Inductively coupled plasma?mass spectrometry (ICP-MS) is one of the primary techniques for trace elemental analysis. This Technology Forum discusses the effects of carbon in samples and the use of correction equations to account for interferences. Participants in this Forum include Deborah Bradshaw, consultant and Spectroscopy Editorial Advisory Board member; Shane Elliott, ICP-MS Products Global Business Manager for Bruker CAM; and Charles A Schneider, Inorganic Product Planning Manager at PerkinElmer.
Inductively coupled plasma–mass spectrometry (ICP-MS) is one of the primary techniques for trace elemental analysis. This Technology Forum discusses the effects of carbon in samples and the use of correction equations to account for interferences. Participants in this Forum include Deborah Bradshaw, consultant and Spectroscopy Editorial Advisory Board member; Shane Elliott, ICP-MS Products Global Business Manager for Bruker CAM; and Charles A Schneider, Inorganic Product Planning Manager at PerkinElmer.
While we know the effects that high carbon has in a sample in terms of spectral interferences (for example, a high result for 52Cr due to 40Ar12C), have you seen other effects on the intensity data when carbon is at a high concentration in the samples?
Bradshaw: In my limited experience with high carbon matrices using ICP-MS, I have noticed an enhancement in the intensities of most analytes. This would be expected if the carbon were from an organic solvent, but I see the effect whatever the source of the carbon.
Elliott: Carbon enhancement is a well known and widely reported phenomenon. Samples with high levels of carbon can show sensitivity enhancement for difficult-to-ionize elements (high IP elements), and in fact some users use this effect to improve sensitivity on some of these elements by spiking carbon into their solutions (or even adding carbon to the carrier or plasma gases).
Schneider: Yes. For elements with high ionization potentials (such as Zn, As, Se, and Hg), the presence of carbon in a sample increases the ionization efficiency, leading to higher intensities. This will cause problems for samples containing significantly more carbon than the calibration blank and standards, as the resulting apparent concentrations will be higher than the true values.
To correct for the potential variability of carbon content in samples, an excess amount of clean carbon is typically added to each sample to force a carbon leveling effect. Typically, ultrapure methanol or isopropanol is added to produce a carbon concentration of 0.5–1% (v/v) for calibration and sample solutions. This will ensure that the ionization efficiency of the high ionization potential elements doesn’t change throughout the analysis. The source of clean carbon must be carefully chosen. Semiconductor grade reagents will have the lowest trace metal contaminants. Other grades of alcohols can contain significant levels of trace elements and contaminate the solutions.
For elements like Cr and V, what is your experience with using correction equations (that is, those supplied by EPA methods) as compared to using reaction cell and collision cell correction?
Bradshaw: I have found that in many instances, the correction equations work very well. However, if there is any interference on one of the masses used in the equation, then there is a problem. For example, the correction equation for the determination of 75As (to correct for Cl, forming 40Ar35Cl) uses mass 82. If there is Br in the sample, then 81Br1H produces an affect on mass 82, affecting the result for As.
Elliott: Correction equations are a viable option for correcting for known interferences, even when no interference management (CRI, CC, DRC, and so forth) is used — but they are limited in their application. If the correction required is significant, compounding errors can lead to poor precision and inaccurate recoveries. In the case of Cr and V, if residual interferences remain when using an interference management technology, a suitable correction equation is a viable option for correcting for small residuals.
Schneider: Correction equations work well as long as the interferent concentrations are within a finite range. Once the concentration of carbon or chloride exceeds a few hundred ppm, the concentration accuracy from the correction equation begins to decline. Cell-based ICP-MS instruments work to improve the signal-to-noise of Cr and V by several orders of magnitude. They reduce the BECs caused by interferences through chemical reactions or energy filtering. As a result, cell-based instruments can produce accurate results for V and Cr in much higher concentrations of carbon or chloride.