Fractionation fluid inclusion oil

Usually, fluid inclusion oils are measured by GC-MS as whole oils (i.e., without prior fractionation), because extraction yields are usually very low (too low for gravimetric analysis). However, fluid inclusion oils can be fractionated into total hydrocarbon fractions and nitrogen-sulfur-oxygen (NSO) fractions to avoid unwanted interference of NSO compound peaks with aliphatic and aromatic hydrocarbon peaks. Furthermore, samples that have significant n-alkane interference, in particular in their biomarker distributions, can be further separated into n-alkanes and branched and cyclic hydrocarbon fractions by using molecular sieves (e.g., ZSM-5 sieve such as silicalite, a zeolitlic form of silica). 

The branched and cyclic hydrocarbon fraction is isolated from the fluid inclusion oil by trapping n-alkanes in the molecular sieve. The fractionation is conducted in a Pasteur pipette plugged with glass wool and dry-packed with 5 cm of silicalite (about 650 mg). To ensure adequate packing, the pipette is gently tapped while loading, since too rapid elution may result in incomplete removal of C15+ n-alkanes. The fluid inclusion oil sample is then introduced onto the column in a minimum amount of solvent (n-pentane), and is allowed to adsorb onto the molecular sieve before the addition of solvent, dropwise at first. The branched and cyclic hydrocarbon fraction is then slowly eluted with 4 mL of n-pentane. If required, the n-alkanes may be retrieved from the molecular sieves by hydrofluoric acid digestion. 

This technique is widely used in petroleum geochemistry to correlate crude oils and source rocks [72,73]. More recently, a wider range of isotopes (H, N, and O) and applications have been developed (for a current literature snapshot, see Reference [74] and other papers in that special issue). However, CSIA has had rather limited application for analysis of fluid inclusion oils. This is because at least 0.5 nmol C entering the mass spectrometer (as CO2) is required to get a reasonably precise isotopic measurement [70], and this is roughly equivalent to a small size peak on a GC-FID. As fluid inclusion oils are commonly not detectable by GC-FID, but only by more sensitive GC-MS, this means that many cannot be isotopically analyzed. However, some success with richer fluid inclusion oils has been reported [75,76]. If possible, it is better to fractionate fluid inclusion oils prior to CSIA so as to obtain a cleaner n-alkane fraction, as this enables better and cleaner resolution of these compounds [76]. Purified n-alkane fractions can be obtained by adduction onto silicalite packed in a Pasteur pipette, followed by digestion of the silicalite sieve in hydrofluoric acid [77,78]. 

FIGURE 30.10 Carbon isotopic compositions of /i-alkanes in the Jabiru fluid inclusion oil, production oil, and urea nonadducted fraction of the production oil (modified from Reference [75]). 

Second, it is known that fluid inclusion oils are sometimes enriched in polar compounds compared with crude oils from the same reservoir [46,141], although this is not always the case [142]. This is likely due to fractionation of the oil in the reservoir, with preferential adsorption of polar compounds onto mineral surfaces [143-146]. The possibility of carryover of polar compounds highly adsorbed onto grain surfaces into inclusion oils is limited by the extensive cleanup steps advocated and can be checked by outside rinse blanks (see Section 30.3.1) and/or by the use of surrogate extraction standards [50],... 

---