Gas flow patterns and longitudinal distribution of free selenium atoms in quartz tube atomizers for hydride generation atomic absorption spectrometry

Gas flow patterns in unheated quartz tube hydride atomizers were studied by light scattering on ammonium chloride particles and by observation of iodine vapors. The flow pattern is basically laminar, with a turbulent region at the T-junction and irregularities at the tube ends. We studied the free atom distribution over the length of the atomizers optical bar under various conditions by measuring the selenium signal with the tube perpendicular to the radiation beam, through the tube walls. In the whole range of atomization conditions used (Se 60–400 μg l−1, 0.8–2 l min−1 of H2 as purge gas, 0–50 ml min−1 O2,), there is a sharp maximum of free atom concentration in the tube centre, and all the free atoms disappear before they reach the tube ends. The absorption profiles are essentially the same in unheated as in flame-heated tubes, and there is no difference between heated flame-in-tube and externally heated arrangement. With decreasing purge gas flow rate, free atoms occupy a shorter part of the tube. Free atoms start to decay immediately after entering the optical tube of the atomizer. In both heated and unheated atomizers, enhanced free atom decay at higher concentrations contributes to the curvature of calibration graphs. As all free atoms disappear before reaching the ends of the tube even at relatively high purge gas flow rates, it is not possible to obtain higher sensitivities by extending the atomizer length. There is no difference in the free atom distribution in the central part of the tube if oxygen is prevented to enter the tube from the ends (by quartz windows or flames burning at the tube ends), so the free atom decay is not caused by O2 diffusing from the ambient atmosphere. With flames burning on tube ends, free Se atoms were reappearing close to the ends, proving the possibility of reatomization of species formed by free atom reactions. When no measures against O2 penetration from the tube ends of the heated atomizer are taken, signal reappearance takes place at the borders of the hot part of the atomizer, contributing substantially to the overall sensitivity. The results obtained correspond very well to the present picture of hydride atomization by H-radicals and their subsequent decay.