Absolute dating of plasters and mortars clearly represents a key information to study important structures and buildings that may have undergone a difficult story starting from their construction. This is for instance the case of the architectures in the archaeological site of Hierapolis (Denizli, Turkey). However, when discussing about the possibility to apply radiocarbon (14C) dating, in this site the presence of different sources of contaminants, due to the geological and geochemical conditions and to the used raw materials, prevents the binder dating. As an alternative, we thus decided to focus on the small fragments of straw that had been used as additives in the mortar/plaster matrices. The fragments were identified, selected and dated using a 14C experimental set-up specifically optimized for microgram-sized samples. The obtained results were satisfying, even though the measured 14C ages also pointed out some possible criticalities in dating such small samples collected from a carbonaceous matrix.

When dating documents by radiocarbon (14C), what we typically measure is the concentration of the support (e.g. paper, parchment, or papyrus). This can however lead to a possible misinterpretation of results because the support may be older than the writing itself. To minimize such a possible ambiguity, the ideal approach would be direct dating of the ink or color (if organic). Here we propose a feasibility study to date carbon-based black inks when deposited on papyrus, one of the most widespread writing supports used in the past. We prepared test samples, using a commercial papyrus and a homemade black ink, obtained combining modern charcoal fragments and Arabic gum. Even though the ink binder might have represented the best candidate to be dated, we verified by FTIR that the molecular composition of its soluble fraction is very similar to papyrus extractives, thus identifying the residual charcoals recovered after extraction as the most suitable material for the measurement. Enough charcoal material was extracted from the test samples and processed using our new setup optimized for microgram-size samples. The overall experimental procedure was found to be reproducible, and measured 14C concentrations were coherent with the data obtained from larger samples and raw materials.

Graphitization of 0.5–1.5 mg C, and of smaller samples to a lesser extent, is routinely done at our Facility by reduction over zinc. The method yields low background, good accuracy but offers a limited throughput, requires dedicated equipment and considerable operator time. Sealed-tube graphitization is faster, easier and cost-efficient producing as many graphites as CO2 can be purified in one day with low background, good accuracy and precision, provided precise measurements of δ13C values can be attained by accelerator mass spectrometry (AMS) to correct for isotope fractionation (Xu et al. 2007). We tested sealed-tube graphitization on 0.1 to 1.0 mg C samples and found that while we were able to obtain low backgrounds of >57,000 ±1000 yr BP for ∼1.7 mg C and 41,230 ± 430 yr BP for ∼0.09 mg C (0.0008 ± 0.0001 and 0.0059 ± 0.0003 Fraction Modern, respectively), results were variable for sample sizes <0.5 mg C. Measurements of FIRI Belfast Cellulose and TIRI Barleymash showed 0.3–0.6% precision and 1% accuracy for most sample sizes. We found better results in our laboratory by introducing the following modifications: (1) shorter inner tube (2 cm long), (2) short flame-seal length (∼7–8 cm) and (3) keeping the inner tube with iron separate from the outer tube containing zinc and titanium hydride during cleaning.