U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
| REPORT |
| This report is an archived publication and may contain dated technical, contact, and link information |
|
 |
| Publication Number: FHWA-HRT-14-021 Date: January 2014 |
Publication Number: FHWA-HRT-14-021 Date: January 2014 |
A total of 15 samples of AASHTO M247 Type I beads acquired from State transportation departments were evaluated as part of this study. The sample size used in this study was calculated at the 95 percent confidence interval based on the sample size determination formula provided in figure 3, where n is the sample size, σ is the underlying standard deviation of metals concentrations in the beads based on past analysis, and B is the specified error of estimation.
Figure 3. Equation. Formula for determining sample size.
The underlying standard deviation of metals concentrations for this estimation was taken from previous analysis of arsenic content in glass beads during the TAMU/TTI AGBMA funded study. The mean ± standard deviation arsenic contents from Batch 1 and Batch 3 beads in this previous study were 83.3 ± 1.42 ppm and 393 ± 6.93 ppm, respectively. Based on an allowable margin of error of 1 percent, the corresponding sample sizes were determined to be 11.2 and 12, respectively. A final selected sample size of 15 satisfied the statistical criteria based on 95 percent confidence interval and 1 percent allowable error.
Based on the final selected sample size, Dr. Carlson made the request for samples from his contacts within State transportation departments. All samples from the transportation departments were shipped directly to Dr. Carlson in quart- to gallon-size resealable zipper bags packaged in boxes. Each of the samples received were cataloged upon receipt. Once all the samples had arrived, a team member collected the samples, took out 100 g subsamples of each bulk supplied material, and renamed the samples to blind the laboratory staff running the sample extractions and analysis from the identity and location of origin of the provided samples.
Three subsamples from each of the 15 bead samples were used to determine the total, extractable, and bioaccessible contents of arsenic and lead in the beads. KOH fusion digestion was carried out to assess the total arsenic and lead content, EPA Method 3050B was carried out to assess the extractable arsenic and lead content, and the in-vitro oral bioaccessibility method was used to assess the bioaccessible arsenic and lead content in the beads. The resulting solutions from each of these methods were analyzed by ICP-MS. Because the ICP-MS analytical method is common throughout, it is detailed first below, followed by the three processing methods.
Deionized (DI) water was produced in the laboratory using a Barnstead Nanopure™ DI water system. DI water was used for preparing reagents, generating standards, and conducting experiments. Specpure® analytical standards for arsenic and lead were purchased from Alfa Aesar. Nitric acid (American Chemical Society (ACS) grade), hydrochloric acid (ACS grade), potassium nitrate (ACS grade), potassium hydroxide (ACS grade), and sodium hydroxide (ACS grade) were purchased through Fisher Scientific. Oxalic acid (ACS grade) was purchased from VWR International. Glycine (98.5–101.5 percent) and hydrogen peroxide (ACS grade, high purity) were purchased from JT Baker. SRM 612—Trace Elements in a Glass Matrix, was purchased from NIST.
Analysis was carried out as described in EPA Method 6020A: ICP-MS. An ELAN® DRC II ICP-MS system housed within TAMU’s Center for Chemical Characterization was used to quantify the concentration of arsenic and lead in solutions produced from the total metal, extractable metal, and orally bioaccessible metal extractions. All samples were preserved in 1 percent (volume/volume) nitric acid and kept at 4 °C while in storage. Samples were allowed to come to room temperature before analysis.
The Method Detection Limit (MDL) for each of the eight heavy metals using ICP-MS was determined according to 40 Code of Federal Regulations (CFR) Appendix B to Part 136, “Definition and Procedure for the Determination of the Method Detection Limit.” The MDL is the minimum concentration of a substance that can be measured and reported with 99 percent confidence that the analyte concentration is greater than zero. The practical quantitation limit (PQL) was then set based on the greater of the MDL or the lowest analyzed calibration standard. A four-point calibration was used to quantitate analytes in a range from 1 to 100 µg/L. Samples in which the analytes were present at concentrations above the highest calibration standard were diluted to within the calibration range and reanalyzed.
The KOH fusion process (procedure number APSL-03) developed by the PNNL was used to dissolve a portion of the subsampled beads for total bulk bead metals content analysis.(2) In this procedure, recycled glass beads from each batch were crushed using a porcelain mortar and pestle and passed through a #140 U.S. mesh sieve (< 100 µm). Researchers weighed 0.25 ± 0.075 g of the crushed beads and transferred them into a 7 mL carbon crucible. Approximately 1.8 ± 0.4 g of potassium hydroxide and 0.2 ± 0.1 g of potassium nitrate were added to the crushed beads, and the contents were mixed by swirling. The crucible and its contents were heated using a Bunsen burner until the mixture melted and all visible effervescence subsided. Allowing all the effervescence to subside was a minor modification to the original method that allowed for reproduction of the arsenic and lead concentrations in the SRM.(2) The digestate was then allowed to cool to room temperature.
Approximately 5 mL of DI water was added to dissolve the cake-like crystalline melt, and the resulting solution was transferred to a 1,000 mL volumetric flask. Additional 5-mL aliquots of DI water were repeatedly added to the crucible until all of the melt was dissolved. The solution in the flask was diluted to approximately 500 mL total volume using DI water and acidified using 25 ± 5 mL of concentrated nitric acid to dissolve any precipitate. Researchers added 0.3 ± 0.1 g of oxalic acid crystals to dissolve any additional observed precipitate that was not dissolved with the nitric acid addition. The flask contents were then filled to 1,000 mL mark with DI water. A 15-mL aliquot of the sample was extracted from the flask, labeled appropriately, and stored at 4 °C until ICP-MS analysis.
The total metals content of glass beads was calculated from the measured concentrations using the formula in figure 4:
Figure 4. Equation. Formula to calculate total metals content.
Where:
C = concentration of metal in fusion solution, µg/L.
∀ = volume of solvent, L.
M = mass of beads, g.
The extractable metals content of the glass beads was analyzed by EPA method 3050B.([7]) High purity ACS grade nitric acid and high purity ACS grade hydrogen peroxide were used for this test. Refluxing columns were used as vapor recovery devices, and a water bath capable of heating up to 100 °C was used as the heat source. ICP-MS was used for analysis of samples. Researchers measured 1.0 ±0.01 g of each glass bead sample and placed them in a circular flask/digestion vessel into which a 1:1 (volume:volume) mix of nitric acid was added. The sample was heated to 95 °C in a water bath and then refluxed for 15 min.
After cooling, an additional 5 mL of concentrated nitric acid was added, and the sample was heated to 95 °C in the water bath and refluxed for 30 min. Because fumes were not observed, additional nitric acid addition and refluxing was not performed, and the sample was heated for 2 h without boiling at 95 °C and then cooled. Researchers added 2 mL of DI water and 3 mL of 30-percent hydrogen peroxide solution to the flask, and this was heated to 95 °C until effervescence was minimal. Upon cooling, the 30‑percent hydrogen peroxide solution was added in 1-mL aliquots, and the above procedure repeated until effervescence was not observed. No more than 10 mL of hydrogen peroxide solution was added to the samples. The sample was heated at 95 °C without boiling for 2 h. The sample was then cooled and filtered using Whatman filter paper no. 41 and diluted to 100 mL. A 15-mL aliquot of the sample was extracted, labeled appropriately, and stored at 4 °C until analysis was conducted by ICP-MS.
The extractable metals content of glass beads was calculated from the measured concentrations using the formula in figure 5:
Figure 5. Equation. Formula to calculate extractable metals content of glass beads.
Where:
C = concentration of metal in extraction solution, µg/L.
∀ = volume of solvent, L.
M = mass of beads, g.
The in-vitro oral bioaccessibility method developed by Kelly et al. was used to determine the bioaccessibility of arsenic and lead in the glass bead samples.(1) This method was found to directly correlate results from in-vivo bioavailability testing protocols for heavy metals. The Kelley et al. method was selected for use in evaluation of the accessibility of metals for biouptake because the in-vitro test is faster and eliminates in-vivo testing.
The in-vitro oral bioaccessibility method was run in a 0.4 M glycine solution with a pH adjusted to 1.5 ± 0.05 with hydrochloric acid. Researchers weighed 0.1 g of each bead and placed them in a high-density polyethylene bottle containing 100 mL of the pH-adjusted glycine solution. The reactors were tied using zip-ties to rotators (Barnstead Thermolyne LABQUAKE®) and placed in an environmentally controlled, orbital shaker chamber (LabLine Orbit Environ-Shaker, Model # 3948, Lab-Line Instruments Inc., Melrose Park, IL). The rotators were operated at their maximum rotation speed of 8 rpm (which was a modification of the original 30 rpm used in the Kelly et al. method), and the chamber temperature was maintained at 37 ± 5 °C for an hour. After 1 h, the rotators were turned off, and the samples were allowed to settle for 5 min. A 15‑mL aliquot of the solid free supernatant was extracted from each bottle, labeled appropriately, and stored at 4 °C before ICP-MS analysis.
The oral bioaccessible content of arsenic and lead in the glass bead samples was calculated from the measured concentrations using the formula in figure 6:
Figure 6. Equation. Formula to calculate the oral bioaccessible content of arsenic and lead in glass bead samples.
Where:
C = concentration of metal in glycine solution, µg/L.
∀ = volume of solvent, L.
M = mass of beads, g.
Four methods were compared for their ability to evaluate arsenic and lead contents in glass beads used in pavement markings: 1) the KOH fusion method followed by ICP-MS, 2) EPA Method 3052 (microwave assisted hydrofluoric acid digestion) followed by ICP or AAS performed at EPA to evaluate metals in siliceous solids, 3) bench-top XRF performed at FHWA, and 4) FP-XRF performed at Florida Department of Transportation (FDOT). Subsamples from glass beads (with the exception of sample AA) received at TAMU were sent to each agency for testing. SRM 612 was also analyzed for total arsenic and lead by EPA and TAMU.(2, 3)
The retroreflective performance measurements were conducted by creating pavement marking samples containing glass beads on metal sheets. The metal sheets were painted using a shoe to put down the paint. The shoe was dragged along the metal sheet to spread the paint with a uniform thickness of 15 mils over the entire length of each pavement marking sample.
Immediately following application of the paint, glass beads were applied on the surface using a bead dispenser for even, but random and dispersed, application on the paint. Three replicate markings were used to assess the retroreflectivity of each glass bead sample. After curing, the markings for 24 h, a Delta LTL-X retroreflectometer was used to measure the retroreflectivity of the pavement-marking samples in units of millicandela per square meter per lux (mcd/m2 lx). The retroreflectivity was measured in two directions: in the direction of the paint application and the opposite direction. The retroreflectometer was used to take five independent measurements from each direction, which were summed to determine the final retroreflectivity value for each sample.
Researchers placed 80-g subsamples of glass bead samples AA, DC, and EA in an up-flow cartridge reactor. The cartridges, previously described by Boulanger, were 25 mm in diameter and 265 mm in length.(5) pH-adjusted DI water at a pH of 7 was passed through the cartridges at a flow rate of 30 mL/h. Researchers sampled 10 mL subsamples of the cartridge effluents after 1 h. Samples were preserved with 1 percent hydrochloric acid and refrigerated at 4 °C until they were placed in a cooler and shipped to EPA’s NRMRL overnight. Received samples were transferred to a refrigerator at EPA prior to analysis using a coupled high-performance chromatography (HPLC) ICP-MS system. All samples were analyzed within 2 weeks of their receipt.
Arsenic speciation was completed in the EPA’s NRMRL under the guidance of Dr. Mallik Nadagouda. An Agilent 1100 series high-performance chromatograph was used to chromatographically resolve arsenic species on a ZORBAX Eclipse® XDB-C18 (5 mm by 4.6 mm (inside diameter) by 250 mm) column using a binary eluent system. The binary mobile phase consisted of DI water with either 5 mM tetrabutylammonium hydroxide (eluent A) or 2.5 mM ammonium phosphate (eluent B) that is adjusted to a pH of 6.0. pH adjustment was accomplished for eluent A using phosphoric acid and eluent B using ammonium hydroxide. A 100 µL injection introduced the sample to the front of the column, and a linear elution gradient with a flow rate of 1 mL/min was used. The linear gradient program used included 0 percent B from 0 to 1.5 min, ramped to 50 percent B at 4 min, held at 50 percent until 6 min, ramped to 100 percent B at 8 min, and returned to 0 percent B at 12 min.
The HPLC’s effluent was sent to an Agilent 7500cc ICP-MS using 0.25 mm inner diameter polyetheretherketone tubing. The ICP-MS was operated in standard resolution mode for arsenic (m/z 75). Eluent flow was introduced into the ICP-MS through a micro-concentric nebulizer cell. Full operation parameters of the ICP-MS used in this method can be found in the Almassalkhi reference.([8]) A seven-point external calibration curve ranging from 1 to 150 µg/L (As3+/As5+) was used to quantify the analytes.
Approximately 50 g of each site soil sample was subsampled and weighed. A particle size distribution using a series of U.S. Sieve #30, #40, #50, and #80 was performed, and the fraction of soil retained on each sieve was weighed and kept separately. An inclined plane made of a strip of Plexi-Glass® and a light table was used to manually separate glass beads, based on roundness, color, and translucence, from a representative portion of each fraction of soil. The glass beads from each fraction were weighed and collected. The total glass bead mass in 50 g of soil sample was calculated using the equation in figure 7.
Figure 7. Equation. Formula to calculate total mass of glass beads in a representative sample.
The glass bead content in the site soil samples were calculated using the equation in figure 8:
Figure 8. Equation. Formula to calculate the percentage of glass beads within a site soil sample.
The respirable fraction (< 10 µm size) of soil samples and blank samples was obtained by modifying the wet sieving process, which is a common procedure to extract dust from soil samples.([9], [10], [11]) Approximately 50 g of site soil sample was wet sieved using a U.S. Sieve #10, #50, #230, and #800 using DI water. Water and soil particles passing through U.S. Sieve #800 were collected and stored in a glass beaker. Water from each sample was allowed to evaporate, leaving behind flakes of soil particles. The soil was scraped out with a spatula and stored in polypropylene tubes, and analyzed for total arsenic using KOH fusion digestion and ICP-MS analysis.
QA/QC efforts focused on several areas, including prevention of cross contamination, ensuring a representative subsampling from the initially provided samples, experimental controls and replicates, and QA/QC related to instrumental analysis. Cross contamination prevention included controls on sample handling that involved marking the subsamples. Any materials coming into contact with the glass beads during the experiment were also pre-screened for the likelihood of cross contaminating the glass beads. The DI water used in all laboratory experiments, the 1 percent nitric acid solution used for diluting samples, and the glycine solution used in the bioaccessible extraction were also evaluated for their background arsenic and lead content.
Experiments were carried out in triplicate to produce data between environmental factors that could be compared using statistical approaches. For every extraction procedure, a method blank, which consisted of analysis without using any sample, was generated. DI blanks and method blanks were also generated for all experimental procedures. The total content of arsenic and lead in the SRM 612 (glass wafer) was determined using the KOH fusion process.
Instrumental QA/QC followed the guidelines outlined in EPA Method 6060A, and the MDL was determined as described in 40 CFR Appendix B to Part 136. Interferences were not observed for arsenic and lead, and the instrument limit of detection and resulting MDLs were able to observe quantifiable concentrations of these two components within the experimentally derived samples.
