Sonochemical degradation of bisphenol A: a synergistic dual-frequency ultrasound approach Files / folders included in the dataset: BPA_Data Software required to open the files in the dataset: Excel Other information: Please use tabs in the excel sheet to navigate between different data sets. Shaun Fletcher,a Lukman A. Yusuf,a, Zeliha Ertekin,a and Mark D. Symes*a aWestCHEM, School of Chemistry, University of Glasgow, Glasgow, G12 8QQ, United Kingdom *Email: mark.symes@glasgow.ac.uk Reagents: Bisphenol A (?99% purity, Supelco, CAS Number: 80-05-7) was used in all experiments without further purification. For analytical calibration, five solutions with concentrations in the range of 10 - 30 mg L?1 were used. For degradation studies, aliquots of 20 mg L?1 (87.6 ?M) BPA solution were used as model pollutant samples, which were treated for 10, 20, 30, and 40 minutes under varying ultrasound conditions. For iodometric dosimetry, potassium iodide (99%, Alfa Aesar) was used. Experimental Setup: Ultrasonic irradiation of BPA solutions was achieved with two distinct devices operating at different frequencies. An Elmasonic P30H ultrasonic bath (120/100 W ultrasonic power effective at 37 kHz / 80 kHz, respectively) was used to generate switchable 37 kHz or 80 kHz frequencies. The bath was filled with 600 mL of deionized water, and a glass sonoreactor (built in-house from borosilicate glass, cylindrical in shape, with inner diameter = 50 mm, wall thickness = 3 mm, and height = 68 mm) containing 25 mL of sample was positioned ca. 10 mm from the base. In this way, ultrasonic energy was transferred to the sample indirectly via the surrounding water. A Branson 450 Digital Sonifier (400 W) was used to introduce 20 kHz ultrasound irradiation to the solution through a titanium probe (diameter = 6.35 mm, length = 230 mm). The tip of the probe was placed ca. 10 mm beneath the surface of the analyte solution to prevent any agitation or aerosoling at the top layer. Two dual frequency regimes, in which the bath operated at i) 37 kHz and ii) 80 kHz (in each case operating alongside the probe at 20 kHz), were used in this work. For the study of BPA degradation, experiments under each ultrasound regime were performed in triplicate. Additional analyses, whereby only one source of ultrasound was used for sonication, were performed to assess the synergistic effect. It is important to note that the input power of an ultrasonic device does not translate perfectly to the acoustic power dissipated in a sonicated sample. The acoustic power density afforded by the assembly used in this work has been estimated using calorimetry. The temperature of the liquid inside the sonoreactor was recorded for 200 seconds, with the rate of temperature increase then being used to derive the power density. The data for the relevant groupings of power and frequency are shown in Table 1. This serves as an approximation for the acoustic power afforded by the apparatus, and does not account for minor variations due to the inhomogeneous acoustic field or heat loss from the free surface. The ultrasonic probe was set at 80% of its maximum power in order to minimize any instabilities in the energy output over extended periods of time. Table 1: Acoustic power density delivered by various configurations of the sonication apparatus. The probe was always used at 80% of its maximum rated power and the bath was always used at 100% of its maximum rated power. Entry Probe frequency / kHz Bath frequency / kHz Acoustic Power Density / W L?1 1 20 - 146.08 ± 4.83 2 - 37 131.18 ± 2.14 3 - 80 87.03 ± 1.95 4 20 37 272.08 ± 3.92 5 20 80 209.63 ± 6.94 Thermal control of the reactor was necessary to mitigate unwanted heating of the sample during the sonication process. The temperature of a given sample was monitored throughout each experiment with a thermocouple (Pico Technology TC-08) and kept to a maximum of 25 °C with the periodic addition of fresh cold water to the ultrasonic bath. This is analogous to the use of cooling jackets employed in many examples of sonochemical reactors. Analytical methods: Measurement of BPA concentration in stock and treated samples was performed with an Agilent 1290 Infinity high performance liquid chromatograph (HPLC) attached to an Agilent 6125B Single Quad mass spectrometer (MS). A Phenomenex 15 cm C18 column was used as the stationary phase. The mobile phase was a 50/50 (v/v) mixture of 0.1% formic acid in (a) water and (b) acetonitrile (Fisher, HPLC grade), with a flow rate of 0.5 mL min?1 and a sample injection volume of 10 ?L. An ultraviolet detector monitoring 273 nm was used to identify fractions. The concentration of BPA in each treated sample was determined by the area of the characteristic chromatographic peak (at retention time, rt = 2.3 min) compared to the chromatograms of the stock solutions (see Supplementary Figure 1). Validation of the transformation products of the sonochemical process was achieved with an Agilent 6546 Q-TOF-MS high resolution accurate mass spectrometer. To quantify the production of oxidative species by cavitation bubbles, dosimetry with potassium iodide solution was employed. The oxidation of iodide in solution generates molecular iodine (Equation 1). When excess iodide ions are present, molecular iodine forms the triiodide ion (I3?) (Equation 2). 2I? ??" oxidation " I2 (1) I2 + I? ? I3? (2) Triiodide exhibits a strong absorbance peak at 350 nm. In the context of ultrasonic irradiation of KI, the concentration of I3? serves as an analogue for the oxidising capability of the system via reactive oxygen species generated by the processes in Equations 1-4. For each of the combinations of ultrasonic bath and horn used in this work, 20 mL of 0.1 M KI solution was subjected to ultrasonic irradiation for 5 minutes. The concentration of triiodide was determined by measuring the 350 nm absorbance peak of the treated solutions with an Agilent Cary 60 UV–visible spectrophotometer. A calibration of absorbance at 350 nm vs triiodide concentration is provided in Supplementary Figure 2. To determine the extent of BPA degradation during ultrasonic treatment, chemical oxygen demand (COD) techniques were used. A photometer (Hanna Instruments, HI-97106) with accompanying reagent test kits (Hanna Instruments, HI-93754D-25) was used to obtain the COD (in mg L?1) of treated BPA samples. The COD removal was calculated with Equation 3 by comparing the COD to that of the stock solution. ?("Extent of chemical oxygen demand removal = 1" -(("CO" "D" _"sample" )/("CO" "D" _"stock" ))" " #(3) ) Supplementary Figure 1: Calibration of BPA analytical methods. LC-MS calibration for quantitative BPA analysis: HPLC chromatograms of 10 mg L?1 (red), 15 mg L?1 (blue), 20 mg L?1 (green), 25 mg L?1 (purple) and 30 mg L?1 (orange) solutions of BPA with absorbance at ? = 273 nm are plotted vs. retention time (a). The area of the characteristic peak (rt = 2.3 min) is plotted vs. concentration of BPA, with linear regression used to define the relationship (b). The corresponding mass spectrum for the characteristic peak of the chromatogram shows signals for both the singly deprotonated ([M-H]?, m/z = 227) and doubly deprotonated ([M-2H]2?, m/z = 113) molecular ion of BPA (c). Supplementary Figure 2: Calibration of iodide dosimetry. UV-visible spectroscopy calibration for triiodide: UV-visible spectra of stock solutions of triiodide (a). The absorbance at ? = 350 nm is plotted vs. concentration of triiodide, with linear regression used to define the relationship (b). Supplementary Figure 3: Chromatograms of degraded bisphenol A (37/20 kHz). Representative HPLC chromatograms of 20 mg L?1 BPA solutions following treatment with 37/20 kHz dual frequency ultrasound. Absorbance at ? = 273 nm is plotted vs. retention time in the case of five samples treated for 0 – 40 mins, with chromatograms presented in series with increasing sonication time. The grid line at rt = 2.3 min is provided as a reference for the characteristic chromatographic peak of BPA. Supplementary Figure 4: Chromatograms of degraded bisphenol A (80/20 kHz). Representative HPLC chromatograms of 20 mg L?1 BPA solutions following treatment with 80/20 kHz dual frequency ultrasound. Absorbance at ? = 273 nm is plotted vs. retention time in the case of five samples treated for 0 – 40 mins, with chromatograms presented in series with increasing sonication time. The grid line at rt = 2.3 min is provided as a reference for the characteristic chromatographic peak of BPA. Supplementary Figure 5: Rate analysis of bisphenol A removal. Pseudo first-order rate analysis for the removal of bisphenol A by dual frequency ultrasonic treatment. Rate constants have been calculated as the slope parameter of linear regression modelling of ln (C0/C) versus time in each case, determined to be 0.064 ± 0.004 min?1 and 0.048 ± 0.002 min?1 for treatment with 37/20 kHz (blue) and 80/20 kHz (burgundy) frequencies, respectively. Supplementary Figure 6: Single frequency degradation rate analysis. Pseudo first-order rate analysis for the removal of bisphenol A by single frequency ultrasonic treatment. Rate constants have been calculated as the slope parameter from linear regression modelling of ln (C0/C) vs. time in each case, determined to be 0.0034 ± 0.0002 min?1 for 20 kHz ultrasound (a), 0.0056 ± 0.0002 min?1 for 37 kHz ultrasound (b), and 0.0025 ± 0.0002 min?1 for 80 kHz ultrasound. Supplementary Figure 7: Iodide dosimetry of sonoreactor. The average UV-visible spectra for 0.1 M potassium iodide solutions following sonication for 5 mins with 80 kHz (green), 37 kHz (red), 20 kHz (blue), 80/20 kHz (orange), and 37/20 kHz (purple) frequencies are plotted vs. wavelength. The absorbance at ? = 350 nm was used to quantify the concentration of triiodide formed via oxidation of iodide by in situ generated reactive oxygen species. Supplementary Figure 8: Mass spectrometry of degraded bisphenol A. Full mass spectrum of degraded BPA (after 37/20 kHz irradiation for 40 mins) (a), with selected portions of the spectrum enlarged to show the molecular ion (m/z = 227) (b), in addition to the proposed oxidative degradation products with m/z = 241 and 243 (c), m/z = 257 and 259 (d), m/z = 275 (e), m/z = 151 (f), m/z = 133 and 135 (g), and m/z = 115 (h). The most likely structures of the molecules corresponding to the observed signals are inset for reference.