Search

The spatial dependence of spectral and matrix interference effects known to affect ICP emission spectrometry has been studied. The emissions of atoms, ions, and molecules were resolved on both a vertical and radial basis. Interference and sample volatility effects were found to be limited to specific spatial regions. Molecular background emission was dominated by the entrainment of air and was limited primarily to the outer radial regions of the plasma. In contrast, the radial displacement of analyte molecular emission from more slowly atomized samples varied with height. In the case of the easily ionized element interference on Ca atomic/ionic emission, evidence was found indicating at least two different mechanisms of interaction. The ability to attain radial resolution was instrumental in assigning causes to interference effects.

The consequences of using N2 as a nebulizer gas during operation of a direct current plasma were evaluated with respect to its utility as a practical analytical procedure. Performance of the N2 system vs. that of an all-Ar plasma was compared on the basis of background plasma emission, atomization efficiency, detection limits, sensitivity, plasma electron density, and immunity from the easily ionized matrix element interference effect. Observations with regard to the mechanism of the EIE effect were also made. It was concluded in many cases that the use of N2 as a nebulizer gas does not significantly degrade analytical performance. An exception to this conclusion was the finding that refractory samples did not atomize as efficiently in the N2 plasma system.

The observation of molecular fluorescence caused by high-power pulsed laser excitation has been found to be an accurate technique for the determination of relative OH radical concentrations on a spatially resolved basis in a circular air-acetylene flame. It was found that the high power of radiational excitation used allowed fluorescence saturation to be approached. Theory then predicts fluorescence intensities to be independent of temperature and collisional quenching effects. However accurate estimates of relative concentrations can not be obtained unless corrections are made for incomplete saturation and postfilter effects. Procedures for conveniently applying such corrections are reported.

Spatially resolved flame temperatures are spectroscopically measured using the slope method, over small (≃0.7 μl), relatively homogenous volumes of flame gases. The spatial resolution is uniquely obtained by introducing Co as a thermometric species into isolated volumes via the use of a droplet injection technique. By this method the emission of light is restricted to a limited volume, whose position in the flame can be accurately determined and controlled. Vertical resolution is determined by the width of the entrance slit of the monochromator employed (100 μm in this study), and horizontal resolution is limited by the width of the emission cloud formed by the injected droplets (1 to 3 mm). The possibility of self-absorption effects are greatly reduced because of the short radiation path length involved. The performance of the method is illustrated by its application to the spatial temperature mapping of the secondary reaction zone of a cylindrical air-acetylene flame. The effects of N2 as a sheathing gas and flame stoichiometry on the radial and vertical temperature distribution of the flame are also investigated. It is determined that a large, virtually isothermal, central zone exists in the flame.

A direct reading emission spectrometer system is described which permits dynamic background correction measurements at each analytical wavelength. The system utilizes a refractor plate mounted on a tuning fork as a rapid means for square-wave shifting the spectral band pass from each analytical wavelength to the adjacent background wavelength position. A laboratory computer controls the system, receives the data, and performs the background subtraction. Evaluation of the system has shown that photon counting statistics are applicable and that reliable background corrections are obtained. Comparisons of this correction approach with others that have been used have further demonstrated its advantages for multielement analysis by emission spectrometry.