XRF/XRD combined spectroscopy for material characterization in the fields of Material science and Cultural heritage

Every investigation technique has its specific advantages: this is the reason why, in modern research, it is common to combine many investigation techniques – especially the non-destructive ones - to achieve deeper structural information about a sample. X-ray diffraction (XRD) and fluorescence (XRF) techniques are useful non-destructive analytical techniques, with applications not only in industrial field and mining but also in environmental control and cultural heritage monitoring and conservation. In the present research, the advantages of a combined approach with XRF and XRD techniques are considered, due to their complementarity, and a new method of combining data is presented, executing the simultaneous computation of the refinement both for XRF and XRD. In this case, instead of the common approach with an iterative refinement, passing from XRF to XRD and vice versa, both XRF and XRD data are processed simultaneously with a combined Rietveld refinement. This innovative approach has been implemented in the program MAUD, combining original XRD algorithm with the XRF module implementation from the GimPy and JGIXA programs, creating comprehensive radiation–matter interaction model, which takes care of both elastic scattering and photoelectric absorption/fluorescence. Moreover, through a plugin-based application container, Eagle-X, specifically developed for this research project, some easy external wizards have been developed using JAVA language for preliminary XRF analysis and model set-up, which will be in the next future integrated into the MAUD current interface. This new approach has been applied to two case studies. The first study was in the cultural heritage field with the analysis of ancient Venetian coins, called sesini, which were never investigated before. These coins were widely used in the Venetian Republic over a time span ranging from the second half of the 16th until the early years of the 17th century. The rationale of the study was to establish a multilayer model that once validated could be used for fully non-destructive characterization of similar items. The approach, applied to 20 samples from different time periods, has given interesting results. First, the actual composition of the copper-based alloy used for these specific types of Venetian coin has been measured for the first time, using a three-layer model, with also direct measurements on the coin cross-section for validating the data obtained. Second, the detailed characterization of the coins provided essential background knowledge for fully non-destructive characterization of the same kind of coins. Third, the data obtained were very interesting from a historical point of view, because the silver depletion, which this research has investigated over the coin series, reflects a political and economic situation in strong evolution for the Venetian Republic in the second half of 16th century. Political and economic competitors and a continuous effort in military confrontations obliged Venice to revise its coin system and values not only for sesini but also for the other silver-based coins, with larger value, in a process called debasement. The second application of the combined approach regarded an industrial application concerning a sintered titanium alloy, Ti6AlV4, that has the widest use (about 45% of the total production), because of good machinability and excellent mechanical properties. This is an alloy which contains the two allotropic forms of Ti, the Ti-alpha, which has compact hexagonal cell, stable at room temperature, and Ti-beta phase, which has a body-centred cubic lattice, stable over 882°C. The presence of the two phases is related to the presence of atomic elements which are alpha- and beta-stabilizers. In this case study six samples, produced with Selective Laser Melting (SLM) technology, with different production parameters, has been considered, and a model based on a surface layer of compact oxide and a bulk with the alloy only has been adopted. The model has evidenced the presence of the TiO2 oxide on the surface, as attended from existing literature, and confirmed the quality of the alloy because for all the samples, the investigated areas report Al e V content inside the ranges required by ASTM and ISO specifications. The analysis has allowed also to investigate the presence of contaminants like copper due to the cutting process by Electrical Discharge Machining (EDM), and to find a correlation between the content of Ti-beta phase inside the samples and the combined presence of iron and silicon, which increases as soon as increase also the two elements. Moreover, the increase of Ti-beta phase is boosted by the contemporary increase in energy density during SLM production process. This is consistent with the fact that higher energy allows a higher localized temperature in SLM process and the equilibrium fraction of beta phase rises at high temperatures. This then leads to a higher fraction of alpha+beta phases at room temperature and, because the cooling rate was the same for all samples, this means a higher fraction of  phase at room temperature. The application of the technique to the two case studies is very productive from the informational point of view, but a critical aspect for a successful application of the technique is the sample. No preparation is virtually needed for analysis but, of course, this is immediately true for industrial components as soon as they are produced, but it is not so true for archaeological artefacts, where the condition of production, history and store conditions are unknown. Corrosion patinas can alter the read of the data, and some care must be taken for analysis, not only because the patinas may not be homogenous, but also because the depth of penetration for XRF and XRD are not the same, respect to the same substrate. The cleaning of the artefacts is not always possible or desired by the owners, so this can at first stage complicate the approach to combined analysis, regarding the model to be adopted in material simulation for data interpretation. In any case, the combined analysis remains a valid approach provided that the user is conscious of the limits in terms of depth of analysis, linked to the analysis tool (X-ray beam, detector, etc…) and to the surface conditions of the sample.