Materials Chemistry and Physics 81 (2003) 345–348

SEM–EDX—a useful tool for forensic examinations

G. Zadora∗, Z. Brozek-MuchaInstitute of Forensic Research, 9 Westerplatte St., Cracow 31033, Poland

Abstract

There are two main aims of forensic examination of the physical evidences. The first aim is comparison of the evidence with the referencematerial (called discrimination). The task is to find out whether they could have come from the same object. The second aim, when thereis no comparative material available, is a classification of the evidence sample into a group of objects taking into account its specificchemical and physical properties. Scanning electron microscopy with energy dispersive X-ray spectrometry (SEM–EDX) is a powerfultool for forensic scientists to classify and discriminate evidence material because they can simultaneously examine the morphology andthe elemental composition of objects. Moreover, the obtained results could be enhanced using some methods of chemometric analysis. Afew examples of problems related to the classification and discrimination of selected types of microtraces are presented.© 2003 Elsevier Science B.V. All rights reserved.

Keywords: SEM–EDX; Forensic examination; Microtraces; Chemometry

1. Introduction

Fragments of various materials such as glass, paints, fi-bres and gunshot residues are frequently present at the sceneof such events as car accidents, burglaries, fights or crimescommitted with the use of fire arms. Very small fragmentsof those materials that arise during these events can be re-covered from hair, hands, clothing and shoes of participantsand witnesses of the events and provided as an evidence ata court.

One of the tasks of forensic examinations is identificationand classification of objects. For instance, it is necessary toknow whether gunshot residues are present in a sample takenfrom suspect’s hands or that a glass fragment recoveredfrom suspect’s clothes could originate from a car windowor another group. The second task of forensic examinationsis a comparison of traces collected from the scene of crimewith the evidence materials (e.g. microtraces recovered froma suspect’s clothes). The task is to find out whether theycould have came from the same object.

In order to solve the above-mentioned problems it is nec-essary to obtain information on both the morphology and theelemental contents of the analysed objects. An observationof the morphology of the sample is carried out, using vari-ous optical microscopes, that enable one to observe objectsat the magnifications of hundreds of times. In case when the

∗ Corresponding author. Tel.:+48-12-422-87-55;fax: +48-12-422-38-50.E-mail address: gzadora@ies.krakow.pl (G. Zadora).

information on morphology is insufficient, some physicaland chemical examinations are to be performed. For thesetasks a number of analytical methods can be utilised[1–3].

Traces such as fragments of glass, paints or gunshotresidues most often occur in very small quantities. Thus,sensitive analytical methods are required in order to ob-tain satisfactory results from small amounts of sample. Forthe selection of proper analytical methods one should takeinto account the fact that the method is not allowed to de-stroy the samples because the material might be re-used.One of these non-destructive methods is scanning electronmicroscopy with an energy dispersive X-ray spectrometry(SEM–EDX). It is a powerful tool for forensic purposesbecause one can examine objects considering their mor-phology and the elemental composition. Moreover, theobtained results can be evaluated by using suitable methodsof chemometric analysis.

The aim of this work is the presentation of an SEM–EDXmethod for application in forensic purposes at the Instituteof Forensic Research, Cracow, Poland.

2. Analytical equipment

The examination of gunshot residues, glass and car paintswere carried out, using a scanning electron microscopeJSM-5800, made by JEOL, Japan, with an energy dispersiveX-ray spectrometer Link ISIS 300, Oxford Instrument, UK.The samples were covered with carbon, using a sputteringunit—SCD 050, BAL-TECH, Switzerland.

0254-0584/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved.doi:10.1016/S0254-0584(03)00018-X

346 G. Zadora, Z. Brozek-Mucha / Materials Chemistry and Physics 81 (2003) 345–348

Fig. 1. Inorganic gunshot residue of 20�m in diameter.

2.1. Gunshot residues

Gunshot residues can be of organic and inorganic nature.Organic residues originate from the propellant: unburnedand partially burned gun powder particles, some productsof their transformation and also particles of lubricants. Inor-ganic residues, mostly metallic, originate from the primer aswell as from metallic parts of cartridge and the weapon it-self. Among them only particles originating from the primerreveal specific chemical content and characteristic morphol-ogy: regular or distorted spheres of linear dimension of mi-crometers (Fig. 1). Their morphology is an effect of extreme

Fig. 2. Dendrogram as a result of cluster analysis of various types of ammunition: B, Browning 7.65 mm; b, Browning 6.35 mm; L, Luger 9 mm; M,Makarow 9 mm; S, sporting 5.6 mm; T, Tokarev 7.62 mm. Numbers denote individual experiments within the same type of ammunition.

conditions taking place during explosions of the primer andpropellant initiated by the hit of the firing pin in the primer(high temperature and high pressure, and then fast expansionand cooling), whereas the chemical contents is a derivativeof the content of the primer.

Most frequently, gunshot residues are collected from thesuspect’s hands, where the concentration of these traces ismaximal immediately after shooting. In case more time haselapsed since the event traces should be collected ratherfrom the hair, face and clothes where they usually remainfor longer than on hands.

Particles detected in the material studied are classified tak-ing into account their elemental content[3]. It has been es-tablished from empirical studies that only three-componentparticles containing lead, antimony and barium are uniqueprimer residues. Accompanying two- and one-componentparticles called indicative are also characteristic gunshotresidues, but particles of a similar elemental contents maybe created in other circumstances. Particles containing iron,chromium, nickel, copper, zinc and other elements are typi-cal for the case, projectile and its jacket as well as the barrel.However, they can also originate from subjects of everydayuse and thus they cannot be considered as an evidence offirearm shooting.

Physicochemical examinations of gunshot residues, calledalso chemical ballistics, are helpful, e.g. in identification ofdamages and injuries as the effect of the use of firearms(with indicating the entrance and exit of projectile), estima-tion of the shooting distance and also establishing, whethera person has used a firearm. Fulfilling these tasks is veryhelpful for the reconstruction of an investigated crime. How-ever, new challenges for gunshot residues examiners arise.With a growing frequency the administration of justice asksabout the type of ammunition, and so the firearm used in

G. Zadora, Z. Brozek-Mucha / Materials Chemistry and Physics 81 (2003) 345–348 347

cases when the only accessible for examinations evidenceare gunshot residues. Thus one more task can be formulated,i.e., the identification of an ammunition from the gunshotresidues detected.

One of the attempts to solve this problem is a systematicstudy of gunshot residues, originating from various ammu-nition types and statistical and chemometric evaluation ofthe analytical data[4,5].

The frequency of occurrence of primer residues, contain-ing a combination of lead, antimony and barium, was takeninto account. The performed cluster analysis[6,7] utilisingthese features showed a promising tendency of grouping am-

Fig. 3. A glass classification scheme—a non-statistical approach: c, car windows; ecb, external glass of car bulbs; eob, external glass of ordinary lightbulbs; h, car headlamps; icb, internal glass of car bulbs; iob, internal glass of ordinary light bulbs; o, optic glass; p, container glass; t, tableware; w,window panes.

munitions according to their types (Fig. 2). This can con-tribute to at least a group identification of an ammunition incases, when the only evidence accessible for investigationsare gunshot residues collected at the scene of crime.

2.2. Glass

The importance of glass as evidence was recognisedmany years ago. As it was mentioned before, glass as the ev-idence material often occurs in very small quantities. Thus,investigations of glass samples require sensitive analyticalmethods providing satisfactory results from small amounts

348 G. Zadora, Z. Brozek-Mucha / Materials Chemistry and Physics 81 (2003) 345–348

of the examined material such us the quantitative elementalanalysis using an SEM–EDX method. The elemental com-position of glass strongly depends on the properties of glassproducts. Taking into account the difference in the elemen-tal contents of samples collected from various groups ofglass objects determined by means of SEM–EDX method,a scheme of classification and discrimination of glass wasobtained at our institute[7–9].

Inspection of the obtained results allowed to find elements,whose ranges of concentration in the considered groups didnot overlap. The effect of such a non-statistical approach tothe problem of classification of glass samples collected inPoland, is presented inFig. 3.

Only sets 3–5 were homogenous. The classification of ob-jects in sets 6 and 7 was not possible with the non-statisticalapproach. The cluster analysis was used to solve the problemof classification in the case of sets 6 and 7. As a result a clas-sification scheme was created that made it possible to clas-sify most of the glass microtraces examined by means of theelemental analysis with SEM–EDX method[8]. Moreover,a differentiation of glass objects revealing the same quali-tative elemental composition (determined with SEM–EDXmethod) can be achieved using the presented approach basedupon a “cord distance” and Student’st-test[9].

2.3. Car paint

The paint coat of a car body consists of a number ofsuccessively overlaid paint layers. These layers differ fromeach other in terms of their ingredients, i.e., resin, pigmentsand fillers. The number of layers making up a car coveringdepends on its type. In brand new cars and in those that havenot been repainted there are only three to four layers. Paintcoverings of renovated cars consist of a larger number oflayers (sometimes even more than a dozen), including notonly enamels, but also putties, painters’ putties and groundundercoats.

In identification and comparative studies of paint chips,scientists define their macroscopic properties—colour, shade

and texture—and their microscopic properties relating totheir morphology (the number and sequence of layers, theirthickness and colour). The next stage is a detailed analy-sis of the chemical content of each layer, including identi-fication of the binder, pigments and fillers. SEM–EDX canbe very useful in the case, when the compared paint sam-ples are similar in the microscopic properties relating totheir morphology (the number and sequence of layers, theirsthickness, colour) and the results of infra-red spectroscopyanalysis. Thus, the identification of a particular paint layercan be carried out comparing the contents of elements sincethey are characteristic for a given layer and do not repeat[10].

3. Conclusions

The usefulness of SEM–EDX methods in solving prob-lems of analysis of microtraces for forensic purposes wasconfirmed in the course of international tests, e.g. “Foren-sic Testing Program” (Collaborative Testing Services,USA). SEM–EDX can be successfully used for foren-sic purposes since sample preparation is simple and non-destructive.

References

[1] P. Burke, C.J. Curry, L.M. Dawies, D.R. Cousins, Forensic Sci. Int.28 (1985) 201.

[2] J.A. Buscaglia, Anal. Chim. Acta 288 (1994) 17.[3] H. Meng, B. Caddy, J. Forensic Sci. 42 (1997) 553.[4] Z. Brozek-Mucha, A. Jankowicz, Forensic Sci. Int. 123 (2001) 39.[5] Z. Brozek-Mucha, G. Zadora, Prob. Forensic Sci. XLVI (2001) 281.[6] B.S. Everitt, Cluster Analysis, Arnold, Oxford, 1993.[7] G. Zadora, Z. Brozek-Mucha, Prob. Forensic Sci. XL (1999) 33.[8] G. Zadora, Z. Brozek-Mucha, A. Parczewski, Prob. Forensic Sci.

XLVII (2001) 137.[9] G. Zadora, Z. Brozek-Mucha, A. Parczewski, Prob. Forensic Sci.

vXLVII (2001) 144.[10] J. Nieznanska, J. Zieba-Palus, P. Koscielniak, Prob. Forensic Sci.

XXXIX (1999) 77.