Thickness determination of gold layer on pre- Columbian objects and a gilding frame, combining pXRF and PLS regression Fabio Lopes, a * Fabio Luiz Melquiades, a Carlos Roberto Appoloni, a Roberto Cesareo, b Marcia Rizzutto c and Tiago F. Silva c In conservation, restoration and characterization studies of art and archaeological objects, the improvement of analytical tech- niques is a tendency. X-ray fluorescence (XRF) is a versatile technique, and it has been widely used in the last decades for charac- terization of a great variety of materials (metals, glass, paints, inks, ceramics, etc.) applied to cultural heritage studies. Besides the chemical composition, it is possible to infer the layer thickness through XRF, enabling a general knowledge of the manufacturing techniques implemented by the culture of origin, as well as the association with the technological level reached for the production of each kind of artefact. The aim of this study is to introduce an alternative way for gold thickness determination of coatings in cultural heritage objects, combining portable XRF data and partial least square regression. As a case of study, we present the use of this methodology in portable XRF measurements performed in situ on a gilding frame in Brazil and in two pre- Columbian artefacts from Chavin culture in Peru. Gold layers with thicknesses determined by Rutherford backscattering spectrom- etry (RBS) were used as standards to perform a calibration model and to check the methodology before its application to unknown artefacts. Copyright © 2016 John Wiley & Sons, Ltd. Introduction Many analytical techniques have been widely used in the last de- cade for the studies and characterizations of archaeological and historical objects of art, especially the X-ray fluorescence (XRF), its variant portable XRF (pXRF), particle-induced X-ray emission and instrumental neutron activation analysis), which are all adequate techniques for these kinds of studies due to the quantity and quality of information that can be obtained on the elemental composition of objects. [1–3] In this list, XRF is a versatile technique that combines a non- destructive, well-established technique, with high analytical sensitivity that allows simultaneous multi-elemental analysis with low cost and simple instrumentation. Thus, the technique is partic- ularly suitable for archaeological studies where the characterization of materials or process is specially important. [4–6] In recent years, the use of pXRF has increased in museums, churches and excavations due to its portability. [7–9] In fact, the characterizations of metallic pieces are among the most prominent applications of PXRF in archaeometry and cultural heritage, in particular characterization of pre-Columbian alloys, golden or silvered objects and pigment analysis. [10–17] However, it is also possible to analyse the layer thicknesses, allowing the study of manufacturing techniques or even the tech- nological level reached to produce each kind of artefact. A method commonly used for thickness determination is differential attenua- tion, in which the net intensity of Kα and Kβ peaks of the elements of interest in the layer and in the base of the work of art is determined. [18–23] Another possibility is the use of multivariate statistics to obtain the same information. It is known that the concentration and other sample information are proportional to the peak intensity and sometimes to the scattering peak intensity. The use of multivariate statistics improves data interpretation, enabling the determination of implicit information. This study aim is to introduce an alternative way for gold thick- ness determination of coatings in cultural heritage objects, combin- ing pXRF data and partial least square (PLS) regression. As a case study, we present the use of this methodology in pXRF measure- ments performed in situ on different objects, including a gilded frame in Brazil that was in restoration process and knows that the thickness is important for proper restoration, and in two pre- Columbian artefacts from Chavin culture in Peru, where there was a doubt about the method of manufacturing. Gold layers with thick- nesses determined by Rutherford backscattering spectrometry (RBS) were used as standards to perform a calibration model and to check the methodology before its application to unknown artefacts. * Correspondence to: Fabio Lopes, Laboratório de Física Nuclear Aplicada, Departamento de Física, Universidade Estadual de Londrina, Campus Universitário, Caixa Postal 10011, CEP: 86057-970, Londrina, PR, Brazil. E-mail: fabiolopes@uel.br a Laboratório de Física Nuclear Aplicada, Departamento de Física, Universidade Estadual de Londrina, Campus Universitário, Caixa Postal 10011, CEP: 86057- 970, Londrina, PR, Brazil b Instituto di Matematica e Fisica, Università degli Studi di Sassari, Via Vienna 2, 07100, Sassari, Italy c Instituto de Física da Universidade de São Paulo, Rua do matão, travessa R 187, CEP 05508-090, São Paulo, Brazil X-Ray Spectrom. 2016, 45, 344–351 Copyright © 2016 John Wiley & Sons, Ltd. Research article Received: 22 June 2016 Revised: 22 August 2016 Accepted: 23 August 2016 Published online in Wiley Online Library: 27 September 2016 (wileyonlinelibrary.com) DOI 10.1002/xrs.2711 344