31P FT-NMR of Concentrated Lake Water Samples

Abstract

The use of phosphorus-31 Fourier Transform nuclear magnetic resonance (31P FT-NMR) spectroscopy for the study of dissolved organic phosphorus (DOP) in fresh water has been recently established by Nanny and Minear. The fact that NMR is an element-specific technique, is nondestructive, and has the ability to differentiate between similar phosphorus compounds makes it invaluable for the identification and characterization of DOP. Such information regarding DOP is required in order to understand aquatic nutrient cycling. The difficulty with using 31P FT-NMR spectroscopy for such studies is the extremely low DOP concentration; usually ranging from < 1 μg P/L in oligotrophic lakes to approximately 100 μg P/L for eutrophic systems. Nanny and Minear raised the DOP concentration into the NMR detection range, which is on the order of milligrams of phosphorus/liter, by concentrating large volumes of lake water with ultrafiltration (UF) and reverse osmosis (RO) membranes. Volume concentration factors of several ten thousand fold provided DOP concentrations of up to 60 mg P/L. Other DOP concentration methods such as anion exchange, lanthanum hydroxide precipitation, and lyophilization require severe chemical and/or physical transformations of the sample and/or they need long processing times, all of which increase the risk of DOP hydrolysis. Sample concentration with UF and RO membranes does not require the sample to undergo these major changes and is also a relatively rapid concentration method. In addition to these concentration capabilities, the use of ultrafiltration and reverse osmosis membranes permitted fractionation of the DOP samples according to molecular size. Nanny and Minear used three membranes in series with decreasing pore size: 30kDa (kilodaltons), 1 kDa, and RO (95% NaCl rejection) to separate the high-molecular-weight, intermediate-molecular-weight, and low-molecular-weight DOP species. In the intermediate-molecular-weight fraction, Nanny and Minear observed the presence of monoester and diester phosphates. Spectra from ten samples collected over a year typically consisted of a large broad signal in the monoester phosphate region spanning from a chemical shift of 2.00 ppm to −0.50 ppm. The maximum of this signal was usually in the range of 1.00 to 1.50 ppm. This broad signal had a shoulder in the diester phosphate region which sometimes was intense enough to appear as an individual signal.