Electronic Supplementary Material 2 to the Article Submitted to Anthropological And

Electronic Supplementary Material 2 to the article submitted to Anthropological and

Archaeological Sciences

Colour in context. Pigments and other coloured residues from the Early-Middle Holocene site of

Takarkori (SW Libya)

Savino di Lernia (1, 2, *), Silvia Bruni (3), Irina Cislaghi (3), Mauro Cremaschi (4), Marina Gallinaro

(5,6), Vittoria Gugliemi 3), Anna Maria Mercuri (7), Giansimone Poggi (5), Andrea Zerboni (4)

(1) Dipartimento Scienze dell’Antichità, Sapienza Università di Roma. Via dei Volsci 122 – I- 00185 Roma, Italia. Phone: + 39 0649697092. (2) School of Geography, Archaeology and Environmental Studies, University of the Witwatersrand, Private Bag 3, Johannesburg 2050, South Africa. (3) Dipartimento di Chimica, Università degli Studi di Milano, via C. Golgi 19, I-20133 Milano, Italy. (4) Dipartimento di Scienze della Terra ‘Ardito Desio’, Università degli Studi di Milano, via L. Mangiagalli 34, I-20133 Milano, Italy. (5) The Archaeological Mission in the Sahara, Sapienza Università di Roma, Italy (6) Dipartimento di Storia, Scienze dell’Uomo e della Formazione. Università di Sassari, Italy. (7) Laboratorio di Palinologia e Paleobotanica, Dipartimento di Scienze della Vita, Università di Modena e Reggio Emilia, Viale Caduti in Guerra, 127, I-41121 Modena, Italy. * Corresponding author: [email protected] – www.acacus.org

SILVIA BRUNI, IRINA CISLAGHI, VITTORIA GUGLIEMI

METHODS OF CHEMICAL ANALYSIS

Spectroscopic techniques

Fourier-Transform Infrared (FTIR) spectra were collected on KBr discs of the powdered samples between 4000 and 400 cm-1 by means of a Jasco FT-IR 620 spectrometer. Spectra were acquired as sum of 32 scans and resolution was set at 4 cm-1.

1 Two different instruments collected Raman spectra. More precisely, a Jasco TRS-300 spectrometer equipped with a diode array detector was used, employing for excitation the 676 nm-line of a Kr+ laser. In order to limit interference due to fluorescence, the 785 nm-line of a diode laser was also used. In this case the Raman spectrum was collected by a Jasco RMP 100 microprobe equipped with a 50x objective and connected by fibre optics to the laser and to a Lot Oriel MS25 spectrometer with a CCD detector. The laser power at the sample was in all cases of few milliwatts.

Visible-NIR reflectance spectra were acquired on powdered samples between 400 and 2400 nm by a Jasco V-570 spectrophotometer equipped with an integrating sphere internally coated with

BaSO4.

X-ray Diffraction (XRD)

XRD analyses were performed using a PHILIPS PW 1820 vertical scan powder diffractometer, with

Cu-K incident radiation. Step-scanned data (2 = 0.02°) were collected in the 5° - 65° (2) range, in the  - 2 mode, with time per step ranging from 1 to 4 s.

For pigments scraped from grinding equipments, the few available grains were placed on a quartz single crystal sample holder and dispersed in amyl acetate.

Scanning Electron Microscope – Energy Dispersive X-ray (SEM-EDX) analysis

SEM-EDX analysis was performed by a JEOL 5500 LV EDS-equipped instrument (20 kV, 10–11 A, 2

μm beam diameter).

Gas Chromatography–Mass Spectrometry (GC-MS)

For the GC-MS analysis of the lipid fraction of ten pigment lumps, some hundreds of milligrams of each of them, from ca. 110 mg to 1 g according to the available amount, were extracted twice, each time using an volume of a mixed solvent chloroform/methanol 2:1 approximately double

2 than that of the powdered sample and sonicating for about 20 min (Evershed et al. 1990). The combined extracts were centrifuged to remove the suspended particulate matter and filtered using syringe microfilters. The solvent was then removed under a gentle nitrogen stream. The dried extract was treated with 20 L of BSTFA (N,O-bis(trimethylsilyl)trifluoroacetamide) with 1% TMCS

(trimethylchlorosilane), Fluka) for 5 min, then 50 or 100 L of 2,2,5-trimethylpentane were added, and kept for 1 h at 70 °C under agitation to obtain trimethylsilyl (TMS) derivatives. Tridecanoic acid was added to the extract as an internal standard prior to derivatization so that its concentration in the final solution corresponded to 0.1 g/L.

One L of the resulting solution was then analysed by a Gas Chromatograph coupled with a

Quadrupole Mass Spectrometer (Shimadzu GC-MS QP 5050). The MS transfer line temperature was kept at 260 °C. The chromatograph was equipped with a Equity-5 (Supelco) fused silica capillary column (poly(5% diphenyl/95% dimethylsiloxane, 30 m  0.25 mm i.d., 0.25 m film thickness). The injector was used in splitless mode at 270 °C and the detector temperature was also 270 °C . The chromatographic conditions were: 60 °C isothermal for 1 min, 20 °C/min up to 70

°C, 10 °C/min up to 240 °C, 4 °C/min up to 285 °C and isothermal for 40 min. The carrier gas was helium at a constant flow rate of 0.7 mL/min. The quantitative analysis of fatty acids was performed in Selected Ion Monitoring (SIM) mode and based on the preliminary analysis of a standard mixture.

The analysis was repeated on different layers of the samples with similar results.

For the GC-MS analysis of the protein fraction of the same pigment lumps samples, a procedure suggested in the literature to reduce the interference of some pigments in amino acid analysis was used (Colombini et al. 1999). In detail, some hundreds of milligrams of each sample, from ca. 100

to 800 mg according to the available amount, were extracted twice by 2 mL aqueous NH3 2.5 M in an ultrasonic bath for 2 hours each time. The dried extract was treated with 100 L of 6 M HCl in

3 0.3 mL microvials under Ar atmosphere in a heating block at 110 °C for 24 hours (Gimeno-

Adelantado et al. 2002). Following the literature (Colombini et al. 1998, 1999), 1 L of the

hydrolysate was dried under a gentle stream of N2 and then 15 L of N-methyl-N-(t- butylmethylsilyl)trifluoroacetamide (MTBSTFA), 40 L of pyridine and 2 L of triethylamine were added. 1 L of a 2.510-3 M solution of norleucine in 0.1 M HCl was also added as an internal standard. The derivatization was completed in 30 minutes at 70°C in a water-bath. 1 L of the derivatized sample was injected in the Gas-Chromatograph.

The analysis was performed using the same instrumental equipment described in the previous section. The chromatographic conditions were: initial isotherm at 100 °C for 2 min; from 100 °C to

280 °C at 6 °C/min; final isotherm at 280 °C for 10 min. All other experimental parameters were as described above. The Mass Spectrometer was operated in SIM modality. The relative weight percentages of 14 amino acids were thus determined: alanine, glycine, valine, leucine, isoleucine, proline, methionine, serine, threonine, phenylalanine, aspartic acid, hydroxyproline, glutamic acid, tyrosine.

The results obtained on archaeological pigment samples were compared to a laboratory database containing the results of the GC-MS analyses on reference proteinaceous binders. More precisely, the reference materials included: proteins such as collagen (Sigma), albumin (Fluka) and casein

(Bresciani); simple proteinaceous substances such as milk, egg yolk, egg white, rabbit skin glue

(Wien, Beck, Koller & Co and Lefranc & Bourigliois), fish glue; more complex binders obtained as mixtures of the above mentioned substances and other components (e.g. oil, resin, vinegar) prepared by restorer Desirée Snider according to the recipes from ancient treatises. Pure proteinic binders were analysed both as such and after drawing small fragments from samples of paintings again prepared by D. Snider with various inorganic pigments. The more complex binders, prepared as mixtures of more than one ingredient, were instead analysed exclusively from samples of

4 paintings. Linear discriminant analysis was performed using DTREG software in order to classify the binder in the sample from a rock painting.

References

Colombini MP, Fuoco R, Giacomelli A, Moscatello B (1998) Characterization of proteinaceous binders in wall painting samples by microwave-assisted acid hydrolysis and GC-MS determination of amino acids. Stud. Conservation 43:33-41

Colombini MP, Modugno F, Giacomelli A (1999) Two procedures for suppressing interference from inorganic pigments in the analysis by gas chromatography-mass spectrometry of proteinaceous binders in paintings. J. Chromatogr. 846:101-111.

Evershed R, Heron C, Goad LJ (1990) Analysis of organic residues of archaeological origin by high- temperature gas chromatography and gas-chromatography – mass spectrometry Analyst 115: 1339-1342

Gimeno-Adelantado JV, Mateo-Castro R, Domenech-Carbo MT, Bosch-Reig F, De la Cruz-Canizares J, Casas-Catalan MJ (2002) Analytical study of proteinaceous binding media in works of art by chromatography using alkyl chloroformates as derivatising agents Talanta 56:71-77.