Ion Composition Elucidation (ICE)
Instrumental Techniques: Mass Spectral Determination
Andrew H. Grange and William C. Brumley
Characterization Research Divison,
NERL, U.S. EPA,
PO Box 93478, Las Vegas, NV 89193-3478
Mass Peak Profiling from SIR Data
To plot the mass peak profile of an ion, mass-to-charge (m/z) ratios are monitored across the profile. In Figure 1, the areas of the chromatographic peaks in the ion chromatograms for each m/z ratio, which were normalized to the largest signal in Figure 1b-1f, were plotted to provide the profile in Figure 1a.
Figure 1. A mass peak profile (a) and the ion chromatograms from which it was constructed (b-f).
Plotting peak areas rather than peak heights improves the signal-to-noise ratio through signal averaging. The weighted average of several points across the top of the profile provides the exact mass. Preparation of SIR descriptors and processing of the data are done automatically by procedures written in RSX (the computer's operating system) and VG code on the data system and in Lotus 1-2-3 (2.2), Reflection 4.2 and WordPerfect 5.1 on an ancillary personal computer. By comparison with conventional full-scan modes, mass peak profiling from selected-ion recording data (MPPSIRD) provides important advantages for determining exact masses: speed, sensitivity and higher mass resolution. Using full scanning and perfluorokerosene as the calibrant at 10,000 resolution, a time of 2.7 seconds is required to record a mass spectrum, including two calibration ions. With MPPSIRD, only 0.8 second is required to monitor 17 points across the mass range, including the full profile of the target ion and 5 points across the top portion of a single calibration ion. This threefold advantage in scan speed permits the chromatographic peaks of partially coeluting components from complex environmental extracts to be better delineated. It also provides better deconvolution of the ions from each component. At 10,000 resolution, MPPSIRD provided a detection limit of 6 femtograms (fg) for a tetrachlorobiphenyl.(1) This was lower by a factor of 170 than that obtained with full scanning. Running routinely at 20,000 resolution in order to exclude more mass interferences sacrifices a factor of 10 in sensitivity, but still leaves a 17-fold advantage in sensitivity.
Precision and accuracy are comparable for the two scan techniques. With MPPSIRD, average exact masses from triplicate determinations are correct to within 2.5 ppm. Because the number of possible elemental compositions for an exact mass increases rapidly with the mass of the ion and the error limits of the measurement, high accuracy greatly reduces the number of compositions possible for an exact mass.
Applications: Confirmation and Rejection
In our research, MPPSIRD is most frequently used to determine compositions for ions from environmental extracts.(1) Library matches can be confirmed or rejected by determining the composition of the molecular ion or of a large fragment ion if no molecular ion is present in the mass spectrum. Part of a total-ion chromatogram for a dark-colored extract that contained a high level of polychlorinated biphenyls (PCBs) is shown in Figure 2.
Figure 2. Part of a total-ion chromatogram for an extract of a Superfund site sample. The shaded peak was investigated (see Figure 3).
To illustrate the importance of determining an exact mass, the small, shaded peak due to a trace component was investigated. For trace components that partially coelute with larger amounts of other compounds, a background-subtracted mass spectrum of poor quality often is obtained - one that can contain additional masses or can lack some that are normally present. Commonly, several poor library matches for different compositions are found, as is the case in Figure 3, where the background-subtracted mass spectrum for the shaded peak in Figure 2 (a) and four library matches featuring molecular ions at m/z 266 are displayed.
Figure 3. The background-subtracted mass spectrum for the shaded chromatographic peak in Figure 2 (a) and four library matches (b-e).
Plotted in Figure 4 are the four mass peak profiles corresponding to the molecular ions that were calculated at a mass resolution of 20,000 using an in-house model based on the Gaussian distribution.
Figure 4. Four mass peak profiles calculated for 20,000 resolution and a mass peak profile constructed from SIR data.
The mass scale in the left inset is magnified for the two ions having the most similar masses. The boxes on top of the two profiles correspond to error ranges of + to -2.5 ppm. Because the ranges do not overlap at 20,000 mass resolution, any of these four ions could be distinguished from the others. Shown in the right inset is the experimental mass peak profile for the target ion acquired at 22,500 resolution. When the profile was plotted on the mass scale of the four candidate ions, it was clear that all four library matches were incorrect. Indeed, the only composition that corresponded to the exact mass obtained within the error limit of the measurement and that did not violate the nitrogen rule was C20H26 which could be an alkylated biphenyl. MPPSIRD is a technique for proving that no compounds having targeted compositions are present in complex mixtures down to very low detection limits.
MPPSIRD also is a quality assurance tool. Before using SIR to monitor the apex of a mass peak profile in quantitative analyses, it is important to document that no interferences can contribute signal at the m/z ratio monitored for each analyte. Figure 5 shows mass peak profiles for the molecular ion of acridine on the left and for molecular ions of phenanthrene/anthracene, containing a 13C atom, on the right.
Figure 5. Mass peak profiles for the molecular ion of acridine (left) and the molecular ion of phenanthrene/anthracene containing a 13C atom.
These compounds were in an extract from a creosote-containing soil, which was analyzed with a mass resolution of 22,000. If analysis for acridine were performed at lower resolution, large amounts of coeluting phenanthrene or anthracene would interfere and provide inflated concentrations for acridine. In addition, any effect on resolution caused by coelution of large amounts of other compounds can be documented. Measuring the width of a mass peak profile at 5% of its maximum provides real-time verification of the mass resolution at the time an analyte entered the ion source.
The exact mass determined in Figure 4 provided a single composition for an unidentified ion that weighed less than 300 amu (atomic mass units, or Daltons). For heavier ions, however, multiple compositions often are possible within the error limit of the mass determination. In Table 1, 12 possible compositions based on the exact mass of the molecular ion are listed for an unidentified compound detected in a sample extract from a Superfund site.
Table 1. Twelve possible compositions for an unknown from a Superfund site.
More criteria were required to select the correct composition from this list. The relative abundances of M+1 and M+2 ions in low-resolution mass spectra free of interferences long have been used to help determine the elemental compositions of molecular ions.(2) High resolution removes interferences with the M+1 and M+2 ions from the target compound, and SIR provides ample sensitivity to observe the less abundant M+1 and M+2 ions. At high resolution, five criteria are available for differentiating between the possible compositions. In addition to the abundances of the M+1 and M+2 profiles relative to the molecular profile, the exact masses of the M+1 and M+2 profiles and the shape of the M+2 profile are useful. M+2 profiles were calculated and plotted for the 12 compositions at a resolution of 20,400. The profile shapes for the last three compositions were unique and easily distinguishable from the others. The experimental and calculated M+2 profiles are shown in Figure 6.
Figure 6. The M+2 profile from SIR data for an unidentified compound from a Superfund site acquired with 20,400 resolution (a) compared with the calculated M+2 profile (b).
Their similarity identified the last composition as correct. The C28H41N34S+ ion and 13C2C26H41NS+ ion are responsible for the first and second maxima in the profile, respectively.
When exact masses and relative abundances of the M+1 and M+2 profiles are needed to reject one or more compositions, a SIR descriptor that monitors the top 60% of the M, M+1, and M+2 profiles is used.(3) Plotting of partial profiles permits simultaneous monitoring of multiple ions, despite the limited number of m/z ratios permitted by the instrument's software. The model predicts the masses and relative abundances on the basis of partial profiles. These additional criteria have been used to identify correctly the compositions of four standards with molecular weights between 536 and 766 amu. For the largest ion, 15 determinations, which provide a mass error limit of + to -1.0 ppm for the profiles of the M, M+1, and M+2 ions, were required to reject the composition most similar to the correct one. If more than one composition for a high-mass ion remained, one could similarly identify fragment (F) ions, resorting to study of F+1 and F+2 ions, when multiple compositions for F were possible. Exact masses of neutral-loss fragments are determined by difference. The molecular ion would be identified as the sum of its parts.
Grange, A.H., J.R. Donnelly, W.C. Brurmley, S. Billets and G.W. Sovocool. "Mass measurements by an accurate and sensitive selected ion recording technique" Analytical Chemistry Vol. 66,; pp. 4416-4421, 1994.
McLafferty, F.W. Interpretation of Mass Spectra 3rd ed. University Science Books. 1980.
Grange, A.H., J.R. Donnelly, W.C. Brumley and G. W. Sovocool "Determination of an elemental composition from mass peak profiles of the molecular ion (M) and the M+1 and M+2 ions" Analytical Chemistry Vol. 68, pp. 553-560, 1996.
Moseley, M., L. Deterding, K. Tomer and J .Jorgensen "Nanoscale packed-capillary liquid chromatography coupled with mass spectrometry using a coaxial continuous-flow fast atom bombardment interface" Analytical Chemistry Vol. 63, p. 1467, 1991.
Note: The U.S. EPA, through its Office of Research and Development, partially funded and collaborated in the research described here. The article has been subjected to the agency's peer review and has been approved as an EPA publication. The instrument used for this work was a VG70-250SE double-focusing mass spectrometer.