The procedure is strictly analogous to that used for absorbance measurements in UV and visible
molecular spectrometry (p. 355). To avoid interference from emission by excited atoms in the flame and
from random background emission by the flame, the output of the lamp is modulated, usually at 50 Hz,
and the detection system tuned to the same frequency. Alternatively, a mechanical 'chopper' which
physically interrupts the radiation beam, can be used to simulate modulation of the lamp output.
Broad band absorption by volatile molecular species can often give rise to a high and variable
background. Various instrumental strategies have been used for making the appropriate corrections.
One popular method has been the use of a deuterium continuum source as shown in Figure 8.26. A half-
silvered mirror enables light from both the sharp line source and the continuum source to be directed
through the sample vapour at the same time. This mirror is coated with small reflecting circles, thus
leaving a light path for the hollow cathode radiation, and with the continuum source modulated 180°
out of phase with the sharp line source the two signals may be electronically distinguished. The
continuum radiation will be diminished to a negligible extent by sharp line absorption but to a much
greater extent by the broad band absorption. Thus its intensity can be used to compute a background
correction for the analytical measurement. An alternative way of producing broad band radiation is the
Smith-Hieftje technique. This uses the hollow cathode lamp as a source, by running it at high currents
when broad band radiation is emitted. The lamp is pulsed alternately at high current, and at lower
currents for sharpline emission.
Figure 8.26
Practical system for flame atomic absorption spectrometry including
a deuterium background corrector.