Noise in inductively coupled plasma mass spectrometry: some preliminary measurements

Abstract

Fundamental limiting noise sources in a non-commercial inductively coupled plasma mass spectrometer were investigated, using colour photographic evidence and signal processing which is manipulated via computer software. Noise sources were found to be 50 Hz from the mains power supply, peristaltic pump rotation, plasma audiofrequency peaks and their harmonics, and 1/f noise (which is mainly caused by the sample introduction system). These noise sources adversely affect the precision of inductively coupled plasma mass spectrometry (ICP-MS) for a given measurement bandwidth. To reduce their contributions in the noise power spectra (NPS) of the signal, the use of a bonnet device was investigated. This reduced the 1/f noise and peristaltic pump rotation frequency and completely removed the plasma audio frequency peak in the NPS. This latter frequency has been observed by a number of workers in ICP emission spectrometry and is associated with instability at the boundary of the plasma, where it enters the surrounding atmosphere. This audiofrequency peak has been removed in emission spectrometry by using an extended torch arrangement. In this present work the position of the plasma audiofrequency peak in the NPS changed with sampling distance, radiofrequency power and coolant gas flow rate and the peak was removed by the use of a torch extension bonnet.