A microwave oven is used to pretreat carbonate samples prior to graphitisation and radiocarbon dating by accelerator mass spectrometry. The method reduces the risk of contamination of small carbonate samples and provides a fast and convenient method for the acid evolution of CO2.

Calcareous tufa from five sites in southern Poland, representing several most typical conditions of tufa sedimentation, were chosen for 14C, 13C, and 18O measurements. These tufas were deposited in a high-energy turbulent stream (the Raerawka site), in streams with moderate but variable flow (Rzerzusnia and Trzebienice), and in semilimnic conditions (Sieradowice site). Sediments of the Gliczarow site represent spring travertines. In all but the latter site, direct comparison of 14C dates of carbonate and organic fractions was possible, leading to an estimate of initial apparent age of carbonate sediments. Clear correlation was found between the value of initial apparent age of tufas and the hydrodynamic conditions of sedimentation. Corresponding values range from ca 3900 yr (Raetawka) to 910 yr for semilimnic sediments (Sieradowice). Intermediate, almost identical values, equal to 2460 +/- 200 yr and 2100 +/- 160 yr, were obtained for tufas from Rzerzusnia and Trzebienice, respectively. Detailed sedimentologic classification of tufaceous deposits is presented and some primary and secondary factors affecting the accuracy of radiocarbon dates of various types of tufas are also discussed.

Traces of uranium and radium within the 14C sample generate radon (Rn) which gets occluded during the benzene synthesis, thus generating false (extra) counts within the 14C counting window. This, if undetected, gives rise to erroneous 14C age determinations. The application of simultaneous alpha and Beta liquid scintillation spectrometry will enable a mathematical evaluation of the 14C signal unaffected by alpha and Beta particle emissions from radon decay daughters.