Experiments with tungsten furnaces

For certain applications, such as the determination of rare-earth elements, tungsten may be a superior substrate to graphite in terms of atomization efficiency. In addition, tungsten can be heated to higher temperatures than graphite, and since the mass of the tubes can be kept small it is easy to reach high heating rates [10]. Nevertheless, there are drawbacks resulting from possible chemical reactions of the tube material with the analyte and sample matrix. However, the ability of a furnace to accomodate both graphite and metal tubes enhances the versatility of the furnace technique, since more options are available to provide optimum conditions for any given analyte-matrix combination. Interchanging furnace materials is straightforward since, in the present configuration (see Fig. 1), the tubes are contacted by maintaining a vertical pressure at the terminal blocks over the cuvette contact areas, whether they be metal or graphite. For the thinner contact areas of the tungsten tubes adaptor blocks have to be inserted.

Figure 4 exhibits a signal trace for 4 ng of vanadium atomized from a tungsten platform. The shape of the signal shows that there is no interaction of vanadium with the tube surface, thus carry over contamination should be small. The possibility to determine involatile elements as well using the platform technique is advantageous with respect to gas phase interference effects [4], since higher gas phase temperatures can be achieved.

Fig. 4 Vanadium signal transient obtained with a tungsten platform equipped tungsten tube. The ashing temperature was 1000 oC. Upper gas flow was 10 % hydrogen in argon. Sample volume 2 µl. Fig. 5 Atomic emission signals for lead at two temperatures using a tungsten platform equipped tungsten furnace. The ashing temperature was 400 oC. Other conditions as given in Fig. 4.

The emission signals for lead shown in Fig. 5 confirm that the gas phase temperature in the platform equipped tungsten furnace is sufficient to increase the population of the excited states as a function of temperature in accordance with the Boltzmann distribution. To obtain the best atomic emission detection limits it is necessary to establish the temperature conferring the highest signal to noise ratio. However, it is evident that the signal to noise ratio is not improved at the higher temperatures. This furnace can therefore provide sufficiently high temperatures for optimum atomic emission detection limits, even for elements with relatively high (4.38 eV) excitation potentials. In addition, since the tungsten tube is spatially isothermal, the susceptibility to self-absorption effects should be less pronounced than in furnaces exhibiting temperature gradients [13]. The possibility to generate atomic emission signals without resorting to more complex probe [3] or two-step [7] atomizing systems offers interesting prospects which merit further investigation.


The full potential of the L’vov platform or indeed the probe technique cannot be fully utilized in end-heated Massmann-type furnaces due to the occurrence of temperature gradients.

Acknowledgements:- This work was supported by grants from the Swedish Natural Science Research Council. We thank B. Hütsch of Ringsdorff-Werke for supplying graphite parts, Prof. H. Ortner of Plansee Werke for the tungsten material, Dr. B. Welz of Bodenseewerk Perkin-Elmer GmbH for providing the AA-3030 and HGA-500 on loan and Dr. J. Moore of Varian Techtron Pty Ltd. for providing the SpectrAA-30 and GTA-96 on loan.

Previous page
Return to the Conference Presentations page, or the Atomic Spectrometry Group home page