The measurement of a radiocarbon date is a complex business and ORAU is involved in all stages of the process from advice given prior to submission of samples right through to help with statistical interpretation and publication.
Radiocarbon measurements are difficult to make with good precision and accuracy. This is because carbon is such an abundant element in the environment and so contamination from material of a different age is always possible.
For these reasons the methods employed at a radiocarbon lab must be rigorous, well tested and reproducible. All of these processes are carefully controlled at ORAU through our quality management system. The unit uses some thirty different pre-treatment methods for samples of different kinds and in different states of preservation. These methods have been developed, over the last thirty years, as part of the laboratory's research programs. They cover most of the materials suitable for radiocarbon measurement.
The methods range from simple acid/base/acid treatments to separation of amino acids from proteins using HPLC.
Once material has been purified for radiocarbon dating, the samples are combusted and their stable isotopes analysed. This is an important part of the quality control. The experience of the Unit in stable isotope work enables us to use both the stable isotope ratios (δ13C and δ15N) and the ratios of carbon to nitrogen to look for signs of impurities.
This data can also provide useful information on food webs. In the case of human bone, this can be important for the dating itself since some aquatic diets can affect radiocarbon dates significantly.
The unit has four stable isotope mass spectrometers. Two Europa Scientific instruments are used in conjunction with automated gas collection systems for radiocarbon samples. Another two (one from Finingan and one from Europa Scientific) that are used for stable isotope research. All are usually operated in continuous flow mode (where the sample is transported in a stream of inert He gas). They are set up to measure the carbon and nitrogen compositions and stable isotope ratios. Precisions are usually better 0.1 to 0.2 per mil for δ13C and 0.3 per mil for δ15N.
At this stage, we reject a significant minority of samples as being unsuitable for dating. Because a wrong radiocarbon date is usually more misleading than no date at all, we never date material if we are unsure of its chemical origins. In borderline cases we will measure the radiocarbon but inform the submitter of the possible uncertainties.
(See also the introduction to AMS)
We have two purpose-built gas-collection systems, designed at ORAU, which semi-automate the collection of samples from continuous-flow combustion, and mix the samples with the correct quantities of hydrogen for conversion to graphite.
Once the sample has been processed to either carbon dioxide or elemental carbon (in the form of graphite). The proportion of radiocarbon atoms in the sample is determined by AMS.
The AMS system at Oxford was built specially for us for radiocarbon dating by High Voltage Engineering Europa BV. The instrument is designed to give very high precision measurements on a routine basis.
At ORAU, measurements are always made in conjunction with six standards of known composition and two samples of known age. The measurements on standards and known age material allow a constant check to be kept on the accuracy of the dates.
The radiocarbon concentration can then be used to calculate the radiocarbon age of the sample.
The precision of radiocarbon dating depends on age. For recent material ORAU aims to get a precision of about 0.3% (typically 25-30 years).
It is important to note that radiocarbon calibration usually reduces the precision of measurements considerably.
Individual measurements are unlikely to be able to give a range at two standard deviations (or 95% confidence) of better than 120 years. In many cases, the overall range is closer to 200 years and in some cases greater still. This is one reason why it is important before dating to ask advice on whether or not radiocarbon has the resolution to answer particular chronological questions.
In some cases, further statistical analysis can improve the precision of the method where there are several dates.
All radiocarbon measurements measured at ORAU are scrutinised by the academic members of the Unit in a weekly meeting to assess the data from the chemical pre-treatment, the stable isotope measurements and AMS measurement itself.
Results are only released once we are satisfied with the details of the process. If any uncertainties remain these are reported to the submitter in the results letter.