The GSC Radiocarbon Dating Laboratory participated in a program that was designed to exhaustively investigate all aspects of C-14 dating (Scott et al, 1990). A three year study, with Dr. E.M. Scott as principal investigator, was undertaken by the University of Glasgow in 1986. The study had four main aims, namely:

• to provide important information for C-14 users by quantifying the uncertainties on the results routinely obtained from C-14 laboratories;
• to provide an objective cross-check and verification of different laboratory techniques;
• to provide quantative insight into the validity of routinely quoted errors; and
• to assign and numerically assess the contribution of the various experimental processes to the overall error.

The study was carried out in three stages to systematically assess various aspects of the radiocarbon dating technique.

The figure below (figure 8, Scott et al., opt. cit), which includes data from all 3 stages, compares the GSC laboratory with other laboratories worldwide.

larger image [GIF, 19.2 kb, 344 X 368, notice]

The offset for the GSC laboratory, to the left of the Y-axis, indicates a negative bias of less than 20 years, thus GSC age estimates may be slightly younger than the true age of the sample. The very small confidence interval attests to the high precision of the GSC laboratory.

Stage 1 was designed to investigate variations in the different counting techniques (gas proportional, liquid scintillation, and accelerator mass spectrometry). The samples underwent no laboratory treatment, therefore only samples of calcium carbonate for gas proportional and AMS laboratories, or benzene for the liquid scintillation laboratories were dated.

Stage 1

Stage 2 was designed to investigate the effects of different laboratory processing techniques on various sample materials (pre-treated algal carbonate, wood cellulose, and humic acid). The GSC laboratory selected algal carbonate and wood cellulose to test our acid digestion and burning techniques, respectively. The samples were pre-treated in Glasgow before aliquots were sent to the participating laboratories. The pre-treatment for carbonate and wood cellulose are noted below:

• carbonate: a bulk sample of a marine alga (Lithothamnion) was collected and hand sorted to remove shell fragments; leached with dilute (0.5M) HCl to remove approximately the outer 10% of the material; washed with distilled water and dried. The sample was thoroughly mixed during washing and after drying; and
• wood cellulose: prepared from 20 rings (dendrochronologically dated) by successive boiling in 2M potassium hydroxide, followed by bleaching in sodium chlorite, and then washed with distilled water and dried. The sample was thoroughly mixed by tumbling.

Stage 2

Stage 3 was designed to investigate the effects of complete laboratory treatment on the dating of natural samples of wood, shells, and peat. During Stage 3 of the study GSC dated all the types of material available, i.e. wood, peat, and shells, as well as a marine sediment. Each gas preparation was counted in both the 2 L and 5 L counters except for the marine sediment, and one of the shell samples (GSC-4666), because the gas leaked from the cylinder before it was re-counted.

All the study data has been summarized in a table.

The figures for each stage indicate that the GSC age estimates compare very favourably with the expected ages (cf. table listing). In stage 1 and 2 the GSC laboratory tended to provide ages slightly younger than the expected ages, but in stage 3 of the study, in which each laboratory carried out the complete treatment on natural samples, the GSC age estimates were excellent.

Stage 3