Comments on Instrumental Response Factors in Graphite Furnace Atomic Absorption Spectrometry

Abstract

The observations made in at least three recent publications, regarding accurate tracking of transient peaks in graphite furnace atomization, are well founded. Sturgeon and Chakrabarti found, using a Varian AA-5 and an electronic integrator of their own design, that an amplifier time constant of more than 18 msec caused severe distortion of transient peaks. Siemer and Baldwin pointed out that the presence of a 300-msec time constant in a Varian AA-6 analog amplifier circuit caused serious distortion and suppression of transient peaks. Moreover, the position of the analog filtering components in the signal processing sequence determined the extent of peak distortion. Erspamer and Niemczyk described convenient modifications to a Varian AA-6 which reduced the effects of slow analog signal processing on transient peak shapes. These workers also pointed out that microprocessor-based spectrophotometers can potentially provide signal handling suitable for use with graphite furnaces, if analog signal processing is sufficiently fast and the digital electronics sample the signal frequently enough. Liddell performed experiments to determine the electronic requirements for the processing of transient signals from a carbon rod atomizer. His findings were as follows: (a) A modulation frequency of at least 40 Hz is required. (b) The background absorption signal must be measured within 2 msec of the uncorrected total absorption signal. (c) The emission signal must also be measured within 2 msec of the uncorrected total absorption signal. (d) The signal processor must be able to respond to the individual pulses and process them without bias from earlier signals.