Determination of Endocrine Disrupting Chemicals in Drinking Water at Sub ng/L Levels Using Direct Injection and 6495 Triple Quadrupole Mass Spectrometry

Poster Presentation

Prepared by C. Marvin1, A. Fandino1, L. Tölgyesi2, B. Wuest2, D. Yang3
1 - Agilent Technologies, Inc, 5301 Stevens Creek Blvd., 3U-WQ, Santa Clara, CA, 95051, United States
2 - Agilent Technologies Sales & Services GmbH & Co., Hewlett-Packard-Str. 8, Waldbronn, 76337, Germany
3 - Agilent Technologies, Inc., 5301 Stevens Creek Blvd., 3U-WQ, Santa Clara, CA, 95051, United States

Contact Information:; 302-636-8423


Endocrine disrupting chemicals (EDCs) at sufficient amounts can interfere with the endocrine system in mammals. EDCs of different origins in aquatic streams may lead to the potential contamination of the drinking water. As a result, EDC levels in municipal water supplies are regulated by several agencies down to ng/L levels (EPA Method 539, EPA Method 1698). A high-end triple quadrupole has been modified to enhance sensitivity in positive and negative ion modes. Improvements include an optimized Q1 ion transfer optics and a novel ion detector that uses a high voltage conversion dynode with low noise characteristics.

Tap water (Santa Clara, California) was spiked with androstenedione, equilin, 17-?-estradiol, estriol, estrone, 17-?-ethynylestradiol and testosterone at different concentration levels. 900 µL were injected into the system by direct injection. Chromatography was performed using gradient separation with water and acetonitrile/methanol as mobile phases. Mass spectrometer was operated in MRM and fast polarity switching mode. Two transitions were monitored per compound.

In comparison to formate/acetate or ammonia buffers, the addition of 0.4 mM ammonium fluoride in the aqueous mobile phase enhanced the signal response for most of the hormones. The mobile phase gradient was found to be critical to separate the matrix interfering materials from the analytes. Both Poroshell 120 Phenyl-Hexyl (2.1x100 mm, 2.7 µm) and HSS T3 columns (2.1x100 mm, 1.8 µm) were evaluated. Most of the hormones were detected at sub ng/L levels in spiked drinking water.

Linearity was evaluated for target compounds from 0.1 to 35 ng/L. Correlation coefficients (R2) for were higher than 0.99. Reproducibility was studied by performing repeatability studies (n=5), expressed as area RSD%. Excellent precision and accuracy within 80 and 120% was achieved. Sensitive instrumentation allows the quantitation of hormones at sub ng/L levels using direct injection without the need of time consuming offline SPE.