Review of U.S. EPA Methods for Perchlorate Using Suppressed Conductivity Detection

Perchlorate is identified as an environmental contaminant found in drinking, ground, and surface waters. The origins of perchlorate are both natural and anthropogenic. In its natural origin, perchlorate was found to originate from atmospheric deposition, possibly from chloride aerosol being exposed to an electrical discharge in the presence of ozone.1 The nitrate deposits of the Atacama Desert in Chile are another natural source.2 Perchlorate contamination is often attributed to the manufacture and use of ammonium perchlorate in solid propellant for rockets, missiles, and fireworks, and in the production of matches, flares, pyrotechnics, ordnance, and explosives.3 Research on perchlorate’s prevalence in the environment has received a great deal of attention primarily because perchlorate poses a human health concern, as it impairs normal thyroid function by interfering with iodine uptake by the thyroid gland.4

In February 2011, the U.S. EPA issued a press release announcing the decision to develop a national primary drinking water regulation for perchlorate, with its anticipated publication in 2013. The EPA has determined that perchlorate meets the federal Safe Drinking Water Act’s (SDWA) three criteria for regulating a contaminant: 1) Perchlorate may adversely affect public health; 2) there is a substantial likelihood that perchlorate frequently occurs in public water systems at levels of health concern, as monitoring data show that over 4% of public water systems have detected perchlorate; and 3) there is a meaningful opportunity for health risk reduction for the 5.2–16.6 million people who may be getting drinking water that contains perchlorate.

Prior to issuance of its regulatory determination, the U.S. EPA issued a recommended Drinking Water Equivalent Level (DWEL) for perchlorate of 24.5 μg/L. In early 2006, the U.S. EPA issued Cleanup Guidance for this same amount. Both the DWEL and the Cleanup Guidance were based on research by the National Academy of Sciences (NAS).5 This followed numerous other studies, including one that suggested that human breast milk had an average of 10.5 μg/L of perchlorate.6 In February 2008, the U.S. FDA said that U.S. toddlers on average are being exposed to more than half of the EPA’s safe dose from food alone. In March 2009, a Centers for Disease Control study found 15 brands of infant formula contaminated with perchlorate. Combined with existing perchlorate drinking water contamination, infants could be at risk for exposure to perchlorate above the levels considered safe by the U.S. EPA.7

Several states have enacted a drinking water standard for perchlorate, including Massachusetts in 2006. California’s legislature enacted AB 826, the Perchlorate Contamination Prevention Act of 2003, requiring California’s Department of Toxic Substance Control (DTSC) to adopt regulations specifying best management practices for perchlorate and perchlorate-containing substances. The Perchlorate Best Management Practices were adopted on December 31, 2005, and became operative on July 1, 2006. In 2007, the California Department of Public Health issued drinking water standards, which currently have a maximum contamination level (MCL) of 6 μg/L, with a proposal to decrease the acceptable MCL to 1 μg/L in the future. Several other states, including Arizona, Maryland, Nevada, New Mexico, New York, and Texas, have established unenforceable advisory levels for perchlorate.

Ion chromatography methods for perchlorate analysis

Ion chromatography (IC) has been recognized as an effective tool for the determination of perchlorate in drinking water and other sources. U.S. EPA Methods 314.0, 314.1, and 314.2 describe the determination of trace perchlorate in drinking water using IC with suppressed conductivity detection. The method detection limits (MDLs) of U.S. EPA Methods 314.0 and 314.1 are 0.53 μg/L and 0.03 μg/L, respectively. These methods are also described in Dionex™ (now part of Thermo Fisher Scientific, Sunnyvale, CA) Application Update 1488 and Application Note (AN) 176,9 which demonstrate MDLs of 0.10 μg/L and 0.02 μg/L, respectively.

U.S. EPA Method 314.0 is subject to interferences and loss of sensitivity caused by the presence of high concentrations of the common matrix ions chloride, sulfate, and carbonate. This is illustrated in Figure 1, which shows the effects of increased matrix anions on the perchlorate peak. For this reason, U.S. EPA Method 314.0 recommends determining the sample conductivity prior to analysis, and shows an 86% recovery for a 4-μg/L spike in a high-ionic-strength synthetic inorganic water with a conductivity of 4200 μs/cm. The method provides two options for samples with high total dissolved solids (TDS): dilution and the use of trap columns. Though both strategies are effective, dilution will result in an increased method reporting level (MRL), while trap columns add a sample preparation step. In addition to high-concentration matrix anions, some anionic compounds, such as chlorobenzene sulfonates, are known to elute at a similar retention time as perchlorate, and can therefore lead to a false positive result.

Figure 1 – Effect of matrix concentration on perchlorate peak shape.

To avoid these complications, Dionex Corporation and the U.S. EPA collaborated to develop U.S. EPA Method 314.1. This method uses a preconcentration column to trap perchlorate from the matrix, followed by matrix elimination with 10 mM sodium hydroxide. Perchlorate is then separated using a 2-mm primary column. To minimize the identification of a false positive peak, a second analytical column (with a different selectivity that yields an improved separation of perchlorate from the chlorobenzene sulfonates) is used as the confirmatory column to verify the presence of perchlorate. Although the method is effective and can be used for compliance monitoring, the authors recommend the method described in Dionex AN 17810 to determine trace concentrations of perchlorate in high-ionic-strength matrices. AN 178 was also produced during collaboration with the U.S. EPA on the development of U.S. EPA Method 314.2.

Method 314.2 improves on Methods 314.0 and 314.1, and was specifically developed for the determination of perchlorate in high-ionic-strength samples using a two-dimensional IC approach. In this approach, a portion of the first-dimension separation is diverted to a concentrator column, the contents of which are then eluted to a second-dimension column to separate perchlorate from the other ions in the diverted portion of the first dimension. There are several advantages of the 2-D matrix diversion approach. First, initial sample loading onto the 4-mm column allows a large sample injection volume (large amount of sample) due to the high capacity of the analytical column and higher selectivity for analytes of interest relative to the matrix ions. Second, the analyte peak that is partially resolved in the first dimension is focused onto a concentrator column prior to the second dimension. The diverted portion of the first dimension has passed through a suppressor to convert the hydroxide eluent to water, which provides the ideal environment for retention and focusing on the anion concentrator. Third, the second-dimension column has a smaller cross-sectional area relative to the first dimension, thereby enhancing the detection sensitivity. Finally, this approach allows the combination of two different chemistries in two dimensions, enabling selectivity not possible when using a single-chemistry dimension.

In Method 314.2 and AN 178, perchlorate is partially resolved on a 4-mm Thermo Scientific Dionex IonPac™ AS20 column in the first dimension, collected onto a concentrator column, and then resolved in the second dimension with a 2-mm Dionex IonPac AS16 column. Method 314.2 has a lower MDL, ranging from 0.012 to 0.018 μg/L, depending on the sample volume analyzed. The improved selectivity eliminates the possibility of a false positive, such as a chlorobenzene sulfonate.

Improved perchlorate analysis: 2D-IC in a capillary format

Since the development of Reagent Free Ion Chromatography™ (RFIC™) systems and their utility for 2D-IC, the need to improve performance, resolution, and sensitivity has led to the evolution of smaller-scale, capillary IC systems. Capillary IC scales down traditional IC from 10-fold to 100-fold by using 0.4-mm columns and 10-μL/min flow rates. The benefits of scaling down are the ability to use less eluent and the option of leaving the system constantly running, which is much more efficient and economical. With an RFIC system continuously on, there is no need for startup or equilibration time; less calibration is required; and there is no eluent preparation for those unplanned, last-minute samples. Capillary IC falls under the scope of the U.S. EPA’s flexibility rule, and is therefore accepted for compliance monitoring.

Figure 2 – Schematic diagram of the analytical/capillary 2-D system: Perchlorate is resolved from the matrix on a 2-mm Dionex IonPac AS20 column set, concentrated on a monolithic capillary concentrator, separated on a 0.4-mm Dionex IonPac AS16 column set, and detected by suppressed conductivity detection.

Thermo Scientific AN 102411 demonstrates the improved determination of perchlorate using a capillary format in the second dimension of a 2D-IC system. Figure 2 shows the schematic of a 2D-IC system for trace determinations of perchlorate. The 4-mm and 2-mm columns used in AN 178 are replaced with 2-mm and 0.4-mm columns, respectively. The system with a smaller-diameter column is operated at a lower flow rate, thus requiring much less reagent consumption and reducing system maintenance. In this method (as compared to U.S. EPA Method 314.2), there is a much greater decrease in cross-sectional area from the first dimension to the second dimension, and therefore improved sensitivity can be obtained (i.e., there is a 25-fold increase from a 2-mm column to a 0.4-mm column as compared to a fourfold increase from a 4-mm column to a 2-mm column using the flow rates in U.S. EPA Method 314.2). AN 1024 demonstrates an MDL of 0.005 μg/L, and good recovery results for perchlorate were obtained with different sample matrices. The results of a 2-D separation on a capillary format from a high-ionic-strength (matrix) sample are shown in Figure 3.

Figure 3 – Chromatograms of synthetic high inorganic water containing 1000 mg/L each of chloride, sulfate, and bicarbonate fortified with 0.2 μg/L perchlorate in (a) first dimension and (b) second dimension.

Conclusion

IC with suppressed conductivity is the preferred method for the determination of perchlorate in drinking and high-ionic-strength waters. RFIC systems provide the sensitivity and ease of use for samples with moderate ionic strength. For more complex, high-ionic-strength samples, the use of validated 2D-IC provides many benefits, including on-line matrix elimination, increased sensitivity, and elimination of false positives. An even further improvement has been the evolution of capillary IC, which, when executed on a RFIC system in 2-D format, can achieve ppt-level detection.

References

  1. Dasgupta, P.K.; Martinelango, K. et al. The origin of naturally occurring perchlorate: the role of atmospheric processes. Environ. Sci. Technol. 2005, 39, 1569–75.
  2. Ericksen, G.E. The Chilean nitrate deposits. Am. Sci. 1983, 71, 366–74.
  3. Trumpolt, C.W.; Crain, M. et al. Perchlorate: sources, uses, and occurrences in the environment. Rem. J. 2005, 16, 65–89.
  4. Stanbury, J.B.; Wyngaarden, J.B. Effect of perchlorate on the human thyroid gland. Metabolism1952, 1, 533–9.
  5. Health Implications of Perchlorate Ingestion; Washington, DC: The National Academies Press, 2005. ISBN 10: 0-309-09568-9. http://books.nap.edu/catalog.php?record_id=11202.
  6. McKee, M. Perchlorate found in breast milk across US. New Scientist2005 [online]; www.newscientist.com/article/dn7057-perchlorate-found-in-breast-milk-across-us.html (accessed Jan 2, 2013).
  7. Jacob, A. CDC scientists find rocket fuel chemical in infant formula: powdered cow’s milk formula contains thyroid toxin. Environ. Working Group2009 [online]; www.ewg.org/report/CDC-Scientists-Find-Rocket-Fuel-Chemical-In-Infant-Formula (accessed Jan 2, 2013).
  8. Dionex (now part of Thermo Fisher Scientific) Application Update 148: Determination of perchlorate in drinking water using Reagent-Free™ Ion Chromatography, Sunnyvale, CA, 2006 [online]; www.dionex.com/en-us/webdocs/7225-AU148_V30_released121506.pdf (accessed Jan 2, 2013).
  9. Dionex (now part of Thermo Scientific) Application Note 176: Determining sub-ppb perchlorate in drinking water using preconcentration/matrix elimination IC with suppressed conductivity detection by U.S. EPA Method 314.1, Sunnyvale, CA, 2007 [online]; www.dionex.com/en-us/webdocs/48972-AN176_released20070312.pdf (accessed Jan 2, 2013).
  10. Dionex (now part of Thermo Scientific) Application Note 178: Improved determination of trace concentrations of perchlorate in drinking water using preconcentration with two-dimensional ion chromatography and suppressed conductivity detection, Sunnyvale, CA, 2006 [online]; www.dionex.com/en-us/webdocs/48897-AN178_V30_released121506.pdf (accessed Jan 2, 2013).
  11. Thermo Scientific Application Note 1024: Improved determination of trace perchlorate in drinking water using 2D-IC, Sunnyvale, CA, 2012 [online]; www.dionex.com/en-us/webdocs/113927-AN1024-IC-Trace-Perchlorate-Drinking-Water-2D-IC-AN70213_E.pdf (accessed Jan 2, 2013).

Richard F. Jack, Ph.D., is North America Market Manager; Lillian Chen, Ph.D., is Senior Applications Chemist; Brian De Borba is Applications Lab Manager; and Jeffrey Rohrer, Ph.D., is Director of Applications Development, CMD Strategic Marketing, Thermo Fisher Scientific Inc., 1228 Titan Way, Sunnyvale, CA 94085, U.S.A.; tel.: 408-965-6408; e-mail: [email protected].