Photophysics and photochemistry of natural waters with emphasis on radical probe development and application

Abstract

The work presented here consists of a literature review and calculations to estimate the importance of photochemistry to carbon cycling in the oceans, followed by a photophysical study of a series of stable nitroxide radical probes that have been used for the quantitative detection of individual carbon-centered radicals and reducing species in natural waters. Two appendices follow. The first contains preliminary experiments utilizing one of the nitroxide probes in an investigation of hydroxyl radical production rates and steadystate concentrations in seawater. The second consists of an investigation of the singlet lifetimes of humic acids (HA), in order to aid in understanding their photochemical cycling and nt1uence on other compounds. The impact of photochemical reactions on global oceanic carbon cycling was calculated from literature values. The results indicate that between 1 and 13% of all dissolved organic carbon in the oceans is oxidized photochemically. This is a significant flux term, much larger than that of riverine input for example. A photophysical study of nitroxide radical probes was undertaken. For all of the compounds studied, steady-state absorption and fluorescence spectra were identical to those of the parent fluorophores. A decrease in fluorescence lifetime and quantum yield of tens- to hundreds-fold was observed for the paramagnetic compounds relative to their diamagnetic counterparts. Very rapid fluorescence quenching rates (3 to 80 x 1010 s-1) were calculated for the fluorescamine moiety of the paramagnetic nitroxide compounds in a variety of solvents. Calculated energy minimized geometries were very similar for all compounds which implies that geometric differences are not responsible for the variations found m fluorescence lifetimes and quantum yields between compounds. Calculated Forster and Dexter overlap integrals do not support deexcitation by these mechanisms. Time-resolved absorption measurements resulted in no evidence for transient species due to either intersystem crossing to the triplet state or charge transfer. Of the mechanisms considered, direct internal conversion to the ground state, is most likely given our results. An investigation of the utility of 3-(aminomethyl)-2,2,5,5-tetramethyl-1- pyrrolidinyloxy free radical (3-amp) for detection and quantification of hydroxyl radicals in natural waters found that the addition of primary probe compounds resulted in the generation of secondary carbon-centered radicals that were successfully trapped by 3-amp. Competition kinetics experiments with dimethyl sulfoxide resulted in a natural scavenger rate constant that matched previous literature results for coastal seawater. As expected, the addition of formate resulted in decreases, and the addition of nitrite in increases, in the hydroxyl radical trapping rate by this method. The resulting quantum yield values were about an order of magnitude higher than previous literature results. However, probably due to the use of different latitudes at which to estimate the incident solar radiation at the sea surface, hydroxyl radical production rate and steady-state concentrations calculated were about an order of magnitude lower than literature results. One experiment showed no increase in the hydroxyl radical production rate from Milli-Q water to oligotrophic and coastal seawater although the sample absorption coefficients increase by a factor of more than 20. However a single experiment comparing three different coastal seawater samples did show a correlation between absorption and hydroxyl radical production rate. More detailed work is needed to recognize the full potential of this method. Marine HA fluorescence lifetime measurements utilizing time-resolved single photon counting revealed a large portion of chromophores with very short (20-60 ps) lifetimes and low quantum yields. At least three distinct lifetimes could be distinguished by iterative deconvolution, although they probably result from the grouping of a multitude of individual chromophores. The theory of calculating the quantum yields of individual chromophores measured in a mixture is developed and calculations are made, although from an incomplete data set. Shorter fluorescent lifetimes for a given chromophore center within HA result in smaller quantum yields and are thought to be caused by very rapid competing intramolecular dark pathways such as energy or electron transfer Preliminary work investigating changes in time-resolved fluorescent lifetimes due to different sources of HA (Orinoco vs. Suwanee Rivers) and solution types (seawater vs. standard buffer) showed little variability.