M. Járó - L. Költő szerk.: Archaeometrical research in Hungary (Budapest, 1988)
Dating - CSAPÓ János, KÖLTŐ László , PAP Ildikó: Archaeological age determination based on the racemization and epimerization of amino acids
4.5. Calibration curve for the determination of age After the selection of the analytical methods and the correction of the errors we could start to determine the ages of the archaeological findings. As already mentioned in the introduction, temperature needs special attention. This means that those température conditions need careful consideration which the sample has been subjected to, following the death of the organism. As we have only a limited knowledge of the temperature changes and fluctuations which have occurred, the temperature of the reactions during the racemization or epimerization process can only be estimated. It N is not possible to determine it exactly; the only exceptions are some extreme cases, e. g. the deep waters oi the oceans. With this assumption is mind we have looked for a solution as to how we could compare the composition of the sample with an unknown age with another sample whose age has already been determined. Of course special attention was paid to the fad that the history of the two samples to be compared should preferably be as similar as possible. The most important factors were the depth from which the sample was excavated (temperature) and the type of soil (pH), as the process of racemization is mainly determined by the temperature and the pH value. For this purpose we collected 100 bone samples of known age from various Hungarian museums. These ages ranged from two thousand to five hundred thousand years. We determined the raw protein content oi these samples, as well as the amino acid composition (D-alloisoleucine) and the ratio oi D- and L-amino acids. Based on the degree of racemization we have categorized amino acids into three groups. The first was made up of amino acids with a quick racemization time, which could be used for the age determination of young bone samples (2000 to 20 000 years). The second group contained amino acids where the half-time of racemization lay between 20000 and 100 000 years, and which could be used for bone samples aging between 100000 and 200000. The last group consisted of those amino acids which could be used for determining the ages of samples over 200000 years. For all three groups for all amino acids the D/L ratio was illustrated as a function of time, thereby enabling us to draw 4—5 calibration curves for each group. The first group does not contain amino acids with a long racemization period, while the third one does not contain those with a short period. The second group contains some overlap, with the other two groups. After estabhshing the time function of the D/L amino acid ratio of the groups with different ages, the methods became applicable for the determination of the ages of samples with conditions similar to those of the calibration curve. By determining the D/L amino acid ratio of the unknown bone sample, it was immediately clear to which group the sample belongs. By comparing the D/L amino acid ratio of the calibration curve with the D/L ratio of the unknown sample, the age of this latter can immediately by read. For one unknown sample we usually used 3-5 amino acids to determine the age, and by averaging these ages we were able to determine the real age of the sample. By applying the caUbration curve, the errors introduced by the temperature and pH could be avoided. However some errors due to the other method of age determination were introduced. The method of age determination based on the racemization of amino acids becomes absolute if we assess the temperature conditions, as in this case the D/L amino acid ratio of the sample is the unknown value, while all the other data — such as the age of the sample - can be calculated from the equation on the speed of the reaction. 5. Re sul ts an d c onclusi ons By following sample preparation procedure described in the "materials and methods" section we get two fractions from the sample to be analysed. The "free amino acid fraction" obtained by 0.1 mole decomposition is not very applicable for determining age due
