Comparison of deuterium arc and Smith-Hieftje background correction for graphite furnace molecular absorption spectrometry of fluoride and chloride

Abstract

Molecular absorption spectrometry (MAS) in a graphite furnace, which is a sensitive method for the determination of halides, involves the formation of a diatomic molecule containing a metal reagent and a halide (the analyte). Other reagents, called chemical modifiers, are also added to increase the number of molecules formed. However, the utility of the technique for practical analysis has been limited by spectral and chemical interferences. Large amounts of reagents (μg amounts) produce relatively large background signals. Previous MAS work has employed deuterium arc background correction to account for these signals, but this method has been shown to be unable to accurately correct for large backgrounds. In this work, Smith-Hieftje background correction was employed for the first time with molecular absorption spectrometry for the determination of fluoride and chloride. For fluoride, a platinum hollow cathode lamp (HCL) was employed as the light source; aluminum was added as a reagent to produce AlF, the molecule whose absorbance was measured; and barium was added as a chemical modifier. The chemical and furnace conditions were optimized, but no AlF signal was obtained with Smith-Hieftje background correction. The absence of a signal was attributed to the involatility of platinum, which is poorly self-reversed during the Smith-Hieftje correction step, the use of a high level HCL line that was poorly self-reversed, and to the large width of the AlF molecular line (0.08 nm). Using deuterium arc background correction, a detection limit of 200 pg, a characteristic mass of 370 pg, and a linear dynamic range of 1.6 orders of magnitude were obtained. This technique was employed for the determination of fluoride in oyster tissue and dental rinse, but poor recovery (< 50%) was obtained. Chloride was determined with a lead HCL, the AlCl molecule, and cobalt and strontium as modifiers. Smith-Hieftje and continuum source background correction were compared for the determination of chloride; characteristic masses were 0.6 and 1.0 ng, respectively. The successful use of Smith-Hieftje background correction for chloride was attributed to the increased self-reversal of the lead HCL compared to the platinum HCL employed for fluoride. The addition of nitric acid to aqueous standards was shown to severely degrade the absorption signal. Smith-Hieftje background correction was shown to be more accurate than deuterium arc background correction for the determination of chloride in a simulated rain water standard reference material.