The analysis of coal without sample dissolution was investigated by introducing coal slurries, into a variety of atom cells including electrothermal atomisation atomic absorption spectroscopy (ETAAAS), inductively coupled plasmas and direct current plasmas for atomic emission spectroscopy (ICP-AES, DCP-AES) and inductively coupled mass spectroscopy (ICP-MS). All the analyses were calibrated using aqueous standards. Slurries were injected into an electrothermal atomiser. The effects of furnace programme, background correction and air ashing were investigated. As, Se, Cd and Sb were successfully determined in a variety of certified reference material coals. For As the coal was slurried in nickel nitrate, magnesium nitrate, nitric acid and ethanol. Continuiun source and Smith-Hiefije background correction were compared for correction of a broadened Al line interference at the As (193.7 nm) line. Only the latter was effective. Smaller Se signals were obtained from coal compared to aqueous solutions. Iron coal produced structured background and hence overcorrection. The successful method introduced air into the ash stage. Both DCP and ICP techniques yielded good agreement with certificate values provided that the particle size was reduced to below 16 um and 10 um respectively. Simplex optimisation identified the critical parameters for aluminium determinations as being high injector flow rate and low observation height. Preliminary investigations of slurry atomisation using ICP-MS and 0.2% m/v slurried coal gave no blocking. Contamination from zirconia grinding elements used to comminute coal was investigated using laser ablation ICP-MS. Excitation temperature (Texc), ionisation temperature (Tion) rotational temperature (Trot) and electron number density (ng), were measured for different slurry concentrations (1-30% m/v) in the ICP. Depending on themometric species used, Texc may decrease with slurry concentration, but there were no similar decreases in Tion, T(rot) and ng. Observed decreases in analyte emission with increased sample loadings (> 10%) were shown to be caused by transport effects.

Document Type


Publication Date