Sarah E. Page

Assessing the Contribution of Free Hydroxyl Radical in Organic Matter-Sensitized Photohydroxylation Reactions

Photochemical formation of reactive oxygen species from dissolved organic matter (DOM) is incompletely understood, especially in the case of hydroxyl radical (•OH) production. Many studies have used various probes to detect photochemically produced •OH from DOM, but the fundamental reactions of these probes are not necessarily specific for free •OH and may also detect lower-energy hydroxylation agents. In this study, two tests were applied that have previously been used as a diagnostic for the presence of free •OH: methane quenching of •OH and hydroxybenzoic acid (hBZA) product yields. Upon application of these two tests to a set of five DOM isolates, it was found that methane quenching and the hBZA ratio results were not necessarily consistent. Overall, the results provide compelling evidence that all isolates studied photochemically produce free •OH. The hydroxylating acitivity of Elliot Soil Humic Acid and Pony Lake Fulvic Acid, however, also had a significant contribution from a photochemically generated hydroxylating agent that is lower in energy than free •OH. Catalase quenching experiments were conducted to assess whether hydrogen peroxide was the immediate precursor to hydroxyl in these systems. In all cases, catalase addition slowed photohydroxylation of terephthalate, but the contribution of hydrogen peroxide photolysis was determined to be less than 50%.

Terephthalate as a probe for photochemically generated hydroxyl radical

Hydroxyl radical ( OH) production in sunlit natural waters is known to be an important factor in pollutant degradation and nutrient cycling, and various probes have been developed to study this species in aquatic systems. Many of these probes have limitations in sensitivity and/or selectivity for OH, leaving room for improvement. Terephthalate (TPA) is a known probe for OH that produces a fluorescent product, hydroxyterephthalate (hTPA), upon hydroxylation, but its suitability for studying photochemical OH production has not been fully tested. To that end, the photochemical behavior of TPA and hTPA was characterized. TPA and hTPA react with OH with rate constants of (4.4 +/- 0.1) x 10(9) M(-1) s(-1) and (6.3 +/- 0.1) x 10(9) M(-1) s(-1), respectively. They were found to react with singlet oxygen ((1)O(2)) with significantly lower rate constants of <<10(4) M(-1) s(-1) and (5.0 +/- 0.1) x 10(4) M(-1) s(-1), respectively, indicating that TPA is selective for OH versus(1)O(2). TPA did not undergo direct photolysis, while hTPA was shown to undergo direct photochemical degradation with a Phi(direct,365nm) of (6.3 +/- 0.1) x 10(-3). TPA was applied to monitoring photochemical OH production by nitrate, nitrite and dissolved organic matter (DOM), and OH quenching rate constants measured for DOM were similar to results from previous studies. TPA provides a stable and sensitive probe for OH under significantly shorter photo-exposure times than similarly structured probe molecules, such as benzoate. However, the photoinstability of the analyte, hTPA, makes TPA an unsuitable probe for the study of photochemical systems under ultraviolet irradiation with wavelengths shorter than 360 nm.

Hydroxyl Radical Formation upon Oxidation of Reduced Humic Acids by Oxygen in the Dark

Humic acids (HAs) accept and donate electrons in many biogeochemical redox reactions at oxic/anoxic interfaces. The products of oxidation of reduced HAs by O(2) are unknown but are expected to yield reactive oxygen species, potentially including hydroxyl radical (·OH). To quantify the formation of ·OH upon oxidation of reduced HAs by O(2), three HAs were reduced electrochemically to well-defined redox states and were subsequently oxidized by O(2) in the presence of the ·OH probe terephthalate. The formation of ·OH upon oxidation increased with increasing extent of HA reduction. The yield of ·OH ranged from 42 to 160 mmol per mole of electrons donated by the reduced HA. The intermediacy of hydrogen peroxide (H(2)O(2)) in the formation of ·OH was supported by enhancement of ·OH formation upon addition of exogenous H(2)O(2) sources and by the suppression of ·OH formation upon addition of catalase as a quencher of endogenous H(2)O(2). The formation of ·OH in the dark during oxidation of reduced HA represents a previously unknown source of ·OH formation at oxic/anoxic interfaces and may affect the biogeochemical and pollutant redox dynamics at these interfaces.

Sensitive and selective time-gated luminescence detection of hydroxyl radical in water

A terbium probe for the selective time-gated detection of hydroxyl radical is presented. This probe demonstrates high sensitivity with an 11.0-fold increase in time-gated luminescence intensity for an hour-long exposure to hydroxyl radical concentration in the femtomolar range, and high selectivity for HO* over other reactive oxygen species.

Dark Formation of Hydroxyl Radical in Arctic Soil and Surface Waters

Hydroxyl radical (•OH) is a highly reactive and unselective oxidant in atmospheric and aquatic systems. Current understanding limits the role of DOM-produced •OH as an oxidant in carbon cycling mainly to sunlit environments where •OH is produced photochemically, but a recent laboratory study proposed a sunlight-independent pathway in which •OH forms during oxidation of reduced aquatic dissolved organic matter (DOM) and iron. Here we demonstrate this non-photochemical pathway for •OH formation in natural aquatic environments. Across a gradient from dry upland to wet lowland habitats, •OH formation rates increase with increasing concentrations of DOM and reduced iron, with highest •OH formation predicted at oxic-anoxic boundaries in soil and surface waters. Comparison of measured vs expected electron release from reduced moieties suggests that both DOM and iron contribute to •OH formation. At landscape scales, abiotic DOM oxidation by this dark •OH pathway may be as important to carbon cycling as bacterial oxidation of DOM in arctic surface waters.

Evidence for dissolved organic matter as the primary source and sink of photochemically produced hydroxyl radical in arctic surface waters

Hydroxyl radical (˙OH) is an indiscriminate oxidant that reacts at near-diffusion-controlled rates with organic carbon. Thus, while ˙OH is expected to be an important oxidant of dissolved organic matter (DOM) and other recalcitrant compounds, the role of ˙OH in the oxidation of these compounds in aquatic ecosystems is not well known due to the poorly constrained sources and sinks of ˙OH, especially in pristine (unpolluted) natural waters. We measured the rates of ˙OH formation and quenching across a range of surface waters in the Arctic varying in concentrations of expected sources and sinks of ˙OH. Photochemical formation of ˙OH was observed in all waters tested, with rates of formation ranging from 2.6 ± 0.6 to 900 ± 100 × 10(-12) M s(-1). Steady-state concentrations ranged from 2 ± 1 to 290 ± 60 × 10(-17) M, and overlapped with previously reported values in surface waters. While iron-mediated photo-Fenton reactions likely contributed to the observed ˙OH production, several lines of evidence suggest that DOM was the primary source and sink of photochemically produced ˙OH in pristine arctic surface waters. DOM from first-order or headwater streams was more efficient in producing ˙OH than what has previously been reported for DOM, and ˙OH formation decreased with increasing residence time of DOM in sunlit surface waters. Despite the ubiquitous formation of ˙OH in arctic surface waters observed in this study, photochemical ˙OH formation was estimated to contribute ≤4% to the observed photo-oxidation of DOM; however, key uncertainties in this estimate must be addressed before ruling out the role of ˙OH in the oxidation of DOM in these waters.

On the use of hydroxyl radical kinetics to assess the number-average molecular weight of dissolved organic matter.

Dissolved organic matter (DOM) is involved in numerous environmental processes, and its molecular size is important in many of these processes, such as DOM bioavailability, DOM sorptive capacity, and the formation of disinfection byproducts during water treatment. The size and size distribution of the molecules composing DOM remains an open question. In this contribution, an indirect method to assess the average size of DOM is described, which is based on the reaction of hydroxyl radical (HO(•)) quenching by DOM. HO(•) is often assumed to be relatively unselective, reacting with nearly all organic molecules with similar rate constants. Literature values for HO(•) reaction with organic molecules were surveyed to assess the unselectivity of DOM and to determine a representative quenching rate constant (k(rep) = 5.6 × 10(9) M(-1) s(-1)). This value was used to assess the average molecular weight of various humic and fulvic acid isolates as model DOM, using literature HO(•) quenching constants, kC,DOM. The results obtained by this method were compared with previous estimates of average molecular weight. The average molecular weight (Mn) values obtained with this approach are lower than the Mn measured by other techniques such as size exclusion chromatography (SEC), vapor pressure osmometry (VPO), and flow field fractionation (FFF). This suggests that DOM is an especially good quencher for HO(•), reacting at rates close to the diffusion-control limit. It was further observed that humic acids generally react faster than fulvic acids. The high reactivity of humic acids toward HO(•) is in line with the antioxidant properties of DOM. The benefit of this method is that it provides a firm upper bound on the average molecular weight of DOM, based on the kinetic limits of the HO(•) reaction. The results indicate low average molecular weight values, which is most consistent with the recent understanding of DOM. A possible DOM size distribution is discussed to reconcile the small nature of DOM with the large-molecule behavior observed in other studies.

Seasonal dynamics in dissolved organic matter, hydrogen peroxide, and cyanobacterial blooms in Lake Erie

Hydrogen peroxide (H2O2) has been suggested to influence cyanobacterial community structure and toxicity. However, no study has investigated H2O2 concentrations in freshwaters relative to cyanobacterial blooms when sources and sinks of H2O2 may be highly variable. For example, photochemical production of H2O2 from chromophoric dissolved organic matter (CDOM) may vary over the course of the bloom with changing CDOM and UV light in the water column, while microbial sources and sinks of H2O2 may change with community biomass and composition. To assess relationships between H2O2 and harmful algal blooms dominated by toxic cyanobacteria in the western basin of Lake Erie, we measured H2O2 weekly at six stations from June – November, 2014 and 2015, with supporting physical, chemical, and biological water quality data. Nine additional stations across the western, eastern, and central basins of Lake Erie were sampled during August and October, 2015. CDOM sources were quantified from the fluorescence fraction of CDOM using parallel factor analysis (PARAFAC). CDOM concentration and source were significantly correlated with specific conductivity, demonstrating that discharge of terrestrially-derived CDOM from rivers can be tracked in the lake. Autochthonous sources of CDOM in the lake increased over the course of the blooms. Concentrations of H2O2 in Lake Erie ranged from 47 ± 16 nM to 1570 ± 16 nM (average of 371 ± 17 nM; n = 225), and were not correlated to CDOM concentration or source, UV light, or estimates of photochemical production of H2O2 by CDOM. Temporal patterns in H2O2 were more closely aligned with bloom dynamics in the lake. In 2014 and 2015, maximum concentrations of H2O2 were observed prior to peak water column respiration and chlorophyll a, coinciding with the onset of the widespread Microcystis blooms in late July. The spatial and temporal patterns in H2O2 concentrations suggested that production and decay of H2O2 from aquatic microorganisms can be greater than photochemical production of H2O2 from CDOM and abiotic decay pathways. Our study measured H2O2 concentrations in the range where physiological impacts on cyanobacteria have been reported, suggesting that H2O2 could influence the structure and function of cyanobacterial communities in Lake Erie.

Corrigendum: Seasonal Dynamics in Dissolved Organic Matter, Hydrogen Peroxide, and Cyanobacterial Blooms in Lake Erie

“Concentrations of H2O2 in Lake Erie ranged from 2± 24 nM to 1140± 240 nM (average of 162± 11 nM; n= 221). Materials and Methods, Section 2.6 H2O2 Concentrations, Paragraph Number 1: Results, Section 3.4 H2O2 Concentrations in Lake Erie, Paragraph Number 1: Hydrogen peroxide concentrations in the surface waters of Lake Erie varied by over an order of magnitude during the study period, from 2 ± 24 to 1,140 ± 240 nM (average ± SE from triplicate measurements of each water sample), with an overall average of 162 ± 11 nM (average ± SE, n = 221; Figure 10D). Results, Section 3.4 H2O2 Concentrations in Lake Erie, Paragraph Number 3: The largest difference in H2O2 concentration between surface water and depth was observed when H2O2 at 21m was nearly double the surface concentration at the same site under mixed conditions (312 ± 53 nM vs. 148 ± 42 nM H2O2 in bottom vs. surface, respectively at site 949 in the central basin; Figure 8C). Discussion, Section 4, Paragraph Number 1: The average and range of hydrogen peroxide (H2O2) concentrations in Lake Erie (162 ± 11 nM, up to a maximum of 1,140 ± 240 nM; Table 1 and Figure 10D) were higher than the range previously observed at one station in the western basin of Lake Erie in August 1987 (100–200 nM; Cooper et al., 1989), but within the wide range of H2O2 concentrations observed in lakes (∼10 nM to >2μM; (Cooper and Zika, 1983; Cooper et al., 1989; Häkkinen et al., 2004; Febria et al., 2006; Khan et al., 2013; Scully et al., 1996). Discussion, Section 4, Paragraph Number 2: Using the average calculated photochemical production rate of 67 ± 3 nM h−1 H2O2 by CDOM, it would take almost 3 h of mid-day light to produce the observed average concentration of H2O2 (162 ± 11 nM) assuming no other sources and no sinks. All water samples were collected between 09:00 to 15:30 h, with the majority of samples collected between 10:00 h and 12:30 h, suggesting that there could have been sufficient time and UV light for photochemical production to generate the observed concentrations if H2O2 sinks were minimal. Discussion, Section 4, Paragraph Number 3: Photochemical production of H2O2 by CDOM could account for the observed H2O2 if CDOM in the lake consistently had ∼ 3-fold higher apparent quantum yields (8λ; Equation 3) than we used. Discussion, Section 4, Paragraph Number 7: For example, given that the depths of bottom water sampled (4– 61m) were also greater than the depth of UV light penetration (depth of 1% light was 1.5 ± 0.1m for 412 nm), there was not enough UV light to produce the 30–300 nM H2O2 observed at depth (Figure 8C). However, in this study, similar magnitudes of H2O2 concentrations were observed between surface and bottom waters even at depths >20m during stratified conditions (Figure 8C). Conclusions and Implications, Section 5, Paragraph Number 1: This study demonstrated that CDOM and H2O2 concentrations were not related and that even the upper estimates of photochemical production of H2O2 by CDOM were likely too low to account for all H2O2 (especially at depths below the photic zone). These results, combined with measured and estimated rates of biological production of H2O2 that can equal or exceed photochemical production (this study; Marsico et al., 2015), strongly suggest that biological activity contributes substantially to H2O2 concentrations in Lake Erie. Conflict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Copyright