Step I: sampling and the issue of chronology questionsAccurate dating of mortar in masonry requires a comprehensive approach involving collaboration between experts in mortar analysis, archaeologists and architects who understand wall stratigraphy20,21. Precise sampling is crucial and begins with well-defined research questions related to chronology, along with documentation and historical research and analysis of the masonry22 (Fig. 1, step I). For example, if the aim is to determine the age of construction, it is important to avoid areas of repair or renovation. However, if the focus is on determining the period in which the building was in use, these repaired or renovated areas may be of greater importance. Mortar between stone blocks is likely to be original if the stone block overlies the mortar; on the other hand, if the mortar protrudes beyond the stone block, this indicates a later intervention. Care should be taken when selecting samples from the ground, i.e. from collapsed ruins, as these may have been transported or weathered, thus being their association with the event to be dated not accurate.Fig. 1Graphical representation of the new multi-analytical procedure for radiocarbon dating of historical mortars.Mismatches in the dating results can arise due to different factors, such as mortar constituents (the type of binder and aggregate) or environmental factors (the state of preservation, which may be due to e.g. recrystallization and delayed hardening).For instance, bedding mortars or core mortars are generally less altered over time than plaster and are less exposed to the external environmental parameters23.To minimize secondary carbon sources, sampling sites should be carefully selected, favoring areas that are less exposed to weathering over exterior surfaces20,21. Analyzing samples from greater depths and intermediate heights helps to mitigate the influence of ambient water, which can introduce younger samples through rainfall or surface water or obvious aging effects from dissolved geological carbonates in groundwater and soil moisture.As the slaked lime (Ca(OH)2) absorbs CO2 from the atmosphere, the setting and hardening process starts from the surface and progresses inwards. Delayed hardening in the inner parts of thick walls can lead to inaccuracies in dating results24. Optimal samples should be taken at a depth close to the wall surface, deep enough to avoid problems with the surface, but not too deep to have problems with delayed hardening. If carbonate aggregates are present, careful sampling is essential to limit dispersion.In summary, careful sampling and consideration of various factors are crucial for successful 14C dating results in mortar. These considerations and methods contribute to the robustness and reliability of mortar dating in archaeological investigations.The evaluation of the degree of carbonation of the mortar with the phenolphthalein test is the first mandatory characterization step. Phenolphthalein indicates the presence of calcium hydroxide in the mortar. A sample that is not fully carbonated must be excluded for 14C dating. The test can be carried out in situ on the masonry or in the laboratory on a sample.Step II: analytical procedure to characterize mortars for datingTo characterize mortars and select those materials that can be suitable for dating purposes, it is essential to determine the composition of all the constituents of the mixture, their relative amounts (binder/aggregate ratio—B/A), the nature of binder and aggregates, the constituents within the binder, as well as the degree of carbonation. In fact, the information we can get from different analytical techniques can give us hints about the manufacturing process of the materials. In particular, it can suggest us whether the basic conditions for applying radiocarbon dating are respected and can support us to choose the best approach to select the most appropriate fraction to be dated. For example, the aforementioned B/A ratio allows us to understand how much material we have to sample to get enough mass at the end of the selection procedure.For a comprehensive characterization, several investigations must be performed, each useful in reconstructing the overall picture and providing key information to select or exclude material for dating (Fig. 1, step II). The complementarity of multiple investigations is crucial for an accurate and full understanding of the material. Indeed, the investigations make it possible to determine the relative chronology of different construction phases within a building or site25,26. Here following the summarized description of the analytical techniques proposed for characterization.OMThin-section observation of mortar under an OM in transmitted light provides essential insights into the nature of binder, aggregate and lumps19,27.For the binder, OM provides information on the texture (micritic, microsparitic, sparitic), the mineralogical composition, the birefringence colors, the structure and the interactions with the aggregate. Moreover, OM allows us to classify the binder as: air lime, natural hydraulic lime, air lime with addition of pozzolanic materials (i.e. cocciopesto, volcanic ash and clay minerals) and modern hydraulic binder28,29.The description of the aggregate is crucial for the evaluation of contamination sources, taking into account mineralogical composition, particle size distribution, amount of binder with respect to aggregate (B/A) ratio, macroporosity and alteration products.Petrography is also beneficial for the identification of lumps and organic fragments in the mortar. OM observation make it possible to recognize the type of lumps and distinguish between residues of stones used to make binders and binder residues.The observation of lumps with OM allows us to recognize their types and origins, achieving information on the rock used to produce lime as well as suggestions on production technologies27.Petrographic observation contributes to assess the uniformity of the binder and to identify zones of different crystallinity due to partial recrystallization by circulating water. Sources of contamination, such as recrystallization of calcite and carbonate aggregates, can lead to exclude samples from 14C dating21.Modern binders should be eliminated from dating, since the dating principle is not applicable to these types of binders. Particular attention should be paid to magnesium binders30. The 14C dating outcomes may be affected by the presence of much younger 14C, due to the properties of minerals produced upon carbonation (such as magnesite and nesquehonite).XRPDXRPD analysis of bulk samples includes the mineralogical composition of both the binder and the aggregate, which can be integrated with the identified phases in thin sections. Single lumps and binder-rich portions can be also analyzed. All these data yield crucial information, revealing whether the mortar is non-carbonated (portlandite), if the sample contains magnesium lime (brucite, hydromagnesite, magnesite), or if the binder exhibits hydraulic properties31 (tobermorite, hydrogarnet), or if secondary reactions occurred which lead to the formation of new phases (gypsum, hydrotalcite, hydrocalumite). The presence of these latter two phases in mortar binders strongly influences the radiocarbon dating of lime mortars, because of their high (CO3)2– anion capture capability32,33. The presence of gypsum indicates that the binder has altered, suggesting an open system and therefore a context subject to contamination from the external environment34.SEM-EDSObservations under the optical microscope can be further enhanced and supplemented by SEM–EDS which combines microscopy and X-ray spectroscopy to obtain detailed information on the morphology and elemental composition of mortar constituents. Semi-quantitative elemental analysis is useful for: (1) estimating the provenance of raw material through the analysis of residues of stones used for lime production; (2) obtaining information on the hydraulic index (HI)35 and the overall composition of the binder, including the possible presence of Ca and Mg based binder, and of silico-aluminates ferriferous phases; (3) evaluating changes in elemental composition within reaction rims areas; (4) characterizing lumps, especially if they have a heterogeneous texture; (5) achieving micro-chemical information about the aggregate and providing hypothesis on its provenance.TGATGA is used in the analysis of historical mortars for evaluating hydraulic behavior; it involves subjecting a sample to controlled temperature changes while measuring its mass as a function of temperature. TGA serves for characterization of binder materials (air binder, hydraulic binder, gypsum, etc.)36,37. Moreover, the TGA results can be integrated with the HI value calculated from punctual micro-chemical analyses carried out with SEM-EDS38.Step III: Selection and characterization of the powder for the screening of CaCO3 originUpon assessing that the sample exhibits datable characteristics, as a consequence of all the analyses performed in Step II, the following process involves the selection and further characterization of the carbonate fraction. The binder calcite has the same chemical composition as burned carbonate rocks or carbonate aggregates, but different textural, isotopic signatures and mechanical properties.A mechanical separation of binder-rich bulk and lump was performed, starting from a selection under stereomicroscope. For bulk samples, a portion enriched with binder and lumps is separated, then sieved to 63 µm and lightly crushed.Our approach aims at finding non-destructive techniques able to determine the origin of the calcite in the powder samples selected for dating (Fig. 1, step III). Non-destructive techniques allow the preservation of the sample mass so that the same sample can be subjected to several analytical procedures and treated for 14C analysis (Fig. 1, step IV).The different origin of carbonates (geogenic and anthropogenic) can be detected by the different distortions in the lattice structure within small crystallites. In principle, different types of calcite interact with electromagnetic radiation in a way that depends on the atomic arrangement. FTIR and Raman spectroscopies can be used to identify short-range order at the molecular level. In addition, CL analysis, which is conventionally used to assess the origin of calcite, in our approach is combined with ATR-FTIR and micro-Raman.The most important advantage in this non-destructive approach is that the exactly same powder is analyzed in OM-CL, ATR-FTIR and micro-Raman; and if the sample is mainly constituted by anthropogenic calcite, it is used for step IV.OM-CLCL is a petrographic technique which represents an additional way of examining thin sections or powder samples of carbonate specimens39. The phenomenon of CL of mortars has been discussed since 1997 and has been used in numerous studies to evaluate the origin of carbonates13,40,41. Different densities and distribution of atomic defects in the calcite crystal structure serve as markers to identify the origin of calcite. Considering this principle, geogenic calcite and anthropogenic calcite may have different luminescence intensities due to the different formation process.The phenomenon can be easily observed with petrographic microscopes equipped for CL analysis (OM-CL), this instrumentation is relatively inexpensive and easy to use. For the non-destructive analysis of powders, we used OM-CL. The disadvantage of this technique lies in the resulting color hues, especially when multiple emissions of the same powder result in a composite hue. Typically, a qualitative analysis was performed, just attributing “hues” to the different observed colors (see for example tile red, dull purple, brown, dark brown, grey, dull grey and black). In such a framework, interpretation of data could be influenced by the operator him/herself20,42,43. This problem can be solved by combining several analytical techniques to obtain a validated and unambiguous result.FTIRIn the context of mortar dating, spectroscopy has already been used to distinguish the origins of calcite. As demonstrated in previous studies44,45,46 conventional Fourier transform infrared spectroscopy in transmission mode with KBr pellets can be employed for rapid sample analysis, using the heights of v2 and v4 bands.In order to use non-destructive analysis and preserve the sample for further analysis and dating, ATR-FTIR was tested on samples with known composition and origin to establish whether this mode could lead to the same results as the FTIR technique with KBr pellet47.Since it has been shown that differences in grinding degree affect peak widths and relative heights of carbonate archaeological materials34,48, samples with same preparation procedures were analyzed to replicate the typical pre-treatment that might be carried out on unknown samples for dating purposes.The distinct trend lines highlight the systematic differences in v2 versus v4 peak heights in ATR-FTIR mode for calcites formed through various processes. Two trend lines were created (geogenic and anthropogenic calcites), which can help to determine the origins of unknown samples, offering preliminary insights into their formation. The ability to discern calcite origins through the ATR technique is particularly advantageous in the field of mortar dating, as powdered samples can be collected and reused for dating if they contain anthropogenic calcite.Micro-RamanMicro-Raman spectroscopy is a valuable tool for the characterization of mortars, enabling high lateral resolution analysis of the mineral phases of aggregates and binder components49,50. So far, some studies have demonstrated that micro-Raman spectroscopy can be successfully used to estimate the content of cations (Mg2+, Fe2+ and Mn2+) in carbonates, as the vibrational frequencies of the translational (T) and librational (L) modes of carbonates are significantly related to their cation composition51,52. Raman spectroscopy has been used to investigate variations in atomic bonding in biogenic calcite crystals and to distinguish the degree of crystallinity of calcium carbonate in biological materials by assessing the frequencies and width of the v1 and v4 bands53. Raman analysis of CaCO3 polymorphs in54 found that the amorphous calcium carbonate exhibits a broad peak in the lattice mode region (below 400 cm−1) and that the most prominent band associated with the carbonate ion at around 1085 cm−1, which appears as broader and significantly less intense than usual, slightly shifts towards lower wavenumbers.We carried out a study to determine the origin of calcite using micro-Raman spectroscopy. The potential to distinguish between geogenic and anthropogenic calcite using micro-Raman spectroscopy was established for the first time by the authors55.Raman spectroscopy and statistical methods have shown that the anthropogenic calcite samples exhibit a broadening of the L, v1 and v4 bands (calculated from FWHMs) compared to geogenic calcite samples.Structural disorder within the calcite crystals or the presence of low crystalline order is reflected in relatively broad FWHM values and wavenumber shifts. The wider and shifted toward lower wavenumber is the spectral band, the lower the crystallinity within the mineral.The influencing parameters (including band position, band intensity, the area covered by the bands and the FWHM values of L, v4 and v1) for distinguishing the origins of calcite were successfully identified and they can be used to determine the origin of calcite in unknown samples intended for dating.The potential of micro-Raman on distinguishing different calcite domains was also confirmed by Toffolo et al.56. In this paper, the micro-Raman analyses were performed on petrographic thin sections in archaeological lime samples.Step IV: carbonate micro-sample preparation and AMS measurementsThe limited sample material due to the high possible level of heterogeneity of the mortars, the sample loss during the characterization step and the highly selective pre-treatment process, motivates us to use the micro-sample 14C preparation.In this framework, the so-called Lilliput graphitization line at the LABEC laboratory in Florence, one of the laboratories of CHNet, the INFN network for Cultural Heritage, was integrated with a reaction chamber designed for the extraction of CO2 from carbonates (Fig. 1, step IV). The Lilliput line is particularly useful in the case of mortar treatment, because it allows managing samples as small as only 50 µg of carbon, well below the limit of the “traditional” larger samples of about 700 µg18,57. Such small processed masses provided the possibility to investigate the feasibility of dating even individual lumps of binder in mortar samples.Typical processing masses for mortar samples are:—approx. 2.5 mg in the case of lump; —approx. 5 mg in the case of bulk mortar.Acid dissolution and Lilliput graphitization reactorsThe extraction of C from the selected inorganic fraction of the mortar is carried out by acid dissolution. The carbonate sample, mechanically separated and previously characterized with non-destructive techniques, is treated with H3PO4 in the acidification line.For bulk samples, 2 evolving CO2 fractions are usually collected per sample: the first in a few seconds (0–10/30 s) and the second thereafter (10/30–60 s). The selected shortened reaction time is intended to avoid the risk of geological contamination, at least in the first fraction, as contaminants may still be present despite mechanical separation. In the case of lump samples, a fraction from the first few seconds of the reaction is collected without the risk of contaminants reacting with the acid.The CO2 extracted from the acidification line is then cryogenically transferred into the graphitization chamber using liquid nitrogen. The amount of CO2 collected is monitored by pressure measurements. Typically, about 100 mbar of CO2 is collected for each sample; this pressure basically corresponds to about 50 µg of graphite at the end of the reaction given the inner volume of the Lilliput reaction chambers. The graphitization reaction occurs on small copper inserts previously prepared with Fe catalyser pressed on them and is triggered at 600 °C in presence of H2 excess; the reaction produces water, which is trapped within the cold finger. After the graphitization process, the copper inserts with the graphite deposited on them are mounted in specially modified aluminum holders that fit into the ion source of the accelerator to measure the radiocarbon concentration.
