Toward Developing Techniques─Agnostic Machine Learning Classification Models for Forensically Relevant Glass Fragments.

Glass fragments found in crime scenes may constitute important forensic evidence when properly analyzed, for example, to determine their origin. This analysis could be greatly helped by having a large and diverse database of glass fragments and by using it for constructing reliable machine learning (ML)-based glass classification models. Ideally, the samples that make up this database should be analyzed by a single accurate and standardized analytical technique. However, due to differences in equipment across laboratories, this is not feasible. With this in mind, in this work, we investigated if and how measurement performed at different laboratories on the same set of glass fragments could be combined in the context of ML. First, we demonstrated that elemental analysis methods such as particle-induced X-ray emission (PIXE), laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS), scanning electron microscopy with energy-dispersive X-ray spectrometry (SEM-EDS), particle-induced Gamma-ray emission (PIGE), instrumental neutron activation analysis (INAA), and prompt Gamma-ray neutron activation analysis (PGAA) could each produce lab-specific ML-based classification models. Next, we determined rules for the successful combinations of data from different laboratories and techniques and demonstrated that when followed, they give rise to improved models, and conversely, poor combinations will lead to poor-performing models. Thus, the combination of PIXE and LA-ICP-MS improves the performances by ∼10-15%, while combining PGAA with other techniques provides poorer performances in comparison with the lab-specific models. Finally, we demonstrated that the poor performances of the SEM-EDS technique, still in use by law enforcement agencies, could be greatly improved by replacing SEM-EDS measurements for Fe and Ca by PIXE measurements for these elements. These findings suggest a process whereby forensic laboratories using different elemental analysis techniques could upload their data into a unified database and get reliable classification based on lab-agnostic models. This in turn brings us closer to a more exhaustive extraction of information from glass fragment evidence and furthermore may form the basis for international-wide collaboration between law enforcement agencies.

Emerging nuclear methods for historical painting authentication: AMS-14C dating, MeV-SIMS and O-PTIR imaging, global IBA, differential-PIXE and full-field PIXE mapping.

There is a considerable interest in developing new analytical tools to fight the illicit trafficking of heritage goods and particularly of easel paintings, whose high market values attract an ever-increasing volume of criminal activities. The objective is to combat the illicit traffic of smuggled or forged paintworks and to prevent the acquisition of fakes or looted artefacts in public collections. Authentication can be addressed using various investigation techniques, such as absolute dating, materials characterization, alteration phenomena, etc.; for paintings this remains a challenging task due to the complexity of the materials (paint layers, ground, varnish, canvas, etc.) and preferable use of non-destructive methods. This paper outlines results from concerted action on detecting forged works of art within the framework of a Coordinated Research Project of the International Atomic Energy Agency (IAEA) called Enhancing Nuclear Analytical Techniques to Meet the Needs of Forensic Sciences1. One of the main objectives is to foster the use of emerging Nuclear Analytical Techniques (NAT) using particle accelerators for authentication of paintings, with potential application to other forensics domains, by highlighting their ability to determine painting authenticity and to track restorations or anachronistic clues. The various materials comprising a test painting were investigated using an array of NAT. Binder, canvas and support were directly dated by 14C using Accelerator Mass Spectrometry (14C-AMS); binder and pigments' molecular composition was determined using Secondary Ion Mass Spectrometry with MeV ions (MeV-SIMS); paint layer composition and stratigraphy were accurately determined using Ion Beam Analysis (IBA) and differential Particle-Induced X-ray Emission (PIXE); and pigment spatial distributions were mapped using full-field PIXE. High resolution Optical Photothermal Infrared Spectroscopy (O-PTIR) molecular imaging was also exploited. Obtained results are presented and discussed. It is shown that the combination of the above-mentioned techniques allowed reconstructing the history of the test painting.

Inter-laboratory workflow for forensic applications: Classification of car glass fragments.

The International Atomic Energy Agency (IAEA) has coordinated a research project titled "Enhancing Nuclear Analytical Techniques to Meet the Needs of Forensics Sciences" (CRP F11021) with the aim of empowering accelerator and reactor based techniques for applications in forensic sciences. One of the key topics of this project was the analysis and classification of forensic glass specimens using Ion Beam Analysis (IBA) techniques and in particular, Particle Induced X-ray Emission (PIXE). To this end, glass fragments from car windows from different car models and manufacturers provided by the Israeli police force were subjected to PIXE measurements at three laboratories to determine their elemental compositions and possible glass corrosion. Major and trace elements were measured and given as an input to machine learning (ML) algorithms in order to develop classification models to determine the origin of the glass samples. First, we have developed ML models based on the results obtained at each lab. These models successfully classified glass fragments into different car models with an accuracy> 80% on external test sets. Next, we demonstrated that following an appropriate pre-processing step, results from different labs could be combined into a single unified database for the derivation of a classification model. This model demonstrates good performances that matches or surpasses the performances of models derived from the individual labs. This finding paves the way towards establishing an international database that is composed of measurements from various PIXE labs. We believe that using this methodology of combining various sources of measurements will improve models' performances and generality and will make the models accessible to law enforcement agencies around the world.

Identification of Synthetic Organic Pigments (SOPs) Used in Modern Artist's Paints with Secondary Ion Mass Spectrometry with MeV Ions.

This work reports on the first systematic study using secondary ion mass spectrometry with MeV ions (MeV-SIMS) for analysis of synthetic organic pigments (SOPs) that can be usually found in modern and contemporary art paints. In order to prove the applicability of the method to different chemical classes of SOPs, 17 pigments were selected for the analyses. The focus was on blue and green phthalocyanines, yellow and red (naphthol AS) azo pigments, red quinacridone, anthraquinone, and diketopyrrolo-pyrrole pigments. Since there are no reference spectra available for this technique, pure pigment powders were measured first to create a database. Simple two-component paint systems were also prepared for testing purposes by mixing synthetic organic pigments with alkyd and acrylic binders. Commercial paints that contain the SOPs with identical C.I. numbers as in the prepared two-component samples were analyzed. All pigments were successfully identified in commercial products in the MeV-SIMS mass spectra through molecular and larger specific fragment ion peaks in the positive-ion mode. The main advantages of MeV-SIMS over other techniques used in SOPs identification, like pyrolysis gas chromatography mass spectrometry (Py-GC/MS), direct-temperature resolved mass spectrometry (DTMS), and laser desorption ionization mass spectrometry (LDIMS), can be summarized as follows: (i) pigments and binders can be detected simultaneously in the same mass spectrum acquired over a short measurement time (up to 500 s), (ii) only small sample flakes are required for the measurements, which are analyzed without any chemical treatment prior to the analyses, (iii) samples are not consumed during the analyses and can be reused for other measurements, e.g., multielemental analysis by other ion beam analysis (IBA) techniques, such as particle-induced X-ray emission (PIXE). Compared to, e.g., Raman spectroscopy, the significant benefit of MeV-SIMS is the exact identification of the SOPs in the paints even if pigments of similar structures are measured.

Investigation of key factors in preparation of alpha sources by electrodeposition.

The electrodeposition for alpha source preparation, using several electrolyte solution-cathode material combinations, is investigated and evaluated. The investigated factors focused on the electrodeposition time, the applied current, electrolyte volume and anode-cathode distances for the conventional electrodeposition cell (with no external stirring or cooling system). The conditions (temperature and the solution pH) during the electrodeposition process were also studied and discussed. The optimized parameters for each system are provided, and evaluated for the usage in determination of actinides (uranium, plutonium, americium and curium radioisotopes) in various samples.

RADIONUCLIDE AND MAJOR ELEMENT ANALYSIS OF THERMAL AND MINERAL WATERS IN CROATIA WITH A RELATED DOSE ASSESSMENT.

Major elements concentrations (Na, K, Ca, Mg) and the activity concentrations of 226,228Ra, 234,238U, 210Po, 210Pb, 40K and 137Cs in northern and eastern Croatian thermal and mineral waters, collected directly from springs (or wells), are presented herein with total effective doses assessed for those waters that are considered as drinking 'cures' and are available for consumption. The methods used for radionuclide determination included alpha-particle spectrometry, gas-proportional counting and gamma-ray spectrometry, while the major element composition was determined by ICP-MS. The activity concentrations of all of the radionuclides were found to be below the guidance levels set by the WHO and EC Directive, with the exception of one water sample that measured 0.26 Bq L-1 of 228Ra. The effective ingestion dose assessment for the consumption of the so-called water 'cures' during 1, 2 or 4 weeks' time period throughout 1 year was well below the recommended 0.1 mSv for drinking water.

Evaluation of several electrolyte mixture-cathode material combinations in electrodeposition of americium radioisotopes for alpha-spectrometric measurements.

Three different types of electrolytes, subsequently modified and adjusted, in combination with three cathode materials used as source backings were analysed for electrodeposition of americium isotopes for alpha-spectrometric measurements. The obtained results are discussed in terms of electrodeposition yield and source quality (source homogeneity and spectral resolution, FWHM). The optimal conditions for source preparation are provided.

Mussels (Mytilus galloprovincialis) as a bio-indicator species in radioactivity monitoring of Eastern Adriatic coastal waters.

Croatian Adriatic coastal waters are systematically monitored within the Mediterranean Mussel Watch Project using mussels (Mytilus galloprovincialis) as a bio-indicator species. The study includes determination of naturally occurring ((7)Be, (40)K, (232)Th, (226)Ra and (238)U), as well as anthropogenic (137)Cs radionuclides. Activity concentrations in dry weight of mussels' soft tissue along the Croatian Adriatic coast are presented, with spatial and seasonal variations given and discussed. Samples were collected in spring and autumn for the period between 2009 and 2013. Radionuclides were determined by gamma-ray spectrometry. Activity concentrations of (7)Be were the highest in spring periods, especially in the areas with significant fresh water discharges. Activity concentrations of (40)K did not vary significantly with season or location. (137)Cs activities were low, while (232)Th, (226)Ra and (238)U activities were mostly below the detection limit of performed gamma-spectrometric measurement.

Levels of major and trace elements, including rare earth elements, and ²³8U in Croatian tap waters.

Concentrations of 46 elements, including major, trace, and rare earth elements, and (238)U in Croatian tap waters were investigated. Selected sampling locations include tap waters from various hydrogeological regions, i.e., different types of aquifers, providing insight into the range of concentrations of studied elements and (238)U activity concentrations in Croatian tap waters. Obtained concentrations were compared with the Croatian maximum contaminant levels for trace elements in water intended for human consumption, as well as WHO and EPA drinking water standards. Concentrations in all analyzed tap waters were found in accordance with Croatian regulations, except tap water from Sibenik in which manganese in concentration above maximum permissible concentration (MPC) was measured. Furthermore, in tap water from Osijek, levels of arsenic exceeded the WHO guidelines and EPA regulations. In general, investigated tap waters were found to vary considerably in concentrations of studied elements, including (238)U activity concentrations. Causes of variability were further explored using statistical methods. Composition of studied tap waters was found to be predominately influenced by hydrogeological characteristics of the aquifer, at regional and local level, the existing redox conditions, and the household plumbing system. Rare earth element data, including abundances and fractionation patterns, complemented the characterization and facilitated the interpretation of factors affecting the composition of the analyzed tap waters.