Sources of mass bias and isotope ratio variation in multi-collector ICP-MS: optimization of instrumental parameters based on experimental observations

In this work, several contributing factors to the observed mass bias in inductively coupled plasma mass spectrometry (ICP-MS) have been identified. Analyses of the isotopic compositions of B deposited on sampler and skimmer cones demonstrate enrichment of 10B on the former and 11B on the latter. Grounding the capacitive discharge system to enhance sensitivity also magnified the level of 11B enrichment on the skimmer cone more than four-fold. This supersonic expansion of the ion beam behind the sampler is confirmed to be an important source of mass bias. Isotopic analyses of the Fe, Zn and Tl leached from used extraction lenses yielded a linear relationship between the levels of lighter isotope depletion and mass ratio. Although consistent with the space-charge effect, the fact that isotopically-heavy deposits were found demonstrates that the ion beam diverges into a relatively wide solid angle in the field-free region behind the skimmer. This severely impairs transmission of, in particular, the lighter isotopes. For a wide range of elements (Li, B, Fe, Ni, Cu, Sb, Ce, Hf and Re), the magnitude of the mass bias was found to be affected by the sample gas flow rate, as well as the distance between the sampler and the end of the torch, i.e., the sampling depth, employed in the Neptune multi-collector ICP-MS instrument. Mathematical analysis of the profiles of intensity variations as a function of these instrumental parameters revealed that the response peaks closer to the torch for the heavier isotopes of all studied elements. Owing to this spatial non-coincidence, tuning for maximum intensity on either isotope will result in sampling from a region where even slight plasma instabilities will be translated into substantial variations in mass bias. Therefore, in-plasma processes also contribute to the degree and temporal stability of mass bias. In light of these findings, recommendations for optimizing multi-collector ICP-MS with respect to obtaining the highest possible precision are presented.