1The purple of the ancients is undoubtedly the most fascinating and mystifyingly complex pigment of all the natural organic colorants. In the past two decades, this mollusc pigment has been the focus of increased historical and chemical research. Classical authors, such as the 4^th century BC Greek philosopher Aristotle1 and the 1^st century AD Roman historian and naturalist Pliny2 wrote about it. Much more recently, D. Cardon3 and R. Haubrichs4 have reviewed its history. The chemistry of this purple pigment has been reviewed by C. J. Cooksey,5 and analytical methods have been developed for multi-component identifications of Muricidae pigments via liquid chromatography.6

2These purple pigments were extracted from the hypobranchial glands of certain Muricidae sea snails inhabiting the Mediterranean and nearby waters. The three main mollusc species that have been associated with purple dyeing in the Mediterranean region have appeared in the literature under various nomenclatures, and are Hexaplex (= Murex = Phyllonotus = Trunculariopsis) trunculus, Bolinus (= Murex = Phyllonotus) brandaris, and Stramonita (= Purpura = Thais) haemastoma.7 They are shown in fig. 1. The extracted colorants were used as a paint pigment and as a textile dye, though the former usage probably chronologically preceded the latter. Purple and violet garments bestowed upon the owner an aura of power and sanctity and therefore these textiles were the prerogative of sovereigns, generals, eminent officials and high priests. Consequently, these textile dyes have been referred to as “royal purple”, “imperial purple”, “priestly purple”, and, most commonly of all, “Tyrian purple”, after one of the main cities of the Phoenicians, who are credited with at least the development of this dye and its trade.

Fig. 1 — Three Muricidae sea snails inhabiting the Mediterranean sea and processed in Antiquity to extract the purple pigment (from left to right): Bolinus brandaris, Hexaplex trunculus and Stramonita haemastoma.

(Courtesy of Eretz Israel Museum, Tel-Aviv.)

3As this paper was originally delivered in Athens, it is interesting to note the interconnection between the Hebrew and Greek versions of the two biblical forms of purple. As such, the names of these two “purples” mentioned in the Hebrew Bible are first cited in Exodus 25: 4 and are denoted in Hebrew as Tekhelet and Argaman. According to traditional accounts, the first translation of the Hebrew Bible into another language –Greek– was begun in the 3^rd century BC. Furthermore, as this project was undertaken by about seventy Rabbis who knew Greek, this body of work is known as the Septuagint. In that translation, Tekhelet is rendered as ὑάκινθον (yakinthon), that is, a hyacinth colour, and Argaman as πορφύραν (porphyran), a purple. There is an interesting Talmudic legend, from some time in the first two centuries AD, regarding this literary venture, and it is described in the Babylonian Talmud (Tractate Megillah 9a) as follows: “it is related of King Ptolemy that he gathered seventy-two Elders and placed them in seventy-two [separate] houses, without revealing to them why he gathered them. He entered each one’s house and said to them: ‘Write [i.e. translate] for me the Torah of Moses your master’. God then gave wisdom in the heart of each one and they all concurred in one identical erudition [i.e. translation]”.

4The references to these two biblical chromatic names are with respect to the colours of woollen dyes produced from the pigments extracted from certain sea snails. The exact colours of these two textile dyes have been disputed for nearly two millennia, although the past century has seen more intensive discussions of this topic, thanks, in part, to the advances made in modern analytical instrumentation.

5This laboratory has conducted intensive research on the purple mollusc pigment for about a quarter of a century in order to decipher the colours and chemical constitutions of these two biblical dyes, as well as, in general, to have a better scientific understanding of the production and usage of this pigment in Antiquity. The analytical investigations involve a multi-stage approach consisting of the following steps:

• Extraction of the purple pigment from the glands of modern sea snails;

• Dissolution of the sample with an optimal solvent;

• Development of an analytical HPLC separation method for the colorants constituting the pigment;

• Building a chromatographic and spectral data base of all the components;

• Application of the method to modern and archaeological pigments;

• Determining the malacological provenance of the ancient pigment.

1. Extraction of the purple pigment from the Muricidae family of sea snails

6The Hexaplex trunculus molluscs produce considerably more pigment than either Bolinus brandaris or Stramonita haemastoma sea snails. The latter two produce red-purple pigments, and the H. trunculus can also produce similarly coloured reddish-purple pigments, but other H. trunculus snails can produce bluish-purple or violet pigments. Analytical research has shown that in ancient times it was the H. trunculus that was mainly used for either red-purple or blue-purple dyes, and is thus the most historically important purple-producing sea snail.

7The chromogenic precursors to the final purple pigment are actually colourless in the hypobranchial glandular fluid of the live snail. This gland can be reached by breaking the snail’s shell with a hard object (e.g. stone or hammer) and exposing the vein. Upon excising the gland with a sharp object (knife or scissors) or by simply puncturing it, the enzyme that is present in a different compartment of the gland comes into contact with the precursors. Consequently, in the presence of air and light, this enzyme transforms the colourless precursors via various complex photo-oxidative processes into the final purple pigment.8 In the H. trunculus mollusc, the first short-lived colour of the exposed chromogens is white, which then spontaneously turns to yellow, then green, and finally purple. The development of the final purple colour stage is greatly hastened if it is exposed to direct sunlight as opposed to room lighting conditions.

8The next step is to analyze this pigment qualitatively and quantitatively.

2. Finding the optimal solvent for the dissolution of purple pigments

9In order to perform any chromatographic analysis via the HPLC method, the sample must be dissolved prior to its injection into the instrument. The major problem with the water-insoluble purple pigment, which has been shown to consist of various indigoids and related colorants,9 listed in table 1, is that these compounds are only sparingly soluble, even in hot organic solvents. In general, the solubilities of the indigoid dyes decrease from indigo to monobromoindigo to dibromoindigo. Previously, in order to dissolve the primarily indigoid-based pigment, glacial acetic acid, pyridine, and dimethyl formamide (DMF) were used for this purpose. However, since 2001, this laboratory has found that dimethyl sulfoxide (DMSO) is the optimal solvent at the elevated temperature of 150°C. The laboratory had already observed that DMSO is an excellent solvent for red safflower dyeing,10 which led to its application to indigoids as well.11 This solvent has a high normal boiling point (189°C) and, with the regular necessary laboratory precautions, is relatively safe to use. This high temperature, for the duration of 5 minutes, increases the dissolution of the sparingly soluble dibromoindigo and, when performed under subdued lighting conditions, this treatment does not cause any detectable amount of debromination.

10For a typical analysis, a micro-sample whose mass can be even less than 10 µg (barely visible to the naked eye) is placed in a 2-mL glass vial; 100-400 µL of DMSO is added to it and placed in a dry-block heater set to 150°C and heated for 5 minutes. Subsequently, microfiltration of the solution is accomplished via centrifugation for 5 minutes by means of a centrifuge tube assembly composed of a nylon filter of 0.2 or 0.45 µm pore size housed in a polypropylene body and inserted into the accompanying Eppendorf vial.

3. Analytical method development for the HPLC separation and detection of all the colourants constituting the purple pigment

11The method by which components are extracted or washed out of a chromatographic column is known as “elution”, and the mobile phase (the solvent) that performs this task is termed the “eluent”. A suitable elution method –combining solvents, concentrations, and gradient times– is needed for the separation and detection of all the possible colorants constituting the purple pigment. For this purpose, a ternary solvent system was developed, consisting of methanol, water, and phosphoric acid (5 % w/v, pH of 1.50 at 25°C).12 This linear step-wise gradient elution method was developed for ten standard purple-related dyes, and it produces good separation for all of the major dyes investigated. The common and abbreviated names for these dyes are listed in table 1. The resulting chromatogram –a graph of the chromatographic separation– is produced in less than 30 minutes, and is shown in fig. 2. The time that each component is retained in the separation column before eluting out is called the retention time, and labelled as t_R or R.T. Under appropriate conditions, it is characteristic of that component and is an aid to its identification.

Fig. 2 — HPLC chromatogram showing the separation of the dyes, where the left vertical scale is for the first three isatinoids, and the right scale is for the other dyes; the abbreviations are explained in table 1.

(Graph: Z.C. Koren.)

Fig. 3 — UV/Vis spectra of the isatinoids in DMSO solution.

Fig. 4 — UV/Vis spectra of the indigoids in DMSO solution (Graph: Z. C. Koren).

Fig. 5 — UV/Vis spectra of the indirubinoids in DMSO solution.

12The advantage of this method over those previously published is that it produces a good separation of the smaller isatinoid molecules that may be present in the pigment, whereas other methods did not include this group in their separation scheme. Although isatin (IS) from H. trunculus snails has been shown to be a truly minor component in the raw unprocessed purple pigment and in dyes subsequently produced from these pigments,13 the raw B. brandaris and S. haemastoma pigments analyzed by this method contain a major quantity of the brominated isatin, 6BIS, which may be useful in differentiating such pigments from the H. trunculus ones.14

13With this method, a spectrometric detection of the eluting components is made by the photodiode array (PDA) detector (also abbreviated as DAD: diode array detector). It produces a UV/Vis absorption spectrum for each dye detected. These are shown in figs. 3-5 for the three related chemical groups that constitute the purple pigment, the isatinoids, indigoids, and indirubinoids. Included in the figures are the molecular formulas and abbreviations of these dyes. From the respective spectra, it can be seen that the isatinoids maintain their colour when dissolved, producing yellowish solutions with a wavelength corresponding to the maximum absorption of visible light, λ_max, of about 415 nm. Similarly, the indirubinoids in DMSO-solution retain their reddish-purple hue in the dissolved state as in the solid state, with an average wavelength at maximum absorption of about 540 nm. While the navy-blue indigo produces blue DMSO-solutions, as expected, however, contrary to expectations, the violet MBI and red-purple DBI both produce bluish solutions when dissolved, and are visually nearly identical in colour to indigo solutions. The λ_max values from these three indigoid solutions hover at around 605 nm.

14The two properties produced from the HPLC-PDA methodology, the chromatographic retention time and the spectrometric UV/Vis absorption spectrum provide a dual means of positive identification of each dye. Thus, by comparing the retention time and the spectrum of each peak in the chromatogram produced from a real sample with those two properties from a standard component, the identity of the unknown component can be identified with a high level of confidence.

4. Application of the analytical HPLC method to the chromatic fingerprinting of modern Muricidae pigments

15In order to determine the zoological provenance of purple archaeological pigments produced from Muricidae snails, it is important to identify and characterize the presence of all the detectable colorants produced from each zoological species. This unique “chromatic fingerprinting” is important in order to perform archaeomalacological provenance determinations of ancient purple pigments and dyes. It is of course assumed that the living species available today were also in existence in Antiquity. Such a chromatographic multi-component characterization of selected samples from modern Muricidae species, H. trunculus, B. brandaris, and S. haemastoma, has already been performed and published.15

16The application of the analytical HPLC method to study the pigments from modern Muricidae snails has led to the following advances and discoveries made in this laboratory:

• Analysis of a raw unprocessed snail pigment;16

• Identification of DBIR17 in a H. trunculus snail;18

• Separation and detection of ten colorants that could constitute a mollusc purple pigment;19

• Detection of the following four dyes in H. trunculus snails: IS, 6BIS, 6MBIR, 6’MBIR;20

• Detection of 6BIS in B. brandaris and S. haemastoma.21

17In addition, the optimization of the all-natural fermentation dye vat consisting of H. trunculus snails that would have been practised in Antiquity was determined and reconstructed.22 A related natural process was also independently rediscovered by the late J. Edmonds23 and by I. Boesken-Kanold.24

5. Archaeometric analyses of residual purple pigments and dyes

Tel Dor – 6^th century BC pigment at a Phoenician dyeing installation

18A dark residual stain on a small pebble-sized piece of limestone, which was found in the conduit between two pits at the 6^th century BC site of Tel Dor in north-central Israel (fig. 6), was analyzed by dissolving it in hot DMF.25 After about a minute, the solution’s colour turned pale blue, indicative of the presence of an indigoid in the residue. A blue colour in a DMF extract may be due to the presence of indigo alone, which would thus indicate that this site was a dyeing installation for producing blue-dyed textiles from a flora source, most probably from the leaves of the woad plant, Isatis tinctoria. Alternatively, if the solution’s blue colour was also a result of the presence of brominated indigoids (see above), which can originate only from malacological sources, then this would indicate that the vat installation was for the production of purple mollusc dyes.

Fig. 6 — The two-pit installation connected via a conduit at Tel Dor.

(Photo: Z. C. Koren.)

Fig. 7 — Purple-stained potsherd excavated at Tel Kabri.

Fig. 8 — The King Darius I stone jar.

19Prior to HPLC analyses, this solution was then analyzed via visible spectrophotometry, whereby a spectrum of light absorption is obtained at various wavelengths for the dissolved sample. While limited in its powers of discernment, it was nevertheless a simple instrumental tool and useful when carefully interpreted. The results of the analyses show a spectrum for the dissolved archaeological residue with a visible wavelength at maximum absorption, l_max, of 600 nm. Comparing this value with the comparable wavelength for indigo (613 nm) and with that for DBI (59 nm) shows that the now soiled residual pigment was much richer in the red-purple DBI than in indigo, and hence the source for this pigment is in fact molluscan. The possible presence of MBI (absorbing at 607 nm) can also not be ignored. Hence, this installation served in fact for the dyeing of real purple textiles. The limestone found between the pits may have been used for the production of an alkaline environment, which is necessary for the reductive dissolution of the dye to its leuco form during the dyeing stage.

20This type of visible spectrophotometric analysis was, as noted above, not detailed enough, and hence the advent of the HPLC method provided the fine qualitative and quantitative tuning that was essential for a complete determination of the dyes constituting archaeological pigments.

Tel Kabri – 7^th century BC purple-stained potsherd

21The first HPLC method to study an archaeological pigment from a dyeing vat was published in 1995.26 The beautiful potsherd sample, excavated at the Phoenician site of Tel Kabri in the North of Israel, is depicted in fig. 7. For this analysis, the detector was a variable wavelength detector, which measured the chromatogram at a fixed wavelength chosen at the beginning of the run. While this detector lacked the universality of the more modern photodiode array (PDA) detectors, it nevertheless provided accurate data for the most optimal wavelength, which in this case is about 600 nm. As previously noted, brominated indigoids and indigo have their absorption maxima near this wavelength. It is gratifying to note that at this wavelength (600 nm) the existence of DBIR is still detected, even though its wavelength at maximum absorption is about 540 nm. This was the first time that DBIR was found in an archaeological pigment.

Purple-painted royal jar of King Darius I – mid-5^th century BC

22The photodiode detector was used for the detection of an unusual royal object.27 This unique marble jar with the name of King Darius carved on it has scattered purple stains on its exterior and even on the base (fig. 8). Upon closer inspection, it can be deduced that the entire object was painted with a purple pigment in either a fresco or secco technique. The mortar used was probably kaolinite, painted afterwards with the purple pigment. The detailed chromatographic fingerprint showed that this pigment was obtained from a H. trunculus snail (or very similar species) that is rich in DBI and poor in IND (indigo). During the Persian Achaemenid period, the kings controlled a vast empire with easy access to Mediterranean and other shores where Muricidae snails were found.

23The function and purpose of this object are not entirely certain, but it was probably a gift from the king to a person whom he favoured. It is interesting to note a biblical parallel here. In the sixth chapter of the Book (or Scroll) of Esther in which, according to tradition, the story unfolds in the royal court of one of the Achaemenid kings, probably Xerxes, it states: “in this manner shall be done to the man whom the king desires to honour”.

Fig. 9 — The miniscule purple fabric found at Masada.

(Photo: Z.C. Koren.)

Fig. 10 — Chromatogram at 600nm of the dye extracted from the purple yarn from the Herodian fabric.

Herodian fabric – late 1^st century BC

24One of the more pleasantly surprising discoveries made in this laboratory was the discovery that a miniscule piece of fabric, measuring about 4×2 mm (fig. 9), found in a refuse ditch at the top of the palatial fortress of Masada in the Judean desert, probably belonged to the royal purple mantle of King Herod.28 This king was undoubtedly one of the most memorable kingly characters who reigned in Judea in ancient Israel. He built a number of palaces and other building projects, which some say were verging on megalo-maniacal in size.29 One of his most outstanding landmarks was the refurbishment of the Second Temple in Jerusalem, of which the famous external Western Wall still survives with its Herodian stone “grandeur”.

25The chromatographic result of an HPLC analysis (fig. 10) on a small dyed yarn from that weave bears the hallmark of a dye produced from a molluscan source. Comparing this dye composition with that of other snails, it is clear that the marine zoological source of this pigment is the H. trunculus snail (or some other species that is nearly indistinguishable from it). This pigment is rich in DBI, which is responsible for its reddish-purple hue. This then is also the colour of the biblical Argaman dye mentioned above, the first such dye found in ancient Israel.

Biblical Tekhelet dye

26Recently, this author analyzed a bluish-purple (violet) textile from Masada and determined that the dyestuff source is the IND-rich H. trunculus sea snail. This, then, is the only biblical Tekhelet dye so far discovered. More details about this analysis will be provided in a future publication.

27The analytical study of purple molluscan pigments and dyes on archaeological artifacts from more than two and a half millennia has shown major advances, especially within the last two decades. The archaeometric discoveries made in this laboratory can be summarized as follows:

• Decipherment of the method by which the ancients performed purple-dyeing by natural means;30

• HPLC analysis of a raw unprocessed purple archaeological snail pigment;31

• HPLC identification of DBIR in a molluscan archaeological pigment;32

• Discovery of the purple pigment as the sole paint pigment on a royal objet d’art from King Darius I;33

• Discovery of the royal purple mantle of King Herod I –the biblical colour Argaman produced from DBI-rich H. trunculus sea snails;34

• Discovery of the biblical blue-purple Tekhelet dye produced from IND-rich H. trunculus snails.