14C - AMS measurements of microgram-sized samples: hardware developments and applications to Cultural Heritage

In the field of Cultural Heritage and archaeology, knowing the period in which an artefact is produced or being able to date biological remains is of fundamental importance. Radiocarbon dating is surely one of the most widespread dating techniques. It allows us to date all those organic and inorganic materials which have been exchanging carbon with a carbon reservoir until a certain moment (e.g. the death of an organism or the definite isolation of the inorganic system from the reservoir itself). One of the basics in radiocarbon dating is the need to collect a sample, whose mass typically varies according to the kind of materials. Considering the possibility to measure the residual 14C abundance by Accelerator Mass Spectrometry (AMS), sample masses are usually relatively small, i.e. of the order of magnitude of tens of milligrams. However, in some particular cases, the mass required for the analysis may become problematic: indeed, there can be some applications in which the amount of mass required would alter the integrity and spoil the legibility of the object to be dated (e. g. when a central portion of the object should be collected), or the amount of mass that we can collect is already very small (e.g. in case of highly degraded materials or samples for which a particular selective pre-treatment is mandatory). For this reason, especially in the most recent times, the interest in reducing the mass required for the 14C measurements has increased. My Ph.D. research has been focused on reducing the amount of mass needed to perform radiocarbon measurements down to about 50 μg of graphite. During these three years, I have worked at INFN-LABEC laboratory (Istituto Nazionale di Fisica Nucleare - Laboratorio di tecniche nucleari per i Beni Culturali), in Florence. At LABEC, radiocarbon measurements are typically performed preparing samples of about 700 μg of C (final carbon mass at the end of the sample preparation process). In order to reduce the mass of the treated samples, the first part of the project has been represented by the upgrade of the pre-existing graphitization system and the following optimization of the new installed experimental set-up for microsamples. We installed new graphitization reactors, reducing their volumes in order to improve the collected pressure to favour the graphitization reaction. New reactors were equipped with a small quartz tube used as the “hot” part and a silver cold finger. We also designed and assembled small ovens and small Peltier-based devices, used to reach the temperature needed to trigger the reaction and to trap the unwanted water produced during the graphitization reaction, respectively. We installed new pressure gauges, sensitive to low pressures, and we assembled a home-made data acquisition system based on an Arduino board. In addition to the optimization of the graphitization set-up, the optimization of the AMS measurements in the Tandem accelerator was necessary as well. Several tests on the new experimental set-up were performed, verifying a good reproducibility, a satisfying background and fine precision and accuracy. Since the new microsamples set-up proved to be reliable, we decided to focus on new possible applications in the field of Cultural Heritage: Radiocarbon dating of mortars: Mortars are very heterogeneous materials and the separation between the carbon fraction of interest (that is the binder) and the possible contaminants (carbonates aggregates) can be very challenging. A crucial point of this application is finding a very selective separation procedure for the removal of the contaminants mentioned above. This procedure should be strongly selective causing a high mass loss, so that using the typical set-up for large samples is not feasible. Our new microsamples set-up allowed us to deal with this problem: in fact, we were able to keep our procedure selective enough for the application dating carbonate samples as low as few milligrams, thus without increasing the initial mass of our collected mortar core (that is usually not possible in the field of Cultural Heritage). Radiocarbon dating of inks: The main issue of this kind of application is the invasiveness of the analysis, especially in case of using the typical set-up for large samples. It is clear that to date the ink, we would collect a portion of the text, thus we would risk to spoil the legibility of the document. For this reason, reducing the mass required for the measurements is mandatory. Our microsample set-up allowed us to perform such task with satisfying precision while reducing by an order of magnitude the mass of the sample to be dated, thus efficiently reducing the invasiveness of the dating procedure. We believe that the optimization of the experimental set-up for the dating of microsamples is necessary and of high importance: the case studies discussed in this thesis prove it, since without the microsamples set-up we could have not performed such analyses.