Banner

How Does Gas Chromatography Work?

How Does Gas Chromatography Work?

by Peter Ridgeon
 
Introduction

Gas Chromatographs first became commercially available in the late 1950s and the technique has been steadily developed and refined since then and has become one of the most important tools in analytical chemistry. The popularity of gas chromatography stems from the fact that the technique offers an extremely wide application range, combined with very high measurement sensitivity and selectivity.

Gas chromatography is a separation technique in which the constituent components of a sample mixture are subjected to a competitive distribution between two phases; one a moving gas stream and the other a stationary liquid or solid. The separation process is performed by introducing a small aliquot of the analysis sample into a gas stream (the carrier gas) flowing through a tube (called the column) containing the stationary phase. Two different separation mechanisms are used. In adsorption chromatography the stationary phase is a powdered adsorbent material such as alumina or silica gel, whereas in partition chromatography the stationary phase is a liquid.

In the early days of gas chromatography, a typical column was about 2 metres long with an internal diameter of 4 millimetres, packed either with a powered adsorbent or with an inert powdered material the particles of which were coated with a film of the liquid stationary phase. It was later found that much higher separating power could be obtained if the stationary phase was contained on the inner walls of a long narrow tube, with the result that capillary columns have almost completely replaced the original packed version.

In both adsorption and partition chromatography the time taken for an individual chemical substance to pass through the column is inversely proportional to the temperature and most analyses are performed with the column operated at elevated temperature. In some cases better separation can be obtained by starting the analysis with the column at low temperature followed by a controlled increase as the analysis proceeds. Temperature programming the column in this way can result in a dramatic reduction in the time taken for an analysis to be performed.

The analysis is initiated by introduction of the sample into the gas stream at the beginning of the column, normally into a specially heated zone so as to ensure rapid evaporation. The various constituents of the sample then migrate through the column at a rate that is determined by their partition coefficients between the two phases. Provided that the partition characteristics of the various components in the sample are sufficiently different, a separation is obtained.

As the individual components emerge from the column in the gas phase, they pass through a detector, which generates an electrical signal proportional in magnitude to the concentration of the substance. The output of the detector is displayed as a 'chromatogram' - a trace showing each sample constituent as a 'peak' on a horizontal baseline. A range of detection systems is available, the choice for a particular application depending on considerations such as the type of sample being analysed and the required detection sensitivity and selectivity.

 

  
 

Qualitative Analysis by Gas Chromatography <top>

Gas chromatography can be used to identify the chemical composition of sample materials. The basis of the use of the technique for this purpose lies in the fact that, under fixed column temperature and carrier gas flowrate conditions, the time taken for a substance to pass through a particular column is a fixed and repeatable characteristic. The 'retention time' of a substance is a parameter equivalent to it's boiling point or melting point in that it can be used as evidence of identity. However, it must be remembered that, as with melting and boiling point data, a substance's retention time may not be unique and care must be exercised in assigning an identity based solely on retention time measurements. The combination of a gas chromatograph, to achieve a separation of a samples components, with a mass spectrometer detection system which gives chemical structure information has become a popular and powerful means of overcoming the limitations of sample identification made solely on the basis of retention time. <top>

 
 

Quantitative Analysis by Gas Chromatography <top>

Quantitative measurements made by gas chromatography are based on the fact that the electrical output generated by the detector is proportional to the concentration of sample analyte in the carrier gas stream. Some detectors offer a wide linear response range, but others are more limited in this respect. For example, the response of the flame ionization detector is linear over a concentration range of more than 10 6 whereas the electron capture detector response flattens off when the concentration of the analyte is a factor of 10 4 greater than the detection limit.

When developing a quantitative method of analysis by gas chromatography, it is important to be aware of the fact that the response of most detectors varies from compound to compound. The electron capture detector, for example, provides extremely high sensitivity for the detection of substances of high electron affinity such as halocarbons, but little or no response to substances such as hydrocarbons. A quantitative method must also take into consideration the fact that it is not possible to precisely inject the same volume of liquid sample each time an analysis is performed.

These difficulties are overcome by the use of the internal standard method of quantitative analysis. This involves finding a substance that is not present in the analysis sample, but with similar chromatographic characteristics and that elutes without interfering with any of the peaks of the components in the sample. This substance (called the 'internal standard') is then used as a reference compound in making accurate quantitative measurements of the concentration of each of the components of interest in the sample. Preparing a calibration mixture containing all of the analytes in the sample, plus the internal standard in measured amounts. Analysis of this mixture then allows the relative response of the detector to each of the analytes compared to the internal standard to be calculated. Analysis of the sample to which a measured amount of the internal standard has been added. Using the results of this analysis and the relative response data derived from the calibration analysis to calculate the concentration of each of the compounds of interest in the sample. 

The use of the internal standard method overcomes the difficulties in quantitative analysis otherwise caused by the fact that the detector response is not the same for all substances and of the inability to inject identical liquid sample volumes repeatedly. 

When analysing gas samples, it is possible to reproducibly inject the same volume of gas. This enables the detector response to be calibrated to each of the sample constituents by a technique known as external standardisation, in which a standard mixture of known concentration of all the components, but without any added internal standard, is analysed. The same volume of the sample gas mixture is then injected and the concentrations of each component calculated by comparison of the peak areas with those of the calibration mixture.

Quantitative measurements made in gas chromatography almost always involve calculations using the area under the 'peak' of the component in the chromatogram as the measure of concentration. A modern gas chromatograph will normally be equipped with a computer based data processing system with a video display of the chromatogram, and with sophisticated software to identify the beginning, retention time, end, and area of each peak.  Once the operator has imputed calibration data, the computer program will automatically calculate sample component concentrations, and will provide analysis reports in whatever format is required in addition to storing the analysis data, including the chromatogram, for future reanalysis, should this be necessary.<top>

 
 

Sample Introduction Techniques <top>

A variety of different techniques are available for introduction of the sample into the chromatographic column. If the analysis requires the entire aliquot of the sample to be injected into the column, this can be done in the case of liquids by use of a micro syringe. Solid samples can be treated in the same way, after dissolution in a suitable solvent. Gas samples can be injected by syringe, but normally are introduced by means of a valving arrangement which can be easily automated and that ensures reproducible injection volumes.

If the analysis requires the measurement of volatile substances in a mainly involatile sample matrix, it is often preferable to use a 'head space' method of sample introduction. This involves maintaining the sample at a defined elevated temperature in a sealed container for a period of time to enable a situation to be established in which the concentration of the volatile substances in the vapour phase are in equilibrium with their concentration in the sample matrix. An aliquot of the 'headspace' vapour above the sample is then removed from the container and injected into the chromatographic column. This technique ensures that only the volatile sample constituents are injected and can be very beneficial in preventing low volatility materials from contaminating the column. Measurement of the ethyl alcohol concentration in blood samples is a good example of the use of the headspace technique.

'Purge and trap concentration' is another technique used for injection of volatile sample constituents into the chromatographic column. This involves placing the sample material in a container and purging it with a flow of an inert gas to remove volatile constituents. The purge gas is passed through a bed of absorbent material, which retains any volatile material present. The adsorbent bed is then connected into the path of the carrier gas flow and is rapidly heated to release any volatile compounds into the column. Purge and trap techniques have been widely used for analysis of contaminants in drinking water.

The aim of the injection technique is to introduce a representative aliquot of the analysis sample in a narrow band into the beginning of the column. Although the modern capillary column can provide very high separating efficiency, it will only do so if the amount of sample injected is small enough to ensure that none of the analytes exceed the maximum capacity of the stationary phase. The minimum volume that can be dispensed reliably from a micro syringe is about 0.1 microlitres and this is often greater than the capacity of the column. To overcome this problem, liquid samples containing high concentrations of analytes are normally introduced into a capillary column using a 'split' injection technique. This involves allowing only a small fraction of the injected sample to enter the column, the bulk being vented from the injector, out to atmosphere.

In the analysis of very dilute samples, a 'splitless' injection technique may be used. In this method, the split vent referred to above is closed during the injection, so that the entire sample enters the column. It is usual to reopen the split vent at some stage after the injection has been made and the analytes have entered the column, so as to flush the injector of any residual solvent vapour.

In some capillary column applications it is preferable to use a Programmed Temperature Vaporising (PTV) injector, rather than the hot, isothermal type. When using the PTV, the sample is introduced into a cold injector, which is then heated at a controlled rate so as to vaporise the analytes and transfer them to the column. This technique ensures that the sample components are transferred into the column at the minimum possible temperature, and is particularly suitable for the analysis of samples containing thermally unstable compounds. The PTV injector can be used in both split and splitless modes. In addition, it can also be used for large volume sample injections. In many applications of gas chromatography, the analysis sample consists of a dilute solution of analytes in a volatile solvent. Using a conventional hot capillary column injector, it is not possible to inject sample volumes of more than about 5 microlitres without causing a reduction in column separating efficiency. 

However, a PTV injector can handle comparatively large volumes of such samples, with dramatic benefits in terms of the limits of detection. When operated in the large volume injection mode, the PTV is maintained at a low temperature, with the split vent open during the sample introduction phase. The sample is injected relatively slowly, so as not to flood the injector and to allow the volatile solvent to vaporise and exit to atmosphere, via the split vent. The vent is then closed and the injector temperature programmed so as to transfer the analytes into the column. Sample volumes in excess of 100 microlitres can be injected using this technique.

It is also possible to inject the sample using a cool on column technique. This involves using a syringe with a needle diameter smaller than the bore of the column so that the sample can be introduced directly into the column. The column oven is initially maintained at a temperature lower than the boiling point of the solvent used to dissolve the sample. After the injection, the solvent evaporates, leaving a narrow band of sample material. The column oven is then temperature programmed, and the sample constituents transfer into the gas phase as the temperature rises. This technique ensures that a truly representative aliquot of the sample is introduced into the column and minimises thermal degradation problems. In order to avoid deposition into the capillary column of any involatile material present in the sample, the on column injection is normally made into a short length of pre-column that can be easily replaced, when required.

Thermal desorption techniques are widely used in gas chromatography for the measurement of trace impurities in air samples. In this technique a large volume of sample air is passed through a trap consisting of a tube containing a powdered adsorbent, such as charcoal. Contaminants in the air are trapped in the adsorbent bed. After collection of the sample, the trap is then inserted into the carrier gas path at the inlet to the chromatograph, and rapidly heated to release the analytes into the column. <top>

 
 

The Column <top>

The basis of a satisfactory method of analysis by gas chromatography is the successful separation of the various components of the sample mixture. Much of the method development involved in gas chromatography is concerned with the selection of a suitable column for the application.

Although there are still a limited number of applications of gas chromatography that are best performed using packed columns, the bulk of analyses now use capillary columns. The separating efficiency of a capillary column is much greater than that of a packed column, and the capillary column offers improved detection limits because of the sharp peaks obtained and by minimising the effect on stability by vaporisation of the stationary phase ('bleed') into the detector.

Capillary columns are manufactured from fused silica tubing, and are typically 10 - 50 metres long with an internal diameter in the range 0.1 - 0.53 millimetres. A polyimide coating is applied to the outside of the silica tubing in order to improve its mechanical robustness.

A film of the stationary phase, typically 0.1 - 5 microns thick, is coated on the inside wall of the column. Narrow bore columns with thin stationary phase coatings provide fast analysis times, but the sample capacity of such columns is very small. Overloading of the stationary phase, resulting in a reduction in column efficiency, occurs if too large a sample is injected. Larger diameter columns contain more stationary phase, and give longer analysis times but are less prone to sample overload problems. The stationary phase must be chemically inert, involatile, and thermally stable at the column temperature employed for the analysis. A wide range of capillary columns is available, mainly using high boiling point organic polymers as the stationary phase. As a general rule, non-polar sample materials such as saturated hydrocarbons, are best analysed using non-polar stationary phases, of the 100% dimethyl polysiloxane type, which separate on the basis of the differences in the boiling point of the analytes. Polar sample materials, such as alcohols and ketones, require a polar phase such as polyethylene glycol that separates on the basis of hydrogen bonding interactions. In between these extremes there are many other stationary phases with different polarities, suitable for a wide range of applications.

The choice of carrier gas has an important bearing on the performance achieved by a capillary column.  Nitrogen is generally unsuitable as a carrier for capillary column applications; it's use resulting in long analysis times and poor column efficiency. Helium is a better choice and is widely used, although the fastest analysis times and highest column efficiencies are obtained when using hydrogen as carrier gas. <top>

 
 

Detectors <top>

The detection systems used in gas chromatography are required to respond to the various sample constituents as they emerge in the carrier gas stream at the end of the column. In some applications it is necessary for the detector to respond to all of the sample constituents, but in others it may be more desirable to use a detector that selectively responds only to particular substances. The detector is required to provide adequate sensitivity to meet the requirements of the analysis, and to be capable of giving an output that is linearly proportional to the concentration of the analyte in the sample.

A wide range of detectors have been developed for use in gas chromatography some of these are listed below:

  
 

Flame Ionization Detector (FID) <top>

Very few ions are produced in a clean hydrogen flame and the current flow between the two electrodes is very small. Typically the baseline current of an FID is in the order of 1 x 10-11 amps. When an organic substance is eluted from the column it is burnt in the flame and produces a large increase in ions. This results in an increase in the current flowing between the two electrodes and this is amplified by the detector's electronics and displayed as a 'peak'.

 

The FID is the most widely used gas chromatography detector because of it's high sensitivity (detection limit approx 1 x 10-12g/sec), wide linear response range (106), low cost, and ease of use.

The FID responds to all organic substances with the exception of formic acid, formamide and formaldehyde, but does not respond to He, Ne, Ar, Kr, Xe, H2, 02, N2, H2S, CO, C02, COS, S02, NH3, NO, N02, N20 or water.<top>

 
 

Thermal Conductivity Detector (TCD) <top> 

Four tungsten filaments are mounted in a steel block. Two of the filaments are in a chamber in the block through which the carrier gas flows after emerging from the end of the column, and the other two are in a second chamber through which a reference stream of carrier gas passes. The filaments are electrically connected in the arms of a Wheatstone Bridge network. Electrical power is applied to the bridge and the filaments become hot due to the current flow. The actual temperature achieved by the filaments is a function of the power applied to the bridge, and the thermal conductivity of the gas passing through the detector.

Under stable conditions the same gas is flowing through both chambers of the detector and the bridge attains equilibrium, resulting in a steady output. When a sample component is eluted from the column, this results in a change in thermal conductivity of the gas flowing over F3 and F4 causing a change in the temperature of the filaments. This results in a change in filament resistance, causing the Wheatstone Bridge to become unbalanced and to output a signal proportional to the out of balance. The performance of the thermal conductivity detector can be improved by the use of a system in which the filaments are maintained at constant temperature during the elution of the sample components, by variation of the applied power. The power required to maintain the filaments at constant temperature is outputted to the data system and produces a chromatogram. Compared with the variable temperature type of thermal conductivity detector the constant filament temperature system gives greater sensitivity and improved linear range.

The thermal conductivity detector responds to any sample component that has a different thermal conductivity to that of the carrier gas. Thus the detector will respond both to organic and inorganic materials and its main application is for the detection of those substances (mainly permanent gases and water) that cannot be detected with the FID.

The sensitivity of the thermal conductivity detector is a function of the difference in thermal conductivity between the carrier gas and that of the sample components. Helium or hydrogen carrier gas provides the highest sensitivity, but argon is suitable for some applications.<top>

 
 

Selective Detectors <top>

Both the FID and thermal conductivity detectors are normally regarded as being non-selective. i.e. they respond to all of the sample components. However, the FID can behave as a selective detector in some cases. For example, it can be used for measurement of very small concentrations of organic contaminants in air, the detector giving good response to the organic substances, but no response to air.
Selective detectors commonly used in gas chromatography include the following:
 

Electron Capture Detector (ECD) <top>

The electron capture detector consists of a chamber containing a radiation source and a pair of electrodes across which a polarizing voltage is applied. The radiation source normally used in the ECD is Ni 63, a Beta particle emitter. Nitrogen is used as the carrier gas and is ionized by the radiation, producing high energy electrons. These electrons generate a current flow between the two electrodes, and this is outputted to the data system.

Some chemical substances exhibit a high affinity for electrons and if a material of this type is introduced into the ECD a decrease in the number of electrons in the chamber will occur. This results in the current flowing between the two electrodes being reduced. On the other hand, many substances do not have such an affinity for electrons and compounds of this type pass through the detector with little or no effect on the standing current. Thus, the ECD provides a selective response to electron capturing species.

In its original form the ECD was polarized with a DC voltage but improvements in performance were shown to be possible if a pulsed voltage was used instead. More recently it has been shown that further improvements in performance can be obtained if the detector is operated under constant current conditions; the frequency of application of the polarizing voltage pulses being altered during the sample elution in order to maintain a constant current. The use of the pulse-modulated system improves the linearity of the ECD and also the ability of the detector to tolerate contamination.

The ECD responds strongly to many substances containing halogens, to polycyclic aromatic hydrocarbons, polychlorinated biphenyls, tin and lead organic compounds etc. Among the applications of the ECD is the detection of traces of pesticides in crops and water samples and in the measurement of atmospheric pollutants causing ozone depletion. Typical detection limit of the ECD is 2 x 10-14g/sec with a linearity range (pulse modulated system) of 104. <top>

 
 

Flame Photometric Detector (FPD)<top>

The FPD is based on the fact that when burnt in a hydrogen rich flame, phosphorus and sulphur containing substances give a characteristic light emission. Phosphorus compounds emit light at 526 nm and sulphur compounds at 394 nm, due to the formation in the flame of the HPO and S2 species respectively. In the FPD, an optical filter, designed to block light other than at 526 or 394 nm is placed between the flame and a photomultiplier detector. In this way, a selective response to phosphorus and/or sulphur compounds is obtained.

The response of the FPD to phosphorus compounds is linearly proportional to sample concentration, but in the case of sulphur, since the response is due to formation of the S2 species, it follows an approximately square law relationship. Main applications of the FPD include the detection of traces of organo phosphorus and sulphur pesticides and for the determination of H2S, SO2, mercaptans etc.<top>

 
 

Nitrogen Selective Detector (NPD)<top>

The NPD is a modified flame ionization detector in which an alkali metal halide is continuously vaporized into the flame. In the presence of the alkali metal halide vapour in the flame, organic phosphorus or nitrogen containing substances burn to produce ionic species which are unusually stable. In a normal FID the ions produced by combustion of the sample rapidly decrease in concentration in the area above the flame. If the collector electrode is moved away from the immediate vicinity of the flame little or no response is obtained. In contrast, in the NPD , the ions formed from organo phosphorus or nitrogen substances persist much longer and, if the collector electrode is placed in a suitable position, a selective response to phosphorus or nitrogen compounds can be obtained.

Applications of the NPD include measurement of traces of pesticides in crops and water samples and of drugs in body fluids. <top>

 
 

Mass Spectrometer Detector (MS) <top>

The mass spectrometer has become the most important detector used in gas chromatography. The reasons for this are that the detector offers very high sensitivity, but most importantly it enables a positive identification of the sample constituents to be made. In a mass spectrometer detector, the column effluent is passed into an ionization chamber, and bombarded with high energy electrons from a hot tungsten filament. Any organic substances present in the carrier gas are fragmented by collision with these electrons. The ionic species resulting from this process are accelerated into a magnetic sector or quadrupole spectrometer where they are characterised on the basis of their mass to charge ratio. The mass spectrum resulting from this process is characteristic of the sample component and can be used for identification purposes.

In some applications the electron bombardment from a tungsten source may be too energetic in that complete disintegration of the sample molecule may occur resulting in a mass spectrum that provides little useful information. In such cases, an alternative chemical ionization technique may be used. In this method, a gas such as methane is fed into the mass spectrometer inlet where electron bombardment results in the creation of low energy ions that then interact with the sample components to produce a more useful mass spectrum. <top>