Experiment 2: Conclusions

UV-visible spectroscopy is a valid, simple and cost effective method for determining the concentration of absorbing species if applied to pure compounds, and used with the appropriate standard curve. A standard curve relating absorbance to concentration can be developed for any compound, and used to determine the concentration of samples containing the same compound. The analysis should be done at a wavelength with maximum absorption, and located in relatively flat region of the spectra so that absorbance will be high and constant in a narrow range around the chosen wavelength. The optimal wavelength should ensure good absorbance of the analyte and low absorbance by other species in the solution. This wavelength will allow valid absorption measurements to be made on analyte samples that contain mixtures of materials.

These requirements make determination of TOC in natural water samples containing NOM difficult. This experiment showed that the use of a standard curve made from a single, pure compound is not appropriate for natural water samples. The heterogeneous nature of NOM makes representation by a single compound difficult as evidenced by our erroneous results for soil extract and wetland water samples. A more reliable standard curve would be made using actual samples. The TOC of various dilutions of the sample could be determined by a TOC analyzer and then related to the corresponding absorbance. This standard curve could be used reliably to determine the concentration in similar samples. Since NOM can vary structurally due to seasonal change or variable inputs, such a standard curve would have to be periodically verified, and probably would not remain valid over the entire year. The relationship, however, is purely empirical and cannot be applied to samples form other sources nor to the same samples if the relative quantities of absorbing constituents do not remain constant. Examples in the literature document the success of this type of analysis if done appropriately.

The acidification of standards and samples seemed to introduce error into the system as judging from the large confidence limits on the acidified sample concentrations. UV-visible spectroscopy can also be used to characterize humic substances in a general way as shown by similar E4/E6 ratios for NOM from different sources. However, only empirical correlations seem possible, and there is much disagreement over applying these correlations. Thus, further interpretation of our data would necessitate a more detailed study of the samples in order to determine molecular weight, source and % aromaticity. The characterized samples could be used to determine empirical relationships between the absorbance at various wavelengths. These relationships could be aided in identifying unknown characteristics of similar samples. Howver, they would be strictly sample specific and could not be used on a large range of NOM samples.