IDENTIFICATION OF THE BYZANTINE ENCAUSTIC MURAL PAINTING IN EGYPT

Mediterranean Archaeology and Archaeometry, Vol. 15, No 2,(2015), pp. 243-256 Copyright © 2015 MAA Open Access. Printed in Greece. All rights reserved. 10.5281/zenodo.16612 IDENTIFICATION OF THE BYZANTINE ENCAUSTIC MURAL PAINTING IN EGYPT Basem Gehad*1, Mona Foad Aly2 and Hussein Marey2 1The Grand Egyptian Museum, Conservation Center, Cairo, Egypt 2, 3 Faculty of Archeology, Conservation Center, Cairo University, Egypt Received: 15/12/2014 Accepted: 29/04/2015 Corresponding author: Basem Gehad (basemgehad2013@gmail.com) ABSTRACT Encaustic painting uncovered in the hermitage of Apa Apollo at Baouit- Assuit, was studied by means of spectroscopic, chromatography as well as elemental x- ray fluorescence, the analysis re- vealed unique information’s about the composition of the organic binding medium, as well as it deformation and alteration pattern. The elemental analysis highlights also the types of pigments used in the artistic palette used by the painter in order to execute his paintings. Beeswax was proved to be used, from bees which feed on sunflower, lead was the major com- ponent of the orange pigment indicating the usage of minium, arsenic sulphide mixed with hema- tite was used for the brownish red color, a copper based blue pigment, probably the Egyptian blue as well as the green earth mixed with Attachmate was also used for green colors. The results of the study gives a new information’s about a unique paintings executed with a rare technique, in Egypt. KEYWORDS: Encaustic, Monastery, Mural Painting, Byzantine, Assuit, Beeswax, FTIR, GC-MS. 244 B.GEHAD et al. 1. INTRODUCTION was excavated. The monastery was founded at It is unknown exactly when and where wax the end of the fourth century, and the was first used as painting medium, but recent excavation showed that it had become studies highlight a lot of different examples prosperous by the sixth century. The monastery around the Mediterranean basin, and it is as- was destroyed by the twelfth century. sumed that the encaustic paintings started as a In 2005-2006 the IFAO mission uncovered a pure Greek painting tradition. unique encaustic painting inside the walls of the The Roman writer Pliny the Elder (1st centu- monks’ houses in the northern part of the her- ry AD) states that from the oldest Greek mas- mitage of the Apa Apollo at Baouit (Benazeth, ters –such as Polygnotus– onward, encaustic 2010 ) (Fig.1) was a common painting technique in ancient Greece and Rome. This text appears in chapters "on the artists who painted with encaustic by means of palette knife or brush" –as stated in the treatise’s index, which is in fact a history of ancient Greek and Roman painting (Rackman, 1952). Unfortunately most of the mural paint- ings painted in wax in Egypt cannot now be lo- cated, only few of them were uncovered in Figure 1 Encaustic mural painting on the northern wall, building no.1 , Baouit. Baouit – Assuit as well as the wadi el Natron. The problem is exacerbated by the lack of pub- A complete painted room known as sale 7 lished scientific reports for this kind of paint- was uncovered in 4 seasons of excavation and ings. On the other hand, another source of en- given conservation treatment. Unfortunately caustic paintings should be considered, as it the paintings were destroyed in 2010, The de- forms a part of the development of encaustic tachment and rescue operation was performed painting technology in Egypt, which is the by Christophe Guilbaud,Bruno Szkotnicki, Ab- mummy portraits or the so-called the fayoum ied Mahmoud and Basem Gehad in April portrait. Studies of these encaustic objects in- 2010,due to the collapsing of the mud brick form us about the technology of encaustic paint- walls of the house; now they are detached and ing in Egypt from the beginning of the 1st cen- kept in trays in the store rooms. tury C.E to the end of the 3rd century C.E . The paintings were uncovered in a construc- In 1991 the Institut Français d’Archéologie tion used as housing, nominated as Building 1, Orientale (IFAO) uncovered a hitherto un- excavated by Marie-Hélène Rutschowscaya and known painting of the annunciation in the Ramez W. Boutros, consisted of several rooms western semi dome of the church of the holy set around a courtyard. The most southerly of virgin in Dier el Surian (Innemee , 1999). During these rooms controlled access to the building, a continuation of the work in 1995 different with a door opening to the outside and another painting stratifications were identified; the four opening onto the courtyard. On the north side, layers identified are from different phases of the two rooms were separated by a corridor that led church. to a kitchen. On the east side stood an ensemble Another encaustic mural painting was of five vaulted halls. The walls and the vault of identified in one of the monks’ houses at baouit. the biggest of these (Hall 7) were decorated The town of Bawit (or Baouit in French) is locat- with mural paintings (Fig.2). ed between Dayrut and Asyut, 1.25 miles (2km) On the north side of this room, the paintings west of the village of Bawit and about 17 miles of the vault depicted episodes from the birth of (28 Km) south of El Ashmunin. Christ: the dream of Joseph, the voyage to Beth- The site is famous from the excavations lehem, the nativity with the miracle of Salome carried out there at the beginning of the and the shepherds, the presentation in the twentieth century. The monastery excavated at Temple and the adoration of the Magi. Below, at Baouit was dedicated to saint Apollo and called the base of the vault, is a geometric frieze (Bé- Apa Apollo.Not more than 10 percent of the site nazeth et al, 2008). Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256  245 IDENTIFICATION OF THE BYZANTINE ENCAUSTIC MURAL PAINTING IN EGYPT The hypotheses of painting using molten wax (or the application of hot instruments for painting with the wax) or the use of the Punic wax (saponification wax) mentioned by Pliny has been argued between different scholars. At present the most widely accepted hypothesis states that the ancient technique of encaustic painting consisted of the application of colored beeswax in its molten state, a theory which is based upon ancient literary sources. As the Roman writer Seneca wrote that the choosing and applying of the paint was performed rapidly (Doxiadis 1995), this would be possible Figure 2 Room no.7 , building no.1 , during excavation, only with the hot or molten wax. Petrie also photo to the north -west ( Courtesy of IFAO). mentioned that the environment of Egypt On the south side of the room, the vault bears would keep the wax melted or near melting the representations of a series of characters in point (Petrie 1911). which one can recognise two of the founders of But at the same time some scholars have not Bawit monastery and nine prophets, each of accepted this idea, disagreeing with the idea of whom holds a scroll inscribed with a passage using molten wax for the following reasons: from their prophecy. The base of the vault is molten wax cannot produce the long thin and decorated with a frieze of meanders alternating diluted brush strokes observed, as molten wax with birds. The wall itself is decorated with a sets and dries very fast, giving only thick and pattern of diamond shapes each holding a green short brush strokes, the faces indicate not brush leaf. Another geometric pattern, covered with marks but palette knife marks, which could large blooming flowers, is painted on the west only be reproduced by saponification of wax, wall. The east wall, of which only the lower part the terminological meaning of the word remains, holds three niches. The decoration of encaustic could refer to the method and the the largest, set in the middle of the wall, was tools but not the molten wax . found in the debris. It depicted Christ The second hypothesis is the use of water surrounded by seraphim and the symbols of the soluble encaustic or punic wax. According to four evangelists. ancient recipe given by Pliny, after bleaching A great effort is being made to reassemble the the wax using sea water and sun light, the wax fragments of the encaustic painting, which is a was then boiled with an alkaline solution representation of the saints and the holy family produced from the burning of ash, , as the in Egypt in a Coptic and byzantine style. water filtered from burned ash contains high The precise composition of the medium has proportions of potassium hydroxide and minor proved to be a continual source of disagree- proportions of sodium hydroxide. ment. The etymology of the word encaustic in The idea of using saponified wax or punic wax ancient lexicons and dictionaries is derived has met with great criticism, as there was no from the Greek word enkaio or enkaiein which direct reference to it, as well as problems con- means to burn or to be executed by fire (Lieber, cerning the color fading if the saponified wax 1840).Tthe meaning given by the lexicon indi- with its salts was used. cates the method used in the preparation, which The word sapo, Latin for soap, first appears in depends mainly on heating the wax before the Pliny the Elder's Historia Naturalis (Rackman, application. Pliny refers to the impregnation of 1952), which discusses the manufacture of soap the pigments with the wax by melting and from tallow and ashes, but before this a refer- preparing cylindrical shapes or color cakes. ence to soap production appears in The Ebers (Doxiadis, 1995). papyrus from Egypt, 1550 BC, indicating that the ancient Egyptians bathed regularly and combined animal and vegetable oils with alka- Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256 246 B.GEHAD et al. line salts to create a soap-like substance ( Scholl, (Ramer, 1979; (Meilunas, Bentsen & Steinberg, 2002). 1990; Pitthard, Vak, Griesser, Stanek, & The pigments used by the encaustic painters Laubenberger, 2007; Kühn, 1960), (Moretto, were, according to the recent analytical results Orsega, & Mazzocchin, 2011). Infrared spectra performed in both encaustic mummy portrait were obtained in the mid-range using JASCO- and paint saucers from the roman period, red 6100 FTIR spectroscopy. About 1 to 3 mg from lead (minium) for the orange red color, the archeological samples were ground with 99- Egyptian blue for the blue pigment, white lead 97mg of KBr in an agate mortar, Percentage or gypsum for the white color, ochre mixed transmittance was collected in the range of with jarosite and in some cases madder lake for 4000–400 cm−1 with 4 cm−1 resolution. the rose and the flesh tones, and ochre mixed. 2.5 Gas chromatography–mass spectrometry (gc-ms) 2. MATERIAL AND METHODS Gas chromatography – mass spectrometry is 2.1 X-Ray Fluorescence (XRF) one of the most effective methods widely used for the identification of organic material and the Non-destructive chemical- elemental analyses alteration aspect connected to the aging of these were collected for twelve samples using Niton materials. GC-MS has been widely used for the XLt -793 W portable XRF spectrophotometer identification of wax in general and beeswax device, that produced measurements in parts specifically, the method followed in this re- per million (ppm). The NITON XLt x-ray tube search was described in different research as based analyzer is a completely portable instru- (White, 1978), (Bonaduce & Colombini, 2004) ment with a one hand trigger operation and a and (Maia & Nunes, 2013). Type of GC/MS: touch screen with full navigation, a complete thermo scientific trace GC ultra-coupled with energy spectra view, and an RS232 download ISQ single Quadra pole MS, type of column TG5 port. All samples were exposed for a minimum MS, working protocol followed: initial tempera- of 180 seconds. XRF charts were produced us- ture 50 0C held for 2 min, 50 0C / min ramp to ing NITON xrf software. The principle compo- 180 0C, 3 0C / min ramp to 300 o C, 6 0C / min nents and the cluster analyses were performed ramp to 320 0C. using PAST statistical software program, ver- For sample preparation, approximately 3 mg sion 2.10. of beeswax was dissolved in 4 ml of chloroform; 2.2 SEM-EDX the solution was mechanically shaken for 2 min The scanning electron microscope used was a to complete dissolution of beeswax. Quanta 3D 200 I) (FEI Philips – Holland) cou- 3. RESULTS pled with EDX. Column pressure 60 PA, low 3.1 Plaster and paint substrate vacuum, In back scattered mode (BSED). The investigation of the cross sectioned sam- 2.3 X- Ray Diffraction (XRD) ples, using SEM, indicates six layer in the com- For the x-ray diffraction samples were analysis position of the mural painting subjected for the using panlytical X-pert pro with a Cu anode, study showed that the paint layer could be de- working at 40 mA / 45 kV. The samples were scribed according to the SEM-BSED investiga- analysed in a nondestructive mode without any tions as following:(Fig.3 ): sample preparation. An approximately flat sur- (a) Silty sand render with a thickness around faced archeological sample was attached into 2.5mm, this layer is reinforced with strips of the sample holder inside the xrd apparatus. The chopped straw, the strip length could be from data were interpreted using the embedded 300 to 500 µm, according to the SEM investiga- software. tion, the straw could be from Barley straw 2.4 Fourier Transform IR spectroscopy (FTIR) (Hordeum vulgare L.) (b) A thick lime based render with a thickness Fourier transform infrared spectroscopy has about 1.5mm, this layer is also reinforced with been widely used in the study of wax and pig- chopped straw, and the layer is rich with large ments, the method used was in accordance with Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256  247 IDENTIFICATION OF THE BYZANTINE ENCAUSTIC MURAL PAINTING IN EGYPT grain sand particles ranging from 100 to 150 3.2.1 Orange paint layer µm. Two samples from the orange paint layer (c) Moderate lime based coat, the thickness of were investigated and analysed by means of this layer is about 150 µm. SEM coupled with EDX,-indicating the presence (d) Two fine coats with very fine sand particles, of highly back scattered elements in the paint the inner is about 45 µm while the outer is layer, while elemental analysis using XRF and about 120 µm. EDX proved the presence of Lead (Pb) as a ma- (e) The paint layer which ranges from 5 to 20 jor element as well as Iron (Fe) and Mn,Cu,As. µm. (Fig. 5,6) Figure 5 XRF pattern for the orange pigment, lead was the major element indicated in the sample. Figure 3 SEM indicating the stratigraphic composition and the thicknesses of the encaustic paint layer and its substrate, Baouit, Building no.1. The X-ray diffraction analysis of the paint layer proved that the white preparation layers are mainly calcium carbonate (Fig. 4), which could be interpreted as a lime based plaster, executed on a silty sand render. The main components of Figure 6 EDX analysis for the orange paint layer indicat- ing the presence of Pb, Fe. Al, Ca and traces from Mn the plaster were quartz and calcite, while the and K. silty sand render represented by albite as one of its main components. From the XRD results, no The presence of red lead was confirmed by alumino-silicate materials were found (ex: Poz- means of XRD analysis, where both minium zolan) and that confirms that air lime and not and calcite were identified (Fig.7). hydraulic lime was used. 3.2.2 Yellow paint layer XRF analysis of three samples analysed from the yellow paint layer indicated that iron (Fe) was the most abundant element, with some traces of Ca, Cu. Two samples (bulk and powder) were ana- lysed using XRD. The analysis revealed the presence of natrojarosite, hydrated sodium iron sulphate, [NaFe3(SO4)2(OH)] for the lemon yel- low pigment used (Fig.8), a mineral previously Figure 4 XRD pattern for the encaustic mural painting reported as a yellow pigment in ancient Egypt substrate indicating the presence calcite and quartz, in- by Le fur (Le fur , 1994) and Colinart (Colinart, dicating the presence of lime based plaster , with no 2001). alumino -silicate materials. Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256 248 B.GEHAD et al. Figure 7 XRD pattern of the orange pigment indicating the presence of minium Pb3O4 and calcite CaCO3. Figure 8 XRD pattern for the yellow paint layer revealed the presence of natrojarosite, hydrated sodium iron sulphate, [NaFe3(SO4)2(OH)] for the lemon yellow pigment The pigment was also identified in different cannot be relied on as a solo analysis technique, painting examples from the old kingdom (Ak- whether the green pigment was iron based or hethotp mastaba 6th 6 dyn. Saqqara) towards the copper based was not clear. middle kingdom period, (ex.Karanak, Luxor), as well as some other examples from the Ptolemaic period. 3.2.3 Green paint layer The olive green pigment used in the monastery of Baouit was prepared in thin sections, which allowed a clear observation of large and well crystallized grains, embedded in a deep green matrix. Analysis of the green pigment using XRF indi- Figure 9 XRF result of the green paint layer indicating the presence of Cu, Ca Fe, and Pb. cated the presence of copper and iron as major components (Fig.9), with some traces of magne- EDX analysis revealed the presence of iron sium and Arsenic. The XRF elemental analysis (Fe), Calcium (Ca), Aluminum (Al), Magnesium Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256  249 IDENTIFICATION OF THE BYZANTINE ENCAUSTIC MURAL PAINTING IN EGYPT (Mg), potassium (K), Chloride (Cl), Copper (Cu). suggest that glauconite was used rather than These results confirm the presence of green Celadonite (Figs. 10, Table 1). earth, while the ratio of Iron (Fe) to (Mg) would Table no.1: elemental relative percentage resulted fom both XRF and EDX analysis of the green sample from Baouit encautic mural painting. XRF analysis EDX analysis Element Concentration in PPM Element Percentage (%) Sb 51 C 45 Sn 97 O 30 Pb 32 Mg 0.53 As 234 Al 1.22 Cu 2554 Si 3.59 Fe 1820 S 2.62 Mn 97 Cl 0.49 K 0.40 Ca 11.78 Fe 1.37 Cu 1.76 Figure 11XRD result for the first sample of the green Figure 10 EDX analysis for the green paint layer revealed paint layer indicating the presence of glauconite, calcite the presence of Fe, Ca, Al, Mg, K, Cl and Cu. and attachamite. Figure 12 XRD pattern for the second green sample indicating the presence of glauconite, calcite and traces of cela- donite Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256 250 B.GEHAD et al. The usage of green earth was conformed us- Table 2 relative percentage elemntal result using xrf and ing XRD analysis. Two samples were analysed, edx for the blue paint layer. in one sample glauconite and atacamite were EDX (%) XRF(P.P.M) present, and in the second sample both glauco- Element Per- Ele- Per- nite and Celadonite were present (Figs.11, 12). centage ment centage O 44 Cu 13194 3.2.4 Blue paint layer Si 13.8 Fe 4916 An investigation of the blue pigment in thin Ca 30 Pb 1030 section showed the presence of anisotropic Cu 6.4 Zn 891 crystals with tubular shape, and refractive in- Fe 1.5 Sn 281 dex less than 1.66.compaed to the canada bal- S 0.7 Sb 124 sam medium used in the sample preparation. Al 1.44 these properties are typical of cuprorivaite, the K 0.7 crystal shape was also observed under sem in- vestigation (fig no.13) i.e. egyptian blue. ele- mental analysis using xrf analysis indicated the 3.2.5 BROWNISH RED presence of copper as a main component; iron The elemental analysis using portable X-ray also being present. the analysis also indicated fluorescence for the brownish red sample indi- the presence of lead (fig no.14). the presence of cates the presences of Arsenic (As 26772 ppm) lead and traces of other elements as tin and and minor amounts of iron (Fe 3974 ppm) and zinc is of paramount importance as this may calcium (Ca 4251ppm). The XRF result may in- indicate the reuse of bronze alloy as a source of dicate the use of arsenic sulphide, or orpiment copper for the fabrication of the egyptian blue. (fig no.15). The xrd pattern of the brownish red pigment indicates orpiment, calcite and hema- tite in the composition of the paint layer (fig no.16). This might explain the dark Chroma of this color, being due to the mixture of orpiment and hematite. Figure 13 SEM image of the blue paint layer revealed the presence of coarse grains and tubular shaped crystals . Figure 15 XRF result of the brownish red paint layer in- dicating the presence of As and traces of Fe. Figure 14 XRF result for the blue paint layer indicating Cu as a major elemental component of the blue pigment Figure 16 XRD pattern for the brownish red paint layer indicating the presence of orpiment, hematite and cal- cite. Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256  251 IDENTIFICATION OF THE BYZANTINE ENCAUSTIC MURAL PAINTING IN EGYPT 3.3 Organic binding medium ester (Fig.18). The band at 1512 cm-1 could be The identification of the organic binding me- assigned to carboxylate salts or fatty acid salts dium was based on the vibrational signature of (Mirghani, Che Man, Jinap, Baharin, & Bakar, the functional groups for beeswax, using both 2002), while alcohols bands are present in the Fourier transform infrared analysis, and gas OH stretching at 3407 cm-1 and the alcohols OH chromatography mass spectrometry analysis. bending at 1122 cm-1 and 1056 cm-1 . Further information obtained using these two The ftir spectrum of the green sample repre- methods enabled us to understand the original sents a wax based paint layer, in which a partial technique used for the execution of the encaus- saponification process in indicated by the pres- tic technique and the alteration of the material ence of the carboxylate and alcohol functional over time. groups. This could be due to natural aging of wax, as the Punic process of saponification will 3.3.1 FTIR analysis result in the complete saponification of wax. Two samples were analysed by means of The small peak at 958 may indicate the presence FTIR analysis, from the red and green paint of Si-O, which could be from the green earth layers. The results were compared to two con- pigment. trol samples of bees wax - molten and Punic The FTIR analysis of a sample from a red beeswax (fig no.17) were prepared- in order to flaking paint layer proved the presence of wax compare the results with the archeological as a binding medium; this is demonstrated by samples in order to identify both the presence the presence of the methylene group of the wax of the wax and the technique used. hydrocarbon, as well as the carbonyl group of the wax ester. A significant broad and large band could be seen in the region of 3450 cm-1 due to the OH group, probably indicating the presence of tri- glyceride alcohols. A small band at 1107 and 1028 cm-1 for the long chain alcohols indicates that small quantities of alcohol could be in the sample. Crystalized areas of wax represented by the small band for C=O at 1735 cm-1 and another one for the C –O at 1170 cm-1, indicates that cer- Figure 17 FTIR pattern for two control samples of molten tain alterations could have occurred in the and Punic ( saponified) wax sample composition. Inorganic materials were indicated by the The main functional groups obtained from presence of a secondary absorption band of the the contemporary beeswax sample were two CO3 -2 group of calcium carbonate in the region main doublets representing stretching in plane of 2516 cm-1, 1793 cm-1. A broad band at 1458 and out of plane in the region of 2950 and 2850 cm-1 overlapped the CH band in the same re- cm-1, as well two other sharp doublets at 1460 gion and bending of the CO3-2 is seen at 874 cm- and 1470 cm-1. The peak in the region of 715 – 1 (Derrick, stulk, & Lundry, 1999).Small peaks in 720 cm-1 indicates the CH2 torsion bend; the C- the region of 1107 and 1028 cm-1 indicated the O of the wax ester usually appears in the 1165 – presence of long chain alcohols (Mirghani, Che 1170 cm-1region; C=O of the organic fatty acids Man, Jinap, Baharin, & Bakar, 2002), and could appears in the 1715 -1720 cm-1 region. be connected to the broad band of the OH at The FTIR results of the paint layers showed 3454 cm-1. two strong and sharp methylene stretching symmetrical and asymmetrical bands at 2919 and 2849 cm-1 respectively, as well as two bands at 721 cm-1 , 874 cm-1 and 922 cm-1, a strong C=O at 1735 cm-1 and C-O at 1169 cm-1 of the Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256 252 B.GEHAD et al. Figure 18 FTIR pattern of two samples from green and red paint layer, Baouit encaustic mural painting. The green sample was compared to both a tation and the separation of the main organic molten bees wax sample and a Punic bees wax components of the wax identified as seen in sample. The comparison of the ftir charts (Fig.20) mainly hydrocarbons, esters and fatty (Fig.19) proved that the archeological samples acids, clearly represent the wax alteration. The were similar to a large extent to the molten con- alteration is shown by: trolled sample. Some alterations appear in the  Presence of long chain alcohols. archeological paint layer profile, but it does not  Depletion of hydrocarbon profile of wax. resemble totally and intentionally saponified  Presence of phenolic compounds. wax (Punic sample), which means that the pres-  Alteration of esters and appearance of gly- ence of small bands related to long chain alco- ceride esters and fatty acids. hols or carboxylates, could be interpreted as a result of natural aging and alteration of wax due to oxidation and water hydrolysis in an ar- id and exposed environment. Figure 19 comparison between FTIR pattern of the green archaeological paint layer with the two control samples Figure 20 GC-MS result for the beeswax contemporary of the molten and Punic wax. sample and the archaeological green sample. 3.3 GC-MS analysis The four previous aspects are shown in the fol- The GC-MS analysis for one of the wax based lowing Table 3, taking into consideration the paint layers indicates the state of alteration that relative intensity of the separated compounds: the binding medium had reached. The fragmen- Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256  253 IDENTIFICATION OF THE BYZANTINE ENCAUSTIC MURAL PAINTING IN EGYPT Table 3 GC-MS analysis results including Retention time and compounds identified from both control beeswax sample and the archaeological green sample. No Rt Compound Chemical formula Area Comments % 1 5.20 3Oxo20methyl C22H31NO2 0.42 Steroid compound result from the hydroxyconanine1,4diene reaction of lipid with heavy metal as lead 2 7.99 Octadecanal C18H34O 0.17 Long chain alcohol 3 17.29 2-4Chlorophenyl C17H10ClN3O2S 0.52 1. Phenolic compounds thiazolylamino,benzoxazine 4 23.16 Hexadecanoic acid C16H32O2 11.48 2. PALMITIC ACID 5 24.24 2,5Dimethylhexane C8H18O4 2.78 Peroxides 2,5dihydroperoxide 6 27.83 Dimethyl glycerol ether C27H56O5 0.20 Ether accompanied with carboxylic acid 8 28.08 Heneicosane C21H44 1.3 hydrocarbons 9 28.68 PHENOL, C15H24O 0.24 Phenol based compund 2,6BIS (1,1DIMETHYLETHYL) 4METHYL 10 33.31 Tricosane C21H44 7.8 Hydrocarbons 11 33.44 Tricosane C21H44 9.6 Hydrocarbons 12 35.14 Tetracosane C24H50 3.2 Hydrocarbons 13 35.94 Ethyl iso allcohlate C26H44O5 0.21 14 36.18 Nonacosanal C29H60O 0.7 Fatty Alchol 15 37.4 Heptacosane C27H70 6.9 Hydrocarbons 16 37.6 Hexacosane C26H54 3.4 Hydrocarbons 17 38.7 Heptacosane C27H70 29.3 Hydrocarbons 18 39.5 Octacosane C28H58 2.28 Hydrocarbons 19 40.43 Nonacosane C29H60 14.2 Hydrocarbons 20 41.23 Dimethyl Glycerol Ether C27H56O 6.28 Ether accompanied with carboxylic acid 21 41.20 Dimethyl Glycerol Ether C27H56O 1.16 Ether accompanied with carboxylic acid 22 41.77 Dimethyl Glycerol Ether C27H56O 5.8 Ether accompanied with carboxylic acid 23 43.2 Octadecanoic acid hydroxyl C20H40O 2.38 caprylic acid ester 24 47.8 Dimethyl Glycerol Ether C27H56O 5.8 Ether accompanied with carboxylic acid 25 53.15 Ethyl Iso allcholate C26H44O5 0.2 Alcohols 26 55 Dotriacontane C32H66 0.8 Hydrocarbons 27 62.3 TetrteraAcontane C44H90 5.44 Hydrocarbons 30 25 20 15 10 5 CW AW 0 Octadecanoic acid hydroxyl … Chlorophenyl Octadecanal Phenol Dimethyl glycerol ether Hexacosane Octacosane Heneicosane Lucenin 2 Hexadecanoic acid Dotriacontane Tricosane TetrteraAcontane Tetracosane Heptacosane Nonacosanal Dimethyl Glycerol Ether Nonacosane Ethyl iso allcohlate peroxide Figure 21 Comparison between the main component of contemporary beeswax and the archeological sample from Baouit Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256 254 B.GEHAD et al. 4. DISCUSSION light richness of the byzantine palette used for the encaustic painting in Egypt, during the 6th The analysis performed on the samples from century. the unique encaustic painting from the monas- The FTIR analysis in the mid-range (400-4000) tery of Baouit explains the formation of the for five archaeological samples, proved the painting palette. The pigments identified were presence of methylene and ester functional mainly types identified previously from the groups identical to those for the wax ftir profile. roman period, only the brownish red is an ex- This result confirms the usage of wax and more ception. The palette shows the variety of likely beeswax as a binding medium in the mu- sources where these pigments were obtained; ral painting of baouit. both natural and artificial sources were used. The samples indicate evidence of hydrolysis The integrated elemental analysis using EDX or partial saponification of the organic medium, and XRF confirms the usage of a green earth the presence of OH vibration bands in the 3400 based pigment in combination with a copper region as well as 1035 to 1060 cm-1 region pigment, the XRF shows major amounts of Fe indicates the presence of long chain alcohols or and Cu as well as minor amounts of Sb, Sn, Pb glycerides. One should also mention the ob- and As, while the EDX gives major Ca peaks served depletion of the methylene or the with minor percentage of Al, Fe, Cu, as well hydrocarbon from the weakness of the peak of traces of Mn, K, Cl and S. The results match the characteristic group which may be due to those of green samples from different periods the sublimation of the lower molecular (pharonic and roman period green samples) hydrocarbons. The same is true for those peaks from an ancient shrine, analysed by Berry (Ber- representative of the ester functional group, as ry, 2002) as well mineralogical identification they were shifted and weakened, and and description for green earth (both glauconite sometimes the intensity of the carbonyl peak and celadonite) in the literature (Buckley, Bevan, increased, which indicates the occurrence of Brown, & Johnson, 1978) (Bearat, 1996). hydrolysis and formation of carboxylate, for The presence of both copper based pigment which some small representative peaks were and green earth in the XRD pattern, explains the observed in the archaeological sample charts. result of the investigation of the thin sectioned It’s hard to say whether the saponification sample for the green paint layer , with large identified on the ftir spectrum for the green particles embedded in very fine green archaeological samples was intentionally made matrix. by the artist in order to prepare the water The EDX and the XRF elemental analysis of soluble encaustic, or that the presence of the the red pigment indicates lead as a main ele- saponifaction was due to the water hydrolysis ment, with traces of iron and calcium, which as well as other alteration processes, which could be traces from the white wash. The min- could have taken place in the wax composition eralogical composition of the pigment was iden- during rancidification . tified using XRD qualitative analysis as Minium It’s well known that rancidifcation is the de- (Pb3O4), as well as calcite indicating the lime caying of unsaturated fatty acids by an oxida- based render. tion process; this process could be promoted by The absence of wallastonite in the qualitative the presence of trace metals as Ca, Cu, Fe, and XRD result of the blue pigment indicates the Zn (Megahed, Nashy , & Ashkar, 2011), which temperature used in the manufacture of the already exist in the paint layer from the pig- Egyptian blue was lower than 950 oC (Riederer, ments, as well as the presence of heat. These 1997). As no green particles could be observed two factors existed within the context of the in the stereo microscopic investigations, the mural painting at baouit. theory is supported by well identified sharp The Chromatogram of the archaeological wax edge crystals observed under the SEM-BSED based paint layer shows the highly deteriorated investigations. pattern of the wax medium. This was demon- The identification of pigments like natrojaro- strated by the presence of long chain alcohols, site and mixture of orpiment and hematite high- palimtates and mainly the glycerols of the Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256  255 IDENTIFICATION OF THE BYZANTINE ENCAUSTIC MURAL PAINTING IN EGYPT hydrolysed triglycerides represented by the that the honey and its wax were naturally ester of fatty acids. formed by bees which fed on sunflowers. Oxidation or photooxidation may be one of the main reasons for the alteration of the wax; 5. CONCLUSION other components found in the archaeological sample such as phenolic compounds may The painting palette of the encaustic mural indicate that this process took place. painting at Baouit was identified as: lead red The profile of the hydrocarbon was also (minium), Egyptian Blue, natroJarosite, a mix- strongly altered. The main hydrocarbon ture for the brownish red (red ochre and orpi- heptacosane concentration was 29 % in the ment), and green (glauconite and atacamite contomperoray wax sample but only 6.9 % in with traces of Celadonite). The mixtures high- the archaeological wax. This indicates that the light how the ancient artists adapted their mate- hydrocarbons which represent the main rials to obtain the colors. backbone of the wax were totally depleted and The identification of beeswax proves the exist- changed as a result of aging starting with the ence of the encaustic technique as an ongoing lower molecular weight hydrocarbons. tradition for painting in Egypt during the By- One of the interesting pieces of information zantium period. The analysis shows also that that came from the GC-MS analysis, is the the wax was probably not treated with alkali in presence of lucenin, which is a luteolin flavone order to prepare the Punic; the presence of long glycoside found in the sunflower family (Bohm chain alcohols could be a result of alteration & Stussey, 2002) with a molecular composition through natural water hydrolysis and oxidation (Luteolin 6, 8-di-glucoside), which informs us as a part of the aging process. ACKNOWLEDGEMENTS We thank the anonymous reviewers for constructive comments. I would like also thank Dr.Giselle Hadje, director of the Baouit excavation mission, Dr.Anita Quiles from IFAO, Dr Mo- hamed Abd el Rohman, from SCA. Kate Fulcher from UCL Institute of Archaeology - British Mu- seum- department of Conservation Science and Research. REFERENCES Bearat,H. (1996) Chemical and mineralogical analyses of * gallo-roman wall painting from dietik- on, switzerland, Archaeometry 38, 1 81-95. Bénazeth,D., Herbich,T. (2008) le kôm de baouît : étapes d’une cartographie, BIFAO 108. Benazeth,D. (2010) Nouvelles Campagnes de fouilles a Baouit(2005, 2006). Etudes Coptes XI,Paris, pp17-25. Berry, M. (2002) The Study of Pigments from Shrine I at Ismant el-Kharab, in C. A. Hope and G. E. Bowen (eds), Dakhleh Oasis Project: Preliminary Reports on the 1994–1995 to 1998– 1999 Field Seasons, 53–60. Binda, L., Saisi, A., Tiraboschi, C., Valle, S., Colla, C. and Forde, M. C. (2003) Application of sonic and radar tests on the piers and walls of the Cathedral of Noto. Construction and Building Materials, vol. 17, 613–627. Bohm, B. A., & Stussey, T. F. (2002). Flavonoids of sun flower family. Biologia plantarum, 45(2), 226. Bonaduce, I. and Colombini, M. P. (2004). Charchterisation of beeswax in works of art by gas chrimatography - mass spectrometery and pyrolysis - gas chromatography- mass spectrometery procedures. Journal of chromatography, 1028, 297-306. Buckley, H. A, Bevan, J. C., Brown, K. M., and. Johnson, L. R ( 1978) Glauconite and celadonite" two separatemineral species Mineralogical Magazine, September, vol. 42, pp. 373-82. Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256 256 B.GEHAD et al. Cartwright, C., Middleton, A. (2008) Scientific aspects of ancient faces: mummy portraits from egypt , in the british museum technical research bulletin, vol2. Cosentino, P. and Martorana, R. (2001) The resistivity grid applied to wall structures: first results. Proceedings of the 7th Meeting of the Environmental and Engineering Geophysical Society, Euro- pean Section, Birmingham, U.K. Derrick, M., stulk, D., & Lundry, J. (1999). Infra red spectroscopy for conservation science. los angeles: Getty conservation institute. Doxiadis, E. (1995). The Mysterious Fayoum Portraits, faces from ancient egypt. Cairo : the American University in cairo press. Innemee , K. (1999). Encuastic painting in egypt, painting in Dier el surian. 2, 131-139. Kühn, H. (1960). Detection and Identification of Waxes, including Punic Wax, by Infra-Red . Meilunas, R., Bentsen , J., & Steinberg, A. (1990). Analysis of Aged Paint Binders by FTIR Spectroscopy . Studeis in Conservation , 33-51. Mirghani, M., Che Man, Y., Jinap, S., Baharin, B., & Bakar, J. (2002). FTIR Spectroscopic Determination of Soap in Refined Vegetable Oils. JAOCS, 79(2), 113. Moretto, L. M., Orsega, E. F., & Mazzocchin, G. A. (2011). Spectroscopic methods for the analysis of celadonite and glauconitein Roman. Journal ofCulturalHeritage, 3-11. Moretto, L. m., Orsega, E. F., & Mazzocchin, G. A. (2011). Spectroscopic methods fro the analysis of celadonite and glauconite in roman wall paintings. Journal of cultural Heritage .temperature gas chromatography and chemometric analysis. food chemistery, 136, 961 - 968. Megahed, M., Nashy , H., & Ashkar, E. (2011). Evaluation of fried edible oil and determenation of elemnts content. Agriculture and biology journal of north america, 687-692. Lieber, F. (1840). Encyclopædia Americana: a popular dictionary of arts and sciences (Vol. IV). Philadelphia. Maia, M., & Nunes, F. M. (2013). Authentication of beeswax ( apis mellifera )by high Petrie, W. (1911). Roman portrait and Memphis (IV). London: school of Archeology in Egypt. Pitthard, V., Vak, B., Griesser, M., Stanek, S., & Laubenberger, M. (2007). Fayoum Portraits from the collection of greek and roman antiquites, Kunsthistorisches museum, Vienna. Technologishe Studien, 4, 11-29. Rackman , H. (1952). Pliny the elder natural history, chapter XXXV. Cambridge . Ramer, B. (1979). The Technology, Examination and Conservation of the Fayoum portraits in the Petrie Museum. Studies in Conservation, 1-13. Spectrography. Studies in Conservation, 5, 71- 81. Scholl, Reinhold. (2002) Der Papyrus Ebers: die grösste Buchrolle zur Heilkunde Altägyptens. Walker, A. (2012) The Emperor and the World: Exotic Elements and the Imaging of Middle Byzantine Im- perial Power, Ninth to Thirteenth Centuries C.E. New York, Cambridge University Press. White, R. (1978). The application of gas chromatography to the identifcation of waxes. Studies in conservation, 23(2), 57. Mediterranean Archaeology and Archaeometry, Vol. 15, No 2, (2015), pp. 243-256