Blank corrections for ramped pyrolysis radiocarbon dating of sedimentary and soil organic carbon.

Ramped pyrolysis (RP) targets distinct components of soil and sedimentary organic carbon based on their thermochemical stabilities and allows the determination of the full spectrum of radiocarbon ((14)C) ages present in a soil or sediment sample. Extending the method into realms where more precise ages are needed or where smaller samples need to be measured involves better understanding of the blank contamination associated with the method. Here, we use a compiled data set of RP measurements of samples of known age to evaluate the mass of the carbon blank and its associated (14)C signature, and to assess the performance of the RP system. We estimate blank contamination during RP using two methods, the modern-dead and the isotope dilution method. Our results indicate that during one complete RP run samples are contaminated by 8.8 ± 4.4 μg (time-dependent) of modern carbon (MC, fM ∼ 1) and 4.1 ± 5.5 μg (time-independent) of dead carbon (DC, fM ∼ 0). We find that the modern-dead method provides more accurate estimates of uncertainties in blank contamination; therefore, the isotope dilution method should be used with caution when the variability of the blank is high. Additionally, we show that RP can routinely produce accurate (14)C dates with precisions ∼100 (14)C years for materials deposited in the last 10,000 years and ∼300 (14)C years for carbon with (14)C ages of up to 20,000 years.