153 Research Article Received: 3 September 2010 Revised: 27 November 2010 Accepted: 28 January 2011 Published online in Wiley Online Library: 28 March 2011 (wileyonlinelibrary.com) DOI 10.1002/xrs.1322 PIXE analysis of multilayer targets M. A. Reis, a,b* N. P. Barradas, a,c P. C. Chaves a,b,d and A. Taborda a,b Particle induced X-ray emission (PIXE) analysis is generally accepted to be a depth insensitive technique and most of the work up until recently has been oriented to the analysis of in-depth homogeneous samples. Up until recently, multilayer targets were normally looked upon as a source of problems, implying a limitation to qualitative or semi-quantitative approaches. The growing need to analyse layered targets having complex structures or close Z elements, which provide unsolvable Rutherford backscattering spectrometry (RBS) spectra, pointed out the need to change this situation. The coupling of two software codes, namely NDF and DATTPIXE, was thus carried out a few years ago based on the development of a PIXE yields simulation library, LibCPIXE. Since then, developments were introduced at various levels, from the software code level to the level of analytical methodologies and even equipment. Starting in the concepts of equivalent depth and differential PIXE, and ending in the possibilities opened up by the use of microcalorimeter high-resolution detectors within this context, an overview of the current status of development of the analytical technology involved is made. Copyright c  2011 John Wiley & Sons, Ltd. Introduction The simplest way to use particle induced X-ray emission (PIXE) [1] is in the analysis of thin samples. In this case, the calibration of the technique is quite simple if one uses thin film standards like certified Micromater  films. [2] In this case, using the same beam energy to analyse both the standards and the unknown samples [3] leads to a very simple quantitative analytical procedure. On the other hand, in the other extreme of quantitatively analysable complexity one can find, presently, the analysis of multilayer medium thick films having light, medium and heavy elements in their composition. [4–6] A similar situation is to be found in Rutherford backscattering spectrometry (RBS). In this case, the simple problem is the determination of the areal mass of thin films of heavy elements deposited on light matrices substrates, while the complex problems include conditions as disparate as quantum dots layers, roughness or interface diffusion conditions and medium thick layers of light or medium heavy elements laying deep in the target or existing on top of heavy element films or substrates. It is thus seen that in respect to multilayer samples, different types of difficulties emerge both for PIXE as well as for RBS. It is nevertheless a fortunate condition the fact that most of the parameters badly determined (although needed) by performing one of the techniques are frequently parameters well determined by the other. This is for instances, the case of the depth location of elements, information accessible by RBS and needed by PIXE to properly quantify the expected X-ray yields of elements in the case of samples that are not depth homogeneous. In the case of films composed of close atomic number elements, an example of the opposed situation is found. In fact, in this case, PIXE will provide information on the individual areal mass of each element, an information needed to properly calculate the shape of the RBS spectra without over or underestimation of either film width or film roughness. This highly complementary nature of PIXE and RBS therefore allows overcoming various problems faced in the analysis of multilayer samples, as long as the techniques are used in an holistic coupled mode. In respect to the specific applications, these properties already are important for many different fields. Two examples of these are (1) human heritage, where the analysis of artefacts covered by thin metallising layers, or the characterisation of paintings layers [6] are an important issues and (2) energy production, where the characterisation of films involving close atomic number elements and used in the production of thermophotovoltaic devices, [7–9] is a major problem. Still, these are just specific cases of the more general condition where complex in-depth inhomogeneous samples must be characterised. In this work, a short historical resume is made of the develop- ments that were carried out in this context at Instituto Tecnol ´ ogico e Nuclear (ITN), and in particular those that we consider as rep- resentative of important forward steps in PIXE or more generally in Ion Beam Analysis (IBA), such as the coupling of DATTPIXE [10,11] and NDF [12] algorithms via LibCPIXE [13,14] open source library, or the use of microcalorimeter high-resolution energy dispersive Spectroscopy (EDS) X-ray detectors in this context. In respect to the use of microcalorimeters, a reanalysis of a previ- ous studied sample is presented and the use of this type of X-ray de- tector coupled to a new fitting code, is shown to be of the up most importance for the identification and quantification of the contri- bution of radiative Auger emissions (RAEs) [15,16] for the spectra. History Historically, the PIXE work on depth inhomogeneous samples at ITN can be considered to start in the study of the diffusion of gold on magnesium alloys, carried out in the beginning of the ∗ Correspondence to: M. A. Reis, Instituto Tecnol´ ogico e Nuclear, EN10 Sacav´ em, Apartado 21, 2686-953 Sacav´ em, Portugal. E-mail: mareis@itn.pt a Instituto Tecnol´ ogico e Nuclear, EN10 Sacav´ em, Apartado 21, 2686-953 Sacav´ em, Portugal b Centro de Física At´ omica da Universidade de Lisboa, Av. Prof. Gama Pinto 2, 1649-003 Lisboa, Portugal c Centro de Física Nuclear da Universidade de Lisboa, Av. Prof. Gama Pinto 2, 1649-003 Lisboa, Portugal d Instituto Superior T´ ecnico da Universidade T´ ecnica de Lisboa, Av. Rovisco Pais, 1049-001 Lisboa, Portugal X-Ray Spectrom. 2011, 40, 153–156 Copyright c  2011 John Wiley & Sons, Ltd.