A Study on Accurate Radiocarbon Dating of Ancient Cultural Properties with Accelerator Mass Spectrometry

本研究利用Ngaoya大学测年与材料研究中心的Tandetron加速器质谱计，对古代文物和文献进行高精度放射性碳测年的一些基本问题进行了研究。首先，研究了以Fe为催化剂，H_2为还原剂，以Co_2为原料制备石墨的方法。当使用石墨化样品时，一次测量所需的碳量减少到1毫克，并且σ误差^<14>C可追溯到<正负>40年。证明了样品石墨化对文物的^<14>C定年是不可缺少的，但样品数量有限。作为第二步，我们建立了一个程序，通过校正碳样品的碳同位素分馏的影响来获得更准确的^<14>C年代。对每个古文化样品制备的几种石墨材料进行了再现性测试。试验表明，^<14>C的测量应重复3次以上才能获得高度可靠的结果。此外，使用主要生长在北美和欧洲的木材样品的树木年代学开发的从^<14>C日期到日历年的校准程序已成功应用于日本样品。对一些日本文物样品进行了放射性碳定年，如旧纸质文件、文件容器盒、木制雕像、古铁剑、铁雕像、丝绸样品、古船碎片、旧骨样品等。这些样品的日历年，通过校准用Tandetron AMS测量的^<14>C日期估计，与历史估计一致。放射性碳定年法也被证明对找出仿制品很有用。

This study focussed on some basic problems to attain a high accuracy radiocarbon dating of ancient cultural properties and documents, by using a Tandetron accelerator mass spectrometer of the Dating and Materials Research Center, Ngaoya University.As a first step, a sample preparation method was developed to produce graphite from Co_2, using Fe as a catalyst and H_2 as a reducer. When a graphitized sample is used, carbon amount necessary for one measurement was reduce to 1 mg, and one sigma errors of ^<14>C dates to <plus-minus>40 years. It proved that the sample graphitization is indispensable for ^<14>C dating of cultural properties with rather limited sample amount available. As a second step, we have established a procedure to obtain a more accurate ^<14>C dates by correcting the effect of carbon isotopic fractionation of carbon samples.Repoducibility tests are conducted for several graphite materials prepared from each ancient cultural sample. The tests indicated that ^<14>C measurements should be repeated more than 3 times to get a highly reliable results. In addition, a program for calibration from ^<14>C date to calendar year, developed using dendrochronology of wood samples grown mainly in North America and Europe, has been applied successfully to the Japanese samples.Radiocarbon dating was conducted for some Japanese cultural-property samples, such as old paper documents, document container boxes, wooden statues, ancient iron swords, iron statues, silk cloth samples, ancient ship fragments, old bone samples, etc. The calendar years of those samples, estimated by calibrating the ^<14>C dates measured with a Tandetron AMS,are consistent with historical estimations. The radiocarbon dating is also proved to be useful to figure out imitations.