Elsevier

Applied Radiation and Isotopes

Volume 85, February 2014, Pages 92-95
Applied Radiation and Isotopes

Portable XRF and principal component analysis for bill characterization in forensic science

https://doi.org/10.1016/j.apradiso.2013.12.004Get rights and content

Highlights

  • The paper is about a direct method for bills discrimination by EDXRF and principal component analysis.

  • The bills are analyzed directly, without sample preparation and non destructively.

  • The results demonstrates that the methodology is feasible and could be applied in forensic science for identification of origin and false banknotes.

  • The novelty is that portable EDXRF is very fast and efficient for bills characterization.

Abstract

Several modern techniques have been applied to prevent counterfeiting of money bills. The objective of this study was to demonstrate the potential of Portable X-ray Fluorescence (PXRF) technique and the multivariate analysis method of Principal Component Analysis (PCA) for classification of bills in order to use it in forensic science. Bills of Dollar, Euro and Real (Brazilian currency) were measured directly at different colored regions, without any previous preparation. Spectra interpretation allowed the identification of Ca, Ti, Fe, Cu, Sr, Y, Zr and Pb. PCA analysis separated the bills in three groups and subgroups among Brazilian currency. In conclusion, the samples were classified according to its origin identifying the elements responsible for differentiation and basic pigment composition. PXRF allied to multivariate discriminate methods is a promising technique for rapid and no destructive identification of false bills in forensic science.

Introduction

Counterfeiting of money is a problem faced by governments in most countries. Counterfeiters have improved their techniques of forgery, “imitating” the various security features that are present in paper money, added to the printing process. Bills chemical profile becomes an efficient methodology for criminal investigations (Wanderson et al., 2011), including the combat to counterfeiting of banknotes.

Considering the analytical techniques capable to analyze this kind of sample, Raman spectroscopy is one of the most used for ink analysis (Geiman et al., 2009, Muehlethaler et al., 2011) and Mass Spectrometry (MS) is very important to obtain the chemical profile very accurately (Eberlin et al., 2010). A very promising technique in forensic applications is portable X-ray Fluorescence (XRF). Although XRF will not replace atomic spectrometry techniques such as MS for sub-parts per million level analyses, it offers a number of significant advantages including minimal or no sample preparation, high sample throughputs, rapid and definitive identification of many elements, and accurate no destructively quantitative results. As should be obvious from many recent news reports on elevated levels of toxic elements in children's lunchboxes, toys, and supplements, field-portable XRF analyzers can fill a very important niche and are becoming increasingly popular for a wide variety of elemental analysis applications (Palmer et al., 2009, Trombka et al., 2002).

Indeed, by using multivariate analysis, the range and complexity of problems that can be solved by analytical spectroscopy may be increased (Marten and Dijksra, 1976), which is a relatively new discipline applied to XRF data, used for management, analysis, and visualization of large data sets in order to extract relevant physical and chemical information from extensive and complex multivariate data such as spectroscopic measurements: EDXRF spectra carry various (explicit and implicit) signatures on the properties of analyzed sample matrices (Kaniu et al., 2011).

Counterfeit 100-dollar bill using XRF was first measured by Ninomiya et al. (1995), using a fine beam system. The discrimination of true and false bills was obtained comparing the spectra and the relative intensity of the available peaks. So, counterfeit bills which would be made exquisitely could be discriminated non destructively and easily from a true bill using XRF.

The objectives of this study were: (a) to demonstrate the potential of Portable X-ray Fluorescence technique and multivariate analysis method of Principal Component Analysis (PCA) for classification of bills in order to use it in forensic science; (b) to identify qualitatively the elements responsible for pigmentation in the different regions of the same sample due to its colors.

Section snippets

Experimental

Bills of Dollar, Euro and Real (Brazilian currency) were measured directly at different colored regions, without any previous preparation. All bills were new ones, bought directly from the Bank of Brasil. One 50 Dollar bill, one 50 Euro bill and six Real bills (2, 5, 10, 20, 50 and 100), totalizing 8 samples and 59 measurements, were analyzed. In each bill, from 6 to 8 measurements in regions of different colors were performed.

The portable homemade EDXRF system consists of a mini X-ray tube (Ag

Results and discussion

Spectra interpretation allowed the identification of Ca, Ti, Fe, Cu, Sr, Y, Zr and Pb. Calcium, Ti and Fe were identified in all samples. Yttrium was found only in the Euro bill. Strontium and Zr were measured just in Real samples. Pb was found only in the 100 Real bill. Fig. 1, Fig. 2, Fig. 3 present the spectra of the bills by its origin, showing the bills differentiation by its PXRF spectra.

The blank paper of the three groups of banknotes presents the pigment white of titanium, although

Conclusion

The samples were classified from the X-ray fluorescence spectra of each bill, in a fast and non destructive way using portable equipment. Besides, elements from each different pigmentation region allowed inferring its basic composition. The samples were discriminated according to its origin identifying the elements responsible for the differentiation as follows: Ti for Euro bill, Fe for dollar and Ca for Real. EDXRF allied to multivariate discriminate methods is a very promising technique for

References (9)

There are more references available in the full text version of this article.

Cited by (33)

  • X-ray fluorescence spectroscopy

    2023, Non-Destructive Material Characterization Methods
  • pXRF in tropical soils: Methodology, applications, achievements and challenges

    2021, Advances in Agronomy
    Citation Excerpt :

    Among them, portable X-ray fluorescence (pXRF) spectrometry has gained in popularity as a quick, inexpensive, reliable, non-destructive, and environmentally-friendly method to determine multi-elemental content in different matrices like rocks, soils, metals, food, and water (Weindorf et al., 2014). As a consequence, pXRF systems have been increasingly used in selected applications for in situ or ex situ analyses of a wide variety of research fields such as geochemistry (Al-Musawi and Kaczmarek, 2020), geology (Steiner et al., 2017), archaeology (Scharlotta and Daniels, 2020), plant material characterization (Borges et al., 2020a; Guerra et al., 2018; McGladdery et al., 2018), water analysis (Pearson et al., 2018), forensic science (Appoloni and Melquiades, 2014; Schweitzer et al., 2005), animal science (Hoffmann et al., 2020), and environmental quality assessment (Mukhopadhyay et al., 2020). Amid these research fields, pXRF units have attracted great interest of soil scientists.

  • Determining old Chinese non-circulating paper money's authenticity using μ-Raman and MA-XRF spectroscopies

    2020, Journal of Cultural Heritage
    Citation Excerpt :

    The literature does not report many studies in which the MA-XRF technique has been applied for studying paper money. However, many studies on banknotes have been carried out using conventional X-ray fluorescence (XRF) that consists of the analysis of a limited number of regions (or spots) of the sample, for example [21–25]. In recent times, XRF imaging configured as Micro-XRF imaging (MXRF) or Macro-XRF imaging (MA-XRF) has been widely used to investigate the chemical nature of inks or pigments present in paper (such as stamps, manuscripts, parchments, etc.) with the advantage of characterising the whole surface of the sample instead of just a few sparse points [26–29].

View all citing articles on Scopus
View full text