Measuring the phenolic content of dissolved organic matter in wetlands

Background/Question/Methods

Wetlands play an important role in the production and transformation of watershed-derived dissolved organic matter (DOM). Indicators suggest the amount of wetlands’ DOM has been increasing for several decades. Climate Change has been proposed as a potential contributor to the trend, and under this mechanism, the phenolic content of DOM may be experiencing a disproportionally large increase. We explored the possibility of assessing the phenolic character of DOM using multidimensional fluorescence spectroscopy as a more convenient alternative to “wet chemistry” methods. In this work, parallel factor analysis (PARAFAC) was applied to fluorescence excitation-emission matrices (EEMs) of humic samples in an attempt to analyze their phenolics.  PARAFAC results were correlated with phenol concentrations derived from the Folin-Ciocalteau reagent-based method. The reagent-based method showed that the phenolic content of five International Humic Substance Society (IHSS) DOM samples varied significantly (approximately 5 to 22 ppm Tannic Acid Equivalents) in phenol concentration. Results of the study show the potential for PARAFAC analysis of multidimensional fluorescence data to be a tool for monitoring phenolics in DOM.  Applications include assessing potential for the formation of disinfection byproducts upon chlorination/treatment of drinking water sources and monitoring the impact of Climate Change on the phenolic character of DOM.

Results/Conclusions

A five-component PARAFAC fit was applied to the EEMs of the IHSS sample dataset and it was determined by PARAFAC score correlations with phenol concentrations from the reagent-based method that components C1 (R²=0.78), C4 (R²=0.82), and C5 (R²=0.88) have the highest probability of containing phenolic groups. Furthermore, when the scores of components C4 and C5 were summed, the correlation improved (R²=0.99). Likewise, when the scores of C1, C4, and C5 were summed, their correlations were stronger than their individual parts (R²=0.89). Since the reagent-based method is providing an indicator of “total phenol” amount, regardless of the exact molecular structure of C1, C4, and C5, it seems reasonable that each of these components individually contributes a portion to the summed “total phenol” profile, and that the sum of their phenol-related spectral parts represents a larger portion of the “total phenol” index. However, when the sum of all five components were plotted against the reagent-based phenol concentrations, due to the considerable impact of largely non-phenolic components C2 (R²=0.23) and C3 (R²=0.35), the correlation was quite poor (or no correlation at all with R²=0.10).