Specific chemical reactions after the CC separation of initial mixtures into their components allow one to advance from the identification of groups in compounds (typical of classical qualitative analysis) to the identification of particular substances.
In this instance chromatography yields additional information about the nature of substances by means of retention volumes. In combination with qualitative reactions performed after chromatographic separation these data allow one to identify unam- biguously almost all volatile compounds. At the same time chromatography serves as a micro-preparative method of isolating individual components. We can distinguish the two most useful procedures of performing rapid qualitative analysis. The first consists
REACTIONS AFTER SEPARATION
A
Retention VOlUme Number of carbon atoms
Fig. 9.1. (A) Chromatogram of a tencomponent mixture of organic compounds of various classes.
Temperature, 125°C; stationary phase, squalane; column length, 1 m; sample volume, 1 ~ 1 . (B) Charac- teristic graphs for identifying the compounds corresponding to the chromatographic peaks. 1 = Alcohols; 2 = ketones; 3 = carboxylic methyl esters; 4 = methyl ketones; 5 = mercaptans; 6 = alde- hydes; 7 = aromatic hydrocarbons. Reprinted with permission from ref. 74.
in dividing an eluate leaving a column into several approximately equal flows, which are collected in test-tubes containing reagents selective towards certain classes of compounds [74]. A change in colour in one of the test-tubes at the instant a peak appears indicates that the substance concerned belongs to a particular class. The results of qualitative functional group analysis of the compounds corresponding to a certain chromatographic peak serve as a basis for choosing a calibration graph of the logarithm of the retention value (or Kovits retention index) of the compounds of the given homologous series as a function of the number of carbon atoms in the compound (see, for instance, ref. 75).
Relationships of this type are widely used for identifying organic compounds. As an example, Fig. 9.1A illustrates a chromatogram of a ten-component system and Fig. 9.1B
shows the characteristic curves for identifying the compounds corresponding to the chromatographic peaks [74] . Chemical determination of the functional groups showed that peak I corresponds to alcohol and ketone, peak I1 to ketone and ester, peak 111 to mercaptan, peak lV to aldehyde, alcohol and aromatic hydrocarbon, and peak V t o aldehyde and aromatic hydrocarbon. With the aid of characteristic linear dependences (Fig. 9.1B) it was concluded that peak I corresponds to ethanol and acetone, peak I1 to methyl propionate and methyl ethyl ketone, peak 111 to propylmercaptan, peak IV t o butanol, pentanal and benzene and peak V to hexanal and toluene.
The practical application of this method may present certain difficulties as different groups are characterized by the same group reaction (for instance, aromatic and aliphatic unsaturated compounds react positively t o a Lerozene test). If two groups give a positive reaction towards the same reagent, the additional problem arises of whether a certain chromatographic zone corresponds to one bifunctional compound or to two different monofunctional compounds with similar chromatographic properties. The characteristic curves usually answer these questions. Thus, when studying a certain zone, positive reactions towards Schiff and Lerozene reagents were obtained. The characteristic curves for unsaturated aldehydes, olefins and saturated aldehydes have shown that the above chromatographic zone can correspond only to an unsaturated aldehyde with five carbon atoms, as the other two curves contain no compound with such a retention time. The main idea of Walsh and Merritt [74] has been presented here in some detail as other methods are based on the same combination of functional group analysis and retention data for identifying the compounds.
The Walsh and Merritt method [74] is advantageous in that it is possible to study each peak for different functional groups simultaneously and to determine the nature of the micro-components by subsequent injection of several samples without replacing the absorptive reagents, provided that the colour of the solution is stable for a sufficiently long period of time. However, the low sensitivity of colour reactions requires large aniounts of sample (not less than 20-IOOpg) [74], which considerably limits the appiicability of the method.
This method can be applied to the analysis of relatively volatile compounds. For hgh-boiling compounds the procedure can be changed. An additional intermediate stage is performed, namely trapping of the chromatographic zones separated in the column for their subsequent study by qualitative reactions. Sometimes it is useful to use tubes with a selective reagent for trapping and group determination. Thus, Chriswell and Fritz [76] used a column filled with a selective complexing agent containing copper(I1) for trapping amines. The trapping of amines is accompanied by the formation of a coloured zone whose length is proportional to the amount of amines absorbed.
The detection limit is several parts per million. The absorbed amines can be isolated by treating the sorbent with a potassium hydroxide solution and then studied by GC.
The sensitivity of the method based on partition of an eluate flow can be enhanced by concentrating the substance in short glass porous traps filled with aluminium oxide [ 7 7 ] . When the flow-rate through the trap is about 5 ml/min, most of the compounds (aldehydes, ketones, alcohols, esters, aromatics and sulphur-containing compounds) are trapped by a coat 2 - 3 mm long in the form of concentrated zones, which decreases
295 the detection limit to 0.1-l.Obg. Also, although the sensitivity remains far from that required, the procedure allows smaller amounts of the compounds to be used.
Cronin and Gilbert [78] developed a technique for detecting nitrogen-containing Compounds after GC separation by hydrogenolysis and colour reactions. Colour reactions for detecting ammonia formed during hydrogenolysis are conducted with the aid of reagents applied to the walls of the column.
Chemical reagents for determining the nature of the functional groups in the com- pounds eluted from a gas chromatograph were used by Gine and Cerda [79]. The limit of detection of carboxylic acids, amines and thiols was 1-3Opg.
Casu and Cavallotti [80] developed a sophisticated device for conducting qualitative reactions after chromatographic separation, in which the compounds separated by GC are automatically analysed with the aid of a ribbon impregnated with reagents and moving relative to the chromatograph outlet If the eluate leaving the column contains a compound of a particular class, a coloured spot appears on the sorbent layer. The class of compound corresponding to each peak can easily be determined by changing the reagent and comparing the results. The qualitative reactions conducted after GC sepa- ration were reviewed by Walsh and Merritt [74].
A useful tool in functional group analysis is a combination of gas chromatography with thin-layer chromatography [80-83 J . An eluate leaving a chromatographic column passes through a heated connecting tube and is directed to a plate coated with a sorbent layer which moves continuously or stepwise relative to the chromatograph outlet. Thus, the components separated by GC are adsorbed on the start line and can be further analysed by thin-layer chromatography. Such a combination has additionally at least two merits. The first is a more complete separation of the components which, for some reason, cannot be attained by GC. The second resides in the use for functional group analysis of optical methods that have been well developed in thin-layer chromatography (fluorescence, formation of coloured compounds, etc.), which makes it possible to identify both whole molecules and their components by choosing a specific chemical reaction.
Of great interest is the combined use of crystallographic and colorimetric methods by Janik [84] for identifying nano- and picogram amounts of compounds in GC eluates.
The fractions of high-boiling compounds (diphenyl, acenaphthene, fluorene and phenan- threne) leaving a chromatographic column condensed on a pure glass plate moving in a perpendicular direction (for more volatile compounds a cooled metal plate was used) as well separated zones consisting of microcrystals. The distribution of crystals inside the zones coincides with a record of the chromatographic peak: the zones are bounded by a fine crystal ‘colour’ whereas in the centre of the zone well separated microcrystals are located whose type and shape are typical of the given compound. When ‘chromatograms’
were obtained on the surface of the plate, the crystals in amounts of several nanograms were transferred to a microscope with a special glass micro-spatula.
A procedure was developed [84] for performing. colour reactions in a microscope.
A crystal under study is placed on a silica gel granule and dissolved in&enzene with the aid of a micropipette with an inner diameter 0.07 mm. The solution is instantly absorbed on silica gel and in 5-l0sec the solvent (benzene) is evaporated. The ratio between the amount of the substance and silica gel is 1:102 which exceeds by three orders of
magnitude the visibility limit of the colour reaction of the formation of a tetracyano- ethylene complex with aromatic compounds on a silica gel surface. After adding a benzene solution of tetracyanoethylene (0.05-0.1 111) to the sample compound absorbed by silica gel, an intensely coloured complex is formed whose colour can be reliably measured with a microscope even if the weight of the sample compound is no more than 20ng. The typical colours observed for some classes of the compounds (hydro- carbons and oxygen-, nitrogen- and sulphur-containing compounds) were given [ 8 4 ] .
A similar technique making use of a moving plate in liquid chromatography was proposed I851.