Evaluation of space charge effects in the second vacuum stage of a commercial inductively coupled plasma mass spectrometer by planar laser-induced fluorescence imaging

https://doi.org/10.1016/j.sab.2012.06.028Get rights and content

Abstract

The effect of matrix on the formation and focusing of a Ca ion beam in the second vacuum stage of an inductively coupled plasma mass spectrometer has been evaluated with the use of planar laser induced fluorescence. A cross section of the beam was imaged near the entrance to the mass analyzer of a commercial instrument. Characteristics of the beam from a solution containing only the Ca analyte closely matched those predicted by simulation software. The individual addition of three matrix species, Mg, Cs, and Pb, had minor effect on beam shape. Cs and Pb both affected the beam trajectory. The most pronounced effect was with the Pb matrix, which caused an order-of-magnitude drop in the Ca signal intensity at the electron multiplier of the mass spectrometer. The loss in signal was due primarily to a shift in the direction and location of the Ca ion beam that caused it to miss the entrance into the mass analyzer.

Highlights

► We image a Ca ion beam in the second vacuum stage of an ICP-MS using PLIF. ► Ion beam shape and size is accurately predicted by SIMION. ► Added matrix species do not dramatically change beam shape. ► Heavy matrix species change beam direction and position.

Introduction

Following its commercial introduction in the early 1980s, inductively coupled plasma mass spectrometry (ICP-MS) was the subject of a number of studies by early adopters that focused on the new technique's analytical figures of merit. The reports from those studies consistently included discussions of “non-spectroscopic” matrix effects that were significant in magnitude and different in character than those observed for the more-familiar ICP-optical emission spectrometry [1], [2], [3], [4]. Tan and Horlick [2] noted in particular that heavy matrix elements caused more severe changes in the signal than light elements, and that the lightest analytes were most severely affected by the addition of matrix species. They speculated that high concentrations of matrix species were disrupting transport of the analyte ions through the mass spectrometer's ion optics.

Gillson et al. [5] and Tanner [6] provided plausible theoretical explanation for some of the observed matrix effects by calculating the influence of space charge effects on beam formation in typical ICP-MS ion optics. Consistent with experimental observations, the space charge calculations indicated that the heaviest matrix species would have the largest effect on the lightest analytes. The mass dependence of this particular form of matrix effect has important implications, not only for routine analysis by ICP-MS, but also for isotope ratio measurements, where variable mass biases can undermine precision measurements [7]. Tanner's calculations provided convincing qualitative evidence that space charge affects beam formation in ICP mass spectrometers. However, quantitative modeling of space charge effects would require a comprehensive model of both the ICP and the vacuum interface between the ion source and the mass spectrometer, and no such model exists.

Experimental characterization of the effects of space charge on ICP-MS beam formation has proven to be extremely challenging. Measurements of changes in signal levels in response to changing sample composition are inconclusive because they reflect not only changes in ion transport caused by space charge, but also matrix effects caused by the sample introduction system, the plasma itself, and the ion extraction process. Several different approaches have been taken to isolate the contribution of space charge to non-spectroscopic matrix effects from other contributions. Three groups, including our own, attempted to profile the ion beams in the second vacuum stages of the ICP-MS interface by placing targets in the ion beam, depositing ions for a period of time, and then performing post-deposition, spatially-resolved analysis on the targets [8], [9], [10], [11]. The beam profiles recorded in these experiments followed expected trends, but the experiments suffered from some important limitations. They were highly invasive, placing a solid target or mesh in the beam path. The offline analysis of the target was time consuming, and there were no real-time data from the experiments. In the experiments from our lab [11], at least, it was clear that beam chemistry was having a large effect on the size and shape of the deposits, and that the distributions of elements in the deposits did not necessarily represent the distribution of ions in the ion beam.

Two groups used time-resolved signals from streams of monodisperse droplets to study the effect of space charge on ion transport in the ICP-MS. Olesik and Dziewatkoski [12] examined the effect of a lead matrix on the shapes of analyte ion peaks, and Stewart and Olesik [13] expanded on the study to look at changes in the time-dependent signals from argon and other non-analyte species induced by the passage of the ion cloud from a single analyte particle. Allen et al. [14] split the ion beam between dual quadrupole mass analyzers and compared the shapes of peaks from two different ions contained within a single droplet. Both groups saw clear evidence that the space charge associated with clouds of ions from single droplets distorted the ion beam in the instrument, both along the beam's axis of propagation and perpendicular to it.

Tanner, et al. [15] and Praphairaksit and Houk [16], [17], [18] provided additional evidence of the effects of space charge on ICP-MS performance in two different approaches to reducing those effects. Tanner's group built an instrument with a three aperture interface that reduced plasma gas flow into the ion optics with an attendant reduction in the effects of space charge on ion beam formation. Praphairaksit and Houk placed a heated filament at the base of the skimmer cone to inject electrons into the ion beam and reduce the buildup of positive charge. Both approaches reduced matrix-induced mass biases in the research instruments, but neither approach has been adopted commercially.

In our research group we used laser-induced fluorescence (LIF) to record the spatial distribution of Ba ions and the effect of high concentrations of Mg and Pb on that distribution [19]. Mg narrowed the beam slightly, while Pb suppressed Ba+ densities across the beam. As was the case with the previous studies, our experiments demonstrated clearly that the character of the ion beam in the second vacuum stage of an ICP-MS changes with changes in matrix composition. However, because the instrument used in the study had no ion optics, our measurements did not provide a direct link between the changes that we observed and the performance of a working commercial instrument. The use of LIF to profile the ion beam offered some important advantages over the deposition experiments that had been performed earlier. LIF is highly selective and non-invasive. We were unable at the time, however, to devise an effective approach that would allow us to use LIF to profile the beam in the cramped second vacuum stages of existing commercial instruments.

The introduction by Varian of an instrument with a novel right-angle mirror geometry prompted us to revisit the LIF experiments in the second vacuum stage of an ICP-MS. As will be described in detail in the experimental section of this paper, the open geometry of the Varian instrument allowed us to train both excitation and emission optics on the entrance aperture to the third vacuum stage of the instrument with only minor modifications that had no effect on instrument performance. The resulting images of the ion beam cross section provide direct measures of the effect of space charge on ion beam formation and focusing in a working ICP-MS.

Section snippets

ICP-MS

All experiments were performed on a Varian model 820 ICP-MS. The Varian ICP-MS line was acquired by Bruker (Fremont, CA) in 2010, and the successor to our instrument is now marketed as the Bruker Aurora M90. The instrument has two nominal operating modes: normal sensitivity and high sensitivity. Instrument parameters for the two modes are presented in Table 1. The modification to the instrument is visible in Fig. 1. Normally, the turbo pump for the second vacuum stage is mounted directly to the

Results and discussion

Composite ion beam cross sections for the four test solutions and the two instrument operating modes are presented in Fig. 3. The corresponding MS signals at mass 43 are summarized in Table 2. The false color scales along the rows are the same, but differ between the rows by approximately an order of magnitude. In other words, the beam produced in the high sensitivity mode is more compact, and is brighter by more than a factor of ten than the normal-mode beam. The ratio between the mass

Conclusions

We have demonstrated that the intense ion beam in the second vacuum stage of an ICP-MS can be focused effectively and that the beam shape and size are very close to those predicted by a SIMION model, despite the fact that the model does not account for space charge effects. Space charge effects are manifest as lateral shifts in ion beam position, and the ion beam profiles suggest that the dramatic drop in signal caused by the addition of a Pb matrix is due primarily to the analyte ions being

Acknowledgment

The authors acknowledge the Chemical Sciences, Geosciences and Biosciences Division, Office of Basic Energy Sciences, Office of Science, U.S. Department of Energy (ICP-MS equipment, Alisa Edmund, and Paul Farnsworth) and the National Science Foundation, grant no. PHY-0969856 (laser equipment, Mary Lyon, and Scott Bergeson), for support of this research.

References (22)

  • R.S. Houk

    Mass spectrometry of inductively coupled plasmas

    Anal. Chem.

    (1986)
  • Cited by (8)

    • Ion sampling and transport in ICP-MS

      2017, Spectrochimica Acta - Part B Atomic Spectroscopy
      Citation Excerpt :

      Rather, the measurements from both groups were consistent with space-charge induced changes appearing at the point of charge separation, downstream from the skimmer tip and in the vicinity of an extraction lens. The Farnsworth lab has published a series of measurements of ion distributions downstream from the skimmer cone based on laser-induced fluorescence (LIF) [57–60]. The LIF measurements have the advantage of being less invasive and more selective than the deposition experiments, but they have been limited to a very narrow range of analyte species and ion optical configurations.

    • The effects of analyte mass and collision gases on ion beam formation in an inductively coupled plasma mass spectrometer

      2016, Spectrochimica Acta - Part B Atomic Spectroscopy
      Citation Excerpt :

      The final concentrations were 1.25 mM Ba with 1.5 mM Mg, Cs, or Pb respectively. The matrix concentrations are the same as were used in our published Ca study [6]. Gases used in the CRI study were 99.999% helium and 99.995% hydrogen (Airgas, Radnor, PA)

    • Fluorescence imaging of ion distributions in an inductively coupled plasma with laser ablation sample introduction

      2015, Spectrochimica Acta - Part B Atomic Spectroscopy
      Citation Excerpt :

      One such gap, an understanding of vaporization patterns in the ICP that are specific to LA generated aerosol, is the focal point of this work. Recent research in our lab, often utilizing laser-induced fluorescence (LIF), has been focused on characterizing ion transmission through the vacuum interface of solution nebulization (SN) ICP-MS instruments [2,3,6–10]. Because LIF is noninvasive it is a powerful technique to directly probe analyte species just upstream and downstream from the interface.

    View all citing articles on Scopus

    This paper is dedicated to Gary M. Hieftje, on the occasion of his 70th birthday, in recognition of his boundless contributions to spectroscopy and analytical chemistry.

    1

    Current address.

    View full text