Wolfgang Frech and Douglas C. Baxter

Department of Analytical Chemistry, Umeň University, S-901 87 Umeň, Sweden


One limitation in the use of Massmann-type graphite furnaces arises from the fact that, during atomization, temperature gradients develop along the tube. Falk et al. [1] have shown that the evolution of these gradients is a function of heating rate, starting temperature and time. For ramp rates above 1000 oC/s the axial temperature distribution along the central part (▒ 10 mm) of a Perkin-Elmer 2200 tube is less than 200 oC during ramping up to 2000 oC provided that heating commences from ambient temperature. However, when the tube centre reaches the final temperature heat losses due to conduction give rise to gradients of more than 400 oC over the central region (▒ 10 mm) and even lower temperatures at the tube ends. This temperature distribution may result in:

  • Losses of analyte by condensation at cooler tube parts

  • Gas phase interference effects due to low vapour temperatures

  • Memory effects since it is difficult to remove analyte from cooler parts

  • Problems in vaporizing analyte which has spread to the tube ends

    The extent to which developing temperature gradients affect the atomization efficiency should depend on the appearance time of the analyte. Lesser effects are expected for volatile elements atomized from the tube wall since vaporization occurs prior to the development of severe gradients. On the other hand, if vaporization is delayed due to the volatility of the element or by using platform [2] or probe [3] techniques, the effects of spatial non-isothermality will be more pronounced. Nevertheless, the analyte is then volatilized at a higher tube temperature which normally results in reduced interference effects [4].

    Problems due to spatial non-isothermality have been addressed by heating tubes from the side through one pair of electrodes [5]. However, with this configuration, only relatively short tubes can be heated isothermally. Longer tubes, side-heated by two pairs of contact electrodes, have also been used by several workers [6, 7] but such systems are difficult to work with, particularly at high temperatures, since even minute misalignments between tube and contact electrodes lead to overheating of the tube/electrode interface. This results in reduced tube lifetime. To avoid this problem side-heated cuvettes with tube and contacts in one piece (ICC) have been constructed [8]. Here we demonstrate some of the advantages with spatially isothermal tubes for atomic spectrometry. The results obtained are compared with conventional Massmann-type furnaces. In addition, the possibilities of interchanging the ICC with tungsten tubes as described by Sychra et al. [9] and of making atomic emission measurements are outlined.

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