*4.2. Characterization of the Inorganic Substrate and Discussion of the Preparation Method of the Pigment as a Branch in the Operating Chain of the Purple Dyeing Industry*

In the microRaman spectra, Figure 3, the characteristic vibrational band at 1085 cm−<sup>1</sup> evidenced the presence of calcium carbonate in the composition of all the examined purple samples, either those from the pigment lumps (AKR-10891, AKR-10882, and TRI-PR 13) or the sample from the Raos wall paintings (RAOS-KAL1). This band, assigned to the ν<sup>1</sup>

symmetric stretching of CO3 2- in the calcium carbonate molecule, is rather similar for the two calcium carbonate allotropes (~1086.2 cm−<sup>1</sup> for pure calcite and 1085.3 cm−<sup>1</sup> for pure aragonite) [56]. This slight difference is smaller than the instrument's spectral resolution (~3 cm<sup>−</sup>1) and therefore could not be used for the distinction of the two polymorphs in the archaeological samples. Given that the resolved band at 152 cm−<sup>1</sup> (Figure 3) assigned to lattice vibrational mode is also common for calcite and aragonite, it is customary to focus on other additional, though weak, vibrational bands for the identification of the two allotropes. The most characteristic differences in the Raman spectra of calcite and aragonite are (a) the relative integrated intensities (with respect to the ν<sup>1</sup> band) of the bands at ~283.6 cm−<sup>1</sup> for aragonite and ~281.2 cm−<sup>1</sup> for calcite (for aragonite the integrated intensity of this band is at least 20 times weaker than the corresponding one of calcite, for which it is the second strongest peak in the spectrum) and (b) a band at 214.7 cm−<sup>1</sup> in the aragonite Raman spectrum, which is specific and may as well be used for its identification [56]. If the 1085 cm−<sup>1</sup> band in the spectra shown in Figure 3 was assigned to calcite, a corresponding band at ~280 cm−<sup>1</sup> (characteristic also for calcite) should have been resolved in the Raman spectra, which was not the case. The latter indirectly suggests aragonite as the calcium carbonate allotrope identified here. Additionally, a weak band resolved at ~210 cm−1, whose intensity increased proportionately to the respective one of 1085 cm<sup>−</sup>1, agreed with the aforementioned suggestion.

Complementary data were collected with the corresponding FTIR spectra acquired from a few particles of each of the investigated samples. It should be mentioned that the FTIR spectra refer to a higher amount of sample, and thus are more representative of the bulk pigment than the Raman spectra acquired on single particles focused under the 100× objective of the Raman microscope, which allows a spatial resolution close to 1 μm. In the FTIR spectra shown in Figure 5a, it is possible to identify with certainty both aragonite and calcite in the inorganic substrate of the pigment based on the bands that are specific to one of the calcium carbonate polymorphs. In the case of the samples taken from wall-painting details, calcite may have derived also from the lime-based ground or may refer to chalk white as a pigment in the painting. However, the identification of both calcium carbonate allotropes and aragonite in particular in the purple pigment found in lump form puts in evidence the marine origin of the calcium carbonate substrate of the organic pigment. In the FTIR spectra, the well-resolved bands at 1475, 1081, 860, 711, and 700 cm−<sup>1</sup> perfectly reproduce the respective characteristic bands of pure aragonite. Among the IR absorption bands of the carbonate ion, the split peaks at 711 and 700 cm−<sup>1</sup> (ν<sup>4</sup> mode) and the peak at 1081 cm−<sup>1</sup> (ν<sup>1</sup> mode) are specific to the aragonite structure whereas the peak assigned to the ν<sup>2</sup> mode appears with a difference of ~15 cm−<sup>1</sup> for the two polymorphs, namely, at 860 cm−<sup>1</sup> for aragonite versus 875 cm−<sup>1</sup> for calcite [57,58]. Additionally, the stronger and broad absorption band assigned to the ν<sup>3</sup> mode, presenting a significant shift for calcite (at 1421 cm−1) compared to aragonite (at 1475 cm−1), further confirmed the presence of both calcium carbonate polymorphs in the composition of the purple pigments. For the identification of the DBI compound, only the two main bands at 1635 and 1612 cm−<sup>1</sup> were barely resolved on the slope of the main broad carbonate band, Figure 5b, [26,43,45].

Examining a sample of the purple pigment from Trianda (TRI-PR13) under microscope, it was possible to observe the aragonite crystals with the characteristic needle-shaped form and clear appearance, as shown in Figure 6 [58,59]. In agreement with the presented FTIR data, in an older study of the purple lump (AKR-10891) with X-ray diffraction analysis [60], magnesium-rich calcite was also identified, which together with aragonite strengthened the assumption that the inorganic base of the pigment was possibly obtained from crushed and ground shells. The "recycling" of shells, of any species, either food residues or debris of purple production (*Muricidae* family), was a well-known practice not only for obtaining the base of the purple pigment but also for their use as coarse aggregate in the floor mortars or as raw material for the production of lime.

**Figure 5.** (**a**) FTIR spectra acquired in transmission mode on KBr pellets of samples from the purple pigments (AKR-10882, AKR-10891, and TRI-PR13) found free of any support. Both aragonite and calcite are identified in the inorganic substrate of the pigment. (**b**) Zoom in on FTIR Scheme 13. The major characteristic vibrational bands of the 6, 6 -DBI compound, namely, at 1635, 1612, 1313, and 1157 cm<sup>−</sup>1, are barely resolved on the slopes of the main broad carbonate band.

**Figure 6.** Microphotographs in different magnifications (original magnification 200×, 500×) of the purple pigment TRI-PR13 from Trianda in white light. The transparent longitudinal crystals of aragonite are evident in the purple matrix in the higher magnification, in particular.

The composition of the purple pigment consisting of a calcium carbonate substrate to which the purple dye was adsorbed, possibly by immersion in the alkaline juice as described by Pliny the Elder [21] in the procedure for preparing the pigment *purpurissum,* attests to its preparation in a vat, rendering improbable the application of a direct "primitive" technique. These data allow the establishment of vat-dyeing know-how to be traced at least back to the Late Bronze Age in the Aegean [23,61,62].

Reviewing the published results of analyses of shellfish purple pigment found in different archaeological contexts over time, in different applications and substrates, it appears that the inorganic base of the pigment can vary [32]. The choice of the inorganic base may have been related not only to the availability of the raw material but may also have been dictated by the substrate to which it was applied. For the wall paintings, the calcitic basis of the purple presents a perfect affinity with the lime-based ground and the calcium white used in a mixture with the rest of the pigments for rendering the light colors. An interesting alternative instance of shellfish purple was identified in the paint decoration of a funerary *kliné*, in Daskyleion (5th century BCE), with kaolinite being the main constituent of the inorganic base of the organic pigment, applied in that case on a white ground made of gypsum [63]. In the same paper, the HPLC data acquired from samples extracted from the purple paint detail on the *kliné* were discussed in comparison with the data acquired from an extract from a purple dyed-textile find, found within the same tomb. The perfect match of the HPLC profiles of the two extracts, from the purple pigment and the purple-dyed textile in the same context, constitutes a strong analytical demonstration of the argument suggesting the production of the purple pigment was integrant to the purple dyeing industry of textiles, according to Pliny's text.

Moving even further onward in time, and referring to a shellfish purple pigment identified in the paint decoration of Hellenistic terracotta figurines excavated from a rockcut tomb in Chania, Crete (ca. 300 BC) [64], another white earth, huntite was identified in association with the purple pigment. Although in that case huntite was not directly attributed to the inorganic base of the purple pigment, this should be considered as a possibility. Interestingly, in a more recent investigation [65], huntite was also identified in association with the molluscan purple color revealed among bones and fragments of unidentified materials preserved in a gold larnax of the chamber of Tomb II, in the Macedonian royal tomb complex at Aegeae discovered by Prof. M. Andronikos. These data provide evidence for a wider range of white earth that apparently was known to be appropriate for use in the purple dyeing/pigment preparation operating chain.
