Abstract
Bisphenol A (BPA) is a compound widely used in plastics and other industrial consumer products. However, BPA has been found to show toxicity and endocrine disrupting effects and so the presence of BPA in consumer products is of great concern. As current detection methods of BPA are slow, a fast and reliable technique to measure BPA is needed. It has been reported that BPA shows fluorescence in a variety of organic solvents and thus it could be determined spectrofluorometrically. But the fluorescence of BPA is too low to be detected easily in aqueous solution (Olmo and Zafra, 2000). As BPA is commonly found in water a method for using florescence to measure BPA in aqueous solvent could be useful. The first part of this thesis studied the solvent effects on the florescence of BPA which was carried out by measuring the fluorescence of BPA solutions with varying percentage of acetonitrile in water. It was found that BPA does show weak fluorescence in water, especially compared to ACN. This study also found that the ratio of water and ACN in the solvent has little to no effect on stokes shift of the fluorescence emission peak of BPA but does affect fluorescence intensity. If BPA is to be detected using fluorescence, selective fluorescence methods must be developed to distinguish BPA from interfering compounds with similar spectral behavior. In the second part of this thesis constant wavelength synchronous fluorescence spectroscopy (CWSFS) was used to find a pair of excitation and emission wavelengths that can be used to detect BPA without interference. The fluorescence of BPA and its analogs in ACN solutions were compared using CWSFS and constant wavelength analysis. A selective pair of wavelengths was found with an excitation λ 280 nm and emission λ 304nm. Once the selective wavelength pair was selected analyses of BPA in three different brands of bottled water samples in PET plastic bottles was carried out. Liquid-liquid extraction (LLE) was used for pre-treatment of the samples. For each brand one sample was spiked with BPA standard to determine recovery of the method. After LLE the reconstituted extracts were measured using fluorescence with the selective wavelength pair as well as a traditional wavelength pair used to detect BPA in ACN solutions. The reconstituted extracts were also analyzed using HPLC, which is an established method for detection of BPA. Both the HPLC method and the fluorescence methods were found to be valid producing calibration curves with R2 values above 0.999. The recoveries for the method were found to be very high, ranging from 98.1% to 132.3%. The calculated BPA concentrations were found to vary greatly for the HPLC and the fluorescence methods. The analytic results using the selective fluorescence wavelengths were found to be much closer to those by the HPLC than by the traditional wavelengths method. It was also found that the analytic accuracy was high at higher BPA concentrations. Thus, this means the method can be greatly improved upon simply by increasing the pre-concentration of BPA in water sample extracts.