No abstract

The flame-sealed tube zinc reduction graphitization method has been successfully adapted and optimized for radiocarbon measurements on EnvironMICADAS in the Institute for Nuclear Research of the Hungarian Academy of Sciences, Hertelendi Laboratory of Environmental Studies (ATOMKI HLES). To reduce the cost and treatment time of producing graphite targets from samples of about 1 mg carbon content, we have omitted the titanium hydride (TiH2) reagent and used a decreased amount of zinc as the sole reductant in our new method. These changes have also helped to eliminate methane formation during the graphitization processes as well as to recover higher ion current at the same background level. These conditions have led to improved efficiency in the 14C measurements; furthermore, the instrument background level remained sufficiently low (<49,000 yr BP). After determining the optimum parameters of the new Zn graphitization method (2.5 mg Fe powder, 15.0 mg Zn powder, 10 hr graphitization at 550°C in heating block, reaction cells with reagent pretreated at 300°C for 1 hr), verification of the accuracy was carried out by the preparation and measurement of IAEA standard samples (C2, C6, C7, C8) with known 14C activity. The sensitivity of the method for gas contamination was tested and determined by comparing the results to measurements of reserved portions of previously processed real samples.

No abstract

A coupled accelerator mass spectrometer–gas interface system has been successfully operating at the Hertelendi Laboratory of Environmental Studies, Debrecen, Hungary, since 2013. Over the last 6 years more than 500 gas targets were measured below 100 µg carbon content for carbon isotopic composition. The system was tested with blanks, OxII, IAEA-C1, IAEA-C2, and IAEA-C7 standards. The performance of our instrumentation shows good agreement with other published gas-interface system data and also shows a quite good agreement with the nominal value of international standard samples. There is a measurable but quite small memory effect after modern samples, but this does not significantly affect the final results. Typical ion currents at the low energy side were between 10–15 µA with a 5% CO2 in He mixing ratio. The relative errors average ±6% for samples greater than or equal to 10 µgC sample with mean count rates of 300 counts per microgram C for OxII. The blank is comparable with other systems, which is 0.0050 ± 0.0018 F14C or 34,000–47,000 yr BP, which allows for the routine measurement of both of small environmental and archeological samples.