No abstract

A previously unrecognised source of noise in inductively coupled plasma (ICP) atomic emission spectrometry was identified with a combination of high-speed motion picture photography and noise spectrum analysis. The noise mechanism is a fluid mechanics phenomenon and involves axisymmetric oscillations of the plasma as the plasma gases flow from the torch into the surrounding static atmosphere. The oscillations develop into vortex rings with increasing height above the torch. As these plasma oscillations pass through the optical axis of the measurement system they produce periodic variations in the analytical signal, typically in the 100-600 Hz range. The frequencies of the oscillations observed in the films agreed with the frequencies of the -major noise peaks i n the noise power spectra. Knowledge of this noise phenomenon is relevant to studies of the fundamental properties of the ICP and its applications.

The noise power spectra of 85Rb+ signal and MNb+ signal from an Inductively coupled plasma mass spectrometer were measured at the same plasma conditions as were those of Sr II emission from the plasma Itself. Comparison of these spectra showed that discrete frequency noise In the emission at the mass spectrometer sampling orifice Is nearly Identical with that In the mass spectrometrlc signal and that white noise In the mass spectrometer signal was higher than that found In the emission signal. The dependence of noise frequencies on plasma operating conditions was generally the same for both measurements and was generally the same as that expected of emission from the plasma alone, l.e., when the plasma was not being sampled for mass spectrometry. However, discrete frequency noise In emission from the plasma alone differed substantially In frequency from that in the mass spectrometrlc signal. These results Indicate that the plasma Is the source of discrete frequency noise In the mass spectrometrlc signal and that the discrete noise frequencies can be affected by changes In plasma gas dynamics due to interaction between the plasma and the mass spectrometer sampling Interface. The major source of signal Instability In this particular Inductively coupled plasma mass spectrometer was found to be 1H noise.

No abstract