Wired Chemist

Gas Chromatography/Mass Spectroscopy

Chromatography is the separation of a multi-component mixture by distribution between two phases, one stationary and one moving. In GC, the stationary phase is a liquid and the moving phase is a gas (the carrier gas). The one limitation on use of GC is that the liquid to be analyzed must be sufficiently volatile that, when it is injected into the heated GC (up to temperatures of 300°C), the sample will vaporize so that it can mix with the carrier gas. As the sample moves into the column containing the stationary liquid phase, the components of the samples interact with the stationary phase to different extents. For the nonpolar components found in oil, the separation is controlled by the van der Waal (London) forces that influence the boiling point order. Hence, the GC elution order parallels the fractional distillation process (Tables 2,3).

In GC/MS as each individual component of the oil exits the separation column, it is ionized by the mass spectrometer. The MS detector converts the intensity of the charge to the relative concentration of each component, the pattern of which confirms the sample origin, thus providing a "hydrocarbon fingerprint." The identity of an unknown can be determined by matching the sample's mass spectrum with that of a known compound via computer-assisted mass spectral library searches and probability-based matches.¹⁰

High-Performance Liquid Chromatography (HPLC)

Compounds that are either thermally labile or nonvolatile are more amenable to analysis by non-thermal methods. In this case, HPLC is the method of choice. As in GC, the sample to be analyzed (analyte) is partitioned between a stationary and a moving (mobile) phase, which travels through a column. Whereas the mobile phase in GC is a gas, in HPLC it is a liquid. Because most HPLC instruments are equipped with UV detectors, HPLC is a sensitive method for the analysis of aromatic hydrocarbons, which absorb UV light (vide supra).

Components of oil induce mixed-function oxidase (MFO) activity in vertebrate animals. Bile metabolites are the products of MFO activity in the liver. HPLC analyses of fish bile have been conducted to detect PAH compounds. Using fluorescent detection, the light is tuned to one or two wavelengths characteristic of PAHs. The strength of the signal emitted correlates roughly with the general level of bile exposure.

If bile metabolites are detected, then analyses can be performed for MFO enzymes. John Stegeman, a biochemist at Woods Hole Oceanographic Institution, has developed a MFO assay based on cloned antibodies to particular MFO enzymes. His test for MFO was used extensively during the spill. Results of his tests revealed that MFO induction was higher in oiled prickleback fish samples than in the unoiled controls. However, the significance of this induction is not easily interpreted.

After the Exxon Valdez accident, sediment samples were collected from ten sites in PWS to determine the degree of oiling. Sixty sediments were analyzed for Prudhoe Bay crude oil (PBCO), the most abundant oil on the North Slope, using a rapid HPLC screening method with fluorescence detection. The predominant aromatic compounds in PBCO are one- to three-ring alkylated ACs. However, as crude oil weathers or degrades with time, the dominant components in the aromatic fraction are the naphthalenes, phenanthrenes, and dibenzothiophenes. Accordingly, the two fluorescence wavelengths selected for analysis of the post-spill sediment samples were the composite for naphthalenes/dibenzothiophenes (290/335 nm) and that for phenanthrenes (260/380 nm). The HPLC assay was preferentially chosen because GC/MS analyses are expensive and time-consuming. However, selected GC/MS analyses were performed for comparison with the HPLC screening method. Intertidal sediments from four heavily oiled sites in PWS that had been treated with high-pressure, hot-water washes exhibited HPLC patterns similar to that of weathered PBCO. In contrast, sediments from three unoiled sites in PWS produced HPLC chromatographic patterns that more closely approximate that of diesel fuel (Figure 6). GC/MS analyses confirmed that PBCO was a primary source of contamination in many of the sediments collected in PWS. GC/MS results from the sediment samples contaminated with diesel fuel, a common fuel for fishing and pleasure boats, are consistent only with a diesel fuel refined from a crude oil low in dibenzothiophenes, such as Cook Inlet crude.¹⁷