After the cores were recovered, they were cut into sections at the catwalk, and the sediment samples within the plastic liners were deep frozen in liquid nitrogen immediately. Unfortunately, the cooling chain was interrupted during transport to Germany. For this reason the samples reached temperatures above zero for a few days. They were kept in a deep freezer (at ~-20ºC) again until the laboratory investigation commenced. For the analysis of the nonvolatile organic matter, the samples were freeze dried and aliquots were impregnated with a resin for organic petrographic investigations. After having polished the blocks, maceral composition and vitrinite reflectance were measured under the microscope using incident light of 546 nm and fluorescence illumination with blue light irradiation. Due to the scarcity and the small size of the huminite only 2-14 readings per sample were achieved.
The composition of the structured organic matter is determined by a semiquantitative estimation method that is based on a procedure applied successfully for vegetation surveys (Hufnagel and Porth, 1989). The contents of the main components (liptinite, bituminite, vitrinite, inertinite, and fecal pellets) are given in percent, and the composition of the liptinite (sporinite, alginite, and faunal liptinite) is similarly expressed. For organic-geochemical analysis, the dried samples were crushed in a disc mill. Total carbon and sulfur content were analyzed in the untreated sample with a LECO CS-444 analyzer. Prior to total organic carbon (TOC) analyses and the determination of 13C/12C ratios of kerogens, carbonate carbon was removed with hot 2N HCl (80ºC, 3 hr). Calibration was done with LECO steel rings and a homemade rock standard. The error associated with the TOC analysis is ±3% relative.
The crushed samples were extracted for 3.5 hr in a Soxtec device with acetone/hexane (50:50, v:v) as solvent, along with activated copper granules as a sulfur-removing agent. Due to the low amounts of soluble organic matter obtained, no further separation by quantitative liquid chromatography was done. After precipitation of the asphaltenes with n-pentane, the remainder was cleaned up by filtering it over aluminum oxide. Gas chromatographic separation of the nonaromatic fractions was done on a 30 m DB 5 fused silica column in a Hewlett Packard (HP) 5890 gas chromatograph connected to a HP 5972 mass selective detector (MSD). For the identification of normal- and cyclo-alkanes, mass spectra and retention times were used.
Extraction of hydrocarbon gases from the wet samples was done with a vacuum-degassing system (similar to that described by Faber and Stahl, 1983). About 150 g of sample material was heated to 100ºC in a closed bulb (~1 L), and phosphoric acid was added through a valve. The "combined" gases were extracted from the unsieved samples, consisting of "free" and "sorbed" hydrocarbon gases (e.g., Faber et al., 1990). Free gases are considered to consist predominantly of bacterial methane, thermal gases are mainly found in the sorbed phase, and combined gases are often a mixture of bacterial methane and thermal hydrocarbon gases (Faber et al., 1997, Whiticar and Suess, 1990).
Gas concentrations are
measured using a conventional gas chromatograph. Data is given in parts per
billion by weight (ppbw). Gas ratios (e.g., C1/[C2+C3])
are based on volumetric concentrations. For isotope analyses, methane (C1),
ethane (C2), and propane (C3) are separated on a GC column
and oxidized to CO2 and H2O. The CO2 is used
for stable carbon isotope ratio determination with an isotope-ratio mass
spectrometer (Finnigan MAT 251 IR-MS). The isotope ratios are given in the 
-notation
and are related to the Peedee belemnite (PDB) standard. The standard deviations
of the isotope values for the sediment gases are in the order of ± 1
for 13C1,
13C2,
and 13C3
(see Dumke et al., 1989). Minimum sample quantities are about 1 µL for 13C1,
13C2,
and 13C3.
Deuterium measurements could not be performed due to the insufficient gas
quantity.
