Functional principles
M4 TORNADO
25
Procedure Spectrum
4 Deconvolution
Especially for multi-element
samples, overlaps of the individual
fluorescence peaks in the
spectrum occur because of the
line diversity and the finite detector
resolution. To allocate the correct
intensities to the individual
elements, the spectrum must be
separated into the individual
element lines by a complex
mathematical process
(deconvolution or fitting with pure
element peaks).
The result is the peak area
(intensity) for each element of the
individual not overlapped
fluorescence peaks.
5 Statistical error Sigma
For each deconvoluted
fluorescence peak the absolute
standard deviation on the basis of
the statistical error of the complete
peak and the background is
calculated.
BGii
NN 2+=
δ
δ
i
= standard deviation for the peak area
N
i
= net peak area for the element i
N
BG
= background area
6 Fit quality (Chi
2
)
The fit quality is a statistical
parameter, which allows
conclusions about the quality of
the deconvolution. The
standardized square sum of the
differences between the measured
and the calculated, deconvoluted
intensities is calculated for all
channels.
The value for the fit quality should
preferably be smaller than 10.
High values (>10) are an indicator
for misidentified or non-identified
elements or for an inaccurate gain
correction.
=
−
−
=
2
1
2
2
12
2
))1((
11
n
ni
i
i
yy
nn
δ
χ
To Next Page
Loading...
Need help?
General
