Chapter I

Basic principles of gas chromatography

The development of gas chromatographic (CC) methods has led to revolutionary

changes in analytical chemistry and also in experimental methods of physical chemistry
and methods for the separation of volatile compounds, as GC has a number of important
advantages over other methods [ l ] .
Firstly, GC is universal and can be applied to the analysis of a very wide range of
substances from hydrogen isotopes to oligomers and metals, with boiling points within
the range -250-1000°C. The method enables one to obtain, in a single analysis,
information not only on a single substance but also on the contents of all (or most) of
the components present in a complex mixture.
Secondly, the practical application of the method is simple, the equipment used being
of a standard type with automatic recording of the results of an analysis.
Thirdly, GC ensures a high efficiency of separation within a relatively short analysis
time (1 -30 min).
The development of GC is still continuing. About 1500-2000 papers are published
annually in this field, and the total number of publications up to 1970 was about 20,000 [2].
A study of the published information indicates the major role of gas chromatography
among methods for the analysis of organic compounds and gases. Thus, for instance, in
1974, the papers in the principal analytical journals (Analytical Chemistry, Zeitschriff
fur Analytische Chemie, Analyst (London),Zhurnal Atialiticheskoi Khimii and
Zauodskaya Laboratoriya) were distributed as follows: all types of chromatography,
58.8%; GC, 34.6%;spectral methods, 25.7%; electrochemical methods, 6.6%; other
methods (colunietric, gravimetric and others), 8.9%. Thus, in the analytical chemistry of
organic compounds and gases, the total number of papers on GC exceeds those on
spectral methods, and one in three papers is based on GC.
The development of CC began after James and the Nobel Laureate Martin published
the first work on gas-liquid chromatography [3] . Chromatography as a general method
of separation, however, was first discovered in 1902 by the Russian botanist Mikhail
Seniyonovich Tsvet (1 872-1919), who proposed an adsorption chromatographic method
of separation in the liquid phase and described its application to the analysis of
chlorophyll in plants [4].
The basic units of the chromatograph are the chromatographic column and the
detector. The column (Fig. 1 .I, C) separates the test mixture into its components and
the detector (Fig. 1 . l , D) records (in the carrier gas flow) the concentrations of the
separated Components. The results of the separation are recorded automatically.
Fig, 1.I shows schematically the separate stages of the chromatographic separation of
a three-component mixture, illustrating the positions of the chromatographic zones in the
column at definite time intervals, and the relationship between the separation process
and the recorded chromatogram. At the moment of injection of the test mixture, the
zones of all three substances are located at the head of the column. Under the action of
the flow of carrier gas, the components of the mixture begin to move along the column

Keferences p. 29
 :
2 BASIC PRINCIPLES 01;GC

(a)

i t t t i t

1:ig. 1.1. Schematic representation of the Chromatographic separation of a three-component mixture.

(a) Dynamics of chromatographic process (position of chromatographic zones in column at definite
time intervals); (b) recorded chromatogram.

at different speeds, which are determined by the nature of the compounds being
separated and the type of sorbent used. The first t o be eluted from the column is
component 1 (sloping hatching) then component 2 (vertical hatching) and finally
component 3 (horizontal hatching).
Separation in GC is based on different distributions of the molecules of the com-
ponents being separated between the mobile gas phase and the stationary phase. A dynamic
equilibrium is established between these phases for each component of the test mixture.
Under the action of the flow of carrier gas, the components of the test mixture move
along the chromatographic column with different speeds. The speed of this motion
depends, for each component, on its distribution constant between the gas and stationary
phases. The speed of motion of a chromatographic zone is inversely proportional to the
distribution constant, i.e., readily sorbable Components move along the sorbent layer
more slowly that sparingly sorbable components.
A quantitative description of the elution process in GC can be obtained most readily
by kinetic treatment of the e!ementary processes of the motion of the molecules of the
 BASIC PRINCIPLES OF C C 3

test compounds in the column. It is assumed that the following conditions are fulfilled
in a chromatographic separation: (1) the molecules of the test compounds are in dynamic
equilibrium between the gas and stationary phases, and this equilibrium is independent
of the presence of other components in the sample; (2) the molecules of the test
conipounds move along the column only in the gas phase; ( 3 ) the carrier gas velocity, the
temperature and the properties of the sorbent are constant along the length of the
column and across its section, and the pressure drop can be neglected.
In the course of separation in the column, a definite proportion of the niolecules of a
given coniponent is in the gas phase at any instant, namely nm/(n, + n m ) , where nm and
n, are the numbers of molecules of the given component in the mobile and stationary
phases, respectively. Consequently, if the total retention time of this type of molecule
in the column is t R , the average retention time of a molecule in the gas phase is
tRn,,/(n, + nm). As the molecules are moving along the column only in the gas phase,
the molecules of the test compound will pass through a Column of length L , with a mean
carrier velocity zl, within a time f, (n,/n,+ n,). Therefore, the following equation holds:

Using eqn. 1.1, one can determine the retention time as the average time for the
passage of the molecules of the test substance from the head to the end of the column:
L n
tR = Y - ( l + 2 )
Id
m
Considering that the gas hold-up is L/ii and nJnm = K,(V,/V,), where K , is the
distribution constant of the substance between the liquid and gas phases and V,/V,, is
the ratio of the volume of the stationary phase t o that of the mobile phase in the colurnn,
we obtain, after simple rearrangements:

or

where tN is the net retention time and F is the volume velocity of the carrier gas as
measured at the column temperature.
Eqn. 1.4 can be used to obtain an expression for the net retention volume, V N . The
retention volume is the volume of the carrier gas necessary to elute a compound from
the column under given conditions:

I/,= V,V,=t,F=K,V,
- (1.5)
where VN is the net retention volume, VM is the hold-up volume of the column and VR
is the retention volume (in the absence of a pressure drop).
Hence, under standard experimental conditions, the retention volume (adjusted in
ternis of the hold-up volume of the column) is directly proportional to the distribution

References p. 29
4 BASIC PRINCIPLES OF GC

constant of a given compound between the mobile and stationary phases and is a value
characteristic for a given compound. This means that each substance, independent of its
concentration in the sample, will be eluted from the column after a definite time, which
is characteristic of the substance. The retention time is the same type of constant for a
given compound as other widely used characteristics such as the boiling point and
specific gravity. Eqn. 1.5 substantiates the use of GC in the qualitative analysis of the
components of a mixture and in measuring distribution constants.
Fig. 1.2 exhibits the relationship between the distribution constant and the retention
values of the substances being analyzed. Fig. 1.2a shows the distribution isotherms of
substances A and B, and Fig. 1.2b indicates the position of the chromatographic zones
in the column a certain time after the beginning of the analysis. Substance B is sorbed
better than substance A by the stationary phase (KO,,, KD,*). Therefore, most of the
molecules of B will be in the stationary phase and a smaller proportion in the carrier gas
flow. For substance A, we have the opposite situation. Therefore, the zone of substance
A will move along the column faster than the zone of substance B. At the column outlet,
the separated components of the test mixture proceed, in the carrier gas flow, t o the
detector, the response of which is proportional to the concentration or flow-rate of the
components of the test substance in the carrier gas.
The detector readings are recorded automatically by an electronic potentiometer. The
diagram obtained, which reflects the results of the chromatographic separation, is
called a chromatogram. A typical chromatogram of a hydrocarbon mixture is shown in
Fig. 1.3. On the basis of the chromatogram one can determine the qualitative composition
of the mixture analyzed.
Let us consider the basic elements of a chromatogram [5]. The baseline is the portion
of the chromatogram (for instance, between peaks 1 and 2) obtained when only the
carrier gas is eluted from the column. A chromatographic peak is the portion of the
chromatogram corresponding to the detector signal when one or several components are
eluted from the column. The retention time ( t R ) is the time elapsed from the moment
the sample is injected into the column to the appearance of the peak maximum. The
hold-up time (tM)is the retention time of a compound that is not sorbed by a given
stationary phase. The adjusted retention time It;) is the total retention time less the
hold-up time ( t i = t R - tM).The peak width 010) is the segment of the zero line obtained
(a)
I (b)

0 a

Fig. 1.2. Distribution isotherms (a) and chromatographic separation (b) of a two-component mixture
of A and B. a and c are the concentrations of the compounds analyzed in the stationary and mobile
phases, respectively.
 BASIC PRINCIPLES OF GC 5

\
1I-
U-..L',L
12 8 4
-, 0

Tlme (min)

Fig. 1.3. Parameters of a chromatogram. Separation conditions: chromatograph, Tsvet-4, column

200 x 0.4 cni filled with 10%Apiezon K o n Chronlosorb P; temperature, 6 5 ° C . Peaks: 1 = air;
2 = cyclohexadiene; 3 = cyclohexane; 4 = rnethylcyclohexane. The designations of the parameters
are explained in the text.

by interpolation of the baseline in the interval from the beginning to the end of the peak.
In chromatographic practice, use is generally made of the peak width at half-height
(/A, s), which is easier to determine from the chromatogram. The peak height ( h ) is the
distance from the peak maximum to its base, measured in a direction parallel to the
detector signal axis. The peak area (S) is the area enclosed between the line bounding the
peak and its base.
Qualitative analysis in CC is carried out on the basis of measurement of the retention
times (or retention volumes). It is more convenient to identify unknown components by
using relative, rather than absolute, retention values. The relative retention time (or

References p. 29
6 BASIC PRINCIPLES OF GC

relative retention volume), r,,,, is determined by the equation

- "R.A = 'R,A- tM,A - VN , A - K D,A
r~~~-
tlR,B tR,B-tMB vN9B KO,E
Relative retention values are determined exclusively by the distribution coefficients
of a given compound and the compound used as the standard; they are independent of
such experimental conditions as the carrier gas velocity, the amount of the stationary
phase, the column length and the sample size. The literature contains a great number of
experimental data on relative retention values of various compounds on different
stationary phases (see, for instance, refs. 6 and 7). Peaks of unknown components of the
mixture being analyzed are identified by comparing the relative retention times of the
maxima of these peaks, which are determined directly from the chromatogram, with the
tabulated values for known compounds.
In developing the chromatographing procedure, after the problem of the chromato-
graphic separation of the final sample components has been solved, their quantitative
determination must be carried out.
The size of the signal given by the detector used in GC is directly proportional to the
concentration of a component in the carrier gas under constant experimental conditions:
h(t) = l/Ric(t) (1 *7)
where h is the detector signal, c(t) is the concentration at time t and Ri is a constant.
Therefore the amount, i, of the component being analyzed is directly proportional to the
area of its chromatographic peak:
q i= J c(t)dt = J Rih(t)dt = R, J h(t)dt = R , S (1.8)
The content of the jth component in the mixture can be calculated by the following
equation:
KS..100
p. = 1 1
ZKiS,
This is the simplified scheme for the interpretation of chromatograms. In conclusion, we
shall give some examples that characterize the main types of application of the method.

EXAMPLES OF APPLICATIONS

Analysis of complex mixtures

GC is widely used for the analysis of complex mixtures, beginning with the separation
of methane molecules containing different hydrogen isotopes (Fig. 1.4a [8] ) and ending
with high-boiling oligomers of organosilicon compounds (Fig. 1.4b [9] ). Usually, the
chromatographic analysis lasts a few minutes, tens of minutes or more rarely, hundreds
of minutes. Some mixtures, however, can be analyzed within a few seconds (Fig. 1 . 4 ~[lo]).
It should be noted that GC is used not only for laboratory analyses but also for the
control and regulation of engineering processes. Fig. 1.4d [ 111 shows a chromatogram
 EXAMPLES OF APPLICATIONS I

of an analysis of a sample collected from a flow of a polymerizate of a synthetic

ethylene-propylene rubber (SKEP) using a Soviet-designed process chromatograph (KhP-2 16).
The problems involved in the analysis of complex mixtures have been discussed in
books [12, 131.

Determination of physicocheinical quantities

GC methods are widely used for detennining physicocheniical characteristics such as

distribution coefficients, activity coefficients, heats of solution, heats of adsorption,
adsorbent surface areas, coefficients of diffusion in gas and liquid phases and rate
constants of heterogeneous and homogeneous reactions [ 14-17] .
Fig. 1.5 [ 181 illustrates two chromatograms of the decomposition products of
tert.-butyl hydroperoxide. Chromatograni (a) was obtained at the beginning of the
reaction of the decomposition of the peroxide, and chrotnatogram (b) during the course
of the reaction. The decomposition of the rut.-butyl hydroperoxide leads to shrinking of
the peak of the hydroperoxide (5) and to the appearance of new peaks of the reaction
products on the chromatograni (1-3).
Chromatographic study of changes in the concentrations of the initial compounds
and reaction products with time enables one t o obtain, in most instances, the infomiation
necessary for describing the kinetic behaviour of the system.
GC methods are also widely used for studies of adsorption phenomena and measurement
of the surface areas of soIids. As an exaniple, Table I .1 [ 191 lists data that characteriLe
the good agreement between surface areas measured by classical methods and b y GC.
GC enables one t o obtain within a much shorter time, results that are comparable in
accuracy with those of classical methods.

Preparation techniques

GC can be used for isolating pure components from a mixture. Automated preparative
equipment is available in which sample injection, separation and collection of pre-assigned

rABLk 1 1
COMPARISON or suRr ACL AREAS or CATALYSTS AS D L T L R M I N L D BY CHROMATO-
GRAPHIC AND STATIC MrTHODS I’ROM ARGON ADSORPTION [ 191
Saniple Specific surface area (mz/g) Difference (%)
Volume niethod (on Chromatographic method
vacuiini set-up) of therriial desorption

Chelate polynicr 0.08 0.085 1-6.3

Molybdenum f o 11 0.01
Titanium oxide 4.2 4.2 0.0
Silica gel 20.3 20.3 0.0
A h niin a 135 132 -2.2
Aluniinosilica tc 336 332 -1.2
Silicoii-magnesia catalyst 414 398 -3.9
__ -. -

References p. 29
8 BASIC PRINCIPLES OF GC

(a)

I I I I I I
300 310 320 330 340
Time (mtn)

2 2

1
k
I l
6
I

Time bet)
I I
12
I
60
I
40
Time (min)
I
20
 EQUIPMENT 9

1 - L 1 1 ~~ -1
1 2o 1 J
40 20 40
Time i m i n )

Fig. 1.5. Chromatogram of a solution of ferf.-butyl hydroperoxide in the presence of manganese

stearate (a) at the beginning of and (b) during the reaction. Conditions: column, 120 X 0.4 cm;
sorbent, 30% dinonyl phthalate o n Celite-545; temperature, 50°C. Peaks: 1 = acetone; 2 = fer?.-butanol;
3 = ?err.-butyl peroxide; 4 = toluene (internal standard); 5 = ?err.-butyl hydroperoxide;6 = chlorobenzene
(solvent).

fractions is carried out automatically. In order to obtain pure reactants, including

monomeric solvents, methods of preparative chromatography are used on an industrial
scale [20-231.

GC EQUIPMENT

Fig. 1.6 [24] shows schematically a simple gas chromatograph. The carrier gas
proceeds from a cylinder (1) through a reducer (2), a pressure regulator ( 3 ) and a flow
stabilizer (4) to the reference cell of the detector (6), then through the sample intro-
duction system (7) to the chromatographic column (9), which is located together with

Fig. 1.4. Examples of the application of gas chromatography. (a), Chromatogram of isotopic
molecules of methane [8]. Conditions: capillary glass column (47 m X 0.22 m m I.D.), on the inner
walls of which was a layer of active silica formed as a result of treatment of the glass capillary with a
10% solution of sodium hydroxide at 100°C; temperature, 77°K; carrier gas, nitrogen-helium (7:3);
flow-rate, 1 ml/min. The upper chromatogram was obtained with the use of an ionization chamber
as detector and the lower chromatogram with a flame-ionization detector. Peaks: 1 = I4CH,;
2 = CH,,H; 3 = CH,'H,; 4 = CH'H,; 5 = C3H,; I ' = "CH,; 2' = I3CH,; 3' = CH,IH; 4' = CH 12H1 '3
5' = CH,,H,; 6' = C'H,. (b), Chromatograni of high-boiling organosilicon compounds [ 9 ] :
CH,
(C,H,),CH,Si-fO-Si< 1,-0-SiCH,(C,H,), .Conditions: column, 100 X 0.4 cm; sorbent,
C*HS
6.85%PFMS-6 on INZ-600 treated with dimethyldichlorosilane vapour; temperature, 354°C.
Peaks: 1, n = 0; 2, n = 1; 3, n = 2; 4, n = 3. (c), Chromatogram of the rapid separation of a gas
mixture [ 1 0 ) . Conditions: column, 100 X 0.4 cm; sorbent, 20% molecular sieves CaA o n Celite;
temperature, 20°C. Peaks 1 = hydrogen; 2 = oxygen; 3 = nitrogen; 4 = methane; 5 = carbon
monoxide. (d), Chromatograms of the analysis of a flow of the polymerizate of an ethylene-
propylene copolymer n4th a KhP-216 process chromatograph. Peaks: 1 = ethylene; 2 = propylene.

References p. 29
10 BASIC PRINCIPLES OF GC

----
-7 !----- -1
I
I
4-

I’

Fig. 1.6. Simple gas chromatograph. For explanation, see text.

the detector in a thermostat (10). The pressure at the column inlet is measured by a
pressure gauge (5) and the volume velocity of the carrier gas is checked periodically by
a foam meter (1 1). The sample is injected with a syringe (8) into the flow of carrier gas
upstream of the column through a sample injector (7). The flow of carrier gas carries the
sample to the column where its components are separated into separate zones. The
separated substances (chromatographic zones) enter the detector (6), which determines
the concentration (or mass flow) of the components in the carrier gas. The detector
signal, which is proportional to the concentration (or mass flow), is automatically
recorded by a potentiometer (12). Detailed consideration of the separate units can be
found elsewhere [25,26].

Sample injectors

In laboratory practice, gaseous and liquid samples are usually injected into the
chromatograph with syringes similar to medical syringes. A simple device for introducing
the sample into the chromatograph with a syringe is shown in Fig. 1.7a. The syringe
needle is introduced through a rubber gasket into a heated evaporator through which a
continuous flow of carrier gas is passing. The sample to be injected rapidly evaporates
and is transferred to the chromatographic column in the vapour state.
In chromatography, gaseous samples are usually injected by means of a system with
detachable tubes of known volume [27] , a diagram of which is given in Fig. 1.7b. The
same idea was used in developing an automated system for injecting gaseous samples
with diaphragm valves [28]. For automatic injection of samples from a liquid flow, use is
 EQU IPM I:NT 11

(b)

14

Pig. 1.7. Sample injectors. (a), Sample injector,for introducing samples with a syringe: 1 = carrier gas
inlet; 2 = rubber seal; 3 = heater; 4 = to chromatographic column. (b), Injector forgaseous samples:
1 = Ilow of gas to be analyzed; 2 = Ilow of carrier gas; 3 = chromatographic column; 4 = saniple
injection loop of known volume. (c), Automatic sample injector with moving rod forliquid samples:
1 = seal; 2 = membrane seal; 3 = moving rod; 4 = injection volume; 5 = body; 6 = sample flow;
7 = flow of carrier gas.

made of sample introduction systems with a moving rod (Fig. 1 . 7 ~ )The. role of the
calibrated sample injection volume is played by the channel in the rod. As the rod moves,
a definite volume of the liquid sample filling the calibrated channel is transferred from
the flow of sample into the flow of carrier gas, where the liquid sample is evaporated
and the vapour is carried to the chioinatographic column by the flow of carrier gas.
Capillary colunins require only very small samples for analysis. This problem is
usually solved not by designing miniature sample introduction systems, b u t by using the
flow division method. Thus, only a very small portion of the carrier gas containing part of
the sample is directed t o the capillary column.

References p. 29
12 BASIC PRINCIPLES OF GC

Columns

In a chromatographic column, the components of a mixture are divided into separate

zones. At present, two basic types of chromatographic column are used, namely packed
and capillary columns [29,30]. The packed columns can be subdivided into preparative
columns (diameter greater than 10 mm), analytical columns (diameter 3-6 mm) and
capillary packed columns (diameter 0.5-2.0 mm). The length of the column with the
packing is 0.8-10 m. Capillary columns (diameter 0.2-0.6 mm) are generally used
without a packing, the inner walls being coated with a film of the liquid stationary phase.
The length of capillary columns is 20-100 m. In spite of their high efficiency, capillary
columns are used much more rarely than packed columns because high-efficiency
capillary columns with reproducible characteristics are more difficult to prepare.
The material of construction of chromatographic columns must be resistant to
adsorption and catalytically inert. In most instances, columns made of stainless steel,
glass, polymers and copper are used.
Packed columns are filled with narrow fractions of either solid adsorbents with a
developed surface or solid supports whose surface is coated with a layer of the liquid
stationary phase. In order to reduce the adsorption and catalytic activity, mineral solid
supports are treated with the vapour of dimethyldichlorosilane or hexamethyldisilazane,
which deactivate the hydroxyl groups of the surface. Diatomite supports are used most
often and for polar compounds good results are obtained by using inert polymer supports.
Detailed information on solid supports is given elsewhere [31,32].

Detectors

A detector determines quantitatively the concentration (mass flow) of the test

components in the carrier gas after they have been separated in the chromatographic
column. The characteristics of the detector largely determine the accuracy and sensitivity
of the entire analysis and the detector is therefore one of the most important units of
the chromatographic installation. Hence “the history of development of gas chromato-
graphy is to some extent the history of development of the detector” [33].
Two types of detectors are widely used in GC: concentration detectors, whose
readings depend on the concentration of the substance in the carrier gas, and mass (flow)
detectors, whose readings are determined by the rate of feed (mass flow) of the test
substance carried to the detector in the flow of carrier gas.
The readings of the concentration detector depend only slightly on the flow-rate,
whereas those of the flow-rate detector change sharply with the flow-rate. An example
of the first type of detector is the katharometer, and of the second type the flame-
ionization detector. We shall now consider the different types of detectors.

Katharometer

The principle of operatir. of the katharometer is based on the change in the electric
resistance of the sensor (filament, coil) in relation to the heat conductivity of the gas
leaving the GC column. The heat conductivity of the gas in the low-concentration range
 THEORETICAL CONCEPTS OF THE SEPARATION PROCESS 13

depends linearly on the concentration of the eluted substances in the flow of carrier gas.
Therefore, the resistance of the sensor changes linearly with the concentration of the
detected substance in the flow of gas leaving the column.
Use is generally made of the differential method, in which the working and reference
cells of the detector are wired as a Wheatstone bridge, and a flow of pure carrier gas
passes through tlie reference cell. When pure carrier gas passes through both cells, the
bridge is in equilibrium. When a zone of a substance is eluted from the column, the
composition of the gas mixture in the working cell changes, and so do the temperature
and resistance of the sensor wire, and the potentiometer records the imbalance of the
bridge, which is proportional t o the concentration of the substance in the carrier gas.
The main advantage of the katharometer is its universality. The katharometer can be
used for detecting permanent gases, various inorganic compounds (including such
aggressive coinpounds as nitrogen dioxide, hydrogen chloride and fluoride gases, if a
katharometer of special design is used) and vapours of organic compounds. In quanti-
tative calculations, it is necessary to take into account'that the detector signal depends
on the type of compounds being examined [34-391.
Wide use is also made of the gas density balance [40-421, the sensitivity of which is
slightly lower than that of the katharometer. The gas density balance, however, has the
following advantages over the katharometer: (1 ) no preliminary calibration is necessary
for qualitative analysis; (2) analysis of more aggressive gases is possible, because the
vapours of the test substances do not come into contact with the sensitive elements; and
(3) readily available gases are used as carrier gases.

E'larn~-ionizutb~~
detector

The flame-ionization detector [43,44] is widely applied for detemiining organic

coinpounds and especially impurities. The principle of operation is based on a sharp
decrease in the electric resistance of a hydrogen flame when trace amounts of organic
compounds, which forni ions in tlie course of oxidation, are introduced into it. These
ions are collected at the electrodes, one of which is usually a burner nozzle. The very low
ionization current that then arises is amplified and recorded by a potentiometer. A pure
hydrogen flame usually generates a background current of the order of 10-"-10-12 A;
when the test organic substances are introduced into the flame, currents of 10-12-10-7A
are generated. Fig. 1.8 shows a flame-ionization detector of the type DIP-2.
In recent years, selective ionization detectors [45-481 have been increasingly applied:
the electron-capture detector for determining halogen-containing compounds [49], the
themionic flame detector for phosphorus- and nitrogen-containing compounds [SO]
and the mass spectral detector for most compounds [ S l , 521.

THEORETICAL CONCEPTS OF THE GC SEPARATION PROCESS

I n GC, when the Lones of the test compounds are nioved along the sorbent layer by
the flow of carrier gas, two opposite effects occur simultaneously; the distance between
the concentration maxima of the chromatographic zones of adjacent components

References p. 29
14 BASIC PRINCIPLES O F GC

36 mrn
I I

Fig. 1.8. Flame-ionization detector, type DIP-2: 1 = body; 2 = electrode/burner; 3 =diffuser for air
supply; 4 = electrode/collector; 5 = upper detachable cup; 6 = air inlet orifice; 7 = inlet of eluate with
hydrogen.

increases (this effect improves the separation) and so does the width of the chromato-
graphic zones (this effect impedes the separation).
The theory of GC explains the relationship between separation and the experimental
parameters, and also the observed regularities in the two basic chromatographic
characteristics, i.e., the retention value and peak broadening.
 THEORETICAL CONCEPTS OF THE SEPARATION PROCESS 15

The retention volume (see eqn. 1.5) is directly proportional to the distribution
constant of the test c o n ~ p o u n dbetween the liquid and gas phases and t o the volume of
the liquid stationary phase in the column. This equation has been used for determining
the distribution constant of organic con~poundsin the gas-liquid system. The values of
the distribution constants are in good agreement with the corresponding values obtained
by the static method [53], thus supporting the validity of eqn. 1.5, which is the basic
equation of GC.
It should be noted, however, that in the general case, when the distribution isotherpi
is non-linear, the distribution constant in the retention volume equation is a function of
the concentration. In this instance the retention volume also depends on the concen-
tration, and this effect leads t o the formation of asyminetrical Chromatographic zones.
When determining retention values in CC it is necessary to take into account the
compressibility of the carrier gas [3], due to which the flow-rate, pressure and density
of the carrier gas vary according t o a definite law along the length of the column. Let us
determine the retention value, making an allowance for the compressibility of the
carrier gas. Taking into account eqn. 1.3, we can write the following expression for the
retention time:

(1.10)

To calculate the integral, we shall use Boyle’s law:

UP = UOPO (1.1 1)
and Poiseuille’s l a w

(1.12)

where u and p are the linear velocity and pressure of the carrier gas at some point x in
the chromatographic column, uOand p o are the linear velocity and pressure, respectively,
of the carrier gas at the column outlet, kpem,,is the permeability constant and 77 is the
viscosity of the carrier gas.
Expressing u and dw from eqns. 1.1 1 and 1.12, we can calculate the integral

(1.13)

where p i is the pressure at the column inlet. The value of the term kpenn./77u0p0can be
found from the equation

(1.14)

or
kpenn. = 2L
- - (1.15)
77UOPO P’- P i

References p. 29
16 BASIC PRINCIPLES OF GC

Thus

(1.16)

and hence

(1.17)

or

(1.18)

where j is the pressure-gradient correction factor of the gas:

(1.19)

tk is the adjusted retention time taking into account the pressure drop in the column.
The corrected retention volume,yi, and the corrected gas hold-up volume of the column,
pM, can be calculated by similar equations:
= VRj (1.20)
v$= VMj (1.21)
V i is the retention volume that would have been measured if the carrier gas were
incompressible. Because the correction for compressibility is less than unity, the corrected
retention volume is less than that observed at the column outlet.
Fig. 1.9 shows the dependence of the pressure-gradient correction factor on the ratio
of the pressures at the inlet and outlet of the column. It can be seen that the pressure-
gradient correction factor greatly depends on this pressure ratio.
As an illustration, let us consider the determination of the corrected volume from the
experimental results. As a result of experimental measurements, the following experimental
parameters were determined: retention time of the test substance, 6.0 min; retention
time of helium (non-sorbable component), 0.5 min; carrier gas velocity measured at 25°C
(298°K) and at an atmospheric pressure of 760 mmHg, 50 ml min; column temperature,
100°C (373°K); water vapour pressure at 25"C, 24 mmHg; gauge pressure at the column
inlet, 900 mmHg. First, we shall determine the true velocity in the chromatographic
column. In measuring the carrier gas volume velocity, Fo, with a foam velocity meter,
one must take into consideration the correction for the water vapour pressure, p H I O ,at
the measurement temperature, T,, and the correction for the gas volume due to the
difference between the column temperature, T,, and that of the velocity meter, T , :

(1.22)

In this instance:

F O = 5 O (298
=) [l- (g)]=60.5cm2/min
 THEORETICAL CONCEPTS OF THE SEPARATION PROCESS 17

c
L-
-b
5
i

1 3 3 4 6

PI /Po

Fig. 1.9. Dependence of the correction for the compressibibty of the carrier gas o n the ratio of the
pressure at the inlet and outlet of the column and o n the pressure drop. The pressure at the inlet is
1 atm.

Let us calculate the retention volume, V A = fRFo= 6.0 min X 60.5 ml/min = 363 ml,
and the corrected retention volume, taking into account that p i / p o = 1.18, j = 0.98 and
= j V R = 334 ml.
Subtracting the corrected gas hold-up volume (V; = jFdM = 29.6 ml) from the
corrected retention volume, we obtain the value of the net retention volume, VN = 304 ml.
Absolute retention values are used mainly in determining physicochemical quantities
(activity coefficient, distribution constant, etc). As noted above, for identification
purposes use is made of relative retention values or functions of relative retention values.
The second basic characteristic of the chromatographic process of separation of a
compound is the broadening of its chromatographic zone. A sufficiently general and
formal description of zone broadening was given by Martin and Synge [54] on the basis
of the theory of theoretical plates. In this'theory the chromatogi iphic column is
considered as a system consisting of a set of successive sections. Each section is a
theoretical plate in which equilibrium of the test compound between the liquid and gas
phases is established instantaneously. The chromatographic process is simulated as two
operations that are repeated many times, namely (1) instantaneous transfer of the mobile
phase from a given plate to the next plate in the absence of mass exchange between the
phases, and (2) establishment of equilibrium of the test compound between the gas and
stationary phases on each plate.
For instance, let a column contain five theoretical plates (Fig. 1 .lo) [ 5 5 ] . In the
initial position, each plate is filled with a gas, and the zero plate contains a sample of the
test compounds A and B, one of which (A) is not sorbed in the stationary phase, while
the other (B) is sorbed; the mass distribution ratio D, = 1, i.e., half of the molecules
of substances B are in the mobile phase and the other half in the stationary phase.
We introduce into the column a volume of pure carrier gas equal to the volume of the
gas phase of the plate. The gas phase of the zero plate (together with substances A and B

References p. 29
18 BASIC PRINCIPLES OF GC

.......
........ o c 0 0
I .. .. .. .. O Q
0.3

, . .. ..
3
. . . . . . 0 0
' 0 0
. . . . . .

.... . . o o
. . 0 " . . . . . . . 0 0

.. .. .. .. I I I . . . . . .
I I I I
1
.. .. . "
0 0
0 . . . . . . . 0 0

.. .. .. .. 0 0

.. .... .. 0 0

.. .. ..

Pig. 1.10. Schematic representation of the chromatographic process based o n the theory of theoretical
plates for compounds A (not retained by the stationary phase) and B (retained by the stationary phase).
0-4, successive stages of the chromatographic process. The plot depicts the distribution of concen-
trations of compounds A ( 0 ) and B ( 0 ) after five stages of gas phase transfer. Upper part of column,
gas phase; lower part, liquid stationary phase. Q = Amount of substance analyzed; 1 = distance from
the column inlet.

which are located in it) will go to the first plate, the gas phase of the first plate to the
second, and so on. In the zero and first plates the substance will be distributed between
the two phases in equilibrium. This process will be repeated upon injection of each new
portion of carrier gas.
After four elementary volumes of carrier gas have passed through the column, the
substance will start to be eluted from the column. The elution curve of substance B is
shown in Fig. 1.1 1 [ 121 . The asymmetry of the elution curve is due to the excessively
small number of plates (N). At N > 50, the peaks are already almost symmetrical, and
at N > 100 the chromatographic zones correspond to the Gauss equation.
The solution of the problem leads to the following equation for the chromatographic
zone based on the theory of theoretical plates:

(1.23)

where n is the number of theoretical plates in the column, q is the size of the sample
analyzed, and V is the volume of the carrier gas that has passed through the column.
From eqn. 1.23, it follows that

I ' R C (1.24)
v; (2?T)f
c
max. u"u max.

In estimating the number of theoretical plates in the column, the real process is
compared with the above-described ideal process of separation, which yields the same
results. The characteristics of zone broadening in the theory of theoretical plates is the
number of theoretical plates of the chromatographic column.
010 .
I
T H E O R r T I C A L CONCI PTS O F T H E SEPARATION PROCFSS

“v
19

rig. 1 . 1 1. Elution curve tor n column of five theorctical plates calculated o n tlie basis of the theory of
theoretical plates. cy = Fraction of substance emerging from column; r i v = number of volumes of
carrier gas passed through column and corresponding to one theoretical plate.

Using eqn. 1.23, one can propose a method for detemiining the number of theoretical
plates. The peak width at a height c = c,,,,,/e can be calculated from the equation

(1.25)

Solving this equation, we obtain

n=2 ( v:: (1.26)

v; ve
-~

Considering that V g = V e = (1/2)be, where b, is th width of th chromatographic peak

at a height h,,a,./e, we have

(1.27)

I n practice, the following equation is usually used for calculating the number of
theoretical plates :

(1.28)

where x is the chart distance from the moment of sample injection t o the emergence of
the peak maximum and)’ is the peak width at half-height.
The efficiency of a chromatographic column increases with the length of the column
used. Therefore, a inore invariant value is the height equivalent to a theoretical plate
(HETP), o r h).
h = L/n (1.29)
The concept of the chromatographic process in the theory of theotetical plates is
rather fomial; this theory does not consider the actual causes of chromatographic zone
broadening. Van Deeniter er al. [56] and Klinkenberg and Sjenitzer [57] developed a
velocity theory in which Lone broadening in a packed column is attributed to a number
of kinetic causes.

References p. 29
20 BASIC PRINCIPLES OF GC

v (rnl/rnin)

Fig. 1.12. Dependence of HETP on linear velocity of carrier gas. 1 =Contribution to HETP from
molecular diffusion; 2 = contribution from mass transfer; 3 = contribution from eddy diffusion.

According to the velocity theory, the dependence of h on u is expressed by the

equation
h = A + (B/u) i-
Cu (1.30)
The first term of this equation ( A ) reflects the contribution to HETP from eddy
diffusion, the second (Blu) from the molecular longitudinal diffusion and the third (Cu)
from the resistance to mass transfer.
Fig. 1.12 illustrates the dependence of HEPT on the carrier gas velocity and shows the
contributions corresponding to each process that causes zone broadening. Further
development of the theory of spreading in chromatography was carried out by Giddings
[ 5 8 ] . We shall now consider in more detail the separate groups of processes that lead to
zone broadening .

Eddy diffusion

In any packed column, zone broadening is due, in particular, to the many possible
routes, of different lengths, by which the molecules of the test substance move along the
column in the flow of carrier gas, Le., to the multiplicity of channels along which the
carrier gas moves in the packing. Therefore, depending on the length of the route, some
molecules will reach the end of the column earlier and others later compared with the
average molecule hold-up time in the column. Thus, the multi-route progress of the
carrier gas through the packing layer leads to broadening of the chromatographic zone.
This cause of broadening is called eddy diffusion. In Van Deemter’s equation, eddy
diffusion is expressed by the term A , and
A = 2hdp (1.31)
where A is a coefficient characterizing the shape of the particles and the uniformity of
their packing in the column, while dp is the average diameter of the sorbent grains.
The value of A is independent of the nature of the test compound, its retention value
and the nature of the carrier gas. A more rigorous description of eddy diffusion was
given by Giddings [ 5 8 ] .
 THEORETICAL CONCEPTS OF T H E SEPARATION PROCESS 21

Molecular diffusion

In the course of separation, there is always a concentration gradient in the gas phase
of a chromatographic zone and molecular diffusion therefore always takes place in the
gas phase, leading to peak broadening. The contribution from molecular diffusion is
reflected in Van Deeniter’s equation by the second term

(1.32)

where y is the obstruction factor, which takes into account the sinuosity of the diffusion
paths in the packing and Dg is the diffusion coefficient of the test substance in the carrier
gas. The factor y is equal to or less than unity [59]. Molecular diffusion depends on the
properties of both the test substance and the carrier gas. When the principal cause of
broadening is the longitudinal molecular diffusion, it is expedient to use dense gases for
reducing the zone broadening, i.e., it is preferable to use nitrogen or argon rather than
hydrogen or helium as the carrier gas.

Resistance to mass transfer

During the motion of a chromatographic zone along the column, the front edge of the
zone is predominantly characterized by the process of sorption, i.e., transfer of
molecules from the gas to the stationary phase. After the maximum, the opposite
phenomenon is observed, namely desorption takes place at the rear edge, i.e., transfer
of molecules to the test compound from the stationary to the gas phase. Both of these
processes occur rapidly, although not instantaneously. Therefore, the zone of the
substance in the gas phase slightly leads the zone of the substance in the stationary phase,
which also contributes to peak broadening [60].
The third term of Van Deemter’s equation reflects the contribution from this process
to the HETP:

(1.33)

where k = KV,/Vg is the extraction coefficient, d, is the thickness of the liquid

stationary phase (LSP) film and D, is the diffusion coefficient of the test compound in
the LSP. The value of C depends on various factors, the most important being the
thickness of the LSP fdm. The value of C increases directly proportionally to the square
of this thickness.
It should be noted that the value of h , although it is an important characteristic of the
column, which defines the broadening of a chromatographic zone, cannot be regarded as
the only value that determines the possibility of solving a particular analytical problem.
In developing an analytical procedure, the problem very often reduces to the separation
of at least two compounds with similar properties.
For a quantitative assessment of the separation of the chromatographic zones of two
compounds present in a mixture in similar concentrations, a number of separation
criteria have been proposed that are a function of the difference between the retention

References p. 29
22 BASIC PRINCIPLES OF GC

values and the widths of the chromatographic zones. The IUF'AC Committee [61]
recommended the use of the following value as a separation criterion:
R , = 2Y10, - Y*) (1.34)
where y is the chart distance between the peak maxima of compounds 1 and 2 and y A
andyB are the peak widths of compounds 1 and 2 at the base of the peaks. The value of
R , varies from 0 to "0; the peaks are completely separated at R , = 1 . In the Soviet
literature, the quantity K 1 is usually adopted as a separation criterion [33], where
K = 1/2R, (1.35)
The peak resolution is determined by the sorbent selectivity, a (a= f R , J f R , J , the
column efficiency, (N), and the mass distribution ratio, D, (see ref. 62):

(1.36)

(1.37)

where fi is the volume ratio of the mobile to stationary phase in the column, K is the
distribution constant and a,K and N refer to the second, dower component.
The development of a satisfactory separation procedure can often be reduced to the
determination of the optimal conditions under which the value of the separation
criterion for a pair of compounds that are difficult to separate would be the highest.
The dependence of the separation coefficient on the experimental parameters is
considered in detail elsewhere [33].

QUALITATIVE AND QUANTITATIVE ANALYSIS

The advances made in GC were due largely to the development of efficient identifi-
cation methods, the characteristic feature of which is a wide use of a combination of
various physical and chemical methods for identifying the peaks in a chromatogram
[48,63-651. The general scheme of application of some widely used identification
methods in GC is shown in Fig. 1.13.'

Qualitative analysis

Qualitative analysis often includes the following stages: (1) preliminary preparation of
the sample, ( 2 ) chromatographic separation with the use of chemical reactions and
selective detectors, (3) isolation and physicochemical study of separate fractions, (4) GC
re-examination of separation fractions. Thus, in order to determine the composition of
a test mixture, bothchromatographic methods based on the measurement of the retention
values and methods based on the physicochemical properties of the test components are
applied. In the following sections we shall consider the principal' methods of identification.
 QUALITATIVE A N D QUANTITATIVE ANALYSIS 23

- 3 - 4 5 -

r 7
J
1;ig. 1.1 3. General scheme of qualitative analysis in gas chromatography. 1 = Sample to be analyzed;
2 = preliminary preparation of sample (separation, chemical treatment, physicochemical investigation);
3 = chromatographic separation, chemical analysis; 4 = selective detectors; 5 = physicochernical study
of separated fractions; 6 = gas chromatograph; 7 = re-cxamination of separate fractions.

Staridard compound method

This method is based on the introduction into the test mixture of standard substances
that are assunied to be already present in the mixture. The coincidence of the retention
times is usually the basis for identifying the peak of the test compound as the standard
coinpound. I t should, however, be noted that this condition is not sufficient for the
qualitative identification of a compound, because identical (or very similar) retention
times may characterire several substances. The reliability of this method increases with
the use of more efficient columns and columns with different phases, the nature of
which detenniiies the sequence of emergence of the coniponents and their retention
values.
As many compounds are not readily available, in piactical chromatography reaction
methods are used for the preparation of standard mixtures [66].

Metliod using fabiilafed data

The determination of the qualitative composition of a mixture by this method is

based on a cornparison of expeiiinentally detemiined retention values of peaks with
tabulated retention values for known coinpounds. Standard compounds with tabulated
retention values are introduced into the test mixtuie; in order to increase the
reliability of the method, several control measurements must be made for corn pounds
with diffeient structures, the presence of which i n the sample being analyzed is assumed,
so as to establish the identity of the chromatographic properties of the given column
with that used for tabulating the data on the retention values. Tabulated retention values
have been published [6, 71.

Refcrenccs p. 29
24 BASIC PRINCIPLES OF GC

Method of several phases

In this method, an unknown mixture is analyzed, not on a single column, but on

several columns with different phases [66]. This technique increases the reliability of the
chromatographic identification of the substance and permits the type of compound
being analyzed to be determined. Fig. 1.14 [67] shows the logarithmic dependence of
the retention times of various compounds for two stationary phases, paraffin oil (tR,J
and tricresyl phosphate (tR,2).It follows from the data presented that compounds of
the same type are characterized by straight lines that do not pass through the origin. This
is an important regularity, which is utilized for determining the type of test compound.

Computational methods and correlation ratios

When tables of retention values lack data on some compounds;it may be useful to use
correlation equations that relate the logarithm of the retention values with the properties
of the test compounds [3] (for instance, the number of carbon atoms or the boiling
point). In many instances, in order to determine retention values, one can use com-
putational methods based on the additive scheme [68-701. For instance, eqn. 1.38
is valid for the retention values of alkanes:
log V = Enii rii (1.38)
where riiis the increment of the logarithm of the retention value corresponding to a
definite combination of bonds (the structural element) and nii is the number of structural
elements of type ij in the molecule of the compound.
The additive scheme is based on the assumption that the molecular interaction of the
test compound with the liquid stationary phase can be regarded as an interaction of a sum
of definite structural elements of the mdecule, each of which is characterized by a
definite contribution to the retention value.
Fig. 1.15 depicts the relationship between the calculated and experimental values of the

$;,:
logarithms of the retention times of alkanes. The agreement is satisfactory.
d

/p
cc

B loo

I lo-
/,$7O !
-
O8
P

il0I,
/
1-
A0
0 1 I I ‘6 I O6 ,
 QUALITATIVE A N D QUANTITATIVE ANALYSIS 25

6 3 0

Fig. 1.16. Separation of mixture of hexafluoroben7ene ( I ) , propyl chloride (2) and n-heptane (3) with
the use of a selective flarne-emission detector. (a), Chlorine; (b), fluorine; (c), carbon.

Physicochemical methods
This group of methods is based on joint utilization of chromatographic and physico-
chemical methods. The application of selective detectors, which record only compounds
of one or several definite classes, pennits information on the nature of the test compounds
t o be obtained; this information, coupled with the chromatographic data, enables one
reliably t o identify the components of the text mixture. As an example Fig. 1 .I6 [71]
shows three chromatograms of the same mixture (hexafluorobenzene, propyl chloride
and ti-heptane, recorded with flame-emission detector. Each chromatogram was obtained
by recording the intensity of a definite spectral line which was selective for chlorine-,
fluorine-, or carbon-containing compounds. This technique made it possible t o establish
the elemental composition of the compounds being analyzed.
Methods of analytical reaction GC are also widely applied for identification purposes
[64, 721. these methods use chemical reactions in a unified chromatographic scheme for
analytical purposes.
The deduction method [ 7 3 ] is often used for group identification. In this method,
two chromatographic analyses of the initial mixture are carried out: one is an ordinary
analysis without the use of chemical reactions, while the other utilizes in the chromato-
graphic scheme a reactor containing an absorbent (reagent) that forms non-volatile
compounds with certain classes of chemical substances. Therefore, the chromatogram of
the second analysis shows no peaks of the reacting compounds (the chromatogram of
the second analysis can be obtained from that of the first analysis by deducting the
peaks of the reactants), which indicates that they belong t o compounds of a definite class.
This method was first used for determining the content of unsaturated compounds in
hydrocarbon mixtures. Unsaturated compounds were absorbed in the reactor by concen-
trated sulphuric acid deposited on silica gel.

Preliminary treatment of the sample

I n thus method, both chemical and physical techniques can be used. For instance, after
a reaction of a mixture of fatty acids with bromine, compounds with one unsaturated

References p. 29
26 BASIC PRINCIPLES OF GC

bond are determined by displacement of the peaks on the chromatogram, and

compounds with several double bonds are determined by the disappearance of their
peaks, as the introduction of bromine into the molecule of an organic compound sharply
reduces its volatility.
Extraction is also used as a subtraction method. For example, an efficient method for
identifying alcohols is preliminary extraction with propylene glycol of components of a
sample dissolved in carbon tetrachloride. Propylene glycol efficiently extracts alcohols,
but not aldehydes, ketones, hydrocarbons or esters, which remain in the carbon tetra-
chloride solution. Acids, phenols and amines, which are also readily soluble in propylene
glycol, can be removed by treatment with an alkali or acid.
The above general identification methods open up wide opportunities for establishing
the composition of unknown mixtures.

Quantitative analysis

An important stage in chromatographic analysis is the qualitative interpretation of

chromatograms, as a result of which the quantitative contents of the components in the
test mixture can subsequently be determined. The accuracy of the results obtained
depends on a number of factors, in particular on the method of analysis selected, the
characteristics of the detector used, the method of calibration and calculation and the
nature of the components being analyzed [74].
In accordance with eqn.l.8, the amount of a substance in a chromatographic zone is
directly proportional to the area of the chromatographic peak in the chromatogram. In
this connection we shall consider methods for determining the area of a chromatographic
zone, assuming that it is a Gaussian curve.
(1) The area of a chromatographic zone is commonly expressed as the product of the
peak height, , and the peak width at half-height:

(1.39)
A more general expression for determining the area of chromatographic peaks, which
permits the calculation of the area of partially separated peaks, has been suggested [75] :
= K8 hmax. P6 (1.40)
If 0 = 0.5,0.75 or 0.9, then K , = 0.941, 1.66 or 2.73, respectively.
(2) The area of a chromatographic zone can be determined as the product of the peak
height and the retention time [76] :

(1.41)
(3) The area of a chromatographic zone is proportional to the height of the chromato-
graphic peak:

(1.42)
 QUALITATIVE AND QUANTITATIVE ANALYSIS 27

Therefore, the amount of a substance in a chromatographic zone is directly proportional

to the following values, which are determined directly from the chromatogram: hmau.pe ;
hmax,xm,, or h,,,,,. However, if in the first method the coefficient of proportionality
between the amount of substance and the value being measured is largely determined b y
the metrological characteristics of the detector and its opening conditions, in the second
and third methods this coefficient also depends on the conditions of the chromatographic
experiment. In particular, in the second method the value of the area depends o n the
number of theoretical plates, which, in general. changes from one substance to another.
In the third method, the coefficient of proportionality depends on the retention value
and separation efficiency.
It should be noted that with large samples, deviations from the above linear dependence
also occur and the retention times change. These deviations must occur in all instances
when the ratio of the volume of the test sample to the peak width (in volume units), as
measured with a vanishingly small sample siLe, exceeds 0.4 [77]. When using ‘manual’
methods for increasing the accuracy, the peak width must be measured by means o f a
measuring magnifying glass with reticule divisions of 0.1 mni [78] .
In recent years, electronic integrators have been widely used in chromatography
[79-811. The use of integrators considerably reduces the processing time, ensures a high
accuracy and reduces the cost of processing. Table 1.2 [82] conipares different methods
of chromatogram processing. I t can be seen that the use of electronic integrators improves
the accuracy and reduces the processing time.
The amount of a substance in a chromatographic Lone is detennined both by the
characteristics of the detecting systeni and by the parameters of the chromatographic
peak. Therefore, in order t o detennine the content of a component in a sample, it is not
sufficient to find the area (or other parameters) of the chromatographic peak of the
substance, but it also necessary t o determine the coefficient of proportionality, which
depends on the type of the detecting system, the experimental conditions, the nature of
the test sample, and so on. Let us consider the quantitative methods for determining the
content of the coniponents in the test mixture.

TABLE 1.2
COMPARISON 01 SOML CIIROMATOGKAhl PROCFSSING METIIODS
~~ .~ -

Characteristic Manual methods Electro- Electronic

____
mechanical digital
Use o i Triangulation Area Cutting o u t
integrator integrator
planimeter method calculation* and weighing

Chromatogram 45- 60 45-60 50-60 100-200 15-30 5-10

processing time
(min)
Reproducibility 4.06 4.06 2.58 1.74 1.29 0.44
(%,)

*Height multiplied by width at half-height.

References p. 29
28 BASIC PRINCIPLES OF GC

Internal normalization method

The modern version of this method in GC was suggested by Keulemans and co-workers
183,841. The calculation is made with eqn. 1.9 taking into account the correction
coefficients for the separate compounds (see, for instance, ref. 82).

Absolute calibration method

In this method, the dependence of the area (or another parameter) of the chromato-
graphic peak on the absolute amount of the test substance is determined experimentally
for each component, Le., one of the following dependences is determined:
qi = Ks = K h = Khph,,,,kJ (I .43)
The absolute calibration method is widely used in chromatography (see for instance
refs. 85 and 86.
It should be noted that absolute calibration must be checked periodically; the
frequency of this checking must be determined empirically. Usually, in re-calibrating,
one can restrict the check to a few points on the calibration graph.

Internal standard method

This method was first used in GC by Ray [87]. A mixture of unknown composition,
into which a known substance is specially introduced at a concentration R , is analyzed.
The concentration of the standard is calculated with reference to the entire test mixture,
which is taken as 100%. The content of the components (P)in the test mixture is
calculated by the equation
P =A S i R or p = fhih8 (1.44)
f s t d . 'std. f h std. hstd.
where fi and f s t d . are the correction coefficients depending on the individual sensitivity
of the detector to the component i and the standard. If R is constant, calibration graphs
are obtained of percentage of impurity versus the ratio of the peak height of the
component i to that of the standard.
The use of relative calibration methods enables one greatly to improve the accuracy
of measurements, because the effect of the experimental conditions on the analytical
results is reduced as the change in the analytical parameters usually affects the retention
time of the standard substance and that of the sample components equally (although there
are exceptions). The corresponding equation can be obtained on the basis of eqns. 1.40-
1.42 and 1.44:

(1.45)

where f i = I / ( q - CUO) and fstd.= l/(astd.- ao).Note that the relative values in this
equation are much less dependmt on the experimental conditions than are the absolute
values of the height and area of the chromatographic peak. Another advantage of the
method is that it is no longer necessary t o measure accurately the volume of the test
 REFERENCES 29

sample except when calibration with a constant sample size is used in order to take into
account the non-linearity of the detector. The method can also be used t o measure the
content of individual components even when not all o f the compounds are recorded on
the chromatogram. In choosing the standard compound, one must ensure that it is
compatible with the sample being analyzed. In order t o increase the accuracy of the
analysis, it is desirable that the substance used as the standard should be similar t o the
test components in terms of the retention value and their content in the mixture being
analyzed.
The method automatically takes into account possible losses of the test con1 ponents
during preliminary preparation of the sample [88] (for instance, during extraction,
distillation, adsorption and other extraction and concentration procedures).
Many of the above calculation and calibration methods are based on the assumption
that the components of the test sample are not adsorbed and remain unchanged during
the analysis, the detector readings are linear over a wide range of concentrations and
are independent of the presence of other components in the sample and the values of the
correction coefficients published in the literature are adequate. Unfortunately, many of
these conditions are not fulfilled in practice. Therefore, accurate quantitative results can
be obtained only by calibrating a given device with a mixture of known composition
that contains all of the compounds present in the test mixture. The calibration of the
device must be checked periodically.
In conclusion, we wish t o emphasize the necessity of applying statistical methods
[65,89,901 for estimating the accuracy of the chromatographic procedures used.

REFERENCES

1 J. Janik, Chromatogr. Rev., I 1 (1969) 203.

2 J. I. Walraven, Joint Symposium o n Accurate Methods of Anaiysis f o r Major Constituents, London,
April 1570, Paper 1 1.
3 A. T. James and A. J. P. Martin,RiocAem. J., 50 (1952) 679.
4 M. S. Tsvet, Khromatograficheskii Adsorptsionnyi Analiz (Chronlatographic Adsorption Analysis),
Academy Press, Moscow, 1946.
5 R. P. W. Scott (Editor), Gas Chromatography 1560, Butterworths, London, 1960.
6 J. S. Lewis, Coinpilafion of Gas Chromatographic Data, ASTM Special Technical Publication
No. 343, American Society for Testing and Materials, Philadelphia, Pa., 1963.
I W. 0. McReynolds, Gas Chrornatograpkic Retention Data, Preston Technical Abstracts Co.,
Evaston, Texas, 1966.
8 F. Bruner, C. P. Cartoni and M. Posanzini,Anal. Chem., 41 (1969) 1122.
9 G. N. Turkeltaub and B. M. Luskina, Zavod. I,ab., 35 (1969) 545.
10 V. G. Berezkin and N. S. Nikitina, Zh. Fiz. Khitn., 42 (1968) 2942.
1 1 V. L. Kepke and G. 1:. Sokolin, Zavod. Lab., 36 (1970) 1301.
12 S. Dal Nogare and R. S. Juvet, Gas- Liquid C/iromatograp1iy, Theory and Practice, Intersciencc,
New York, 1962.
1 3 A. V. Kiselev and Y. I. Yashin, GasAdsorption Chromatography, Plenum Press, New York, 1969.
14 C. E. Doring, in E. Leibnitz and H. G . Struppe (Editors), Handbirch der Gas-C/iromatographie,
Acadeniische Verlagsgesellscliaft, Leipzig, 1966, p. 737.
15 M. S. Vigdergauz and R. I. IzrnayIov, Prinieneniye Gazovos Khrornatografii dlya Opredeteniya
Fiziko-khimicheskikh Svo-vsrv Veshchests (Application of Gas Chroniatograpliy for the Determination
of Physicochcniical Properties of Substances). Nauka, Moscow, 1970.
30 BASIC PRINCIPLES OF GC

16 S. Z. Roginsky, M. I. Yanovsky and A. D. Berman, Osnovy Primeneniya Khromatografii v Katalize

(Fundamentals of Application of Chromatography in Catalysis), Nauka, Moscow, 1972.
17 A. B. Kiselev and V. P. Dreving (Editors), Experimentalnyie Methody v Adsorbtsii i Molekuliarnoy
Khromatografii (ExperimentalMethods in Adsorption and Molecular Chromatography), Moscow
University Press, Moscow, 1973.
18 D. A. Vyakirev, N. F. Shushunova, I. I. Chuyev, M. K. Shchennikova and G. G. Kuravskaya,
Neftekhimiya, 9 (1969) 861.
19 N. E. Buyanova, G. B. Gudkova and A. P. Karnaukhov, Kinet. Katal., 6 (1965) 1085.
20 K. V. Aleksyeyva, V. G. Berezkin,S. A. Volkov and E. G. Rastyannikov, Polucheniye bChistykh
Veshchestv Merodom Preparativnoy Gazovoy Khromatografii (Obtaining Pure Substances by
Preparative Gas Chromatography), TSNIITNeftekhimprom, Moscow, 1968.
21 A. Zlatkis and V. Pretorius (Editors), Preparative Gas Chromatography, Wiley-Interscience, New
York, London, Sydney, Toronto, 1971.
22 K. I. Sakodynsky and S. A. Volkov, Preparativnaya Gazovaya Khromatografiya (Preparative Gas
Chromatography), Khimiya, Moscow, 1972.
23 V. A. Averin, V. Ya. Shnol' and B. G. Distanov, Gazov. Khromatogr., No. 9 (1969) 142.
24 M. Taramasso, Gas Chromatographia, Franco Angeli Editore, Milan, 1966.
25 J. Krugers (Editor), instrumentation in Gas Chromatography, Centrex, Eindhoven, 1968.
26 B. V. Stolyarov, I. M. Savinov and A. G. Vitenbeg, in B. V. Ioffe (Editor),Bukovodstvo k
Prakticheskini Rabotam PO Gazovoy Khromatografii (Manual on Practical Work in Gas Chromato-
graphy), Leningrad University Press, Leningrad, 1973, p. 284.
27 F. van de Craats, Anal. Chem., 14 (1956) 136.
28 J. Hoomeijer, A. Kwantes and F. van de Craats, in D. H. Desty (Editor), Gas Chromatography 1958,
Butterworths, London, 1958, p. 288.
29 W. R. Supina, The Packed Column in Gas Chromatography, Supelco, Bellefonte, Pa, 1974.
30 L. S. Ettre, Open Tubular Columns in Gas Chromatography, Plenum Press, New York, 1965.
31 D. M. Ottenstein,J. Gas Chromatogr., 1 (1963) 1 1 .
32 V. G. Berezkin, V. P. Pakhomov and K. 1. Sakodynsky, Tverdyie Nositeli v Gazovoy Khromatografii
(Solid Supports in Gas Chromatography), Khimiya, Moscow, 1975.
33 A. A. Zhukhovitskii and N. M. Turkeltaub, Gazovaya Khromatografiya (Gas Chromatography),
Gostopekhizdat, Moscow, 1962.
34 L. C. Browning and J. 0. Watts, Anal. Chem., 29 (1957) 94.
35 E. G. Hoffman,Anal. Chem., 34 (1962) 1216.
36 A. E. Messner, D. D. Rosie and P. A. Argabright,Anal. Chem., 31 (1959) 230.
37 K. L. Grob, D. Mercer, T. Gribbon and H. Wells, J. Chromatogr., 3 (1960) 545.
38 H. Veening and G. D. Dupre,J. Gas Chromatogr., 4 (1966) 153.
39 L. D. Hiushaw, J. Gas Chromatogr., 4 (1966) 300.
40 A. J. P. Martin and A. T. James, Biochem. J., 63 (1956) 138.
41 A. G. Nerheim,Anal. Chem., 35 (1963) 1640.
42 A. A. Datskevich, Gazov. Khromatogr., No. 1 (1964) 79.
43 J. Harley, W. Nel and V. Pretorius, Nature (London), 181 (1958) 177.
44 J. G. McWilliain and R. A. Dewar, Nature (London), 182 (1958) 1664.
45 E. R. Adlard, Crit. Rev. Anal. Chem., 5, May (1975) 13.
46 V. V. Brazhnikov, Differentsialnyie Detektory dIya Gazovoy Khromatografii (Differential
Detectors f o r Gas Chromatography), Nauka, Moscow, 1974.
47 V. A. Botin, Radioionizatsionnoye Detektirovaniye v Gazovoy Khromatografii (Radioionization
Detection in Gas Chromatography), Atomizdat, Moscow, 1974.
48 L. S. Ettre and W. H. McFadden (Editors), Ancillary Techniques of Gas Chromatography, Wiley-
Interscience, New York, London, Sydney, Toronto, 1969.
49 J. E. Lovelock and S. R. Lipsky,J. Amer. C h e m SOC., 82 (1960) 431.
50 A. Karmen,Anal Chem., 36 (1964) 1416.
51 D. Henneberg, Z. Anal. Chem., 183 (1961) 12.
52 V. L. Talroze, V. V. Reznikov and G. D. Tantsyrev, Dokl. Akad. Nauk SSSR, 159 (1964) 182.
53 J . C. Giddings and R. L. Mallik,i,:d Eng. Chem., 59 (1967) 18.
 REFERENCES 31

54 A. J. P. Martin and R. L. M. Synge,Biochem. J . , 35 (1941) 1385.

55 M. Singliar, Pracrica Oomatografiei de Gaze, Trad. Dinliniba Slovaca, Buchurest, 1963.
56 J . J. van Deemter, F. .I.Zuiderweg and A. Klinkenbcrg, Cheni. Eng. S c i , S (1956) 271.
57 A. Klinkenherg and I:. Sjenitzer, Chem. Eng. Sci., 5 (1956) 258.
58 J . C. Giddings, Dynamics of Chromatography, Part 1, Marcel Dekker, New York, 1965.
59 E. Glueckauf, in D. 11. Desty (Editor), Gas Orromatography 1958, Academic I’rcss, New York,
1958, p. 33.
60 G. Schay, Theoretische Gnindlageti der Gas-Chroniatoxraphie, VEB Deutscher Verlag der
Wissenschaften, Berlin, 1960.
61 IUPAC, Pure Appl. Chem., 37 (1974) 445.
62 Cs. Horvath, in 1,. Ettre and A. Zlatkis (Editors), The Practice of’
Gas Chromatography, Interscience,
New York, London, 1967, p. 129.
6 3 D. A. Leathard and R. C. Shurlock, Identification Techniques in Gas Chronzatography, Wiley,
London, New York, Sydney, Toronto, 1970.
64 V. ti. Berezkin, Analytical Reaction Gas Chromatography, Plenuni Press, New York, 1968.
65 R. Kaiser, Chromatoyraphie in der Gasphase, Bibliographes Institu!, Mannlieim, 1969.
66 V. G. Berezkin, L. Sojik and I. UndeovB, J. Chromatogr., 9 8 (1974) 157.
67 E. Bayer, in D. H. Desty (Editor), Gas Chromatography, Buttcrworths, London, 1958, p. 333.
68 V. G. Rerezkin, Neftekhimija, 1 (1961) 169.
69 V. G. Berezkin and V. S. Kruglikova, Neftekhiniiya, 2 (1962) 845.
70 G. Schoniburg, Advan. Chrornatogr., 6 (1968) 21 1.
71 A. J. McCormack, S. C. Tong and W. D. Cook, Anal. C%em.,3 7 (1965) 1470.
72 M. Reroza and M. N. Inscoe, in L. S. Ettre and W. H . McFadden (Editors), Ancillary Techniques of
Gas Chromatography, Wiley-Interscience: New York, London, Sydney, Toronto, 1969.
73 R. L. Martin,Anal. Chem., 32 (1960) 336.
74 L. A. Kogan, Kolitscizesti~entzayaGazovaya Khromatografiya (Quan titative Gas Chroma tography),
Khimiya, Moscow, 1975.
75 A. A. Zhukovitskii. B. A. Kazanskii, 0. D. Sterligov and N. M. Turkeltaub, Dokl. Akad. Nauk
SSSR, 123 (1958) 1037.
76 J. Novik, Chetn. Listy, 59 (1965) 1021.
77 V. I. Kallnanovski and A. A. Zhukovitzki,J. Chromafogr., 18 (1965) 243.
78 V. G . Baranova, A. G. I’ankov and Ya. I. Turian, Ostrovy F’izikokhimicheskikh Metadov Analiza i
Kontrol Proizvodstva Izoprena (Fundamentals of’f)lzysicochernicalMethods o f Analysis and
Control of Isoprene ProducfionJ, NIITEKhim, Moscow, 1965.
79 J. M. Gil1,J. Chromatogr. Sci., 10 (1972) 1.
80 ti. G. Struppe (Editor), .4spects of Gas Chromatography, Deut. Akad. Wiss., Berlin, 1971, p. 144.
81 Chromatographia, 7, No. 9 (1972).
82 H. M. McNair and E. J . Bonelli, Basic Gas Chromatography, Varian Aerograph, Walnut Creek,
Calif., 1967.
83 A. 1. M. Keulenians and A. Kwantes, Anal. Chim. Acfa, 13 (1955) 357.
84 A. I. M. Keulernans, A. Kwantes and G. W. Rijndcrs, Anal. Chim. Acta, 16 (1957) 29.
85 N. Brenner and L. S. Ettre,Anal. Chem., 31 (1959) 1815.
86 A. I. Dolgina, A. D. Alekseyeva and A. N. Meshcheryakova, Gazov. Khromatogr., No. 4, (1966)
120.
87 N. If. Ray, J. AppL Chetn., 4 (1954) 21.
88 11. P. Burchfield and E. E. Storrs, Biochemical Applications of Gas Chromatography, Academic
Press, New York, 1962.
89 A. N. Zaydel, Elementarnyiye Otsenki Oshihok Izmerenii (Elementary Estimates o f Measurements
ErrorsJ, Nauka, Leningrad, 1967.
90 K . Doerffel, Statistik in der Analytischen Chemie, Deutscher Verlag fiir Grundstoffindustrie,
Leipzig, 1966.