Page 120 - Inorganic Mass Spectrometry : Fundamentals and Applications

110 Olesik

have been obtained for elements that suffer from extensive molecular ion spectral
overlaps [ 1531.
The multico~ponent spectral fitting approach has been the most successful
for a broad range of sample types [ 153,1541. A linear combination of spectra of all
of the individual components (elemental ions and polyatomic ions) is optimized
for best fit of the experimental spectrum. “Model” spectra are determined from
the natural abundances of the elements and polyatomic “model” spectra are
of
calculated from the isotopic abundances the individual element isotopes. There-
fore, experimental “model” spectra are not required.
Several problems can be encountered with multicomponent spectral fitting.
If the signal is too large to be measured directly at some masses, those masses must
not be included. An initial mass scan can be used to identify regions of the
spectrum that should be skipped [ 1531. The initial mass scan might also be used to
determine whether certain molecular ions can be excluded from the model be-
cause they are at insignificantly low levels. If there are more individual compo-
nents in a segment of the spectrum than masses, there are more variables than
unknowns and a unique solution to multicomponent fitting is not found. This can
be dealt with by establishing relationships between components and constraining
the signals due to certain ions in order to have at least as many equations as
unknowns (independent variables). Ions such as oxides and hydroxides containing
a co~mon element (e.g., CaO+ and CaOH+) should maintain a relatively constant
relationship [154]. Blank spectra can be used to establish limits for ArN+ com-
pared to the signal at mass 54, ArO+ compared to the signal at mass 56, and
ArOH+ compared to the signal at mass 57. A lower limit for Kr in relation to the
signal at mass 84 and value for the “continuum” background (measured at mass
a
220) can also be set from the blank spectrum. Mass bias must be determined and
included. Experimental spectra form Ca and sulfate standard solutions can be used
to establish limits for the CaO+-to-CaOH+ ratio and the ratios for S species [S,+/
SO,+, SO,H+/(S,+ + SO2+), and S03+/(S2+ + SO2+)] [154]. An Excel based
spreadsheet template is available for download from a S~ectroc~i~ic~ Acta
~lectro~ica
article by DeBoer [ 1541 at http://www.elsevier.nl:8O/inc~omepa~e/
saalsabl (download file 520389197).

Many experimental parameters and components affect sensitivity, including the
analyte transport efficiency of the sample introduction system and the mean size
and size distribution of the aerosol entering the IC€? The plasma torch design, rf
generator, load coil, interface between the atmospheric pressure ICP and mass
spectrometer, ion optics, mass spectrometer itself, and detector also affect sensi-
tivity.