The effect of locating the limits of integration on the accuracy of measuring and calculating statistical moments for chromatographic data is discussed. Definitive criteria are presented for choosing the limits of peak start and peak end such that a fixed error level may be predetermined by correctly choosing the number of data points between the limits of integration. In this manner, the rate of analog-to-digital conversion may also be optimized. The effect of noise on the precision of the moment measurements is shown and correlated with the limits of integration

No abstract

No abstract

The low concentration of hydrocarbons anticipated in pollution baseline studies necessitates the development of analytical techniques sensitive at the sub-microgram per kilogram concentration level. The method of analysis developed in this laboratory involves dynamic headspace sampling for volatile hydrocarbon components of the sample, followed by coupled-column liquid chromatography for the non-volatile components. These techniques require minimal sample handling, reducing the risk of sample component loss and/or sample contamination. Volatile sample components are separated from the matrix in a closed system and concentrated on a TENAX-GC packed pre-column, free from large amounts of solvent and ready for GC/GC-MS analysis. Non-volatile compounds, such as the benzpyrenes, may be extracted from large volumes of water and concentrated on a Bondapak C18 packed pre-column for coupled-column liquid chromatographic separation and analysis. Results of the application of these techniques to the analysis of samples from sites of known low level hydrocarbon contamination are presented and discussed.

No abstract