The California Air Resources Board (CARB) has proposed procedures for the analysis of non-methane organic gases (NMOG) to determine the ozone forming potential (OFP) of automotive exhaust. For realization of these methods two differently configured GC systems are necessary. In order to reduce the efforts concerning costs, maintenance and quality control of two analytical instruments, a single run method is developed for routine analysis. This method allows identification and quantification of individual hydrocarbons (IHC) in the range of carbon numbers C2 to C12.
Analytical problems arising from high contents of water and carbon dioxide in exhaust samples are discussed. Water reduction is obtained by a Nafion® Dryer by means of membrane diffusion of polar compounds. Contamination as well as memory effects due to this sample work up are described. Sample pre-concentration of 50-200 mL diluted automotive exhaust is performed using a triple phase “mixed bed” adsorption tube at O°C. The enriched IHC are thermally desorbed, cryogenicly focused on a glass beads filled cryo trap and subsequently injected into the GC system. To avoid plugging of the cryo trap by freezing CO2 and remaining H2O the adsorption tube is pre-flushed with helium at slightly elevated temperatures prior to the actual thermal desorption. Adequate separation efficiency over the wide range of C2-C12 in a single run requires the application of two separation columns combined via the “Deans type” column switching. Light-end species (C2-C5) are separated on a Al2O3/KCl PLOT column, whereas mid-range species (C6-C12) are eluted simultaneously from a CP-Sil-5 CB column. The moment of column switching is adjustable according to the analytical problem. During method development the identification of the individual components is performed by means of single gas standards (C ≤ 4), GC/MSD and / or comparison with chromatograms of GC standard reference mixtures (C ≥ 5) respectively. A chromatogram of an automotive exhaust sample as well as limits of detection and blank contributions of the analytical system are presented. Scope and limitations of the method are described. In spite of an already semi automatic application delivering results of high analytical quality, some further work has to be done on additional identification of compounds, optimization of adsorbent mixture, and solving the problem of too high ethyne quantification.