Pyrolysis-gas chromatography/mass spectrometry of a coal extract and its fractions separated by planar chromatography: correlation of structural features with molecular mass.

The structural characterisation of a coal liquefaction extract and its three fractions separated by planar chromatography has been described. Size exclusion chromatography showed the molecular mass distributions to become progressively larger with decreasing mobility on the plate. UV-fluorescence spectroscopy of the fractions indicated parallel increases in the sizes of polynuclear aromatic ring systems. Analysis by probe-mass spectrometry of the 'whole' coal extract showed the expected array of small polynuclear aromatic groups extending to m/z 450. The probe mass spectra of the lightest fraction ('mobile in pyridine and acetonitrile') showed similar features, except for effects due to vacuum drying to remove solvent. In sharp contrast, the two heaviest fractions ('mobile in pyridine and immobile in acetonitrile' and 'immobile in pyridine') showed no significant ions other than those from residual NMP solvent (m/z 98 and 99). Pyrolysis-gas chromatography/mass spectrometry of these two heaviest fractions showed only traces of aromatic compounds or fragments. The aromatic pyrolysis products of these fractions were too large and involatile to pass through the GC column. The major components observed in the pyrolysis-gas chromatography/mass spectrometry of the two heavy fractions were alkanes and alkenes, ranging between C10-C25. Since none of the samples contained free alkanes, alkenes or cycloalkanes before pyrolysis, they were generated during the pyrolysis step. The shifts of UV-fluorescence spectral intensity to shorter wavelengths with decreasing size indicated by size exclusion chromatography (SEC) provide direct evidence of differences in structure with changing molecular mass. This evidence strongly suggests that species identified as being of large molecular mass in this extract sample are not composed of molecular aggregates. It remains difficult to establish whether and when it would be legitimate to invoke molecular aggregates to explain the large molecular masses (MMs) identified here and in other work. Copyright 2000 John Wiley & Sons, Ltd.