YOSHISUKE NAKAMURA, KATSUYUKI TAKAHASHI, OSAMU KUJIRAI and HARUNO OKOCHI
ETV using a tungsten boat furnace was studied in combination with an axially viewed (end-on) horizontal ICP for improvement in sensitivity. A high-dispersion echelle spectrometer in which wavelength modulation and second-derivative signal detection are incorporated was used. Time-resolved spectra were obtained by using an emission intensity–time–wavelength profiling method. In order to obtain a high S/B, the heating programme of the ETV system, the flow rate of the transport gas and the delay time are important in addition to the plasma conditions. Once the optimum conditions have been established, reproducible peak appearance times can be obtained. The net emission intensity, repeatability and limit of detection (LOD) obtained by ETV end-on ICP-AES were compared with those obtained by end-on pneumatic nebulization ICP-AES. The improved LODs were close to those obtained by pneumatic nebulization ICP-MS. Discrimination between the Fe impurity in the tungsten boat and spiked Fe was possible using time resolution. Trace amounts of Mn and Al in a large amount of Mo and Fe, respectively, were time-resolved at neighbouring wavelengths. The chemical species of the vaporized materials were identified by X-ray microanalysis. The main disadvantage of the method is that nitric and sulfuric acids cannot be used.