"The Basis for Compositional Bullet Lead Comparisons," Forensic Science Communications, July 2002
July 2002 - Volume 4 - Number 3
Research and Technology
The Basis for Compositional Bullet Lead Comparisons
Charles A. Peters
Forensic Physical Scientist
Materials Analysis Unit
Federal Bureau of Investigation
When the physical markings of a fired bullet recovered from a crime scene are too mutilated for visual comparison or the firearm used in the crime is not recovered, the bullet can be compared with other bullets associated with a suspect by its elemental composition. When a crime-scene bullet contains the same analytical elemental concentrations (i.e., match in composition) as the bullets from known cartridges, a single source for these bullets cannot be excluded. During the manufacturing processes, thousands of lead specimens (bullets and bullet cores) are produced with analytically indistinguishable compositions. However, those lead specimens that share the same composition are generally packaged within the same box of cartridges, or in boxes of cartridges of the same caliber and type at the same manufacturing plant, on or about the same date. When the differences in element concentrations are small but analytically significant, a comparative examination can be used to differentiate among bullets made of different alloys or to exclude a single source for bullets of the same alloy.
Comparative bullet lead analysis was developed in the early 1960s by researchers at General Atomic (now General Activation Analysis, Incorporated, Encinitas, California) under a federal grant to develop uses for neutron activation analysis (NAA). Researchers developed procedures for analyzing such materials as gunshot primer residues, glass, paint, and bullet lead. The results of their research were published in U.S. Atomic Energy Commission Reports (Lukens et al. 1970; Lukens et al. 1970), the Journal of Radioanalytical Chemistry (Guinn 1982; Guinn et al. 1987), and the Journal of Forensic Sciences (Lukens and Guinn 1971). In one research effort, the group acquired and analyzed samples from bullet lead manufacturers. The results of these analyses confirmed that a cast billet poured from a pot of molten lead is relatively homogeneous, but that leads poured from separate molten batches are distinguishable. As a result, comparative bullet lead analysis has been adopted by laboratories and accepted by courts internationally (Andrasko et al. 1993; Blacklock and Sadler 1978; Brandone and Piancone 1984; Capannesi and Sedda 1992; Cohen et al. 1988; Desai and Parthasarathy 1983; Dufosse and Touron 1998; Gillespie and Krishnan 1969; Guy and Pate 1973; Kishi 1987; Krishnan 1973; Krishnan and Jervis 1984; Sankar Das et al. 1978; Sreenivas et al. 1978; Suzuki and Yoshiteru 1996).
The NAA technique used at many laboratories has been replaced by inductively coupled plasma-optical emission spectroscopy (ICP-OES), previously known as inductively coupled plasma-atomic emission spectroscopy (ICP-AES) (Peters and Koons 1988). OES was adopted because people confused AES with auger electron spectrometry (Boss and Fredeen 1997). Since the 1970s, ICP-OES has been widely accepted and is the method of choice for most inorganic analyses (Koons 1993; Montaser and Golightly 1987). One advantage of ICP-OES is its ability to determine the concentrations of as many as 70 elements simultaneously in some samples. ICP-OES instrumentation is used in environmental, manufacturing, research, and forensic laboratories throughout the world and has been used by the FBI Laboratory in casework for the past 12 years. The ICP-OES procedure currently used in the FBI Laboratory can determine the concentrations of seven elements (antimony, arsenic, copper, bismuth, silver, tin, and cadmium) in most bullet leads. The main disadvantage of ICP-OES is that it is a destructive technique, requiring acid digestion of approximately 60 milligrams of each replicate sample of bullet lead.
Lead used in the bullet-manufacturing process is generally obtained from secondary lead smelters where the raw material is made primarily of recycled automobile batteries. Under stringent environmental regulations, these smelters separate the batteries into plastic, acid, and lead components. This lead is then mixed with lead from other sources and melted in kettles with capacities of 75 to 100 tons. This scrap lead is reprocessed into ingots (also called pigs). Elements such as copper and tin may be present but are controlled within limits determined by the economics of the process and use of the product. For bullet manufacture, there are few physical requirements for the lead. Chiefly, the lead must be processable. Antimony may be added to harden the alloy, but its level will also vary with the requirements of the product and the economics of its use. Hardened lead is generally used in non-jacketed bullets, whereas soft lead (i.e., lead where antimony has not been added) is generally used in jacketed bullets. The other elements are present in trace amounts and can vary.
Lead is generally delivered to the bullet manufacturers in several forms: ingots which are 65 to 80 pounds (Figure 1); billets which are 100 to 300 pounds (often 125); and sows which are approximately 2,000 pounds (1 ton). If delivered in ingots or sows, the lead is remelted in 7- to 10-ton pots (Figure 2) along with lead waste from the manufacturing process that may include rejected bullets (coated or uncoated), excess lead from bullet shaping, or any other scrap lead in the factory. The molten lead is then poured into a billet mold (Figure 3) and allowed to cool and solidify (Figure 4). Wire is extruded from the billets and cut into slugs (Figure 5). The slugs are formed into bullets by swaging, then tumbled for smoothness (Figure 6), and loaded along with gunpowder into primed cartridge cases (Figure 7). The cartridges are then loaded into boxes, which are stamped with a packing code (also called lot number) (Figure 8).
|Figure 1 Ingots of lead from secondary smelter Click to enlarge image.||Figure 2 Pot containing source (melt) of lead at bullet manufacturer Click to enlarge image.|
|Figure 3 Lead being poured into billet molds Click to enlarge image.||Figure 4 Solidified billets of lead Click to enlarge image.|
|Figure 5 Billets being extruded into wire and then cut into slugs Click to enlarge image.||Figure 6 Bullets are tumbled and polished Click to enlarge image.|
|Figure 7 Bullets, cartridge cases, and powder assembled to make cartridges Click to enlarge image.||Figure 8 Cartridges packed into boxes stamped with coded assembly date Click to enlarge image.|
The composition of lead reflects its inevitable heterogeneity at the secondary smelter, where the source material is usually a variable mixture of virgin and scrap lead. Differences in each batch may be attributed to environmental contamination, variations in mold-erosion rates, and temperature variations. Typically, the extracted metal must be processed further before its final use. However, the ultimate goal is to produce an acceptable product at the lowest possible cost. One consequence of the economics is that variations in composition are tolerated as long as they do not adversely affect the physical properties of the products being manufactured. Maximum levels of certain deleterious impurities are defined and not exceeded; at the same time, alloying elements are kept between pre-established minimum and maximum levels.
When processing the lead to produce wire for bullets, the ammunition manufacturer may add rejected lead from previous runs, lead trimmings, rejected bullets (including copper-plated rounds), and virtually any other source of lead in the plant that may be recycled into the pot with the lead ingots. If it was not recycled, the scrap would become an environmental hazard. Thus, with the proportions of recycled materials undoubtedly varying from batch to batch, the composition of the lead mixture will inevitably vary.
This lead mixture occurring both at the smelter and the ammunition manufacturer provides meaningful information to forensic scientists. The homogeneity of each melt supplies an identity to a batch while it provides the ability to distinguish between batches. This enables bullets to be compared by the different mean concentrations of the elements in each. The variation of the concentrations within a source depends on both the homogeneity of the source and the analytical reproducibility of the instrument making the measurements. The number of distinguishable compositions that can occur in a given concentration range increases as the variability of the measurement is decreased. For example, if the antimony level in a melt of lead were known (with 95 percent confidence) to be 0.12 % +/- 0.1, it would not be possible to distinguish (with 95 percent confidence) an alloy that contained 0.20 percent antimony from this melt on the basis of the antimony level. On the other hand, if the variability was only +/-0.001, 28 distinguishable antimony levels could exist. With modern ICP-OES instrumentation, the high precision achieved (3-5 percent relative standard deviation) in determining most elements in lead results in millions of potentially distinguishable lead compositions. It is this ability to distinguish small differences, in fairly narrow composition ranges (e.g., 0.01-0.05 percent) of the seven elements determined, that results in a high degree of discrimination between different melts.
The overall composition of the lead product is fixed after the billet formation has cooled (Figure 4). At most manufacturers, other scrap is frequently added to the lead in the melting pot throughout the dynamic process of bullet lead formation. As a result, bullets made from continuous pours may be analytically indistinguishable over only one to two tons. In one study, five billets from each of two melts produced on consecutive days were sampled at Winchester Western Company in 1974 and were analyzed by NAA. The measured percentages of antimony, copper, and arsenic determined in these samples are presented in Table 1. These results show that, for each melt, the five billets made from that melt are indistinguishable in their concentrations of all three elements. The billets from different melts are readily distinguishable by the concentrations of antimony and copper, which are significantly higher in pour one than they are in pour two, and by the concentrations of arsenic, which are slightly lower in pour one than in pour two. In another study, a single billet was extruded into a wire that was subsequently divided into the top, middle, and bottom portions of a billet. These samples were collected and analyzed by ICP-OES, the results of which are presented in Table 2. The concentrations of each of the three elements exhibit no measurable variation among the samples, indicating that this billet is homogeneous from top to bottom with respect to the measured element concentrations.
The variability of the final product can be affected by the final step in cartridge manufacturing, the packaging of cartridges into boxes. As a result of casework, in which lead from many boxes of cartridges from major ammunition manufacturers was analyzed, it has been widely demonstrated that most boxes of ammunition contain bullets from more than one melt. As previously discussed, bullets produced from a single wire (i.e., from the same billet) are analytically indistinguishable. However, during the processes of cutting, swaging, finishing, and jacketing bullets; assembling cartridges; and packaging boxes; bullets from various melts are intermingled. This was demonstrated in a published study involving 200 bullets from each of four manufacturers (Peele et al. 1991). A small part of the results of this study is shown in Tables 3 through 6. For one box of cartridges from each manufacturer, the average concentrations of five elements in each of the distinctive compositional groups are shown. The results are typical of those found in the larger study. One conclusion from this study is that for the ammunition studied (.38 special caliber cartridges loaded with lead round-nose bullets) within each box of 50 cartridges, Federal has one or two compositional groups, Remington and CCI (Cascade Cartridge Incorporated) have approximately five compositional groups, and Winchester has as many as 15 compositional groups. Although each alloy is specified by the manufacturer to contain certain antimony content, its concentration varies from 0.58 to 0.81 percent in the box of Remington ammunition and from 0.24 to 0.66 percent in the box of Winchester ammunition shown in the tables. These levels of variability in antimony and the other trace elements account for a large number of distinct compositions of bullet lead.
Compositional bullet lead comparisons are possible because each melt of lead has its own characteristic composition. There are enough identifying elements with concentrations that are measurable with good precision in the lead alloy to distinguish among most melts. Years of analysis in the FBI Laboratory have demonstrated that the distinctiveness of a melt is defined not only by the number of elements measured but also by the relative scarcity of other alloys in that melt. Not all measured elements are equally effective at discriminating among lead sources, however. In general, for most lead products, the relative source discrimination power of the measured elements decreases in the following order: copper, arsenic, antimony, bismuth, and silver (Peele et al. 1991). Tin is not included in this list because in many lead sources it is not present at detectable levels. However, when tin is present, it provides excellent discrimination among melts of lead. Antimony, specified by the ammunition manufacturers, is alloyed with lead in order to harden the bullets. The other elements are present in trace amounts and can vary from one product to another. Bullet leads analyzed from CCI, Federal, Remington, and Winchester have contained up to 0.42 percent arsenic, 6.8 percent antimony, 2.5 percent tin, 0.2 percent bismuth, 0.22 percent copper, 0.031 percent silver, and 0.011 percent cadmium. The wide ranges in concentrations of all of these elements within sources provide for thousands of distinguishable melts of bullet lead at any one time.
The composition of a molten pot of battery lead can change because of volatilization of selected elements, segregation during solidification, as well as other factors (Schmitt et al. 1989). However, in experimental studies of bullet lead ingots, no compositional variations have been observed. That is, once a composition is created, it does not change appreciably merely by being held at the pouring temperature. Even if there are several compositions within a melt due to the factors cited, the probability of a random match between unrelated melts of material would still be low because of the huge number of compositions that could potentially occur.
The more practical reason for a compositional match is that the material more likely is derived from consecutively poured billets than from a random match among the millions of possibilities among unrelated melts of material. Accordingly, the assumption of homogeneity of the melt is a conservative approach because it results in an overestimate of the number of analytically indistinguishable bullets produced.
In order to assess the significance of a compositional match, it may be helpful to know the number of bullets that can be manufactured from a homogeneous melt. A simple calculation can determine the number of bullets that can be produced from one ton of lead, though the number per ton will vary according to the weight of the bullet. For example, a .38 caliber lead round-nose bullet typically weighs 10.23 grams, which is equivalent to 158 grains. There are 454 grams in a pound, and therefore 44 bullets of this caliber are produced per pound. Because as much as 20 percent of the lead can be lost as waste during production, only 35 bullets are actually manufactured per pound, or about 71,000 bullets per ton. A .22 caliber long rifle lead round-nose bullet weighs 2.6 grams, which is equivalent to 40 grains. Therefore, with allowance for waste, approximately 140 bullets per pound or 280,000 bullets per ton can be produced. To appreciate the significance, compare this with the fact that there are approximately 9 billion cartridges produced annually by ammunition manufacturers in the United States.
In addition to the number of bullets manufactured within one melt, other factors must be considered, such as the distinctiveness of the melt, the distribution of ammunition, the relative concentration of the elements, and the date of manufacture. A manufacturer's distribution of cartridges throughout the United States is generally on a case-by-case basis. Since a single melt can be represented across many boxes of ammunition, it is expected that one source of lead can be distributed to more than one geographical area. Exceptions to this distribution might be bullets produced for law enforcement and the U.S. military, who order large amounts of ammunition at one time. If a packing code (lot number) is found on a box of ammunition, then the assembly date may be obtained from the manufacturer. Another factor that must be considered is a case where multiple shots of various calibers, manufacturers, and compositions are fired at a crime scene. If multiple compositions present in the crime-scene lead are analytically indistinguishable from lead groups in partial boxes of ammunition, it is much more likely that the crime-scene bullets came from those boxes than it is when only one compositional group is present.
All the aforementioned factors are considered when interpreting the compositional analysis data to determine if there is an association between specimens. If such an association exists, an example of the conclusion reached by the FBI Laboratory may read as follows, "The bullet removed from the victim and 10 of the 15 analyzed cartridges from the suspect residence are analytically indistinguishable from one another. Therefore, they likely originated from the same manufacturer's source (melt) of lead." This conclusion does not associate a bullet to a box but rather to a melt of lead that has bullet specimens within that box and perhaps other boxes.
Continuing and expanding on the early work by General Atomic and others, the FBI Laboratory has successfully defended challenges to the scientific validity of compositional bullet lead comparisons and its application to individual cases in federal, state, and local court systems since the 1970s. Recently, as a result of a Daubert hearing, the United States District Court in Columbia, South Carolina, admitted the technique in United States v. Jenkins 1997; and as a result of a Frye hearing, New York admitted the technique in People v. McIntosh 1998. The admissibility of this examination in court has also been affirmed on an appeal by the New Jersey Supreme Court in State v. Noel 1997.
Andrasko, J., Kopp, I., Abrink, A., and Skiold, T. Lead isotope ratios in firearm investigations, Journal of Forensic Sciences (1993) 38(5):1161-1171.
Blacklock, E. C. and Sadler, P. A. Shot-gun pellet identification and discrimination, Forensic Science International (1978) 12(2):109-117.
Boss, C. A. and Fredeen, K. J. Concepts, Instrumentation, and Techniques in Inductively Coupled Plasma Optical Emission Spectrometry. Perkin-Elmer Corporation, Wellesley, Massachusetts, 1997.
Brandone, A. and Piancone G. F. Characterization of firearms and bullets by instrumental neutron activation analysis, International Journal of Applied Radioactive Isotopes (1984) 35(5):359-364.
Capannesi, G. and Sedda, A. F. Bullet identification: A case of a fatal hunting accident resolved by comparison of lead shot using instrumental neutron activation analysis, Journal of Forensic Sciences (1992) 37(2):657-662.
Cohen, I. M., Pla, R. P., Milam M. I., and Gomez, C. D. Activation analysis of trace elements in bullet lead samples and characterization by multivariate analysis applied to a criminal case, Journal of Trace and Microprobe Techniques (1988) 6(1):113-124.
Desai, H. B. and Parthasarathy, R. A. Radiochemical neutron activation analysis method for the determination of tin, arsenic, copper, and antimony for the forensic comparison of bullet lead specimens, Journal of Radioanalytical Chemistry (1983) 77(1):235-240.
Dufosse, T. and Touron, P. Comparison of bullet alloys by chemical analysis: Use of ICP-MS method, Forensic Science International (1998) 91:107-206.
Gillespie, K. A. and Krishnan, S. S. Analysis of lead shot: A comparison of analyses using atomic absorption spectrophotometry and neutron activation analysis, Journal of the Canadian Society of Forensic Science (1969) 2(4):95-103.
Guinn, V. P. NAA of bullet-lead evidence specimens in criminal cases, Journal of Radioanalytical Chemistry (1982) 72(1-2):645-663.
Guinn, V. P., Fier, S. R., Heye, C. L., and Jourdan, T. H. New studies in forensic neutron activation analysis, Journal of Radioanalytical and Nuclear Chemistry (1987) 114(2):265-273.
Guy, R. D. and Pate, B. D. Studies of the trace element content of bullet lead and jacket material, Journal of Forensic Sciences (1973) 18(2):87-92.
Kishi, T. Forensic activation analysis: The Japanese scene, Journal of Radioanalytical and Nuclear Chemistry (1987) 114(2):275-280.
Koons, R. D. ICP atomic spectrometry in the forensic laboratory, Spectroscopy (1993) 8:16-21.
Krishnan, S. S. Trace element analysis by atomic absorption spectrometry and neutron activation analysis in the investigation of shooting cases, Journal of the Canadian Society of Forensic Science (1973) 6:55-77.
Krishnan, S. S. and Jervis, R. E. Characterization of shotgun pellets and gunshot residue by trace element concentration patterns by neutron activation analysis using the slowpoke reactor, Journal of the Canadian Society of Forensic Science (1984) 17(4):167-181.
Lukens, H. R., Schlesinger, H. L., Guinn, V. P., and Hackleman, R. P. Forensic Neutron Activation of Bullet-Lead Specimens. U.S. Atomic Energy Commission Report GA-10141, 1970, pp. 1-48.
Lukens, H. R., Schlesinger, H. L., Guinn, V. P., Hackleman, R. P., Graber, F. M., Hiatt, M. A., and Yamaguchi, T. Applications of Neutron Activation Analysis in Scientific Crime Investigation (Bullet Identification). U.S. Atomic Energy Commission Report GA-10276, 1970, pp. 62-82.
Lukens, H. R. and Guinn V. P. Comparison of bullet lead specimens by nondestructive neutron activation analysis, Journal of Forensic Sciences (1971) 16(3):301-308.
Montaser, A. and Golightly, D. W. Inductively Coupled Plasmas in Analytical Atomic Spectrometry. VCH, New York, 1987.
Peele, E. R., Havekost, D. G., Halberstam, R. C., Koons, R. D., Peters, C. A., and Riley, J. P. Comparison of bullets using the elemental composition of the lead component. In: Proceedings of the International Symposium on the Forensic Aspects of Trace Evidence. U.S. Government Printing Office, Washington DC, 1991, pp. 57-68.
People v. McIntosh, Ind. No. 146/96, County Court of New York, Dutchess County, 178 Misc. 2d 427; 682 N.Y.S. 2d 791; 1998 N.Y. Misc. Lexis 486, August 5, 1998, Decided.
Peters, C. A. and Koons, R. A. Multielement analysis of bullet lead by inductively coupled plasma-atomic emission spectroscopy, Crime Laboratory Digest (1988) 15(2):33-38.
Sankar Das, M., Venkatasubramanian, V. S., and Sreenivas, K. Isotopic analysis of bullet lead samples, Journal of the Indian Academy of Forensic Sciences (1978) 15(1):15-20.
Schmitt, T. J., Walters, J. P., and Wynn, D. A. Analysis of pure lead and lead alloys for the automotive lead/acid battery industry by inductively coupled plasma emission spectroscopy, Applied Spectroscopy (1989) 43(4):687-696.
Sreenivas, K., Venkatasubramanian, V. S., and Sankar Das, M. A multitechnique approach for bullet characterization, Journal of the Indian Academy of Forensic Sciences (1978) 17(1):1-10.
State v. Noel, A-143 September 1997, Supreme Court of New Jersey, 157 N.J. 141; 723 A.2d 602; 1999 N.J. Lexis 80, September 29, 1998, Argued, February 10, 1999, Decided.
Suzuki, Y. and Yoshiteru, M. Determination of trace impurities in lead shotgun pellets by ICP-MS, Analytical Sciences (1996) 12:129-132.
United States v. Terry Charles Jenkins, United States District Court, Columbia, South Carolina, 3:96-358 (1997).