Reversed-Phase Liquid Chromatography Tandem Mass Spectrometry And Elisa Determination Of Caffeine As A Marker For Sewage Contamination Of Wilgreen Lake

Date of Award

January 2011

Degree Type

Closed Access Thesis

Degree Name

Master of Science (MS)



First Advisor

Susan E. Godbey

Department Affiliation



The primary objective of this study was to determine the concentration of caffeine in Wilgreen Lake and its tributaries in order to establish the contribution of domestic wastewater effluents to the lake's impairment. This was achieved by collecting samples from different locations and depth profiles from the lake and its tributaries, and analyzing them using liquid chromatography-tandem mass spectrometry coupled to an Electrospray Ionization (ESI) Interface. An isotope dilution method using 13C3-labeled caffeine as the internal standard was adapted from known methods and employed to determine and correct the recovery of native caffeine in water samples. The presence in the lake of seven other human pharmaceutical compounds (carbamazepine, sulfamethoxazole, trimethoprim, cotinine, paraxanthine, acetaminophen and fluoxetine) belonging to various therapeutic categories in the lake was also investigated using the LC-MS/MS method of analysis.

Since caffeine exists in surface waters at low concentrations, solid phase extraction (SPE) was used to pre-concentrate the samples to achieve concentrations detectable by the instrumentation. SPE also served as a sample clean-up step to minimize matrix effects. A multi-point standard calibration curve was generated using peak area ratios of the most abundant product ions from standards containing caffeine in the range of 6.25 - 2500 ng/ml and a stable 13C labeled isotope of caffeine as the internal standard. 13C3-atrazine was used as an injection standard. Using the developed method, caffeine was detected in the range of 66 - 417 ng/L in water samples collected from various sites in Wilgreen Lake. The recoveries of 13C-labeled caffeine were from 40 to 126%, while caffeine's recovery ranged from 105-116%.

A secondary objective was to analyze caffeine using a recently manufactured caffeine Enzyme-Linked Immunosorbent Assay (ELISA) kit, and compare its effectiveness in the determination of caffeine in surface water samples to LC-MS/MS. Results from ELISA showed high concentrations of caffeine in the lake water samples ranging from 537 - 1080 ng/L but its performance was inferior to LC-MS/MS analysis; the assay had a lower reliability, sensitivity and precision in the analyses. The ELISA method has the advantages of quick assay requiring only a small amount of sample, relatively low expense, and simple technique. LC-MS/MS offers advantages in terms of simultaneous determinations of several analytes present in water samples, high sensitivity, and high quantitative reliability.

This document is currently not available here.