Abstract
This paper focuses on describe the effect of the physical phase of linseed oil over the color and brightness of natural lakes manufactured with cochineal and Brazilwood. The lakes were made from scratch following existing recipes for the powder pigments and lately mixed with two different types of linseed oil: thickened boiled and cold-pressed raw. In order to investigate properties such as deposition on paper, color, and brightness, we measured their rheological properties to establish the relationship between the physical me- chanical properties of the lake pigments and the final results in a painting of each formula. Many pigments used in the food industry, pharmaceuticals, and even those in contact with animal consumption products have a synthetic origin. Therefore, understanding the process by which natural pigments are synthesized can help reduce health risks. This represents a promising area of research.
Keywords:Lake pigments; Cochineal; Brazilwood; Rheology
Introduction
Pigments are the core of a painting and a predominant material in the art and restoration industries, among others. By definition a pigment is a fine milled powder with a particle size between 1 and 0.3 micrometers, must be insoluble in water and can be used to paint when mixed with an adhesive media Eastaugh and Walsh [1], however, nowadays nano-pigments are the rage of modern pigment technology, such as vantablack or thin layer UV coatings. But what makes a pigment suitable for use? That depends effectively on its physical properties, being one of the most important, which will help to determine color, durability, toxicity, brightness, and texture.
In the history of pigments, those that produce brilliant colors tend to capture the interest of not only artists but also merchants [2] that propelled the pigment industry to find richer shades from natural sources and also artificial means. Egypt was historically the first place where artificial lake pigments made from organic dyes were formulated [3]. However, the exact origin and recipes for organic lake pigments remain unclear [4-5].
A lake pigment is an artificial pigment made by coloring a white inert powder using a dye with the help of a fixative [2]. Being translucent in nature and mixed with oil, these pigments have been used to obtain the final touches of light and details, for example, drapes on textiles in many paintings [6] or folds, illumination or light effects in historical manuscripts [7-8]. Among colors, red has been of great importance in many cultures, depicting status and power in part due to the high cost of dying clothes [6]. The main organic dyes to obtain red were obtained from an insect called ”Kermes” Kermes vermillio and a plant called “Madder” Rubia tinctorum, however the over exploit of these natural sources made them almost extinct and therefore difficult to attain [3]. After America’s discovery two new sources of red dyes rose, cochineal and Brazilwood and during the nineteenth century became especially popular between the impressionist and pot impressionist painters [3].
Nowadays cochineal Coccus cacti and Brazilwood Paubrasilia equinata are two of the most popular and important dyes, the chemical molecule responsible for the color is carminic acid and brazilin, these used not only in textile and art production, but also in the color, food and cosmetic industries [9-14]. Numerous analysis of cochineal have been performed as a food deterrent to ants [12], food pigment [10-13], and consequently there are improvements in the extraction process [11]. Similar evaluations have been done for the Brazilwood [9], as being very resistant to decay by fungi and tested to repel the attack of dry wood termites [15] or used as the main material for violin bow production due to its unique vibrational properties found in wood [16].
The final color of the lake pigments depends on the chemical reaction between the inert powder and the metallic salt used to synthesize the compound; therefore, the range of color in these particular cases oscillates between mild orange, pink, purple, and crimson red, despite having two different origins, an insect and a plant [1-4]. To elaborate the lake pigments, this paper uses techniques attributed to the alchemical practice on pigment development found in Mappae Clavicula [2] as well as ancestral recipes used in textile dying [17] and old traditional painting recipes [18].
In this report, we present the alchemical process we followed to obtain four different lake pigments made from cochineal and Brazilwood. Afterward, a study of the rheological properties was performed to characterize the behavior of the four lake pigments in addition to two different linseed oils: boiled thickened and raw cold pressed used as agglutinate medium. Through this particular study, we want to establish a relation between the rheology of our lake pigments and the behavior as paint over paper, the final color, and the brightness.
Materials and Methods: Synthesis of pigments
The elaboration of the red lake pigments was done, as discussed previously, using already existing traditional recipes with the addition of novel materials, and the processed appeared to be the tradition of Mexican textile dying [17-18].
To extract the dye from our sources, we need to follow a simple set of instructions. First, both cochineal and Brazilwood must be reduced to a minimal size, the cochineal is milled finely, while the Brazilwood is reduced to chips. Secondly, they are boiled with 1000 ± 20 ml of distilled water in glass beakers for at least half an hour until the dye is released. After the solution has cooled down, it is important to filter using a cotton cloth to prevent small particles that could affect the final color and consistence of the lake [18].
To achieve natural red pigments, two types of powders were used: potassium carbonate (K2CO3) and calcium carbonate (CaCO3). Alum was used to fix the lake pigment, which is one of the preferred mordants in many of the Mexican recipes for textile dying. Lake pigments typically use a core base of white mineral powder (such as clay or calcite), which is dyed using a mordant substance (alum, iron, copper, or other metallic salts) that acts as a fixative of the dye on the powder [3]. The common name for the aluminum potassium sulfate compound is [Kal (SO4)2·12H2O], however is also called tlalxocotl and is normally found in its crystallized form [17]. The dyes were made from ”silver cochineal”, an insect that is from the D. opuntiae subespecies found mainly in central Mexico (Queretaro, Mexico state, Mexicocity, Morelos and Jalisco). The insect must be dried in the sun for a few weeks before use [8-19]. The Brazilwood wood chips Paubrasilia equinata were obtained directly from a natural three from Viζosa, Brazil.
Lake Pigments with Potassium and Calcium Carbona
As mentioned above, these two powders were chosen to be dyed to make the final lake pigment. The process is the same in both cases, once the dye is filtered, each powder is placed separately inside the beaker with 500 ± 10 ml of the dye in each case while being stirred and heated. Alum is slowly added, making an effervescence reaction that will stop until the solution reaches a pH scale of 7, measured with the help of Hydrion pH strips, meaning that the solution has reached a balance and is a neutral substance. When the powder begins to decant leaving clear water on top, it is time to start the filtration process of the final solution and let it dry naturally for three to four days.
Lake pigments made with potassium carbonate (K2CO3) will end up darker than those made with calcium carbonate (CaCO3). For Brazilwood, the potassium carbonate lake will be a dark brownish red color, while the calcium carbonate lake will be pink with orange undertones. For cochineal lake pigments, the final powder of potassium carbonate will be a darker magenta color and the calcium carbonate a pale pink color.
Results
The pigments obtained with Cochineal and Brazilwood prepared with CaCO3 and K2CO3 were mixed separately with linseed oil (also known as flax oil Linum usitatissimum) in two physical states or phases: raw cold pressed and thickened boiled (also called spessito). This oil was chsosen due to its popular use in oil painting techniques [2] and also because linseed oil has an unique property of polymerize faster than other oils (around 800C). The behavior of oil change according to the temperature. When is boiled changing its structure [20] from the raw version giving the final texture of a thickened oil. Also there is a notable difference of its amber color [20-21] which allow us to analyze the lakes without changing the kind of oil in all the cases.
A total of eight lakes were made with linseed oil: four with the raw state and four with the boiled state. The raw linseed oil was from the commercial brand Atl, while the boiled was from the Winsor and Newton brand. In all cases 2.5 g of fine powdered pigment mixed with 5 ml of linseed oil were used (Figure 1).
Raw oil tends to dry faster; meanwhile, boiled oil took more than a few days to settle down and dry. The powder and oil were mixed with the help of an agate mortar until it acquires a uniform texture and color. After drying on paper, the samples made with the raw oil acquire a ”matte” final brightness. However, boiled oilprepared lakes obtain a ”satin” or “lustrous” finish.

Rheology Characterization of the Red Lakes
Steady shear flow tests of Brazilwood lakes and cochineal lakes with K2CO3 and CaCO3 in raw and boiled linseed oil were measured with the help of a Discovery Hybrid Rheometer (DHR-3, TA instruments) at room temperature (250C)) and humidity 40%. All tests were carried out with plate-plate geometry with shear rate deformation of 0.01-300 1/s. The temperature was controlled with a Peltier system (0.10C).
Figure 2 shows the shear stress τ plotted against the shear rate γ˙for the Brazilwood lakes. In the figure, the pure linseed oil, raw and boiled, i.e. without any pigment, is plotted to have a viscosity reference. The results are fitted to the well-known power law model τ = μγ˙n [22], where τ is the shear stress, μ is the viscosity constant, γ˙ is the shear rate and n is the fitted power law index. The power law index for all test lakes is close to unity (n ∼ 1), which indicates that all test lakes behave like a Newtonian fluid characterized by constant viscosity and their viscosity function η can be considered to be η = μ.
From Figure 2 we obtained the next results: (i) the boiled oil increases the viscosity in all the Brazilwood lakes. This can be observed directly by the comparison between the solid symbols with the open symbols and; (ii) the nature of the carbonate powders used in the lake preparation affects directly on their viscosity, lakes prepared with calcium carbonate are more viscous than those obtained with potassium carbonate.
Figure 3 displays the plot of shear stress as a function of the shear rate for the cochineal lakes. The behaviour of the lakes with boiled oil reached a relatively high viscosity than those ones with raw oil, similar to what we observe on the Brazilwood lakes. Amid the preparation of the final lake pigments a granular consistency was observed, particularly in the cochineal lake made of potassium carbonate, this characteristic seems to affect the viscosity at low shear rate of deformation as this declined at increasing rates of deformation. On the figure are plotted the symbols united with lines which is the fit of the power law model τ = μγ˙n except for lakes with K2CO3.
From Figure 3 we can observe: (i) the boiled thickened linseed oil increases the viscosity while mixed with the lakes; (ii) the addition of the different kind of carbonate powders used in the lakes recipes influences directly their viscosity. However, the behavior for lakes with K2CO3 is not as predictable as the CaCO3 this mainly attributed to the size and physical properties of the powder pigment not being as smooth in this case.
(iii) the behavior of the mixture of raw linseed oil and lakes with CaCO3 can be adapted to the law model τ = μγ˙n; however, for lakes made with K2CO3 this can not be possible. Lakes with potassium carbonate present a decrease in viscosity through the shear rate increase. The Sisko model explains this particular behavior because it is considered as a generalized power law model that includes a Newtonian component. The general form of the model can be written as τ = μγ˙n + μ∞γ˙, where μ∞ is the infinite shear viscosity [23-24]. There are few references that support this idea and could be a line for future research. In summary, on Table 1 the viscosity of all the lakes is presented (Table 1).



Behaviour as paint, brightness and color of the lakes
The final test of this report consists of a simple prove of our lakes and the observation of the behavior while being used as a painting. Simple brush lines using No. 5 brushes made of gray squirrel fur were used to paint over the paper chosen due to its fabric-like surface texture to enhance the drying time. We focused on two main variables: is it easy to paint with this lake? and what is the change on the color before and after drying. The next photographs are presented to show the results, the raw (upper) and boiled (down) linseed oil in the fresh (left) and dry (right) state.
Figure 4 shows the brush strokes made with Brazilwood K2CO3 lake. In picture [1] the brush strokes made with raw linseed oil in the fresh state show a dark magenta color that tends to a wine red color in the dry state (photo [2]). In picture [3] the brush strokes made with boiled linseed oil in the fresh state show a sparkly dark red tone. The tone changes to a diffusive red in the dry state as shown in the picture [4]. Figure 5 shows the brush strokes made with the Brazilwood CaCO3 lake. In picture [1] the brush strokes with raw linseed oil in the fresh state show a pink purple color, tending to a pale pink color in the dry state (photo [2]). In picture [3] the brush strokes with fresh boiled linseed oil show a sparkly orange tone. The tone changes to a pale orange in the dry state as shown in the picture [4].



Figure 6 shows the brush strokes made with Cochineal K2CO3 lake. In picture [1] the brush strokes with raw linseed oil in the fresh state show a dark purple color, tending to a dark pink (“mexican pink”) in the dry state (photo [2]). In picture [3] the brush strokes made with boiled linseed oil in the fresh state show a sparkly red tone which tends to a dark red on the dry state shown in picture [4]. Figure 7 shows photographs of the Cochineal CaCOlake3. In picture [1] the color of the brush strokes with raw oil are pale “mexican pink” wich tend to a pale pink orange color (photograph [2]). In picture [3] one can see that brush strokes with boiled linseed oil on a sparkly red color tend to a red dark red on the dry state (photograph [4]).

As seen in the pictures, we have a difference in texture, brightness, and color between the two forms of lakes depending on the oil used; in general there is a significant change between the fresh and the dry forms. Regarding ease of application on our surface, lakes made with raw linseed oil and CaCO3 behave better and dry faster; also, these lakes have a smoother texture than the K2CO3, the increase in the viscosity of the boiled linseed oil formula makes it harder to apply on the surface. Still, after drying, the lakes with the raw formula appear with a matte finish, while the boiled formula tends to look brighter. Finally, regarding the color the K2CO3 formula produce darker tones between red and purple, while the CaCO3 produce lighter tones between pink shades. The Brazilwood lake with CaCO3 and boiled linseed oil turned out to be the optimal product for the paintings. The Cochineal lake with K2CO3 and boiled linseed oil turned out not to be optimal due to the behavior seen in the rheology analysis mainly due to the nature of the powder particles.
Conclusion and Remarks
We had a total of eight different lake pigments made of cochineal and Brazilian wood. The main variables that influence the viscosity of our final lake pigments are both the nature of the oil used and the size and form of the particles. The K2CO3 formula overall behavior, for both cases, is difficult to predict, on the other hand the CaCO3 formula behaves similarly in all four cases of lake pigment. This result will result in the way the final lake pigment can be used, the boiled oil formula is thicker and a little more difficult to use, still the CaCO3 formulas are more viscous than the K2CO3 in both cochineal and Brazilwood, the same can be observed in the raw linseed oil formulas. Regarding the final texture of the lakes after drying, the boiled thicker formulas appear with a softer and uniform finish, while the raw linseed oil formulas have a rougher appearance.
Regarding the final color and brightness, because the raw linseed oil is less viscous than its boiled counterpart, it dries faster, giving a matte finish and changing to paler colors afterwards, the boiled samples take up to a few days to dry completely, but the final appearance is a brighter satin finish, and the colors, as they also fade a little, do not change considerably from the original paint. In general the K2CO3 formulas are darker and harder than the CaCO3 formulas, leaving a range that goes from dark red, dark magenta to light orange pink and light pink. This results open up further investigation regarding the optical behavior of our formulas in the future.
Acknowledgement
M.W., A.B. and E.M.C.M thank the technical assistance M.A.Garcıa, M.A. Maynez and the equipment used in the Van de Graaff 0.7 MeV (Maria) at the Instituto de Fisica UNAM and the Sociedade de Investigacoes Florestais, Departamento de Engenharia Florestal, Universidad Federal de Viζosa (UFV) which donates.
dropdown menuReferences
- Eastaugh N, Walsh V, Siddall R, Chaplin C (2004) The Pigment Compendium: optical microscopy of historical pigments. Elsevier Butterworth-Heinemann, 1st edition.
- Merrifield MP (1854) Dress as a fine art. John P. Jewett and Co, 1st edition.
- Delamare F, Guineau B (2014) Colour, Thames and Hudson, 1st edition.
- Herbst W, Hunger K, Wilker G, Ohleier H, Rainer W (2004) Industrial Organic Pigments. Wiley, 3rd edition.
- Malkin E (2017) When europe’s art turned red. New York edition c pp.2.
- Bomford D, Ashok R (2009) A closer look colour. The National Gallery London, 1st edition.
- Klockenkamper R, Bohlen A, Moens L (2000) Analysis of pigments and inks on oil paintings and historical manuscripts using total reflection x-ray fluorescence spectrometry. X-Ray Spectrometryx 29: 119-129.
- Fernandez FM, Roque G, Useda E, Rangel MH, Casanova M, et al. (2017) ROJO MEXICANO La grana cochinilla en el arte. Instituto Nacional de Bellas Artes, 1st edition.
- Oliveira LFC, Edwards HGM, Velozo ES, Nesbitt M (2002) Vibrational spectroscopic study of brazilin and brazilein, the main constituents of brazilwood from brazil. Vibrational Spectroscopy 28: 243-249.
- Lloyd AG (1980) Extraction and chemistry of cochineal. Food Chemistry 5(1): 91-107.
- Borges ME, Tejera RL, Díaz L, Esparza P, Ibáñez E (2012) Natural dyes extraction from cochineal (Dactylopius coccus). New extraction methods. Food Chemistry 132(4): 1855-1860.
- Eisner T, Nowicki S, Goetz M, Meinwald J (1980) Red cochineal dye (carminic acid): Its role in nature. Science 208(4447): 1039-1042.
- Baranyovits FLC (1978) Cochineal carmine: an ancient dye with a modern role. Endeavour 2(2): 85-92.
- Arshid FM, Desai JN, Duff DJ, Giles CH, Jain SK, et al. (1954) A study of certain natural dyes I-the adsorption of brazilwood and logwood colouring matters by fibres. Coloration technology 70(9): 392-401.
- Silva CA, Monteiro MBB, Brazolin S, Lopez GAC, Richter A, et al. (2007) Biodeterioration of brazilwood Caesalpinia echinata Lam. (Leguminosae-Caesalpinioideae) by rot fungi and termites. International Biodeterioration and Biodegradation 60(4): 285-292.
- Matsunaga M, Sakai K, Kamitakahara H, Minato K, Nakatsubo F (2000) Vibrational property changes of spruce wood by impregnation with water-soluble extractives of pernambuco (guilandina echinata spreng) ii: structural analysis of extractive components. Journal of Wood Science 46: 257-353.
- Arroyo L (2014) Tintes Naturales Mexicanos. Universidad Nacional Autonoma de Mexico, 2nd edition.
- Verhecken A, Kirby J, Bommel M (2014) Natural Colorants for Dyeing and Lake Pigments practical recipes and their historical sources. Archetype publications, 1st edition.
- Chavez-Moreno CK, Tecante A, Fragoso-Serrano M, Pereda-Miranda R (2010) Metabolic profiling of dactylopius (hemiptera: Dactylopiidae) species pigments by geographical origin and hosts using multivariate data analysis. Biochemical Systematics and Ecology 38(4): 671-679.
- Lazzari M, Chiantore O (1999) Drying and oxidative degradation of linseed oil. Polymer Degradation and Stability 65(2): 303-313.
- Vitorino T, Melo MJ, Carlyle L, Otero V (2016) New insights into brazilwood lake pigments manufacture through the use of historically accurate reconstructions. Studies in Conservation 61(5): 255-273.
- Barnes HA, Hutton JF, Walters K (1989) An Introduction to Rheology, Elsevier Science BV 1th edition.
- Lupi FR, Gabriele D, Cindio B, Sanchez MC, Gallegos C (2011) A rheological analysis of structured water-in-olive oil emulsions. Journal of Food Engineering 107(3-4): 296-303.
- Nindo CI, Tanga J, Powers JR, Takhar PS (2007) Rheological properties of blueberry puree for processing applications. LWT - Food Science and Technology 40(2): 292-299.