The enigmatic mechanism of the flame ionization detector: Its overlooked implications for fossil fuel combustion modeling

The flame ionization detector (FID) has been a commercial analyzer now for about 50 years. It still finds significant use as a sensitive quantitative monitor of organic compounds in gas chromatography and for monitoring mixtures of hydrocarbons. Its carbon counting ability to integrate, for example, total unburned hydrocarbon emissions from a source, now is accepted without question. This is especially noteworthy as the fundamental chemistry on which the instrument is based has always been uncertain. Although now largely overlooked, its mechanism has significant implications and suggests that there is an underlying simplicity to hydrocarbon combustion. As a result, in the light of recent discoveries concerning the very rapid formation of a pool of hydrocarbon radicals in hydrocarbon combustion, a re-examination of the chemi-ionization mechanisms in hydrocarbon flames has been undertaken. Many of the previous speculations have been scrutinized and it is confirmed that the primary chemi-ionizing reaction of CH(X-2 Pi) with O atoms is most likely the sole source in combustion including the FID. The oft suggested roles of electronically excited states of CH now are ruled out but with some slight uncertainty remaining on the still unknown importance of the metastable CH(a(4)Sigma(-)) state in flames. The reason for the equal per carbon response of the FID with any hydrocarbon finally has been resolved. From isotopically labeled studies and measurements of the concentrations of CH and C-2 it is seen, under the same conditions, that different hydrocarbons do produce approximately the same levels of CH on a unit carbon basis. This results from the very rapid destruction and reformulation kinetics in the reaction zone of flames, and formation of a hydrocarbon radical pool that constitutes the unburned carbon. These radicals then are gradually eroded by the continuing oxidation or by soot precursor growth. As a result, the nature of the carbon in a hydrocarbon fuel is mainly irrelevant, only its quantity. The one well-documented exception has always been C2H2 but the data now show this so-called anomalous behavior to be no more than a reflection of its uniquely slower combustion nature in the reaction zone. It is not apparent in substituted acetylene fuels. Close to the reaction zone its kinetics produce a larger profile of unburned carbon that is evidenced by enhanced levels observed for CH and C-2. The nature of the specific responses of the FID to other organic structural categories also is a reflection of their primary combustion breakdown and a measure of the initial pool of unburned carbon. Exactly similar responses are seen in both the FID and in soot formation tendencies. The connection though is indirect in that both processes relate to and result from the same pool of non-oxidized carbon, rather than any implied inceptive role. As a result, the observed sensitivities previously recorded with the FID now can be a useful aid in validating the primary dominant steps in combustion mechanisms and the example of dimethyl ether combustion is used as an illustration. At present, this rich analytical database could be particularly useful in modeling the more complex partially oxygenated fuels that now are being extensively studied. (C) 2007 Elsevier Ltd. All rights reserved.