Instrumental correction for background absorption using a double beam instrument or a continuum
source has already been discussed (p. 325). An alternative is to assess the background absorption on a
non-resonance line two or three band-passes away from the analytical line and to correct the sample
absorption accordingly. This method assumes the molecular absorption to be constant over several band
passes. The elimination of spectral interference from the emission of radiation by the heated sample and
matrix has been discussed on page 324 et seq.
Chemical effects include stable compound formation and ionization, both of which decrease the
population of free atoms in the sample vapour and thereby lower the measured absorbance. Examples of
compound formation include reactions between alkaline earth metals and oxyanions such as aluminates,
silicates and phosphates, as well as the formation of stable oxides of aluminium, vanadium, boron etc.
In the former case, releasing agents, typically strontium or lanthanum salts which themselves form
stable oxysalts, are added in excess. Stable oxides can sometimes be dissociated by using a hotter
flame. The addition of easily combusted complexing agents such as EDTA to the sample solution and
the use of fuel rich flames with a low oxygen content are ways of minimizing oxide formation. The use
of inert atmospheres for flameless vaporization has previously been mentioned (p. 328). Ionization of
sample atoms can be suppressed by adding an ionization buffer, i.e. an easily ionizable metal such as
lithium or lanthanum, in excess. This is particularly necessary if the hot (3200 K) nitrous oxide-
acetylene flame is used. The degree of ionization for several elements in this flame is included in Table
8.7. It should be noted that lanthanum salts serve the dual function of releasing agents and ionization
buffers.
Two important types of physical interference are observed. One of these derives from the formation of
solid solutions of one element within another, e.g. chromium in iron. The effect of this is to modify the
volatilization profile of the analyte, and to make it dependent upon the matrix composition. Use of a
releasing agent will usually obviate the problem for flame volatilization, but these interelement effects
have proved much more intractable for flameless methods. Careful matrix matching of samples and
standards can be used to reduce the interference. It should be remembered that the analytical
measurement for a flameless technique is of the height of a transient peak (p. 329) and that the peak
shape will be changed by alterations in the volatilization profile. Measurement of peak areas rather than
peak heights can improve reproducibility. However, even with the application of all these remedies,
interelement effects remain major contributors to the poorer precisions observed in flameless atomic
absorption measurements. Where aspiration of a sample into a flame is employed, variable aspiration
rates due to changes in the surface tension and viscosity with solution composition can occur. Careful
matching of samples and standards with regard to acidity, total salt concentration and other major
constituents is needed to overcome this source of error.