Lead Contamination from Shooting Sports

Performed by Dean Moore

Chemistry 4181, University of Colorado, Spring 2004

Experiment performed April 3 through April 9, 2004

Most gun ammunition is made of lead or lead alloys, and the purpose of this research was to determine levels of lead contamination resulting from shooting sports. The paper uses terminology pertaining to guns without definition; see [1], [2] and [3] for explanations. Flame atomic absorbance spectroscopy at 283.3 nm was used to determine the lead content of samples. After the right-handed firing of about fifty rounds of .38 Special half-jacketed ammunition from a Smith & Wesson double-action .357 magnum revolver at a Boulder-area indoor range, for a single right-hand sample, 880±119 mg of lead was found on the right hand; assuming a hand area of exactly 12 square inches and scaling to a square foot, this corresponds to 10600±1430 mg of lead per square foot. The left hand yielded 443±152 mg of lead; this scales to 5320±1820 mg of lead per square foot. Wall and soil samples varied widely, and the standard deviation of the average was calculated by using the largest RSD of all samples and the average. Two samples were taken from the wall of the same indoor Boulder-area target range as the hand samples, and yielded an average of 16700±1620 mg of lead per square foot. Of soil samples, five samples from the middle of an outdoor Boulder county trap range gave negative values, and are taken as indeterminate. Two samples taken by the firing line of a Boulder county trap range gave meaningful numbers of high standard deviation and an average of 239±215 ppm. Six soil samples taken from the impact zone of an outdoor target range in Boulder county had values ranging from 57800 ±11400 ppm of lead to 106000±20600 ppm of lead, and an average of 84600±16670 ppm lead was derived. No conclusions may be drawn of wall and hand samples, as there were too few data points. It may be concluded that soil lead concentrations far above EPA limits were found for the impact zone of the Boulder-area target range, but it may have been a concentrated sample and more investigation is needed.

Introduction

The objective of this research was to measure the lead contamination from shooting sports by using flame atomic absorbance spectroscopy. Two types of samples were analyzed: soil samples from outdoor target ranges, and swipes from hands used in shooting and swipes from walls.

Lead is toxic to nearly all life, and has been part of the environment for thousands of years ([4]); it is mentioned in the Book of Exodus (15:10) in the Old Testament. Lead’s toxicity is well-documented [5]; it has been implicated in reducing IQ scores of children, in kidney disease, high blood pressure, anemia, and many forms of damage to the reproductive systems of both sexes, including impotence, miscarriage, and stillbirth [6].

Lead is and has been introduced to the environment in many ways: from discarded lead-acid batteries, from the computer industry, from power plants, leaded gasoline (banned in the United States since 1986, [7]), natural sources, and others, including target shooting. Shooting ranges are almost entirely unregulated, and the EPA does not consider the firing of bullets and/or shot to be “discarding” lead, hence the EPA does not regulate shooting sports [8]. A Colorado study [9] found that after a three-month period of firearms instruction at an indoor range, some police trainees had blood lead levels above OSHA limits. Lead slowly oxidizes [10] when exposed to air, and dissolves in acid water or soil, polluting soil and groundwater. It has been estimated [11] that in soil of pH 5.5 that a shotgun pellet will dissolve completely in 100-300 years.

Shooting contributes lead to the environment in two ways: bullets and shot, and also primers, which are often made with lead styphnate [12].

Shooters may inhale lead, and lead may be absorbed through pores on the skin [13]. Spent ammunition and lead from primers may affect humans, and may kill birds including waterfowl [14] that mistake shotgun pellets for seeds, and predator/scavengers such as eagles [15], [16].

Outdoor shooting sports’ contribution of lead to the environment is highlighted in the next graph [17]:

Figure 1: Outdoor target ranges put more lead into the environment than nearly any other major industrial sector in the U.S., yet they remain almost entirely unregulated.

Published EPA levels [5] of lead follow:

Table 1: EPA lead limits

Floors	40 mg of lead in dust / square foot
Interior window sills	250 mg lead / square foot
Bare soil in children's play areas	400 parts per million (ppm) of lead
Bare soil in the rest of the yard	1200 ppm average

Flame atomic absorbance spectroscopy was used due the method’s simplicity, familiarity, and lead's strong response to the technique; lead’s strong line at 283.3 nm was used.

Experimental

This research was adapted from the Lead in Soil experiment [4]. Target ranges from which samples were taken are not identified by name.

Lab Procedure: Due to the use of concentrated acid, safety measures were taken. Gloves and lab glasses were used throughout, and waste was properly disposed of. Water used was 18 MW, and all glassware was rinsed with 1% nitric acid before use.

Standard Preparation: A stock solution of 103±0.631 ppm lead was prepared by dissolving 0.0165±0.0001 g of lead nitrate to the mark in a 100 mL volumetric flask with 1% nitric acid. Eight dilutions in 50 mL volumetric flasks were prepared by adding 1% nitric acid to stock solution: 0.516±0.021 ppm, 1.03±0.022 ppm, 2.06±0.024 ppm, 5.16±0.061 ppm, 10.3 ±0.076 ppm, 25.8±0.180 ppm, 51.6±0.330 ppm, and 103±0.631 ppm of lead, the last being undiluted stock solution.

Sample Gathering: Soil samples were collected in sandwich bags from outdoor target ranges. Surface samples of walls were collected by taking a picture frame, holding it on the wall, and swiping the area with a baby wipe [18]. Hand swipes were taken in the bathroom immediately after shooting, putting a lab glove on one hand, and swiping the other with a baby wipe. A new glove was put on for the other hand, storing both in plastic sandwich bags.

Samples Ran: These consisted of one swipe from the right hand after the right-handed firing of about 50 rounds of .38 Special half-jacketed ammunition from a Smith & Wesson .357 magnum double-action revolver at an indoor range, one from the left hand. Unfortunately, in the rush the type of ammunition was not collected. Two swipes were taken perhaps a foot apart from the wall by the firing line of the same range; the range obviously had poor ventilation and had dirty walls. Five samples of soil were ran from the middle of a Boulder-area outdoor trap range, and three samples were taken from by the firing line of the same Boulder-area trap range; the area has been a trap range for “over forty years” according to a bystander. Six samples were ran from soil taken where bullets and shot impact the side hill at a Boulder-area outdoor range; one of the previous six samples was found to have two shotgun pellets in it, and the area has been a target range “a long time,” according to a University of Colorado student familiar with the area.

Sample Preparation Samples were prepared in two ways, due to the use of both soil samples from outdoor ranges and baby wipes for surfaces in indoor ranges: soil samples were first ground with a mortar and pestle, and weighed to close to 0.5 g. Both soil samples and baby wipes were placed in 100 mL beakers and digested in 20 mL of concentrated nitric acid, and were heated on a hot plate for an hour, swirling with a glass rod at the midpoint. The samples were cooled in an ice bath and filtered into 100 mL volumetric flasks through Whatman-41 filter paper. The filter paper was washed with 18 MW water, and the volumetric flask was diluted to the mark with 1% nitric acid. A soil blank was prepared with about 0.5 g of Ottawa sand, and given the same treatment as other soil samples. The soil spike was prepared by adding 5 mL of lead stock to ~0.5 g of Ottawa sand. A baby wipe blank was prepared with an unused baby wipe; a baby wipe blank spike was prepared as the soil blank spike, minus the weighing. Samples were slowly collected over several days, and calibration standards and samples were all read the same afternoon so as to have only one calibration curve.

Atomic Absorbance: Standards and samples were measured on the University of Colorado’s Thermo Jarrell Ash Video 12 Spectrophotometer at 283.3 nm using a lead hallow cathode lamp and a bandwidth of 0.5 nm. The standards were measured, and dynamic linear range was determined to be from 0.516 to 51.6 ppm lead. All samples were then measured, and several samples were diluted with 1% nitric acid to bring them into dynamic linear range, and read a second time.

Results and Discussion

All calculations were done on Excel, at 95% confidence where applicable.

Lead Calibration Curve: The dynamic linear range of the eight standards was 0.516 to 51.6 ppm lead; the highest concentration of 103 ppm, that of the stock solution, was not used, as the correlation coefficient dropped dramatically when it was included. Denoting absorbance A and concentration C, the calibration curve was

A = (5920 ± 160)(mL/g)C + (0.008±0.004) (unitless)

See Figure 2. The correlation coefficient was r = 0.998. The concentration error bars were nearly invisible. Of note, the intercept was large.

Limit of Detection: This was problematic, owing to the calibration curve’s large intercept, and the blank’s absorbance had a standard deviation of zero. The calibration standard’s absorbance reading at 0.516 ppm lead yielded negative lead concentration, while that at 1.03 ppm yielded positive lead concentration, and 1.03 ppm lead may be taken as a realistic limit of detection.

Calculations of Extraction Efficiencies: For baby wipes, this came in at 69.7±0.056%, and for soil samples, it was 66.4±0.126%.

Calculations of Lead Concentrations: Refer to tables 2 and 3 below for results. Note results are divided into areas and soil samples, and it is important not to compare the two. For Lead per Square Foot calculations of hands, a hand area of exactly 12 square inches is assumed. Note Lead in Grams is meaningless for wall samples. The two wall readings varied by a factor of about 1.79, and the standard deviation was derived by using the largest RSD of samples and average. For soil samples, results varied widely, and low, high, and average values are given; standard deviations are calculated by using the largest RSD of all samples and the average. See also Figure 1 above for EPA lead limits.

Table 2: Areas

Sample	Lead in Grams	Lead per Square Foot
Right hand, one sample	880 ± 119 mg of lead	10600 ± 1430 mg /sq. ft
Left hand, one sample	443 ± 152 mg of lead	5320 ± 1823 mg of lead/ sq. ft
Indoor range walls, two samples	--	16700±1620 mg of lead/ sq. ft

Table 3: Soil Samples

Sample	Lead concentrations in ppm, low value	Lead concentrations in ppm, high value	Average
Middle of outdoor Boulder-area trap range, five samples	Negative	Negative	Meaningless
By the firing line, same outdoor Boulder-area trap range as above, two samples	217±194 ppm	262±214 ppm	239±215 ppm
From where the shots impact the side hill, Boulder-area outdoor target range	57800±11400 ppm	10600±20600 ppm	84600±16700 ppm

Conclusions: For the hand samples, only one data point was gathered for each hand, and one data point is meaningless. The fact that the left hand’s lead reading was roughly half the right’s reading is a “red flag,” but more research is needed before drawing any conclusions. Searches on the web as well as the University of Colorado’s Chinook system yielded discussions of lead contamination and shooting, but nothing on levels of lead on hands.

Five samples from the middle of a Boulder-area trap range came in negative, and are taken as indeterminate. Perhaps the solutions should have been concentrated.

The samples from where the shots impact the side hill at a Boulder-area range cannot be easily dismissed. As six samples were run, this has significance. These values are within the order of magnitude of the figures 4700 to 57000 ppm as one source [11]] reports. The presence of two shotgun pellets in the highest-reading sample is a “red flag,” and perhaps the sample was drawn from an area of very high lead content. More research and samples from a broader area are needed.

The large intercept of the calibration curve could have come from many sources. Perhaps the instrument was not zeroed properly, or perhaps the nitric acid or the 18 MW water was contaminated. Of note, the nitric acid had a guaranteed lead level below 0.05 ppm.

A number of improvements and enhancements to this research suggest themselves, in no particular order:

Take many soil samples from a grid at an outdoor target range, carefully mapping lead concentrations throughout the range. Also check different depths of soil.
Keep better track of ammunition used. It would be useful to know if the primers contained lead styphnate.
Try to ascertain why the calibration curve’s intercept was so large, and work to avoid this in the future.
Try taking a sample from a non-shooting bystander before and after shooting at the indoor range, to determine how much lead a shooter gathers from ambient air.
Try taking a sample the hands after handling a recently cleaned gun before shooting.
Try sampling a gun surface, both before and after shooting.
Read soil pH. This is important in how fast lead degrades.
Get more wall samples. This is problematic; the wall samples in this research were gathered clandestinely and rather quickly, hence no ranges are identified; several bystanders at the trap range would only allow sample gathering if the range was guaranteed to be anonymous. People into guns can be extremely paranoid, and one finds references on the Internet like “No target range would allow us to take such samples.”
There are now “zero-lead-emission firing ranges” [19]. It would be worthwhile to check out how “zero-emission” they are.
No rifles or shotguns were fired, and nothing was fired at an outdoor range. This would be worth checking out.
No air samples were run. This requires equipment that the University of Colorado does not currently have, and requires dealing with potentially paranoid people. Proper equipment and permission would be required.
To check blood lead levels of regular target shooters versus people who never shoot might be interesting.
Several standard deviations were quite large. Part of this was due to absorbance readings close to the bottom of the calibration curve, and hence the inverse function of a ‘small” number was taken, and the same small number was used in division necessary in calculation of standard deviation. Getting a better calibration curve and concentrating the samples would be helpful here.