Chapter 27


Gas Chromatography


  1. Instrumentation
  2. The Carrier Gas
  3. Longitudinal Diffusion Term
  4. Retention Volumes
  5. Injectors
  6. Column Configurations and Ovens
  7. Detection Systems
  8. Detectors
  9. Gas Chromatographic Columns and Stationary Phases
  10. Capillary/Open Tubular
  11. Solid Support Materials
  12. Selection of Stationary Phases
  13. Liquid Stationary Phases
  14. Adsorption of solutes on Solid Support and Column Walls
  15. Gas-liquid chromatography (GLC)
  16. Gas-solid chromatography (GSC)
  17. Temperature Programming
  18. Qualitative and Quantitative Analysis
  19. The Retention Index
  20. Interfacing GC with other Methods



Gas chromatography is a technique used for separation of volatile substances, or substances that can be made volatile, from one another in a gaseous mixture at high temperatures. A sample containing the materials to be separated is injected into the gas chromatograph. A mobile phase (carrier gas) moves through a column that contains a wall coated or granular solid coated stationary phase. As the carrier gas flows through the column, the components of the sample come in contact with the stationary phase. The different components of the sample have different affinities for the stationary phase which results in differential migration of solutes, thus leading to separation.


This separation technique was introduced by martin and James in 1952 which is the latest of the major chromatograhpic techniques. However, by 1965 over 18000 publications in gas chromatography (GC) were available in the literature. This is because optimized instrumentation was feasible. Gas chromatography is good only for volatile compounds or those which can be made volatile by suitable derivatization methods or pyrolysis. Thus, about 20% of chemicals available can be analyzed directly by GC.

Gas chromatography can be used for both qualitative and quantitative analysis. Comparison of retention times can be used to identify materials in the sample by comparing retention times of peaks in a sample to retention times for standards. The same limitations for qualitative analysis discussed in Chapter 26 also apply for separations in GC. Quantitative analysis is accomplished by measurement of either peak height or peak area.


Gas - Solid Chromatography (GSC)


The stationary phase, in this case, is a solid like silica or alumina. It is the affinity of solutes towards adsorption onto the stationary phase which determines, in part, the retention time. The mobile phase is, of course, a suitable carrier gas. This gas chromatographic technique is most useful for the separation and analysis of gases like CH4, CO2, CO, ... etc. The use of GSC in practice is considered marginal when compared to gas liquid chromatography.


Gas - Liquid Chromatography (GLC)


The stationary phase is a liquid with very low volatility while the mobile phase is a suitable carrier gas. GLC is the most widely used technique for separation of volatile species. The presence of a wide variety of stationary phases with contrasting selectivities and easy column preparation add to the assets of GLC or simply GC.




It may be wise to introduce instrumental components before proceeding further in theoretical background. This will help clarify many points which may, otherwise, seem vague. The following schematic shows a typical GC instrument. It should also be noted that a detector will require special gas cylinders depending on the detector type utilized. The column temperature controller is simply an oven, the temperature of which can be varied or programmed.



Three temperature zones can be identified:

a.    Injector temperature, TI, where TI should allow flash vaporization of all sample components.

b.    Column temperature, Tc, which is adjusted as the average boiling points of sample components.

c.    Detector Temperature, TD, which should exclude any possible condensation inside the detector.


Generally, an intuitive equation can be used to adjust all three zones depending on the average boiling point of the sample components. This equation is formulated as:

TI = TD = Tc + 50 oC


The Carrier Gas


Unlike liquid chromatography where wide varieties of mobile phase compositions are possible, mobile phases in gas chromatography are very limited. Only slight changes between carrier gases can be identified which places real limitations to chromatographic enhancement by change or modification of carrier gases. A carrier gas should have the following properties:

1.    Highly pure (> 99.9%)

2.    Inert so that no reaction with stationary phase or instrumental components can take place, especially at high temperatures.

3.    A higher density (larger viscosity) carrier gas is preferred.

4.    Compatible with the detector since some detectors require the use of a specific carrier gas.

5.    A cheap and available carrier gas is an advantage.



Longitudinal Diffusion Term


This is an important factor contributing to band broadening which is a function of the diffusivity of the solute in the gaseous mobile phase as well as the molecular diffusion of the carrier gas itself.

HL = K DM /V


Where; DM is the diffusion coefficient of solute in the carrier gas. This term can be minimized when mobile phases of low diffusion, i.e. high density, are used in conjunction with higher flow rates.


The same van Deemter equation as in LC can be written for GC where:


H = A + B/V + CV


The optimum carrier gas velocity is given by the derivative of van Deemter equation

Vopt = { B/C }1/2

However, the obtained velocity is much greater than that obtained in LC.


The carrier gas pressure ranges from 10-50 psi. Higher pressures potentially increase compression possibility while very low pressures result in large band broadening due to diffusion. Depending on the column dimensions, flow rates from 1-150 mL/min are reported. Conventional analytical columns (1/8) usually use flow rates in the range from 20-50 mL/min while capillary columns use flow rates from 1-5 mL/min depending on the dimensions and nature of column. In most cases, a selection between helium and nitrogen is made as these two gases are the most versatile and common carrier gases in GC.


Retention Volumes


The volume of the mobile phase needed to elute a solute is called the retention volume (VR) of that solute. The retention volume can be calculated by multiplying the retention time times the flow rate. The dead or void volume can also be calculated by multiplication of the dead time times flow rate. The average flow rate (F) of carrier gas inside a column can be calculated from the relation:


F = Fm (Tc/T) {(P PH2O)/P}


Where; Fm is the measured flow rate at the column outlet, Tc is the column temperature in Kelvin, T and P are the ambient temperature and pressure, and PH2O is the vapor pressure of water when Fm is measured by a soap-bubble meter. The specific retention volume of a solute, Vg, can be calculated from the relation:


Vg = {jF(tR tM)/m}(273/Tc)


Where, m is the mass of stationary phase and j is a pressure drop factor defined as:

j = 3[(Pi/P)2 -1]/ 2[(Pi/P)3 -1]

Where; Pi is the inlet pressure and P is the ambient pressure (outlet pressure).


Relationship between Vg and Distribution Constant


The specific void volume which uses corrected retention volumes is defined as:


We have from previos derivations that k' = (tR tM)/tM. Substitution in the Vg equation gives:


Vg = {jFtMk' /m}(273/Tc)

VoM = jtMF

Vg = {VoMk' /m}(273/Tc)

VoM is the same as the carrier gas volume and equivalent to VM hence k' = KVS/ VoM


Vg = {KVs/m}(273/Tc)


However, the density of the stationary phase, r, is defined as:

r = m/VS

Substitution gives:


Vg = {K/r}(273/Tc)


This relation implies and proves that the retention volume of a specific solute is related to its distribution constant as well as column temperature and density of the stationary phase.




Septum type injectors are the most common. These are composed of a glass tube where vaporization of the sample takes place. The sample is introduced into the injector through a self-sealing silicone rubber septum. The carrier gas flows through the injector carrying vaporized solutes. The temperature of the injector should be adjusted so that flash vaporization of all solutes occurs. If the temperature of the injector is not high enough (at least 50 degrees above highest boiling component), band broadening will take place. A schematic representation of a septum injector can be shown in the figure below:




There are several types of injectors and injection techniques. However, due to the limited time and nature of this course no further discussion of these techniques will be presented here.


Column Configurations and Ovens


The column in chromatography is undoubtedly the heart of the technique. A column can either be a packed or open tubular. Traditionally, packed columns were most common but fast developments in open tubular techniques and reported advantages in terms of efficiency and speed may make open tubular columns the best choice in the near future. Packed columns are relatively short (~2meters) while open tubular columns may be as long as 30-100 meters. Packed columns are made of stainless steel or glass while open tubular columns are usually made of fused silica. The temperature of the column is adjusted so that it is close to the average boiling point of the sample mixture. However, temperature programming is used very often to achieve better separations. The temperature of the column is assumed to be the same as the oven which houses the column. The oven temperature should be stable and easily changed in order to obtain reproducible results.


Detection Systems


Several detectors are available for use in GC. Each detector has its own characteristics and features as well as drawbacks. Properties of an ideal detector include:

  1. High sensitivity
  2. Minimum drift
  3. Wide dynamic range
  4. Operational temperatures up to 400 oC.
  5. Fast response time
  6. Same response factor for all solutes
  7. Good reliability (no fooling)
  8. Nondestructive
  9. Responds to all solutes (universal)




Several detectors are used in GC with varying characteristics, advantages, and limitations. Two widely used detectors will be briefly discussed.


a. Thermal Conductivity Detector (TCD)


This is a nondestructive detector which is used for the separation and collection of solutes to further perform some other experiments on each purely separated component. The heart of the detector is a heated filament which is cooled by helium carrier gas. Any solute passes across the filament will not cool it as much as helium does because helium has the highest thermal conductivity. This results in an increase in the temperature of the filament which is related to concentration. The detector is simple, nondestructive, and universal but is not very sensitive and is flow rate sensitive.



Note that gases should always be flowing through the detector including just before, and few minutes after, the operation of the detector. Otherwise, the filament will melt. Also, keep away any oxygen since oxygen will oxidize the filament and results in its destruction.


Remember that TCD characteristics include:


Wide dynamic range (105)


Insensitive (10-8 g/s)

Flow rate sensitive


b. Flame Ionization Detector (FID)


This is one of the most sensitive and reliable destructive detectors. Separate two gas cylinders, one for fuel and the other for O2 or air are used in the ignition of the flame of the FID. The fuel is usually hydrogen gas. The flow rate of air and hydrogen should be carefully adjusted in order to successfully ignite the flame.



The FID detector is a mass sensitive detector where solutes are ionized in the flame and electrons emitted are attracted by a positive electrode which is shown as a current and a signal is obtained.

The FID detector is not responsive to air, water, carbon disulfide. This is an extremely important advantage where volatile solutes present in water matrix can be easily analyzed without any pretreatment.


Remember that FID characteristics include:



Sensitive (10-13 g/s)

Wide dynamic range (107)

Signal depends on number of carbon atoms in organic analytes which is referred to as mass sensitive rather than concentration sensitive

Weakly sensitive to carbonyl, amine, alcohol, amine groups

Not sensitive to non-combustibles - H2O, CO2, SO2, NOx



Electron Capture Detector (ECD)


This detector exhibits high intensity for halogen containing compounds and thus has found wide applications in the detection of pesticides and polychlorinated biphenyls. The mechanism of sensing relies on the fact that electronegative atoms, like halogens, will capture electrons from a b emitter (usually 63Ni). In absence of halogenated compounds, a high current signal will be recorded due to high ionization of the carrier gas which is N2, while in presence of halogenated compounds the signal will decrease due to lower ionization. The detector can be schematically represented by the figure:


Qualitative and

Remember the following facts about ECD:

1. Electrons from a b-source ionize the carrier gas (nitrogen)

2. Organic molecules containing electronegative atoms capture electrons and decrease current

3. Simple and reliable

4. Sensitive (10-15 g/s) to electronegative groups (halogens)

5. Largely non-destructive

6. Insensitive to amines, alcohols and hydrocarbons

7. Limited dynamic range (102)

8. Mass sensitive detector



Important Other Detectors:


AES, AAS, chemiluminescent reaction (S), mass spectrometer, FTIR


Gas Chromatographic Columns and Stationary Phases


Packed Columns


These columns are fabricated from glass, stainless steel, copper, or other suitable tubes. Stainless steel is the most common tubing used with internal diameters from 1-4 mm. The column is packed with finely divided particles (<100-300 mm diameter) which is coated with stationary phase. However, glass tubes are also used for large scale separations. Several types of tubing were used ranging from copper, stainless steel, aluminum and glass. Stainless steel is the most widely used because it is most inert and easy to work with. The column diameters currently in use are ordinarily 1/16" to 1/4" 0.D. Columns exceeding 1/8" are usually used for preparative work while the 1/8" or narrower columns have excellent working properties and yield excellent results in the analytical range. These find excellent and wide use because of easy packing and good routine separation characteristics. Column length can be from few feet for packed columns to more than 100 ft for capillary columns.


Capillary/Open Tubular


Open tubular or capillary columns are finding broad applications. These are mainly of two types:


Wall-coated open tubular (WCOT) <1 mm thick liquid coating on inside of silica tube

Support-coated open tubular (SCOT) 30 mm thick coating of liquid coated support on inside of silica tube


These are used for fast and efficient separations but are good only for small samples. The most frequently used capillary column, nowadays, is the fused silica open tubular column (FSOT) which is a WCOT column. The external surface of the fused silica columns is coated with a polyimide film to increase their strength. The most frequently used internal diameters occur in the range from 260-320 micrometer. However, other larger diameters are known where a 530 micrometer fused silica open tubular column was recently made and is called a megapore column, to distinguish it from other capillary columns. Megapore columns tolerate a larger sample size.




It should be noted that since capillary columns are not packed with any solid support, but rather a very thin film of stationary phase which adheres to the internal surface of the tubing, the A term in the van Deemter equation which stands for multiple path effects is zero and the equation for capillary columns becomes


H = B/V + CV


Solid Support Materials


The solid support should ideally have the following properties:

  1. Large surface area (at least 1 m2/g)
  2. Has a good mechanical stability
  3. Thermally stable
  4. Inert surface in order to simplify retention behavior and prevent solute adsorption
  5. Has a particle size in the range from 100-400 mm


Selection of Stationary Phases


General properties of a good liquid stationary phase are easy to guess where inertness towards solutes is essential. Very low volatility liquids that have good absolute and differential solubilities for analytes are required for successful separations. An additional factor that influences the performance of a stationary phase is its thermal stability where a stationary phase should be thermally stable in order to obtain reproducible results. Nonvolatile liquids assure minimum bleeding of the stationary phase.


It was suggested that a stationary phase similar in chemical structure to solutes should be used. This means that polar solutes are best separated on polar stationary phases and vice versa.

Usually, the liquid stationary phase is dissolved in a suitable solvent, mixed with the solid support and is packed through the chromatographic column. Evaporation of the solvent leaves the solid support coated with a thin film of the liquid stationary phase. The amount of stationary phase is critical and is usually carefully chosen where:


Weight of liquid stationary phase * 100%

% Loading = __________________________________

Weight of stationary phase plus solid support


Increasing percent loading would allow for increased sample capacity and cover any active sites on the solid support. These two advantages are very important, however increasing the thickness of stationary phase will affect the C term in the van Deemter equation by increasing HS, and therefore Ht. A compromise is always necessary depending on the nature of the sample in question. If the sample is volatile, a relatively high percent loading (about 20%) may be necessary while about 3% loading is enough for materials of low volatility. This helps good mass transfer and better resolution. Generally, the film thickness primarily affects the retention character and the sample capacity of a column. Thick films are used with highly volatile analytes, because such films retain solutes for a longer time and thus provide a greater time for separation to take place. Thin films are useful for separating species of low volatility in a reasonable time. On the other hand, a thicker film can tolerate a larger sample size. Film thicknesses in the range from 0.1 5 mm are common.



Liquid Stationary Phases


In general, the polarity of the stationary phase should match that of the sample constituents ("like" dissolves "like"). Most stationary phases are based on polydimethylsiloxane or polyethylene glycol (PEG) backbones:



The polarity of the stationary phase can be changed by derivatization with different functional groups such as a phenyl group. Bleeding of the column is cured by bonding the stationary phase to the column or crosslinking the stationary phase.


In summary, immobilized Liquid Stationary Phases should have the following characteristics:

Low volatility

High decomposition temperature (thermally stable)

Chemically inert (reversible interactions with solvent)

Chemically attached to support (to prevent bleeding)

Appropriate k' and a for good resolution



Bonded and Crosslinked Stationary Phases


The purpose of bonding and cross-linking is to prevent bleeding and provide a stable stationary phase. With use at high temperatures, stationary phases that are not chemically bonded or crosslinked slowly lose their stationary phase due to bleeding in which a small amount of the physically immobilized liquid is carried out of the column during the elution process. Crosslinking is carried out in situ after a column is coated with one of the polymers listed in the table below.


Modification of Stationary Phases


Stationary phases cab be modified by derivatization or addition of chemical reagents, for instance a stationary phase can be modified to allow for more selective separations where silver modified stationary phases show excellent characteristics for the separation of unsaturated compounds. Chiral groups can be included in a stationary phase to improve separations of enantiomeric mixtures. Stationary phases that have chiral (optically active) groups are called chiral stationary phases and have proved very useful for separation of chiral compounds. In some situations, analytes can be derivatized to form optically active reagents that can be separated on a chiral column.



Examples of Liquid Stationary Phases and Their Applications


Stationary Phase

Common Trade Name

Maximum Temperature (oC)

Common Applications

Polydimethyl siloxane

OV-1, SE-30


General-purpose nonpolar phase; hydrocarbons; polynuclear aromatics; drugs; steriods; PCBs

Poly(phenylmethyldimethyl) siloxane (10% phenyl)

OV-3, SE-52


Fatty acid methyl esters; alkaloids; drugs; halogenated compounds

Poly(phenylmethyl) siloxane (50% phenyl)



Drugs; steriods; pesticides; glycols

Poly(trifluoropropyldimethyl) siloxane



Chlorinated aromatics; nitroaromatics; alkyl-substituted benzenes

Polyethylene glycol

Carbowax 20M


Free acid; alcohol; ethers; essential oils; glycols

Poly(dicyanoallyldimethyl) siloxane



Polyunsaturated fatty acids; rosin acids; free acids; alcohols



In summary, stationary phases are usually bonded and/or crosslinked and the following remarks are usually helpful:


1. Bonding occurs through covalent linking of stationary phase to support

2. Crosslinking occurs through polymerization reactions to join individual stationary phase molecules

3. Nonpolar stationary phases are best for nonpolar analytes where nonpolar analytes are retained preferentially

4. Polar stationary phases are best for polar analytes where polar analytes are retained preferentially



Adsorption of solutes on Solid Support and Column Walls


A real problem in both LC and GC has been the physical adsorption of polar or polarizable analyte species such as alcohols or aromatic hydrocarbons, on the silicate surfaces of column packings or capillary walls. Adsorption results in distorted peaks, which are broadened and are tailed. Adsorption is the consequence of residual silanol groups that are present on the surface of silicates. The silanol (SiOH) groups on the support surface or walls of capillary columns have a strong affinity for polar organic molecules and tend to retain them by adsorption. Adsorption on support materials can be deactivated by silanization with dimethylchlorosilane (DMCS).






Gas-liquid chromatography (GLC)

Packed columns are fabricated from glass, metal, or Teflon with 1 to 3 m length and 2 to 4 mm in internal diameter. The column is packed with a solid support (100-400 mm particle diameter made from diatomaceous earth) that has been coated with a thin layer (0.1-5 mm) of the stationary liquid phase. Efficiency increases with decreasing particle size as predicted from van Deemter equation. The retention is based on absorption of analyte (partition into the liquid stationary phase) where solutes must have differential solubility in the stationary phase.

Open tubular capillary columns, either WCOT, SCOT are routinely used. In WCOT the capillary is coated with a thin film (0.1-0.25 mm) of the liquid stationary phase while in SCOT a thin film of solid support material is first affixed to the inner surface of the column then the support is coated with the stationary phase. WCOT columns are most widely used. Capillary columns are typically made from fused silica (FSOT) and are 15 to 100 m long with 0.10 to 0.5 mm i.d. The thickness of the stationary phase affects the performance of the column as follows:

  1. Increasing thickness of stationary phase allows the separation of larger sample sizes.
  2. Increasing thickness of stationary phase reduces efficiency since HS increases.
  3. Increasing thickness of stationary phase is better for separation of highly volatile compounds due to increased retention.

Much more efficient separations can be achieved with capillary columns, as compared to packed columns, due to the following reasons:

1.      Very long capillary columns can be used which increases efficiency

2.      Thinner stationary phase films can be used with capillary columns

3.      No eddy diffusion term (multiple paths effect) is observed in capillary columns


Gas-solid chromatography (GSC)


Gas-solid chromatography is based upon adsorption of gaseous substances on solid surfaces. Distribution coefficients are generally much larger than those for gas-liquid chromatography. Consequently, gas-solid chromatography is useful for the separation of species that are not retained by gas-liquid columns, such as the components of air, hydrogen sulfide, carbon disulfide, nitrogen oxides, and rare gases. Gas-solid chromatography is performed with both packed and open tubular columns. For the latter, a thin layer of adsorbent is affixed to the inner walls of the capillary. Such columns are sometimes called porous layer open tubular columns (PLOT columns). Two common types of adsorbents are molecular sieves and porous polymers.


Molecular Sieves


Molecular sieves are metal aluminum silicate ion exchangers, whose pore size depends upon the kind of cation present, like sodium in sodium aluminum silicate molecular sieves. The sieves are classified according to the maximum diameter of molecules that can enter the pores. Commercial molecular sieves come in pore sizes of 4, 5, 10, and 13 angstroms. Molecular sieves can be used to separate small molecules from large ones.


Porous Polymers


Porous polymer beads of uniform size are manufactured from styrene crosslinked with divinylbenzene. The pore size of these beads is uniform and is controlled by the amount of crosslinking. Porous polymers have found considerable use in the separation of gaseous species such as hydrogen sulfide, oxides of nitrogen, water, carbon dioxide, methanol, etc.

Temperature Zones in GC


Three temperature zones should be adjusted before a GC separation can be done. The injector temperature should be such that fast evaporation of all sample components is achieved. The temperature of the injector is always more than that of the column which depends on the operational mode of the separation. The detector temperature should be kept at some level so as to prevent any solute condensation in the vicinity of the detector body.


Temperature Programming


Gas chromatographs are usually capable of performing what is known as temperature programming gas chromatography (TPGC). The temperature of the column is changed according to a preset temperature isotherm. TPGC is a very important procedure which is used for the attainment of excellent looking chromatograms in the least time possible. For example, assume a chromatogram obtained using isothermal GC at 80 oC, as shown below:



The third component elutes 11 minutes after the second component and an analyst has to wait all this time to finish the analysis. In TPGC, the separation is conducted at 80 oC for the first 8 min and then the temperature is increased several degrees per minute till a chromatogram like that appearing in the figure below is obtained.


Temperature isotherms of practically any shape can be constructed to affect suitable separations. If it is required to separate some overlapping peaks, it is always wise to reduce the temperature rather than flow rate because flow rate is troublesome to adjust. It is also observed that broad peaks obtained in isothermal separations are changed into sharp looking peaks in TPGC which is also an additional advantage of the technique.

An experimental example for a TPGC separation is also shown below:






Qualitative and Quantitative Analysis


GC is an excellent quantitative technique where peak height or area is proportional to analyte concentration. Thus the GC can be calibrated with several standards and a calibration curve is obtained, then the concentration of the unknown analyte can be determined using the peak area or height. The detector response factor for each analyte should be considered for accurate quantitative analysis.

Gas chromatographs are widely used as criteria for establishing the purity of organic compounds. Contaminants, if present, are revealed by the appearance of additional peaks. Qualitative Analysis is usually done by comparison with retention times of standards which are very reproducible in GC provided good injection practices are followed. Injection should be done with a suitable Hamilton type syringe through the heated septum injector till all needle disappears, then the needle is drawn back as steadily and fast as possible. This is important for reproducible attainment of retention times.


The Retention Index


The retention index, I, was first proposed by Kovats in 1958 as a parameter for identifying solutes from chromatograms. The retention index for any given solute can be derived from a chromatogram of a mixture of that solute with at least two normal alkanes (chain length >four carbons) having retention times that bracket that of the solute. That is, normal alkanes are the standards upon which the retention index scale is based. By definition, the retention index for a normal alkane is equal to 100 times the number of carbons in the compound regardless of the column packing, the temperature, or other chromatographic conditions. The retention index system has the advantage of being based upon readily available reference materials that cover a wide boiling range. The retention index of a compound is constant for a certain stationary phase but can be totally different for other stationary phases. In finding the retention index, a plot of the number of carbons of standard alkanes against the logarithm of the adjusted retention time is first constructed. The value of the logarithm of the adjusted retention time of the unknown is then calculated and the retention index is obtained from the plot.


The adjusted retention time, tR, is defined as:


tR = tR - tM





Interfacing GC with other Methods


As mentioned previously, chromatographic methods (including GC) use retention times as markers for qualitative analysis. However, this characteristic does not absolutely confirm the existence of a specific analyte as many analytes may have very similar stationary phases. GC, as other chromatographic techniques, can confirm the absence of a solute rather than its existence. When GC is coupled with structural detection methods, it serves as a powerful tool for identifying the components of complex mixtures. A popular combination is GC/MS. This interface has been used for the identification of hundreds of components that are present in natural and biological systems. For example, these procedures have permitted characterization of the odor and flavor components of foods, identification of water pollutants, medical diagnosis based on breath components, and studies of drug metabolites. Another popular interface is that of GC/FTIR where the IR spectra of analytes are compared to standards for identification.




Since GC is utilized in the analysis of all types of volatile samples including carcinogens and poisons, the gas chromatograph should be placed in a well ventilated atmosphere.