Indirect
chromatographic method for vapor pressure determination
Abstract: Indirect chromatographic methods for evaluation of vapor pressures from retention times
(GLC-VP) are fast and commercial gas liquid chromatographs can be used without any modifications. As show by our group, vast majority of previously published data is burdened with significant systematic errors. Combination of GLC-VP method with our four self-made apparatuses seems to be promising in selected important applications.
Indirect chromatographic methods for evaluation of vapor pressures from retention times (GLC-VP) used to be very popular in the community of environmental chemists. This can be easily understood, as "classical" methods are typically not suited for measurements in millipascal to micropascal pressure range required for environmental applications (most frequently for evaluation of chemicals fate in the environment). Moreover, GLC-VP method has obvious advantages over traditional methods; impurities do not influence results, small amount of sample is required and measurements are relatively fast (one experimental point can be measured within hours or even minutes). Remarkably, it yields vapor pressures for supercooled liquid, required in environmental applications and not accessible by direct methods; recalculation from solid state to supercooled liquid state is however impossible without uncertain extrapolations or some approximations [1; 2]
Though sporadically used in sixties, GLC-VP paramount was in nineties. Review papers by Koutek et al. [3] , Letcher and Naicker [4], Delle Site[5] and Ruzicka et al.[6] list most of GLC-VP papers up to ca 2010. In last decade, number of authors using GLC-VP was relatively small (with the exception of Chickos and coworkers [7]).
Main methods for extraction of vapor pressures from GLC retention times
Roughly, indirect chromatographic methods for evaluation of vapor pressures (GLC-VP) can be divided into three main categories:
- Single reference approach [8] (GLC-RT1S in the remainder of the text)
- Double reference approach [9] (in most cases, Kovats indices are used, GLC-RTKI in the remainder of the text)
- Multiple reference approach (in most cases, so called "correlation gas chromatography" , its performance is now under investigation in our laboratory)
Serious systematic error present in most papers using Single reference approach
While analysis of double reference approach shown that in some cases (missing members in homologous series) the uncertainty od few percent can be achieved [9], single reference approach is prone to significant errors.
Some of those errors could be removed:
(i) selection of better data for reference compound(s) could improve results a
bit [6];
(ii) presmoothing suggested by Hamilton [10]
should be avoided (it can hide low quality chromatographic datasets/datapoints;
our recommendation is quite opposite - use Arc plot representation to reveal
suspected datapoints)
(iii) avoiding long extrapolation could yield better data (albeit not at
preferred temperature 298 K; thermodynamically controlled extrapolation
(SimCor) should always be preferred to simple extrapolation [6; 8]).
(iv) for recalculation from sublimation pressures to supercooled liquid vapor
pressures, entropy of melting should not be taken as a constant.
In majority of papers, retention times as a function
of temperature are however not published, preventing any such corrections.
There is however one more serious issue. When using
single reference compound (e.g. n-eicosane) for several unknown samples, it is
not possible to expect that the resulting error will be the same for all
unknown samples.
This is illustrated at the figure below.
Figure 1: Relative errors in vapor pressures ∆prel obtained with GLC-RT1S methodology [8] for selected n-alkanes with n-alkanes as reference compounds. ∆prel stands for 100(pGC−prec)/prec, where pGC is the vapor pressure obtained with GLC-RT1S methodology and prec is the recommended vapor pressure. CX stands for n-alkane with X carbon atoms.
This figure is based of Figure S2 in [8] and demonstrates that using n-hexadecane as a reference compound would lead to errors -10%, +25%, and +55% for n-pentadecane, n-octadecane, and n-eicosane, respectively.
As summarized in Conclusion section of Koutek et al. [8] GLC-RT1S approach lead to errors of ca 20 % for n-alkanes, 1-chloroalkanes, alkylbenzenes, and SOME chlorobenzenes, while errors were much higher for MOST studied chlorobenzenes and completely unacceptable for polar compounds (alcohols, nitriles).
Potential for correct extraction of vapor pressures from GLC retention times using Double reference approach
Though we
have shown that vast majority of data obtained using these methods was
seriously biased due to improperly selected reference data [6] or inappropriate methodology [8], we have found that GLC-VP method can be
effectively used for evaluation of missing data in homologous series [9] and we presume that it can be very useful
when combined with traditional (time and effort demanding) methods. This topic is being investigated in our laboratory and results for polyaromatic hydrocarbons are promissing [11]. Our yet unpublished data also suggest that GLC-VP method can be used for extrapolation of reliable VP obtained via reliable direct methods.
Summary of recommendations for extracting vapor pressures from GLC retention times
Some findings of our studies can be summarized as practical recommendations for those using GLC-retention time techniques to characterize the thermodynamics of GLC retention and/or extract the vapor pressures:
(i) Prioritizing the high quality vapor pressure data for reference compounds (taken preferably from acknowledged original sources) along with reliable vapor pressure equations used for their temperature dependence and, if necessary, extrapolation.
(ii) Using a broad experimental temperature range with low end of the temperature scale as close to 298 K as possible should be given the priority in the experimental GLC setup to avoid the necessity of long-range temperature extrapolations.
(iii)
Consideration of the difference in heat capacity between the liquid and solid
states should be preferred in transforming solid state vapor pressures to
subcooled vapor pressure data ( and vice versa).
References
[1] Allen, J. O.; Sarofim, A. F.; Smith, K. A., Thermodynamic Properties of Polycyclic Aromatic Hydrocarbons in the Subcooled Liquid State. Polycyclic Aromatic Compounds 1999, 13, 261 - 283.
[2] van Noort, P. C. M., Semi-empirical estimation of organic compound fugacity ratios at environmentally relevant system temperatures. Chemosphere 2009, 76, 16-21.
[3] Koutek, B.; Cvačka, J.; Streinz, L.; Vrkočová, P.; Doubský, J.; Šimonová, H.; Feltl, L.; Svoboda, V., Comparison of methods employing gas chromatography retention data to determine vapour pressures at 298 K. J. Chromatogr. A 2001, 923, 137-152.
[4] Letcher, T. M.; Naicker, P. K., Determination of vapor pressures using gas chromatography. J. Chromatogr. A 2004, 1037, 107-114.
[5] Delle Site, A., The vapor pressure of environmentally significant organic chemicals: A review of methods and data at ambient temperature. Journal of Physical and Chemical Reference Data 1997, 26, 157-193.
[6] Růžička, K.; Koutek, B.; Fulem, M.; Hoskovec, M., Indirect Determination of Vapor Pressures by Capillary Gas-Liquid Chromatography: Analysis of the Reference Vapor-Pressure Data and Their Treatment. Journal of Chemical & Engineering Data 2012, 57, 1349-1368.
[7] Fischer-Lodike, C.; Zafar, A.; Chickos, J., The vapor pressure and vaporization enthalpy of pristane and phytane by correlation gas chromatography. J. Chem. Thermodyn. 2020, 141, 105931.
[8] Koutek, B.; Mahnel, T.; Šimáček, P.; Fulem, M.; Růžička, K., Extracting Vapor Pressure Data from GLC Retention Times. Part 1: Analysis of Single Reference Approach. Journal of Chemical & Engineering Data 2017, 62, 3542-3550.
[9] Koutek, B.; Fulem, M.; Mahnel, T.; Šimáček, P.; Růžička, K., Extracting Vapor Pressure Data from Gas-Liquid Chromatography Retention Times. Part 2: Analysis of Double Reference Approach. Journal of Chemical & Engineering Data 2018, 63, 4649-4661.
[10] Hamilton, D. J., Gas chromatographic measurement of volatility of herbicide esters. J. Chromatogr. A 1980, 195, 75-83.
[11] Koutek, B.; Pokorný, V.;Mahnel, T.; Štejfa V., Řehák, K., Fulem, M.; Růžička, K.,
Estimating Vapor Pressure Data from Gas–Liquid Chromatography Retention Times: Analysis of Multiple Reference Approaches, Review of Prior Applications, and Outlook
Journal of Chemical & Engineering Data 2022, 67(9) 2017–2043 https://doi.org/10.1021/acs.jced.2c00236
.