A Preliminary Study Of Radiocarbon Dating (¹⁴C) On Phytoliths

Radiocarbon dating is a challenging work in northwestern China, because thereliable samples of materials such as charcoal, wood or other terrestrial macrofossilsare difficult to collect. On the other hand, phytoliths are especially abundant and canbe preserved in great amount in most loess and loess-like sediments which are widelydistributed in arid-semiarid areas in China.The AMS¹⁴C dating of phytoliths couldprovide a promising approach for dating Quaternary sediments. This work focus ontwo key aspects of AMS¹⁴C dating, i.e. high-quality graphite target preparationtechniques and the reliability of¹⁴C dating which have great impact on¹⁴C dating ofphytoliths.The AMS¹⁴C dating is a complicated process which involves in samplepretreatment, graphite target sample preparation and AMS measurement. The AMS¹⁴C dating can reach maximum100ka BPon the performance of the accelerator massspectrometer, but the current AMS¹⁴C dating is still unable to break through the limitof50ka years. The main reason is that the graphite target sample line can introducesmall amount of comtamination into the graphite which affects the AMS¹⁴C results.Therefore, the performance of the graphite target line largely determines the quality ofthe age of phtoliths.A new AMS graphite target preparation line based on the Fe/H₂reduction methodhas been developed in the¹⁴C laboratory at Institute of Geology, CEA, China. Thisthesisanalyzes its performance based on comparative experiments on standardsamples, background smples and nearly100graphite samples with known ages. Someconclusionsare summarized below1) The best experimental conditions for good graphites are that the ration of H₂toCO₂is2; the temperature of the synthetic graphite is550℃and the Fe powder isfisrly activated with oxygen and subsequently is reduced with hydrogen gas whichmakes the experimental process stable.2) The blank value of this line is between0.0045and0.0049pMC with mean0.0047±0.0001（1σ） PMC, corresponding to the limited age of45,700yr, if the machine background is subtracted. The recent blank of AD2013could reach0.0031pMC which reaches the limited age of50,000yr. This work has also tested thereproducibility and memory effects of this line. Furthermore, it has carried out testson dendro-dated tree ring samples and the made comparison of different labs withsame samples to test the reliability. These good results allow us to use the line inroutine¹⁴C dating.The other key aim of this work was to examine the realiablity of AMSradiocarbon dating with the contemporary phytolith extracted from2011rice strawstem and the sample preparation protocols by a natural blank phytoliths from thebottom of L1-3loess in Weilan. Some conclusionsare as follows:3)The initial determination of phytolith from contemporary straw indicates the¹⁴C content occluded into phytoliths is in equilibrium with the¹⁴C activity of theatmosphere. The fraction of phytoliths combusted at900°C could yield more reliable¹⁴C ages than at above900°C temperature although the AMS¹⁴C date of phytolithswould be200yrs oldetr than its true age.4) The fraction modern of background phytolith from L1-3loess with OSL date ofover71ka BP is close to the background of graphite system and suggests that thecarbon from plant cells trapped inside of phytoliths is free from the contamination forrelatively long periods.The result also indicates that the laboratory protocols in the labare designed to minimize the possibility of introducing contaminant carbon,specifically modern carbon, to the phytolith sample during extraction, graphitization,and measurement.Hence, phytolith provides a new alternative source for¹⁴C dating toestablish the reliable age control for late Pleistocene when the common samples suchas charcoal or wood areabsent in arid areas of northern China.