δ(18) O analysis of organic compounds: problems with pyrolysis in molybdenum-lined reactors

Effective δ(18) O determinations of organic compounds have been made using high-temperature pyrolysis (HTP) units that employ molybdenum (Mo) foil as an oxidation barrier instead of glassy carbon. In this investigation we evaluated the performance of a Mo-lined reactor during δ(18) O determinations from benzoic acid using a standard high-temperature conversion elemental analyser (TC/EA) unit. Our Mo-lined reactor was associated with poor performance statistics. It was hypothesised that this was as a consequence of the partitioning of oxygen within our system.

In order to test this hypothesis scanning electron microscopy energy-dispersive X-ray spectroscopy (SEM-EDX) and Raman spectroscopy was conducted on two Mo-lined reactors, which had been used for the analysis of benzoic acids.

The EDX analysis revealed that the Mo metal, which had become severely pitted and corroded in the region corresponding to the hottest part of the furnace, was associated with Mo-oxide patinas throughout its entire length.

A complex circular reaction between CO, Mo and C is likely to be responsible for the sequestering of a portion of sample oxygen into Mo-oxides, fractionating our sample. Accurate analysis of benzoic acid and other compounds can therefore only be achieved if sample oxygen is quantitatively returned to gaseous CO, preventing the buildup of Mo-oxides within the reactor over time. This is most likely to be achieved if sample gases have a sufficient residence time within the reactive glassy carbon bed, optimising the recovery of sample oxygen. However, such conditions are unlikely to have been met within our standard TC/EA system resulting in poor performance of the Mo-liner compared with other HTP units.