A Wet Oxidation Method for AMS Radiocarbon Analysis of Dissolved Organic Carbon in Water

We present a method for the extraction of dissolved organic carbon (DOC) from water. The method is adapted from Burr et al. (2001) using the basic steps: 1) sample filtration; 2) acidification to liberate and remove dissolved inorganic carbon (DIC); 3) evaporation of the sample to isolate salts that include trace quantities of carbon; 4) combustion of the salts; and 5) purification of the CO2. Two significant improvements have been made to the earlier method. The first is to use wet oxidation with potassium permanganate to oxidize organics in place of the combustion step and the second is the development of a reduction/oxidation purification procedure to remove sulfur and nitrogen oxides that may form during the oxidation step. Wet oxidation has a practical advantage over the previous method because it proceeds at low temperature (70 ° C). The original method required quartz vessels to oxidize the salts at 900 ° C. At this temperature, salts in the samples formed gases that interfered with the isolation of CO2 and the quartz vessels degraded with each combustion, affecting their structural integrity. The expensive quartz vessels could only be used for a limited number of samples, whereas Pyrex vessels used for wet oxidation are inexpensive and can be used indefinitely. The blank fraction modern carbon ( f ) and its mass dependence for the refined technique was determined from repeat analyses of salicylic acid produced from petrochemicals. For samples with a mass m above 0.5 mg, F = 0.0083 ± 0.0011. For samples below 0.5 mg, the blank follows a 1/ m dependence as observed for other accelerator mass spectrometry (AMS) 14C measurements (Donahue et al. 1990). The reproducibility of the method is demonstrated using repeat measurements from a variety of samples, including a sample measured with the former high-temperature 900 °C combustion technique. The virtues of the wet oxidation method are that it is economical, produces a low blank, and provides good reproducibility. DOI:  10.2458/azu_js_rc.55.16277