Solid-state Forensic Finger sensor for integrated sampling and detection of gunshot residue and explosives: towards ‘ Lab-on-a-ﬁ nger ’

Increasing security needs require ﬁ eld-deployable, on-the-spot detection tools for the rapid and reliable identi ﬁ cation of gunshot residue (GSR) and nitroaromatic explosive compounds. This manuscript presents a simple, all-solid-state, wearable ﬁ ngertip sensor for the rapid on-site voltammetric screening of GSR and explosive surface residues. To fabricate the new Forensic Fingers , we screen-print a three-electrode setup onto a nitrile ﬁ nger cot, and coat another ﬁ nger cot with an ionogel electrolyte layer. The new integrated sampling/detection methodology relies on ‘ voltammetry of microparticles ’ (VMP) and involves an initial mechanical transfer of trace amounts of surface-con ﬁ ned analytes directly onto the ﬁ ngertip-based electrode contingent. Voltammetric measurements of the sample residues are carried out upon bringing the working electrode (printed on the index ﬁ nger cot) in direct contact with a second ﬁ nger cot coated with an ionogel electrolyte (worn on the thumb), thus completing the solid-state electrochemical cell. Sampling and screening are performed in less than four minutes and generate distinct voltammetric ﬁ ngerprints which are speci ﬁ c to both GSR and explosives. The use of the solid, ﬂ exible ionogel electrolyte eliminates any liquid handling which can resolve problems associated with leakage, portability and contamination. A detailed study reveals that the ﬁ ngertip detection system can rapidly identify residues of GSR and nitroaromatic compounds with high speci ﬁ city, without compromising its attractive behavior even after undergoing repeated mechanical stress. This new integrated sampling/detection ﬁ ngertip strategy holds considerable promise as a rapid, e ﬀ ective and low-cost approach for on-site crime scene investigations in various forensic scenarios.


Introduction
2][3][4] Traditionally, such analysis has been done in a central laboratory, which mandates time-consuming sampling, transportation and storage steps. 1,5These limitations result in sparse deployment of equipment and can cause delays in crime scene investigations.Furthermore, transportation and storage steps may cause contamination/degradation of the collected samples, thus jeopardizing the forensic investigation and the effective administration of justice. 1 Currently, analytical methods such as X-ray Fluorescence (XRF) 6,7 and Raman spectroscopy 8 have come closest to the realization of portable devices for forensic analysis.However, these devices are still complex, cumbersome and costly.Laser Induced Breakdown Spectroscopy (LIBS) is also an attractive technique for developing a eld-deployable sensing system, but such a device is yet to be demonstrated. 9,10On the other hand, electroanalytical devices require simple instrumentation that can be consolidated into a small footprint due to advances in microelectronics, while generating reproducible and specic signals towards electroactive analytes.Electroanalysis can thus be exploited to develop portable analytical tools using sensitive and inexpensive sensors that offer attractive opportunities for diverse decentralized forensic applications, ranging from 'alternative-site' testing (e.g., at a crime scene) to police-station screening. 1,4Recently, we demonstrated a protocol based on voltammetry of microparticles (VMP) (formerly known as abrasive voltammetry) for detecting GSR with screen printed sensors. 11However, this technique does not involve a wearable device, and still mandates the use of aqueous media and related liquid handling for detection, making the process somewhat cumbersome.An all-solid-state wearable sensor could provide a suitable, robust solution to this limitation.
Two major challenges confound the development of an allsolid-state wearable forensic sensor: (i) the rst seventy-two hours are the most critical period in forensic investigations as potential evidence not collected initially is oen lost; 7 rapid sample collection is thus a crucial step. 2,11,12(ii) Many electrochemical techniques mandate the use of a solvent electrolyte.This is a major inconvenience as liquid handling systems can suffer from problems associated with leakage and contamination.Hence, a solid-state electrolyte, incorporated onto the wearable substrate, can circumvent the need for carrying additional aqueous reagents, and facilitate rapid decentralized forensic investigations.
This article describes an innovative approach that leverages ngertip sensors for rapid on-site detection of explosives and gunshot residue (GSR).The new single-use wearable ngertip sensor, called a Forensic Finger, includes an entire electrochemical cell (electrode contingent + solid-state electrolyte) fabricated on stretchable and exible nger cots (Fig. 1A).The electrode contingent is fabricated on the tip of a nger cot using well-established screen printing technology.The new concept builds upon our expertise in printed, wearable sensors 13 and electroanalysis of forensic samples. 1,11Wearable electronic devices have received tremendous attention over the last decade.Researchers have successfully developed electronic textiles for monitoring vital physiological parameters, such as heart rate, body temperature, ECG and patient movement, 14,15 sweat rate 16 and sweat electrolytes. 17,18Epidermal sensors have also recently gained importance with groups demonstrating tattoo-based devices for monitoring physiologically-relevant physical 19 and chemical 20 parameters.However, a wearable chemical forensic device for detecting explosives and GSR is yet to be demonstrated.
For simple and reliable sampling the new printable Forensic Finger utilizes the VMP method.VMP involves mechanical transfer of materials directly onto the surface of a solid electrode followed by voltammetric measurements of the collected sample. 213][24] To obviate the need for an aqueous electrolyte, we synthesize a solid-state ionogel, which is cast directly onto the tip of a nger cot.6][27][28] These represent attractive electrolyte materials owing to their low cost, mechanical exibility and ionic conductivity.
Following the fabrication of the Forensic Finger, the user adorns the electrode-printed nger cot on the index nger and the ionogel-modied nger cot on the thumb (Fig. 1A).To investigate a surface for possible GSR/explosives residues, the user gently abrades the index ngercontaining the nger cot with the printed electrodeson the surface (Fig. 1B) and then brings it in contact with the ionogel-coated thumb to complete the electrochemical cell (Fig. 1C).The sample is then analyzed using rapid square wave voltammetry with a eld-portable electrochemical analyzer.The complete process can be carried out independently by the user within four minutes.
This printable ngertip sensor was characterized for the detection of GSR (from live bullet casings) and 2,4-dinitrotoluene (DNT) powder residues.Detailed studies demonstrate the high specicity of the Forensic Finger sensor towards the target analytes, the stability of the ionogel, and the robustness of the system towards mechanical stress.Finally, the Forensic Finger was employed for on-site detection of GSR at a local ring range.The new forensic tool thus aims to support on-site criminal investigations, by rapidly identifying suspected explosive or gunshot residues (without a detailed quantitation).Its attractive analytical performance, illustrated in the following sections, during laboratory and in-eld testing, along with its ergonomics, low-cost, and robustness, opens up new avenues in wearable devices for decentralized crime scene investigation.
All laboratory-based electrochemical measurements were performed using a CH Instruments (Austin, TX) model 630C electrochemical analyzer while a portable CH Instruments model 1230A electrochemical analyzer was used for eld-based GSR assessment.

Fabrication and detection protocol for Forensic Finger
The sensors were fabricated using an MPM SPM semi-automatic screen printer (Speedline Technologies, Franklin, MA).The sensor patterns were designed in AutoCAD (Autodesk, San Rafael, CA) and outsourced for fabrication on 75 mm-thick stainless steel stencils (Metal Etch Services, San Marcos, CA).A separate stencil pattern was created for each layer (Ag/AgCl, carbon, insulator).Prior to screen printing the sensors, the nger cots were stretched over an alumina substrate to provide a at surface for high resolution printed electrodes.They were secured using polyester tape.The rst step in the sensor fabrication process involved the printing of a layer of transparent dielectric ink.This layer was printed to impart further mechanical resiliency to the underlying nger cot and the electrode system.This was followed by printing layers of Ag/ AgCl, carbon, and nally a transparent dielectric ink to dene the active electrode area.Following each routine, the ink was cured at 90 C for 10 min.The contacts for the sensor interface to the electrochemical analyzer were obtained by screen printing Ag/AgCl on a exible PET substrate, which was subsequently cured at 90 C for 10 min.The Ag/AgCl-coated PET substrate was later cut into rectangular strips and xed to the printed sensor on the nger cots using conductive silver adhesive (MG Chemicals, Ontario, Canada).The conductive adhesive was cured at 65 C for 10 min.
The ionogel was synthesized according to a reported method. 29In brief, 20 wt% PEDGA and 2 wt% initiator were mixed with C 2 mimBF 4 .The composition was sonicated for 30 min to obtain a uniform mixture.Later, 10 mL of the mixture was spread over 1 Â 1 cm 2 area of the nger cot (stretched and attached to an alumina substrate, similar to the screen printing step) and irradiated with a UV light using a PortaRay 400R curing system (Uvitron International, Inc., West Springeld, MA) for 30 s, leading to photo-crosslinking of the PEGDA to obtain the ionogel.Fig. 1A shows the actual photo of both the nger sensor and the ionogel.
The nger cot containing the printed sensor was worn on the index nger and the ionogel-coated nger cot was worn on the thumb of the same hand for the detection of GSR or explosives (Fig. 1A).The index nger (with printed sensor) was abrasively rubbed over a surface (possibly containing the analyte powder) in a manner which allowed some of the powder to be mechanically transferred onto the active area of the sensor (Fig. 1B).For GSR analysis the Forensic Finger was abrasively rubbed over Remington UMC Target 45 automatic ammunition, on which GSR was present.In the case of explosives detection, the sensor was abraded in a similar fashion over a DNT powder (as received form the manufacturer).Later, the thumb (with ionogel) was brought in contact with the index nger (Fig. 1C) to achieve a complete electrical circuit.Each Forensic Finger was used for one-time analysis.Separate sensors were used for the detection of DNT and GSR.

Square-wave anodic stripping voltammetry
Square Wave Stripping Voltammetry (SWSV) was employed to characterize the electrochemical signature of GSR.A potential of À0.95 V was applied for 120 s, and a scan to a nal potential of 0 V vs. Ag/AgCl was performed.An accumulation time was implemented for deposition of metal ions present in GSR alongside metallic species.Square-Wave Voltammetry (SWV) was employed to characterize the electrochemical signature of DNT.The voltammograms were scanned from an initial potential of 0 V to a nal potential of À1.75 V vs. Ag/AgCl.All scans were performed at a frequency of 25 Hz, an amplitude of 25 mV, a potential step of 4 mV and using the ionogel described in Section 2.2.

Results and discussion
In this work, GSR and DNT were electrochemically analyzed at a bare carbon electrode, screen-printed onto a disposable nger cot substrate.Fig. 1 outlines the sequence for sampling and analysis at these new Forensic Finger sensors.In Fig. 1A, we observe the three electrode cell printed onto the disposable nger cot worn on the index nger.The working and counter electrodes are comprised of carbon ink while the reference electrode is comprised of Ag/AgCl ink.The fabrication of this three-electrode cell is discussed in Section 2.2.We also observe the ionogel electrolyte, which is immobilized upon a nger cot worn on the thumb.Fig. 2B illustrates the abrasive sampling method which involves 'swiping' the index-nger electrode over the surface of interest, transferring the target sample directly upon the working electrode, which is then immediately ready for analysis.This 'swiping' protocol coupled with analysis is described in Section 1 and labeled voltammetry of microparticles (VMP).By placing the ionogel-electrolyte thumb in direct contact with the three-electrode index nger, the electrochemical cell is completed and is ready for immediate analysis, as illustrated in Fig. 1C.
Utilizing SWSV, as outlined in Section 2.3, we observe a distinct pattern corresponding to traces of GSR from the surface of ammunition (described in Section 2.1).Fig. 2A outlines a voltammogram for the ionogel in the absence of GSR (black) as well as the voltammetric ngerprint recorded subsequently to swiping a GSR-rich surface (red).The voltammetric pattern for this GSR sweep is very distinct and has been shown to be characteristic from scan to scan.We observe three voltammetric signals at potentials À0.6 V, À0.4 V and À0.2 V, which are attributed to Pb, Sb, and Cu, respectively.Previous data demonstrates that these metals strip at similar potentials on a similar electrode surface. 11We also attain the voltammetric signature of DNT by scanning reductively upon swiping the Forensic Finger over a DNT-rich surface.Fig. 2B displays voltammograms for the reductive scan of the ionogel in the absence (black) and in the presence (red) of The electrochemical ngerprint of this nitroaromatic compound is very distinct compared with that of the blank ionogel scan due to the presence of easily-reducible nitro groups.Three signals are observed for this reductive scan at potentials À0.9 V, À1.2 V and À1.6 V.The rst two signals at À0.9 V and À1.2 V are attributed to the stepwise reduction of the two nitro groups of DNT to hydroxylamine groups, while the third signal (at À1.6 V) is attributed to the reduction of one of the hydroxylamine groups to an amine. 30,31mploying this system, several studies were conducted to examine the specicity of the Forensic Finger sensor toward the target analytes, the stability of the ionogel over several days, and the robustness of the system towards mechanical stress.Finally, real GSR samples were examined from different surfaces at a ring range to demonstrate the eld-deployable nature of this system.

Specicity of Forensic Finger toward GSR and DNT
We examined the specicity of the Forensic Finger toward both GSR and DNT compared with VMP of other surfaces to ensure that the voltammetric signals are due to the target components and not other contaminants that may be routinely encountered.These control surfaces were obtained from random objects in the laboratory, and no attempt was made to clean or alter them.In Fig. 3(A-E), we examine the specicity of the system for the anodic scan comparing a GSR-rich surface with other surfaces upon which potential contaminants could be immobilized.Fig. 3C shows SWSV for the oxidation of species found in GSR samples.Three voltammetric signals are observed upon swiping the GSR-rich surface.The signal at À0.6 V is attributed to Pb, the signal at À0.4 V is attributed to Sb, and the signal at À0.2 V is attributed to Cu, as observed in Fig. 2A.Fig. 3A, B, D and E display VMP signals of wooden, plastic, paper and metal surfaces, respectively.A featureless baseline is observed for samples from each of these control surfaces, clearly indicating the absence of false response and substantiating that the voltammetric ngerprint observed in Fig. 3C is due to the presence of GSR.These specicity tests outline the suitability of this protocol for the detection of GSR in forensic scenarios.
Similarly, in Fig. 3(A 0 -E 0 ), we demonstrate the specicity of the Forensic Finger in the identication of DNT powder.For example, Fig. 3C 0 displays reductive SWV recorded following swiping of a DNT powder sample.The distinctive reduction signals of DNT are observed at potentials À0.9 V, À1.2 V and À1.6 V, corresponding to the stepwise reduction of the nitrogroups to hydroxylamine groups and further reduction to an amine, as previously discussed in Section 3.This voltammetry can be compared to the featureless responses obtained for samples of wooden, plastic, paper and metal surfaces, provided in Fig. 3A 0 , B 0 , D 0 and E 0 , respectively, which again indicates the absence of false signals.Based on the data represented in Fig. 3, it is clear that the Forensic Finger is well suited for eld-based analysis of various security-related compounds.

Examination of stability of ionogel for detection of GSR and DNT
The advantages of utilizing ionogels as electrolytes for wearable electrodes are outlined in Section 1.One of the most signicant advantages of the use of an ionogel electrolyte is the thermal and kinetic stability of this media due to the negligible vapor pressure of the ionic liquid. 32In order to fully evaluate this with respect to the Forensic Finger, we examined the ionogel's performance on the day of fabrication (Day 1), and 6 days thereaer (Day 7), to evaluate stability and shelf-life.The results of this study are outlined in Fig. 4. Fig. 4A & B display the results for examination of the GSR electrochemical signature corresponding to Day 1 and Day 7 aer fabrication, respectively.A featureless baseline is obtained at both time intervals for the ionogel in the absence of the GSR sample (black line).This substantiates the ionogel's stability over this period.We also observe a clear voltammetric ngerprint for GSR sampling (red line) for Days 1 and 7.The three voltammetric signals corresponding to Pb, Sb and Cu, observed in previous sections, are clear for each period, albeit the levels of the species present vary from scan to scan.This reects the limited control over the levels of sample that are transferred using the swipe sampling technique.However, this protocol is offered as a rapid eld screening tool, leading to a distinct voltammetric signature for suspected powders whereby a threshold level is set for the presence or absence of the target analyte, and no attempt at quantication is made.
Fig. 4A 0 and B 0 display the results for examination of the DNT signature for Day 1 and Day 7, respectively.As per the GSR test, a featureless baseline response of the ionogel alone (black line) is recorded for each of the time intervals showing no degradation of the electrolyte within the gel matrix.In addition, clear reduction signals are observed for DNT (red line) for each time interval, illustrating the appreciable thermal and kinetic stability of the ionogel.The three characteristic reduction signals for DNT, observed in previous sections, remain clear and well-dened (although the actual peak heights change, as expected from changes in the amount of collected DNT).These observations highlight the signicant advantage of implementing ionogels into a eld-deployable tool, as negligible care is required for the storage of this electrolyte.While this study was not extended beyond seven days, it is still expected that the ionogel would retain its stability given the thermal and kinetic properties as well as negligible vapor pressure of RTILs in general. 32

Stress study to examine strength of Forensic Finger
The Forensic Finger is presented as a wearable sensor suitable for eld-deployable analysis.Therefore, the sensor itself must be resilient against the movements of the wearer.In this study, we examined the effects of mechanical stress upon the response of the electrode system for the detection of both GSR and DNT.
The stress applied to the electrode involved iterations of opening and closing a st, while wearing the nger cot electrode, as shown in Fig. 5A (open) and B (closed).Both GSR and DNT were sampled aer applying mechanical stress iterations to the sensor.A nger cot electrode was subject to 10 exions of the wearer's hand, aer which GSR was sampled.A voltammetric scan was then taken of the GSR sample and the result is  shown in Fig. 5C (black).A second nger cot electrode was subject to 50 exions of the wearer's hand, aer which GSR was again sampled.Subsequent voltammetry was implemented on the sample and the response is also displayed in Fig. 5C (red).From an inspection of these scans, one can notice that the distinct electrochemical signature for GSR does not appear to be affected by the repeated mechanical stress upon either of the electrodes, regardless of exion number.There is negligible change of the GSR signature with increasing mechanical stress iterations (10-black and 50-red), underscoring the robust nature of the electrode.A similar experiment was performed utilizing DNT as the target analyte, and the results are provided in Fig. 5D.As with the GSR, negligible degradation of the DNT prole was observed with increasing mechanical stress for varying iterations (10-black and 50-red).The voltammetric ngerprint for each species is consistent with each stress evaluation, substantiating the stability and practicality of this ngertip electrode as a wearable sensor in a variety of practical scenarios.

In-eld analysis of Forensic Finger toward detection of gunshot residue
A study was executed at a shooting range to examine the application of the Forensic Finger in a real-life scenario involving GSR detection.The effect of the VMP technique for the detection of GSR over different surfaces was examined, as well as variations in voltammetric GSR ngerprints from subjects who handled a rearm.All the studies were performed using a portable electrochemical analyzer interfaced with a notebook computer, as displayed in Fig. 6A.The rst surface was sampled outside the shooting range, where trace GSR may be found, but where no rearms were being discharged.The voltammetry obtained for this sample with the ngertip sensor is shown in Fig. 6B (black) and displays a relatively featureless baseline within the scope of the experiment.A trace Cu signal, observed at À0.15 V, may be due to the transport of GSR particles outside of the range facility.The second sample was taken over a wooden surface within the shooting lanes of the range, and is also shown in Fig. 6B (red).This scan displays a signicant increase in the levels of GSR, as expected from a GSR-rich environment.Two voltammetric signals are observed at potentials À0.6 V and 0 V, and are attributed to Pb and Cu, respectively.A clear signal is not observed for Sb, however, contributions from Sb are indicated by the shi of the Cu peak potential to a more positive value. 11his voltammetry demonstrates the potential use of the Forensic Finger in a scenario whereby a more GSR-rich environment at a crime scene can be identied, signifying the discharge of a rearm in one particular location.The ability of the VMP technique to detect GSR from the hand of an individual, before and aer handling a rearm, has also been investigated.Samples were obtained from a subject at different instances during the experimental process: in the laboratory, prior to any contact with GSR, named Nno contact, and having handled and loaded the rearm, named L loading.These control scenarios were sampled from two different subjects, and the results are shown in Fig. 6C and D. Fig. 6C illustrates the control samples taken from one subject's right hand, and Fig. 6D displays the samples taken from a second subject's right hand.The Nno contact voltammetry (black) displays a featureless baseline within the scope of the experiment, indicating the absence of any GSR-relevant components on the subject's hands.The Lloading voltammetric response (red) displays two distinct signals at À0.6 V and 0 V which are attributed to Pb and Cu (but with contributions from Sb as previously outlined), respectively.The Pb and Cu signals are similar and consistent with those observed in Fig. 6B and as before, display a signicant ngerprint for GSR when compared to the 'no contact' scenario.The increase in these signals is consistent over two different subject samples and demonstrates that GSR can easily be transported from the surface of a rearm and ammunition to a subject merely by handling and loading.Similar voltammetric ngerprints for these control scenarios have been reported at screen-printed electrodes utilizing a liquid aqueous electrolyte. 11Variations in the current output of the signatures also demonstrate that sampling can vary, and different amounts of GSR can be obtained at different times.However, the aim of this study is to demonstrate a rapid screening tool, and it is clear that the variation in voltammetry prior to and following handling a rearm is dramatically different.The Forensic Finger concept thus represents a promising new avenue towards the on-site detection of GSR.

Conclusions
We have demonstrated the fabrication and characterization of a wearable ngertip sensor -Forensic Fingerfor eld-deployable, on-the-spot analysis of GSR and explosive agents.The new concept offers a convenient, integrated sampling and analysis routine, which can be performed within minutes, and obviates the necessity for liquid handling to realize a complete, userfriendly device.The Forensic Finger consists of electrodes, screen printed upon a stretchable nger cot substrate and is complimented with a conductive, exible ionogel electrolyte.Sampling is implemented by utilizing the simple and efficient VMP method, whereby target analytes are mechanically transferred directly onto the electrode surface via swiping the sensor over the area of interest.The Forensic Finger exhibits noteworthy sensitivity and selectivity toward both GSR and DNT.Studies have demonstrated that the ionogel electrolyte is stable over a week-long period.We demonstrate the robustness of the threeelectrode Forensic Finger sensor through mechanical stress studies and illustrate that the characteristic voltammetry of both GSR and DNT is retained.The integrated sampling and analysis steps, along with the removal of liquid handling and rapid square-wave voltammetry, ensure results within a four minute time frame.We have demonstrated the practical application of this ngertip sampling/detection system by presenting the distinct voltammetric response for GSR in a 'GSR-rich' environment as well as the voltammetric ngerprint of GSR immobilized from the hand of a subject subsequent to the handling of a rearm.The new concept holds considerable promise as a portable, eld-deployable screening method aimed at the rapid identication of a security threat or providing forensic evidence from either rearms or explosives (without detailed quantitative information).We anticipate future integration of the controlled electronic backbone in the form of a wristband, wristwatch, or ring.With the rapid development of wireless communications, the transmission of results to a smartphone or centralized database would be of substantial utility to on-site forensic investigations.

Fig. 1
Fig. 1 Schematic delineating voltammetry of microparticles at a wearable Forensic Finger.(A) The Forensic Finger exhibiting the three electrode surface screen-printed onto a flexible nitrile finger cot (bottom left inset), as well as a solid, conductive ionogel immobilized upon a similar substrate (top right inset); (B) 'swipe' method of sampling to collect the target powder directly onto the electrode; (C) completion of the electrochemical cell by joining the index finger with electrodes to the thumb coated with the solid ionogel electrolyte.

Fig. 2
Fig. 2 Voltammetric response obtained at Forensic Finger sensor/ionogel interface in the absence (black) and in the presence (red) of (A) GSR & (B) DNT.Voltammetric parameters are outlined in Section 2.3.

Fig. 3
Fig. 3 Specificity tests for the Forensic Finger toward analysis of GSR-and DNT-rich surfaces.Voltammetric response of the Forensic Finger at (A and A 0 ) wooden; (B and B 0 ) plastic; (C) GSR-rich; (C 0 ) DNT-rich; (D and D 0 ) paper and (E and E 0 ) metal surfaces.

Fig. 4
Fig. 4 Week-long stability study of the ionogel.Response in Days 1 (A and A 0 ) and 7 (B and B 0 ), voltammetry of bare sensor (black line), GSR sampling (red line -A & B), and DNT sampling (red line -A 0 & B 0 ).

Fig. 5
Fig. 5 Effect of mechanical stress applied to the Forensic Finger.(A) Opening and (B) closing of a fist, while wearing the finger cot electrode.(C) SWSV scans of GSR sampled from a GSR-rich surface subsequent to 10 (black) and 50 (red) applications of mechanical stress to the electrode.(D) SWV scans of a DNT sample subsequent to 10 (black) and 50 (red) applications of mechanical stress.

Fig. 6 (
Fig. 6 (A) VMP detection of GSR sampled at the Forensic Finger performed using a portable electrochemical analyzer (CH Instruments model 1230A) interfaced with a notebook computer.(B) VMP of GSR samples taken from two different surfaces on location in a firing range: from a railing outside the firing range (black) and from a surface inside the shooting lanes of the range (red).(C and D) VMP of GSR samples from two subjects for two different conditions: Nno contact and Lloading.Electrochemical parameters, as in Section 2.3.See Experimental section for other details.