Info

number of carbon atoms

Fig. 4.4. Relationships between the number of carbon atoms in a homologous series of compounds and the calibration constants for the GUT permeation passive samplers equipped with PDMS membranes of 50 mm thickness.

number of carbon atoms

Fig. 4.4. Relationships between the number of carbon atoms in a homologous series of compounds and the calibration constants for the GUT permeation passive samplers equipped with PDMS membranes of 50 mm thickness.

Fig. 4.5. Relationship between the molecular mass of a compound and the calibration constant for the GUT permeation passive sampler equipped with the PDMS membrane of 50 mm thickness. Equation of the regression line is k = —0.00019x+0.330 (r2 = 0.824), where x is molecular mass of the analyte.

4.5.2 Molecular mass

Physico-chemical properties of a compound, including its molecular weight and boiling point, are more closely related to the structure of a molecule than just the number of carbon atoms. Thus, they should be more useful when trying to predict the calibration constants for a broader range of compounds (not necessarily members of a homologous series).

Figure 4.5 presents the relationship between the molecular mass of a compound (for all four classes of compounds) and the calibration constant k [13]. In general, the relationship was linear, with a high value of the linear correlation coefficient (r2 — 0.824). The confidence band of the calculated calibration constants is also plotted to help visualize the estimated range of values that an unknown compound might have [25,26]. It should be emphasized that for the relationship obtained, none of the compounds included in the study fell outside the 95% confidence band.

4.5.3 Boiling point temperature

Boiling point of a compound depends mostly on the strength of intermolecular interactions in the liquid phase, which are determined by the structure of a compound and the polarity of the functional groups present in the molecule. The same factors affect the solubility of a compound in PMDS, therefore the calibration constant of a permeation passive sampler should be correlated to the boiling point of the analyte. Figure 4.6 shows the relationship between the experimentally determined calibration constants and the boiling points of the analytes [13]. Again, a linear relationship was obtained, with the linear correlation coefficient of r2 = 0.814.

The results presented in Figs. 4.5 and 4.6 demonstrate that the calibration constant k of a permeation passive sampler can be estimated with reasonable accuracy from the molecular weight of a compound or its boiling point, without the need for experimental calibration for each individual compound. Also, the confidence band of the calculated calibration constants is plotted to help visualize the estimated range of values that an unknown compound might have. Again, none of the studied compounds fell outside the 95% confidence band.

The choice of one of the two descriptors (molecular weight or boiling point temperature) is somewhat arbitrary and depends on the nature of the compound. In fact, the best results were obtained by averaging the k values obtained by the two methods. Our experiments were carried out for analytes whose molecular weights ranged from 72 (n-pentane) to 156 (n-undecane) and the boiling points ranged from 36°C (n-pentane) to

Fig. 4.6. Relationship between the boiling point of a compound and the calibration constant for the GUT permeation passive sampler equipped with the PDMS membrane of 50 mm thickness. Equation of the regression line is k = —0.00096x+0.250 (r2 = 0.814), where x is the boiling point temperature of the analyte.

Fig. 4.6. Relationship between the boiling point of a compound and the calibration constant for the GUT permeation passive sampler equipped with the PDMS membrane of 50 mm thickness. Equation of the regression line is k = —0.00096x+0.250 (r2 = 0.814), where x is the boiling point temperature of the analyte.

183°C (butylbenzene). These ranges cover many organic compounds relevant to air analysis. The correlations established could probably be applied for compounds outside of these ranges, but care should be exercised when using this approach. The course of the regression line obtained for the relationship between the boiling point of a compound and the calibration constant of the passive sampler indicates that this regression can only be used for compounds whose boiling point is below 250°C.

4.5.4 Linear temperature-programmed retention index system

In field measurements, very often the identity of all compounds present in the sample is not known. The approaches proposed thus far cannot be used for obvious reasons for unknown analytes, which normally precludes even rough estimation of the total load of organics in the air when using permeation passive samplers. It would be very advantageous to be able to estimate the calibration constants for all compounds present in a sample. The knowledge of the calibration constants could then be used to quantify the unknown compounds provided that the response factors of the detector towards these compounds were known [13,16].

In the process of identification of organic compounds in complex mixtures, the use of retention indices is becoming more important [27]. This is a result of the outstanding stability of fused silica capillaries and the excellent reproducibility of the now available GCs [16]. Retention index (RI) is a measure of the retentiveness of a compound relative to straight (normal) chain hydrocarbons under given set of chromato-graphic conditions. In 1958, Kovats proposed the use of the homologous series of n-alkanes as retention markers [28]. The original Kovats retention index system was applicable to isothermal separations only. Since most separations in GC are carried out these days under temperature-programmed conditions, the linear temperature-programmed retention index system (LTPRI) proposed by Van den Dool and Kratz [29] is used much more often today. LTPRI of a substance is calculated according to the following formula:

0 0

Post a comment