A molecular probe, 3-amino-2,2,5,5,-tetramethy-1-pyrrolydinyloxy (3ap), was employed to determine the formation rates of one-electron reducing intermediates generated photochemically from both untreated and borohydride-reduced Suwanee River fulvic and humic acids (SRFA and SRHA, respectively). This stable nitroxyl radical reacts rapidly with reducing radicals and other one-electron reductants to produce a relatively stable product, the hydroxylamine, which can be derivatized with fluorescamine, separated by HPLC and quantified fluorimetrically. We provide evidence that O2 and 3ap compete for the same pool(s) of photoproduced reducing intermediates, and that under appropriate experimental conditions, the initial rate of hydroxylamine formation (RH) can provide an estimate of the initial rate of superoxide (O2–) formation. However, comparison of the initial rates of H2O2 formation (RH2O2) to that of RH show far larger ratios of RH/RH2O2 (∼6–13) than be accounted for by simple O2– dismutation (RH/RH2O2 = 2), implying a significant oxidative sink of O2– (∼67–85%). Because of their high reactivity with O2– and their likely importance in the photochemistry of CDOM, we suggest that coproduced phenoxy radicals could represent a viable oxidative sink. Because O2–/phenoxy radical reactions can lead to more highly oxidized products, O2– could be playing a far more significant role in the photooxidation of CDOM than has been previously recognized.

The hydroxyl radical (•OH) is the most reactive oxidant produced in natural waters. Photoproduction by chromophoric dissolved organic matter (CDOM) is one of its main sources, but the structures responsible for this production remain unknown. Here, a series of substituted phenol model compounds are examined to test whether these structures could act as a source of •OH. We find that many of these compounds do produce •OH with quantum yields (Φ) ranging from ∼10–4 to ∼10–2. In particular, two compounds that have hydroxy groups and carboxyl groups in a para relationship (4-hydroxybenzoic acid and 2,4-dihydroxybenzoic acid) exhibit relatively high Φ values, ∼10–2. For 2,4-dihydroxybenzoic acid, the formation of •OH was confirmed through the use of competition kinetics and reaction with methane. We conclude that these types of structures, which may derive from polyphenolic source materials such as lignins, tannins, and humic substances, could be an important source of •OH in natural waters.

Addition of a series of phenol electron donors to solutions of humic substances (HS) enhanced substantially the initial rates of hydrogen peroxide (H2O2) photoproduction (RH2O2), with enhancement factors (EF) ranging from a low of ∼3 for 2,4,6-trimethylphenol (TMP) to a high of ∼15 for 3,4-dimethoxyphenol (DMOP). The substantial inhibition of the enhanced RH2O2 following borohydride reduction of the HS, as well as the dependence of RH2O2 on phenol and dioxygen concentrations are consistent with a mechanism in which the phenols react with the triplet excited states of (aromatic) ketones within the HS to form initially a phenoxy and ketyl radical. The ketyl radical then reacts rapidly with dioxygen to regenerate the ketone and form superoxide (O2–), which subsequently dismutates to H2O2. However, as was previously noted for the photosensitized loss of TMP, the incomplete inhibition of the enhanced RH2O2 following borohydride reduction suggests that there may remain another pool of oxidizing triplets. The results demonstrate that H2O2 can be generated through an additional pathway in the presence of sufficiently high concentrations of appropriate electron donors through reaction with the excited triplet states of aromatic ketones and possibly of other species such as quinones. However, in some cases, the much lower ratio of H2O2 produced to phenol consumed suggests that secondary reactions could alter this ratio significantly.

Absorption of sunlight by chromophoric dissolved natural organic matter (CDOM) is environmentally significant because it controls photic zone depth and causes photochemistry that affects elemental cycling and contaminant fate. Both the optics (absorbance and fluorescence) and photochemistry of CDOM display unusual properties that cannot easily be ascribed to a superposition of individual chromophores. These include (i) broad, unstructured absorbance that decreases monotonically well into the visible and near IR, (ii) fluorescence emission spectra that all fall into a single envelope regardless of the excitation wavelength, and (iii) photobleaching and photochemical quantum yields that decrease monotonically with increasing wavelength. In contrast to a simple superposition model, these phenomena and others can be reasonably well explained by a physical model in which charge-transfer interactions between electron donating and accepting chromophores within the CDOM control the optical and photophysical properties. This review summarizes current understanding of the processes underlying CDOM photophysics and photochemistry as well as their physical basis.

The mass spectra acquired by ESI FT-ICR MS of untreated, borohydride-reduced, and borodeuteride-reduced samples of Suwannee River fulvic acid (SRFA) and a C18 extract from the upper Delaware Bay were compared to one another. Treatment of these samples with sodium borodeuteride was shown to produce unique mass labels for species which contain one or two ketone/aldehyde moieties. Approximately 30% of all identified peaks in the two samples were shown to comprise ketone/aldehyde-containing species. The molecular formulas of the majority of these species had O/C and H/C molar ratios typically attributed to lignin-derived compounds and/or carboxylic rich alicyclic molecules (CRAM). However, the significant loss of UV–vis absorption following reduction supports a lignin-based origin for the optical (and photochemical) properties of these samples. The mass-labeling method described and tested herein shows great promise as a means to further characterize the structure and composition of complex natural samples, especially in terms of identifying specific subsets of chemical species that contribute significantly to the optical and photochemical properties of such samples.

Extensive data exist on the optical properties of CDOM from terrestrial and coastal environments, yet the open oceans have been historically under-sampled. Consequently, the source and structural basis of marine CDOM optical properties are still debated. To address this issue, detailed optical measurements were acquired for both untreated and sodium borohydride (NaBH4) reduced natural waters and C18 extracts (C18-OM) across the Equatorial Atlantic Ocean. Except in regions of upwelling or in the vicinity of the Congo River outflow, CDOM absorption coefficients and visible emission intensity were far smaller for surface waters (aCDOM(355): 0.057–0.162 m− 1; λex/λem = 350/450 nm: 0.396–1.431 qse) than for waters below the mixed layer (aCDOM(355) 0.084–0.344 m− 1; λex/λem = 350/450 nm: 0.903–3.226 qse), while spectral slopes were higher (surface: 0.019 to 0.025 nm− 1; deep: 0.013 to 0.019 nm− 1), consistent with photobleaching of CDOM in surface waters. Distinct emission bands were observed in the ultraviolet, primarily at excitation/emission wavelengths (λex/λem) = 280/320 nm, but also at λex/λem = 300/340, 300/405 and 320/380 nm for some stations and depths. In contrast, visible emission exhibited maxima that continuously redshifted with increasing λex (> 330 nm), a property characteristic of CDOM from estuarine and coastal environments. Further evidence that CDOM in the offshore waters of this region is composed of a major terrestrial component includes: 1) similar spectral dependencies of the emission maxima and fluorescence quantum yields; 2) a large Stokes shift in the emission maxima with short-wavelength excitation (λex = 280 nm); 3) correlation of visible emission intensities with absorption at λex = 280, 320 and 450 nm, with absorption to fluorescence ratios comparable to those found in estuarine and coastal environments; 4) affinity of C18 cartridges for the long wavelength (visible) absorbing and emitting material, but not the UV emitting material; 5) preferential loss of visible absorption and substantially enhanced blue-shifted emission in the visible following borohydride reduction of both the Equatorial Atlantic waters and the C18-OM of these waters. These results support the occurrence in offshore waters of a major terrestrial CDOM component that absorbs in the UV and visible and emits in the visible, as well as marine CDOM components that absorb and emit in the UV. The results further demonstrate that the simultaneous acquisition of complete spectral absorption and emission properties, combined with chemical tests (C-18 extractions, borohydride reduction) can provide a far clearer picture of the sources and cycling of CDOM within the oceans.Highlights► Complete optical properties of CDOM and C18-OM were acquired across the Equatorial AO. ► C18 extraction and NaBH4 reduction impact on oceanic CDOM optical properties was probed. ► Region exhibits properties similar to those of other terrestrial/coastal regions. ► Data provides evidence of a major terrestrial CDOM component at this locale.

The mechanism(s) by which hydrogen peroxide (H2O2) is photoproduced by humic substances and chromophoric dissolved organic matter was probed by examining the dependence of the initial H2O2 photoproduction rate (RH2O2) and apparent H2O2 quantum yields on dioxygen concentration for both untreated and borohydride-reduced samples. Although borohydride reduction substantially reduced light absorption, the RH2O2 values were largely unaffected. Apparent monochromatic and polychromatic quantum yields thus increased following reduction. The results indicate that light absorption by charge-transfer states or by (aromatic) ketone/aldehydes does not lead to significant H2O2 photoproduction. High concentrations of triplet quenchers relative to that of dioxygen produced only small decreases (sorbic acid) or small increases (Cl– and Br–) in RH2O2, indicating that neither 1O2 nor excited triplet states of quinones contribute significantly to H2O2 photoproduction. The dependence of RH2O2 on O2 concentration provides evidence that the intermediate(s) reacting with O2 to produce superoxide are relatively long-lived (approximately tens of microseconds or more). Evidence of the photochemical formation of O2-reducing intermediates under anaerobic conditions was also obtained; these reducing intermediates appeared to be relatively stable in the absence of O2. Our data suggest that these O2-reducing intermediates are generated by intramolecular electron transfer from short-lived excited states of electron donors to ground-state acceptors.

To probe the mechanism of the photosensitized loss of phenols by humic substances (HS), the dependence of the initial rate of 2,4,6-trimethylphenol (TMP) loss (RTMP) on dioxygen concentration was examined both for a variety of untreated as well as borohydride-reduced HS and C18 extracts from the Delaware Bay and Mid-Atlantic Bight. RTMP was inversely proportional to dioxygen concentration at [O2] > 50 μM, a dependence consistent with reaction with triplet excited states, but not with 1O2 or RO2. Modeling the dependence of RTMP on [O2] provided rate constants for TMP reaction, O2 quenching, and lifetimes compatible with a triplet intermediate. Borohydride reduction significantly reduced TMP loss, supporting the role of aromatic ketone triplets in this process. However, for most samples, the incomplete loss of sensitization following borohydride reduction, as well as the inverse dependence of RTMP on [O2] for these samples, suggests that there remains another class of oxidizing triplet sensitizer, perhaps quinones.

Treatment of Suwanee River humic (SRHA) and fulvic (SRFA) acids, a commercial lignin (LAC), and a series of solid phase extracts (C18) from the Middle Atlantic Bight (MAB extracts) with sodium borohydride (NaBH4), a selective reductant of carbonyl-containing compounds including quinones and aromatic ketones, produces a preferential loss of visible absorption (≥50% for SRFA) and substantially enhanced, blue-shifted fluorescence emission (2- to 3-fold increase). Comparison of the results with those obtained from a series of model quinones and hydroquinones demonstrates that these spectral changes cannot be assigned directly to the absorption and emission of visible light by quinones/hydroquinones. Instead, these results are consistent with a charge transfer model in which the visible absorption is due primarily to charge transfer transitions arising among hydroxy- (methoxy-) aromatic donors and carbonyl-containing acceptors. Unlike most of the model hydroquinones, the changes in optical properties of the natural samples following NaBH4 reduction were largely irreversible in the presence of air and following addition of a Cu2+ catalyst, providing tentative evidence that aromatic ketones (or other similar carbonyl-containing structures) may play a more important role than quinones in the optical properties of these materials.

Fluorescamine derivatized 3-amino-2,2,5,5,-tetramethyl-1-pyrrolidinyloxy (I) is shown to undergo an irreversible reaction with peroxyl radicals and other radical oxidants to generate a more highly fluorescent diamagnetic product (II) and thus can be used as a highly sensitive and versatile probe to determine oxidant production optically, either by monitoring the changes in fluorescence intensity, by HPLC analysis with fluorescence detection, or by a combination of both approaches. By changing the [O2]/[I] ratio, we show that peroxyl radicals can be detected and quantified preferentially in the presence of other radical oxidants. Detection of photochemically produced peroxyl radicals is achieved by employing 3-amino-2,2,5,5,-tetramethyl-1-pyrrolidinyloxy (3-ap) alone, followed by derivatization with fluorescamine. With employment of HPLC analysis, the detection limit of II at a S/N of 2 is ∼3 nM for a 125 μL injection. Preliminary applications include the detection of peroxyl radicals generated thermally in soybean phosphatidylcholine liposomes and produced photochemically in tap water.

The spectral dependencies of absorption and fluorescence emission (emission maxima (λmax), quantum yields (ϕ), and mean lifetimes (τm)) were acquired for a commercial lignin, Suwannee River humic (SRHA) and fulvic (SRFA) acids, and a series solid phase extracts (C18) from the Middle Atlantic Bight (MAB extracts). These parameters were compared with the relative average size and total lignin phenol content (TLP). TLP was strongly correlated with absorption at 280 and 355 nm for the MAB extracts, SRHA, and SRFA. The spectral dependence of λmax, ϕ, and τm was very similar for all samples, suggesting a common photophysical and thus structural basis. A strong decrease of ϕ and τm with increasing average size indicates that intramolecular interactions must be important. When combined with previous work, the results lead us to conclude that the optical properties commonly associated with terrestrial humic substances and chromophoric dissolved organic matter arise primarily from an ensemble of partially oxidized lignins derived from vascular plant sources. They further provide additional support for an electronic interaction model in which intramolecular energy transfer, excited-state electron transfer, as well as charge transfer likely play important roles in producing the observed optical and photochemical properties of these materials.

Absorption of sunlight by chromophoric dissolved natural organic matter (CDOM) is environmentally significant because it controls photic zone depth and causes photochemistry that affects elemental cycling and contaminant fate. Both the optics (absorbance and fluorescence) and photochemistry of CDOM display unusual properties that cannot easily be ascribed to a superposition of individual chromophores. These include (i) broad, unstructured absorbance that decreases monotonically well into the visible and near IR, (ii) fluorescence emission spectra that all fall into a single envelope regardless of the excitation wavelength, and (iii) photobleaching and photochemical quantum yields that decrease monotonically with increasing wavelength. In contrast to a simple superposition model, these phenomena and others can be reasonably well explained by a physical model in which charge-transfer interactions between electron donating and accepting chromophores within the CDOM control the optical and photophysical properties. This review summarizes current understanding of the processes underlying CDOM photophysics and photochemistry as well as their physical basis.