Fractionation during the geochemical transfer of carbon in nature produces variation in the equilibrium distribution of the isotopes of carbon (12C, 13C and 14C). Craig (1953) first identified that certain biochemical processes alter the equilibrium between the carbon isotopes. Some processes, such as photosynthesis for instance, favour one isotope over another, so after photosynthesis, the isotope C13 is depleted by 1.8% in comparison to its natural ratios in the atmosphere (Harkness, 1979). Conversly the inorganic carbon dissolved in the oceans is generally 0.7% enriched in 13C relative to atmospheric carbon dioxide. The extent of isotopic fractionation on the 14C/12C ratio which radiocarbon daters are seeking to measure accurately, is approximately double that for the measured 13C/12C ratio. If isotopic fractionation occurs in natural processes, a correction can be made by measuring the ratio of the isotope 13C to the isotope 12C in the sample being dated. The ratio is measured using an ordinary mass spectrometer. The isotopic composition of the sample being measured is expressed as delta13C which represents the parts per thousand difference (per mille) between the sample carbon 13 content and the content of the international PDB standard carbonate (Keith et al., 1964; Aitken, 1990). A d13C value, then, represents the per mille (part per thousand) deviation from the PDB standard. PDB refers to the Cretaceous belemnite formation at Peedee in South Carolina, USA. This nomenclature has recently been changed to VPDB (Coplen, 1994).
In summary, then, isotopic fractionation refers to the fluctuation in the carbon isotope ratios as a result of natural biochemical processes as a function of their atomic mass (Taylor, 1987). Variations as such are unrelated to time and natural radioactive decay. It is common practice in radiocarbon laboratories to correct radiocarbon activities for sample fractionation. The resultant ages are termed "normalised", meaning the measued activity is modified with respect to -25 per mille wrt VPBD. Not all laboratories correct for sample fractionation, in these cases, the correction factor must be added or subtracted from the conventional radiocarbon age. Correction factors are available in most C14 texts.
A table showing isotopic fractionation of different substances in nature is shown here.
The deltaC13 value for a sample can yield important information regarding the environment from which the sample comes, because the isotope value of the sample reflects the isotopic composition of the immediate environment. In the case of shellfish for example, marine shells possess a dC13 value of between -1 and +4 per mille, whereas river shells possess a value of between -8 and -12 per mille, therefore, in a case where the precise environment of the shell is not known, it is possible to determine the most likely by analysis of the dC13 result.
Fractionation also describes variations in the isotopic ratios of carbon brought about by non-natural causes. For example, samples may be fractionated in the laboratory through a variety of means. Usually, this is due to lack of attention to detail and incomplete conversion of the sample from one stage to another or from one part of the laboratory to another. In Liquid Scintillation Counting, for example, incomplete synthesis of acetylene during lithium carbide preparation may result in a low yield and concomitant fractionation. Similarly, the transfer of gases in a vacuum system may involve fractionation error if the sample gas is not allowed to equilibrate it a volume. Atoms of larger or smaller mass may be favoured in such a situation. If, however, all of the sample is converted completely from one form to another (eg. solid to gas, acetylene to benzene) then no fractionation will occur.