Page 78 - An Introduction to Analytical Atomic Spectrometry - L. Ebdon

Page 60

removal from the atom cell. If N is the number of atoms at time t, dN /dt is the rate of change of the
number of atoms.

In a graphite atomizer, the atoms will appear according to a kinetic rate equation which will probably
contain an exponential function. As the number of atoms in the atom cell increases, so does the rate of
removal, until, at the absorption maximum (peak height measurement), the rate of formation equals
the rate of removal. Thereafter, removal dominates.
The response function, which is normally a peak and may be distorted to some extent by the
electronics, clearly is the difference between the formation and removal functions at that time. Atoms
leave the atom cell partly by diffusion and according to the velocity of the purge gas. The rate of
formation of atoms is more difficult to identify.
L'Vov first developed a kinetic model for atomization, based on increasing temperature, such as that
found in a rod-type system. Fuller developed a model for atomization under isothermal conditions,
applicable to less volatile elements in tube-type systems. Fuller's model assumes first-order kinetics
and involves a number of other assumptions, but its usefulness has been demonstrated. For example, it
confirms the usefulness of integration when atomization is slow (at relatively low temperatures or
when investigating involatile elements), the enhancement of sensitivity available from stopping the
flow of purge gas during the atomization cycle, and indicates methods for the control of interferences.

It is clear from experimental data that the rate of removal of the analyte can exceed the rate of supply.
Hence there is an advantage to be obtained in rapid heating (e.g. 1000 K s ) and stopping the purge-gas
-1
flow during atomization.

While a full treatment of kinetic theories is beyond the scope of this book, it is clear that they have
added much to our understanding of observed events in tube furnaces. It can be expected that more
sophisticated models will offer more comprehensive explanations of observed behaviour.

Q. Would you expect the electrothermal atomization mechanism for zinc to differ depending on
whether the sample was dissolved in nitric or hydrochloric acid?

Q. How can we explain the observed shapes of peaks obtained using furnace atomizers?

3.5 Interferences

Electrothermal atomizers are usually regarded as being more prone than flames to interferences,
although it has now been clearly demonstrated that