Decadal 10Be, 26Al and 36Cl QA measurements on the SUERC 5Â MV accelerator mass spectrometer

We quantify the routine performance and uncertainties of cosmogenic 10Be, 26Al and 36Cl QA measurements made on the SUERC 5 MV accelerator mass spectrometer since 2004. Our analysis compiles data from primary (NIST SRM4325 for 10Be, Purdue Z92-0222 for 26Al and Purdue Z93-0005 for 36Cl) and secondary (Nishiizumi's series for 10Be, 26Al and 36Cl) reference samples with 10Be/9Be, 26Al/27Al and 36Cl/Cl ratios ranging between 10−11 and 10−13. Our decadal datasets indicate that the 10Be, 26Al and 36Cl secondary standard samples have average standard deviations 1.1%-2.4%, 1.1%-2.8% and 3.0%-3.1%, respectively. The average statistical uncertainties based on counting statistics are 1.0%-1.8%, 0.9%-2.8% and 2.5%-2.7% for 10Be, 26Al and 36Cl, respectively. These indicate additional uncertainties (0.6%-1.6% for 10Be, 0.5%-2.4% for 26Al and 1.4%-1.7% for 36Cl) above those calculated from counting statistics alone. The average differences between the measured and the nominal values are within ±1% in 13 of 14 secondary samples.