Article pubs.acs.org/ac
Novel Fluorometric Method for the Determination of Production Rate and Steady-State Concentration of Photochemically Generated Superoxide Radical in Seawater Using 3′,6′(Diphenylphosphinyl)fluorescein Adebanjo Jacob Anifowose, Kazuhiko Takeda, and Hiroshi Sakugawa* Graduate School of Biosphere Science, Department of Environmental Dynamics and Management, Hiroshima University, 1-7-1 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8521, Japan S Supporting Information *
ABSTRACT: Superoxide radical (O2•−) is an important reactive oxygen species in seawater. Measurements of its production rates and steady-state concentrations generated by photochemical processes have been a Herculean task over the years. In this study, a probe − 3′6′-(diphenylphosphinyl)fluorescein (PF-1) − was used to trap photochemically generated O2•− in seawater, thereby yielding fluorescein. The fluorescein produced was measured by an isocratic fluorescence HPLC at excitation/emission wavelengths of 490/513 nm, respectively. The reaction rate constant of PF-1 with O2•− (kPF‑1) was pH-dependent: (3.2−23.5) × 107 M−1 s−1 at pHTOT 7.65−8.50. By applying appropriate equations, both the production rate and the steady-state concentration of O2•− generated by photochemical reactions in the seawater were quantified. Under the optimized experimental conditions, fluorescein standards (3−50 nM) exhibited linearity in the seawater by HPLC. The photoformation of fluorescein, due to the reaction of PF1 with the O2•− photochemically produced in the seawater, was linear within the 20 min irradiation. The detection limit of the fluorescein photoformation rate was 0.03 pM s−1, defined as 3σ of the lowest standard fluorescein concentration per 20 min irradiation. Using this value, the yield of fluorescein, and the fraction of O2•− that reacted with PF-1 in the seawater, the detection limit of the O2•− photoformation rate was 1.78 pM s−1. Superoxide measurements using the proposed method were relatively unaffected by the potential interfering species in seawater. Application of the proposed method to ten (10) seawater samples from the Seto Inland Sea, Japan, resulted in measured O2•− photoformation rates of 3.1−8.5 nM s−1, with steady-state concentrations ranging (0.06−0.3) × 10−10 M. The method is simple, requires no technical sample preparation, and can be used to analyze a large number of samples. uperoxide radical (O2•−) is one of the most important reactive oxygen species in natural waters. It is an intermediate in the oxygen redox chemistry in seawater. It is produced by photochemical processes1−3 as well as biological processes.4−8 In the sunlit surface of the ocean, superoxide is predominantly produced by photochemical reactions involving oxygen and dissolved organic matter/colored dissolved organic matter (DOM/CDOM). Superoxide concentrations in seawater have been reported to range from picomolar to nanomolar,1,5,6,9−11 while production rates between picomolar and nanomolar per second have also been reported.2,3,6 In seawater, O2•− mediates in redox changes involving various chemical species due to its ability to act as an oxidant and a reductant. Therefore, the presence of O2•− could influence the biogeochemical fates of metals, organic substrates, and other reactive chemical species in seawater. The reaction of O2•− with some organic substrates and metal ions, notably, ions of Fe, Cu, and Mn,12−17 its disproportionation (catalyzed and uncata-
S
© XXXX American Chemical Society
lyzed) to form H2O2, short lifetime,5,18 and low concentration (≤ nM) constitute major analytical challenges to direct and routine measurement of superoxide radical in seawater. There have been some existing methods for the determination of O2•− concentration in aqueous solutions. These include direct spectrometric detection of O2•− absorbance at 245−252 nm in an alkaline medium (pH ≥ 11)19 and spectrometric detection of O2•− absorbance after coupling with a suitable reagent, resulting in absorption change.20 Others include measurement of the fluorescent product at appropriate excitation/emission wavelengths after reaction of O2•− with a specific chemical reagent,21 chemiluminescence detection with the use of luminol (though lacks specificity and not recommended),22 and 2-methyl-6-(p-methoxyphenyl)-3,7Received: March 9, 2015 Accepted: November 16, 2015
A
DOI: 10.1021/acs.analchem.5b00917 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry dihydroimidazol[1,2-α]-pyrazin-3-one (MCLA).23 Except for the use of MCLA, most of these methods lack sensitivity or specificity during field application for direct O2•− measurement in natural waters. Nevertheless, the use of MCLA in the realtime measurement of the O2•− photoformation rate in seawater has not been reported. Until now, the only published method for the direct measurement of O2•− photoformation rates in natural seawater employed the use of isotopically labeled 15NO in trapping the O2•− photogenerated after which concentration of the stable 15NO3− product was measured by mass spectrometry.2 Though the method is specific for the measurement of O2•− photoformation rate, its application to field study and large number of samples is limited due to the laborious techniques involved. Most of the other published data on O2•− photoformation rates were estimated based on either H2O2 photoformation rates in the seawater or O2•− decay rates in the dark. Recently, Powers and Miller1 were able to estimate midday steady-state concentration of O2•− due to photochemical reactions in the open ocean by blending remote sensing data products into a model. Nevertheless, accurate rates of O2•− photoformation in the seawater were elusive as the estimation was based on the assumption that the O2•− photoformation rates were twice that of the measured H2O2 photoproduction rates. Therefore, the need to develop a less complicated and highly sensitive analytical method to quantify both the formation rates and steady-state concentrations of photochemically generated O2•− in seawater without mere estimation cannot be overemphasized. Xu et al.24 developed a phosphinate-based probe compound − 3′6′-(diphenylphosphinyl)fluorescein (PF-1) − for the detection of the superoxide radical in the biological system. According to the study, O2•− reacts with PF-1 to yield fluorescein (Figure 1). The reported sensitivity (detection limit,
analytical grade. The seawater samples used in this study were collected from the Seto Inland Sea during the Hiroshima University’s Toyoshio-Maru Sea Cruise in June, 2014 (except those used for the salinity study which were collected in June, 2015). A sample collection was achieved with Niskin sampling bottles and a Sea-Bird CTD carousel multisampling system (SBE-9Pplus; General Oceanic Inc., USA). The samples were filtered using a precombusted glass fiber filter (Advantec GC50, 0.45 μm nominal rating) into acid-clean 250 mL amber glass bottles and stored in the dark at 4 °C until analysis. A primary stock solution of 10 mM Cu2+ was prepared by dissolving anhydrous CuCl2 in 2 mM HCl. Fluorescein (0.3 mM) was prepared by dissolving an appropriate amount in 0.1 M KOH. An aliquot of the alkaline solution of fluorescein was neutralized with 0.1 M HCl to obtain 6 μM fluorescein stock solution. PF-1 is hydrophobic but soluble in polar aprotic solvents like dimethylformamide (DMF) and dimethyl sulfoxide (DMSO). Therefore, 2 mM PF-1 stock solution was prepared in DMF. Preliminary studies on fluorescein and PF-1 (after appropriate dilutions in Milli-Q water) using a fluorescence spectrophotometer (Hitachi F-7000) confirmed that fluorescein has excitation/emission wavelengths at 490/ 513 nm, respectively, while PF-1 showed no fluorescence property. However, the UV−vis absorbance spectra (UV2400PC spectrophotometer, Shimadzu) of PF-1 showed maximum absorption wavelength (λmax) at 272 nm. pH Measurement. Seawater pH values in this study are reported using the total hydrogen scale (pHTOT), calibrated with Tris buffers.25 pH of the Tris buffers was standardized spectrophotometrically using cresol red according to Millero et al.26 (the multiplication sign in their eq 25 was corrected by replacement with the division sign). The pH value of the low ionic strength buffer (borate buffer) is reported on the NBS scale. All pH measurements were made with a Horiba D-24 pH meter. Photochemical Experiment. In order to mimic the natural solar system during all the photochemical experiments, a solar simulator (Oriel model 81160-1000, Oriel Corp.) equipped with a 300 W Xe lamp (ozone-free, model 6258, Oriel Corp.) was used. Wavelengths
