Tutankhamun’s Polychrome Wooden Shawabtis: Preliminary Investigation for Pigments and Gilding Characterization and Indirect Dating of Previous Restorations by the Combined Use of Imaging and Spectroscopic Techniques

2.2.1 Multispectral Imaging (MSI)

Technical images were captured with a Nikon D90 DSLR (CMOS sensor) digital camera that had been modified for “full spectrum (c. 350–1,000 nm)” by removing the inbuilt UV-infrared reflected (IR) blocking filter and fitted with a Tokina Macro 35 mm F/2.8D DX lens that was used at its widest F-stop setting (2.8) to capture all luminescence images with shutter speeds ranging from 0.8 to 8 s. The camera was operated fully manually and tethered to a computer to allow sharp focusing in nonvisible modes (IR and UV) using live view mode with the picture style set to “Neutral” and an ISO speed of 100. Except for the acquisition of UV-induced visible luminescence images, where it was set to a color temperature of 6,500 K, white balance was set to “Custom.” To calibrate the camera, an X-Rite ColorChecker Passport and a reference gray scale were used (Abdrabou, Hussein, Sultan, & Kamal, 2022; Cosentino, 2015; Dyer, Verri, & Cupitt, 2013).

In each case, the shawabti was illuminated by two radiation sources that were symmetrically positioned at approximately 45˚ with respect to the camera’s focal axis and at approximately the same height. To select the wavelength range of interest, a filter or combination of filters was placed in front of the camera lens. Table 2 summarizes the filters and radiation sources used for imaging techniques. All images were captured as RAW files and converted to TIF (tagged image file) format and a set of recommended presets which turn-off all enhancements (e.g., recovery, fill light, blacks, contrast, brightness, clarity, saturation, as well as setting the tone-curve to linear) were applied using Adobe Photoshop. Post-processing procedures for the calibration of the visible (VIS), ultraviolet induced visible luminescence (UVL), ultraviolet reflected (UVR), visible-induced luminescence (VIL), and IR images as well as the creation of infrared false color (IRFC) images are then carried out using the “British Museum (BM) workspace” and nip2 software (Martinez & Cupitt, 2005) as described by Dyer et al. (2013).

2.2.2 Optical Microscopy (OM)

For detailed observations of the polychrome layers, we used a Zeiss Stereo DV 20 (portable stereomicroscope) equipped with an Axio Cam MRC5 with an optical zoom of 28 up to 560×.

2.2.3 XRF

The painted preparation layers were nondestructively analyzed by a handheld XRF (HH-XRF) spectrophotometer using a Thermo Scientific Niton XL3t with a “GOLDD” detector. It was placed in contact with the selected area, and the X-ray spot size was 3 mm in diameter. The X-ray tube has a Ag anode of 50 kV and 200 μA. The standardization modes selected are the “Cu/Zn Mining”, which includes the elements of interest for the analysis of painted preparation layers (e.g., Ca, S, Si, Cu, Fe, As, Pb, Sn, Cl, and P) and the “precious metal” mode, which includes the elements of interest for the analysis of gilding (e.g., Au, Ag, and Cu). This analysis uses four separate filters to determine the concentrations in percentage of elements: a high filter (30 s counting time), a main filter (30 s), a low filter (30 s), and a light filter (30 s), leading to a total measurement time of 120 s per analysis. The software utilizes the Fundamental Parameters algorithm to determine the concentrations of each element. The spectra obtained from the XL3t were downloaded to a computer for analysis by the Thermo Scientific NDT program. The total points considered for HH-XRF analyses were chosen for different colors and conservation situations following suggestions from the results of imaging analyses.

2.2.4 Vis-RS

A Konica Minolta Spectrophotometer CM-2600d was used to record the reflectance spectra of the painted layers under standard illuminant d65 (the di:8°/de:8° geometry (diffused illumination, 8° viewing angle), including the specular component (measuring area: circular area with a diameter of 5 mm). Spectra are collected along a 400–700 nm wavelength range every 10 nm. The light diffused in the integrating sphere is received by the illumination-monitoring optical system and guided to a sensor. The light reflected from the painted surfaces and the diffused light are divided into each wavelength component by a measuring optical system and illumination-monitoring optical sensor, respectively, and the signals proportional to the light intensity of each component are output to the analog processing circuit.

2.2.5 XRD

PANalytical pro model PW3040 XRD with a Cu anode at 30 mA/40 kV was directly performed on the shawabtis. Each shawabti was placed inside an XRD apparatus on a sample holder, and an approximately flat surface of each color was exposed to the X-ray beam. Diffraction patterns were interpreted using X’Pert HighScore software.

2.2.6 FTIR

FTIR measurements were conducted using an FTIR spectrometer (IRPrestige-21, Shimadzu) in the 400–4,000 cm⁻¹ range, with a resolution of 8 cm⁻¹. Samples of the previous restoration materials were identified by comparing the obtained spectra with literature data (Derrik, Stulik, & Landy, 1999) and standards created in the FTIR laboratory at the GEM CC.

3.1.1 Blue and Greenish Hue Painted Layers

In the VIL images (Figures 2b and 3b), the blue and greenish blue painted layers appeared as bright white, while all other materials appeared dark. The luminescence of such areas indicates the presence of Egyptian blue (Accorsi et al., 2009; Amenta, 2014; Abdallah & Abdrabou 2018). According to XRF spectra of the blue painted layer (Figure 2c), the elements that presented the highest concentrations were silicon (Si), copper (Cu), calcium (Ca), and sulfur (S). This result provides strong evidence for the presence of Egyptian blue pigments, confirming the corresponding visible reflectance spectrum (Figure 2d), which matched published spectra of Egyptian blue with an inflection point at about 630 nm (Abdrabou, et al., 2022; Aceto et al., 2014; Edreira, Feliu, Fernández-Lorenzo, & Martín, 2001). The presence of S in the blue regions would suggest the presence of gypsum, whereas the presence of tin (Sn) impurities in the XRF analysis (Table 1) allowed us to assume that a bronze scrap was used to produce the Egyptian blue pigment (Abdrabou et al., 2018; Jaksch, Seipel, Weiner, & El Goresy, 1983). Applying XRD (Figure 4a) confirmed the presence of gypsum and cuprorivaite (the main component of Egyptian blue) for the blue painted layers in accordance with the attribution made after the VIL and XRF spectra. The Egyptian blue pigment was the first synthetic pigment ever produced by man and dates to the Protodynastic period in Egypt (around 3200–3000 BC). It was found on a Protodynastic period bowl with markings attributed to the Scorpion King and was extensively used from the 4th Dynasty in Egypt until the end of the Roman period (Abdrabou, Abdallah, & Kamal, 2017; Corcoran, 2016; Ganio et al., 2015; Mirti et al., 1995). Egyptian blue is a multicomponent pigment made from locally derived component minerals of copper, calcium, silica (in the form of quartz, SiO₂), and about 1% flux, usually comprising lime (CaO) from calcium carbonate to form calcium copper tetrasilicate crystals (cuprorivaite: CaCuSi₄O₁₀), the chromophore in the Egyptian blue. This material can be characterized by its blue tabular crystals, which have been fritted in an oxidizing atmosphere at 850–1,100°C (Hatton, Shortland, & Tite, 2008; Ismail et al., 2016; Lee, 1997; Pradell, Salvado, Hatton, & Tite, 2006).

Figure 2

Characterization of the blue painted layer (Object GEM 3163, Spot 1): (a) visible image; (b) VIL image; (c) XRF spectrum; and (d) visible reflectance spectrum.

Figure 3

Characterization of the greenish blue painted layer (Object GEM 3173, Spot 9): (a) visible image; (a) VIL image; (c) XRF spectrum; and (d) visible reflectance spectrum.

Figure 4

XRD patterns of: (a) the blue painted layer (Object GEM 3163, Spot 1) and (b) the greenish blue painted layer (Object GEM 3173, Spot 9).

In general, Egyptian blue is a very stable pigment, and there are examples to be seen in Egypt that have been exposed for thousands of years without the loss of color, as seen in Figure 1. However, some objects painted with the Egyptian blue pigment show some discoloration that varies considerably, from a relatively greenish hue to black (Daniels, Stacey, & Middleton, 2004). In the studied shawabtis GEM no. 3429, 3173, and 3503, the microscopic investigation reveals that the greenish hue is a surface phenomenon, with what appears to be fresh, bright blue pigment lying beneath the greenish layer. Moreover, the microscopic investigation did not show any yellow particles in the greenish hue areas, so the hypothesis of a mixture of Egyptian blue and yellow pigments can be discarded. It is possible that the greenish hue is due to the degradation of the Egyptian blue pigment to atacamite or its polymorphs through the process known as “copper chloride cancer” (Abadir, 2014) or due to the darkening of organic materials used as binders or previous consolidation materials (Daniels et al., 2004). According to the XRF results of the greenish hue areas (Figure 3c), the elements that presented the highest intensities were Si, Cu, and Ca, with a significant amount of chlorine (Cl). This result provides strong evidence for the presence of the Egyptian blue pigment, confirming the corresponding visible reflectance spectrum (Figure 3d), which matched published spectra of Egyptian blue, with shifting in the reflectance ranging between 400 and 600 nm in the greenish blue painted layer due to its discoloration. The presence of Cl in the greenish hue areas would probably point to the presence of halite or a chlorine-containing solid phase and, considering the chemical reactivity of cuprorivaite, this solid phase could be atacamite, Cu₂(OH)₃Cl, or one of its polymorphs, paratacamite or clinoatacamite, which are products of the degradation of Egyptian blue when in contact with high-chloride concentration solutions. By applying XRD to the greenish hue areas (Figure 4b), apart from calcite and quartz, cuprorivaite (the main component of Egyptian blue) was detected. The absence of atacamite or its polymorphs in the diffractogram could be due to a lower percentage of atacamite or that the new phase formed on the surface of the blue pigment is amorphous. Consequently, the XRD technique is not able to detect it. Therefore, further analysis using Raman spectroscopy and gas chromatography/mass spectrometry will be necessary to determine its identity more precisely. This phenomenon (the coexistence of Egyptian blue and atacamite on blue-green colors) has already been reported by many authors (Giménez, 2015; Green, 2001; Lee, 1997; Riederer, 1977).

3.1.2 Yellow Painted Layer

The yellow fluorescence of yellow areas of shawabti no. 3528 under UV (Figure 5b) suggests that a yellow painted layer was probably made of arsenic (As)-based pigments such as orpiment since it had fluorescence properties (Abdrabou et al., 2018; Stuart, 2007). In addition, the microscopic examination revealed large yellow particles in a granular form, as well as the presence of some single particles of the blue pigment in the yellow painted layer (Figure 5d). These single particles appeared as bright white in the VIL image (Figure 5c), which refers to the Egyptian blue pigment. Moreover, comparing the VIS and VIL images (Figure 5a and c) allowed for the surviving Egyptian blue pigment to be mapped, and it is clear that its presence is much more extensive than what was discernible with the naked eye. In Figure 5e, the XRF spectrum of the yellow painted layer is shown. It shows a high intensity of As, Ca, Cu, S, Si, and Fe. It is interesting to note that Cu, Ca, and Si are detected in the yellow areas, clearly related to the single particles of Egyptian blue in accordance with the attribution obtained by VIL imaging. The presence of these blue single particles in the yellow paint layer could be accidental impurities due to the use of a dirty brush or could be added to achieve a higher intensity of the yellow sensation. The high intensity of As and S provides strong evidence for the presence of an arsenic sulfide pigment that is most likely orpiment (As₂S₃) in a nearly pure form, as confirmed by Vis-RS. In Figure 5f, the inflection point at 480 nm and the slope of the reflectance curve of the yellow layer fit with published reference spectra for orpiment is shown (Aceto et al., 2014; Cavaleri, Giovagnoli, & Nervo, 2013). Orpiment is found locally in Egypt associated with gold and silver ores (in Edfu and the Eastern Desert) or in copper ores (Saini mines), in hot-spring deposits, and volcanic sublimation, but others have reported that orpiment was imported from Syria or Asia Minor (Abdrabou et al., 2017; Bonizzoni et al., 2011; Bonizzoni et al., 2018; Bracci et al., 2015; Davies, 1995; Gard et al., 2020; Sakr, Ghaly, Abdulla, Edwards, & Elbashar, 2020; Scott et al., 2004; Scott, Warmlander, Mazurek, & Quirke, 2009). Historically, its earliest findings are dated to the 2nd Dynasty (Colinart, 2001) and several later findings in Egyptian artifacts have been reported (Abdrabou et al., 2017; Abdrabou, et al., 2022; Ambers, 2004; Amenta, 2014; Bonizzoni et al., 2011; Brunel-Duverger et al., 2019; Colinart, 2001) particularly during the 18th Dynasty, when it was extensively used to give a brilliant yellow, gold-like color with a shiny glaze, so it has been described as royal yellow (Sakr et al., 2020).

Figure 5

Analytical characterization of the yellow painted layer (Object GEM 3528, Spot 21): (a) visible image; (b) UV luminescence image; (c) VIL image; (d) an optical micrograph; (e) XRF spectrum; and (f) visible reflectance spectrum.

3.1.3 Red Painted Layers

All the red painted layers appeared darker in the UV-induced luminescence (UVL) images, which may suggest that the red pigment is red ochre, consistent with the strong quenching properties of the iron-based pigment (McCarthy, 2001). The microscopic examination showed the presence of some single particles of yellow pigment in some red paint areas (Figure 6a), which may be due to the use of dirty brushes or could be added to achieve a higher intensity of the red sensation. The results of XRF analysis for some points on the red painted layers (Figure 6b) showed the presence of Fe, Ca, As, Si, and S with a high intensity, in addition to a small amount of aluminum (Al). The presence of As and S is clearly related to the single grains of orpiment that are shown by the microscopic examination. The presence of Fe, Ca, Si, and Al provides strong evidence for the presence of red ochre, as confirmed by Vis-RS (Figure 5c), which displayed a sigmoid shape with an absorption band between 400 and 500 nm, a sharp positive slope between 550 and 600 nm, and an inflection point at 580 nm, which are indicative of red ochre (Abdrabou et al., 2022; Aceto et al., 2014; Cavaleri et al., 2013; Edreira et al., 2001; Guglielmi, Andreoli, Comite, Baroni, & Fermo, 2021; Miriello et al., 2018). Finally, applying XRD to the red painted layers showed that, apart from quartz, the presence of hematite is detected, in accordance with the attribution made by Vis-RS and XRF. The red areas are all colored with hematite (α Fe₂O₃), the main chromophore found in red ochres, and they contain a small proportion of other crystals, including quartz, and are better described as ochres rather than as pure hematite. This is in keeping with the other evidence from other Egyptian contexts, where red ochres are by far the most commonly reported red pigment. Red ochre was used to create a red color on wooden artifacts in ancient Egypt, according to many works (Bonizzoni et al., 2018; Lee & Quirke, 2000; Pagès-Camagna & Guichard, 2010).

Figure 6

Characterization of the red painted layer (Object GEM 3511, Spot 18): (a) An optical micrograph; (b) XRF spectrum; and (c) visible reflectance spectrum.

3.1.4 White Painted Layers

The microscopic examination revealed the presence of some single particles of the blue and yellow pigments in the white painted layer (Figure 7a). The blue single particles appeared as bright white in the VIL image, which refers to the Egyptian blue pigment. The results of XRF analysis for some points on the white painted layers (Figure 7b) showed the presence of Ca and S with a high intensity, in addition to a small amount of Si, As, Fe, and Cu. The presence of Ca and S provides strong evidence for the presence of calcium-based pigments such as calcium sulfate. The presence of As and S is clearly related to the single grains of orpiment that are shown by the microscopic examination, while silicon and copper are detected in the white areas, clearly related to the single particles of Egyptian blue that are shown by VIL imaging. At some times, the ancient Egyptians added Egyptian blue to the white colors to achieve a higher intensity of the white sensation (Edreira, Feliu, Fernández-Lorenzo, & Martín, 2003). Finally, applying XRD to the white painted layers (Figure 7c) showed that, apart from quartz and weddellite (calcium oxalate), the presence of gypsum is detected, in accordance with the attribution made by XRF. Calcium-based pigments such as calcite, gypsum, and huntite were the most commonly used white pigments in Egyptian painting and have been reported in many works to create a white color on wooden artifacts (Ambers, 2004; Bonizzoni et al., 2011; Heywood, 2001).

Figure 7

Characterization of the white painted layer (Object GEM 3503): (a) An optical micrograph; (b) XRF spectrum; and (c) XRD pattern.

3.1.5 Black Painted Layers

IR images strongly suggest the presence of carbon-based black in the black paint areas, as carbon is opaque under infrared emission and appears dark (Abdrabou et al., 2017). The microscopic investigation of the black pigment indicated the fineness and evenness of particles and did not show any fibrous structure, so it is possible to exclude the burnt vegetable origin of the black pigment (Abdrabou et al., 2018). In addition, some shawabtis (GEM 3503–3511) showed some single particles of orpiment pigments in the black painted layer, which may be due to the use of dirty brushes. In XRF analysis, phosphorus was detected in the black painted layers of shawabtis GEM no. 3503 and 3511. This result provides evidence for the presence of carbon obtained from animal origins that is most likely bone black, which is one of the oldest pigments known to humans and was originally made by charring animal bones (Abdrabou et al., 2018; Mahmoud, 2014), while phosphorus was not detected in the XRF results from the black painted layers of the other shawabtis, making it most likely that they are of lamp black or soot, produced by the combustion of vegetable matter or oil (Ambers, 2004; Lee & Quirke, 2000).

3.1.6 Gilded Layers

The majority of the gold used in ancient Egypt was obtained from alluvial deposits and from quartz rock found between the Nile and the Red Sea (Colinart, 2001). Their composition ranged from very pure to those containing at least 40% of silver (electrum) and copper, with percentages not exceeding 1.5%. The majority of Tutankhamun’s golden objects are gilded, either with gold foil or gold leaf applied over a firm support to achieve the rich appearance of solid gold (Abdallah, Abdrabou, & Kamal, 2018; Abdrabou et al., 2018; James, 1972). There are some gilded wooden shawabtis among them, with varying degrees of color ranging from bright yellow to reddish brown. According to the semiquantitative XRF results, the gilded layers with a bright yellow color have a similar composition and are composed of pure gold (Au) ranging from 99.2 to 98.5%, with less than 1% of Ag and Cu. The gilded layers with a reddish brown color exhibit a high Ag content that reaches 9%. The XRF spectra obtained for the yellow and dark colors were compared. XRF spectra (Figure 8) showed an increase in the Ag intensity in the dark areas of the gilded layers, with no significant change in the S and Cu intensities. The increase of Ag on the dark color of the gilded layers with the presence of S suggests the formation of distinct silver sulfide complexes, namely, AgAuS and Ag₃AuS₂, which correspond with previously published data for some dark gold Egyptian leafs/foils (Abdrabou, El Hadidi, Hamed, & Abdallah, 2018; Frantz & Schorsch, 1990; Hatchfield & Newman, 1991; Rifai & El Hadidi, 2010; Tissot, et al., 2015).

Figure 8

Comparing the obtained XRF spectra for yellow and dark areas of gilding.