figure 8.
`
`
`
`Figure 8 Measurement of Peak Width
`
`The measurements should be taken from the inside of one line to
`the outside of the adjacent line, in order to eliminate errors resulting
`from the finite width of the ink line drawn on the chart. The
`measurement should be repeated using the alterhate edges of the line
`and an average taken of the two readings to avoid errors arising from
`any variation in line thickness. It is recommended that three replicate
`runs should be made and the three values should not differ by more
`than 3%. Most data acquisition systems have software that will also
`measure peak widths but this will be discussed later.
`
`FRESENIUS KABI 1021-0023
`
`

`
`Processing Chromatographic Data
`
`119
`
`Peak Area Meastrrernemfs
`
`The area of a peak is the integration of the mass per unit volume
`(concentration) of solute eluted from the column with respect to time.
`If the flow rate is constant, and the detector is a concentration sensitive
`device,
`the integration will also be with respect to volume flow of
`mobile phase through the column. It follows that the area of the peak
`is proportional to the total mass of solute. Measurement of peak area
`accommodates peak asymmetry and even peak tailing without
`compromising the simple relationship between peak area and mass.
`Consequently, peak area measurements give more accurate results
`under conditions where the chromatography is not ideal,
`the peak
`profiles are not truly Gaussian or Poisson, or described by some other
`specific function.
`
`Unfortunately, whether a chart recorder is used and measurements are
`made directly on the chart, or computer data acquisition and processing
`are available, both systems are based on time as the variable and not
`
`volume of mobile phase flowing through the column. it follows that
`when using peak area measurements, a high quality flow controller (in
`the case of GC) or a high quality high pressure pump (in the case of
`LC) must he used to ensure that all time measurements are linearly
`related to volume flow of mobile phase. It is interesting to note that
`the flame ionization detector (FIB) (the most commonly used detector
`for quantitative analysis in GC) responds to mass of solute entering it
`per unit time and thus the peak area integrated with respect to time is
`truly related to solute mass. This highly desirable property makes the
`accuracy of the quantitative analysis relatively indifferent to small
`changes in flow rate. Unfortunately, there is no equivalent detector in
`LC or, at least, not one that is commercially available.
`
`the
`Peak areas can be measured manually in a number of ways,
`simplest being the product of the peak height and the peak width at
`0.6065 of the peak height (26). This does not give the true peak area,
`but providing the peak is Poisson, Gaussian or close to Gaussian it will
`always give accurately the same proportion of the peak area. One of
`the older methods of measuring peak area was to take the product of
`
`FRESENIUS KABI 1021-0024
`
`

`
`120
`
`Quantitative Chromatographic Analysis
`
`the peak height and the peak width at the base (40). The peak width at
`
`the base is the distance between the points of intersection of the
`
`tangents drawn to the sides of the peak with the baseline produced
`
`beneath the peak. This technique has the disadvantage that the tangents
`to the elution curve must also be constructed manually, which
`introduces another source of error into the measurement. Another
`
`method involves the use of a planirneter (an instrument that provides a
`
`numerical value for the area contained within a perimeter traced out
`
`by a stylus). This is a very tedious method of measuring peak area and
`
`(partly for the same reason) is also not very accurate. The most
`
`accurate manual method of measuring peak area (which, unfortunately
`is also a little tedious) is to