Ten shipboard samples (four from Site 1257 and six from Site 1258) were chosen for detailed investigation of their biomarker compositions, partly to cover the stratigraphic range of the black shales but primarily on the basis of their TOC contents and the results of Rock-Eval analysis. Because the extractions were performed on samples used for shipboard carbonate analysis, the size of the rock samples was limited.
For the series of extractions on samples from Site 1257, only 1 g of rock powder was used, which sometimes yielded sample concentrations that were too low for optimal gas chromatography-mass selective detector (GC-MSD) analysis (see "Gas Chromatography-Mass Selective Detector Analysis"). More material was extracted for the batch of samples from Site 1258, giving better results on the GC-MSD but creating apparent differences in sample quality that affect comparison of results from the two sites. Quantification of compounds could not be done because suitable internal standards were not available on the ship. Thus, samples are evaluated in qualitative terms by employing biomarker ratios.
All samples were extracted ultrasonically; freeze-dried, powdered sediment (1–6 g) was placed in a 100-mL centrifuge tube with ~8 mL of dichloromethane (DCM). The samples were sonicated for 30 min and allowed to settle. The clear supernatant layer was transferred into a clean vial. The residue was extracted a second time, and both extracts were combined. The extract was filtered through a small Pasteur pipette plugged with precleaned cotton wool into a clean preweighed vial. A small piece of clean aluminum foil was placed under the screw lid to prevent sample contamination. The solvent was evaporated under nitrogen, and the weight of the total lipid extract was measured.
The total lipid extracts were treated with acid-activated copper to remove elemental sulfur, which could interfere with GC-MSD analysis. Copper wire normally employed in the Rock-Eval instrument was placed in an Erlenmeyer flask, and a small amount of concentrated (37%) hydrochloric acid was added. When the copper reached its bright typical color, the acid was decanted and the copper was rinsed with nanopure water until pH neutral, then with methanol (seven times), and finally with DCM (seven times).
Small amounts of activated and cleaned copper were added to the vials containing the extracts in DCM, and the samples were magnetically stirred overnight (~8 hr). Blackening of the copper indicated reaction with elemental sulfur present in the sample. Each extract dissolved in DCM was then filtered over a cotton wool–plugged Pasteur pipette to remove the copper and was subsequently evaporated to dryness under nitrogen.
Silica gel (Fisher 100–200 mesh) was activated by heating at 120°C overnight, cooled in a dessicator, and made into a slurry with hexane. Pasteur pipettes were plugged with cotton wool and filled with the slurry. Extracts were dissolved in hexane and added to the top of the column. Fractions of increasing polarity were collected by sequential elution with 4 mL each of hexane (aliphatic fraction), DCM/hexane (1:1) (aromatic fraction), and DCM (polar fraction). The eluates were collected in clean preweighed vials, the solvent was removed under nitrogen, and the mass of each extract was recorded as precisely as possible under shipboard conditions. No internal standard was added.
The aliphatic, aromatic, and polar fractions were dissolved in hexane and transferred to autosampler vials for analysis on the GC-MSD, a Hewlett-Packard (HP) system consisting of an HP 6890 GC with an HP 5973 MSD and an HP 7683 automatic liquid sampler (ALS). The GC was equipped with an electronic program controlled (EPC) split-splitless injector and an HP capillary column (5% phenyl methyl siloxane; 30 m x 0.25 µm) programmed from 70° to 130°C at 20°C/min, then at 4°C/min to 320°C, and held at 320°C for 20 min. Helium was used as the carrier gas. The transfer line was set at 280°C, and the MSD source was set at 230°C. A heated interface connects the capillary column from the GC to the mass spectrometer, where electron impact (EI)-mass spectra were acquired at 70 eV in full scan mode (scan range from m/z = 27–800). An electron multiplier system served as detector. Data were acquired and processed by HP MS-ChemStation software. Compounds were identified by their relative retention times and by comparison with mass spectra reported in the literature.