Absolute analysis in electrothermal atomization atomic absorption spectroscopy-an atomization system for ~ o~ ning all the atoms injected in the opticaf beam zyxwvutsrqponmlkjihgfedcbaZYXWV G. TORN* and P. RESCHIGLIAN Department of Chemistry “G. Ciamician”, Via Selmi, 2 - I-40126 Bologna, Italy and F. FAGIOLI and C. LOCATELLI Department of Chemistry, Universitk Degli Studi, Via Borsari, 46 - I-44100 Ferrara, Italy (Received 17 July 1992; accepted 22 December 1992) Abstract-The apparatus presented in this paper is able to simultaneously confine all the atoms injected into the atomizer within the optical beam. As a consequence, during the time interval fulfilling this condition, the absorbance is constant. Experimental absorbance vs time curves, with Cd as analyte, show a constant value for more than 0.2 s. The curve shapes are in agreement with those theoretically derived for pure diffusion. This finding should lead to a method for absolute analysis that is more accurate than the methods zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA use d in ETA-AAS up to now. In fact, with this approach, one has to measure a constant signal while in the case of area meas~ements one has to integrate a signal that ex~nentially decays toward the baseline value. Data presented in this paper refer to the most volatile elements (Cd, Hg and Pb) in simple matrices. The possibility of extending the presented method to complex matrices and less volatile elements is dependent on the availability of both a more powerful power supply system and well-suited atomizer design. AN ABSOLUTE method of analysis has been defined [l, 21 as a method through which a signal can be related to the concentration or the quantity of an analyte by a theoretical equation sufficiently reliable to allow a direct calculation in absolute units of the desired quantity from a single measurement. Electrothermal atomization atomic absorption spectroscopy (ETA-AAS) is a technique well-suited for absolute analysis because the signal is related to the quantity of analyte by the well-known Lambert-Beer law. This possibility has been recognized from the beginning [2] and has been pursued up to now with increasingly refined adjustments and corrections to the experimental data. For obvious practical reasons, all the atomizers are semi-enclosed systems from which the atoms, once atomized, readily escape, If we assume that the atoms that reach the extremities of the atomizer disappear from the optical beam, then the Lambert-Beer law can be written as: zyxwvutsrqponmlkjihgfed A, = K zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON N,& (1) where A, is the absorbance at time t, K (cm”) a spectroscopic constant [3, 41, S, (cm*> the cross-section of the atomizer and N, the number of atoms of the analyte in the vapour phase present in the atomizer. Equation (1) is valid if the atoms are homogeneously distributed in planes perpendicular to the optical beam [5]. The use of Eqn (1) for absolute analysis is possible if the relation between N, and NO, the number of atoms injected in the atomizer, is known. This relation can be obtained through models of which a large number of increasing complexity are available. The model that best describes the experimental results 13, 61 is a convolution integral of the type: * Author to whom all correspondence should be addressed. 681