Introduction

Teotihuacan, located in central Mexico, was from ca. 250 to 650 CE a state capital, a multiethnic city, an economic center, and a place with a strong manufacturing activity. The city stands out from its contemporaries for being the first complex state developed in the Mexican plateau. By ca. 450 CE, the city covered about 20 km2 with an estimated population around 125,000 people, organized in multi-family housing compoundsFootnote 1 (Cowgill 2015; Froese et al. 2014; Manzanilla 2006, 2008, 2014, 2017).

The socio-economic strategy developed in Teotihuacan had a corporative character. When this is added to the size of the city, its urban planning, the articulation of a lax and discontinuous state with strategic enclaves in areas rich in raw materials, Teotihuacan arises as a unique case in the Mesoamerican region, at a time when the Maya and Zapotec areas displayed different political strategies characterized by its exclusionism (Blanton et al. 1996; Hirth et al. 2020; Manzanilla 2006, 2009; Sugiyama 2004).

Among the many features of Teotihuacan, the use of color was among its most iconic and characteristic cultural expressions. Part of its relevance is related with the variety of media in which they expressed their appreciation for color, as it was applied on pottery, lapidary, bones, cloths, and the human body itself. However, it is in its application on Teotihuacan mural paintings where color acquires its greatest expression and dimension.

Teotihuacan mural painting is characterized by its bright appearance, color saturation, and precision of compositions and shapes. The two-dimensional design with its distinctive saturated red background, the use of clean strokes, and geometrical patterns are part of Teotihuacan stylistic and formal features (Lombardo de Ruiz 1996; Magaloni Kerpel 1996, 2003). The typical color palette is composed of seven basic colors: the characteristic red, orangish red, pink, yellow, green, blue, and, to a lesser extent, gray and black (Littman 1973; Magaloni Kerpel 1996, 2017; Torres Montes 1972).

Interest on Teotihuacan coloring materials began in the first decades of the twentieth century with the work of Manuel Gamio (1922), who proposed the geographical origin of the raw materials of several pigments (Torres Montes 1972, p. 22). Since then, the investigation of color in Teotihuacan has focused on two main aspects: a symbolic aspect and a technological one. The symbolic study of color is centered on its role in the Teotihuacan cosmovision and in the transmission of a political ideology, a mythical-religious thought and a way of life (Angulo Villaseñor 1987, 1995, 1996; Armillas 1944; Berrin et al. 1988; Kubler 1972; Miller 1973; Millon 1972, 1973; Villagra Caleti 1952, 1954, 1971).

A few archaeometric studies from the seventies constitute the first approaches to Teotihuacan color technology found in some of the neighborhoods from the city. Early studies on pigments from Tetitla, Atetelco and Tepantitla compounds identified materials that would later become common in this field: hematite (Fe2O3) for red, limonite (FeO(OH)) for yellow, azurite (Cu3(CO3)2(OH)2) for blue, malachite (Cu2CO3(OH)2) for green, and charcoal for black. In particular, three results stand out from these early studies: the identification of celadonite (K(Mg,Fe2+)(Fe3+,Al)Si4O10(OH)2) in some of the greens from Tetitla and Tepantitla, and the presence of cinnabar (HgS) as well as an unidentified—probably organic—grayish blue pigment (Littman 1973; Torres Montes 1972). Cinnabar has also been identified in a series of mural paintings from the Quetzapapálotl Palace (Goguitchaichvili et al. 2018), while the use of an organic blue pigment was further proposed for Teopancazco, another Teotihuacan neighborhood, as well as iron oxides for reds and yellows (Martínez García et al. 2002; Martínez García et al. 2012).

In one of the most comprehensive studies on Teotihuacan color technology, Magaloni Kerpel (1996) proposed four technical phases in the color technology of Teotihuacan mural painting, based on the analysis of the palette from the Tetitla murals. Focusing on the nature of the pigments used and the contour line, these four phases can be clustered in two main stages. The first one is characterized by the pictorial palette of the first technical phase, with hematite reds, iron oxide oranges, lepidocrocite yellow (γ-Fe3+O(OH)), malachite green, pyrolusite black (MnO2), and charcoal. The second stage begins with the Tlamimilolpa period, at around 200 CE, when the pictorial palette is expanded—along with the raw materials used—adding new colors such as a dark blue, an unidentified grayish blue, and various hues and shades of green. The change from the first to the second stage is also defined by the use of a red instead of the black contour line characteristic of the early stage. The identification of chalcanthite (CuSO4·5H2O) in the dark blue of Tetitla, along with the absence of cinnabar and the use of mixtures of raw materials for the green pigments, stand out for its singularity (Magaloni Kerpel 1996).

Recent studies have focused on color production in Teotihuacan. A productive assemblage for the manufacture and storage of red pigments was registered in site 46C:N4E2, in the NE sector of the city. The discovery of lithic instruments such as mortars and grinders with traces of cinnabar, as well as 19 kg of hematite found in different productive stages, were clear indicators of the storage and processing of raw materials. So far, this is the only study at Teotihuacan that delves into the location of the geological sources of minerals used to manufacture pigments, finding a correlation between the hematite discovered in this productive space and samples of natural red earth from the nearby Sierra Patlachique (Sánchez Morton 2013).

In addition to the aforementioned study, pigments were found in several productive stages at the palace of Xalla (López Puértolas et al. 2020), from raw materials stored in pottery fragments, pigment residues in polished lithic instruments and pigments in intermediate production states, to ready-to-use pigment nodules. In addition, different typologies of production tools—crushers, grinders, and pestles—were discovered. The information collected led to the proposal of a four-phase pigment production process. The results obtained include the identification of new materials, such as conichalcite (CaCuAsO4(OH)) for the production of greens, and bone white (Ca10(PO4)6(OH)2) used to lighten an iron oxide orangish red pigment. Furthermore, the authors reported the inclusion of fillers with sparkling properties such as specular hematite (Fe2O3), quartz (SiO2), and mica.

Recently, in a first approach to mural fragments from Techinantitla site, earth pigments were identified on the yellow and red areas, and the authors report two types of blue pigments, an azurite blue and an unidentified grayish blue pigment, as well as the use of malachite mixed with earths for the green pigments. They proposed technological similarities between Techinantitla pigments and Tetitla. Finally, the authors discuss the possible origin of minerals such as azurite, malachite, or specularite (Ruvalcaba Sil et al. 2021).

In general, these previous studies applied one or more analytical techniques, combining imaging, microscopy and spectroscopic analyses. Such techniques include scanning electron microscopy (SEM), X-ray fluorescence (XRF), X-ray diffraction (XRD), fiber-optics reflectance spectroscopy (FORS), Fourier transform infrared spectroscopy (FTIR), Raman spectroscopy, and particle-induced X-ray emission (PIXE), among others. These analyses were performed to characterize the main materials that give rise to the color, with lesser emphasis on the mixture of materials, the identification of fillers or in understanding the pigment palette through time. Furthermore, these studies did not compare the different architectonic compounds and they did not integrate the data acquired into the past historical dynamics of Teotihuacan.

In the present work, we applied a series of non-destructive and non-invasive analytical techniques to a selection of mural painting fragments from three Teotihuacan architectural contexts. We compared the results with those from previous works on other Teotihuacan compounds from synchronic and diachronic perspectives, allowing us to evaluate the development of color technology in Teotihuacan, and insert the discussion in the history of the city itself. We combined the theoretical frameworks from archaeology of color and the anthropological scope of technology in order to propose a new approach to the technological history of color materials in Teotihuacan.

New approaches to Teotihuacan color technology: the archaeology of color and the application of the concept of technological style

The investigation of color as a social and technological object is one of the aspects covered by the archaeology of color (Jones 2002; Sepúlveda 2021). This line of research places color at the center of the archaeological discussion through two main approaches. In the first, color is studied as an abstract, symbolic phenomenon where it serves as a vehicle loaded with symbolism and coded information that reflects a system of beliefs and concrete values. In the second, color is investigated as a physical phenomenon, where its technology, in its broadest and most general meaning, becomes the main protagonist (Jones 2002, p. 3).

Technology can be understood as a multifaceted cultural phenomenon that encompasses physical, social, economic, and cognitive aspects (Lemonnier 1986, pp. 147–148). It operates as a phenomenon where both material and social agents intervene, providing relevant information from a past society, regarding its knowledge and production processes. In this way, technology is—in its broadest sense—a fully integrated system that manifests cultural values and choices (Lechtman 1977, pp. 3–4; Sillar and Tite 2000). Research on color based on this perspective allows us to look into the raw materials used in the production of pigments, their procurement process and the interregional relations required to obtain them, the manufacturing techniques and tools employed along with the knowledge and skills of the producers, the cultural decisions made in the whole process, the reconstitution of color palettes, and their permanence—or variations—over time and on a regional scale, just to mention a few main aspects. Hence, through the technological study of the ancient world colors, it is possible to reconstruct stories that permeate the social, economic, religious, and political network of a specific culture (Jones and MacGregor 2002; Sepúlveda 2009, 2011, 2021; Sepúlveda et al. 2019).

Furthermore, color materials have a series of definable features and characteristics that must respond to decisions made by the artisans, based on preferences and demands from each moment in time, taking into account the socio-political, economic, and religious context of the culture and generating a particular style of manufacturing. Style, understood in a broad sense as the specific character of an object produced by an individual or social group in a given socio-political, cultural, and temporal context, is mainly reflected in the esthetic and technological qualities that have their own characteristics and are subject to change according to the influences or interests of the societies that produce them over time (Álvarez Icaza Longoria and Escalante 2017). As such, style can be addressed from different fields of analysis, where the esthetic and the technological scopes are the main ones, with this later field being termed technological style (Lechtman 1977). This concept of technological style has been applied to other Teotihuacan industries such as lapidary or the elaboration of stuccoes (Melgar Tísoc and Solís Ciriaco 2018; Melgar et al. 2021; Pecci et al. 2018).

Derived from the above, in order to define a technological style, it is necessary to look for constant attributes that are transmitted over time, representing a preference to do things in one way instead of the other, as well as those that vary over time. Our study on Teotihuacan pigments has led us to identify a series of technological and physical attributes that help define a technological style of color in Teotihuacan: the variety of hues and shades, the raw materials that generate the chroma and the fillers—taking into account their identification, granulometry, morphology, and distribution— the proportion and distribution of the different components in the pictorial layer, and the superimpositions of color to create specific hues or shades.

In order to address the technological aspects outlined above, it is necessary to perform pigment microarchaeology (Sepúlveda 2021), which requires the use of a combination of techniques with the ability to analyze micrometric particles. With this approach, detailed information on the material and technological nature of the pigment matrix can be obtained through a precise identification of the small particles from various raw materials that serve as fillers to generate the different physical and optical properties observed in Teotihuacan colors.

Archaeological contexts

We studied mural painting fragments from three Teotihuacan contexts: Plaza de los Jaguares (part of the Quetzalpapalotl Palace) (19° 41′ 52.4″ N 98° 50′ 45.2″ W)Footnote 2, Amanalco (Techinantitla and Tlacuilapaxco) (19° 41' 52.4″ N 98° 50' 21.9″ W) and two architectural assemblages in the Tlajinga district (19° 40' 12.3″ N 98° 51' 11.6″ W) (Fig. 1). These compounds belong to different temporalities in Teotihuacan history and represent diverse socio-economic levels, as determined by their location, construction materials, the size of their spaces, and the documented material culture. While Plaza de los Jaguares and Amanalco, in the civic-ceremonial heart of the city, would have a higher status, Tlajinga, located on the periphery of Teotihuacan, would have a lower socioeconomic rank than its counterparts in the city core.

Fig. 1
figure 1

Map of Teotihuacan with the architectural assemblages under study marked in green. Other contexts referred to throughout the text are also indicated

Full size image

Plaza de los Jaguares

The oldest of the three contexts, the Plaza de los Jaguares compound, has a late Tlamimilolpa and early Xolalpan occupation (ca. 250–550 CE). It is part of the Quetzalpapalotl architectural compound, located southwest of the Moon Pyramid (Ortega Cabrera 2020). During the course of the Integral Conservation Project of the Patio de los Jaguares Compound between 2012 and 2014 (Rodríguez Torres et al. 2013; Rodríguez Torres 2014), the central, north, and south sections of this place were intervened, revealing substructures associated—by ceramic typology—to a period around 250–350 CE (Tlamimilolpa phase). Archaeological works in this site exposed an open central courtyard, as well as several rooms adjacent to this central space, all of them built in stone and richly painted. In these early stages of the Patio de los Jaguares compound, the murals were painted on stucco supports, as well as on adobe and on the rock itself in some architectural elements.

Tlajinga

Tlajinga is the name given to the cluster of architectural compounds located about 3 km south of the civic-ceremonial center of Teotihuacan, on the west side of the Street of the Dead. Tlajinga occupies an area of 1 km2 and is one of the largest apartment compounds with multifamily character in the southern part of the city. It has been considered as a possible Teotihuacan peripheral district occupied by a population of medium-low socioeconomic status, mostly Teotihuacan, involved in the production of Anaranjado San Martín pottery and slate lapidary. The architectural differences between the structures found in Tlajinga—some made from adobe and others with stone masonry—are evidence (together with other indicators) of the presence of varied socioeconomic levels within the district (Carballo 2017, 2019; Carballo et al. 2019, 2021; Widmer 2019; Widmer and Storey 2012).

Excavations in what was referred to as the East Subdistrict of Tlajinga by Widmer and Storey (2012, p. 106) provided novel data regarding the socioeconomic status and access to sumptuary objects in two apartment compounds: Compound 2:S4W1, a space with a courtyard and several structures with various levels of occupation, adobe, and stone architecture with polychrome mural paintings, and Compound 4:S4W1, a group of several structures around an open patio of approximately 64–80 m2 with mural paintings in red and pink, spanning from the late Tlamimilolpa to the early Xolalpan phases (ca. 300/350–350/450 CE). These two compounds, located next to Street of the Dead, have larger dimensions than the rest of those found in the East Subdistrict and were built in stone masonry, with stucco walls and bichrome and polychrome mural paintings. So far, these are the only mural paintings with complex iconography found in the peripheral area of Teotihuacan (Carballo 2017, 2019; Carballo et al. 2019, 2021).

Amanalco: Techinantitla and Tlacuilapaxco

Located to the northeast of the civic-ceremonial core of the city, Amanalco (ca. 350–550 CE) is the latest archaeological context studied in this work. Amanalco is a set of thirteen structures including Techinantitla and Tlacuilapaxco. Techinantitla is a large architectural assemblage with several structures, including living spaces and a temple (Millon 1991; Millon and Sugiyama 1991). As for Tlacuilapaxco, its spatial organization and size are unknown, as only a portion of the entire site has been excavated. Roughly 1500 fragments of monochrome, bichrome, and polychrome mural paintings applied on both stucco and adobe were recovered during the prospecting works and archaeological excavations, with figurative representations of priests, deities, and mythological flora and fauna. The themes and style of the fragments, as well as their technical complexity, are signs of the high socioeconomic level of the compound. Recent archaeological prospection works at this site suggest that Techinantitla appears to be an assemblage of large rooms around a main temple with a courtyard and exit to the west, similar to the Yayahuala compound but on a smaller scale (Millon 1991; Millon and Sugiyama 1991; Ruvalcaba Sil et al. 2021).

The pictorial fragments recovered in Techinantitla and Tlacuilapaxco comprise one of the richest corpus of Teotihuacan mural painting and can be compared to those of other groups such as Tetitla or Tepantitla. Murals depicting only red and pink hues coexist with others presenting a rich polychromy, with their pictorial palette composed of this red and pink, as well as yellow, green, blue, and black colors.

Materials and methods

Samples

The mural fragments studied (n = 66, Table S1) were recovered during the archaeological campaigns of three different projects: Amanalco Project, directed by Michael Smith (1984–1991), Tlajinga-Teotihuacan Archaeological Project (Dirs. David M. Carballo & Luis Barba Pingarron, since 2012), and the Project for the Integral Conservation of the Patio de los Jaguares Compound, directed by Verónica Ortega (2012–2014). The fragments encompass a temporality of nearly 300 years, from the late Tlamimilolpa (ca. 250–350 CE) to the late Xolalpan (ca. 450–550 CE) phases (Carballo et al. 2021; Ortega Cabrera 2020; Ruvalcaba Sil et al. 2021).

The 23 fragments from the Plaza de los Jaguares Compound (Fig. 2a and b) are the oldest of the group (Tlamimilolpa ca. 250–300 CE in the late Tlamimilolpa phase). The pictorial palette is composed of red, pink, orangish red, yellow, green, grayish blue, and black colors. The iconography is varied and presents the distinctive features of the third iconographic phase of Teotihuacan, especially in the use of red backgrounds with contours and motives of the same hue that links them to the third stylistic phase of Teotihuacan (Lombardo de Ruiz 1996). These colors were applied on a variety of supports, which include stucco plasters, direct painting on rock objects—without a preparative layer—and on adobe surfaces.

Fig. 2
figure 2

Images of selected fragments from a, b Plaza de los Jaguares; c, d Tlajinga; and e, f Amanalco

Full size image

The next set of materials in terms of temporality (Late Tlamimilolpa-Early Xolalpan ca. 300/350–350/450 CE) are those from Tlajinga’s 2:S4W1 and 4:S4W1 Compounds (n = 10) (Fig. 2c and d). The mural paintings of these two groups are elaborated with a more restrained pictorial palette: pink, red, orangish red, and green colors. The red contour line, the use of pink/red bichrome murals, the red backgrounds, and the iconography exhibited place these murals within the third stylistic phase of Teotihuacan (Carballo et al. 2021).

The mural painting fragments from Amanalco (n = 33) (Fig. 2e and f) come from surface surveys and excavations in the Techinantitla and Tlacuilapaxco Compounds. The mural painting fragments studied exhibit a rich polychromy formed by a pictorial palette combining red, pink, yellow, green, blue, and grayish blue colors. As for its esthetic style, from the hues of the colors and the figurative iconography, these paintings belong to the fourth stylistic phase of Teotihuacan (Xolalpan 450–550 CE) (Lombardo de Ruiz 1996).

Analytical methods

Fragments were studied both in situ at the warehouse of the Arizona State University (ASU) Teotihuacan Research Laboratory in San Juan Teotihuacan (Edo. de México, Mexico), and at LANCIC Laboratory (Laboratorio Nacional de Ciencias para la Investigación y Conservación del Patrimonio Cultural), in the Physics Institute of the National Autonomous University of Mexico (UNAM).

Microscopic examination

All fragments were analyzed by digital microscopy, combining both zenithal and oblique light observations, with a portable digital microscope Dino-Lite Edge AF4915ZT at 50X, 100X, 150X, and 220X magnifications. Direct LED illumination was used, with polarized light and extended depth of focus (EDOF).

Scanning electron microscopy coupled with energy dispersive spectroscopy (SEM-EDS) analyses were performed with a Hitachi TM3030 Plus Scanning Electron Microscope with a 15 kV working voltage. Backscattered electron imaging, secondary electron imaging, elemental mapping, and spot elemental analysis were collected.

Color measurement

Color measurements were acquired via a non-contact RUBY spectrocolorimeter with a 400–700 nm spectral range in the the CIE L*a*b* and CIE L* C* h* color spaces. The analyses were performed at an 8 cm working distance with a D65 illuminant on a 4-mm-diameter area, and a CIE 1964 observer. The data in CIEL*a*b* space was converted to CIE L*C*h*, in order to evaluate the saturation of pigments. In this color space, L* indicates brightness (0 to 100), C* saturation (C* = 0 represents a low intensity or gray pigment, and C* ≥ 80 an intense and vivid pigment), and h* the hue of that pigment. The L* and C* nomenclature used to define the color values of the studied materials is presented in Table S2 from a modification of a previous work (Pasco Saldaña 2010).

Fiber optic reflectance spectroscopy (FORS)

An ASD inc. FieldSpec-4 handheld device was used to acquire visible and near-infrared reflectance spectra in the 350 to 2500 nm range. A CIE D65-series external light source with a 6500 K average temperature was used in non-contact mode, with an 8 cm working distance to the fiber optic probe. The diameter of the analyzed area was approx. 3 mm and the spectra were obtained with a 5 s. integration time. Calibration was performed using a certified reflectance standard (AS-02035-000CSTM-SRM-990-362, ASD Inc.). Data was interpreted by comparison with reference spectra found in the literature (Aceto et al. 2014; Cosentino 2014) and with previous results of our group (Ruvalcaba Sil et al. 2021). Additionally, the first derivative of the spectrum was used to determine its inflection point.

X-ray fluorescence (XRF)

Elemental information was obtained through XRF analyses with SANDRA (system for non-destructive X-ray analysis), a portable system developed at LANCIC-IF (Ruvalcaba Sil et al. 2010). This equipment employs a 75 W Mo X-ray tube with a collimator providing a 1-mm-diameter analysis area, and an Amptek X-123 SDD X-ray detector. Analytical conditions were 0.150 mA, 35 kV, and a 200 s acquisition time. Additional measurements were performed on the plasters—taken from the cross section when possible—in order to discern between the elemental information of the lime substrate and that of the pictorial layer. Spectral deconvolution was performed with PyMca software to obtain elemental peak intensities (Solé et al. 2007). Absolute elemental concentrations were not determined.

Raman and infrared spectroscopies

Infrared reflectance spectra were collected using the external reflection mode of a Bruker Alpha FTIR Spectrometer. The system has a 400–4000 cm−1 spectral range with a 4 cm−1 resolution and a 5-mm-diameter analysis area. Compounds and minerals were identified by comparison with reference spectra found in the literature (Chukanov 2014; Vahur et al. 2010).

Two Raman spectrometers were used. The first is a BWTek i-Raman EX portable system, which employs a 1064-nm laser with a 120 mW maximum power and a 4 cm−1 spectral resolution. Measurement times were 5 and 10 s, with either 20 or 50 μm spot sizes. The second Raman spectrometer was a Thermo Scientific DXR model, employing 532 and 785 nm lasers with variable intensity, a 50X magnification, 50 μm slit aperture, and an 8 cm−1 spectral resolution. Three to five repetitions were averaged per acquisition. The spectra obtained were compared with RRUFF database to identify the material composition of pigments (Lafuente et al. 2015).

Results and discussion

Results obtained from the multi-technique characterization of the mural painting fragments are summarized in Table 1. As will be discussed in the following pages, these results allowed shedding new light on Teotihuacan colors and revealed a technological standardization. This standardization is determined by the use of base formulations: a coloring material, a clay matrix, and fillers. These base formulations are adapted in each of the architectural contexts studied, mainly by changing the material responsible for the color. In addition, different pictorial palettes have been observed on the three architectural compounds studied, all of them in accordance with the development of color technology proposed in previous works (i.e. Argote et al. 2020; Littman 1973; Magaloni Kerpel 1996; Martínez García et al. 2002; Martínez García et al. 2012; Ruvalcaba Sil et al. 2021; Torres Montes 1972).

Table 1 Main elements and materials identified in the analyzed fragments and samples from Plaza de los Jaguares, Amanalco (Techinantitla and Tlacuilapaxco) and Tlajinga neighborhoods by XRF and SEM_EDS, FORS, Raman, and FTIR
Full size table

We found a number of relevant characteristics in the pictorial palette of the Plaza de los Jaguares compound. Orangish red, a color related to the earliest phases of the city, is widely used. We also found a grayish blue while the dark blue characteristic of Xolalpan was not detected. Of note in both Tlajinga compounds was the absence of colors such as yellow or grayish blue, commonly found in other contemporary architectural compounds, along with the presence of orangish red. Finally, Amanalco showed the typical color palette of the Xolalpan phase (Magaloni Kerpel 1996), where orangish red was rare—only observed in one mural fragment—and a dark blue was identified, as opposed to Plaza de los Jaguares and Tlajinga. Another peculiarity in Amanalco is the use of an underlying pink layer, which serves as a preparation layer for the subsequent pictorial layers.

Red color

Red, with several shades and hue variations, is the color more often depicted in the three compounds studied. It is composed of four hues: Teotihuacan red, orangish red, pink, and light pink. These colors are present in all of the compounds, with the red and pink exhibiting remarkable technological standardization. The exception was found in the orangish red, as the three compounds displayed singular technological characteristics, such as the absence of fillers or the fineness of its grains.

Regarding the material composition, Teotihuacan red and the two pinks share the same color raw materials: they are composed of matrices rich in red earths, calcite, and kaolinite. Their identification was based on XRF’s detection of elementsFootnote 3 related with these minerals (Fe, Ca, Si, Al, and Sr), as well as on the characteristic bands of kaolinite and calcite observed in the FTIR spectra (Table 1 and Fig. 3) (Chukanov 2014, p. 473; Vahur et al. 2010). The use of hematite-containing natural earths is common in other areas of the city since it was an accessible and abundant raw material with several sources located in the surrounding areas of Teotihuacan (Littman 1973; Magaloni Kerpel 1996; Sánchez Morton 2013; Torres Montes 1972). Additionally, higher calcium content is observed in the light pink and pink pigments, indicating the use of calcite to lighten the shade.

Fig. 3
figure 3

Results from red color. a Comparison of XRF elemental peak intensities of red areas in representative samples from the three compounds, presented as a percentage of the sum of all the elemental peak areas. b Representative FTIR spectra for the red colors from Plaza de los Jaguares compound

Full size image

The technological standardization found in the red and pink colors was not observed for the orangish red, which had a smaller but still continuous presence in all the contexts. The orangish red found in Plaza de los Jaguares and Amanalco mainly contains Fe, Hg, and S, which is confirmed by the presence of Raman bands from cinnabar (252 and 345 cm−1) and hematite (224, 290, and 410 cm−1), implying a mixture of these minerals (Fig. 4a and 4). This result also confirms the use of cinnabar in Teotihuacan mural painting, at least during the Tlamimilolpa period (170/200–350 CE), which was also reported for the mural painting of the Feathered Snails compound, corresponding to the Early Tlamimilolpa phase (Argote et al. 2020).

Fig. 4
figure 4

Results from orangish red color. a XRF elemental peak intensities (normalized to Ca) of orangish red areas in representative samples from Amanalco and Jaguares compounds, presented as a percentage of the sum of all the normalized elemental peak areas. b Representative Raman spectra from Plaza de los Jaguares compound confirming the presence of cinnabar and hematite

Full size image

The archaeological register also reported the presence of an orangish red in Tlajinga’s mural paintings. This classification was based on visual comparison with the other reds found in Teotihuacan. However, its composition suggests a red hue. Elemental analysis of this pigment yielded high amounts of Ca and Fe, along with characteristic elements from clays (Fig. 5a), while the presence of Cu, As, Zn, and P is most likely related to an underlying green pigment shown in Fig. S1. Molecular spectroscopies only identified the presence of hematite, anatase, and montmorillonite (Fig. 5b–d), with this last mineral identified through its FORS and FTIR bands at 2214 nm and 427, 467, 535, 787, 904, 1007, 1044, and 1121 cm−1, respectively (Chukanov 2014, p. 487-488). The presence of this montmorillonite-type clay diverges from the common presence of clays of the kaolinite or kaolinite-montmorillonite type in Teotihuacan pigments (López Puértolas et al. 2020). While anatase was also found in other mural painting samples from all the colors and compounds, the higher titanium content found in this pigment when compared to the other reds could explain the slight difference in hue. An additional unique feature of the “orangish red” from Tlajinga is the presence of a less compacted and significantly flimsier matrix, without fillers of any kind.

Fig. 5
figure 5

Tlajinga “orangish red” color. a XRF elemental intensities (normalized to Ca), presented as a percentage of the sum of all the normalized elemental peak areas. b Raman spectra of hematite and anatase. c Characteristic FORS and d FTIR spectra with the main bands of montmorillonite clay

Full size image

The compositional differences discussed above are reflected in the chromatic diagram shown in Fig. 6. The red and pink pigments, with similar composition, are clustered in lower a* and b* values, the cinnabar-containing orangish red from Plaza de los Jaguares and Amanalco compounds display higher a* and b* values, and the “orangish red” from Tlajinga (orange triangles in Fig. 6) falls closer to the red-pink cluster, confirming what was suggested by the compositional analysis.

Fig. 6
figure 6

Chromatic diagram a* b* of the red, orangish red, and pink colors for all the studied compounds

Full size image

The use of cinnabar in the early Teotihuacan mural paintings was abruptly interrupted after 350 CE, as suggested by both our data and previous reports (Argote et al. 2020; Gazzola 2009; Guzmán García Lascurain 2019; Littman 1973; Magaloni Kerpel 1996, 2017; Magaloni Kerpel et al. 2020; Torres Montes 1972). Although cinnabar is no longer found in mural paintings after ca. 350 CE, it continues to be used as a pigmenting material in funerary and ritual contexts and decorated pottery. Cinnabar has been found in funerary contexts from Teopancazco (ca. 350 CE) and La Ventilla (350–550 CE) (Doménech Carbó et al. 2012; Ejarque Gallardo 2018; Vázquez de Ágredos and Manzanilla 2016), in stuccoed painted ceramics from La Ventilla and other contexts from the Xolalpan period (Conides 2018; Magaloni Kerpel et al. 2020; Maruf Martínez 2019), where the use of this material is recurrent for painting both details and backgrounds.

This continued use of cinnabar in different contexts and substrates places the question of whether it was no longer used in mural painting due to a technical matter concerning its stability, or if it is related to a constriction, reduction, or restructuring of the supply networks of this material from 350 CE onward. Around this time, a full renovation of the city’s architecture took place, combined with desacralization and iconoclastic processes. A great number of termination rituals appear in contexts from this time, suggesting relevant sociocultural, political, or religious changes that could have affected the arrival of raw materials to Teotihuacan (Cabrera Castro et al. 1991; Ejarque Gallardo 2018; Manzanilla 2000, 2012).

Along with the pigment raw materials, Teotihuacan artisans included fillers aiming to obtain different hues and generate different esthetic or physical properties of the pigments. Carbon particles are often found in the reds in variable amounts and, in combination with calcite and red earths, led to the red gradations recorded for the three studied compounds. Fine-grained carbon particles (ca. 25–45 μm, and identified through Raman spectroscopy, Table 1) are disseminated throughout the matrices of Teotihuacan red and the two pink hues, with higher amounts of particles scattered in the matrix of the pink pigments (Fig. S2).

Other common fillers are specular hematite and quartz. The first was usually added to Teotihuacan red in the form of small, flat, angular, black lamellar particles (Fig. 7a–c). It was found in the Teotihuacan red pigment from all the contexts with varied sizes, ranging from ca. 25–35 μm to 150–200 μm, which is considerably larger than the grain size of the matrix. This difference in size and shape of the specularite particles with respect to the red matrix points to different grinding and preparation processes of this material. The aim was to obtain flat fragments of different sizes to generate—once the pigment is applied and burnished—the sparkling and iridescent effect characteristic of Teotihuacan mural painting. On the other hand, transparent quartz particles (identified by Raman spectroscopy, Fig. S3) substituted specularite in the pink colors, suggesting a technological choice to give this typical sparkling appearance.

Fig. 7
figure 7

Microscopic images (100X) of specular hematite particles in the red colors from a Plaza de los Jaguares compound, b Tlajinga and c Amanalco. d Raman spectra of hematite

Full size image

Yellow color

Yellow pigments were found in Plaza de los Jaguares and Amanalco compounds, showing a remarkable technological uniformity in both contexts. They present light tonalities and medium saturation, with a tendency to a red hue. In general, the yellow color consists of a whitish calcite matrix colored with yellow earths containing goethite and kaolinite, as determined by Raman and FTIR (Fig. S4). To this finely ground matrix, considerably larger particles of the same raw materials (from 30 to over 100 μm in size) were later added, probably to increase the saturation of the pigment (Fig. 8). This particular technological approach has already been reported for other Teotihuacan contexts and substrates, such as ceramics or the human body itself (Gazzola 2009; Littman 1973; López Puértolas et al. 2020; Magaloni Kerpel 1996; Magaloni Kerpel et al. 2020; O’Neil 2017; Torres Montes 1972; Vázquez de Ágredos and Manzanilla 2016; Vázquez de Ágredos Pascual et al. 2018). Other fillers observed were carbon particles and hematite. Carbon in different particle sizes can change the tone of the pigment, while hematite gives the characteristic reddish hue to this yellow pigment.

Fig. 8
figure 8

Microscopic images (100X) displaying a green, b goethite, and c goethite and carbon particles embedded in the yellow matrix of the color

Full size image

Slight variations to the “recipe” of natural yellow earths, calcite and carbon particles can be found in certain architectural compounds. A variant of yellow color was found in the mural of the Great Birds from Amanalco, with the addition of green particles (probably malachite) in order to achieve a greenish hue (Fig. 8a). The addition of malachite powder was already identified in Tetitla and other compounds from the Xolalpan period—corresponding to the Third technical phase—(Magaloni Kerpel 1996), providing further evidence of the technological continuity shown by the Teotihuacan color industry through time. Recently, the same variation was found in the yellow pigments of mural paintings from Plaza de las Columnas compound (Guzmán García Lascurain 2019) and in a sculpture of a jaguar at Xalla palace (Muiños Barros 2019) reinforcing the idea that it was a shared technological knowledge, used in compounds from the city’s main core.

Green color

The close relationship between raw materials, mixtures, and technological continuity is fully expressed in the range of greens. The use of a common “base” formulation with variations in the color material is clear for these pigments. In all cases, this base is composed of a calcite and clays matrix with fillers, and a green material. However, there are certain particularities on each of the architectural compounds that allow discriminating among the green pigments.

Two common formulations of green color were found in the Tlamimilolpa phase of Plaza de los Jaguares, which were later detected in the Xolalpan phases of Amanalco and Tlajinga compounds. The first consists of a homogeneous pigment with a whitish matrix, additions of green raw materials and crystalline fillers (Green 1). The second is a heterogeneous pigment produced from a light green matrix—made of a calcareous compound and green raw materials—with yellow, black, and crystalline particles added to generate the typical sparkling effect of the Teotihuacan pigments (Green 2). Magaloni identified the two formulations in the murals from Tetitla dating to the Miccaotli times (100–170 CE), naming them Bright Green (Green 1) and Dry Green (Green 2) (Magaloni Kerpel 1996).

Spectroscopic data shows that Green 1 was prepared from a calcite and kaolinite base (Fig. 9a), with either malachite (Fig. 9b) or pseudomalachite (Cu5(PO4)2(OH)4) powder (Fig. 9c) to generate the green hue and quartz powder for the sparkling effect. The colored material of Green 2 can also be either malachite or pseudomalachite, spread in a calcite and kaolinite matrix with powdered quartz. A difference with Green 1 is the addition of goethite-containing yellow earths and carbon particles to the matrix. Goethite was identified in Green 2 pigments—for the three studied compounds—from its characteristic Raman bands at 297, 395, 470, and 543 cm−1 (Fig. 9d). Additionally, Green 1 is more saturated than Green 2, due to the higher density and larger size of the malachite particles. The use of pseudomalachite is not as common as that of malachite for Teotihuacan greens, with only one previous report of its presence in mural paintings from La Ventilla (Gazzola 2009).

Fig. 9
figure 9

Material characterization of the green colors, displaying a FTIR bands associated with kaolinite and calcite, and characteristic Raman spectra of b malachite, c pseudomalachite, and d goethite

Full size image

The combination of malachite and pseudomalachite was only found in Plaza de los Jaguares. Furthermore, Green 1 from this compound was prepared in a way that differs from the one used 150 years later since its matrix is more translucent than that observed for the rest of the pictorial palette (to the point that the matrix is scarcely visible). The detection by SEM-EDS of a continuous calcium layer—with the acicular particles typical of calcite—surrounding the green crystals (Fig. 10) could indicate the use of lime wash as a medium for the malachite/pseudomalachite powder. This preparation and application technique for Green 1 was not observed in any of the later compounds but is mentioned by Torres Montes (1972, pp. 24–26) in his study of Teotihuacan mural painting.

Fig. 10
figure 10

SEM-EDS elemental mapping of copper particles (malachite and pseudomalachite) on a calcite matrix

Full size image

However, 100–150 years later, this preparation technique and the combination of malachite and pseudomalachite was abandoned, or the raw material sources may have changed, as suggested by our results and confirmed in the available literature (Gazzola 2009; Guzmán García Lascurain 2019; López Puértolas et al. 2020; Magaloni Kerpel 2017; Magaloni Kerpel et al. 2020; Maruf Martínez 2019; Ruvalcaba Sil et al. 2021). Only malachite was found in the green pigments from Amanalco, despite the great number of samples analyzed. In all cases, the green pigment in Amanalco is made with a greater amount of malachite and larger particle sizes (80–200 μm, Fig. 11a) than that found in the other two compounds, and could be an indication of the high socioeconomic level of the inhabitants of Amanalco. In comparison, the pseudomalachite particles from Tlajinga (in Fig. 11b), as well as the malachite particles from Tetitla’s murals of the Xolalpan phase—contemporary to Amanalco—are notoriously smaller, ca. < 50 μm (Magaloni Kerpel 1996).

Fig. 11
figure 11

Microscopic images (at 100X) of two green samples showing the difference in size of the green particles that make up the pigment matrix. a Amanalco and b Tlajinga

Full size image

Blue colors

The last evidence of technological continuity in Teotihuacan color can be found in the blues hues. In this research, we have documented two types of blue (Fig. S5): a dark blue 1 and a lighter blue 2. Blue 1 can be related to the Xolalpan phase (from 350 CE) as it was only observed in Amanalco, the latest of the three compounds. Blue 2 has received plenty of denominations: ultramarine blue by Magaloni (Magaloni Kerpel 1996, p. 211), grayish blue by Torres Montes (1972, p. 25), Teotihuacan Blue by Vázquez de Ágredos (Vázquez de Ágredos Pascual et al. 2018, p. 289), or even simply blue by Gazzola (2009, p.62). The production and use of this second blue is well documented since the Tlamimilolpa phase in Tetitla Compound (170/200–350 CE) (Magaloni Kerpel 1996) where it is scarcely used. In our research, this blue was present in Amanalco, and in a limited form in Plaza de los Jaguares (with a similar temporality as Tetitla).

Both types of blue colors are produced with the same calcite and kaolinite matrix observed in the previous coloring materials, which were characterized in a similar manner as before (see Fig. 3b). Blue 1 has close technological ties with the green copper pigments, as two coloring raw materials are incorporated into the matrix: the blue chroma is provided by azurite, with malachite being added in smaller quantities providing a greenish hue to the pigment. While an initial roughing of the minerals was probably performed individually on these raw materials, the pigment was then homogenized by finely grinding it with mortars in a precise and regular manner. Both minerals are present with grain sizes in the 40–60 μm range and were identified by Raman spectroscopy (Fig. 12) (Frost et al. 2002; Jorge-Villar and Edwards 2021; Lafuente et al. 2015). Edwing Littman (1973) and Luis Torres Montes (1972) identified azurite blue pigments in other Teotihuacan compounds, although without specifying the context. The characteristic dark shade of blue 1 is achieved by the addition of carbon particles by the artisans in a considerably higher amount than in the rest of the pictorial palette analyzed.

Fig. 12
figure 12

Characteristic Raman spectra of a azurite and b malachite from the blues samples

Full size image

Regarding blue 2, previous works describe four features for this pigment: its blue color, grayish hue, fineness of particles with no fillers, and weathering process on the wall. This pigment has been described as composed of the organic dye indigo, chrysocolla ((Cu,Al)2H2Si2O5(OH)4·nH2O), a manganese-containing silicate, or an undetermined organic material (Torres Montes 1972; Littman 1973; Magaloni Kerpel 1996; Ruvalcaba Sil et al. 2021, respectively). Our research confirms the four features, including the grinding process producing fine particles, though we could not measure the grain size. In contrast to the previously proposed compositions from blue 2, we found that this pigment is mainly composed of a mixture of calcite (Raman bands at 152, 278, 710, and 1082 cm−1) and carbon particles, identified by Raman spectroscopy for its characteristic double band around 1360 cm−1 and 1595 cm−1 (Fig. 13) (Gatta et al. 2012; Lafuente et al. 2015; Tomasini et al. 2012, 2015;). No azurite was found, also suggested by the scarce or even null detection of copper.

Fig. 13
figure 13

Raman spectra of various blue type 2 colors displaying characteristic peaks of calcite and carbon

Full size image

The use of bone black was suggested by the presence of phosphorus in the lone sample with blue 2 analyzed by SEM-EDS (Fig. S6). Phosphorus could not be unequivocally detected by XRF in any of the samples, probably due to a combination of factors: the small amount of bone black used together with XRF’s greater depth of analysis, and the superposition of calcium’s escape peak with phosphorus K peak. Raman spectroscopy may identify bone black by the presence of an asymmetric stretching of the calcium phosphate band around 960 cm−1. Although this band is usually lost as a consequence of the combustion process (Correia et al. 2007; Gatta et al. 2012; Tomasini et al. 2012, 2015), it was clearly identified in one of the samples from Amanalco (A64) and was present as a weak signal in the Raman spectra of Plaza de los Jaguares’ blue 2 (Fig. 14).

Fig. 14
figure 14

Raman spectra of the blue 2 color samples with characteristic bands of bone black and calcite

Full size image

It should be noted that bone was a common and widely used material in other Teotihuacan industries, such as the manufacture of clothing and even in pigments, as finely ground calcium phosphate was identified in Xalla as a raw material for the manufacture of orange-tinged pigments (López Puértolas et al. 2020). Along with this, treatises on paintings such as that of Max Doerner (2011) specify the bluish hue generated by the presence of bone black, a fact also reported by Barba Pingarrón and Rodríguez (1990). Nevertheless, further research—including experimental archaeology—is needed to determine the type (or types) of carbon-based pigment that is used to produce this bluish gray color in Teotihuacan.

The formulation of this bluish gray, identified in two compounds with a temporal distance of ca. 150 years, denotes the survival and transmission of the pigment “recipe” between generations of artisans, represented in the raw materials, their degree of grinding, the portion of each material to add, and how to mix them, all seeking to obtain a very specific and defined hue. Finally, this mixture of carbon particles and calcite to obtain bluish hues was also identified in several sites of the Classic Maya Area such as La Sufricaya (ca. 400–450 CE), and the Hunal structure of Copan (ca. 430 CE), and at Bonampak, Uxmal, and Xuelén from Late Classic (ca. 600–950 CE) were, slightly differing from our results, the bluish gray pigment was found with additional traces of azurite (Houston 2009; Magaloni Kerpel 2001, p. 174).

On the differential use of raw materials in Teotihuacan colors

The compositional variety of the pigments in the different compounds gives insights on the mechanisms for procurement and distribution of raw materials in Teotihuacan. In the red colors, these differences can be found in the use of cinnabar or iron oxides with salts or oxides of titanium to shift the red hue towards an orange one, and in the addition of carbon particles, quartz, and specular hematite to generate sparkling optical properties. It is possible that each architectural compound had its own network for the supply of raw materials, as reported before for other activities like the green stone lapidary industry (Melgar Tísoc and Solís Ciriaco 2018; Melgar et al. 2021; Ortega Cabrera et al. 2019) or the production of clothing in Teopancazco, a neighborhood strongly tied to the Veracruz coast in the Gulf of Mexico (Manzanilla 2011, 2017; Manzanilla and Valadez 2017; Manzanilla et al. 2011; Pecci et al. 2018; Rodríguez Galicia et al. 2017). This new data points to a differential or preferential access to raw materials in the city, depending on the demand and certain exchange mechanisms still poorly understood in Teotihuacan.

Furthermore, the standardization of the main raw material chosen by Teotihuacan artisans for elaborating the yellow colors, whether they were intended for mural painting, ceramics, or stone, is interesting in the context of the productive system of color in this Mesoamerican city. The absence of other yellow materials found in previous studies, such as jarosite (KFe3(SO4)2(OH)6), opens new questions about the technological choices of the artisans. Jarosite is a mineral already identified in Teotihuacan ritual contexts, as in the case of the yellow pigment cores deposited inside a Spondylus crassisquama shell and in the “mouth” of a Chama coralloides from a ritual deposit in the palace of Xalla (López Puértolas et al. 2020).

The use of goethite for most contexts and jarosite for ritual ones speaks of an intentional choice, either for economic, socio-cultural, or technical reasons. A hypothesis is that the choice was related to the economic value of the raw material, determined by its ease of access, and to the symbolic values associated with the location where the material was collected. This phenomenon has been observed in later Mesoamerican populations, such as the Nahuas or the Mayas of the Postclassic period (Dupey García 2018; Houston 2009; López Austin et al. 2005; Tokovinine and McNeil 2012).

For the green colors, the selection of malachite for Amanalco, pseudomalachite for Tlajinga and a mixture of both for Plaza de los Jaguares, is another evidence of the differences observed among the architectural compounds. It points to either a change or a differential access to raw materials, as already discussed for the red and yellow colors. The same occurs with the green pigments from Xalla palace, which are made of glauconite ((K,Na)(Fe3+,Al,Mg)2(Si,Al)4O10(OH)2) and celadonite (K(Mg;Fe2+)(Fe3+;Al)Si4O10(OH)2), raw materials with geological deposits in the surroundings of the Central Valleys of Oaxaca (Panczner 1987). Moreover, the second largest deposit of Oaxacan mica in Teotihuacan was found in Xalla, and it was also added to this green pigment, revealing a privileged use of this mineral by the inhabitants of this palace (Manzanilla et al. 2017; Rosales de la Rosa and Manzanilla 2011). Once again, we find technological differences between the pigments from peripheral compounds—such as Tlajinga—and those located in the urban core of the city (Amanalco, Plaza de los Jaguares, and Xalla), further indicating a differential access to raw materials in Teotihuacan (López Puértolas et al. 2020).

Finally, blue 1 found in Tetitla murals presented a similar base formulation, yet its main coloring raw material is a mixture of malachite and a copper sulfate, chalcanthite (a material that has not been identified in any other blue with these characteristics) (Magaloni Kerpel 1996). This use of the same base formulation but with a different color material is analogous to the cases already discussed above. They constitute technological peculiarities that some contexts present within the standardized base formulations.

Proposal for a technological style in Teotihuacan mural painting

Our data, together with previous reports (Argote et al. 2020, Ejarque Gallardo 2018; Gazzola 2009; Guzmán García Lascurain 2019; Littman 1973; López Puértolas et al. 2020; Magaloni Kerpel 1996; Magaloni Kerpel et al. 2020; Martínez García et al. 2002; Martínez García et al. 2012; Maruf Martínez 2019; Ruvalcaba Sil et al. 2021; Sanchez Morton 2013; Torres Montes 1972; Vázquez de Ágredos Pascual 2018), suggest that for the time lapses analyzed, there was a standardized technology for the production of color in Teotihuacan. In contrast, other Mesoamerican regions—such as the Zapotec tombs in Oaxaca and the Mayan mural paintings—show diversity in the pictorial palettes through their color technology, hue, and shade variations and pictorial overlays (Arano et al. 2020; Magaloni Kerpel 2001; Magaloni and Falcón 2008; Vázquez de Ágredos Pascual 2010).

Color technology in Teotihuacan was based on the use of earth pigments, prepared via a basic, rigid formulation, well established, and transmitted over time. Based on our results and those from previous studies, we propose a series of technological markers of a defined technological style in Teotihuacan colors, which can be detected continuously throughout different chronological phases and architectural assemblages in the city. These markers are the continued presence of particular hues in the pictorial palette, the composition of the clay matrices, and the inclusion of carbon particles and materials with sparkling optical properties.

For the first marker, color measurements showed the selection of a number of hues throughout the pictorial palette. These hue preferences were consistently confirmed in the three compounds studied, with no abrupt variations.

The second marker is the clay matrix. By using white, partially translucent clays, Teotihuacan color artisans generated earthy, dense pigments with high covering power. These clays were sometimes combined with calcium carbonates, thus acting as cementing agents and providing a structure to hold the remaining raw materials. The presence of these two types of clay—kaolinite and montmorillonite—in the pigment matrices points to the exploitation of different deposits, to the existence of different supply routes for the raw materials or even to technical choices made by the artisans.

The third technological marker of the Teotihuacan technological style of colors is the use of carbon particles. We found this material in almost all of the pictorial palette studied (only missing in the scarce green 1 samples studied), also showing a temporal continuity in its application, in accordance with previous reports of its use in Tetitla, Plaza de las Columnas and the Xalla palatial compound, among others (Guzmán García Lascurain 2019; López Puértolas et al. 2020; Magaloni Kerpel 1996). Carbon particles were used in two different ways: as a filler to vary the shade in red, pink, yellow, green, and blue colors, and as a material that generates chroma in the blue-gray pigment.

The fourth technological marker we propose is the use of sparkling and iridescent materials. Teotihuacan artisans used three raw materials to achieve this aim: mica (valuable and scarce, mainly present in the green colors), quartz (bright and translucent, used in pink and green colors), and specular hematite (present only in the characteristic Teotihuacan red color).

The four features described above define technological choices, a way of making pigments that Teotihuacan artisans developed during approximately five centuries. During this time span, the artisans went through different processes of exploration, improvement, and mastery of their craftsmanship to the extent that the technique adapted to the prevailing demands, therefore creating and perpetuating a marked technological style.

Conclusions

This work expands and deepens the existing knowledge on Teotihuacan color technology, revealing continuities and discontinuities in the three archaeological compounds studied. It was possible to characterize the matrices of clays and calcite, the pigments in each of the colors, as well as the fillers used in different finishing steps.

The color palette of the three compounds studied showed a temporal continuity, with materials common to other areas of the city, including Teotihuacan red, two kinds of pink and a yellow, all made from earths, a green made from malachite and pseudomalachite, a blue made from azurite, and a grayish blue prepared from carbon and calcite. The materials used to shift red towards an orangish hue displayed a higher variability, as cinnabar and hematite were applied in Plaza de los Jaguares (until ca. 350 CE), and earth pigments in Tlajinga.

Comparing our results with previous studies, the consistency of pigment technology in Teotihuacan became evident. In the time frame approached in our study, Teotihuacan compounds not only share the color palette, but also a standardized color technology. Artisans used a base formulation that was transmitted from one generation to the next, differing only in the materials responsible for the color. This indicated a higher variability among the compounds studied, suggesting the use of different networks and means for supplying the raw materials.

From the point of view of color as a technological and social phenomenon, which reflects cultural choices of the artisans, we were able to propose four technological markers of a Teotihuacan technological style. These markers are shared among all colors in the palette studied and detected in other apartment compounds. The continued use of the same formulations reinforces the notion that the production of pigments in Teotihuacan derived from a well-established production system from, at least, the Tlamimilolpa phase (200–350 CE), similar to the bone, pottery, and lapidary industries.

On the other hand, the conception of color as a technological object embedded in the urban economic system allowed an assessment of the color technology in a synchronic and diachronic way, by comparing our results with those of previous reports.

The findings stress the need for more in-depth studies from a technology perspective on Teotihuacan color materials. New data on the color production system is necessary to better understand the mechanisms for the acquisition of raw materials among the different apartment compounds, locate permanent storage and production areas and extend the research towards other architectural contexts and a wider time frame. By doing so, it would be possible to confirm our proposal of a distinctive technological style and its continuity through time, and of different ways for obtaining the raw materials.