Unambiguous Characterization of Gunshot Residue Particles Using

Feb 1, 2012 - detection and identification of gunshot residue (GSR) particles from firearms discharges has been developed. Tape lifts were used to...
1 downloads 0 Views 1MB Size
Article pubs.acs.org/ac

Unambiguous Characterization of Gunshot Residue Particles Using Scanning Laser Ablation and Inductively Coupled Plasma-Mass Spectrometry Zuriñe Abrego,† Ana Ugarte,† Nora Unceta,† Alberto Fernández-Isla,‡ M. Aranzazu Goicolea,† and Ramón J. Barrio*,† †

Department of Analytical Chemistry, Faculty of Pharmacy, University of the Basque Country, UPV/EHU, Paseo de la Universidad 7, 01006 Vitoria-Gasteiz, Spain ‡ Scientific Police Laboratory, Ertzaintza, Department of the Interior, Basque Government, Erandio, Spain S Supporting Information *

ABSTRACT: A new method based on scanning laser ablation and inductively coupled plasma-mass spectrometry (LA-ICPMS) for the detection and identification of gunshot residue (GSR) particles from firearms discharges has been developed. Tape lifts were used to collect inorganic residues from skin surfaces. The laser ablation pattern and ICPMS conditions were optimized for the detection of metals present in GSR, such as 121Sb, 137Ba, and 208Pb. Other isotopes (27Al, 29Si, 31P, 33S, 35Cl, 39K, 44Ca, 57Fe, 60Ni, 63Cu, 66Zn, and 118Sn) were monitored during the ICPMS analyses to obtain additional information to possibly classify the GSR particles as either characteristic of GSR or consistent with GSR. In experiments with real samples, different firearms, calibers, and ammunitions were used. The performed method evaluation confirms that the developed methodology can be used as an alternative to the standard scanning electron microscopy - energy dispersive spectroscopy (SEMEDS) technique, with the significant advantage of drastically reducing the analysis time to less than 66 min.

T

accurate method for confirming the presence of GSR is from its elemental content.4 Although the criterion has changed over time,1 the ENFSI (European Network of Forensic Science Institute) currently considers spheroid particles between 0.5 and 5.0 μm in diameter with a composition “characteristic of GSR” to have the following elemental profile:4 type 1. Pb−Sb− Ba (particles with this composition may only contain one or several of the following other elements: Si, Ca, Al, Cu, and trace amounts of Fe, S, P, Zn, Ni -in conjunction with Cu and Zn-, K, Cl, and Sn). According to the ENFSI, particles that have a composition “consistent with GSR”, but not unique to GSR, will have one of the following elemental profiles: type 2. Ba− Ca−Si; type 3. Sb−Ba (with no more than a trace of Fe and/or S); type 4. Pb−Sb; type 5. Ba−Al; type 6. Pb−Ba; type 7. Pb; type 8. Ba; or type 9. Sb. Morphologically, the majority of GSR consist of spheroidal particles. Since morphology is dependent upon conditions at the time of impact and the distance from point of production to point of impact, it can vary greatly and should therefore be considered only as a secondary criterion for identification of GSR. The 2010 ASTM standard9 uses an identical classification, which will be used for this study.

he detection and identification of gunshot residues (GSR) from firearms discharges provide valuable investigative information.1,2 According to the theory of GSR particle origin, most GSR leave the firearm after the discharge in the form of a gas. This gas gradually condenses, and individual particles sediment around the discharged firearm, including on the shooter’s hand.3 The majority of GSR particles are typically in the range of 0.5 to 5.0 μm in diameter.4 The information obtained from analyzing GSR is generally used to determine whether a person discharged a firearm and therefore to confirm an alibi, arrest suspects, differentiate homicide from suicide1,5 and to both determine a bullet entrance hole and estimate the firing distance.6 Although organic GSR have been analyzed,7,8 the majority of work performed in this field has focused on the inorganic components of the GSR. Organic compounds mainly originate from propellant and firearm lubricants, taking the form of unburned and partially burned gunpowder particles, some products of their transformation, and hydrocarbons. Inorganic residues such as nitrates, nitrites, and metallic particles originate from the primer and propellant as well as the cartridge case, the projectile jacket, or its core and from the weapon barrel itself.8 Antimony, lead, and barium, derived from antimony sulfide, lead styphnate, and barium nitrate, respectively, from the primer, are the primary elements present in GSR. The most © 2012 American Chemical Society

Received: November 30, 2011 Accepted: February 1, 2012 Published: February 1, 2012 2402

dx.doi.org/10.1021/ac203155r | Anal. Chem. 2012, 84, 2402−2409

Analytical Chemistry

Article

constituents, which are transported to the ICPMS by means of a gas (usually Ar) stream.24 Scanning LA-ICPMS is a technique widely used to determine elements directly from solid samples with minimal sample preparation.25−27 The term “scanning” refers to the uniform translation of the laser spot relative to the sample surface during continuous ablation. Although LAICPMS is increasingly being used in the specialized forensic laboratories,28,29 there are no references of LA-ICPMS based methods for gunshot residue studies in the literature. As important as the employed analytical instrumentation for GSR analysis, sample collection is even more important because it could affect the subsequent analytical determination. In general, tape lifts have proved to be an efficient particle-lifting method.8,19,30 In recent years, devices designed for SEM-EDS based on stubs with carbon adhesive tape have became popular because they allow for the sensitive collection of GSR using an economical and uncomplicated kit. With these devices, particles suspected of containing GSR are collected by dabbing a carbon adhesive tape mounted onto an aluminum stub across the surface of the sample. The primary objective of the present study has been to develop a novel alternate approach to GSR analysis that combines ease and rapidity to provide information about the chemical composition of individual GSR particles. We present a new methodology for the analysis of GSR on the shooter’s hands based on scanning LA-ICPMS, which is a simple and straightforward procedure that identifies GSR on tape-lift collection devices by their elemental content. Four different weapons (two pistols and two revolvers) with four different lead-based primer ammunitions were investigated to verify the usefulness of the scanning LA-ICPMS technique for GSR analysis.

The amount of these elements on the hands of the shooter depends on several factors, such as the type of weapon and munitions used; the age and condition of the weapon; the suspect’s personal hygiene, occupational habits and biometrics (hand measurements); and the time elapsed since firing.10 However, the influence of many of these factors has not been investigated in the present study because the primary objective is to present a new technique for the characterization of GSR. Several techniques have been used over the years to determine these metals, although scanning electron microscopy coupled to energy dispersive spectroscopy (SEM-EDS) is the standard technique used for the analysis and identification of GSR.11 Neutron activation analysis (NAA) has been used since the 1960s, although it requires a nuclear reactor and it is not applicable to lead.12 In 1971, a method was reported for the analysis of GSR by atomic absorption spectroscopy (AAS);13 however, the conventional AAS used only offers single element analysis and was found to be inadequate for Ba and Sb at the levels encountered in hand samples. Many other techniques have been used for GSR analysis, such as adsorptive stripping voltammetry (AdSV),14 inductively coupled plasma atomic emission spectroscopy (ICP-AES),15 inductively coupled plasma mass spectrometry (ICPMS),16 and X-ray microfluorescence.17 Recently highly photoluminescent metal− organic frameworks (MOFs), also known as coordination polymers, in ammunition have been used as markers subjected to irradiation under an UV lamp, to allow for the visual detection of GSR from lead-free ammunition.18 Many of the above-mentioned standard methods (AAS, NAA, AdSV, ICPAES, ICPMS) require time-consuming digestion or extraction of the sampled material prior to determining the total element concentrations. When used for GSR analysis, these techniques provide information about the total composition of the elements in the sample. In contrast to particle analysis, these bulk sample methods sacrifice individual particle identification, which impedes the classification of the GSR using the terminology of the field (as “characteristic” or “consistent” particles), as specific data about the elemental composition of the individual particles are not available. Since the mid-1970s, SEM-EDS has been utilized as a method to identify the presence of GSR by combining information on the morphology and the chemical composition of individual particles. However, SEM-EDS needs to locate individual GSR, and, even with automated search systems, the analysis can take up to several hours per square centimeter of the sample, depending on the surface density of particles with a high average atomic number.19 These automated SEM-EDS systems are available for the detection of GSR using either custom-made or commercial software.20,21 Furthermore, the automated systems may generate false positives, and the interpretation of “compatible” GSR particles should be conducted case-by-case. In addition, because GSR particles as large as 5 μm can initially be hidden by skin debris on the adhesive surface of the sampler used in SEM-EDS,22,23 treatment of the sampler with sodium hypochlorite is, in some cases, recommended to completely digest the organic skin debris. In contrast, combining ICPMS combined with the direct solid-sample introduction technique of laser ablation (LA) requires minimal sample preparation. Conceptually, LA is a simple and straightforward process. A short-pulsed high power laser is focused onto the sample surface and converts a finite volume of the solid sample instantaneously to its vaporized



EXPERIMENTAL SECTION Instrumentation. For the LA-ICPMS measurements a UP266 laser ablation system from New Wave Research Co. (Huntingdon, UK) connected to an Agilent 7500ce ICP mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) was used. The pulsed Nd:YAG laser operated in the Q-switch mode with a wavelength of 266 nm, pulse duration of 4 ns, and a maximum energy of 4 mJ (fluence: 6.97−7.88 mJ cm−2). Typical operating conditions are given in Table S-1 (Supporting Information). The laser-produced aerosol was transported through a 1 m long polyurethane tube (i.d. of 4 mm) by a helium gas stream (0.44 L min−1) and then mixed via a Y-connector with a liquid aerosol (nebulized by means of a microconcentric nebulizer) before entering the plasma. Helium was used as the carrier gas due to its good characteristics as an ablation medium and enhanced transport efficiency for the ablated material.31 The Y-connector was specially designed and manufactured for the physical requirements of our ICPMS system and is shown in Figure S-1 (Supporting Information). This dual-flow introduction system enabled the complete and simple optimization of the LA-ICPMS coupling when nebulizing 1 ng g−1 solution of the elements employed for tuning. Furthermore, during the laser ablation analyses, the plasma was kept under wet conditions by the continuously nebulizing a 1 ng g−1 thallium standard solution, which was also used to monitor the stability of the plasma. To investigate the effectiveness of LA-ICPMS to identify the GSR particles, different scanning-ablation patterns on the surface of the tape-lifted device were studied. The total length of the ablated lines varied from 80 mm (raster ablation) to 2403

dx.doi.org/10.1021/ac203155r | Anal. Chem. 2012, 84, 2402−2409

Analytical Chemistry

Article

51.8 mm (ablation “in concentric circles”) or 36 mm (ablation in “W”). During ablation, the laser was focused on the sample surface with a spot size diameter of 160 μm and operated with a continuous sample translation speed of 20 μm s−1. A helium background measurement was systematically collected before each analysis for the 40 s prior to ablation. Seventeen isotopes corresponding to 15 analytes, 13C, and 205 Tl were qualitatively analyzed. The analytes (27Al, 29Si, 31P, 33 S, 35Cl, 39K, 44Ca, 57Fe, 60Ni, 63Cu, 66Zn, 118Sn, 121Sb, 137Ba, and 208Pb) were chosen because they are the elements that the ENFSI Expert Working Group Firearms/GSR considered to be present in particles either characteristic or consistent with GSR.4 These elements may originate from the cartridge case (Al, Sb, Cu, Ni, P), primer (Al, Ba, Ca, Cu, Ni, Sn, Zn, Si, Sb, S, K, Cl), bullet including jacket (Cu, Ni), or the barrel of the firearm (Fe).8The intensities of these isotopes of interest were systematically normalized against the 13 C signal after subtracting out the mean background signal. Normalization is essential to LA-ICPMS to compensate for variations in the ablation yield because of laser energy drift and sample density. 13 C was selected as the normalizing element because it is homogeneously distributed throughout the polymer of the adhesive matrix. As ablation of the adhesive tape evaporated of the organic compounds in the adhesive, which would eventually be deposited on the inside surface of the sample chamber window, the window was carefully cleaned after every three samples. The cleaning procedure was as follows: first the window was gently wiped with a lint-free cloth wet with 5% nitric acid, then rinsed with water, and finally dried with a clean tissue. The reference method in the skin test analysis was based on SEM-EDS. For this, a JEOL JSM-7000-F (Tokyo, Japan) scanning electron microscope equipped with a EDS detector Energy 350 (Oxford Instruments INCA, Oxfordshire, UK) was used. This equipment provides a proprietary software that allows the automation of scanning electron beam and plate, using the specialized module for GSR. The measurements were performed at 20 kV and using a current of 2 nA. Before the analysis, the samples were carbon-coated to facilitate a conductive surface. Particles candidates for GSR were detected because of its contrast with the background in backscattered electrons image (BEI) taken at 400 X. Subsequently an EDS spectrum of each particle was performed. For each sample, at least 99% of the surface was scanned. Firearms and Ammunition for the Live Firing Tests. The proposed methodology was evaluated using four different weapons. Firearm 1 was a 9 mm Heckler & Koch pistol model USP Compact (Oberndorf/Neckar, Germany) using 9 mm Samson low velocity ammunition (Whately, MA, USA); firearm 2 was a 38 mm Arminius revolver (Hermann Weihrauch Revolver GmbH, Mellrichstadt, Germany) using Aguila 38 SPL-PH+P semijacketed ammunition; firearm 3 was a 22 mm Arminius revolver using Winchester Super-X 22 ammunition (East Alton, IL, USA); and firearm 4 was a Sig-Sauer pistol model 45 Auto (Exeter, NH, USA) using Aguila 45 auto jacketed ammunition (Cuernavaca, México). These firearms were selected because firearm 1 is among the most commonly used by the police forces in Spain, the others came from seizures in criminal acts and they all cover a wide range of calibers. All cartridges came from individual and unopened ammunition boxes. Test shots were undertaken on the test

firing field at the Scientific Police Laboratory (Ertzaintza, Erandio, Basque Country, Spain). Three police volunteers were selected and asked to fire, with their own technique, as many times as required for each experiment. Before the test shots, each firearm was carefully cleaned with a cotton cloth to eliminate any possibility of contamination from previous discharges, and during each shooting sequence, the butt of the gun was protected by covering it with a film (Parafilm M, Pechiney Plastic Packaging Company, Chicago, IL, USA) to eliminate the possibility of transferring material between tests and to avoid contamination from previous shots with the same firearm. For the initial experiments and method optimization, the shooters fired with both hands protected by disposable rubber gloves, and the sampling was performed on the gloved hand, to avoid both permanent skin contamination and carry over effects between experiments. Once the method was optimized, the shootings for later experiments were performed without the rubber gloves, to simulate a more realistic situation. Care was taken by the shooters not to rub or contact other surfaces with their hand during the analysis. Sample Collection. Tape lift kits of 12.7 mm diameter designed for SEM-EDS analysis of gunshot residue (Adhesive Lifts GRA 200, Sirchie Finger Print Laboratories, Youngsville, NC, USA) were used to retrieve the GSR from the shooter’s hands. The collection device consisted of a plastic vial with a tightly fitting cap. The metal stubs in the caps were equipped with a carbon adhesive tape to lift residue. Sample collection was straightforward: simply, the cap containing the carboncovered metal stub was removed, applied to an area on the hand in question, and then returned to its original location in the vial. The number of sampler dabs applied to the hand of a suspect by an officer rarely exceeds 40. For this study 20 dabs were considered adequate, as had been suggested by other authors.32 All of the samples were collected within 5 min of discharging the firearm by repeatedly stubbing separately the back and palm of the shooters’ hand. Although this period of time is not realistic for police departments/forensic scientists, it was chosen to ensure no cross-contamination on the optimization of the analytical method. Once the samples were taken, the stub was covered and clearly labeled on the side to distinguish each sample from the others without compromising the analysis. Finally, the sampler was placed in the ablation cell for laser ablation. Two blank tests were used: one sampling stub applied to the gloved hand of a nonshooter, and another applied to the ungloved hand of the same person.



RESULTS AND DISCUSSION Optimization of Experimental LA-ICPMS Parameters. The helium carrier gas and argon nebulizer gas flows were optimized to improve the sensitivity, reduce the number of doubly charged ions produced, and minimize the fractionation effects. Elemental fractionation, considered to be one of the primary limitations of LA-ICPMS, is the occurrence of nonstoichiometric effects in the transient signals, during which the composition of the analyzed mass is not representative of the bulk sample.33 The optimization procedure was performed using a glass standard reference material (NIST SRM 612), in which the elements U and Th are present at the same concentration level. As both elements have similar ionization potentials, mass-to-charge ratios and most abundant isotope concentrations above 99%, the ratio of their 2404

dx.doi.org/10.1021/ac203155r | Anal. Chem. 2012, 84, 2402−2409

Analytical Chemistry

Article

of the helium flow rate on fractionation was weaker, and the sensitivity decreased slightly at rates above 0.48 L min−1. The conditions with the highest sensitivity and lowest fractionation (with a Th/U ratio approximately 92%) were obtained using flow rates of 0.75 L min−1 for argon and 0.44 L min−1 for helium. The laser energy was optimized to improve the sensitivity with low signal instability from ablating the carbon adhesive in a new sampling stub and monitoring the 13C signal. As can be seen in Figure 1, a good sensitivity and precision were obtained at 55% energy density, so this energy density was selected for the experiments. Higher energy densities, of 70 and 75%, resulted in increased sensitivity, but these conditions increased signal instability as well. GSR Identification Criteria. The primary criterion for identifying a GSR particle using the proposed LA-ICPMS method is the detection, at identical ablation times, of signals corresponding to those of Pb, Sb, and Ba isotopes, which represent the elemental profile of particles characteristic of GSR. Therefore, the ion-time responses showing coincident Pb, Sb, and Ba peaks are considered to be particles characteristic of GSR (Figure 2). As mentioned above, other isotopes have also

Figure 1. Influence of laser output energy in sensitivity and precision (indicated by 1σ error bars). Laser scanning conditions on the surface of the sampling stub: scan speed: 20 μm s−1, repetition rate: 10 Hz, spot size: 95 μm, integration time: 50 ms. Each point is the average of 50 acquisitions.

intensities is used to monitorize the fractionation. Consequently, if a complete atomization is achieved, this ratio should be close to one. Helium flows between 0.4 and 0.6 L min−1 and argon flows between 0.5 and 1 L min−1 were tested. It was observed that the sensitivity for argon flows below 0.6 L min−1 was very low, whereas flows above 0.9 L min−1 increased fractionation and yielded Th/U ratios below 76%. The influence

Figure 2. Characterization of eleven GSR particles (GSR1 to GSR11). The colored vertical bands show abundance peaks of the different metals at the same ablation time. The ablated stub comes from a sample of the hand of a shooter who has made a shot with firearm 1 (9 mm pistol) using 9 mm low velocity ammunition (nonjacketed bullet). 2405

dx.doi.org/10.1021/ac203155r | Anal. Chem. 2012, 84, 2402−2409

Analytical Chemistry

Article

GSR particles have a density34 of approximately 10000 kg m , because the particle mass and size are related to the density ρ, it is useful to write eq I in terms of the particle radius r −3

r=3

3 ∑i m 4πρ

(I)

where ∑im is the sum of the masses of each isotope Pb−Sb−Ba (including other possible majority elements as Sn and Cu) obtained using LA-ICPMS. In the absence of a reference material to be used as a calibrant based on the known concentrations of the target analytes deposited upon tape lifting, another means of estimating the lower limit of quantification related to the particle size is necessary. To this end, it was decided to establish a criterion based on the minimum signal necessary to positively identify each of the isotopes. This minimum signal was defined as 10 times the mean background signal of the corresponding blank sampling stub normalized to the 13C signal. This information is related to the minimum GSR particle size that can be measured using the LA-ICPMS technique. While 0.5 μm diameter particles are considered to be the smallest identifiable by SEM-EDS, in the present study, a particle that generates a signal 10 times the background for each isotope is likely to be characterized as a GSR. Although a less demanding (5 times the background) criteria would significantly increase the number of detected particles (approximately 15−20%), a more restrictive criteria was preferred to obtain results with greater confidence. Scanning Laser Ablation Pattern. The scanning laser ablation pattern is a fundamental parameter to achieve an effective GSR analysis protocol by LA-ICPMS because it must guarantee GSR particle detection in a positive sample. Different ablation patterns were studied, all of which were based on mapping: LA pattern 1 (raster), LA pattern 2 (ablation in W), and LA pattern 3 (concentric circles). Figure S-2 (Supporting Information) shows the aspect of three sample stubs after ablation. Each ablation pattern took a fixed analysis time (pattern 1: 66 min, pattern 2: 30 min, pattern 3: 43 min) to obtain, and the patterns were obtained for all of the firearms investigated after a series of 6 shots to characterize a significant number of GSR particles. As shown in Figure 3A, the raster pattern was the most appropriate, allowing the identification of a larger number of characteristic GSR particles, with all four types of firearms investigated. With a raster pattern, 12.8 mm2 of the stub surface was ablated (10.1% of the total surface), while other patterns ablated 4.5 and 6.5% of the surface (5.76 mm2 with W ablation, and 8.29 mm2 with the concentric circles). Even taking this into account, the relative number of particles detected by raster ablation was significantly higher, and this improvement was not directly linked to the ablated surface or the ablation time. Once the ablation pattern was selected, all the subsequent experiments were performed using a raster ablation. “Characteristic” and “Consistent” GSR. There is no consensus among forensic laboratories about the number of particles required to confirm the presence of GSR,19 and values ranging from 1 to 10 have been suggested. There is significantly more agreement about the wording of reports when a laboratory has decided a result is positive. In such a case, the statement: “the sample is consistent with the suspect having discharged a firearm, having been in the vicinity of a firearm when it was discharged, or having handled an item with a GSR

Figure 3. A. Mean number of GSR particles found in the four types of firearms investigated with three different ablation patterns: raster (pattern 1), ablation in W (pattern 2), and concentric circles (pattern 3). (n = 3, error bars represent 95% confidence intervals). B. Percentage of other elements in the particles characteristic of GSR (Pb−Sb−Ba) with the firearms used. 803 particles characteristic of GSR have been located in the study summarized in Table 2. C. The mean percentage of particles consistent with GSR according to the ENFSI criteria. Base of the study: 72 tape-lifts analyzed in the different trials, with a total of 803 particles characteristic of GSR and a total of 608 particles consistent with GSR.

been monitored using the same criterion (coincident peaks in ion-time responses) to identify particles consistent with GSR. In the case of GSR analysis by LA-ICPMS, the quantitative determination of each of the elements in the GSR particle was not necessary. It would be interesting to have quantitative data related to the concentration of each element in the particles along with their size, assuming the Pb, Ba, and Sb concentration are related to particle size. Therefore, the limit of quantification for this technique related to the particle size could be established, which would enable results corresponding small particles to be considered as positive, as with SEM-EDS techniques. In LA-ICPMS, however, obtaining quantitative results is conditional upon by the availability of an internal standard element whose concentration in both the sample and reference material is known. 2406

dx.doi.org/10.1021/ac203155r | Anal. Chem. 2012, 84, 2402−2409

Analytical Chemistry

Article

Table 1. Number of Particles Characteristic of GSR (Type 1) and Consistent with GSR (Types 2−9) Found Using Raster Pattern for Ablationa 1 shot (n = 3) back of hand Particles Characteristic of GSR Found firearm 1 16 ± 9 firearm 2 5±1 firearm 3 7±2 firearm 4 20 ± 4 firearm firearm firearm firearm a

1 2 3 4

1 17 1 8

± ± ± ±

1 4 1 1

2 shots (n = 3) palm 3 3 5 4

± ± ± ±

9 18 19 8

2 1 3 1

± ± ± ±

back of hand 32 ± 14 11 ± 1 10 ± 3 25 ± 6 Particles Consistent with GSR Found 8 6 3 7

4 8 8 12

± ± ± ±

4 3 2 3

6 shots (n = 3) palm 8 1 3 4

± ± ± ±

4 6 7 7

back of hand

1 1 3 1

± ± ± ±

2 3 3 1

37 14 13 30

± ± ± ±

14 4 2 9

4 8 9 5

± ± ± ±

4 3 5 3

palm 7 3 3 6 6 8 5 4

± ± ± ±

± ± ± ±

1 1 2 4

4 2 5 3

Uncertainty given as 95% confidence intervals.

on it” resembles the phrasing used by most laboratories.19,35 In this paper, the detection of coincident signals corresponding to the Pb, Sb, and Ba isotopes was related to the presence of a particle characteristic of GSR. Under the selected scanning laser ablation conditions, the characteristic signal-time profile of a GSR sample obtained by LA-ICPMS is shown in Figure 2. It can be observed, as the sample stub surface is ablated, that each of the analytes yields a constant background signal. When the laser ablates a GSR particle, Pb, Sb, and Ba peaks are observed at identical ablation times. Table 1 shows the results obtained for the analyses of samples from 1, 2, or 6 shots from the four firearms investigated. As can be observed, even in the limiting case where the samples were obtained after a single shot, the number of particles characteristics of GSR found was significant (>5) enough to provide evidence of the shooting. The relationship between the signal abundance, expressed as the peak heights, for the three characteristic elements (Pb, Sb, and Ba) does not remain constant for each of the firearm assayed. This observation confirms that it is difficult to correlate the composition of each GSR particle with the firearm and/or ammunition used,16 although recently it has been shown by chemometric methods that it is possible to differentiate between primers from different ammunition brands.36 The results shown in Table 1, which concerns the type of firearms, confirm those of other authors19 working with SEMEDS. Pistols generate a much larger number of GSR than the revolvers. However, the number of GSR particles is always significantly higher on the back of the hand than on the palm, as has been repeatedly demonstrated.37 It is well-known to be because the palm is in direct contact with the firearm and only a very small number of GSR particles can transfer to this part of the hand. Although Pb, Sb, and Ba were the three key elements because of the definition of characteristic GSR, other elements were also monitored: 27Al, 29Si, 31P, 33S, 35Cl, 39K, 44Ca, 57Fe, 60Ni, 63Cu, 66 Zn, and 118Sn. The responses obtained for these elements found at the same time of ablation were generally lower than for Pb, Sb, and Ba. Figure 3B shows the presence of other elements in the characteristic particles (Pb−Sb−Ba) with each of the firearms used. The presence of Sn with a weighted average of 38.7% of the characteristic GSR particles was noteworthy for its statistical significance. The metal next in importance was Cu, with 13.5% of mean presence in these

positive samples, which was undoubtedly related to the type of firearm and ammunition used.38 The possibility of detecting Pb−Sb−Ba aggregates from nonfirearm sources has been confirmed by the tests performed on brake pads,39 fireworks, and automobile workers.40 Exposure to sources of potential secondary GSR contamination can occur during the arrest process.41 In the same way, results from the examinations of the number of particles and their chemical proportions revealed a dependence on the distance from the gun muzzle.42 Information on the presence of other elements other than Pb, Sb, and Ba in characteristic GSR particles could settle these questions, but this is obviously an area that exceeds the current study and should be undertaken in other investigations. Similarly, it is obvious that the environmentally friendly lead-free cartridges are beginning to have certain popularity and have become fashionable throughout the world. The GSR particles associated with these types of cartridges can also be analyzed by the proposed method; however, the authors suggest that forensic scientists responsible for routine GSR detection and/or firearms examination should complete these studies. As shown in Figure 3C, type 2 (Ba−Ca−Si) and type 5 (Al− Ba) particles consistent with GSR have not been found in any of the samples analyzed during this assay (with 72 tape-lifts). The highest percentage of particles consistent with GSR correspond those where Pb was the only element (type 7; 28.8%), followed for those composed of Pb−Sb (type 4; 27.9%) or Sb−Ba (type 6; 15.5%). To summarize, it is worth mentioning that a total of 803 particles characteristic of GSR and a total of 608 particles consistent with GSR have been located after analyzing 72 tapelift samples. Skin Tests. Once the validity of the LA-ICPMS method for the characterization of GSR was verified and a study on the nature and composition of these particles had been conducted, several trials were conducted under normal shooting conditions, without using gloves on the hand of the shooter. Sampling by tape-lift and the same protocols for the shooting described above were followed. The results are shown in Table 2. Some authors22,23 have identified the possibility that using the tape-lift method may result in some skin cells obscuring the GSR particles. These particles, completely covered by skin cells, cannot be detected by SEM-EDS, as they lose their morphological information and cannot be located by the image recognition software (SEM provides information only on 2407

dx.doi.org/10.1021/ac203155r | Anal. Chem. 2012, 84, 2402−2409

Analytical Chemistry

Article

Table 2. GSR Particles Found in Samples Taken Directly from the Skin of Shooters after Firing a Single Shot with a 9 mm Pistol (Firearm 1)a palm “characteristic” elemental profile Pb−Sb−Ba Pb Ba Sb Pb, Sb Pb, Ba

LA-ICPMS

SEM-EDS

3 ± 3 (1−4)

8 ± 6 (2−11)

back of hand “consistent”

LA-ICPMS 5 ± 3 (2−8) 2 ± 2 (0−4) 1 ± 1 (1−1) -

“characteristic”

SEM-EDS

LA-ICPMS

SEM-EDS

20 ± 7 (10−25)

32 ± 15 (23−48)

22 ± 8 (18−30) 3 ± 3 (2−6) 1 ± 1 (0−2) 13 ± 2 (12−14) 5 ± 3 (3−8)

“consistent” LA-ICPMS 3 1 2 2

± 3 (2−6) ± 1 (0−2) ± 3 (0−5) ± 3 (0−6)

SEM-EDS 4 ± 4 (2−9) 4 ± 4 (2−8) 1 ± 1 (0−1)

a

Five different shooters fired the same pistol (n = 5) in two sessions, sampling was performed identically for LA-ICPMS and SEM-EDS. Uncertainty given as 95% confidence intervals. In brackets is the range in number of GSR particles detected. A correction factor of 10:1 to the number of GSR particles localized by SEM-EDS has been applied, due to different values of the analyzed surface.

technique (SEM-EDS), which enables high throughput sample analysis. Similarly, the potential problem of concealing GSR particles behind epithelial cells attached to the tapes was removed. Scanning LA-ICPMS contributes little to the sphericity requirement for the GSR particles over SEM-EDS analyses. Nevertheless, it is worth mentioning that the chemical composition and abundance of the particles constitute sufficient evidence for a shot having been fired and are far more conclusive than the evidence based on bulk analyses. This method demonstrated both good specificity and precision and was capable of detecting changes in the GSR profiles of different samples. The sensitivity criterion used was based on the abundances normalized to the 13C signal for each of the 15 monitored isotopes. The results were fully comparable to those from SEM-EDS for both type of firearm used and the side of the hand sampled. The ability to generate qualitative profiles for each particle with scanning LA-ICPMS offers a streamlined strategy for characterizing of GSR and reducing uncertainty in courts and forensic laboratories with a single analytical technique.

the top layer). To avoid this phenomenon, various procedures have been proposed for both destroying any skin cells attached to the tape-lift23 and applying new systems, such as backscattered electron imaging (BEI), to the scanning electron microscope.22 Unlike GSR analysis by SEM-EDS using a tape-lift for sampling, the use of LA allows for the destruction of the epidermal cells because of the penetrating power of the laser. Depending on the ablated material, the energy density (fluence), and the ablation rate, the profiling depth43 can vary from 2 to 10 μm, which is sufficient for the epidermal cells to be destroyed without further treatment. The results shown in Table 2 refer to the sampling carried out on both the palm and the back of the hand after firing a single shot from a 9 mm gun. With respect to the analysis of samples from shooters who wore gloves, the number of positive results on the back of the hand was statistically similar. Similarly, the number of particles identified as consistent with GSR followed the same trend as the case of shooters who wore gloves. The skin tests performed using ten volunteers without any prior contact with firearms were negative in terms of both characteristic and consistent GSR. This again confirms the reliability of the method used. To test the effectiveness of the developed method a set of samples were also analyzed by SEM-EDS, using the same sampling protocol, after firing a single shot from a 9 mm gun. The automatic particle recognition software used by SEM scans the 99% of the stub surface, while the ablated surface using the scanning LA raster pattern is only 10%. For this, the results shown in Table 2 apply a correction factor of 10:1 to the number of GSR particles localized by SEM-EDS. As shown in Table 2 the results obtained by both techniques are similar, including the greater number of GSR particles on the back of the hand than on the palm. The tendency of the SEM-EDS method to locate a greater number of particles in each case may be due to the fact that given the large number of particles, not all of those located by the SEM automatic recognition software can be confirmed by EDS. In the subsequent case-by-case analysis, some of these particles do not reveal a composition compatible with GSR and should be discarded.1



ASSOCIATED CONTENT

S Supporting Information *

Additional figures as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Fax: +34 945014351. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors gratefully acknowledge the Department of the Interior of the Basque Government for use of the facilities afforded by the test firing field at the Scientific Police Laboratory. The authors also acknowledge the technical and human support provided by Alava Central Service of Analysis, and Electronic Microscopy and Material Microanalysis Service, SGIker (UPV/EHU, MICINN, GV/EJ, ESF). The work was funded by the Basque Government, Research Groups of the Basque University System (Project N°. IT338-10).



CONCLUSIONS The developed scanning LA-ICPMS method allows the characterization GSR particles and has several benefits compared to the established method. The use of a raster scanning ablation pattern of the tape-lifts allows for very competitive times, 66 min, relative to those of the reference



REFERENCES

(1) Saverio Romolo, F.; Margot, P. Forensic Sci. Int. 2001, 119, 195− 211. 2408

dx.doi.org/10.1021/ac203155r | Anal. Chem. 2012, 84, 2402−2409

Analytical Chemistry

Article

(2) DiMaio, V. J. M. Gunshot Wounds: Practical Aspects of Firearms, Ballistics, and Forensic Techniques; CRC Press: Boca Raton, FL, 1999; p 424. (3) Schwoeble, A. J.; Exline, D. L. Current methods in forensic gunshot residue analysis; CRC: Boca Raton, FL, 2000; p 192. (4) Niewoehner, L. ENFSI−Guide for gunshot residue analysis by scanning electron microscopy/energy-dispersive X-ray spectrometry, ENFSI Working Group Firearms; ENFSI: Prague, 2008. (5) Meng, H.-H.; Lee, H.-C. Forensic Sci. J. 2007, 6, 39−54. (6) Lichtenberg, W. Forensic Sci. Rev. 1990, 2, 37−62. (7) Northrop, D. M.; Martire, D. E.; MacCrehan, W. A. Anal. Chem. 1991, 63, 1038−1042. (8) Dalby, O.; Butler, D.; W. Birkett, J. J. Forensic Sci. 2010, 55, 924− 943. (9) ASTM, Subcommittee E30.01 on Criminalistics, Standard Guide for Gunshot Residue Analysis by Scanning Electron Microscopy/Energy Dispersive X-Ray Spectrometry (Reapproved 2001), 2010, E 1588-10 ε1. (10) Souza Sarkis, J. E.; Neto, O. N.; Viebig, S.; Durrant, S. F. Forensic Sci. Int. 2007, 172, 63−66. (11) Wolten, G. M.; Nesbitt, R. S.; Calloway, A. R.; Loper, G. L.; Jones, P. F. J. Forensic Sci. 1979, 24, 409−422. (12) Kerr, M. F. RCMP Gaz. 1959, 21, 13−15. (13) Krishnan, S. S.; Gillespie, K. A.; Anderson, E. J. J. Forensic Sci. 1971, 16, 144−151. (14) Erden, S.; Durmus, Z.; Kilic, E. Electroanalysis 2011, 23, 1967− 1974. (15) Koons, R. D. J. Forensic Sci. 1998, 43, 748−754. (16) Udey, R. N.; Hunter, B. C.; Smith, R. W. J. Forensic Sci. 2011, 56, 1268−1276. (17) Brazeau, J.; Wong, R. K. J. Forensic Sci. 1997, 42, 424−428. (18) Weber, I. T.; Geber de Melo, A. J.; Lucena, M. A. M.; Rodrigues, M. O.; Alves, S. Jr. Anal. Chem. 2011, 83, 4720−4723. (19) Zeichner, A. Anal. Bioanal. Chem. 2003, 376, 1178−1191. (20) White, R. S.; Owens, D. A. J. Forensic Sci. 1987, 32, 1595−1603. (21) Germani, M. S. J. Forensic Sci. 1991, 36, 331−342. (22) Zeichner, A. J. Forensic Sci. 2001, 46, 1447−1455. (23) Burnett, B. Skin Debris and Gunshot Residue Samplers: I. The Particle Habitus, 2004. http://meixatech.com/ SKINDEBRISandGSRSAMPLING1.pdf (accessed January 9, 2012). (24) Fernandez, B.; Claverie, F.; Pecheyran, C.; Donard, O. F. X.; Claverie, F. Trends Anal. Chem. 2007, 26, 951−966. (25) Konz, I.; Fernandez, B.; Fernandez, M. L.; Pereiro, R.; SanzMedel, A. Anal. Chem. 2011, 83, 5353−5360. (26) Ugarte, A.; Unceta, N.; Pécheyran, C.; Goicolea, M. A.; Barrio, R. J. J. Anal. At. Spectrom. 2011, 26, 1421−1427. (27) Sanborn, M.; Telmer, K. J. Anal. At. Spectrom. 2003, 18, 1231− 1237. (28) Gallo, J. M.; Almirall, J. R. Forensic Sci. Int. 2009, 190, 52−57. (29) Weis, P.; Duecking, M.; Watzke, P.; Menges, S.; Becker, S. J. Anal. At. Spectrom. 2011, 26, 1273−1284. (30) De Gaetano, D.; Siegel, J. A.; Kiomparens, K. L. J. Forensic Sci. 1992, 37, 281−300. (31) Horn, I.; Günther, D. Appl. Surf. Sci. 2003, 207, 144−157. (32) Zeichner, A.; Levin, N. J. Forensic Sci. 1993, 38, 571−584. (33) Trejos, T.; Almirall, J. R. Anal. Chem. 2004, 76, 1236−1242. (34) Adam, C. Essential Mathematics and Statistics for Forensic Science; Wiley-Blackwell: West Sussex, UK, 2010. (35) Singer, R. L.; Davis, D.; Houck, M. M. J. Forensic Sci. 1996, 41, 195−198. (36) Steffen, S.; Otto, M.; Niewoehner, L.; Barth, M.; Brożek-Mucha, Z.; Biegstraaten, J.; Horváth, R. Spectrochim. Acta, Part B 2007, 62, 1028−1036. (37) Reis Edson, L. T.; Sarkis Jorge, E. S.; Neto Osvaldo, N.; Rodrigues, C.; Kakazu Mauricio, H.; Viebig, S. J. Forensic Sci. 2003, 48, 1269−74. (38) Smith, W. D. Anal. Chem. 2002, 411A. (39) Torre, C.; Mattutino, G.; Vasino, V.; Robino, C. J. Forensic Sci. 2002, 47, 494−504.

(40) Cardinetti, B.; Ciampini, C.; D’Onofrio, C.; Orlando, G.; Gravina, L.; Ferrari, F.; Di Tullio, D.; Torresi, L. Forensic Sci. Int. 2004, 143, 27−46. (41) Berk Robert, E.; Rochowicz Scott, A.; Wong, M.; Kopina Michael, A. J. Forensic Sci. 2007, 52, 838−841. (42) Brozek-Mucha, Z. Forensic Sci. Int. 2011, 210, 31−41. (43) Pisonero, J.; Koch, J.; Waelle, M.; Hartung, W.; Spencer, N. D.; Guenther, D. Anal. Chem. 2007, 79, 2325−2333.

2409

dx.doi.org/10.1021/ac203155r | Anal. Chem. 2012, 84, 2402−2409