Identification of purple by means of pyrolysis gas chromatography/mass spectrometry

Abstract

A multinational EU-sponsored research project [1] was dealing with ancient natural organic colouring materials applied to objects of art and archaeology, whereby the identification of these old dyes was a major aspect. The results on organic dyestuffs found on such antiquities are also of outstanding importance to help restoreres with the application of adequate materials.

For the present studies we chose Tyrian/Royal purple, as this was not only used for dyeing textiles, but also - upon conversion into a pigment - for painting purposes, thus being applied over centuries for a variety of works of art.

The precursors of purple substances can be found in several species of molluscs like Bolinus brandaris and Murex trunculus. After isolation from the mollusc tissue the crude purpuric sap contains a variety of structurally related components (indigotin, indirubin, monobromoindigotin, monobromoindirubin and dibromoindirubin), the main colouring matter of course being 6,6´-dibromoindigotin [2].

Up to now a sample had to be taken irreversibly from the cultural object in question, being big enough to provide - upon extraction - enough material for (e.g.) HPLC analysis, a horror for everybody who wants to avoid even tiny destructions of a culturally valuable object [2]. To cope with such problems our studies were focused on pyrolysis capillary gas chromatography (Py-CGC), thus reducing right from the beginning the necessary quantity of the sample [3-15]. While in the beginning the pyrograms of the samples obtained by FID (flame ionization detector) were straightforwardly identified by comparison with reference compounds [6-8], we only shortly later used a Py-CGC/MS equipment (Py-CGC coupled to a mass spectrometer) [9-11]. Of course this method was tested step by step with pure reference compounds (indigotin, indirubin, monobromoindigotin, monobromo- and dibromoindirubin), with mixtures therof, and with samples of the natural dyestuff in question, in particular with purple (from Murex trunculus). Although all these tests yielded positive results allowing unambiguous identifications, one major problem still had to be overcome: up to this stage the sample demand of this method in fact was already very low, but it could not be considered as non-destructive. So we studied not only the pyrograms of the natural organic dyestuff, but also as a next step those of their possible carriers, e.g., of wool, cotton, silk, as well as of paper and parchment [12-15]. When applying the Py-CGC/MS method to purple simultaneously with its respective carrier we of course had to avoid interferences of the pyrograms. This was accomplished by using SIM (selected ion monitoring) instead of TIC mode (total ion current) of mass spectrometer. Thus we are able now to analyze the colouring materials by Py-CGC/MS while still fixed to their carriers, i.e. without cutting out a large sample (since we now need less than 4 mg of some tiny fibers) and without applying any prior extraction. Our studies showed that the Py-CGC/MS method proved to be useful and almost non-destructive for dealing with such analytical problems. The present case study will explain the method a little more in detail and it will illustrate its efficiency by showing some examples, only for studies in which purple had to be identified. In addition to purple we optimized the Py-CGC/MS method for altogether twenty natural organic dyes like indigo, weld, saffron, turmeric, brazilwood, logwood, madder, cochineal, dyer´s bugloss, Persian berries and safflower.