Chemistry Reference
In-Depth Information
major elements. For instance,
47
Ca (t
1/2
4.5 d) and
32
P
(t
1/2
14.28 d) produce intense Bremsstrahlung radiation,
which prevents the determination of several longer-
lived but low-energy isotopes. Under such conditions,
RNAA has to be done.
Interesting literature is available on the International
Atomic Energy (IAEA) website (2004).
where
S
is the signal response of the detection
method,
k
the calibration factor,
c
the concentration of
the calibrant, and
b
the intersection with the y-axis. The
k
value refl ects the sensitivity. The higher the
k
value for
a given concentration range, the better the sensitivity.
External calibration modes, where the sample and
the corresponding calibrant are separately measured,
can only be used when there is no effect (either quench-
ing or enhancement) by the sample matrix on the ana-
lyte signal. Internal calibration (i.e., standard addition
method) becomes then compulsory to include possible
infl uences of matrix composition on the signal inten-
sity of the detection system (Skoog
et al.
, 1994). Two
or more aliquots of the sample are transferred to volu-
metric fl asks. One is diluted to volume, and the signal
is measured. A known amount of analyte is added to
the second, and its signal is measured after dilution to
the same volume. Data for other additions may also be
obtained. First, it should be checked that a linear rela-
tionship between the signal and concentration exists
(and this should be established by several standard
additions). If several additions are made, the signal can
be plotted versus the added concentrations. The result-
ing straight line can be extrapolated to signal zero. The
intercept with the concentration axis gives the con-
centration of the analyte in the sample. The addition
method has the advantage that it often compensates
for variations caused by physical and chemical inter-
ferences in the sample solution.
In the case of total element determinations using
F-AAS, GF-AAS, HG-AAS, HG-AFS, ICP-AES, and
ICP-MS, spectrophotometry, electrochemical meth-
ods, and biosensors for monitoring metal ions, it may
be possible to calibrate versus the external mode. It
is compulsory, however, to check the validity of this
assumption by comparing the results of at least one
series of samples versus those obtained by the more
labor-intensive standard addition method.
In case of neutron activation analysis, external cali-
bration is the rule. The one-point calibration curve is
applicable, similarly as in isotope dilution mass spec-
trometry. ES-MS-MS is a stand-alone case as far as cali-
bration is concerned, because this method is primarily
used for obtaining structural information of the spe-
cies. Calibration is, of course, evident when calibrants
of the species are available.
Speciation analysis is usually the result of a separa-
tion of species followed by elemental detection. The
standard addition method applied to the original sam-
ple, before the separation or enrichment of the species,
is only justifi ed when an equilibrium has been reached
between the analyte in the sample and the standard.
This also supposes that the calibrant is commercially
available or can be synthesized by the laboratory.
10.2.10 Spark Source Mass Spectrometry
Spark source mass spectrometry (SSMS) used
to be the most sensitive widely used multielement
technique until the 1960s. The basis for SSMS is the
formation of ions when the sample is subjected to a
high-energy discharge (Morrison, 1979). The appara-
tus uses a vacuum spark in which a high-voltage radio
frequency discharge (20-100 kV) is produced between
two closely spaced electrodes containing the material
to be analyzed. This sparking results in vaporization
and ionization of sample constituents. The repetition
rate and duration of the radio frequency spark source
is variable to meet the various analytical requirements.
Nearly all masses are integrated simultaneously over
a period of time to provide high sensitivity. For SSMS,
the sample has to be electrically conductive. This is
achieved for nonconducting biological material by
blending them with high-purity graphite followed by
briquetting to form electrodes that sustain the vacuum
spark. Two major problems arise. First, the graphite
used for making the electrodes also contains trace ele-
ment impurities, and they may exceed the concentra-
tions in the biological sample. The second problem
lies with the interferences caused by the organic ions
of differing complexity obtained from many possi-
ble combinations of C, H, O, N, S, P, etc. Therefore, to
take full advantage of this multielement method, the
samples are ashed before analysis. This will, however,
cause the loss of volatile elements such as mercury and
selenium, among others.
11 CALIBRATION
The calibration of the method forms a very critical
point during the analysis. Calibration will be mostly
done in the relative way (Heumann, 2003). The signal
caused by the detection of the element in a sample will
be compared with a set of calibration samples with
known content. Because of the practical advantages
of linear calibration graphs, they are always favored.
Linear calibration graphs can be obtained by measuring
only a few calibration standards and, in addition, are
easily described by a simple mathematical function:
S =
kc
+
b,
