**Ioannis Karapanagiotis**

Department of Management and Conservation of Ecclesiastical Cultural Heritage Objects, University Ecclesiastical Academy of Thessaloniki, 54250 Thessaloniki, Greece; y.karapanagiotis@aeath.gr; Tel.: +30-2310-301784

Received: 24 May 2019; Accepted: 17 June 2019; Published: 29 June 2019

**Abstract:** Shellfish purple, also known as Tyrian purple and royal purple, has a long history, which has been revealed and documented in recent years through valid physicochemical studies using sophisticated techniques. The aim of the work was to summarize the conclusions of these studies and to describe the results of two unpublished investigations regarding the (i) identification of shellfish purple in a textile (4th century BCE) from ancient Macedonia and (ii) dramatic effect of the dyeing conditions on the composition of the purple dye. Moreover, a critical discussion is included about the discovery of the shellfish pigment and dye based on the available scientific evidence. Previously published reports describing the identification of the shellfish colorant in objects of the cultural heritage were carefully summarized. Shellfish purple was not used only as colorant, but it served other purposes as emphasized in this review. In particular, examples for the use of shellfish purple in medicine, grave goods and fillers and plasters in walls, were described. Examples of materials and methods that were used in the past to produce "fake" purple, imitating the aesthetic result of the valuable royal marine material were summarized. Finally, the solubility of indigoids was discussed using modern approaches of physical chemistry.

**Keywords:** purple; shellfish; mollusk; Tyrian purple; indigo; pigment; dye; cultural heritage; HPLC

### **1. Introduction**

Identification of the materials of cultural heritage objects is important to develop and apply effective conservation strategies, which ensure the preservation of the valuable historical and archaeological objects. Conservation and preservation are probably the most important conditions towards a sustainable market for cultural heritage, along with appropriate policies aiming at promoting the human civilization through the multidisciplinary area of archaeometry and other disciplines [1]. Moreover, the characterization and identification of cultural heritage materials leads to useful insights regarding the technological skills of civilizations and reveals economic, social and historic aspects [2]. For these reasons the application of physicochemical methods and instruments to archaeological research has increased substantially over the last half-century and today represents a major methodological subfield within archaeological science [3].

Ever since the beginning of human existence, humans made use of materials to color virtually everything they used or had: Bodies, caves, pottery, sculpture, stone structures, parchment, paper, clothing and other textiles [4]. The important role of color in human civilization becomes dominant for painting art and textiles. The former is essential for human nature, as evidenced for example by the Lauscax cave paintings, dated from about 17,000 BCE and the hundreds of other paleolithic cave paintings e.g., in Altamira, Chauvet and Cáceres. The astonishing findings of abalone shells that were used to hold ochre mixtures 100,000 years ago suggest that pigments were utilised from the very early days [5]. The vital, timeless and universal power of painting and, more general, pictorial art is

evidenced in many different ways. For example, icons have been used for centuries by the Orthodox Church as a window to communicate with the unseen divine reality whereas the Pioneer plaques are pictorial messages for potential extraterrestrial life. Organic materials such as textiles and their colorants (dyes) are sensitive to degradation effects developed through ageing. Consequently, early known-identified use of dyes dates back to at around 6000 years ago [6], which falls short compared to the iron and manganese pigments found in the aforementioned prehistoric caves.

Among the various colorants, which have been used since the prehistoric times, shellfish purple originating from marine mollusks holds a prominent position [7–9]. Shellfish purple (in Greek, *porphyra* = π*o*ρϕύρα) is known also as Tyrian purple, because of the significant deposits of shells found in the area of Tyre and the important role of the purple dye in the Phoenician history. The material is also known as royal purple and imperial purple, as the expensive purple-dyed textiles were status symbols. Shellfish purple is connected with the incredible story of tekhelet, the elusive sky-blue color mentioned throughout the Bible and the clothing of emperors and high priests in civilizations, which had scattered all around the Mediterranean basin.

The goal of this review is to summarize the current state of the art provided by the Natural Sciences for the shellfish purple in archaeology. Examples of historical written sources are briefly presented; biological mollusk sources and the chemical composition of shellfish purple are described. Critical discussions are included associated with the discovery of the shellfish pigment and dye based on the available scientific evidence. Previously published reports describing the identification of shellfish purple in objects of cultural heritage are carefully summarized, revealing the pros and cons of the physicochemical instruments employed for analytical purposes. Emphasis is placed on the results of shellfish purple in the prehistoric Aegean. Two unpublished studies are included (i) describing the identification of shellfish purple in a textile from ancient Macedonia and (ii) demonstrating the dramatic effect of the dyeing conditions on the composition of the purple dye. Shellfish purple was not used only as a paint and dye, but it served other purposes as described in the present review. Materials and methods that were used in the past to produce "fake" purple, imitating the aesthetic result of the valuable imperial marine material are discussed. Finally, the solubility of indigoids is discussed using modern approaches of physical chemistry.

#### **2. Historical Written Sources and Legends for Shellfish Purple**

The first inscription referring to the royal color is revealed in Amarna letters, dated to the 14th century BCE, in which blue-purple wool objects are described [10]. Another important written source comes from the Mycenaean Linear B clay tablets of Knossos dated to the 13th century BCE [11]. The masculine (po-pu-re-jo) and the feminine (po-pu-re-ja) adjectives are cited in tablets KN X976 and KN L474, respectively [11]. According to the Acts of the Apostles (Chapter 16), Lydia of Thyatira, who was baptized by St. Paul and is the first documented convert to Christianity in Europe, was in the purple-dye trade. Purple produced from mollusks is mentioned several times in the Bible as well as by Homer, Aristotle, Herodotus, Aeschylus and several others [12]. An impressive written example for the use of purple in Byzantine times is the inscription of the mosaic of the Byzantine Emperor, John II Komnenos (΄Ιωάννης*B Ko*μνην*ó*ς) in Hagia Sophia (= Holy Wisdom), which was the cathedral and the spiritual center of the Byzantine empire, as it was the Parthenon for Athens, which was dedicated to the Goddess of Wisdom, Athena. The mosaic in Hagia Sophia is dated to the 12th century CE and the inscription reads: "John faithful in Christ God, emperor of the Romans the Komnenos and king born in the purple".

The history of shellfish purple is full of legends with the most famous being the one describing the discovery of Murex shellfish by Hercules' hound. The scene is represented in the exergue of coins with the ambrosial rocks [13,14], as well as in paintings, such as the oil on panel artwork by the 17th century Flemish master Peter Paul Rubens [8]. Murex shells have been represented in various objects since antiquity, as shown for instance in the Geek krater and coin of Figure 1 [15,16].

**Figure 1.** (**a**) Pottery krater painted with red bands and row of parallel murex shells (13th century BCE). (**b**) Silver coin with the Head of Demeter (observe) and barley–ear with murex (reverse; 4th century BCE). Both objects were found in Greece [15,16]. © The Trustees of the British Museum. Shared under a Creative Commons Attribution–NonCommercial–ShareAlike 4.0 International (CC BY–NC–SA 4.0) license.

#### **3. Biological Sources and Chemical Composition**

Shellfish purple originates from marine mollusks of the Muricidae family, such as *Hexaplex trunculus* L. (*Murex trunculus*), *Bolinus brandaris* L. (*Murex brandaris*) and *Stramonita haemastoma* (*Thais haemastoma*). These three species live in the Mediterranean basin where the use of the purple pigment/paint in prestigious archaeological findings has been revealed. In the Aegean Sea, *M. trunculus* is the most abundant followed by *M. brandaris* and finally *T. haemastoma* species. However, it should be noted that the mollusk populations have been affected by the recent effects of sea pollution and global warming [17]. Surprisingly, *T. haemastoma* mollusks once described in the taxonomic literature as common in the Levant basin were not found in a recent survey [17]. In the same investigation, a decrease in the population of *M. trunculus* was reported. These observations are schematically presented in Figure 2.

**Figure 2.** Past and present abundance categories for *Murex trunculus* and *Thais haemastoma* populations in the Levant basin [17]: 0 = not found; 1 = rare; 2 = common; 3 = very common.

Photographs of *M. trunculus* mollusks, which were collected in the Thermaic (Thessaloniki) gulf in the Aegean Sea, are shown in Figure 3a. The coloring compounds do not exist in the living mollusks, but they are produced by precursors that are called chromogens. The latter are present in hypobranchial glands of the mollusks and undergo spontaneous chemical reactions under exposure to light and air thus yielding to the coloring components of shellfish purple [9,18]. For example, the reaction series for the conversion of tyrindoxyl sulfate (precursor) to 6,6 –dibromoindigotin, DBI (coloring compound) is schematically illustrated in Figure 3b [9,18]. The conversion is fast and it takes about 5–10 minutes after exposure to air and sunlight. By analogous reaction series, the other coloring compounds of shellfish

purple are produced: Indigotin (IND), 6–bromoindigotin (MBI), indirubin (INR), 6 –bromoindirubin (6 MBIR) 6–bromoindirubin (6MBIR) and 6,6 –dibromoindirubin (DBIR).

(d)

**Figure 3.** (**a**) *M. trunculus* mollusks. (**b**) Conversion of tyrindoxyl sulphate to DBI. (**c**) Structures of the coloring compounds of shellfish purple: Indigotin (IND), 6–bromoindigotin (MBI), 6,6 –dibromoindigotin (DBI), indirubin (INR), 6 –bromoindirubin (6 MBIR) 6–bromoindirubin (6MBIR) and 6,6 –dibromoindirubin (DBIR). (**d**) Solutions of the compounds in dimethyl sulfoxide (DMSO), 20 μg mL<sup>−</sup>1.

The compounds of Figure 3c were detected in extracts of *M. trunculus* [19–21] and *M. brandaris* [22]. However, only MBI, DBI and DBIR were detected so far in extracts of *T. haemastoma*. Four more analogues of brominated and unbrominated indirubins, which were detected in *M. trunculus* extracts, were not included in Figure 3c [23]. These isomers have been scarcely reported, as they were found in extremely small amounts [23]. In early reports, DBI was considered to be the marker for the identification of shellfish purple, as this was the first compound isolated from the raw purple pulp [24]. However, none of the brominated molecules of Figure 3c is found in any colorant other than the shellfish purple and therefore any of these molecules can be considered as markers for the presence

of shellfish purple. Moreover, DBI is not necessarily the most abundant component of the purple extract. Chromatographic results have shown that in some *M. trunculus* extracts, MBI is found in larger amounts than DBI [19–21].

Today, the compounds of Figure 3c can be produced in pure forms and in relatively large amounts following synthetic routes [20,25–27]. Photographs of standard solutions of synthesized shellfish components are provided in Figure 3d. Solutions of indigoids (IND, MBI, DBI) are bluish whereas solutions of indirubinoids (INR, 6 MBIR, 6MBIR, DBIR) are reddish. The latter are usually found in shellfish extracts in small amounts. Focusing now on the indigoids it is observed that the DBI solution has a bluish-purple color whereas the MBI and particularly the IND solutions are blue. Consequently, the color of a painted/dyed object can vary, depending on the relative composition of the fixed purple pigment/dye. For example, it was shown that wool dyed with (i) IND is blue, (ii) DBI is purple and (iii) MBI is in between blue-purple [28,29]. Microscopic and chromatographic studies in a 1500-year-old textile dyed with shellfish purple, showed that blue-purple and red-purple fibers were rich in IND and DBI, respectively [30]. In other studies on archaeological textiles, it was suggested that shellfish purple in antiquity was sometimes combined with red anthraquinones, when high proportions of IND were attached on dyed textiles [31–34]. High proportions of IND may have resulted because of the composition of the source mollusk or the dyeing conditions, as both affect the resulting color [28,29]. High proportions of IND may have resulted in purple dyes that looked too blue and therefore mixing with a red colorant was probably necessary to adjust the color and to make it look more purple [31].

#### **4. The Discovery of the Purple Colorant**

The origins of the purple dyeing or pigment production industry could have been in the Persian Gulf [9]. P. G. Moatsos wrote in 1932 that probably the Sumerians discovered the use of shellfish purple as a dye for textiles [35]. However, to date, there is no archaeological or scientific evidence to support this hypothesis. The earliest scientific evidence suggesting purple dyeing in the Gulf area goes back to only 3rd–5th century CE [36]. Fragments of pottery vessels used as dyed baths were analyzed revealing the presence of shellfish purple [36].

The first confirmed use of shellfish purple in wall paintings comes from the analyses of samples found in the islands of Santorini and Rhodes in the Aegean Sea [37–43]. The samples are dated to the 18th–17th century BCE. Among the findings in which the presence of shellfish purple was revealed by chemical analysis was a lump of pigment (Santorini). Chromatographic studies showed that the relative chemical composition of the pigment was similar with the compositions of purple paints removed from local wall paintings, proving indisputably the use of the pigment in the local painting art and suggesting pigment production in the Aegean in the Late Bronze Age [42,43]. The pigment was rich in chalk-aragonite and was presumably prepared by immersion of the chalk in the vat serving for the dyeing of textiles [43]. The procedure for the preparation of the pigment purpurissum, documented by Pliny the Elder [44], is reported to occur in the same dye bath, which also served for the dyeing of purple textiles [45]. It has been therefore suggested that the production of organic pigments in ancient times was probably carried out within the existing dyeing workshops using white inorganic substrates, which gave body to the purple pigment [43,45]. This argument is supported by the chemical results reported for samples that were extracted from a painted burial klinai and a textile fragment, both found within the same tomb, in Daskyleion (5th century BCE) [46]. The relative compositions of the shellfish purple pigment and dye used in the two objects were similar according to chromatographic results [46], thus offering support to the scenario that pigment production and dyeing were probably carried out within the same batch reactor-vessel. If this is true, then it serves as indirect evidence for the presence of a shellfish dyeing textile industry in the prehistoric Aegean where the purple pigment production has been confirmed.

The oldest textile fragments dyed with shellfish purple were found in archaeological sites in Syria and particularly in Chagar Bazar (18th–16th century BCE) [47] and in a tomb complex in Tell Mishrife (Qatna), which had remained sealed since the 14th century BCE, at the time of the Hittite conquest of Syria [48]. The earliest scientific result supporting the development of a shellfish dyeing industry was achieved by analyzing samples of pottery vessels from a Minoan facility (ca. 1800/1700 BCE) in Alatsomouri–Pefka, in the island of Crete in the Aegean Sea [49]. These Cretan findings are earlier than the discovered Phoenician pottery sherds, suggesting that the *Murex* purple of the Minoans became the Tyrian purple of the Phoenicians and later the imperial purple of Byzantium [50]. Consequently, the findings in Chagar Bazar and Alatsomouri–Pefka are the earliest textile fragments and dyeing facilities associated with shellfish purple.

The discussion now focuses on the shell deposits found in the Mediterranean. In the eastern shores of the basin, significant deposits of shells have been discovered in various sites in the Aegean islands, which are dated to the 18th century BCE [51,52]. The earliest archaeological evidence comes from Crete and corresponds to the Old Palace Period and particularly in 19th–18th century BCE [53–55]. There is no evidence suggesting that the famous heaps of mollusk fragments found in Tyre are associated with a dyeing industry, which is dated earlier than the 17th century BCE [56]. Consequently, the available scientific and archaeological evidence suggest that the Minoans in the Aegean and probably the Hurrians in Syria have pioneered the extraction of shellfish purple centuries before the Tyrians.

On the other hand, there is no doubt that the purple dye industry of the Phoenicians became famous and dominated the market from the Homeric times (9th–7th century BCE) onwards [54]. The Phoenicians spread their trade and founded cities, ports and bases up to the Black Sea and across the Mediterranean basin and beyond the Pillars of Hercules. For example, the Phoenicians founded Gadir (Cádiz, Spain) and probably had visited the Canary Islands. In fact, the word "Phoenician" comes from the Greek word "Phoenix" (ϕ*o*ῖνιξ) meaning "red-purple". Consequently, the Greeks had associated the purple pigment/dye with the Phoenicians. However, based on the available archaeological and scientific data, the shellfish purple industry was developed by the Minoans and probably the Hurrians before the Phoenicians.

Finally, it should be stressed that thousands of *M. trunculus* shells were found in Coppa Nevigata, Italy [56–59]. The accumulations of shells are present in the Early Proto-Apennine layers [58,59]. In particular, the collection of Murex began in the 19th–18th century BCE and continued throughout the entire Bronze Age, reaching its highest peak in the 15th and 14th centuries BCE [59]. The large quantity and the condition of the shells in Coppa Nevigata suggest that they were likely to have been used for the extraction of purple [57–59]. This may be the earliest known for the industry of purple, predating that from the Aegean islands [56,58]. It is interesting to note that excavation data indicate the existence of trading activities between Coppa Nevigata and the islands of the Aegean Sea [57].

#### **5. Identification of Shellfish Purple in Archaeological Objects**

### *5.1. Paints, Pigments and Identification Methods*

The goal of this section is twofold: (i) To summarize the scientific identifications of shellfish purple pigments applied as paints on wall paintings and some other objects and (ii) to briefly discuss the analytical techniques used for identification purposes. Both, previously reported identifications and analytical techniques are summarized in Table 1 [37–43,46,60–74]. The latter shows that shellfish purple was used in different objects for painting purposes, uninterrupted by all major pre-Roman civilizations of the Aegean, starting with the Minoan period of the Late Bronze Age up to the Hellenistic period. The use of shellfish purple was disrupted abruptly with the conquest of Constantinople by the Ottomans (1453 CE). At that time the capital of the Byzantine Empire was the major processing center of the valuable material. Around the same time in the West, a decree (1464 CE) of Pope Paul II, stipulated the replacement of shellfish purple by kermes to dye ecclesiastical robes. The two events combined, brought the historical use of shellfish purple to an end. According to Table 1, the most recent historical object painted with true purple is a wall painting in France dated to the 12th century CE [74]. It was suggested that probably the purple pigment was transported there from the eastern coast of the Mediterranean Sea during the time of the Crusades [74].


*Sustainability* **2019**, *11*, 3595

**Table 1.**

Identifications

 of shellfish purple in

pigments/paints

corresponding

 to various historical periods and

geographical

 locations. The techniques used for

pressure

x–ray (EDX) spectroscopy;

photoionization–mass

 spectrometry;

 XRF: X–ray fluorescence;

 MCT:

 n.a.: not available.

Microchemical

 testing; Raman: Raman spectroscopy;

 SEM–EDX: scanning electron microscopy (SEM) coupled to energy dispersive

As described in Table 1 a large variety of techniques were used to identify paints and pigments of shellfish purple. The presence of bromine (Br) in the composition of the purple material (Figure 3c) offers the option to apply elemental analysis techniques, such as X–ray fluorescence (XRF) and scanning electron microscopy (SEM) coupled to energy dispersive x–ray (EDX) spectroscopy (SEM–EDX). The results should be carefully interpreted considering that lichens can contain significant amounts of Br. An example of a SEM–EDX study is provided in Figure 4, which shows the results for a purple painted sample extracted from a burial klinai [46]. The detection of Br is indicative for the use of true purple. Furthermore, the presence of an aluminosilicate compound and gypsum is supported by the results of Figure 4, as Al, Si, Ca and S are included in the detected elements of the EDX spectrum. FTIR measurements confirmed the presence of shellfish purple, kaolinite and gypsum in the archaeological sample [46]. Interestingly, the use of kaolinite in painting techniques of shellfish purple were previously reported in the analyses of the Darius I stone painted jar [62] and the pigment powder of Pompeii [73]. According to Table 1, vibrational spectroscopy, and particularly FTIR and Raman spectroscopy, is another set of techniques which was effective to identify purple pigments and paints in objects of the cultural heritage.

**Figure 4.** SEM–EDX analysis of a purple painted sample, which was extracted from a burial klinai, in Daskyleion, Turkey. Reprinted with permission form reference number [46].

High performance liquid chromatography–diode array detector (HPLC–DAD) is the most efficient among the techniques listed in Table 1 for the analysis of shellfish purple, as it provides: (i) Detection and identification of all seven components within a single measurement-chromatogram, (ii) semi-quantitative results relative to the composition of the detected purple pigment and (iii) complete quantitative results. Examples for these valuable sets of HPLC results are provided next.

The chromatogram for a purple pigment that was found in the Akrotiri settlement is shown in Figure 5. The sample was included in a previous study and labeled Ak3 [42]. The chromatogram is presented herein for the first time. INR was not detected in the sample extract. The other six components of shellfish purple are identified by HPLC, which therefore provides very rich results in terms of detected compounds.

**Figure 5.** Chromatogram of a purple pigment from Akrotiri at 288 nm (Ak3 in [42]).

The HPLC peak areas (semi-quantitative results) are extremely useful to compare shellfish purple pigments from different samples. HPLC peak areas of purple components were reported in the literature for the samples of Minoan [42] and Persian origin [46,62] of Table 1 and these are summarized in Figure 6. In particular, relative integrated HPLC peak areas measured at 288 nm and normalized to the highest value (taken as 100%) are shown in Figure 6 for seven archaeological samples: Ak1 (pigment), Ak2 (wall painting), Ak3 (pigment) come from Akrotiri–Santorini, Ra (wall painting) from Raos–Santorini and Tri (pigment) from Trianda–Rhodes [42]. Samples Dar and Dask were extracted from the Darius I jar [62] and the burial klinai in Daskyleion [46], respectively. The HPLC peak areas for INR, 6'MBIR and 6MBIR were extremely small and therefore the sums of these areas are included in each sample column of the graphs of Figure 6.

The results of Figure 6a show that DBI is the dominant compound in the chromatograms of the five Minoan (Ak1, Ak2, Ak3, Tri and Ra) samples. The HPLC graphs gave moderate areas for MBI and small areas for IND and DBIR. The sum, INR + 6'MBIR + 6MBIR, is negligible or even zero (Ak2 and Ra) as revealed in Figure 6b. A similar HPLC profile was collected for the Dar sample [62], which, as shown in Figure 6, does not have any major difference compared to the Minoan samples. Trade routes between the southern Aegean islands, such as Santorini and Rhodes, and the Phoenician coast were developed since the prehistoric times and this may be depicted in the similar results of the Dar and the Minoan samples. However, the Dask sample is different: The HPLC peak area of DBI is small whereas the areas of IND, MBI, DBIR and the sum of INR + 6'MBIR + 6MBIR are large compared to the Minoan and the Dar samples. Consequently, a different processing method and/or molluskan sources should had been developed/used in Daskyleion, which is located close to the Marmara Sea, in the mainland of Turkey.

The relative peak areas reported in Figure 6 are not actual mass compositions of the samples. As mentioned previously, HPLC can provide complete quantitative results provided, however, that a HPLC calibration method is developed. The coloring components of interest should be available in pure form to be used as standards for quantitation purposes. A calibration procedure for shellfish purple pigment was previously developed and showed that the relative percentage of IND reported commonly using the HPLC peak areas at 288 nm (e.g., Figure 6) is in fact an overestimation of the actual mass content of IND [75]. On the other hand, the corresponding relative percentage of DBI, which is calculated using the HPLC peak areas at 288 nm, is an underestimation of the actual mass content of DBI in the shellfish pigment [75]. Moreover, the calibration process showed that low limits-of-detection (LODs) can be achieved with HPLC for the seven components of shellfish purple, ranging within 0.02 to 0.05 μg mL−<sup>1</sup> [75].

Finally, as described in Table 1, other methods employed to analyze and identify shellfish purple pigments/paints in objects of the cultural heritage are mass spectrometry (LC–MS, direct exposure–mass spectrometry—DE–MS) as well as imaging techniques including FTIR and XRF imaging. The latter are useful to reveal the distribution of shellfish purple on a painted surface, which, in turn, is sometimes useful to understand the applied painting technique.

**Figure 6.** (**a**) Relative integrated HPLC peak areas measured at 288 nm and normalized to the highest value (taken as 100%) are shown for seven archaeological samples. The 18th-17th century BCE samples come from the Minoan sites in Akrotiri–Santorini (Ak1, Ak2, Ak3), Raos–Santorini (Ra) and Trianda–Rhodes (Tri) [42]. Samples Dar and Dask were extracted from the Darius I painted jar [62] and the burial klinai in Daskyleion [46], respectively, both dated to the 5th century BCE. The sums of the HPLC peak areas of INR, 6'MBIR and 6MBIR are included in the graph. (**b**) The upper (80%–100%) part of the (a) graph is enlarged.
