You are here:
LOW-LEVEL DETERMINATION OF PERCHLORATE IN DRINKING WATER USING ION CHROMATOGRAPHY MASS SPECTROMETRY
Citation:
Hedrick, E J., R. Slingsby, D. J. Munch, AND D. P. Hautman. LOW-LEVEL DETERMINATION OF PERCHLORATE IN DRINKING WATER USING ION CHROMATOGRAPHY MASS SPECTROMETRY. Presented at 51st ASMS Conference, Montreal, Canada, June 8-12, 2003.
Impact/Purpose:
The goal of this project is to develop a method for the analysis of perchlorate in drinking water with a method detection limit between 0.02 and 0.1 ug/L (ppb) and a minimum reporting limit between 0.1 and 0.5 ug/L (ppb). [Note: this effort focuses on development of a sensitive and accurate method for laboratory analysis of perchlorate in drinking water: a related NERL task under the RARE program, 12505, focuses on development of a screening method for use with source waters.]
Description:
Perchlorate is a drinking water contaminant originating from the dissolution of the salts of ammonium, potassium, magnesium, or sodium in water. It is used primarily as an oxidant in solid propellant for rockets, missiles, pyrotechnics, as a component in air bag inflators, and in highway safety flares. Based on EPA Information Request Responses, there are 44 states that have reported perchlorate manufacturers or users. From accidental releases and disposal, perchlorate has become a contaminant in surface and ground waters where it is highly mobile and, due to its chemical stability, persists for decades. The primary human health effect is inhibition of iodide uptake by the thyroid gland. By disrupting thyroid hormone production, perchlorate interferes with metabolism and can affect brain development in fetuses and children, leading to mental impairment. The perchlorate anion (ClO4 -) has been found in numerous drinking water supplies at concentrations that recent studies indicate may adversely affect human health. There is an urgent need to be able to confirm and quantify perchlorate at lower concentrations than the currently approved U.S. EPA method allows. In this work, sub-ppb quantitation of perchlorate in drinking waters using ion chromatography with conductivity suppression, electrospray ionization mass spectrometry (IC-ESI-MS) was demonstrated. The primary mass of interest is 99 based on the 75.77% relative abundance of the chlorine-35 isotope. Mass 101, derived from the 24.23% abundance of chlorine-37, is a secondary mass that was also utilized for quantitation and confirmation.
Low-level linear calibrations from 0.01 - 1.0 ppb, yielded R2 >0.999. Method detection limits (MDLs) in deionized and high ionic waters (up to 1000 ppm common anions sulfate, chloride and carbonate) were from 0.03 - 0.11 ppb with no significant difference, at alpha=0.01, between mass 99 and mass 101 MDLs. Precision of replicate injections at 1 ppb, yielded <5% relative standard deviation on mass 99 and <5% on mass 101 on a daily basis. Accuracy as determined by analysis of a certified reference material was +5% of the certified value in deionized water. Ruggedness, as determined by the reproducibility of area counts of 1.0 ppb check standards analyzed periodically over a day of continuous analysis of high ionic matrices, revealed some deterioration of signal intensity (~15% drop).
The major cause of a loss of sensitivity was fouling of the MS sampling cone. Rapid fouling occurred during the first five minutes of analysis due to the elution of cations and common anions such as chloride, sulfate and nitrate. It is also during this time frame that the eluate pH dips to pH 1 due to H+ exchange with monovalent cations that occurs in the electrolytic conductivity suppressor. This highly acidic eluate can damage the stainless steel capillary and sampling cone. To slow column fouling, mass spectrometer manufacturers often recommend diverting the LC flow to the MS until just prior to the elution of the analyte of interest (matrix cutting). Matrix cutting proved to be beneficial for extending the period between cone cleanings and in maintaining signal intensity over the course of a day. A chromatographic problem observed was tailing of large concentrations of the anions sulfate, chloride and nitrate into the elution time of ClO4-. Due to a minor isotope of sulfur, HSO4 - (mass 99) interferes spectrally with ClO4-. The best way to eliminate the problem is to remove the sulfate prior to sample analysis using precipitation with barium.
Recoveries in three different tap waters ranged from 99 -102% based on the average recoveries from quantitation at masses 83, 99 and 101. In-source collisionally activated dissociation of ClO4- to yield ClO3- was done by increasing the cone voltage and was used as additional confirmation of ClO4 - .
Contaminated ground waters with suspect concentrations of ClO4- (based upon previous analyses using IC with conductivity detection) were analyzed by the IC-MS method. The IC-MS method revealed that the suspect contaminants were not perchlorate. One site yielded low spike recovery which is believed to be due to ionization suppression from the combined effect of high sulfate and another contaminant with the same retention time as ClO4 -
Future work will focus on finding a suitable internal standard and in exploring whether other mobile phases and separator columns can achieve the same performance. Isotope dilution calibration using an enriched ClO4- standard (enriched on 18-O) will be tested in the near future.
Overall, IC-suppressed conductivity-ESI-MS proved to be sensitive and specific for ClO4- in drinking and ground waters at sub-ppb concentrations.