Parameter study of self-absorption effects in T otal Re ection X-ray Fluorescence–X-ray Absorption Near Edge Structure analysis of arsenic F . Meirer a, , G. Pepponi b , C. Streli a , P . Wobrauschek a , P . Kregsamer a , N. Zoeger a , G. Falkenberg c a Atominstitut, Vienna University of Technology, 1020 Wien, Austria b Fondazione Bruno Kessler (FBK), via Sommarive 18, 38100 Povo (Trento), Italy c Hamburger Synchrotronstrahlungslabor at DESY, 22607 Hamburg, Germany A B S T R A C T A R T I C L E I N F O Article history: Received 31 October 2007 Accepted 14 May 2008 Available online 29 May 2008 Keywords: TXRF XANES Self-absorption Arsenic Total re ection X-ray Fluorescence (TXRF) analysis in combination with X-ray Absorption Near Edge Structure (XANES) analysis is a powerful method to perform chemical speciation studies attraceelement levels. However, when measuring samples withhigher concentrations and in particular standards, damping of the oscillations is observed. In this study the in uence of self-absorption effects on TXRF–XANES measurements was investigated by comparing measurements with theoretical calculations. As(V) standard solutions were prepared at various concentrations anddried on at substrates. The measurementsshowed a correlation between the damping of the oscillations and the As mass deposited. A Monte-Carlo simulation was developedusing data of the samplesshapes obtained from confocal white light microscopy. The results showed good agreement with the measurements; they con rmed thatthe key parameters are the density of the investigated atom in the dried residues and the shape of the residue, parameters that combinedde ne the total mass crossed by a certain portion of the incident beam. The study presents a simple approach for an a priori evaluation of the self-absorption in TXRF X-ray absorption studies. The consequences for Extended X-ray Absorption Fine Structure (EXAFS) and XANES measurements under grazing incidence conditions are discussed, leading to the conclusion that the damping of the oscillations seems to make EXAFS of concentrated samples non feasible. For XANES “ ngerprint” analysissamplesshould be prepared with a deposited mass and sample shape leading to an acceptable absorption for the actualinvestigation. © 2008 Elsevier B.V. All rights reserved. 1. Introdu c tion Total Re ection X-ray Fluorescence Analysis (TXRF) is a sensitive technique for qualitative and quantitativeelemental determination of traces. Typical TXRF set-ups use multilayer monochromators as preferred compromise among high primary ux and low spectral background [1,2]. Synchrotron radiation TXRF offers detection limits in the fg range for transition metals with a multilayer monochromator and a bending magnet beamline [3,4]. For dilute samples X-ray Absorption Spectroscopy (XAS) is commonly performed in uorescence mode as the uorescence signal is proportional to the absorption coef cient and it gives a better signal to background ratio. With a perfect crystal monochromator TXRF acquisition can be utilized for XAS to gain chemicalinformation on the element of interest. With this modi ed set-up there is a ux reduction of abouttwoorders of magnitude, but it isstill suf cient for the analysis at ppb level [3,5–7]. This approach allows theextension of XAS to traces in droplet samples where only small amounts are available. Such measurements can also be performed for the speciation of light elements using a plain grating monochromator [8]. Fluorescence acquisition of XAS spectra of concentrated sample suffers from self-absorption which causes damping and also broad- ening of the oscillations. Some authors have performed quantitative speciation through analysis of XANES (X-ray Absorption Near Edge Structure) spectra by tting them with analytical functions [9, 10]. Another approach deals with the corrections of the measured spectra to account for self-absorption. Many authors have investigated self-absorption effects in XAS using uorescence acquisition depend- ing on the angle of incidence and detection and have proposed correction models [11, 12]. Due to irregular sample shape and the very shallow angle of incidence these models are not applicable to TXRF. The extreme grazing incidence geometry used by TXRF enhances these self-absorption effects due to theextendedpath length of the incident beam in the droplet. This path length is equivalentto the penetration depth of the incident beam and is therefore energy dependent. As theenergy changes during a XANES scan the size of volume where the uorescence photons originate from is varying. Higher absorption meanssmaller excited volume and therefore less uorescence intensity (and vice versa). Consequently this leads to a Spectrochimica Acta Part B 63 (2008) 1496–1502 This paper was presented at the 12th Conference on Total Re ection X-ray Fluorescence Analysis and Related Methods held in Trento (Italy), 18–22 June 2007, and is published in the Special Issue of Spectrochimica Acta Part B, dedicated to that conference. Corresponding author. Atominstitut, Vienna University of Technology, Stadionallee 2, 1020 Wien, Austria. Tel.: +43 1 58801 54 155; fax: +43 1 58801141 99. E-mail address: fmeirer@ati.ac.at (F. Meirer). 0584-8547/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10. 1016/j.sab.2008.05.004 Contents li sts ava il abl e at Sc i enceDirect Spectrochimica Acta Part B j ourna l homepage: w w w.e l sev i er.com/ l ocate/sab