Micro-invasive approach for non-destructive XRF analysis on light matrix: inside the Opus Lemovicense † Letizia Bonizzoni, a * Anna Galli, b,c,d Maria Pia Riccardi e and Chiara Maggioni f Speed, manageability and above all non-destructivity are the main features that make portable X-ray fluorescence (PXRF) a ver- satile analytical technique widely used in archaeometry. The extensively use of PXRF instruments in recent years during in situ investigations [1–3] makes the scientific community wonder how to use in the best way the data collected, as close attention must be paid when analysing non-standard materials such as ancient handcrafted works. Copyright © 2015 John Wiley & Sons, Ltd. Beside, when using energy-dispersive X-ray fluorescence (EDXRF) for the quantitative analysis of low-Z matrix samples, the problem of evaluating self-absorption effects arises, strongly with portable instruments. The elements with Z < 12 are in fact almost undetect- able, and the exact matrix composition is in general not known. Different approaches have been proposed and applied in the last decades [4–8] ; at the same time, attempts to classify light matrix samples avoiding a precise quantitative determination have been made. [9–11] Going on in the same direction and starting from a case study of light matrix materials (enamels from a Limoges altar cross datable to the end of the 12th century, held in the Poldi Pezzoli Museum in Milan, Italy), this work combines different techniques, all based on the detection of secondary X-rays produced by the interaction of a primary beam with the sample material. EDXRF, portable X- ray fluorescence (PXRF) and electron probe micro-analysis (EPMA) are exploited, each one for its own peculiar trait, to give a complete characterization in terms of composition of low-Z matrix samples. Further, EPMA data – quantitative elemental results at very small spot size, but over a large scale – are the starting point to better understand the limits of the local scale answer from EDXRF and the larger scale one from PXRF, both giving a partial vision of the whole object. This approach can lead to develop a procedure for the most suitable interpretation of X-ray fluorescence (XRF) data to get the best performances from portable spectrometers. Historical introduction 1 The medieval expression Opus Lemovicense indicated the applica- tion of champlevé enamels to gilded copper metalworks, becom- ing shortly a hallmark of the town of Limoges. We do not know for certain where the innovative processing had been developed: the first known attestations, still experimental, were realized at the St. Foy Abbey in Conques (1100–1130 ca). Within a generation, the champlevé technique was widely established in southwestern France (Aquitaine) and northern Spain (Castile), with a particular emergence of the Limousin. [12] The reasons of such an immediate reception can be traced in the cheapness and the simple making of the new procedure compared to the very precious and virtuosic one of High Medieval cloisonné on gold. The support employed – a highly malleable, almost pure copper – allowed to work easily both in engraving and in chiselling, while the cells meant to receive enamels were obtained by champlevage, namely gouging away the surface and carving them in the thickness of the plaque. [13] In 1150s, the ateliers multiplied in the Meuse–Rhine area too: their products distinguished themselves for literate iconographies, classicizing style and harmonious palette. Conversely, Limousin workshops, which were equipped for a large-scale production with a precise division of tasks, aimed at accessible subjects, at a style faithful to Romanesque premises, even if progressively enhanced in dynamism both in compositions and in characters’ attitudes, and at a chromatic brilliance, hinged on the lapis lazuli blue, around which refined variations and gradations were built. [14] The Opus Lemovicense soon acquired European reputation, with exportations of ever larger scope (Spain, Italy, England, Scandinavia, Slavic Countries, up to Russia and China) and more and more considerable quantity: we estimate that the over 10 000 surviving objects represent about 10 of the production effectively realized till * Correspondence to: Letizia Bonizzoni, Dipartimento di Fisica, Università degli Studi di Milano, Via Celoria 16, 20161 Milan, Italy. E-mail: letizia.bonizzoni@mi.infn.it † Presented at the European Conference on X-Ray Spectrometry, Bologna, Italy, 15–20 June 2014 a Dipartimento di Fisica, Università degli Studi di Milano, Via Celoria 16, 20161 Milan, Italy b Dipartimento di Scienza dei Materiali, Università degli Studi di Milano-Bicocca, via R. Cozzi 55, 20125 Milan, Italy c CNR-IFN, Milan, Italy d INFN, Università degli Studi di Milano-Bicocca, via R. Cozzi 55, 20125 Milan, Italy e Dipartimento di Scienze della Terra e dell’Ambiente, Università degli Studi di Pavia, via Ferrata 1, 20700 Pavia, Italy f Scuola di Specializzazione in Beni Storico-Artistici, Università Cattolica del Sacro Cuore, via Lanzone, 29-20123 Milan, Italy 1 Historical introduction and Analysing Opus Lemovicense are edited by C. Maggioni. X-Ray Spectrom. 2015, 44, 169–176 Copyright © 2015 John Wiley & Sons, Ltd. Research article Received: 29 September 2014 Revised: 16 January 2015 Accepted: 16 January 2015 Published online in Wiley Online Library: 26 February 2015 (wileyonlinelibrary.com) DOI 10.1002/xrs.2596 169