GAS Chromatography: Chrom-Ed Book Series

1 1
1
2
3
4
5 Chrom-Ed Book Series

6
7 Raymond P. W. Scott
8
9 GAS
10
11 CHROMATOGRAPHY
12
2This eBook is protected by Copyright law and International Treaties and may be reproduced for proofreading purposes only.
3
4
5 2
5COPYRIGHT @2003 by LIBRARY4SCIENCE, LLC

6ALL RIGHTS RESERVED
7
8
9Neither this book or any part may be reduced or transmitted in any
10form or by any means, electronic or mechanical , including
11photocopying, microfilming, and recording or by any information
12storage and retrieved system without permission in writing from the
13publisher except as permitted by the in-user license agreement.
14
15
16World Wide Web
17http://www.library4science.com/
18
19
6This eBook is protected by Copyright law and International Treaties. All rights are reserved. This book is covered by an End User Licensee
7Agreement (EULA). The full EULA may be seen at http://www.library4science.com/eula.html.
8
9 3
1
2
3
4 Contents
5
6Introduction ............................................................................................1
7 Supplies from Gas Tanks....................................................................6
8 Pure Air Generators.........................................................................6
9 Hydrogen Generators......................................................................7
10 Pressure Controllers............................................................................7
11 Flow Controllers..................................................................................8
12 Flow Programmers............................................................................10
13 Injection Devices...............................................................................14
14 Packed Column Injectors..............................................................15
15 Open Tubular Column Injection Systems.....................................17
16 Retention Gap Sampling...............................................................19
17 Sampling by Solute Focusing........................................................20
18GC Columns .........................................................................................22
19 The Packed GC Column....................................................................22
20 Adsorbents.....................................................................................23
21 Supports for GLC .........................................................................24
22Column Packing....................................................................................28
23 The Capillary or Open Tubular Column...........................................32
24 Static Coating................................................................................35
25 Open Tubular Column Types........................................................38
26 Chiral Stationary Phases...................................................................40
27 The Column Oven and Temperature Programmer............................43
28GC Detectors.........................................................................................43
29 The Flame Ionization Detector..........................................................44
30 The Nitrogen Phosphorus Detector (NPD).......................................46
31 The Electron Capture Detector..........................................................49
32 The Katherometer Detector...............................................................53
33Data Acquisition and Processing..........................................................56
34 The Scaling Ampifier........................................................................57
35 Data Processing.................................................................................60
36Quantitative Analysis............................................................................61
37 Derivatization....................................................................................64
38 Acylation Reactions......................................................................68
11
12
13 4
1Preparative Gas Chromatography.........................................................69

2 The Moving Bed Continuous Chromatography System...................71
3 Lime Oil........................................................................................79
4 The Head space Analysis of Tobacco...........................................80
5 Food and Beverage Products.........................................................83
6References.............................................................................................88
7
8
9
16
17 1
1
2
3Introduction
4
5Chromatography, in one of its several forms, is the most commonly
6used procedure in contemporary chemical analysis and the first
7configuration of chromatography equipment to be produced in a single
8composite unit and made commercially available was the gas
9chromatograph. Gas chromatography was invented by A. J. P. Martin
10who, with R. L. M. Synge, suggested its possibility in a paper on liquid
11chromatography published in 1941 (1). Martin and Synge
12recommended that the liquid mobile phase used in liquid
13chromatography could be replaced by a suitable gas. The basis for this
14recommendation was that, due to much higher diffusivities of solutes in
15gases compared with liquids, the equilibrium processes involved in a
16chromatographic process (see Book 1) would be much faster and thus,
17the columns much more efficient and separation times much shorter. So
18the concept of gas chromatography was envisioned more than fifty
19years ago, but unfortunately, little notice was taken of the suggestion
20and it was left to Martin himself and his coworker A. T. James to bring
21the concept to practical reality some years later in 1951, when they
22published their epic paper describing the first gas chromatograph (2).
23
24The first published gas chromatographic separation was that of a series
25of fatty acids, a titration procedure being used, in conjunction with a
26micro burette, as the detector. The micro burette was eventually
27automated providing a very effective in-line detector with an integral
28response. After its introduction by James and Martin, the technique of
29GC developed at a phenomenal rate, growing from a simple research
30novelty to a highly sophisticated instrument, having a multi-million
31dollar market, in only 4 years. The gas chromatograph was also one of
32the first analytical instruments to be associated with a computer which
33controlled the analysis, processed the data and reported the results.
34
19
20
21 2
1A more sophisticated form of the gas chromatograph was constructed

2by James and Martin and described by James in 1955 (3). The
3instrument was a somewhat bulky device with a straight packed
4column, 3 ft long, that was held vertically and thermostatted in a vapor
5jacket. Initially, the detector was situated at the base of the column and
6consisted of the automatic titrating device, the separation was
7presented as a chromatogram in the form of a series of steps, the height
8of each step being proportional to the mass of solute eluted. The
9apparatus was successfully used to separate some fatty acids, but the
10limited capability of the device to sense only ionic material motivated
11Martin to develop a more versatile detector, the Gas Density Balance.
12
13The gas density balance, was the first detector with a truly catholic
14response that was linearly related to the vapor density of the solute and
15consequently its molecular weight. The gas density balance had a
16maximum sensitivity (minimum detectable concentration) of about 10-
176 g/ml at a signal to noise ratio of two. This detector inspired the
18invention of a wide range of detectors over the next decade providing
19both higher sensitivity and selective response.
20
21The modern gas chromatograph is a fairly complex instrument mostly
22computer controlled. The samples are mechanically injected, the
23analytical results are automatically calculated and the results printed
24out, together with the pertinent operating conditions in a standard
25format. However, the instrument has evolved over many years although
26the majority of the added devices and techniques were suggested or
27describe in the first three international symposia on gas
28chromatography held in 1956, 1958 and 1960.
29
30These symposia, initially organized by the 'British Institute of
31Petroleum' have been held every two years ever since 1956 and the
32meetings have remained the major stimulus for developing the
33technique and extending its capabilities. However, the majority of the
34techniques and devices that have been incorporated in the modern
35chromatograph, were described, reported, or discussed in the first triad
24
25 3
1of symposia. The layout of the modern gas chromatograph is shown as

2a block diagram in figure 1.
G a s S u p p ly U n it
F lo w C o n tr o lle r M icro p ro ce s s o r fo r F lo w
C o n tro lle r a n d
P rogram m e r
F lo w P r o g r a m m e r
S a m p lin g U n it
I n je c t o r I n je c t o r a n d I n je c t o r
(M a n u a l o r A u to m a tic) O v e n C o n tr o lle r
I n je c t o r
O ven
C o lu m n U n it
C o lu m n O v e n
C o lu m n C o n t r o lle r a n d
P rogram m e r
C o lu m n
O ven
D e te c to r E le c tr o n ic s
D e te c to r U n it a n d C o m pu te r D a ta
A c q u is it io n a n d
D e te cto r P r o c e s s in g S y s te m
D e te cto r O v e n
D e te cto r O v e n
C o n tr o lle r
3
4
5Figure 1 The Design of a Modern Gas Chromatograph
6The Modern Gas Chromatograph
7
8Most gas chromatographs consist of four chromatography units,
9supported by three temperature controllers and 2 micro processors
10systems. In some instruments, a single microprocessor unit is employed
11to service the entire chromatograph but this tends to restrict the choice
27
28
29 4
1available for the different parts of the chromatograph. The first unit, the
2gas supply unit, provides all the necessary gas supplies which may
3involve a number of different gases, depending on the type of detector
4that is chosen. For example, a flame ionization detector will require
5hydrogen or some other combustible gas mixture, air or oxygen to
6support combustion and a mobile phase supply that could be nitrogen,
7helium or some other appropriately inert gas. Thus, for the detector
8postulated, a minimum of three different gases would be required
9which will also involve the use of three flow controllers, three flow
10monitors and possibly a flow programmer. In addition the gas supply
11unit would be serviced by a microprocessor to monitor flow rates,
12adjust individual gas flows and, when and if necessary, program the
13mobile phase flow rate.
14
15The second unit is the sampling unit which contains an automatic
16injector which is situated inside a thermostatically controlled
17enclosure. The injector usually has its own oven, but sometimes shares
18the column oven for temperature control. The injector oven, if separate
19from the column oven, is serviced by its own temperature controller
20which both monitors and controls the temperature. There is normally a
21separate controller, usually a microprocessor, that controls the injector
22itself. The injector can range in complexity from a simple sample
23valve, or mechanically actuated syringe to an automatic multi sampler
24that is microprocessor controlled. It can have a complex transport
25system (such as a carousel) that can take samples, wash containers,
26prepare derivatives and, if necessary, carry out a very complex series of
27sample preparation procedures before injecting the sample onto the
28column. Sample preparation is often carried out using a laboratory
29robot which then becomes part of the sampling unit. If a robot is used it
30can be programmed to prepare a wide variety of different samples and
31so software must be written for each type of sample.
32
33The third unit is the column unit which contains the column, the
34essential device that actually achieves the necessary separation, and an
35oven to control the column temperature. It is interesting to note that
36despite the complexity of the apparatus, and its impressive appearance,
32
33 5
1the actual separation is achieved either in a relatively short length of

2packed tube or a simple wall-coated open tube. The rest of the
3apparatus is merely there to support this relatively trivial, but critical
4device. The oven also will contain a temperature sensor and if
5necessary an appropriate temperature programmer. As the mobile phase
6is a gas, there are virtually no interactions between the sample
7components and the mobile phase and thus the elution time can not be
8controlled by techniques such as solvent programming or gradient
9elution. The counterpart to gradient elution in gas chromatography is
10temperature programming. The column temperature is raised
11continuously during development to elute the more retained peaks in a
12reasonable time. It is a similar technique to flow programming but
13decreases the retention exponentially with temperature as opposed to
14linearly with flow rate. The temperature was originally programmed in
15a linear manner using electro-mechanical devices but modern
16temperature programmers contain a dedicated micro processor for the
17purpose. Sometimes all controls are initiated from a central computer
18that is also employed for acquiring and processing the chromatographic
19data.
20
21The fourth unit contains the detector which is situated in its own oven.
22There is a wide range of detectors available each having unique
23operating parameters and its own performance characteristics. The
24detector, and the conduit connecting the column to the detector, must
25be maintained at a temperature at least 15˚C above that of the
26maximum temperature the oven will reach during analysis to ensure no
27sample condenses in the conduits or detector, consequently, separate
28conduit heaters are necessary. Any condensation introduces serious
29detector noise into the system and also reduces the detector response
30thus effecting both the detector sensitivity and the accuracy and
31precision of the results. The detector oven is set at a user defined
32temperature and is operated isothermally, controlled by its own
33detector-oven temperature controller. The output from the detector is
34usually electronically modified and then acquired by the data
35
36
37 6
1processing computer which processes the data and prints out an

2appropriate report.
3
4Gas Supplies
5
6Gases for use with the gas chromatograph were originally all obtained
7from gas tanks or gas cylinders. However, over the past decade the use
8of gas generators have become more popular as it avoids having gases
9at high pressure in the laboratory which is perceived by some as
10potentially dangerous. In addition, the use of a hydrogen generator
11avoids the use of a cylinder of hydrogen at high pressure which is also
12perceived by some as a serious fire hazard despite the fact that they
13have been used in laboratories, quite safely for nearly a century.
14
15Supplies from Gas Tanks
16
17Gasses are stored in large cylindrical tanks fitted with reducing valves
18that are set to supply the gas to the instrument at the recommended
19pressure defined by the manufacturers. The cylinders are often situated
20outside and away from the chromatograph for safety purposes and the
21gasses are passed to the chromatograph through copper or stainless
22steel conduits at relatively low pressure. The main disadvantage of gas
23tanks is their size and weight which makes them difficult to move and
24replace.
25
26 Pure Air Generators
27
28Air generators require an air supply from air tanks or directly from the
29laboratory compressed air supply. The Packard Zero Air Generator
30passes the gas through a 0.5 µ filter to remove oil and water and finally
31over a catalyst to remove hydrocarbons. The hydrocarbon free air is
32then passed through a 0.01 µ cellulose fiber filter to remove any
33residual particulate matter that may be present. The manufacturers
34claim the resulting air supply contains less than 0.1 ppm total
35hydrocarbons and delivers air at 125 psi at flow rates up to 2,500 cc per
36min.
37
38Pure Nitrogen Generators
40
41 7
1
2The nitrogen generator can also operate directly from the laboratory
3compressed air supply. General contaminants are first removed with
4appropriate filters and adsorbents and the purified air passes over
5layers of polymeric hollow fiber membranes through which nitrogen
6selectively permeates. The residual nitrogen-depleted air containing
7about 30% oxygen is vented to atmosphere. The nitrogen produced by
8the Air Products nitrogen generator contains less than 0.5 ppm of
9oxygen, less than 0.5 ppm of water vapor and less than 2.0 ppb of
10halocarbons or hydrocarbons. It can supply up to 1 l/min. at pressures
11from 60 to 100 psi.
12
13 Hydrogen Generators
14
15In the Packard Hydrogen Generator, hydrogen is generated
16electrolytically from pure deionized water. Unfortunately, the
17technology used in hydrogen generators is largely proprietary and
18technical details are not readily available. The electrolysis unit uses a
19solid polymer electrolyte and thus does not need to be supplied with
20electrolytes, only the deionized water. The manufacturers claim the
21device generates 99.999% pure hydrogen with a reservoir capacity of 4
22liter, and an output pressure that ranges from 2 to 100 psi. Other units
23can produce hydrogen flows that range from 0 to 125 ml/min. to 0 to
241200 ml/min. The oxygen, produced simultaneously with hydrogen at
25half the flow rate, is vented to air.
26
27Pressure Controllers
28
29The first control on any gas line is afforded by a simple pressure
30controller. There are a number of pressure controllers associated with a
31gas chromatograph. The reducing valves on the gas tanks are examples
32of simple pressure controllers and the flow controllers that are used for
33detector and column flow control often involve devices based on the
34same principles. A diagram of a pressure controller is shown in figure
352.
36
37
43
44
45 8
1
2
3The pressure controller consists essentially of two chambers separated
4by a diaphragm, in the center of which is a needle valve that is actuated
5by the diaphragm. The diaphragm is held down by a spring that is
6adjustable so that the pressure in the second chamber, and thus the
7outlet flow, can be set at any chosen value. When gas enters the lower
8chamber, the pressure on the lower part of the diaphragm acts against
9the spring setting, and opens the valve. Gas then passes into the upper
10chamber and pressure is built up in the upper chamber to the value that
11has been set at which time the diaphragm moves downward closing the
12valve. If the pressure falls in the upper cylinder, the diaphragm again
13moves upward due to the pressure in the lower chamber, which opens
14the valve and the pressure in the upper chamber is brought back to its
15set value.
N e e dle V a lve
D ia ph ra g m
P re ssu re
A d ju s tm e n t
P2
P1
G as G as
In le t O u tle t
16
17
18Figure 2 The Pressure Controller
19
20Flow Controllers
48
49 9
1
2A constant pressure applied to a column does not ensure a constant
3flow of mobile phase though the chromatographic system, particularly
4if the column is being temperature programmed. Raising the
5temperature of a gas causes the viscosity to increase, and at a constant
6inlet pressure, the flow rate will fall. The reduction in flow rate will be
7related to the temperature program limits and to a certain extent on the
8temperature gradient. To obviate the flow rate change, mass controllers
9are used which ensure a constant mass of mobile passes through the
10column in unit time irrespective of the system temperature. A diagram
11of a mass flow controller is shown in figure 3.
A m p lifie r T o S o le n o id
O p e ra te d C o n tro l
D e te cto r V a lve
V o lta g e H e a te r
R e g u la to r
S e con dary T e m p e ra tu re
F lo w S e n sors
M a in
F low
L a m in a F lo w
12 E le m e n t
13
14 Courtesy of Porter Instrumentation Company Inc.
15
16Figure 3 The Mass Flow Controller
17
18The sensing system consists of a bypass tube with a heater situated at
19the center. Precision temperature sensors are placed equidistant up
20stream and down stream of the heater. A proprietary set of baffles
21situated in the main conduit creates a pressure drop that causes a fixed
22proportion of the flow to be diverted through the sensor tube. At zero
23flow rate both sensors are at the same temperature. At a finite flow rate,
51
52
53 10
1the down stream sensor is heated, producing a differential temperature

2across the sensors. The temperature of the gas will be proportional to
3the product of mass flowing and its specific heat and so the differential
4temperature that will be proportional to the mass flow rate. The
5differential voltage from the two sensors is compared to a set voltage
6and the difference used to generate a signal that actuates a valve
7controlling the flow. Thus, a closed loop control system is formed that
8maintains the mass flow rate set by the reference voltage. The device
9can be made extremely compact, is highly reliable and affords accurate
10control of the carrier gas flow rate irrespective of gas viscosity changes
11due to temperature programming.
12
13Flow Programmers
14
15Flow programming is a procedure where the mobile phase flow-rate is
16increased during chromatographic development. If the mobile phase is
17compressible the relationship between retention volume, flow rate and
18inlet pressure is given by,
19
20 (1)
21
Where (Vr) is the true retention volume of the solute,
(Vr(0)) is the retention volume measured at the outlet.
and (γ) is the inlet/outlet pressure ratio
22
23 (for the derivation of this equation see Book 8, The Thermodynamics
24of Chromatography)
25
26Thus,
27
28Now, from Book where (ε) is a constant
29
30 Thus, (2)
31If (γ) is large compared with unity, Then
32
33 (3)
34
56
57 11
1It is seen that at high values of (γ), the retention time approaches a
2constant value.
3The relationship between and (γ) is depicted in figure 4.
4 Figure 4 shows that there is little advantage in employing inlet/outlet
5pressure ratios much above 5 as values in excess of this do not reduce
6elution time significantly. If the column is very long, and consequently
7has a high flow impedance, higher inlet pressures may be necessary to
8obtain the optimum flow rate but this may not significantly reduce the
9elution time.
10
7 .5 0
5 .0 0
γ2 + γ + 1
2 .5 0
0 .0 0
0 2 4 6 8 10
In l e t e /O u t l e t P r e s s u r e R a t i o ( γ)
11
12
13Figure 4 Graph of against (γ)
14
59
60
61 12
1In figure 5, the log of the retention time is plotted against (γ) for both
2compressible and incompressible mobile phases. It is seen that for a
3compressible mobile phase the retention time falls to a constant level
4when (γ) is about 5 or 6. In contrast, for an incompressible mobile
5phase (i.e. in liquid chromatography), the retention time is continuously
6reduced as (γ) is increased. The advantages of flow programming with
7a compressible mobile phases are much less than for incompressible
8mobile phases. It should be noted, however, that the effect of
9increasing the flow rate above the optimum will progressively
10denigrate the column efficiency, whether the mobile phase is a liquid or
11a gas.
12
1 .5 0
M o b ile P h a s e G a s (C o m p re ss ib le )
1 .0 0
0 .5 0
M o b ile P h a s e L iq u id ( I n c o m p r e s s ib le )
0 .0 0
( 0 .5 0 )
0 .0 0 5 .0 0 1 0 .0 0 1 5 .0 0 2 0 .0 0
In l e t /O u t l e t P r e s s u r e R a t i o ( γ)
13
14
15Figure 5 Graphs of Log Retention Time against Inlet/Outlet
16Pressure Ratio for Compressible and Incompressible Mobile
17Phases
64
65 13
1
2The compressibility of the mobile phase in GC has interesting
3implications for the use of pressure or flow programming. The net
4effect of pressure programming on elution time can be evaluated as
5follows.
6Rearranging equation (2), (4)
7Therefore, during a range of values for (γ) that occur during a pressure
8or flow program, over a time interval (∆t), the contribution of the
9column flow to the retention volume (∆Vr(0)) will be given by,
10
11 (5)
12
13Taking (∆t) as unit time (1 second)
14
15Then (6)
16
17and n = Tt(0), where (Tt(0)) will be the retention time of the solute under
18the defined pressure programming conditions.
19
20Taking a simple practical situation where the retention volume is 1000
21ml on a given column operating at a (γ) value of 2 and the retention
22time of the solute is 10 minutes (600 seconds). The flow properties of
23the column i.e. can be defined in the following manner,
24
25 From equation (3)
26or . Thus,
27
28Substituting for in (6), (7)
29
30Equation (7) can now be employed to calculate the change in retention

31time for a series of solutes separated under pressure programming
32conditions in the defined column
33
34
35Consider five solutes having actual retention volumes of 150, 300, 600,
36900 and 1200 ml eluted under pressure programming conditions where,
37(γ1) is 1.2 and at (p) seconds after the start, γp = γ1 + pa, where (a)
67
68
69 14
1takes values that range from 0.0025/s (0.0375 psi/s) and 0.025/s (0.15
2psi/s) . Employing equation (7) the retention time of the solutes can be
3calculated for the series of different programming rates.
4
800
600
R e t e n t io n T im e ( s e c o n d s )
Vr =1200 m l
400 Vr =900 m l
Vr =600 m l
200 Vr =300 m l
Vr =150 m l
0
0 0 .0 0 5 0 .0 1 0 .0 1 5 0 .0 2 0 .0 2 5
0 .0 7 5 p s i /s 0 .2 2 5 p s i /s .3 7 5 p s i /s
P r o g r a m G r a d i e n t ( γ ) /s
5
6Figure 6 Graph of Retention Time against Pressure Program Rate
7for a Series of Solutes.
8
9The results are shown in figure 6. It is seen that. although the use of
10pressure programming does indeed reduce the retention time of all
11solutes, program rates much above 0.2255 psi/s (13.5 psi/min.)
12provides very little advantage as far as reduction of analysis time is
13concerned.
14
15Injection Devices
16
72
73 15
1The basic injection devices that are used in chromatography, such as

2the external loop valve, have been discussed in book 1. In gas
3chromatography two basic types of sampling system are used, those
4suitable for packed columns and those designed for open tubular
5columns. In addition, different sample injectors are necessary that will
6be appropriate for alternative column configurations. It must be
7stressed, however, that irrespective of the design of the associated
8equipment, the precision and accuracy of a GC analysis will only be as
9good as that provided by the sample injector. The sample injector is a
10very critical part of the chromatographic equipment and needs to be
11well designed and well maintained.
12
13 Packed Column Injectors
14
15In general, the sample injected onto a packed GC column ranges in
16volume from 0.5 µl to 5 µl and usually contains the materials of interest
17at concentrations ranging from 5%v/v to 10%w/v. The sample is
18injected by a hypodermic syringe, through a silicone rubber septum
19directly into the column packing or into a flash heater. Although the
20latter tends to produce broader peaks it also disperses the sample
21radially across the column.
75
76
77 16
S u p p o rtin g M e ta l
D is cs w ith G u id e
H o le s
S y rin g e
C a rrie r
G as
S ilico n e
S e p tu m
H e a te d G la s s O ve n W a ll o r
L in e r O ve n T o p
P ack e d
C o lu m n
1
2
3Figure 7 A Packed Column Injector
4
5Direct injection into the packing constrains the sample into a small
6volume, but can cool the front of the packing. An example of a septum
7injection system used for packed columns is shown in figure 7. The
8silicone septum is compressed between metal surfaces in such a
9manner that a hypodermic needle can pierce it, but when it is
10withdrawn the hole is closed as a result of the septum compression and
11there is no gas leak. The glass liner prevents the sample coming in
12contact with the heated metal wall and thus, reduces the chance of
13thermal decomposition.The glass liner can be fitted with a separate
14heater and the volatalization temperature can, thus, be controlled. This
15"flash heater" system is available in most chromatographs. By using a
16syringe with a long needle, the tip can be made to penetrate past the
80
81 17
1liner and discharge its contents directly into the column packing. This
2procedure is called 'on-column injection' and, as it reduces peak
3dispersion on injection and thus, provides higher column efficiencies,
4is often the preferred procedure.
5
6 Open Tubular Column Injection Systems
7
8Due to the very small sample size that must be placed on narrow bore
9capillary columns, a split injection system is necessary, a diagram of
10which is shown in figure 8.
11
12The basic difference between the two types of injection systems is that
13the capillary column now projects into the glass liner and a portion of
14the carrier gas sweeps past the column inlet to waste. As the sample
15passes the column opening, a small fraction is split off and flows
16directly into the capillary column, ipso facto this device is called a split
17injector. The split ratio is changed by regulating the portion of the
18carrier gas that flows to waste which is achieved by an adjustable flow
19resistance in the waste flow line. This device is only used for small
20diameter capillary columns where the charge size is critical.
21
22The device has certain disadvantages due to component differentiation
23and the sample placed on the column may not be truly representative.
24The solutes with the higher diffusivities (low molecular weight) are
25lost preferentially to those with lower diffusivities (higher molecular
26weights). Consequently, quantitative analyses carried out using the
27high efficiency small diameter capillary columns may have limited
28accuracy and precision, depending on the nature of the sample.
29
83
84
85 18
H o le s S y rin g e
O ve n W a ll o r
C a rrie r S ilico n e O ve n T o p
G as S e p tu m
S plit G a s
S tre a m
H e a te d G la s s to
L in e r W a s te
C a pilla ry
C o lu m n
1
2
3Figure 8 The Split Injection System
4
5This problem was partially solved by using larger diameter columns
6that would permit on-column injection. The columns are constructed to
7have an I.D. of about 0.056 in; which is slightly greater than the
8diameter of a certain hypodermic needles. This injection system is
9depicted in figure 9.
10However, there are also difficulties associated with this type of
11injector. On injection, the sample breaks up into separate portions, and
12bubbles form at the beginning of the column causing the sample to be
13deposited at different positions along the open tube as the solvent
14evaporates. On starting to develop the separation, each local
15concentration of sample acts as a separate injection. As a consequence,
16a chromatogram containing very wide or multiple peaks is produced.
88
89 19
1Procedures have been introduced in an attempt to eliminate sample

2splitting in this manner.
H o le s
S y rin g e
C a rrie r O ve n W a ll o r
G as O ve n T o p
S ilico n e
S e p tu m
W ide B o re
C a pilla ry
C o lu m n
3
4Figure 9 On-Column Injector for Large Bore Open Tubular
5Columns
6
7 Retention Gap Sampling
8
9The first solution to the problem of sample splitting was the 'retention
10gap method' which is depicted in figure 10.
11In this procedure stationary phase is removed from the first few
12centimeters of column. The sample is injected into this section and, if
13the sample becomes split, on commencing development, each split
14portion will still vaporize in the normal way. However, as there is no
15stationary phase present, the solutes will all travel at the velocity of the
16mobile phase until they reach the beginning of the coated section of the
17column. On reaching the start of the coating, the sample will be
18absorbed into the stationary phase and be concentrated at that point. As
19a result the sample is again at one point in the column. The retention
91
92
93 20
1gap procedure is normally used in conjunction with temperature

2programming, the program being initiated at a fairly low temperature.
3
S trip p e d S e ctio n o f
C a pilla ry C o lu m n
L iqu id S a m ple P la ce d
o n S trip p e d S e ctio n
L iqu id S a m ple B re a k s
in to T w o P a rts
S a m p le V o la ta liz e s a n d
M o ve s D o w n th e C o lu m n
S a m p le B e g in s to
A ccu m a la te s o n th e
C o a te d W a lls
S a m ple F o cu sse d a t
O n e S p o t a n d M ig ra te s
N o rm a lly
4
5
6Figure 10. The Retention Gap Method of Sampling
7
8The lower temperature aids in the accumulation of all the solutes where
9the stationary phase coating begins. In order for this method of
10sampling to be successful there must be a significant difference
11between the boiling points of the sample solvent and those of the
12components of the sample.
13
14 Sampling by Solute Focusing
15
16Another method of sampling that avoids sample splitting is the 'solute
17focusing method' which is more effective, but requires more
18complicated and expensive equipment. The injector is designed so that
19there are two consecutive, independently heated and cooled zones
96
97 21
1located at the beginning of the column. A diagram of the solute

2focusing system is shown in figure 11.
Z one 1 Z one 2
S tr ip p e d S e c tio n o f
C a p illa r y C o lu m n
L iq u id S a m p le P la c e d
o n S tr ip p e d S e c tio n
L iq u id S a m p le B r e a k s
in to T w o P a r ts
V o la tile S o lv e n t
R e m o v e d a n d E lu te d
F in a l R e m o v a l o f
S o lv e n t
Z o n e H e a te d a n d S o lu te s
C o n c e n tr a te d o n F r o n t o f
C o o le d S e c tio n
3
4Figure 11 The Solute Focusing Method sampling
5
6Initially the two zones are cooled and the sample is injected onto the
7first zone. The sample usually splits, but the carrier gas is allowed to
8remove the solvent, which is eluted through and out of the column.
9This leaves the sample spread along the first zone in dispersed
10fragments. The first zone is then heated while the second zone kept
11cool.
12
13The solutes in the first zone are eluted through the zone at the higher
14temperature and the sample accumulate at the beginning of the cooled
15second zone. The sample has now been focused as a compact band at
16the beginning of the column. The second zone is now heated and the
17separation developed normally. This technique is more flexible than
99
100
101 22
1the 'retention gap method' but the apparatus is more expensive and the
2procedure more complex.
3
4GC Columns
5
6There are two types of columns in common use in GC and they are the
7conventional packed column and the open tubular column. The former
8are usually 2 to 4 mm I.D. and 1 to 4 meters long and, packed with a
9suitable adsorbent, are mostly used for gas analysis. As a result of the
10simpler injection procedure and the more precise sampling method, the
11packed column tends to give greater quantitative accuracy and
12precision. However, despite its problems with sample injection, the
13open tubular column is seen as the 'state of the art' column and is by far
14the most popular column system in general use. The length of open
15tubular columns range from about 10 m to 100 m and can have internal
16diameters from 100 µm to 500 µm. The stationary phase is coated on
17the internal wall of the column as a film 0.2 µm to 1 µm thick.
18
19The Packed GC Column
20
21Packed columns are usually constructed from stainless steel or Pyrex
22glass. Pyrex glass is favored when thermally labile materials are being
23separated such as essential oils and flavor components. However, glass
24has pressure limitations and for long packed columns, stainless steel
25columns are used as they can easily tolerate the necessary elevated
26pressures. The sample must, of course, be amenable to contact with hot
27metal surfaces. Short columns can be straight, and installed vertically
28in the chromatograph. Longer columns can be U-shaped but columns
29more than a meter long are usually coiled. Such columns can be
30constructed of any practical length and relatively easily installed. Pyrex
31glass columns are formed to the desired shape by coiling at about
32700˚C and metal columns by bending at room temperature. Glass
33columns are sometimes treated with an appropriate silanizing reagent
34to eliminate the surface hydroxyl groups which can be catalytically
35active or produce asymmetric peaks. Stainless steel columns are
36usually washed with dilute hydrochloric acid, then extensively with
104
105 23
1water followed by methanol, acetone, methylene dichloride and n-

2hexane. This washing procedure removes any corrosion products and
3traces of lubricating agents used in the tube drawing process. The
4columns are then ready for packing.
5
6 Adsorbents
7
8There are two types of packing employed in GC, the adsorbents and the
9supports, on which the stationary phase is coated. There are both
10inorganic and organic types of GSC adsorbents, each of which have
11specific areas of application. All are ground and screened to provide a
12range of particle sizes that extend from about 30/40 mesh to 100/120
13mesh. In general, the smaller the particle size the higher the column
14efficiency, but the packing procedure is more difficult. It is also
15essential that the particle size range should be as narrow as possible.
16Packing materials that have a wide size range not only produce
17columns with poor efficiencies, but again, are also far more difficult to
18pack.
19
20Alumina, in an activated form, is used to separate the permanent gases
21and hydrocarbons up to about pentane. Alumina is usually activated by
22heating to 200˚C for about an hour. A common particle size is about
23100/120 mesh and the pore size range from about 1 Å to 100,000Å.
24Silica gel in spherical form (prepared by spraying a neutralized silicate
25solution (a colloidal silica sol) into fine droplets, allowing the silica gel
26to be formed, and subsequently drying the droplets in a stream of hot
27air). Silica is produced with a wide choice of surface areas and
28porosity's, which can range from about 750 m2/g and a mean pore size
29of 22 Å, to a material having a surface area of only 100m 2/g and a
30mean pore diameter of 300 Å. It is used for the separation of the lower
31molecular weight gases and some of the smaller hydrocarbons. In a
32specially prepared form, silica can be used for the separation of the
33sulfur gases, hydrogen sulfide, sulfur dioxide and carbon disulfide.
34
35Molecular sieves are used for the separation of small molecular weight
36gases largely by exclusion. The naturally occurring aluminosilicates are
107
108
109 24
1called zeolites, the synthetic zeolites are the Linde Molecular Sieves of
2which there are a number of different types available for specific
3applications. The zeolites have a crystalline structure which does not
4collapse when dehydrated. When water is removed from the crystals,
5channels of uniform dimensions are left within the structure which
6becomes very porous and the size of the channels changes only slightly
7with temperature. Molecular sieves are used to separate substances of
8different molecular size and shape, e.g. straight chain hydrocarbons can
9be separated from their branched chain isomers. The molecular sieves
10designated 5A and 13X are commonly used for the separation of
11hydrogen, oxygen, nitrogen, methane and carbon monoxide and also
12argon, neon and the other rare gasses.
13
14Carbon is also used as an adsorbent of which there are two types. The
15high surface area active carbon and the graphitized carbon (surface
16areas ranging from 5 m2/g to about 100 m2/g). The high surface area
17carbon, (ca 1000 m2/g) is used for the separation of the permanent
18gases and may need special treatment to modify its activity. The
19graphitized carbon adsorbents are much less active and separations
20appear to be based largely on exclusion. Macroporous Polymers such
21as the packings founded on the co-polymerization of polystyrene and
22divinylbenzene are also popular GC adsorbents. The extent of cross-
23linking determines its rigidity and the greater the cross-linking the
24harder the resin becomes until, at the extreme, the resin formed is very
25brittle. The macro-porous resin consists of resin particles a few microns
26in diameter, which in turn are composed of a fused mass of polymer
27micro-spheres, a few Angstroms in diameter. Consequently, the resin
28polymer has a relatively high surface area as well as high porosity.
29They exhibit strong dispersive type interaction with solvents and
30solutes with some polarizability arising from the aromatic nuclei in the
31polymer.
32
33 Supports for GLC
34
35There have been a number of materials used as supports for packed GC
36columns including, Celite (a proprietary form of a diatomaceous earth),
112
113 25
1fire-brick (calcined Celite), fire-brick coated with metallic silver or

2gold, glass beads, Teflon chips and polymer beads. Today however, the
3vast majority of contemporary packed GLC columns are filled with
4materials that are either based on of Celtic or polystyrene beads as a
5support. Diatomaceous supports comprise the silica skeletons of
6microscopic animals that lived many millions of years ago in ancient
7seas and lakes. As food transfer through the cells could only occur by
8diffusion, the supporting structure had to contain many apertures
9through which the cell nutrients could diffuse. This type of structure is
10ideal for a gas chromatography support, as rapid transfer by diffusion
11through the mobile and stationary phases is an essential requisite for
12the efficient operation of the column. The original Celite material is too
13friable and the brickdust too active, and thus a series of modified
14Celites had to be introduced. There are two processes used to modify
15Celite. One was to crush, blend and press the Celite into the form of a
16brick and then calcine it at a temperature of about 900˚C. Under these
17conditions some of the silica is changed into cristobalite and traces of
18iron and other heavy metals interact with the silica causing the material
19to become pink in color. This material is sold under the trade name of
20Chromosorb P. The second process involves mixing the Celite with
21sodium carbonate and fluxing the material at 900˚C. This causes the
22structure of the Celite to be disrupted and the fragments adhere to one
23another by means of glass formed from the silica and the sodium
24carbonate. As the original Celite structure is disrupted, the material
25exhibits a wide range of pore sizes which differs significantly from the
26material that was calcined in the absence of sodium carbonate. This
27materials is sold under the name of Chromosorb W together with two
28similar materials called Chromosorb G and Chromosorb S. The
29residual deleterious adsorptive properties of the support are due to
30silanol groups on the surface and these can be removed by silanization.
31The support is treated with hexamethyldisilazane which replaces the
32hydrogen of the silanol group with a trimethylsilyl radical. The reaction
33proceeds as follows,
34
115
116
117 26
H H
O O
S i O S i +
C H 3 H C H 3 C H 3 C H 3
C H 3 S i N S i C H 3 C H 3 S i C H 3 C H 3 S i C H 3
C H 3 C H 3 O O
S i O S i
1
2.In this way the strongly polar silanol groups are methylated and
3assume dispersive characteristics that do not produce peak tailing.
4Although the major contributors to adsorption by the support are the
5silanol groups, a residual adsorption results from the presence of trace
6quantities of heavy metals such as iron. which can be largely removed
7by acid washing priorto silanization. All three types of support are
8commercially available. None of these supports, however, are
9completely devoid of adsorptive properties and in may cases the effect
10of the residual adsorption must be further reduced by suitable
11stationary phase additives.
12
13To try to completely eliminate adsorption effects from the support,
14Teflon was explored as a possible alternative to a diatomaceous earth.
15Teflon powder proved to have little adsorption, but also proved to be
16extremely difficult to pack into a column. So difficult, that it is very
17rarely used in general GLC analyses. Its inert character makes it useful
18for the separation of certain highly corrosive materials. It has a
19temperature limit of about 250˚C.
20
21Glass beads have also been used as supports for packed GC columns
22and, if silanized, have little adsorption properties. Being non-porous, all
23the stationery phase must reside on the surface of the beads which
24gives them limited loading capacity. If the loading is increased, the
25stationary phase collects at the contact points of the spheres and form
26relatively thick accumulations, producing a high resistance to mass
27transfer and consequently low column efficiency. Glass beads appears
120
121 27
1to be the worst compromise between a column packed with modified

2Celite and a wall coated glass ,or fused silica, capillary column. The
3macroporous polymer beads are used as supports as well as adsorbents.
4They exhibit significant adsorption as the support itself acts as a
5stationary phase and makes a substantial contribution to retention.
6However, with normal sample loads, the adsorption isotherm is linear
7and so the eluted peaks are symmetrical. Only stationary phases that do
8not affect the polymer in any way can be used with such beads, which
9is a distinct disadvantage. They also have relatively poor temperature
10stability.
11
12Coating the Supports
13
14It is important to have an accurate measure of the amount of stationary
15phase that has been placed on a support to ensure retention time
16reproducibility and qualitative accuracy. The reproducibility of the
17coating procedure may have particular significance when the analytical
18results are to be used for forensic purposes. The material can be coated
19by the direct addition of the stationary phase to the support, by the
20filtration method or by the slurry method. The slurry method of coating
21is the one that is recommended.
22
23Coating by direct addition would appear to be the ideal quantitative
24method of preparing the column packings. A weighed amount of
25stationary phase is added directly to a known mass of support
26contained in a glass flask. The material is well mixed by rotating the
27flask for several hours, but even with extensive mixing, the stationary
28phase being is still irregularly distributed throughout the packing. As a
29result, the efficiency of the column slowly increases with use, as the
30stationary phase distributes itself more evenly throughout the packing.
31It may take several weeks of use for the column to give a constant
32maximum efficiency.
33
34The filtration method gives a packing with the stationary well
35distributed over the support but the loading can not be accurately
36calculated. A known mass of stationary phase is dissolved in sufficient
123
124
125 28
1solvent to provide excess liquid when mixed with a weighed amount of

2the support. The mixture is filtered under vacuum and the volume of
3the filtrate measured. From the volume of filtrate, the amount of
4solvent remaining on the support can be calculated and hence this
5stationary phase loading can be accessed. The bed is then sucked dry,
6the solvent evaporated and the coated support packed into the column.
7The amount of stationary phase on the support is not determined
8accurately by this method due to solvent losses by evaporation.
9In the slurry method of coating, a weighed amount of the support is
10placed in the flask of a rotary evaporator and the required mass of
11stationary phase added. An appropriate volatile solvent is then added in
12sufficient quantity to produce a free flowing slurry. The flask is then
13rotated at room temperature for ten minutes to ensure complete mixing.
14The rotating flask is then heated and the solvent removed by
15evaporation. When the packing appears dry, the material is then heated
16to about 150˚C in and oven to remove the final traces of solvent. This
17method of coating gives an extremely homogeneous surface
18distribution of stationary phase throughout the support and an accurate
19value for the stationary phase loading.
20Column Packing
21
22Short columns are usually straight and can be packed vertically.
128
129 29
1
2
1. 2,2-Dimethylpentane 14.2,2,3-Trimethylpentane
2..2,4-Dimethylpentane 15.3,3-Dimethylhexane
3. 2,3,3-Trimethylbutane 16.2,3,4-Trimethypentane
4. 3,3-Dimethylpentane 17.2,3-Dimethylhexane
5.2-Methylhexane 18. 2-Methylheptane
6. 2,3-Dimethylpentane 19.2-Methyl-3-ethyl-pentane
7.3-Methylhexane 20.2,3,3-Trimethylpentane
8.3-Ethylpentane 21.4-Methylheptane
9.2,3,4-Trimethylpentane 22.3-Methylheptane
10.n-Heptane 23.3-Ethylheptane
11.2,2-Dimethylhexane 24. 2,4-Dimethylhexane
12.2,5-Dimethylhexane 25.#-Methyl-2-ethylpentane
13.2,4-Dimethylhexane 26.n-Octane
3The column temperature was 78.6˚C, and the column packing 2.5w/w Apiezon Oil
4on C22 firebrick 100-120 mesh. The column diameter was 2 mm, the inlet pressure
5200 p.s.i., and the column efficiency 30,000 theoretical plates. The argon detector
6was used and the sample weight was 20 µg.
7Figure 12. Chromatogram from a 50 ft Column Showing the
8Separation of the Isomeric Heptanes and Octanes
9
10The packing is added, about 0.5 ml at a time, and the column tapped
11until the packing had settled. Another portion of packing is then added
12and the process repeated until the column is full. U-shaped columns are
13packed in the same manner. Columns up to 50 ft long can be packed in
14a series of U's and then each U column joined with a low dead volume
15connection. If the columns were glass they were usually filled through
131
132
133 30
1an opening at the top of each U which was terminated in a plug of

2quartz wool and sealed-off in a blow-pipe flame. These long packed
3columns could be operated at a maximum of 200 psi. and could provide
4efficiencies of up to 50,00 theoretical plates. Such columns could
5tolerate charges of several microliters. A chromatogram of the isomeric
6heptanes and octanes obtained from a 50 ft column is shown in figure
712.However, straight columns are clumsy to use and occupy a large
8amount of space which is often difficult to thermostat.
9
F ro m R e du cin g
V a lve
P a ck in g
R e se rvo ir
Q u a rtz
W ool
P ack in g
C o lu m n
V acu u m
Pum p
10
11
12Figure 13 An Example of a Column Packing Apparatus
13The coiled column although more difficult to pack has been readily
14accepted due to the compact nature of their design. To obtain adequate
15efficiencies, however, a special packing procedure had to be
16developed. The apparatus used is shown in figure 13. The packing is
17placed in a reservoir attached to a gas supply that forces the packing
18through the column. The column exit is connected to a vacuum pump.
19A wad of quartz wool is placed at the end of the column, constrained
136
137 31
1by a small restriction, that prevents the wad from being sucked into the
2pump. The vacuum and gas flow are turned on simultaneously and the
3packing is swept rapidly through the column. This causes the material
4to be slightly compacted along the total length of the column and has
5been shown to produce well-packed columns. The procedure is a little
6tedious and the success rate is sometimes less than 90%. In addition,
7the process does not lend itself to automation. The difficulties involved
8preparing packed columns have also contributed to the preferential
9popularity of the open tubular columns.
10
Benzene
T olu e n e
m X yle n e
p X y le n e ο X y le n e
E th y l
Benzene
11
12Figure 14. The Separation of a "Benzole" Mixture on a Packed
13Column, 40 ft Long
14
15The production of capillary columns can be largely automated and
16several columns can be prepared simultaneously. Another example of
17a chromatogram from a 40 ft packed column 2 mm I.D., this time of a
18"benzole" mixture is shown in figure 14.
19
20The column was packed with 5% w/w polyethylene glycol adipate
21coated on deactivated fire brick and operated isothermally at 130˚C
22with an inlet gas pressure of 140 psi. The analysis time was about 3.5
139
140
141 32
1hours. The column efficiency was about 40,000 theoretical plates and
2all the xylene isomers are separated. The two previous off-scale peaks
3are benzene and toluene. This separation could be achieved equally
4well on a open tubular column and probably in less than half the time.
5The advantage of the packed column would be that much higher
6sample loads can be placed on the column and thus the dynamic range
7of the analysis can be made much greater. Components present at a
8level of 0.001% can be easily separated and determined quantitatively
9without any preliminary fractionation or concentration.
10
11The Capillary or Open Tubular Column
12
13Capillary columns are fabricated from stainless steel or quartz. Metal
14capillary columns must be carefully cleaned to remove traces of
15extrusion lubricants before they can be coated, usually by washing with
16methylene dichloride, methanol and then water. After removing oil and
17grease, the columns are washed with dilute acid to remove metal oxides
18or other corrosion products that may remain adhering to the walls,
19washed with water and the again washed with methanol and methylene
20dichloride. Finally the column is dried in a stream of hot nitrogen.
21Metal columns provide the high efficiencies expected from open
22tubular columns and were used for the analysis of petroleum and fuel
23oils, etc. Metal columns, however, have some disadvantages as
24although easily coated with dispersive stationary phases (e.g., squalane,
25Apiezon grease etc.) they are not so easily coated with the more polar
26stationary phases such as CARBOWAX®. In addition, hot metal
27surfaces can cause decomposition or molecular rearrangement of many
28thermally labile materials such as the terpenes contained in essential
29oils. Metal can also react directly with some materials by chelation and
30adsorb polar material which results in asymmetric and tailing peaks.
31Nevertheless, metal columns are rugged, easy to handle and easy to
32remove and replace in the chromatograph consequently, their use has
33persisted in many application areas despite the introduction of fused
34silica columns.
35
144
145 33
1Desty et al. (4), tried to eliminate the activity of the open tubular
2column surface by developing the first silica-based columns and
3invented an extremely clever device for drawing soft glass capillaries.
4Desty produced both circular rigid soft glass and circular rigid Pyrex
5capillary columns, but their permanent circular shape, made them
6difficult to fit to unions connecting columns to injector and column to
7detector. By careful surface treatment the rigid glass tubes could be
8coated with polar stationary phases such as CARBOWAX®.
9Dandenau (5) introduced flexible fused silica capillary columns using
10the quartz fiber drawing technique. The solid quartz rod used in quartz
11fiber drawing was replaced by a quartz tube and the drawing rates
12adjusted appropriately. The quartz tubes had to be coated on the
13outside with polyimide to prevent moisture attacking the surface and
14producing stress corrosion. Coating the capillary tube with a polyimide
15polymer immediately after drawing prevents moisture coming in
16contact with the surface and thus stabilizes the tube. Soft glass
17capillaries can be produced by the same technique at much lower
18temperatures (6) but the tubes are not as mechanically strong or as inert
19as quartz capillaries. Surface treatment is still necessary with fused
20quartz columns to reduce adsorption and catalytic activity and also
21make the surface sufficiently wettable to coat with the selected
22stationary phase. The treatment may involve washing with acid,
23silanization and other types of chemical treatment, including the use of
24surfactants.
25
26Deactivation procedures used for commercial columns are kept highly
27proprietary. However, a deactivation program for silica and soft glass
28columns that is suitable for most applications would first entail an acid
29wash. The column is filled with 10% w/w hydrochloric acid, the ends
30sealed and the column is then heated to 100˚C for 1 hour. It is then
31washed free of acid with distilled water and dried. This procedure is
32believed to remove traces of heavy metal ions that can cause adsorption
33effects. The column is then filled with a solution of
34hexamethyldisilazane contained in a suitable solvent, sealed, and again
35heated to the boiling point of the solvent for 1 hour. This procedure
147
148
149 34
1blocks any hydroxyl groups that were formed on the surface during the
2acid wash. If the column is to be coated with a polar stationary phase, it
3may be advantageous to employ a polar or semipolar reagent as
4opposed to the dispersive silicone to facilitate coating. The column is
5then washed with the pure solvent, dried at an elevated temperature in a
6stream of pure nitrogen and is then ready for coating.
7
8Open tubular columns can be coated internally with a liquid stationary
9phase or with polymeric materials that can be polymerized to form a
10relatively rigid, internal polymer coating. There are two methods for
11coating a capillary column the dynamic method and the static method.
12Dynamic Coating
13
14A plug of solvent containing the stationary phase is placed at the
15beginning of the column. The strength of the solution, among other
16factors, determines the thickness of the stationary phase film. In
17general the film thickness of an open tubular column ranges from 0.25
18µm to about 1.5 µm.
S o lv e n t C a p illa r y
P lu g C o lu m n
19 M o ve m e n t o f P lu g
20
21Figure 15. The Dynamic Coating Procedure for an Open Tubular
22Column
23
24In practice, a 5% w/w of stationary phase in the solvent will produce a
25film thickness of about 0.5 µm. However, this is only approximate, as
26the film thickness is also determined by the physical properties of the
27surface, the solvent and the stationary phase. The coating procedure is
28depicted in figure 15. After the plug has been run into the front of the
29column (sufficient to fill about 10% of the column length), pressure is
30applied to the front of the column to force the plug through the column
152
153 35
1at 2-4 mm per second (it will take about 5.5 hours for the plug to pass
2through a 60 m column). When the plug has passed through the
3column, the gas flow is continued for about an hour. The gas flow must
4not be increased too soon, or the stationary phase solution on the walls
5of the tube is displaced forward in the form of ripples, which produces
6a very uneven film. After an hour the flow rate can be increased and the
7column stripped of solvent. The last traces of the solvent are removed
8by heating the column above the boiling point of the solvent at an
9increased gas flow rate. Complete solvent removal can be identified by
10connecting the column to a detector and observing the baseline drift of
11the detector.
12
13 Static Coating
14
15The entire column is filled with a solution of the stationary phase and
16one end is connected to a vacuum pump. As the solvent evaporates, the
17front retreats back down the tube leaving a coating on the walls. A
18diagram of the static coating procedure is shown in figure 16. The
19column is filled with a solution of stationary phase having a
20concentration appropriate for the deposition of a film of the desired
21thickness. Again the required concentration will depend on the
22stationary phase, the solvent, the temperature and the condition of the
23wall surface. Unfortunately, the optimum solvent concentration is not
24theoretically predictable and requires some preliminary experiments to
25be carried out to determine the best coating conditions.
26
S o lv e n t C a p illa r y
E n d o f C o lu m n P lu g C o lu m n
S e a le d
To
V acu u m
27
28
29Figure 16. The Static Method for Coating Open Tubular Columns
30
31After filling, one end of the column is sealed, and the other end is
32connected to a high vacuum pump and placed in an oven and the
33solvent slowly evaporates and the front retreats leaving a film of
155
156
157 36
1solution on the walls. The solvent then evaporates from this film and
2the stationary phase remains as a thin coating on the wall.
3
4
T im e (m in u te s )
5
6
7 Courtesy of Supelco, Inc.
8
1/ Dichlorodifluoromethane 18/ Bromodichloromethane
2/ Chloromethane 19/ 2-Chloroethyl vinyl ether
3/ Vinyl chloride 20/ cis-1,3-Dichloropropene
4/ Bromomethane 21/ Toluene
5/ 1,1-Dichloroethylene 22/ trans-1,3-Dichloropropene
6/ Methylene chloride 23/ 1-Chloro-2-bromopropane
7/ trans-1,2-Dichloroethylene 24/ 1,1,2-Trichloroethane
8/ 1,1-Dichloroethane 25/ Tetrachloroethylene
9/ cis-1,2-Dichloroethylene 26/ Dibromochloromethane
10/ Chloroform 27/ Chlorobenzene
11/ Bromochloromethane 28/ Ethylbenzene
12/ 1,1,1-Trichloroethane 29/ Bromoform
13/ Carbon Tetrachloride 30/ 1,4-Dichlorobutane
14/ Benzene 31/ 1,1,2,2-Tetrachloroethane
15/ 1,2-Dichloroethane 32/ 1,3-Dichlorobenzene
16/ Trichloroethylene 33/ 1,4-Dichorobenzene
160
161 37
17/ 1,2-Dichloropropane 34/ 1,2-Dichlorobenzene

9Figure 18. The Separation of Volatile Priority Pollutants
10
11The procedure is continued until all the solvent has evaporated and,
12except for the stationary phase, the column is empty. This process may
13take hours to complete. The procedure needs no attention and thus, can
14be carried out overnight. This procedure is more repeatable than the
15dynamic method of coating but, produces columns having similar
16performance to those dynamically coated.
17
18Irrespective of the coating method, column stability depends on the
19stability of the stationary phase film which depends on the constant
20nature of the surface tension forces that hold it to the column wall.
21These surface tension forces can be reduced with an increase in
22temperature or by the solutes passing through the column. As a
23consequence, the film can suddenly break up. Thus, it would be highly
24desirable if the stationary phase was bonded to the column walls or
25polymerized in situ. Such coatings are called immobilized stationary
26phases and can not be removed by solvent washing.
27
28Stationary phases that are polymeric can sometimes be formed on the
29wall surface by depositing the monomers or dimers on the walls and
30then initiating polymerization either by heat or an appropriate catalyst.
31This locks the stationary phase to the column wall and is thus
32completely immobilized. Polymer coatings can be formed in the same
33way using dynamic coating. The techniques used for immobilizing the
34stationary phases are also highly proprietary and little is known of the
35methods used by companies that manufacture the columns. In any
36event, most chromatographers do not want the trouble of coating their
37own columns and prefer to purchase proprietary columns.
38
39Very difficult separations can be achieved using the capillary column,
40and in a relatively short time. An example of the separation of a
41complex mixture on a capillary column is shown in figure 17. The
42column used was designated as a VOCOL column and was 60 m long,
430.75 mm I.D. and carried a film of stationary phase 1.5 micron thick.
163
164
165 38
1The column was held a 10˚C for 6 minutes and then programmed to
2170˚C at 6˚C per minute. The carrier gas was helium at a flow rate 10
3ml/min. The detector employed was the FID. This chromatogram
4demonstrates the clear advantages of capillary columns over packed
5column. Not only does the column produce exceeding high efficiencies
6but they are also achieved with reasonable separation times.
7
8 Open Tubular Column Types
9
10Open Tubular columns are broadly split into two classes, the wall
11coated open tubular columns or WCOT Columns (which have already
12been described and are by far the mot popular,) and the porous layer
13open tubes or PLOT Columns. The two types of column are shown
14diagramatically in figure 18. The PLOT columns are largely used for
15gas analysis and the separation of low molecular weight hydrocarbons.
16The external diameter of PLOT columns range from 320 to 530 µm
17with a porous layer that can be 5 to 50 µm thick.
18
P L O T C o lu m n s W C O T C o lu m n s
P orou s L aye re d O pe n T u be s W a ll C o a te d O p e n T u b e s
C o a tin g C o a tin g
T h ick n e ss T h ick n e ss
5 -5 0 µm 0 .1 - 8 µ m
In te r n a l D ia m e te r
In te r n a l D ia m e te r
1 0 0 -5 3 0 µm
3 2 0 -5 3 0 µm
19
20
21Figure 18. Open Tubular Column Types
22
168
169 39
1The technique of coating the walls with solid particles is again largely
2proprietary but stable and reproducible columns can be prepared and
3are commercially available. An example of the use of a PLOT column
4to separate and determine the impurities in a 2,3-butadiene sample is
5shown in figure 19. The column was 50 m long, 0.32 mm I.D. and
6coated with a 5 µm layer of aluminum oxide modified with potassium
7chloride. The separation was carried out by programming the column
8temperature from 100 ˚C to 200 ˚C at 6 ˚C per minute. The carrier gas
9employed was nitrogen and the gas velocity was 35 cm/sec. The sample
10(100 µl of gas) was placed on the column with a split injector and the
11detector used was the FID. Figure 20 shows an excellent separation
12was obtained with near baseline separation for all solutes. Such a
13separation would allow accurate and precise quantitative assay. The
14analysis of hydrocarbon gasses is an important and is a control assay in
15almost all oil refinery quality control laboratories.
16
5 7 11
1. M eth an e
2. E th an e
3. P ro p an e
4. is o - B u t a n e
5. n -B u tan e
6. 1 -B u tan e
7. t r a n s- 2 - B u t e n e
8. is o - B u t e n e
9. c is - 2 B u t e n e
1 8
6
2 4 10
3 9
0 T im e (m in u te s 15
17
18
19Figure 19. The Separation of the Impurities of 1,3-Butadiene
171
172
173 40
1
2Chiral Stationary Phases
3
4Modern organic chemistry and pharmaceutical research are becoming
5increasingly interested in methods of asymmetric syntheses. This
6enthusiasm has been provoked by the differing physiological activity
7that has been shown to exist between the geometric isomers of
8pharmaceutically active compounds. A tragic example being the drug
9Thalidomide, which was made available as a racemic mixture of N-
10phthalylglutamic acid imide. The important physiological activity
11resides in the R-(+)-isomer and it was not found, until too late, that the
12S-enantiomer was probably tetratogenic and caused serious fetal
13malformations. The separation and identification of isomers can,
14clearly, be very important and chromatography can be very effective in
15the resolution of such mixtures. The use of GC for the separation of
16asymmetric isomers is not as common as LC, but nevertheless there
17some very effective optically active stationary phases that can be used
18in GC for the separation of enantiomers.
176
177 41
H
O
O
OH
O O
O H O
O
OH O H H O
H
O
O
H
O O
O H
OH
H
O
H
O
O HO OH
HO
H
O
O O O
HO
O
OH
1
2Figure 20. The Structure of α − Cyclodextrin
3Some of the more useful GC stationary phases are based on the α- and
4β-cyclodextrins already described. The α-cyclodextrin structure is
5depicted in figure 20. The columns are usually 30 or 60 m long 0.25
6mm .D. and have an operating temperature range of 30˚C to 250˚C.
7Both the α and β forms are commercially available and both have been
8used very satisfactorily for the separation of the optical isomers of
9different flavors and fragrances. In order to employ the cyclodextrins
10as stationary phases for GC the permethylated α- or β-cyclodextrins are
11often embedded in a siloxane matrix (e.g. 35% phenyl-65% methyl
12polysiloxane) which is deposited on the walls of fused quartz capillary
13tubes. The phenyl-methyl-polysiloxane confers onto the column an
14intermediate level of polarity so the separations are basically enthalpic
15due to the dispersive and polar interactions that take place largely with
16the polymer but also entropic resulting from the chiral selectivity of the
17cyclodextrins.
179
180
181 42
1 3
4
2
5
6
1
2
3 Courtesy of Supelco, Inc.
4
Solute Retention Time (min.)
1. R-N-TFA-Amphetamine 10.12
2. S-N-TFA-Amphetamine 10.86
3. S-N-TFA-Methamphetamine 11.41
4. R-N-TFA-Methamphetamine 11.92
5. d-N-TFA--Pseudoephedrine 13.08
6. l-N-TFA-Pseudoephedrine 13.93
6Figure 21. The Separation of Some Chiral Amines

7Derivatization of the base cyclodextrin structure can introduce groups
8to which only one enantiomer can interact, while the other(s) are
9partially or wholly entropically hindered from interaction. This
10increases the differential interaction between the enantiomers and the
11stationary phase, thus, increasing the separation ratio and hence the
12resolution.An example of the use of a proprietary modified
13cyclodextrin in the separation of some chiral amines is shown in figure
1421. It is seen that excellent separations were obtained. A G-PN column
15was used which was 30 m long and 0.25 mm I.D. and operated at
16130˚C employing helium as the carrier gas. The basic materials are
184
185 43
1patented and the technique of bonding and coating the material onto
2the column is extremely difficult and involves much proprietary art.
3
4The Column Oven and Temperature Programmer
5
6The column oven should operate over a fairly wide temperature range
7(e.g. from 5˚C to 400˚C). In practice, however, the maximum oven
8temperature needed is usually less than 250˚C, particularly when
9synthetic stationary phases are being used, as many of them tend to be
10unstable and either decompose or volatilize at higher temperatures.
11Similarly, initial temperatures below 50˚C are also rarely needed. The
12oven usually has air circulation driven by a powerful fan to ensure an
13even temperature throughout the oven. The temperature in any part of
14the oven should be stable to ± 0.5 ˚C and when operating isothermally
15the column temperature should be constant to ± 0.2 ˚C. The oven
16should have a capacity of 1-2 cu. ft. and is supplied with fittings to
17accept more than one column and some switching valves if so desired.
18Such equipment is needed for multidimensional chromatography.
19
20The temperature programmer (hardware and software) usually has a
21range of linear gradients from 0.5˚C/min. to about 20˚C/min. Some
22programmers include nonlinear programs such as logarithmic and
23exponential, but most GC analyses can be effectively accomplished
24using linear programs only. The program rate can be changed at any
25time in the chromatographic development or intermittent isothermal
26periods can be inserted where necessary in the program. The
27temperature programming limits are usually the same as those of the
28oven (viz. 5˚C to 400˚C). All connections between the column and the
29detector, that pass though the column oven wall to the detector oven,
30are supplied with their own heaters so that no part of the conduit can
31fall below the column oven temperature. A cool spot in the conduit
32will cause condensation which can result in broad and distorted peaks.
33
34GC Detectors
35
187
188
189 44
1A large number of GC detectors have been developed and made

2commercially available. In general, GC detectors are 4 to 5 orders of
3magnitude more sensitive than LC detectors and, thus, are ideal for
4trace analysis and environmental monitoring. The detectors with the
5highest sensitivity tend to be specific and sense specific types of
6sample (e.g., halogenated substances by the electron capture detector).
7Conversely, those detectors with a catholic response, although highly
8sensitive compared to LC detectors (e.g. the flame ionization detector)
9are significantly less sensitive than the specific detectors. The detectors
10with a catholic response are the most popular and the majority of GC
11separations are monitored by the flame ionization detector (FID). The
12most commonly used specific detectors are the nitrogen phosphorus
13detector (NPD) and the electron capture detector (ECD) The
14katharometer detector, although having relatively poor sensitivity is
15widely used in gas analysis.
16
17The Flame Ionization Detector
18
19The FID, invented by Harley and Pretorious (7), and separately by
20McWilliams and Dewer (8), evolved from the Heat of Combustion
21Detector developed by Scott (9). The FID detector employs hydrogen as
22the combustion gas which is mixed with the column eluent (helium,
23nitrogen or other appropriate gas) and burnt at a small jet situated
24inside a cylindrical electrode. A potential of a few hundred volts is
25applied between the jet and the electrode and when a carbon containing
26solute is burnt in the jet, the electron/ion pairs that are formed are
27collected at the jet and cylindrical electrode. The current is amplified
28and fed to a recorder or to the A/D converter of a computer data
29acquisition system. A diagram of the basic FID is shown in figure 22.
30During the process of oxidation, oxidized or partially oxidized
31fragments of the solute are formed in the flame which are thought to
32generate electrons by thermionic emission. The background current
33(ions and electrons from the hydrogen flame alone) is very small (1-2 x
3410-12 amperes) and consequently, the noise level is also
35commensurably small (about 10-14 amperes).
192
193 45
E x it G a se s
In s u la te d
C o n n e ctio n In s u la te d
to C o lle cto r C o lle cto r
E le ctro d e E le ctro d e s
F la m e
In s u la tio n
In s u la te d J e t
In s u la te d
C o n n e ctio n
to J e t In s u la tio n
H ydroge n
C a pilla ry C o lu m n
C a rry in g M o bile A ir o r O x y g e n
P h a s e (H e liu m ) fo r
C o m b u s tio n
1
2
3Figure 22. The Flame Ionization Detector

4
5The ionization process is not very efficient, only 0.0018% of the solute
6molecules produce ions, (about two ions or electrons per 10 5
7molecules). Nevertheless, because the noise level is very small, the
8minimum detectable mass of n-heptane is only 2 x 10 -12 g/sec. At a
9column flow rate of 20 ml/min. this is equivalent to a minimum
10detectable concentration of about 3 x 10-12 g/ml. The detector responds
11to mass per unit time entering the detector, not mass per unit volume
12consequently the response is almost independent of flow rate. This is
13particularly advantageous and allows it to be used very effectively with
14capillary columns. Although the column eluent is mixed with the
15hydrogen prior to entering the detector, as it is mass sensitive and not
16concentration sensitive, the diluting effect has no impact on the
17sensitivity. The FID detects virtually all carbon containing solutes,
18with the exception of a small number of small molecular compounds
19such as carbon disulfide, carbon monoxide, etc. In fact, due to its
20diverse and comprehensive response, it is considered a universal
21detector.
22
195
196
197 46
1An example of the use of the FID in a paraffin, isoparaffin, aromatic,

2naphthene and olefin analysis of a hydrocarbon mixture (frequently
3called the PIANO analysis) is shown in figure 23. The column was the
4Petrocol DH 50.2, 50 m long and 0.5 mm I.D. and made from fused
5silica. The column temperature was held a 35oC for 5 minutes and then
6programmed up to 200˚C at 2˚/min. The carrier gas was helium and the
7mobile phase velocity of 20 cm/sec. Many standard tests carried out in
8the hydrocarbon and pharmaceutical industries and for environmental
9testing have been designed to utilize the FID as the detector
10
11
m - X y le n e
p -X y le n e
o -X y le n e
nC 13
nC 6 nC 7 nC 8 nC 9
nC 11 nC 12
nC 5 nC 10
T o lu e n e
B e n ze n e
0 8 16 24 32 40 48 56 64
T im e (m in u te s)
12
13
14
15 Courtesy of Supelco Inc.
16
17
18Figure 23. The Separation of a PIANO Standard Mixture

19
20The Nitrogen Phosphorus Detector (NPD)

21
200
201 47
1The nitrogen phosphorus detector (NPD), is a highly sensitive but

2specific detector and evolved directly from the FID. It gives a strong
3response to organic compounds containing nitrogen and/or phosphorus.
4Although it appears to function in a very similar manner to the FID, in
5fact, it operates on an entirely different principle. A diagram of an NP
6detector is shown in figure 24.
7
A node
+
R u b id iu m o r
C esiu m B ea d
H e a te r
C o n n e c tio n s
F la m e
H e a te r
A ir
H yd rogen
C a rrier G a s
N itr o g e n
8
9
10Figure 24. The Nitrogen Phosphorus Detector

11
12The actual NPD sensor is a rubidium or cesium bead contained inside a
13small heater coil. The helium carrier gas is mixed with hydrogen and
14passes into the detector through a small jet. The bead is heated by a
15current passing through the coil which is situated above the jet, and the
16helium-hydrogen mixture passes over it. If the detector is to respond to
17both nitrogen and phosphorus, then a minimum hydrogen flow is
18employed to ensure that the gas does not ignite at the jet. In contrast, if
19the detector is to respond to phosphorus only, a large flow of hydrogen
20can be used and the mixture burned at the jet. A potential is applied
21between the bead and the anode. The heated alkali bead emits electrons
22by thermionic emission which are collected at the anode and thus
203
204
205 48
1produce an ion current. When a solute containing nitrogen or

2phosphorus is eluted, the partially combusted nitrogen and phosphorus
3materials are adsorbed on the surface of the bead. This adsorbed
4material reduces the work function of the surface and, as consequence,
5the emission of electrons is increased which raises the anode current.
6The sensitivity of the NPD is about 10 -12 g/ml for phosphorus and 10-
711 g/ml for nitrogen). Unfortunately, the performance deteriorates with
8time. Reese (10) examined the function of the NPD in great detail. The
9alkali salt employed as the bead is usually a silicate and Reese showed
10that the reduced response was due to water vapor from the burning
11hydrogen, converting the alkali silicate to the hydroxide.
12
13
1/ Eptam® 2/ Sutan® 3/ Vernam® 4/ Tillam®
5/ Odram® 6/ Treflan® 7/ Balan® 8/ Ro-Neet®
9/ Propachlor 10/ Tolban® 11/ Propazine 12/ Atrazine
13/ Simazine 14/ Terbacil 15/ Sencor® 16/ Dual®
17/ Paarlan® 18/ Prowl® 19/ Bromacil 30/ Oxadiazon
21/ Goal® 22/ Hexazinone
14
16
17Figure 25. The Separation and Specific Detection of Some
18Herbicides Using the Nitrogen Phosphorus Detector
19
208
209 49
1At the operating temperature of the bead, the alkali hydroxide has a
2significant vapor pressure and consequently, the rubidium or cesium is
3continually lost during the operation of the detector. Eventually all the
4alkali is evaporated, leaving a bead of inactive silica. This is an
5inherent problem with all NP detectors and as a result the bead needs to
6replaced fairly regularly if the detector is in continuous use. The
7specific response of the NPD to nitrogen and phosphorus and its high
8sensitivity, makes it especially useful for the analysis of many
9pharmaceuticals and in particular in environmental analyses involving
10herbicides. Employing appropriate columns traces of herbicides at the
11500 pg level can easily be determined.
12
13An example of the separation and identification of a series of
14herbicides employing the NPD is shown in figure 25. An SPB-5
15column was used, 15 m long and 0.53 mm I.D. carrying a 0.5 µ film of
16stationary phase. The column temperature was held at 60 oC for 1
17minute and then programmed at 16 o/min. to 290oC and then held there
18for 5 minutes. The flow rate was 5 ml/min. and the carrier gas helium.
19The sample size was 1 µl of ethyl acetate containing 5 ng of each
20herbicide.
21
22The Electron Capture Detector
23
24The electron capture detector contains a low energy β-ray source which
25is used to produce electrons for capturing by appropriate atoms.
26Although tritium adsorbed into a silver foil has been used as the β
27particle source, it is relatively unstable at high temperatures, the Ni63
28source was found to be preferable. The detector can be used in two
29modes, either with a constant potential applied across the cell (the DC
30mode) or with a pulsed potential across the cell (the pulsed mode). In
31the DC mode, hydrogen or nitrogen can be used as the carrier gas and a
32small potential (usually only a few volts) is applied across the cell that
33is just sufficient to collect all the electrons available and provide a
34small standing current. If an electron capturing molecule (for example
35a molecule containing an halogen atom which has only seven electrons
211
212
213 50
1in its outer shell) enters the cell, the electrons are captured by the
2molecule and the molecules become charged. The mobility of the
3captured electrons is much smaller than the free electrons and the
4electrode current falls dramatically. The DC mode of detection,
5however, has some distinct disadvantages. The most serious objection
6is that the electron energy varies with the applied potential. The
7electron capturing properties of a molecule varies with the electron
8energy, so the specific response of the detector will depend on the
9applied potential
10
11Operating in the pulsed mode, a mixture of 10% methane in argon is
12employed which changes the nature of the electron capturing
13environment. The electrons generated by the radioactive source rapidly
14assume only thermal energy and, in the absence of a collecting
15potential, exist at the source surface in an annular region about 2 mm
16deep at room temperature and about 4 mm deep at 400˚C. A short
17period square wave pulse is applied to the electrode collecting the
18electrons and producing a base current. The standing current, using
1910% methane in argon is about 10-8 amp with a noise level of about 5
20x 10-12 amp. The pulse wave form is shown in figure 26.
21
A p p lie d P o t e n t ia l
A ctive
P e rio d
In a c tive P e r io d
22
23
24Figure 26. Wave form of Electron Capture Detector Pulses

25
26In the inactive period of the wave form, electrons having thermal
27energy only will attached themselves readily to any electron capturing
28molecules present in the cell with the consequent production of
29negatively charged ions. The negative ions quickly recombine with the
216
217 51
1positive ions (produced simultaneously with the electrons by the β

2particles) and thus become unavailable for collection. Consequently
3the standing current measured during the potential pulse will be
4reduced.
5
6The period of the pulsed potential is adjusted such that relatively few
7of the slow negatively charged molecules (molecules having captured
8electrons and not neutralized by collision with positive ions) have time
9to reach the anode, but the faster moving electrons are all collected.
10During the "off period" the electrons re-establish equilibrium with the
11gas. The three operating variables are the pulse duration, pulse
12frequency and pulse amplitude. By appropriate adjustment of these
13parameters the current can be made to reflect the relative mobilities of
14the different charged species in the cell and thus exercise some
15discrimination between different electron capturing materials. A
16diagram of an electron capture detector is shown in figure 27.
17
F lo w
D iffu s e r
I n s u la to r
+
R a d io a c tiv e
S ou rce
N itr o g e n o r H y d r o g e n .
F o r P u ls e d M o d e
O p e r a tio n 1 0 %
18 M e th a n e in A r g o n
19
20Figure 27 The Electron Capture Detector
21
219
220
221 52
1The basic electron capture detector consists of a small chamber one or

2two ml in volume enclosing two metal electrodes. The electrodes may
3be concentric cylinders or metal discs separated by an insulator. The
4cell contains the radioactive source, electrically connected to the
5entrance conduit and to the negative side of the power supply. A gauze
6"diffuser" is connected to the cell exit and to the positive side of the
7power supply. The output from the sensor is processed by suitable
8electronics and the output passed to either a potentiometric recorder of
9a computer data acquisition system. The electron capture detector is
10very sensitive, probably the most sensitive GC detector available (ca.
1110-13 g/ml) and is widely used in the analysis of halogenated
12compounds, in particular, pesticides. An example of a pesticide
13analysis employing an electron capture detector is shown in figure 28.
14
15
16
1 α−BHC 2γ-BHC (Lindane) 3 β-BHC 4 Heptachlor
5 δ-BHC 6 Aldrin 7 Heptachlor 8 Endosulphan
Epox.
9 p,p'-DDE 10 Dieldrin 11 Endrin 12 p,p'-DDD
13 Endosulphan 11 14 p,p'-DDt 15 Endin Aldehyde 16Endosulp. Sulf.
17
19
224
225 53
1Figure 28. The Analysis of Priority Pollutant Pesticides

2
3The column used was a SPB-608 fused silica capillary column, 30 m x

40.53 mm I.D. with a 0.5 µ film of stationary phase. The column was
5programmed from 50oC at 1o/min. to 150oC and then to 260oC at
68o/min. Helium was used as the carrier gas at a flow rate of 5 ml/min.
7The sample consisted of 0.6µl of a solution of the pollutants in n-
8decane. The mass of each pollutant present was about 120 pg.
9
10The Katherometer Detector

11
12The katherometer detector (sometimes spelt catherometer and often
13referred to as the thermal conductivity detector or hot wire detector) is
14relatively insensitive but has survived largely as a result of its catholic
15response and, in particular, its response to the permanent gases.
16Consequently, it is often the detector of choice for gas analysis and
17environmental testing. Its frequent use in these special types of
18application, somewhat surprisingly, has made it the fourth most
19commonly used GC detector. A filament carrying a current is situated
20in the column eluent and, under equilibrium conditions, the heat
21generated in the filament is equal to the heat lost by conduction and
22convection and consequently the filament assumes a constant
23temperature. At the equilibrium temperature, the resistance of the
24filament and thus the potential across it is also constant.
25
227
228
229 54
S e n s o r C o n n e ctio n s R e fe re n ce C o n n e ctio n s
to W h e a ts to n e B rid g e to W h e a ts to n e B rid g e
H e a te d M e ta l
B lo ck R e fe re n ce
F ila m e n t
S en sor
F ila m e n t
C a rrie r G a s R e fe re n ce F lo w
1 F ro m C o lu m n o f C a rrie r G a s
2
3
4Figure 29. The Katherometer Detector ("In-Line Cell")
5
6The heat lost from the filament will depend on the thermal conductivity
7of the gas and its specific heat and both these parameters will change in
8the presence of a foreign gas or solute vapor. The presence of a
9different gas entering the detector causes the equilibrium temperature
10to change, producing a change in potential across the filament. This
11potential change is amplified and fed to a suitable recorder. A diagram
12of the katherometer is shown in figure 329.
13
14The katherometer may have an "in-line" sensor where the column
15eluent passes directly over the filament or an "off-line" sensor where
16the filaments are situated out of the main carrier gas stream and the
17gases or vapors reach the sensing element by diffusion. Due to the high
18diffusivity of vapors in gases, diffusion can be considered as almost
19instantaneous.
20
232
233 55
1. H e liu m
5 2. H ydroge n
6 3. D e u te riu m – H y d ro g e n
4. T ritiu m – H y d ro g e n
3 5. D e u te riu m
6. D e u te riu m – T ritiu m
1 7. T ritiu m
4
7
0 5 10 15 20 25 30
1
2 Courtesy of the Supelco Inc.
3
4Figure 30. The Separation of the Compounds of Hydrogen,

5Deuterium and Tritium
6
7The katherometer detector is very flow and pressure sensitive and the
8sensor must be carefully thermostatted and fitted with reference cells to
9compensate for changes in pressure or flow rate. The filaments of the
10reference and measuring cell are made to form two of the arms of a
11Wheatstone bridge and the out-of-balance signal amplified and fed to a
12recorder or computer data acquisition system.The maximum sensitivity
13will be realized if hydrogen is used as the carrier gas, but, to reduce fire
14hazards, helium is preferred and can be used with very little
15compromise in sensitivity. The katherometer sensitivity is about 10-6
16g/ml with a linear dynamic range of about 500. Although the least
17glamorous, this detector can be used in most GC analyses that utilize
18packed columns and where there is no limitation in sample availability.
19The device is simple, reliable, and rugged and is particularly useful for
20those with limited experience in GC. It is also often the detector of
21choice for process monitoring. An example of the separation of the
235
236
237 56
1various compounds of hydrogen, deuterium and tritium, employing gas

2solid chromatography and using a katherometer detector is shown in
3figure 30. The stationary phase was activated alumina (treated with
4Fe(OH)2), and the column was 3 m long and 4 mm I.D. The carrier gas
5was neon, the flow rate 200 ml/min. (at atmospheric pressure) and the
6column temperature was -196oC.
7
8The four detectors described are well established. reliable and

9generally simple to operate. They are also, probably the most popular.
10The FID, ECD, NPD and the katherometer are employed in over 90%
11of all GC applications. The FID is the most versatile, sensitive and
12linear, and probably the most generally useful. For details of other GC
13detectors see Book 4.
14
15Data Acquisition and Processing

16
17Originally, analytical results were calculated from measurements made
18directly on the chromatogram provided by the chart recorder. This is
19still true for many chromatographs in use today, but analyses obtained
20from contemporary instruments commonly process the results using a
21computer. The output from the detector (which is only rarely the direct
22output from the detector sensor) is usually in millivolts and is suitable
23for direct connection to a potentiometric recorder. This output
24represents a voltage that is linearly related to solute concentration
25being measured by the detector sensor and as the sensor response is
26often nonlinear, the signal usually requires nonlinear processing to
27provide the equired output. This is carried out by the detector
28electronics. The FID is an exception to this, as the ion current from the
29flame itself happens to be linearly related to the mass of carbon passing
30through it per unit time. A block diagram showing the essential
31elements of a data acquisition and processing system is given in figure
3231.
33
240
241 57
F rom
D e te cto r
S ca lin g P o te n tio m e tric

A m p lifie r R e corde r
A /D C o n v e r t e r C o m p u te r P rin te r
1
2
3
4Figure 31. Data Acquisition and Processing System
5
6The Scaling Ampifier
7
8The output from the detector usually passes directly to a scaling
9amplifier that modifies the signal to a range that is appropriate for the
10analog-to-digital (A/D) converter. The output can alternatively pass to a
11potentiometric recorder and produce the chromatogram in real time.
12The computer system can also produce a real time chromatogram but,
13to do so, the data must be processed and the chromatogram presented
14on the printer. The output from most detectors ranges from 0 to 10
15mV? whereas the input required by most A/D converters is
16considerably greater e.g. 0 to 1.0 V. For example, if the FSD of the
17signal is 10 mv, the instantaneous measurement of 2 mV (assumed
18from the detector) must be scaled up to 0.2 volt, which is carried out by
19a simple linear scaling amplifier having a gain of 100.
20
21The A/D Converter

22
23After scaling, the signal must be converted to digital form. There are a
24number of ways to digitize and only one, the simplest will be
243
244
245 58
1described. A diagram showing the operating principle of an

2voltage/frequenc V/F type A/D converter is shown in figure 32.
3
4
F rom
S ca le r
O p e ra tio n a l
A m p lifie r
E le ctro n ic
S w itch
C o m p a ra to r
T o C o u n te r
a n d R e g iste r
5
6Figure 32. The Basic V/F Analog–to–Digital Converter.
7
8The converter consists of an integrator that can be constructed from an

9operational amplifier with a feedback capacitor. The capacitor is
10charged by the voltage from the scaling amplifier through the
11operational amplifier.The output from the integrator is sensed by a
12comparator which activates the electronic switch when the potential
13across the capacitor reaches a preset voltage. The activation of the
14comparator also causes a pulse to be passed to a counter and at the
15same time the capacitor is discharged by the electronic switch. The
16process then starts again. The time taken to charge the capacitor to the
17prescribed voltage will be inversely proportional to the applied voltage
248
249 59
1and consequently the frequency of the pulses from the comparator will
2be directly proportional to the applied voltage. The frequency of the
3pulses generated by the voltage controlled oscillator is sampled at
4regular intervals by a counter which then transfers the count in binary
5form to a register. The overall system is shown diagramatically in
6figure 33.
2 m V
0 .2 v o lt
A /D
C o n ve rto r C o u n te r
R e g is te r
128 64 32 16 8 4 2 1
0V 0V 0V 0V
5V 5V 5V 5V
7 32 + 16 + 2 + 1 = 51
8Figure 33. Stages of Data Acquisition
9
10The output from most detectors ranges from zero to ten millivolts and
11the input range of many A/D converters is from zero to one volt. Thus,
12the instantaneous measurement of 0.2 mV from the detector must be
13scaled up by a factor of 100 to 0.2 volts, which is carried out by the
251
252
253 60
1scaling amplifier in the manner shown. The A/D converter changes the
2analog voltage to a digital number, the magnitude of which is
3determined by the number of "bits" that the computer employs in its
4calculations. If, for example, eight bits are used, the largest decimal
5number will be 255. The digital data shown in figure 33 can be
6processed backward to demonstrate A/D procedure. It is seen that the
7third and fourth most significant "bits" (which are counted from the far
8left) and the two least significant "bits" (which are counted from the far
9right) are at the five volt level (high), which as shown in figure 32 is
10equivalent to 51 in decimal notation (32+16+2+1). It follows that the
11voltage that was converted must be . It should also be noted that
12because of the limitation of 8 "bits", the minimum discrimination that
13can be made between any two numbers is . It follows that 8 bit systems
14are rarely used today and contemporary A/D converters usually have at
15least 12 or 16 bit outputs.
16
17The output from the A/D converter is sampled regularly by the
18computer and the curve relating this data to time will reconstitute the
19chromatogram. The precision of the chromatogram and any
20calculations made with the data will obviously depend on the
21frequency of sampling which is normally user selected.
22
23Data Processing
24
25In the early days of gas chromatography, the associated computers used
26core storage which was bulky, expensive and had a very limited
27capacity (e.g., 8 kilobytes was a large memory). The limited memory
28meant that the programming was limited and had to be written
29extremely economically (i.e. employing the minimum of memory) and
30much of the data processing was done 'on-the-fly'. This meant that after
31each peak was eluted, it retention time and height was noted and its
32area calculated and then the raw data was discarded and only the
33retention time, peak height and peak area were stored. This economic
34processing package could not recalculate the data after the separation
35was complete, it could not reconstitute the chromatogram and it could
36not employ an alternative algorithm for area measurement if the one
256
257 61
1used was not appropriate. These restrictions were entirely a result of

2the cost and size limitation of computer memory at that time. With the
3introduction of cheap, compact solid state memory and the high
4capacity disk memory, the situation has completely changed. 8
5megabytes is now a small memory and disk capacities are now
6measured in gigabytes. All the chromatography data can now be stored
7and reprocessed after the separation as many times as required,
8chromatograms can be reconstituted (with modified axes if necessary)
9and quantitative data manipulated as necessary. In addition, because
10the computer speeds have also increased greatly, on the fly processing
11can be carried out in parallel with normal data processing if required.
12The processing can include a variety of fairly sophisticated
13mathamatical procedures such as base-line correction, peak skimming,
14and multi-peak deconvolution. Techniques used in data processing
15will be discussed in more detail in Book 10.
16
17Quantitative Analysis
18
19There are three important stages in a GC analysis,
20
21 1. The preparation of the sample.

22 2. The development of the separation and the production of the
23 chromatogram
24 3. The processing of the data and the presentation of the results.
25
26Each stage is equally important and if not carried out correctly the
27results will be neither precise nor accurate. Sample preparation can be
28very simple involving no more that diluting a known weight of sample
29with mobile phase or be much more complex including an extraction
30procedure followed by derivatization and then dilution. For some
31samples the preparation can be the most time consuming and difficult
32part of the whole analysis. Details of sample preparation is the subject
33of Book 18 but an example of one of the more complex sample
34preparation methods will be given to illustrate some of the procedures
35that may be necessary.
36
259
260
261 62
1Liquid extraction is a clumsy procedure, particularly when used on the

2micro scale which is often necessary in sample preparation. An
3alernative procedure is solid phase extraction. The procedure is
4relatively simple and involves the use of a short tube packed with an
5appropriate adsorbent such as silica, reversed phase silica or, for some
6applications, macro porous polymer beads. The adsorbent must be
7capable of removing the substances of interest from the liquid medium.
8Extracting trace materials from water (e.g., pollution analysis) a
9reversed phase would be appropriate. Then the substances could be
10displaced into solvents such as n-hexane, methylene dichloride etc. A
11diagram of a simple solid phase extraction tube is shown in figure 34.
12
P o ly pro py le n e
B ody
P o ly e th y le n e F rits
(2 0 µm p o re s ) S ilica B a se d P a ck in g
4 0 m m P a rticle s 6 0 Å P o re s
L u e r T ip
13
14
16
17Figure 34 A Solid Phase Extraction Tube
18
19The extraction tubes are usually made of an inert plastic such as
20polypropylene and have a range of capacities of 1, 2, or 5 ml. The tube
21is one fifth filled with adsorbent and contained by plastic frits at either
22end. The upper part of the tube, above the packing, acts as a funnel or
23container for the liquid to be extracted. The liquid sample is allowed to
24percolate through the adsorbent bed. Sometimes the lower end of the
25tube is connected to a vacuum or the top to a gas supply to increase the
264
265 63
1flow of sample through the bed. The adsobed material is then desorbed
2with an appropriate solvent, the sample diluted to a known volume and
3an aliquot used for analysis. If necessary the extract can be
4concentrated by evaporation and the total concentrate employed for
5analysis.
6
7To avoid breakdown of labile materials, a totally inert extraction
8apparatus can be constructed from Teflon. A diagram of such an
9apparatus, produced by Alltech, is shown in figure 35 which even
10includes a Teflon hypodermic needle.
11
E x tra ct T u b e
S o lid P h a se
E x tra ctio n
P a ck in g T e flo n
F rits
D is p o s a b le T e flo n
T ra n s fe r N e e d le
G la s s C o lle ctio n
V ia l
12
13
14 Figure 35 An All–Teflon Solid Phase Extraction Apparatus
15
16This type of extraction system is useful for biotechnology samples. An
17example of the use of solid phase extraction to determine trace amounts
18(5 ppb) of some chlorinated pesticides in drinking water is shown in
19figure 36. The extraction tube was designated as the Novo-Clean C18.
20It was 47 mm tube long which included the membrane manifold. The
21materials were removed from the water sample by dispersive
22interactions between the solutes and the C18 reversed phase. The tube
23was conditioned before use with 10 ml of methanol, 10 ml of methyl-
267
268
269 64
1tributylether (MTBE), 15 ml of methanol and finally 125 ml of

2deionized water.
0 10 20 30 40
R e te n tio n T im e (m in u te s )
3
4 Courtesy of Alltech Inc
5
1. Lindane–α 9. p,p'-2-chlorophenyl,2-p-chlorpenyl
1,1,dichloroethylene
2. Lindane–β 10. Endrin
3. Lindane–γ 11. p,p'-2,2-bis p chlorpheny1 chlor-
ethylene
4. Lindane-∆ 12. Endrin aldehyde
5. Heptachlor 13. Ensodulfan Sulfate
6. Aldrin 14 p,p'-1'1'1-trichlor, 2,2-bis p chloro
phenyl ethane
7. Heptachlo Epoxide 15. Endosulfan II
8. Dieldrin
6
7Figure 36. Separation of Some Chlorinated Pesticides Removed
8from Drinking Water by Solid State Extraction.
9
10 The water sample was pumped through the extraction tube at a rate of
11100 ml/min. The solutes removed were displaced from the extraction
12tube with 10 ml of methanol followed by 10 ml of (MTBE) and dried
13over anhydrous sodium sulfate. It is seen that all the chlorinated
14pesticides were extracted and concentrations down to 1 ppb could be
15easily identified.
16
17Derivatization
18
19GC samples are usually derivatized to render highly polar materials
20sufficiently volatile so that they can be eluted at reasonable
272
273 65
1temperatures without thermal decomposition or molecular re-

2arrangement. Examples of such materials that need to be derivatized
3are the organic acids, amides, poly hydroxy compounds, amino acids
4etc. In order to render such materials more volatile, they are either
5esterified, silanated or acetylated using one of a number of different
6methods of derivatization.
7
8Acids can be esterified by treating them with an appropriate alcohol
9using an inorganic acid to catalyze the reaction. Hydrochloric acid was
10popular for this purpose because it's strength was adequate and any
11excess could be easily removed. Sulfuric acid is not very suitable as it
12can cause charring and any excess is difficult to remove. Other
13catalysts that have been found effective are trifluoroacetic acid,
14dichloroacetic acid, benzene sulphonic acid, p-toluene sulphonic acids
15and suphuryl and thionyl chlorides. A volatile acid is recommended
16such as hydrochloric acid or thionyl chloride. However, the derivative
17must be must be sufficiently involatile not to allow loss when removing
18the excess alcohol and where appropriate the catalyst itself. A general
19method would be to treat one or two milligrams of the acid contained
20in a small with 125 µl of either methanol or ethanol that contains 3M
21hydrochloric acid and heat at 65˚C for about 35 minutes. A stream of
22nitrogen is bubbled through the reaction mixture to remove the alcohol.
23It is clear that the derivative must be sufficiently involatile, (i.e., has
24an adequately low vapor pressure) to prevent any loss during the
25removal of the alcohol.
26
27Amino acids are more difficult to derivatize but can also be esterified
28in a comparable manner. A few milligrams of the amino acid mixture is
29mixed with 2 ml of 4M alcoholic methanol and heated at 70˚C for 2
30hours. Any excess methanol is then removed by evaporation in a
31stream of nitrogen. Any remaining water is removed by adding a little
32dichloromethane (ca 150 µl) and repeating the evaporation process.
33The derivative is in the form of the hydrochloride and the free base
34must be liberated without causing the ester to be saponified.
35
275
276
277 66
1 Another useful catalyst for esterification is thionyl chloride but the

2thionyl chloride must be purified by distilling from linseed oil before
3use. 10 to 50 mg of acid is placed in a stoppered vial and 200 µl of dry
4methanol is added and cooled in a solid carbon dioxide-acetone bath.
520 µl of thionyl chloride is added with shaking and the vial is the
6warmed to 40˚C and maintained at that temperature for two hours. The
7solution is evaporated to dryness in a stream of nitrogen.
8
9The Lewis acid boron trifluoride or the equivalent reagent boron
10trichloride is also very useful for forming ester derivatives. Boron
11trifluoride is supplied as a 14% solution in methanol. Boron trifluoride
12catalyzed reactions are very fast and can be complete in a few minutes.
131 to 15 mg of the acid are placed in a vial fitted with a ground glass
14stopper and 1 ml of 14% boron trifluoride in methanol added. The
15mixture is heated on a water bath for 2 minutes and then cooled. The
16esters can be extracted with n-heptane with vigorous shaking. Care
17must be taken to extract all the derivative.
18
19The complimentary form of derivatization would be the esterification
20of an involatile alcohol. The normal reagent used for this purpose is an
21acid anhydride which also removes the water as the esterification
22proceeds. There is some competition between the alcohol and the water
23for the anhydride if the conditions are optimized, the anhydride reacts
24preferentially with the water. 10 to 100 mg of acid are placed in a
25stopped vial and 1 molar equivalent of the alcohol is added together
26with 1.2 to 1.4 equivalents of trifluoroacetic anhydride. The mixture is
27warmed and the reaction proceeds rapidly to completion in about 10
28minutes.
29
30One very popular esterifying reagent is diazomethane.
31
32However, diazomethane is carcinogenic and can be extremely
33unstable. All reactions should be carried out in a fume hood and any
34stored solutions of diazomethane in diethyl ether should be restricted
35to a maximum of 100 ml and kept in a refrigerator. The materials must
36never be overheated as there is a risk of explosion.
37
280
281 67
1Despite the dangers, the reagent is very effective. Providing its use is
2restricted to microscale reactions and sensible precautions are taken, it
3is normally safe to use.
4
5Diazomethane is a yellow gas but is used in the form of an ethereal
6solution. Its reacts with an organic acid in the following manner,
7
8 R—COOH + CH2N2 R—COO–CH3 + N2
9
10When the reaction is complete, the yellow color persists and thus the
11reagent acts as its own indicator. An apparatus, developed by Schlenk
12and Gellerman (11) for esterification with diazomethane is shown in
13figure 37.
P o ta ss iu m H y d ro x id e
S o lu tio n
S o lu tio n
of S am ple
N itro g e n
E th e r
D ia z o m e th a n e
G e n e ra tin g
14 S o lu tio n
15
16Figure 37 Apparatus for Generating Diazomethane for
17Esterification
18
19Nitrogen is bubbled through ether so that the gas is saturated with ether
20and then passed through a vessel containing a solution of N–methyl–N–
21Nitroso–p toluene sulfonamide. When potassium hydroxide, is added
22through a dropping funnel, diazomethane is generated. The nitrogen
283
284
285 68
1containing the diazomethane is passed into a solution of the sample

2until a yellow color persists. The sample solution consists of 0.5 to 30
3mg of acid dissolved in 2 ml of a 10% solution of methanol in diethyl
4ether. About 4 ml/ min. of nitrogen is used and after the reaction is
5complete, the solvent is removed by evaporation. Due to the
6methylation of the hydroxyl groups the procedure does not work well
7with phenolic acids. Diazoethane can also be used employing a similar
8technique.
9
10Silyl reagents will react with both alcohols an acids to form
11trimethylsilyl ethers and trimethylsilyl esters respectively. These
12derivatives are volatile and for the most part, are easily separated. The
13two most popular reagents are N,N–bis(trimethyl–silyl)trifluoro-
14acetamide (BSTFA) and bis(trimethylsilyl)–acetamide (BSA). Each
15react rapidly with organic acids to give high yields; the latter reagent is
16often used when an electron capture detector is employed. A few
17milligram of the acid is placed in a vial and about 50 µl of BSA or
18BSTFA added. Reaction can be expected to be complete directly on
19solution, but the mixture can be heated for 5 to 10 min. at 60˚C to
20ensure that reaction is really complete. The reaction mixture can be
21injected directly into the gas chromatograph.
22
23For GC/MS analyses Tert-butyldimethylsilyl esters (TBDMS) are
24recommended. The TBDMS esters are prepared by dissolving about 5
25mg of the acid in 100 µl of dimethylformamide containing 20 µmol of
26imidazole and 10 µmol of TBDMS. The mixture is then heated to 60˚C
27for about 15 minutes, an equal volume of 5% NaCl is added and the
28esters extracted with 1 ml of ether.
29
30 Acylation Reactions
31
32Acylation is also a popular reaction for the production of volatile
33derivatives of highly polar and involatile organic materials. The
34technique, however, has a number of other advantages. In addition to
35improving volatility, acylation reduces the polarity of the substance
36and thus can improve the peak shape and, reduce peak tailing. As a
37consequence amide esters are usually well separated with symmetrical
288
289 69
1peaks. By inserting protecting groups into the molecule, acylation also

2improves the stability of those compounds that are thermally labile.
3Acylation can render extremely polar materials such as sugars
4amenable to separation by GC and, consequently, are a useful
5alternative to the silyl reagents. Acylation is frequently used to
6derivatize amines, amides, alcohols, thiols, phenols, enols, and glycols
7etc..
8
9A typical example of anacylation is the reaction between acetic
10anhydride and an alcohol
11
12 R–CO–CO-R + R'–O–H = R–CO–O–R' + R-CO-O-H
13
14About 5 mg the acid is dissolved in 5 ml of chloroform together with
150.5 ml of acetic anhydride and 1 ml of acetic acid and is heated for 2-
1616 hours at 50˚C. The excess reagents are removed under vacuum and
17the residue dissolved in chloroform and injected directly onto the
18column. Sodium acetate can be used as an alternative to acetic acid in
19which case, 0.3 ml of acetic anhydride is added to 12 mg of sodium
20acetate. The reaction is carried out at 100˚C for about an hour, excess
21reagent is removed by evaporation and the residue taken up in a
22suitable solvent for analysis. An excellent handbook describing a wide
23range of procedures used to produce derivatives for chromatographic
24analysis has been compiled by Blau and Halket (12).
25
26Preparative Gas Chromatography

27
28Gas chromatography has not been used extensively for preparative
29work although its counterpart, liquid chromatography, has been
30broadly used in the pharmaceutical industry for the isolation and
31purification of physiologically active substances. There are a number of
32unique problems associated with preparative gas chromatography.
33Firstly, it is difficult to recycle the mobile phase and thus large volume
34of gas are necessary. Secondly, the sample must be fully vaporized
35onto the column to ensure radial distribution of the sample across the
36column. Thirdly, the materials of interest are eluted largely in a very
291
292
293 70
1dilute form from the column and therefore must be extracted or

2condensed from the gas stream which is also difficult to achieve
3efficiently. Finally, the efficient packing of large GC columns is
4difficult. Nevertheless, preparative GC has been successfully used in a
5number of rather special applications; for example the isolation of
6significant quantities of the trace components of essential oils for
7organoleptic assessment.
8
9The layout of a preparative gas chromatograph is shown in figure38
10
R e du cin g
G as T nnk V a lve
S a m ple
S y rin g e P u m p V a po riz e r
P re p a ra tive C o lu m n
F ra ctio n V a lve
D e te cto r
F ra ctio n
11 R e se rvo irs
12
13Figure 38 Layout of Preparative Gas Chromatograph
14
15Air can not normally be used as the mobile phase due to likely
16oxidation and so either a gas tank or a gas (e.g., nitrogen) generator
17must be used. As the flow rates can be large, more than one generator
18operating in parallel will often be necessary. The sample is usually
19placed on to the column with a syringe pump and rapidly vaporized in
296
297 71
1a suitable heater. Passing the gas in vapor form onto the column helps
2evenly distribute the sample radially across the column. The detector
3that is used must have specifications that are almost opposite to those
4of an analytical detector. It should function well at high concentrations
5of solute, have a generally low sensitivity, if in-line it must be non-
6destructive and have minimum flow impedance. It need not have a
7particularly linear response. The katharometer is one of the more
8popular detectors for preparative GC. The column outlet is passed to a
9selection valve that diverts the eluent to its appropriate collecting
10vessel. The collecting vessel may be cooled in ice, solid carbon dioxide
11or if necessary liquid nitrogen (liquid nitrogen can only be used if a
12low boiling gas such as helium is employed as the carrier gas). In some
13cases the solutes contained in the eluent can be extracted into an
14appropriate liquid or onto the surface of a suitable adsorbent. the
15desired fractions are then recovered by distillation or desorption.
16
17The maximum pressure that can be tolerated by large diameter columns
18is considerably less than their analytical equivalents. Thus to allow
19adequate gas flow rate through the column, the particle diameter of the
20packing must be relatively large. In turn, this means that the efficiency
21obtainable from preparative GC columns is relatively low and, thus, for
22effective separations, the stationary phase must be chosen to have the
23maximum selectivity for the solutes concerned. Compared with
24analytical GC, preparative GC can be far more difficult, The challenge
25is to achieve both adequate resolution and a satisfactory throughput.
26
27The Moving Bed Continuous Chromatography System
28
29The concept of the moving bed extraction process was originally
30introduced for hydrocarbon gas adsorption by Freund et al. (13) and
31was first applied to gas liquid chromatography by Scott (14). A
32diagram of the moving bed system suitable for GC was proposed by
33Scott and is shown in figure 39.
34
35The feasibility of this process was established for a gas
36chromatographic system, subsequently, its viability was also confirmed
299
300
301 72
1for liquid chromatography which will be discussed in Book 19. The

2moving bed system takes a continuous sample feed and operates in the
3following way. The stationary phase, coated on a suitable support, is
4allowed to fall down a column against and upward stream of carrier
5gas. In the original device of Scott, the packing (dinonyl phthalate
6coated on brick dust) was contained in a hopper at the top of the
7column and was taken off from the bottom the column by a rotating
8disc feed table and returned to the hopper by a simple air-lift device.
9
E lu tio n o f
Bed L e s s R e ta in e d
M o ve m e n t S o lu te s
B
S a m ple
Feed
A
C E lu tio n o f S tro n g ly
R e ta in e d S o lu te s
D
H e a te d
S e ctio n
R e co ve ry G a s
F lo w
10
11
12 Courtesy of Butterworths Scientific Publications Ltd. [Ref. 10]
13
14Figure 39 The Moving Bed Continuous Chromatography System
15
16By suitable adjustment of the relative rates of upward carrier gas flow
17and downward stationary phase flow (contained on the falling support)
18some components were arranged to move upward with the carrier gas,
19and others move downwards with the stationary phase. Referring to
20figure 39, if the ordinary chromatogram of the mixture is that depicted
21at (A), the relative speed of the carrier gas and the stationary phase
22defines an imaginary line on the chromatogram. Those components to
23the left of the line, move up with the carrier gas (B) and those
24components to the right of the line, move down with the stationary
304
305 73
1phase (C). The components that move down in the stationary phase are
2stripped out by arranging a portion of the column to be heated and a
3second stream of gas elutes them through a second port (D). Scott and
4Maggs designed a three stage moving bed system to extract pure
5benzene from coal gas. Coal gas contains a range of saturated aliphatic
6hydrocarbons, alkenes, naphthenes and aromatics (benzene, toluene
7and xylenes). The separations they obtained are shown in figure 40.
8
9It is seen that the material stripped from the top section contained the
10alkanes, alkenes and naphthenes and very little benzene. The material
11stripped from the center section consisted of almost pure benzene. The
12residue striped from the lower section contained the toluene, the
13xylenes and even the thiophene which elutes closely to the benzene. To
14eliminate the thiophene, however, it was necessary to loose some
15benzene to the lower stripping section. Nevertheless the separation
16clearly demonstrates the effective use of the moving bed extraction
17technique.
18
307
308
309 74
E x tra ct fro m E x tra ct fro m E x tra ct fro m

S e ctio n 1 S e ctio n 2 S e ctio n 3
Benzene
Benzene
T olu e n e
Benzene
X y le n e s
1
2Figure 40 The Extraction of Pure Benzene from Coal Gas by
3continuous Extraction Using a Moving Bed Technique
4Applications
5
6Gas chromatography has a very wide field of application but its first
7and main area of use is in the separation and analysis of multi
8component mixtures such as essential oils, hydrocarbons and solvents.
9Intrinsically, with the use of the flame ionization detector and the
10electron capture detector (which have very high sensitivities) gas
11chromatography can quantitatively determine materials present at very
12low concentrations. It follows, that the second most important
13application area is in pollution studies, forensic work and general trace
14analysis.
15
16Gasoline is a multicomponent mixture containing a large number of

17hydrocarbons, many of which have very similar molecular weights and
18all are almost exclusively dispersive in interactive character. The
312
313 75
1structure of many of the hydrocarbons are also very similar and there
2are many isomers present. As a consequence, due to their interactive
3similarity the separation factors between individual components is very
4small. It follows that columns of very high efficiency will be
5mandatory to achieve an effective separation. It is clear that open
6tubular columns are ideal for this type of separation problem. In fact, it
7would be impossible to separate the components of gasoline efficiently
8with a packed column, even one that is 50 ft long, and even if the
9inherent long analysis times could be tolerated. In addition this type of
10separation demands the maximum number of theoretical plates and
11therefore not only must open tubes be used but tubes of relatively small
12diameter to produce the maximum number of theoretical plates. In fact,
13several hundred thousand theoretical plates will be necessary and so
14the column must also be very long. As has already been discussed, it is
15necessary to use small radius open tubular columns with a split
16injection system. Furthermore, as a result of the wide range of
17molecular weight of the components present, gasoline has a relatively
18wide boiling range and so will require a temperature program that will
19heat the column to 200 ˚C or more. A thermally stable stationary phase
20must be employed. The individual gasoline components are present
21over a wide concentration range and thus, for accurate quantitative
22results, the linear dynamic range of the detector must also be large.
23These latter demands mandate that the detector must be the FID. The
24separation of the gasoline components is shown in figure 41. The
25stationary phase used was Petrocol (the trade name for a special
26poly(dimethysiloxane)) that is actually intra-column polymerized and
27thus bonded to the surface and, as a result, is very thermally stable. The
28alkane chains in the polymer contribute strong dispersive properties to
29the stationary phase. The necessary high efficiency was obtained by
30using a column, 100 m long, 250 µm I.D. carrying a film of stationary
31phase 0.5 µm thick.
32
315
316
317 76
1
1/ Isobutane 2/ n-Butane 3/ Isopentane
4/ n-Pentane 5/ 2,3-Dimethylbutane 6/ 2-Methylpentane
7/ 3-Methylpentane 8/ n-Hexane 9/ 2,4-Dimethylpentane
10/ Benzene 11/ 2-Methylhexane 12/ 3-Methylhexane
13/2,2,4Trimethylpentan 14/ n-Heptane 15/ 2,5-Dimethylhexane
e
16/ 2,4-Dimethylhexane 17/2,3,4Trimethylpentan 18/ 2,3-Dimethylhexane
e
19/ 2,3-Dimethylhexane 20/ethylbenzene 21/ m-Xylene
22/ p-Xylene 23/ o-Xylene 24/ -Me-3-Ethylbenzene
25/ 1,3,5TriMe-benzene 26/ 1,2,4TriMe-benzene 27/ 1,2,3TriMe-benzene
28/ Naphthalene 29/ 2-Methylnaphthalene 30/ 1-Methylnaphthalene
31/ Dimethylnaphthalene
3 41. A Chromatogram of Gasoline
Figure
320
321 77
1 The column was held at 35oC after in injection for 15 min. and then
2programmed to 200oC at 2oC/min and finally held at 200oC for 5 min.
3To ensure that there was no condensation in the detector, the FID was
4held at 250 oC (50oC) above the maximum column temperature . The
5sample size was 0.1 µl which was split 100-1 onto the column and so
6the total charge on the column was about 1 µg. Helium was used as the
7carrier gas at a linear velocity of 20 cm/sec. The value of the open
8tubular column is clearly demonstrated. An example of the use of the
9packed column in natural product analysis is the separation and
10determination of the free fatty acids in whole milk.
11
1/ n-Valeric Acid 2/ n-Caproic Acid 3/ n-Caprylic Acid
4/ n-Capric Acid 5/ n-Lauric Acid 6/ n-Myristic Acid
7/ n-Palmitic Acid 8/ n-Stearic Acid 9/ n-C16-1 ene
10/ n-Oleic Acid 11 /n-Linoleic Acid 12/ n-Linolenic Acid.
13
323
324
325 78
1 Figure 42. The Separation of the Free Fatty Acids from Milk
2
An3 example of such an analysis is shown in figure 42.
4This analysis requires a rather lengthy procedure for sample
5preparation but, at the same time, avoids a derivatization procedure that
6can easily give incorrect, low values for the fatty acid content. Due to
7their relatively high volatility, the lower fatty acids can be lost as vapor
8during the procedure. Losses can also occur as a result of their
9incomplete derivatization. The sample preparation developed by
10Supelco involved mixing 10 ml of the milk with 10 ml ethanol, 3 ml of
1128% ammonium hydroxide, 25 ml of petroleum ether and 25 ml of
12diethyl ether. The mixture is then well shaken and allowed to stand for
13about 20 minutes. The ether phase is separated and carefully
14evaporated to dryness under a stream of nitrogen. The residue is treated
15with 3 ml of 0.5n NaOH in methanol and heated on a steam bath for 15
16minutes. 5 ml of water is added and then 2N HCl until a pH of about
172.0 is reached. The fatty acids are then extracted with a mixture of 5 ml
18of petroleum ether and 5 ml of diethyl ether. If a quantitative estimation
19is required, then an internal standard would be added and the solution
20diluted to a known volume and an aliquot placed on the column.If an
21external standard is used, then the extract is merely diluted to a known
22volume (e.g., 10 ml) and an aliquot placed on the column. This method
23could be considered as typical of the preparation procedures used in
24GC. It is clear that there can be considerably more time spent on the
25sample preparation than on the actual separation itself. This type of
26separation, however, lends itself to automation either appropriately
27designed hard-wired equipment or by the use of a laboratory robot.
28The hard wired device is generally inflexible, the laboratory robot, on
29the other hand, can be programmed to carry out many different types of
30analysis.
31
32The separation itself has some interesting properties. Free acids are
33very readily adsorbed onto active sites on the support which can result
34in very asymmetric peaks and, as a result of the strong adsorption,
35significant quantitative losses can occur. In the above example, the
36effect of the adsorptive sites on the support is reduced by blocking
328
329 79
1them with phosphoric acid. Phosphoric acid is very involatile and thus
2can tolerate the high temperature and although it is active enough to
3block the adsorption sites it is not active enough to cause sample
4decomposition. It is seen that the peaks exhibit excellent symmetry for
5free acids. Teraphthalic acid has also been used for this purpose to
6deactivate the support. The column was glass, 3 m long and 2 mm in
7diameter and packed with a silicone polymer, SP-216-PS on 100/120
8mesh Supelcoport which is a proprietary support that has already been
9deactivated and treated with phosphoric acid. The column was
10temperature programmed from 130oC to 200oC at 15oC/min. Nitrogen
11was used as the carrier gas at a flow rate of 20 ml/min.. The separation
12is effective, relatively rapid, and accurate quantitative results should be
13easily obtainable from the system. This analysis also demonstrates the
14need for rapid sample preparation techniques as well as rapid
15separations. Fast chromatography is of little use if the chromatograph is
16idle for long periods between samples while they are being prepared.
17
18 Lime Oil
19
20The use of modern GC techniques to separate a sample of lime oil is
21shown in figure 43. A SB–5 column, that contained poly(5%diphenyl-
2295%–dimethylsiloxane) as the stationary phase was used to carry out
23the separation. It is largely a dispersive stationary phase, although the
24diphenyl group will contribute some induced polarizability capability
25to interact with polar solutes. As a consequence substances are eluted
26roughly in order of their boiling points (excepting very polar solutes).
27
28
29The introduction of the diphenyl groups contributes more to phase
30temperature stability than it does to solute selectivity. The column was
3130 m long, 250 µm I.D. carrying a film 0.25 µm thick of stationary
32phase. Helium was used as the carrier gas at a linear velocity of 25
33cm/sec(set at 155˚C). The column was held isothermally for 8 min. at
3475˚C and then programmed up to 200˚C at 4˚c/min. and finally held at
35200˚C for 4 min. The sample volume was 0.5 µl which was split at
36100:1 ratio allowing about 5 µg to be placed on the column. It is seen
331
332
333 80
1from figure 43 that a very good separation is obtained that

2convincingly confirms the complex nature of the essential oil. In
3practice, however, the net flavor or odor impact can often be achieved
4by a relatively simple mixture of synthetic compounds.
5
1 11
5 6 7 8
3
4 18
10
17
2 16
9
13 15
12 14
0 4 8 12 16 20 24 28 32
6 R e te n tio n T im e (m in u te s )
7
1. a–Pinene 7. γ–Terpinene 13. Geraniol
2. Camphene 8. Terpinolene 14. Neryl Acetate
3. β–Pinene 9. Linalool 15. Geranyl Acetate
4. Myrcene 10. Terpinene–4–ol 16. Caryophyllene
5. p–Cymene 11. α–Terpineol 17. trans–α–Bergamotene
6. Limonene 12. Neral 18. b–Bisabolen
8
10
11Figure 43 A Chromatogram of Lime Oil
12
13
14 The Head space Analysis of Tobacco
15
16Tobacco is a herbaceous plant, the leaves of which are harvested, cured
17and suitably prepared for smoking, as cigars or cigarettes, or
336
337 81
1alternatively, chewing or taken as snuff. Its main component, nicotine

2is habit forming and other compounds produced by pyrolysis during
3smoking are carcinogenic and can cause a number of other health
4problems. Tobacco is an extremely valuable export in the United States
5despite the health concern, and its quality is carefully monitored.
6Tobacco can be flue cured, air cured, fire cured or sun cured, but the
7quality of the product can often be monitored by analyzing the vapors
8in the head space above the tobacco. The head space over tobacco can
9be sampled and analyzed using a Solid Phase Micro Extraction (SPME)
10technique. The apparatus used for SPME is shown in figure 44.
11
12The extraction apparatus consists of a length of fused silica fiber,
13coated with a suitable polymeric adsorbent, which is attached to the
14steel plunger contained in a protective holder. The steps that are taken
15to sample a vapor using the apparatus are represented in figure 44.
16
1 2 3 4
P lu n g e r
F ib e r
H e ad Space
C a r r ie r G a s
F ib e r
S a m p le
H e ate r
C a p illa r y C o lu m n
17
339
340
341 82
1
3Figure 44. The Solid Phase Micro Extraction Apparatus
4
5The sample is placed in a small head space vial and allowed to come to
6equilibrium with the air (1). The needle of the syringe containing the
7fiber is the made to piece the cap, and the plunger pressed to expose the
8fiber to the head space vapor. The fiber is left in contact with air above
9the sample for periods that can range from 3 to 60 minutes, depending
10on the nature of the sample (2). The fiber is then removed from the
11vials (3) and then passed through the septum of the injection system of
12the gas chromatograph into the region surrounded by a heater (4). The
13plunger is again depressed and the fiber, now protruding into the heater
14is rapidly heated to desorb the sample onto the GC column. It is
15advisable to arrange for the column is kept cool so the components
16concentrate on the front of the column.
17
1 . B e n z a lde h y de
2 . 6 – M e th y l– 5 – h e p te n e – 2 – o n e 10
3 . P h e n y la ce ta ld e h y d e
4 . N in a n a l
5 . M e n th o l 6
6 . N ico tin e 7
7 . S o la n o n e
8 . G e ra n y l A ce to n e
9 . β– N i c o t y r i n e
1 0 . N e o p h y ta d ie n e
1 1 . F a m e s y la ce to n e
12. C e m bre n e
8
12
3 4
2 5
1 9 11
5 10 15 20 25 30 35
R e te n tio n T im e (m in u te s )
18
19
344
345 83
1Figure 45 A Chromatogram of Tobacco Head Space

2
3When desorption is complete (a few seconds) the column can then be
4appropriately temperature programmed to separated the components of
5the sample. A chromatogram of the head space sample taken over
6tobacco is shown in figure 45. The procedure as outlined by Supelco
7Inc. is as follows. 1 g of tobacco (12% moisture) was placed in a 20 ml
8head space vial and 3.0 ml of 3M potassium chloride solution added.
9The fiber was coated with polydimethyl siloxane (a highly dispersive
10adsorbent) as a 100 µm film. The vial was heated to 95˚C and the fiber
11was left in contact with the head space for 30 min. The sample was
12then desorbed from the fiber for one minute at 250˚C.The separation
13was carried out on a column 30 m long, 250 µm I.D. carrying a 0.25 µ
14m thick film of 5% phenylmethylsiloxane. The stationary phase was
15predominantly dispersive with a slight capability of polar interactions
16with strong polarizing solute groups by the polarized aromatic nuclei of
17the phenyl groups. Helium was used as the carrier gas at 30 cm/sec.
18The column was held isothermally at 40˚C for one minute and the
19programmed to 250˚C at 6˚C/min. and then held at 250˚C for 2 min. It
20is seen that a clean separation of the components of the tobacco head
21space is obtained and the resolution is quite adequate to compare
22tobaccos from different sources, tobaccos with different histories and
23tobaccos of different quality.
24
25 Food and Beverage Products
26
27Due to the likely contamination of food and beverage products with
28pesticides, herbicides and many other materials that are considered a
29health risk, all such products on sale today must be carefully assayed.
30There is extensive legislation controlling the quality of all human foods
31and drinks, and offensives carry very serious penalties. In addition, the
32condition of the food is also of great concern to the food chemist, who
33will look for those trace materials that have been established to indicate
34the onset of bacterial action, aging, rancidity or decomposition. In
35addition, tests that identify the area or country in which the food was
36processed or grown may also be needed. The source of many plants
347
348
349 84
1(herbs and spices) can often be identified from the peak pattern of the
2chromatograms obtained directly from head space analysis. Similarly,
3unique qualitative and quantitative patterns from a GC analysis will
4often help identify the source of many alcoholic beverages.
5Unfortunately, food analysis involves the separation and identification
6of very complex mixtures and the difficulties are compounded by the
7fact that the components are present at very low concentrations. Thus,
8gas chromatography is the ideal (if not only) technique that can be used
9successfully in food and beverage assays and tests.
10
11The potential carcinogenity of the aromatic hydrocarbons make their
12separation and analysis extremely important in environmental testing.
13However, the aromatics can pose some serious separation problems
14(for example, the m- and p-xylenes) due to the closely similar chemical
15structure and characteristics. The xylene isomers differ in structure
16(although not optically active) have similar spatial differences to pairs
17of enantiomers. It follows, chiral stationary phases that separate
18enantiomers can also be used for separating spatial isomers that are not
19necessarily optically active. Nevertheless, the separation ratios of such
20isomeric pairs (even on cyclodextrin stationary phases) is still very
21small, often in the 1.02–1.03 range. As a consequence, the use of high
22efficiency capillary columns is essential, if reasonable analysis times
23are also to be maintained.
24
352
353 85
1 2
1. B enzene
2 . T o lu e n e
3 . n -N o n a n e
4 4 . p -X y le n e
5 . m -X y le n e
5 6 . E th y lb e n z e n e
6 7 . o -X y le n e
7 8 . S ty re n e
9. C um ene
1 0 . α-M e th y ls ty re n e
3 10
9
0 8 16 24 32 40 48 56 64
1
2
4Figure 46 The Separation of Some Aromatic Hydrocarbons
5
6The separation of some aromatics contained in a mixture of
7hydrocarbons is shown in figure 46. A column 30 m long, 0.25 mm
8I.D., carrying a film of permethylated β-cyclodextrin 0.25 µm thick,
9was used by Supelco for the separation. The column was operated
10isothermally at 50˚C and helium was use as the carrier gas at a flow
11velocity of 20 cm/s. It is seen that baseline separation is achieved for
12the m- and p-xylenes and that the separation ratio for the two isomers
13was about 1.03.
14
15Chiral analysis in the drug industry is now extremely crucial. There are
16two factors that have contributed to the importance of chiral GC in
17drug analysis. Firstly, the critical nature of the enantiomeric character
18of a drug has now been well established (largely arising from the
19thalamide disaster). The Food and Drug Administration, as a
20consequence, has mandated that the physiological effect of both or all
21enantiomers of any drug that can exist in chiral form must be
22determined.
23
355
356
357 86
N o rk e ta m in e
t r 1 1 .4 a n d 1 1 .6 8 m i n
K e ta m in e
t r 1 5 .4 0 a n d 1 5 .9 0 m i n
T im e
1
2 Courtesy of Astec Inc.
3Figure 47 The Separation of the Enantiomers of Ketamine and its
4Metabolites Norketamine and Dehydro-norketamine
5
6Moreover, the chiral purity of any commercially available drug must
7also be monitored and controlled. GC is a natural technique for this
8type of work as many modern drugs have relatively small molecular
9weights and consequently are volatile or can easily be made into
10volatile derivatives. In addition, GC capillary columns can easily
11provide the high efficiencies necessary to separate very similar
12compounds with relatively small separation ratios. Ketamine, was
13recently investigated as a potential drug that would reverse the problem
14of protein metabolism in AIDS patients. Unfortunately, the
15determination of the drug distribution in various body fluids by GC
16analysis was complicated by the presence of two chiral metabolites.
17The analysis was successfully achieved using a 30 m long, 250 µm I.D.
18(a Chiraldex G-TA column) operated isothermally at 160˚C using
19helium as the carrier gas with an inlet pressure of 3 Kg/cm2. The
20method could separate all 6 enantiomers as their trifluoryl acetyl
21derivatives as shown in figure 47. The high efficiencies and the general
360
361 87
1versatility of this stationary phase, that provides strong dispersive and

2polar interactions, makes it especially useful for the separation of
3substances with multiple chiral centers and in the presence of
4metabolites. The use of a 5m retention gap method of injection (see
5page 19) allowed the direct injection of 7 µl of plasma.
6
7Essential oils (flavors and perfumes) also contain many chiral
8compounds and one enantiomer may be entirely responsible for a
9particular taste or odor whereas the complementary enantiomer has an
10entirely different olfactory effect. It is clear that the use of chiral
11chromatography can be one of the more useful techniques for the
12analysis of essential oils. A chromatogram of the essential oil vapor
13from White Pine leaves is shown in figure 48.
14
15A head space sample was taken, employing the method previously
16described using 0.5 g of pine leaves contained in a 7 ml vial. The solid
17state extraction procedure employed a glass fiber coated with a
18polysiloxane film which was exposed to the sample vapor at 40˚C for
1920 minutes. Using the special applicator, the fiber was withdrawn from
20the sample vial and placed in a unique capillary column sample device.
21The fiber was then heated to 250˚C for one minute and the vapors
22passed onto the column using a split injector with a 100:1 split. The
23column used was 30 m long 0.25 mm I.D. and carried a film of β-DEX
240.25 µm thick and was programmed from 40˚C to 220˚C at 4˚C/min.
25Helium was used as the carrier gas at a velocity of 35 cm/s. It is seen
26that the sample is broadly separated into two groups, the monoterpenes
27and the sesquiterpenes. The enantiomers of α-pinene and camphene are
28cleanly separated. As these compounds contain no polar groups, the
29chiral selectivity must be based entirely on differential dispersive
30interactions with the derivatized cyclodextrin.
31
363
364
365 88
M o n o te rp in e s S e s q u ite rp in e s
4
2 7
1 8
1 .( - ) - α - P i n e n e
2 .( + ) - α - P i n e n e
3 . β-M y rce n e
4 .( + ) - S a b i n e n e
5 .( + ) - C a m p h e n e
6 .( - ) - C a m p h e n e 5
7 .( + ) - β - P i n e n e 9
8 .3 - C a r e n e 6
0 10 20 30 40
1
2 Courtesy of Supelco
3
4 The columns were 30 m long, 0.25 mm I.D., carrying a film of stationary phase
5 0.25 µm thick of β-DEX™. The column was programmed from 40˚C to 220˚C at
6 4˚C/min. The helium flow velocity was 35 cm/s.
7
8Figure 6.9 Chromatogram of the Essential Oil From White Pine
9Leaves
10It should be noted that whereas the (–)-α-pinene is the first eluted
11enantiomer of α-pinene it is the (+)-camphene that is the first eluted of
12the camphene enantiomers. This tends to indicate that there is no
13rational procedure for predicting the order of elution of an
14enantiomeric pair.
15
16References
17
18 1. A. J. P. Martin and R. L. M. Synge, Biochem. J., 35 (1941)1358.
368
369 89
1 2. A. T. James and A. J. P. Martin, Biochem. J., 50 (1952) 679.

2 3. A. T. James, The Times Science Review, Summer (1966)8.
3 4. D. H. Desty, A. Goldup and B. F. Wyman, J. Inst. Petrol.,
4 45(1959)287.
5 5. R. D. Dandenau and E. M. Zenner, J. High Res. Chromatogr.,
6 2(1979)351.
7 6. K. L. Ogan, C. Reese and R. P. W. Scott, J. Chromatogr. Sci.,
8 20(1982)425.
9 7. J. Harley, W. Nel and V. Pretorious, Nature, London,
10 181(1958)177.
11 8. G. McWilliams and R. A. Dewer, "Gas Chromatography 1958"
12 9. R. P. W. Scott, Nature, London 175(1955)422.
1310 C. H. Reese, Ph. D. Thesis, University of London (Birkbeck
14 College) 1992.
1511. 1.H. Schlenk and J. L. Gellerman, Anal. Chem., 32(1960)1412.
1612. K. Blau and J. Halket, Handbook of Derivatives for
17 Chromatography, John Wiley and Sons, (1993).
1813. M. Freund, P. Benedek and L. Szepesy,Vapour Phase
19 Chromatogrphy, (Ed. D. H. Desty) Butterworths Scientific
20 Publications, London, (1957)359.
2114. R. P. W. Scott, Gas Chromatogrphy 1958, (Ed. D. H. Desty)
22 Butterworths Scientific Publications, London, (1958)287.
23
371
372

GAS Chromatography: Chrom-Ed Book Series

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

GAS Chromatography: Chrom-Ed Book Series

Uploaded by

Copyright:

Available Formats

1 1

5 Chrom-Ed Book Series

5COPYRIGHT @2003 by LIBRARY4SCIENCE, LLC

1Preparative Gas Chromatography.........................................................69

1A more sophisticated form of the gas chromatograph was constructed

1of symposia. The layout of the modern gas chromatograph is shown as

1the actual separation is achieved either in a relatively short length of

1processing computer which processes the data and prints out an

1the down stream sensor is heated, producing a differential temperature

30Equation (7) can now be employed to calculate the change in retention

1The basic injection devices that are used in chromatography, such as

1Procedures have been introduced in an attempt to eliminate sample

1gap procedure is normally used in conjunction with temperature

1located at the beginning of the column. A diagram of the solute

1water followed by methanol, acetone, methylene dichloride and n-

1fire-brick (calcined Celite), fire-brick coated with metallic silver or

1to be the worst compromise between a column packed with modified

1solvent to provide excess liquid when mixed with a weighed amount of

1an opening at the top of each U which was terminated in a plug of

17/ 1,2-Dichloropropane 34/ 1,2-Dichlorobenzene

6Figure 21. The Separation of Some Chiral Amines

1A large number of GC detectors have been developed and made

3Figure 22. The Flame Ionization Detector

1An example of the use of the FID in a paraffin, isoparaffin, aromatic,

18Figure 23. The Separation of a PIANO Standard Mixture

20The Nitrogen Phosphorus Detector (NPD)

1The nitrogen phosphorus detector (NPD), is a highly sensitive but

10Figure 24. The Nitrogen Phosphorus Detector

1produce an ion current. When a solute containing nitrogen or

24Figure 26. Wave form of Electron Capture Detector Pulses

1positive ions (produced simultaneously with the electrons by the β

1The basic electron capture detector consists of a small chamber one or

1Figure 28. The Analysis of Priority Pollutant Pesticides

3The column used was a SPB-608 fused silica capillary column, 30 m x

10The Katherometer Detector

4Figure 30. The Separation of the Compounds of Hydrogen,

1various compounds of hydrogen, deuterium and tritium, employing gas

8The four detectors described are well established. reliable and

15Data Acquisition and Processing

S ca lin g P o te n tio m e tric

21The A/D Converter

1described. A diagram showing the operating principle of an

8The converter consists of an integrator that can be constructed from an

1used was not appropriate. These restrictions were entirely a result of

21 1. The preparation of the sample.

1Liquid extraction is a clumsy procedure, particularly when used on the

1tributylether (MTBE), 15 ml of methanol and finally 125 ml of

1temperatures without thermal decomposition or molecular re-

1 Another useful catalyst for esterification is thionyl chloride but the

1containing the diazomethane is passed into a solution of the sample

1peaks. By inserting protecting groups into the molecule, acylation also

26Preparative Gas Chromatography

1dilute form from the column and therefore must be extracted or

1for liquid chromatography which will be discussed in Book 19. The

E x tra ct fro m E x tra ct fro m E x tra ct fro m

16Gasoline is a multicomponent mixture containing a large number of

1from figure 43 that a very good separation is obtained that

1alternatively, chewing or taken as snuff. Its main component, nicotine

1Figure 45 A Chromatogram of Tobacco Head Space

1versatility of this stationary phase, that provides strong dispersive and

1 2. A. T. James and A. J. P. Martin, Biochem. J., 50 (1952) 679.

You might also like