GSR Across Variable Surfaces and Distances 

Measuring Inorganic Gun Shot Residue: 

Comparisons Across Variable Surfaces and Distances 

Tayfun Sahin

University of Rochester

OPT307: SEM Practicum
Spring 2021
Final Project

Introduction

  1. Gunshot Residue
  2. Project Goals
  3. Experiment Parameters
  4. Limitations
  5. Expectations
Methods and Results
  1. Results Summary
  2. Qualitative Review
  3. EDS X-ray Analysis
  4. Quantitative Review: Particle Density

End Remarks

  1. Discussion and Conclusion
  2. Acknowledgements
  3. References
  4. Comments

 

Introduction

 

1. Gunshot Residue

Gunshot residue ("GSR") is the material discharged after firing ammunition. Typically, GSR lands on the shooter (oftentimes their hands) or the target or nearby surroundings. Scanning electron microscopy is a reliable and commonly used instrument for analyzing GSR which can often determine if a gun was fired, by who, and sometimes other details like the type of ammunition, gun type, distance the gun was fired, etc.

The most significant elements to characterize GSR are usually antimony (Sb), barium (Ba), and lead (Pb). Lead styphnate is set off when the gun's firing pin hits the primer cap. Barium nitrate acts as an oxidizer, and antimony sulfide acts as a fuel, propelling the bullet forward. [1]
Figure 1. GSR discharged from muzzle upon firing. [2]
Figure 2. The deconstruction of a cartridge. [3]

 2. Project Goals

This project has three goals. First, to measure inorganic GSR — especially, Sb, Ba, and Pb. Second, to qualitatively compare GSR on different surfaces and at different distances. Third, to quantitatively compare GSR on different surfaces and at different distances. The quantitative comparison uses particle counting to determine particle density of the samples at selected regions.

 3. Experiment Parameters

This experiment used a long rifle with Federal (American Eagle) ammunition that was 0.22 caliber, 40 gain solid, and had a lead-nose. [4]

There were three types of samples/surfaces where GSR was analyzed: standard printer paper, hydrocarbon polymer paper, and silicon. The silicon sample was a polished single crystal of silicon from a wafer.

These samples were collected at variable distances of 1, 3, and 5 feet away from the rifle's muzzle, denoted as dmuzzle.

Example
Figure 3. Schematic of the experiment setup. The "X feet" represents either 1, 3, or 5 feet from which the muzzle was distanced from the paper. The printer paper was mounted on cardboard, and the hydrocarbon paper and silicon samples were attached to the surface of the printer paper.

The sample preparation involved removing the silicon from the printer paper and mounting it to a stub, while for the hydrocarbon paper and printer paper, a strip was cut out and mounted to a stub. Both paper samples were then placed under a platinum sputter coater for approximately 90 seconds (the silicon sample was already conductive). Given that the rate of deposit is about 1 Å/s at 20 mA current, this resulted in paper samples which had approximately 9 nm of platinum coating. Platinum metal was selected instead of gold for its finer grain size, allowing for better visibility of GSR on the paper samples.

For imaging, the conditions were controlled such that the accelerating voltage (20 kV), working distance (10 ± 0.5 mm), and magnification (300 X) were used throughout the experiment.

Unless otherwise noted, all micrographs used overlayed BSE and SE2 detectors. Typically, though not always, the ratio of the BSD was anywhere from 70-90% and the SE2 detector was 10-30%. (The "BSD" or "SE2" stated on the bottom center of the micrographs only represents which detector was primarily being used; it may have been entirely that detector or both detectors may have still been in use.) While BSD was used to highlight the high density inorganic GSR particles, the addition of the SE2 detector helped depict topography onto the otherwise flat-looking images under the BSD.

 4. Limitations

Generally, quantitative GSR comparisons of the kind used here are discouraged, and case-by-case basis is preferred in forensic applications. [5] That being said, this project merely intends to show trends, and the data is not being used in a setting requiring high standards of reproducibility.

Variability in this experiment can be attributed to multiple factors such as a small sample size, variability of GSR between the bullets fired, and variability of particle density within different regions of a sample, complicating the quantitative comparisons. Particle size also varies, though is assumed insignificant for this project. Lastly, the hydrocarbon paper and silicon were separate distances away from the gunshot hole, dhole, in the printer paper which may have affected results and is considered. The dhole distance only represents the distance between the gunshot hole and the nearest point of the sample; the region viewed in the SEM may still have been at any location within the sample, so the dhole distance is an estimation.

Example
Figure 4. Example of top-down perspective of printer paper where the silicon and hydrocarbon paper samples have different distances from the gunshot hole in the printer paper.

 5. Expectations

For the Qualitative Review: ideal or standard GSR will be spheroid. Some irregular particle morphology is expected on samples at closer distances to the muzzle. [5]

For the EDS X-ray spectroscopy: Sb, Ba, and Pb will be present and detectable. Pb is expected to be the most common element, in part due to the ammunition's lead-nose.

For the Quantitative Review: fewer inorganic GSR particles will be present at further distances (whether distance from muzzle, dmuzzle, and/or distance from gunshot hole, dhole).

 

Methods and Results

 

1. Results Summary

For the Qualitative Review: the GSR on the printer paper was visible to the naked eye, yet not visible on other surfaces or at other distances. As expected, there were more irregularly shaped particles at the samples collected at the shortest distance from the muzzle, dmuzzle = 1 ft, and there were spheroid particles detected throughout all samples and distances. There were also many organic deposits found on the paper samples at dmuzzle = 1 ft. Lastly, the GSR morphology was markedly different between the silicon and paper samples.

For the EDS X-ray spectroscopy: the expectations were met. However, although Sb, Ba, and Pb can be detected, there was less certainty that Sb was present. Other commonly detected elements included C, N, O, Al, Pt, and Si.

For the Quantitative Review: the expectations were generally met. There was an inverse relationship between particle density and distance from the muzzle, however inconclusive results when distance from the gunshot hole varied. Also, there were very few particles on dmuzzle = 5 ft samples.

 

2.Qualitative Review

Out of the three elements of interest, the molecular weights are as follows: Sb (121.76 g/mol), Ba (137.33 g/mol), and Pb (207.2 g/mol). Thus, particles with higher composition of Sb and/or Ba should be darker under BSD. Organic deposits should have even lower molecular weights, and therefore are even darker under BSD.

Typical GSR Particle Morphology:

Typical GSR particle morphology is a spheroid, which can be seen in Figure 5, along with examples of atypical shapes.

Example
Figure 5. (100% BSD) Spherical GSR particles of varying diameters can be seen on this surface of printer paper (dmuzzle = 1 ft, dhole = 0.0 mm). There are also many particles here with irregular morphology.

GSR Differences on Paper v. Silicon:

GSR differences between paper and silicon samples is attributable to the fibers in paper versus the hard solid surface of silicon. The fibers in paper often break down large GSR deposits, but in the worst case (although rarely) a hole was formed; for silicon, large deposits can occur, but some particles don't adhere to the surface as well or at all.

Example
Example
Figure 6. A hole through printer paper (dmuzzle = 3 ft, dhole = 3.0 mm). There are some small GSR particles in the region of the hole. There is some charging around the edges of the hole as well. Figure 7. (100% BSD) GSR on Silicon (dmuzzle = 1 ft, dhole = 3.0 mm), which is primarily Pb. Notably, there's a central area which lacks any GSR, suggesting a high velocity particle hit this area and bounced back.

Additional examples of Silicon, include the following:

Example
Example
Figure 8. GSR on Silicon (dmuzzle = 3 ft, dhole = 3.0 mm). While there's substantially more GSR, it still shows an empty central area where a high velocity particle likely bounced back. Figure 9. GSR on Silicon (dmuzzle = 3 ft, dhole = 3.0 mm), where this image was falsely colorized to emphasize the topography of the sample. As opposed to seeing an area where a particle was deflected, this image shows how some particles skid along the surface before stopping. Spheroid shapes do occur but are less common; in this region, all spheroids were flattened from impact.

Irregular GSR Particle Morphologies:

GSR differences between paper and silicon samples is attributable to the fibers in paper versus the hard solid surface of silicon. The fibers in paper often break down large GSR deposits, but in the worst case (although rarely) a hole was formed; for silicon, large deposits can occur, but some particles don't adhere to the surface as well or at all.
Figure 10. GSR on printer paper (dmuzzle = 1 ft, dhole = 1.5 mm), which is Pb. While the impact clearly caused some distribution, the metal does not appear to have been molten based on its powdery appearance. While other parts of the bullet contain lead (there's a lead-nose), this is perhaps unburnt lead particles from the primer. Figure 11. GSR on printer paper (dmuzzle = 1 ft, dhole = 0.0 mm), which has Pb and a thin layer of unknown composition with holes through it. It's possible that after the unknown layer landed, additional Pb particles hit the surface, and broke through the unknown layer (thus, it has holes); you can partially see Pb particles underneath the layer. The morphology of the unknown layer suggests it is metallic, while its low brightness suggests a lower MW. This image also shows how not all GSR hits the surface at one moment in time — there is delay, however short, before all particles land.

Figure 12. Some areas, such as here, had an exceptionally high density of GSR particles. This area is on printer paper (dmuzzle = 1 ft, dhole = 0.0 mm), and these particles happen to be on top of a large carbon deposit. Figure 13. This area is on printer paper (dmuzzle = 1 ft, dhole = 0.0 mm). Similar to Figure 12, where these particles are on top of a large carbon deposit. Smaller organic deposits can also be identified on the top left.

Figure 14(a). GSR on printer paper (dmuzzle = 1 ft, dhole = 1.5 mm). The metal was likely still molten upon impact, resulting in some flattened areas (esp. if higher velocity on impact) and some areas that clumped together on the paper. (See [6].) This molten-looking GSR was very commonly found at 1 ft paper samples, but not as much or at all for further distances; it's likely the shorter span of time being airborne meant the GSR here had not yet cooled or hardened. Figure 14(b). This represents an overlayed image of the elemental mapping, via EDS X-ray spectroscopy, which primarily consists of Pb (purple) and Carbon (green) from the paper's fibers. (Pb composition of the metal is >98% weight.) Not shown: trace amounts of Ba and Zn were detected as well.


 

3. EDS X-ray Analysis

X-ray spectrometry was used to analyze the elemental composition of various particles. Usually 100 seconds was used per particle or area selected. In terms of characterizing the GSR, lead and barium could be detected with high confidence, however there was greater uncertainty with antimony. Commonly detected elements include: Pb, Ba, C, N, O, Al, Pt, and Si. Other less commonly detected elements include: F, Zn, Sb, Ca, K, etc.

Examples of X-ray Spectra:
Figures 15 (left) & 16 (right). X-ray spectrum and image of GSR on hydrocarbon paper (dmuzzle = 3 ft, dhole = 3.0 mm). X-ray spectrum is focused on the center particle which contains barium with smaller deposits of lead on top.

Figures 17 (left) & 18 (right). X-ray spectrum and image of GSR on printer paper (dmuzzle = 1 ft, dhole = 0.0 mm). The x-ray spectrum is focused on the darker region of the center particle.

Possible Origins of Detected Elements: [1,7-9]
  • Sputter Coat: Pt
  • Primer:
    • Sb, Ba, and Pb
    • Ca from calcium silicide
    • K from potassium chlorate/nitrite/nitrate
    • Al from primer cup
  • Brass metal jacket: Zn
  • Substrates:
    • C, O, Si, and possibly trace Ca
  • N and O are from nitrates/nitrites
  • Unknown Origin / Identification Error: F and Rb
    • Rb is likely an identification error due to its overlap with the silicon peak. It's unknown why fluorine is identified as being present.

4. Quantitative Review: Particle Density

In addition to the parameters stated, all images were taken under high contrast to optimize the images for particle counting in ImageJ. The magnification for all images was taken under 300 X, and there was no overlay between BSE and SE2 detectors—only BSD was used for the following images.

The particle counting was conducting using ImageJ, where particles on the edge were excluded. The watershed function was used to separate overlapping particles.

Ultimately, there were three considerations for how particle density varied: (i) across ammunition rounds,(ii) across distance from the gunshot hole (dhole), and (iii) overall variability across surfaces and distances from muzzle (dmuzzle).

(i) Variability of Ammunition Rounds:

There were three iterations of firing the gun at 1 ft away, with a new piece of printer paper for each time. One strip from each page was cut out, then compared for GSR particle density.

Table 1. GSR Particle Density Across Printer Paper Samples (dmuzzle = 1 ft, dhole= 0.0 mm)
1st Iteration 2nd Iteration 3rd Iteration
7.81 · 10-3 particles/µm2
32.0 · 10-3 particles/µm2
7.37 · 10-3 particles/µm2
Figure 19. GSR on first iteration of printer paper (dmuzzle = 1 ft, dhole = 0.0 mm).
Figure 20. GSR on second iteration of printer paper (dmuzzle = 1 ft, dhole = 0.0 mm).
Figure 21. GSR on third iteration of printer paper (dmuzzle = 1 ft, dhole = 0.0 mm).

The variability here, particularly with the second iteration, could be attributable to the fact that a round of ammunition, or the way in which the gun fires, varies such that the GSR particle density correspondingly varies. However, the large disparity seen in just one sample might be better explained as the arbitrary selection of an area within a sample; this selection ideally would be representative of that sample surface's average GSR distribution, yet this approach is subject to human estimation during the image collection.

(ii) Variability Across Distance from Gunshot Hole:

From one of the pages from the prior test (Figure 20), where the gun was fired 1 ft away, there were three strips of paper cut out: at the gunshot hole, 1.5 mm away from the hole, and 3.0 mm away from the hole.

Table 2. GSR Particle Density Printer Paper Sample (dmuzzle= 1 ft) by Distance from Gunshot Hole
dhole = 0.0 mm dhole = 1.5 mm dhole = 3.0 mm
7.81 · 10-3 particles/µm2
3.61 · 10-3 particles/µm2
3.87 · 10-3 particles/µm2
Figure 19. GSR on the first iteration of printer paper (dmuzzle = 1 ft, dhole = 0.0 mm).
Figure 19(b). GSR on same printer paper at dhole = 1.5 mm.
Figure 19(c). GSR on same printer paper at dhole = 3.0 mm.

While there is a notable decrease in GSR particle density from dhole of 0.0 mm to 1.5 mm, there was no significant difference between 1.5 mm and 3.0 mm. In fact, the trend was slightly opposite of what was expected, although this is more likely attributable to variability of particle density within specific regions of a sample.

(iii) Overall Variability Across Surfaces and Distances From Muzzle:

It's important to restate that this type of quantitative GSR comparison is arbitrary, as seen by the Silicon sample at dmuzzle = 1 ft. This human error highlights the subject nature of selecting a region as "representative" for an image to be used in particle counting.

Image Matrix of GSR Particle Density Across All Sample Surfaces and Distances From Muzzle
dmuzzle = 1 ft dmuzzle = 3 ft dmuzzle = 5 ft
Printer Paper (dhole = 0.0 mm)
Hydrocarbon Paper
Silicon

Table 3. GSR Particle Density Across All Sample Surfaces and Distances From Muzzle
dmuzzle = 1 ft dmuzzle = 3 ft dmuzzle = 5 ft
Printer Paper (dhole = 0.0 mm)

15.7 · 10-3 particles/µm2

(Avg. of 3 iterations)
3.40 · 10-3 particles/µm2
0.181 · 10-3 particles/µm2
Hydrocarbon Paper

5.93 · 10-3 particles/µm2

(dhole = 2.5 mm)

1.15 · 10-3 particles/µm2

(dhole = 3.0 mm)

0.861 · 10-3 particles/µm2

(dhole = 1.2 mm)
Silicon

30.4 · 10-3 particles/µm2

(dhole = 2.7 mm)

2.18 · 10-3 particles/µm2

(dhole = 2.0 mm)

30.2 · 10-6 particles/µm2

(dhole = 3.0 mm)

From this data, as dmuzzle increases, there is less particle density if within the same sample type — even despite variations of dhole. However if dmuzzle is held constant, there are inconsistent trends regarding which surfaces have greater particle density.


 

 

END REMARKS

 

1. Discussion and Conclusion

For the Qualitative Review: the most irregular particle morphologies were present at shorter distances from the muzzle, as expected. There were also many interesting surface topographies, where GSR particle shapes are seemingly dependent on the surface. Based on the results, it appears that GSR can be deflected if it hits hard surfaces such as silicon, or can be flattened.

For the EDS X-ray Analysis: this was a tedious process to do manually, and automated systems may have found particles with composition more characteristic of GSR. One significant limitation is that ammunition primer compositions are proprietary information of the manufacturer. Literature has determined composition for many types of ammunition, but not all, nor all variations within a brand or type. No literature could be found for the specific ammunition used in this experiment. Notably, some 0.22 caliber rimfire ammunition primer discharge contains only lead, or only lead and barium. [9] Even among 0.22 caliber LR rimfire ammunition from the same manufacturer, different products have different compositions than identified here. [10] (And there are plenty of other variations on ammunition such as lead-free types which are not applicable here.) The only thing known for certain is that lead should be present, based on the ammunition package's warning about lead. Barium was experimentally detected at a sufficiently high degree of confidence which strongly suggests it is present. However, antimony was not detected at anything above a trace amount, and oftentimes the EDAX software identified the peak region as calcium, so it's possible that no antimony was actually present.

For the Quantitative Review: Generally, this aspect of the project was limited by the arbitrary selection of regions for particle density, so few larger conclusions may be made. The data suggests that as the distance from the gunshot hole increased, particle density decreased for 0.0 mm to 1.5 mm, but then was approximately the same for 1.5 mm to 3.0 mm. In terms of overall variability across surfaces and dmuzzle distances, there were inconsistent trends between samples when dmuzzle was held constant. However if the sample type is held constant, particle density always decreased as dmuzzle increased, which exemplifies the expected inverse relationship between particle density and distance. In this experiment, it appeared that the distance from the muzzle was more significant for affecting particle density than distance from the gunshot hole (at least for dhole ≤ 3.0 mm).

 

2. Acknowledgements

Thank you to Prof. Brian McIntyre for providing the rifle, ammunition, area to fire rounds, and other resources to complete this project. Also a big thank you to both Prof. McIntyre and the TA Karla Sanchez Lievano for providing tremendous help and instruction on using the SEM this semester.

 

3. References

[1] Trimpe, Michael. The Current Status of GSR Examinations, Crime Scene Resources, Inc., 20 Oct. 2017, www.crime-scene-investigator.net/the-current-status-of-gsr-examinations.html.

[2] "GSR Discharge." Forensic Science - Gunshot Residue Collection, Bureau of Criminal Apprehension, Minnesota Dept. of Public Safety, dps.mn.gov/divisions/bca/bca-divisions/forensic-science/Pages/forensic-programs-crime-scene-gsr.aspx.

[3] Parts-of-a-Bullet-Cartridge. Firearm Review, www.firearmreview.com/beginners-guide-to-reloading/parts-of-a-bullet-cartridge/.

[4] "American Eagle Rimfire 22 LR." American Eagle Rimfire | Federal Ammunition, Federal Ammunition, www.federalpremium.com/rimfire/american-eagle/american-eagle-rimfire/11-AE5022.html.

[5] Dalby, Oliver, et al. "Analysis of Gunshot Residue and Associated Materials-A Review." Journal of Forensic Sciences, vol. 55, no. 4, 8 July 2010, pp. 924-943., doi:10.1111/j.1556-4029.2010.01370.x.

[6] Charles, Sebastien, et al. "Interpol Review of Gunshot Residue 2016-2019." Forensic Science International: Synergy, vol. 2, 12 Mar. 2020, pp. 416-428., doi:10.1016/j.fsisyn.2020.01.011.

[7] Standard Practice for Gunshot Residue Analysis by Scanning Electron Microscopy/Energy Dispersive X-Ray Spectrometry: Designation: E1588-16a. ASTM International, July 2016, materialstandard.com/wp-content/uploads/2019/10/E1588-16a.pdf.

[8] Standard Practice for Gunshot Residue Analysis by Scanning Electron Microscopy/Energy Dispersive X-Ray Spectrometry. Gunshot Residue Subcommittee, Organization of Scientific Area Committees (OSAC) for Forensic Science, Mar. 2020, www.nist.gov/system/files/documents/2020/05/22/OSAC%20GSR%20SEM%20ED%20X-ray%20spec.pdf.

[9] Guide for Primer Gunshot Residue Analysis by Scanning Electron Microscopy/Energy Dispersive X-Ray Spectrometry. Scientific Working Group for Gunshot Residue (SWGGSR), Nov. 2011, www.crime-scene-investigator.net/GSRanalysisguide.pdf.

[10] Jenna, Campbell A. Analysis of Metallic Components of GSR from Various Types of Ammunition and Firearms Utilizing an SEM-EDX. Duquesne Scholarship Collection, 2018, dsc.duq.edu/cgi/viewcontent.cgi?article=1007&context=gsrs. (Note: See Figure 6 and Table 6.)


 

Comments

 

Please enter any comments, criticisms, questions, etc. below.

Your name:

Email address: