Скачать 143.12 Kb.
The Technology of Eighteenth- and Nineteenth-Century Red Lake Pigments
Jo Kirby, Marika Spring and Catherine Higgitt
The eighteenth and nineteenth centuries saw a huge expansion in the use of and demand for natural dyes, in parallel with considerable technological developments in the textile industry. The stimulus for this had its roots in the seventeenth century, with the import of printed cotton fabrics from India. Locally produced cotton calicoes, indiennes, based on the Indian models, became hugely popular across Europe, and although this prompted some resistance and legal restrictions in favour of the wool and silk industries in some countries, in others, notably the Netherlands, cotton dyeing and printing became an important industry. By the late eighteenth century, with the relaxation of the legal restraints, processes for producing textiles were becoming industrialised and very much more complex. In England, for example, the production of cotton cloths became mechanised from the late 1770s, which enabled a very much more efficient production of the final fabric. It is thus no surprise to learn that England had become the dominant producer of printed cottons by the early nineteenth century (note 1). At the same time, the growth in the textile industry was one factor stimulating the development of methods to satisfy the growing need for alkalis and other chemicals: that is, the beginnings of the modern chemical industry (note 2).
The need to meet the ever-increasing demand for printed cottons necessarily led to a greatly increased demand for dyestuffs, notably madder, Rubia tinctorum L., which could give a bright fast red on cotton. Another stimulus for the cultivation of madder was the desire to find an alternative source of red to the heavily used New World cochineal insect, Dactylopius coccus Costa, the import of which was a highly lucrative Spanish monopoly (note 3). Only in 1777 were some of the precious insects obtained by subterfuge by the Frenchman Nicolas Thierry de Menonville, who tried to establish them in Santo Domingo (now Haiti). This experiment was short-lived, but later attempts were made to establish plantations of the nopal cactus and the insects in Spain itself and in southern France. Colonies were later established in other parts of the world, including the Canary Islands (by Spain), Algeria (by France) and Java (by the Dutch, using Spanish stock obtained from Cádiz by industrial espionage). In the early decades of the nineteenth century, other Central American countries such as Peru, Guatemala, Bolivia and Chile also took advantage of the crumbling state of the Spanish empire and developed their own cochineal export industries (note 4).
The search for a method to reproduce the bright red madder colour seen on imported fabrics, the so-called Turkey red (plate 1), and the need to produce dyestuff in such a form that efficient printing could be carried out stimulated research into methods of processing the madder root for efficient dye extraction and the chemistry of the madder dye. The colouring matter of the cochineal insect was studied for similar reasons.
These developments in the understanding of the chemistry of the dyestuffs inevitably had an influence on the production methods for red lake pigments where the colouring matter originated from natural dyestuffs. Not only might the dyestuff itself be processed differently, but also the methods and some of the ingredients used to prepare the pigment were slightly different to those used earlier. Traditionally the alkali used had often been a solution of potassium carbonate, potash, made from wood ash. By the late eighteenth century, the need for alkali had reached a level where it could no longer be met from plant sources: a synthetic process was thought necessary. The Leblanc process, whereby sodium carbonate was obtained from common salt by treating it with sulphuric acid and roasting the sodium sulphate formed with limestone and coal, was devised in response to a competition set by the French government in 1775 and won by Nicolas Leblanc in 1790 (note 5). The process became standard by the mid-nineteenth century and this is reflected in frequent references to soda or sodium salts as ingredients in nineteenth-century recipes for lake pigments, rather than the potassium salts of the eighteenth century.
Changes in the recipes might also lead to differences in the composition of the substrate of red lakes. For example, it has been found that the preparation of amorphous hydrated alumina by the addition of alkali to alum, as became common practice by the late eighteenth century, rather than by the addition of alum to an alkaline solution of the dyestuff, the general practice earlier, results in the incorporation of sulphate ions brought down during the precipitation. The result is a substrate rather similar to the light alumina hydrate described in modern paint technology literature (note 6).
Examination of red lake pigments used in eighteenth- and nineteenth-century paintings in the National Gallery Collection, presented in Tables 1a and 1b, confirms the influence these developments had on their composition. The paintings studied for this survey were all examined during previous conservation treatment or for the preparation of National Gallery Schools catalogues. For the present investigation, paint samples were examined by optical microscopy, energy dispersive X-ray microanalysis in the scanning electron microscope (SEM–EDX), and Fourier transform infrared (FTIR) microscopy and spectroscopy. High-performance liquid chromatography (HPLC) was used to analyse the dyestuffs wherever possible. The results show the wide use of cochineal and madder to prepare pigments, as is perhaps to be expected since they were so heavily used and so intensively researched. Other dyestuffs used in earlier periods for pigment preparation, such as brazilwood, Caesalpinia spp., and lac dye, produced, together with the resin-like shellac, by the Indian lac insect Kerria lacca (Kerr) and other species, were still in use, but they seem to have had a limited application in the preparation of pigments for artists. This is also true of the new, brilliantly coloured, so-called coal tar dyes, although these came to be widely used for other purposes.
The two principal regions of Central and South America where cochineal was to be found were the Peruvian Andes and Mexico – New Spain. The active cultivation of the insect for dyeing and as a source of pigment was already well established in Mexico when the Spanish arrived and this country was the principal source of the cochineal available in Europe until the end of the eighteenth century. Usable quantities probably first arrived in Spain in the 1520s; it was used for dyeing in Florence in 1542, Venice in 1543 and is said to have been in demand in Antwerp by 1550 (note 7). It was available in London by the 1560s.
Cochineal was an unusual commodity in that both farmed – fina – and wild – silvestre – varieties were used (the smaller wild insects making up the silvestre product perhaps including other local Dactylopius species), but the fina form was that destined for export. Apart from precious metals, cochineal was one of the most valuable and substantial export commodities. Quantities exported were very variable; this was a labour-intensive industry and could be affected by factors such as poor harvest of other crops and adverse climatic conditions, thus some periods were very lean. At other times, however, amounts were substantial (note 8).
During the sixteenth century, when the New World dyestuff first came into use, and for many decades subsequently, cochineal lake pigments were prepared by the method used for lakes containing kermes and Old World cochineal dyes: that is, by extracting the dyestuff from dyed textile shearings using alkali and adding alum to precipitate the dyestuff on a substrate of amorphous hydrated alumina. These lakes have been identified in several paintings in the National Gallery Collection, including Joachim Beuckelaer’s The Four Elements: Air, A Poultry Market with the Prodigal Son in the Background (NG 6587), painted in Antwerp in 1570, and Eustache Le Sueur’s Alexander and his Doctor (NG 6576), of about 1648–9. In both these cases, the dyestuff was probably extracted from silk shearings: examination by HPLC revealed the presence of a little ellagic acid, probably derived from galls used to weight the silk before dyeing (note 9).
At the end of the sixteenth century and beginning of the seventeenth, three developments occurred which changed the method of making cochineal pigments fundamentally. In 1612, Antonio Neri published an influential treatise on glass making, L’arte vetraria, which also contained some pigment recipes. Two of these were for making lake pigments from chermesi, a name that then signified crimson, and an Old or New World cochineal insect. In the first, the insects were used to dye wool, but the dyestuff was then re-extracted with alkali and precipitated with alum. The second, a method invented by Neri in Pisa, was more significant: the dyestuff was extracted from the insects with alcohol and an aqueous solution of alum was added to precipitate the pigment; the dyed textile stage was by-passed (note 10).
The second development was the discovery of a method of precipitating the insect dyestuff, predominantly carminic acid, which is present in the insect as the potassium compound (note 11), in the form of a metal complex or salt. The resulting pigment lacks the high proportion of hydrated alumina substrate present in a conventional cochineal lake and became known as carmine. The preparation is feasible because the amount of dyestuff contained by the Mexican cochineal insect is as much as 19 or 20% by weight (note 12), about ten times that contained by the kermes insect, Kermes vermilio Planchon, and considerably more than the Old World sources of carminic acid, Porphyrophora spp.. The discovery was, apparently, a matter of chance. The story is told that, in the 1580s, a Franciscan monk in Pisa making a pharmaceutical preparation with an extract of cochineal and cream of tartar (potassium hydrogen tartrate) inadvertently added acid, making the remedy unusable. It was set aside and after a few days a red precipitate formed. The story continues that a painter found this served well as a pigment and requested that it be made again (note 13). A recipe for the preparation of cochineal carmine, boiling the cochineal with a little kermes, alder (bark, perhaps) and potash alum, filtering off the liquid and leaving it to stand for the carmine to settle out, is said to have been published in 1656 (Table 2) (note 14). Whether or not it is coincidental that both these advances took place in Pisa within about twenty years of one another (Neri was working in Pisa between 1601 and 1603), it is the case that recipes for cochineal (and kermes) pigments quite unlike those of earlier times begin to appear. A mid-seventeenth-century Italian source, for example, includes a direct method, dissolving the dyestuff in alkali and precipitating the lake with alum; another consists of adding alum to a solution of the dyestuff in an extract of fennel seed. ‘Carmine colour’ is made by adding arsenic crystals to an alkaline solution of the dyestuff (note 15).
By the end of the seventeenth century, carmine is recorded as being used for miniature painting; it was generally too expensive for easel painting (note 16). It was made by boiling the ground cochineal in water previously boiled with the astringent seeds of chouan, Anabasis tamariscifolia L., adding a little potash alum, allowing the insect matter to settle out, decanting off the red liquid and leaving it to stand until the carmine settled out. As Table 2 shows, there were various modifications of this method, which is said to have first appeared in 1656 and continued to be cited until the late nineteenth century (note 17). Apart from any astringent property or tannin content the plant matters might contain that might aid precipitation, some also contained yellow or orange dyestuff, which would give a less blue crimson product. During the eighteenth century, other, more predictable methods, appeared. Some involved the addition of alum and cream of tartar, or alum alone, to the boiling cochineal mixture. The earliest examples seem to be found in mid-eighteenth- century German chemistry books and other sources (note 18); the method with alum is in fact described in later French sources as ‘the German method’.
The third development affecting the preparation of cochineal pigments was the introduction of tin salts in the mordanting and dyeing of wool and silk with cochineal; this resulted in a brilliant scarlet colour. Early in the seventeenth century, Cornelius Drebbel discovered that a solution of tin in aqua regia turned an aqueous solution of cochineal dye scarlet. The use of tin salts with cochineal to dye wool and silk scarlet was developed by his sons-in-law, the Kuffler brothers, in their London dyeworks at Bow and spread across Europe by the 1640s (note 19). At some time during the first half of the eighteenth century, the same tin salts were used to precipitate carmine from cochineal solution; this was more scarlet than the crimson-coloured carmine produced by the use of alum and cream of tartar (note 20). The English apothecary and chemist Robert Dossie observed in 1758 that there was a similarity between the preparation of carmine and dyeing, and earlier authors had wondered about a similar connection (note 21). In parallel with this, authors writing on dyeing commented on the tendency of the salts formed by dissolving tin in aqua regia (giving tin[IV] chloride) or nitric acid (giving soluble tin nitrates with cold dilute acid, a form of tin[IV] oxide with concentrated acid) to precipitate in the dye bath, forming a sort of lake (note 22). It thus seems highly likely that the tin-containing cochineals are linked with the technology of the dye bath. In later French recipe collections, the name ‘Chinese’ was associated with the tin[IV] chloride method (note 23).
Carmines were very slow to precipitate and several recipes, first appearing in the late eighteenth century, suggest the addition of a solution of fish glue or beaten egg white in water to help bring it down; these recipes also generally use a little alkali with the alum to precipitate the carmine (see Table 2) (note 24). Pierre-Joseph Pelletier and Joseph-Bienaimé Caventou, who isolated the colouring matter of cochineal in 1818, suggested that the use of a little alkali helped to dissolve out ‘animal matter’, the nature of which they were unable to clarify, associated with the dye; they thought the presence of the animal matter improved the yield of carmine obtained by the addition of acids or acid salts (including alum and cream of tartar). Inevitably, however, the reaction between alkali and alum would bring down the dyestuff in a form more like a deeply coloured hydrated alumina-containing lake than a ‘true’ carmine. In fact, Pelletier and Caventou discovered that many pigments sold as carmines by Paris colour merchants contained alumina (they also discovered that some contained vermilion) (note 25). Nevertheless, this method was one of those tried by J.F.L. Mérimée, who cited Pelletier and Caventou’s work as his source, and modifications of it were used by others (note 26). A sample of carmine from a collection of pigments from J.M.W. Turner’s studio, now in the possession of Tate Britain, London, was found to be of this type; under the microscope it had the appearance of a very deeply coloured purplish-red lake pigment of relatively small particle size (note 27).
The methods of preparation were sometimes misunderstood. For example, to make carmine the editors of the Encyclopédie raisonnée recommended a recipe from Johann Kunckel’s 1679 German edition of Neri’s work: Neri’s recipe for preparing cochineal lake by dyeing wool and re-extracting the colour. This led Robert Dossie to comment:
The compilers of the new French Cyclopedia have given two or three old recipes for the preparation of this colour; and afterwards recommended another; which on examination is only a process for making bad lake of scarlet rags: but rather than to insert such imperfect instructions for the making of an article of great consequence ... I chuse to be silent, and acknowledge my own ignorance in this particular ... (note 28)
While Dossie accepted that the best, most durable, carmine was prepared in France, he disagreed with the suggestion that this was due to ‘some qualities in the air and water’. In fact, up to the end of the nineteenth century, most sources gave the impression that the making of carmine was an art, requiring the best cochineal, iron-free alum, the purest water, the cleanest vessels of porcelain, glass or well-tinned copper and a sunny day: Philippe de la Hire commented that it was impossible to make in cold weather as it would not precipitate and the liquid gelled and went bad (note 29). Over a century later, an Austrian author commented that it was impossible to obtain a good product in England because of the weather (note 30).
It is clear that a good carmine was not easy to make. The early nineteenth-century English colourmaker George Field (1777–1854), in notes on experiments carried out on the permanence of pigments – his own and many made by other London colourmen – during the first quarter of the nineteenth century, described a carmine made by a Mr R. Hancock Senior in 1792 as ‘the most successful of his attempts, more red than carmine’, which suggests that other carmines were less successful (note 31). Study of the surviving archives of English colourmakers such as Lewis Berger and Company Ltd and Winsor & Newton shows that they devoted a great deal of attention to the making of carmines, with many trials (note 32). Lewis Berger was making carmines by the 1770s (the early ones appear to have contained vermilion, clearly a fairly common practice) and these, together with cochineal lakes, remained an important part of his company’s range throughout the nineteenth century. They supplied other colourmakers; Winsor & Newton studied the Berger carmines quite closely (note 33), and the recipes they used were similar. Over this period both companies used versions of the recipes discussed above: alum and cream of tartar, or, later, citric acid; alum with a mild alkali (borax, sodium borate); the use of milk, or other sources of protein (as in Table 2) to improve the yield. Berger used citric acid and milk from the 1830s to produce so-called Orient Carmine, a colour also made by Winsor & Newton, and tried the addition of a little ‘Nitro-Muriate’ of tin (tin[IV] chloride) (note 34). George Field, with whom Henry Newton (1805–1882) had worked, tried tin salts with cochineal to obtain a scarlet colour and Henry Newton himself, writing in 1840, noted that the addition of hydrated tin oxide to carmine (made by one of the other methods) increased the weight of the product without much affecting the colour. He also speculated that contemporary French carmines might contain a tin salt as they were much heavier than English carmines. Tin oxide had the added advantage of apparently protecting the pigments against mould formation (note 35). Winsor & Newton did make a tin-containing ‘French Carmine’ (note 36), but neither Berger nor Winsor & Newton made or sold the number of carmines produced by French colourmakers in the mid-nineteenth century: Jules Lefort wrote in 1855 that there were at least twelve, the number given in a slightly later catalogue of the Paris colour merchants Lefranc et Cie, at prices ranging from 120 francs a kilogram for the finest, Nacarat carmine, to 30 francs a kilogram for the lowest, grade 6, as ‘couleurs en nature’ (note 37).
As Pelletier and Caventou discovered, the methods of preparation involve making the pH of the cochineal extract slightly acidic – this includes Madame Cenette’s method using oxalic acid (see Table 2) (note 38). Carmines prepared in the laboratory following recipes of the type listed in Table 2, using alum and cream of tartar, or alum alone, were precipitated at a pH of about 5, little different to that of an aqueous solution of carminic acid. The yield from carmine preparations was very low, which probably explains the high proportion of insects used, far higher than would be needed for a conventional lake. Comparison of the proportions of the ingredients, shown in Table 2, suggests that a slightly higher amount of the insects (compared to that of alum) was used in the alkali/alum methods than in the alum/ cream of tartar methods (Wood’s method of 1856 is unlike the others in this respect), and considerably more insects with the tin salts methods, although the reason for this is unclear. Tin[IV] chloride gave a much lower pH of about 2 and, as discussed above, a more scarlet carmine. Carminic acid shows marked changes in colour with pH, corresponding to different degrees of proton dissociation in the molecule (note 39). The most marked change takes place over the pH region 4.8–6, for which it can be used as an indicator; here it shows a colour change from orange to crimson (mono-anion to di-anion conversion). Formation of scarlet tin-containing pigments takes place on the acid side of this pH range; aluminium-containing, more crimson carmines form more or less within it. Alumina-containing lakes tend to form on the alkaline side of this range and are a bluer crimson; indeed, as Pelletier and Cavantou observed, they can very easily become purple, particularly if heated (note 40). As well as giving a more scarlet pigment, carmine recipes involving tin salts tend to yield greater amounts of pigment, largely because aqueous solutions of tin[IV] salts readily hydrolyse to an amorphous white hydrous tin[IV] oxide which precipitates (note 41), bringing the dyestuff with it to produce a form of lake. Some eighteenth- and nineteenth-century recipes for cochineal lakes in fact consisted of adding tin salts to the cochineal solution to bring down most of the dyestuff and others involved adding a freshly made precipitate of a form of hydrated tin[IV] oxide to the dyestuff solution. These are less translucent than lakes on a substrate of hydrated alumina.
Modern carmine pigments are carminic acid complexes/salts of aluminium and calcium, and accounts in the modern pigment literature refer to this variety (note 42). However, the precise structure and composition are unclear. Bonding between the metal ion (Al, Ca or Sn) and the dyestuff molecule is likely to involve both salt and chelate formation (the latter by way of one of the quinone oxygens and the adjacent hydroxyl group). The presence of the carboxylic acid group in carminic acid means that interaction between the metal ion and carboxylate function is also possible, as well as hydrogen bonding. The IR spectra of a variety of carmine pigments (Al-, Al/Caor Sn-containing), while showing close similarities to free carminic acid, demonstrate subtle variations in band intensity and position which may throw light on the bonding interactions (note 43).
The appearance of a pure carmine under the microscope is very different from that of a conventional cochineal lake on a substrate of hydrated alumina, which has irregularly sized translucent cherry-red or pink particles (plate 2a). The carmines made in the laboratory have a very fine particle size and an extremely intense colour (plates 2b and 2c); their tinting strength would be very high and the seventeenth- or eighteenth-century painter would have found them very much stronger in colour than any other pigment then available, apart from the blue pigments indigo and, later, Prussian blue. In watercolour, carmine would be easy and economical to use as a wash; in oil, apart from being very expensive, it would be very much easier to use with the addition of a translucent extender. This would reduce the intensity of the colour and bring it closer in its properties to other pigments: simply adding lead white would lighten the colour, but increase its opacity, which was not necessarily desirable. One can imagine that the amount of translucent extender added might determine whether the pigment would be sold as a carmine or a lake.
As only a limited amount of the dyestuff precipitated during a carmine preparation, lakes could be made from the colouring matter left in solution by adding tin salts as described above, or by adding freshly prepared alumina (the same methods are described for preparation of lake pigments even in those cases where no carmine was prepared). Alternatively, alum could be added to the dyestuff solution and alkali added to precipitate the pigment, or vice versa, or even some combination of these methods might be used. If tin salts were used as the precipitating agent, alumina, chalk, china clay or starch, or any combination of these, was often added (note 44). In many cases, the resulting lake pigment could be described as a carmine with a great deal of extender.
In practice, during examination of the samples of cochineal-containing pigments listed in Tables 1a and 1b it was not always easy to make a distinction between a carmine and a lake pigment: in reality a continuum seems to have existed. Furthermore, in addition to the difficulty in deciding whereabouts on the carmine–cochineal lake continuum the pigment under investigation might lie, there is the question of under what name – carmine or lake – the pigment might have been sold and this is almost impossible to answer.
All the samples from eighteenth-century paintings examined during this study contained cochineal dyestuff (confirmed by HPLC or FTIR spectroscopy) (note 45). In some cases, the cochineal pigments were clearly conventional cochineal lakes very similar to those found in previous centuries. In a sample from the deep red chair back in Drouais’s Le Comte de Vaudreuil (NG 4253) of 1758, SEM–EDX examination revealed the presence of aluminium, and small amounts of other elements such as potassium, sulphur, phosphorus, silicon, and calcium, which can originate from the scale insect dyestuff or the ingredients of the substrate (see Table 1a). FTIR microscopy confirmed the presence of a conventional hydrated alumina substrate. Pigments of this type continued to be made into the nineteenth century as can be seen from the pigment used for the executioner’s red tights in Paul Delaroche’s The Execution of Lady Jane Grey (NG 1909), dated 1833. However, in a number of the eighteenth- century works examined, such as An Allegory with Venus and Time (NG 6387) painted by Giovanni Battista Tiepolo in 1754–8 (plate 3), the high sulphur content of the lake pigment suggests a different method of manufacture to that used in previous centuries (plate 4). In these cases the substrate is closer in composition to the light alumina hydrate of the nineteenth century (note 46).
Although of similar appearance to the lakes described above by optical microscopy, EDX analysis showed that the pigment used for the red coat in Lord Heathfield of Gibraltar (NG 111), painted by Sir Joshua Reynolds in 1787, contained little aluminium. Bands are seen in the IR spectrum suggestive of the presence of a pigment rich in cochineal dyestuff, which may be interpreted as a carmine, but this is complicated by the presence of bands at c.1650 and 1550 cm-1 suggesting the presence of a significant amount of protein (it should be noted that the binder is oil). The presence of cochineal dyestuff extracted from wool shearings is unlikely; little or no sulphur is present (note 47). It is more likely that a source of protein such as egg white or gelatine was added during the preparation of the pigment to aid precipitation: Table 2 shows several examples of this type of recipe. Several other similar examples were seen in paintings by Reynolds and Gainsborough, although these contained higher proportions of aluminium-based substrate.
Rounded, translucent inclusions with bluish fluorescence under ultraviolet (UV) illumination were observed in a large number of the cochineal-containing pigments examined, including three eighteenth-century examples (see Tables 1a and 1b). One of these is seen in the red cloak of the servant in Giovanni Antonio Pellegrini’s Rebecca at the Well (NG 6332), painted 1708–13 (plate 5). The strong broad band at c.3400–3300 cm-1 and the series of strong bands at 1154, 1083, 1048 and 1023 cm-1 seen in the IR spectra of these inclusions are typical of starch and in the majority of cases they show a characteristic cross shape under cross polars (plate 6) (note 48). Starch was also mixed with the majority of the examples of nineteenth- century cochineal-based pigments examined (see Table 1b). All those analysed were on tin-containing substrates (sometimes also containing minor amounts of aluminium and/or calcium) and consisted of small, strongly coloured cherry-red particles. An example from Pierre-Auguste Renoir’s A Nymph by a Stream (NG 5982) of 1869–70 (plate 7) is seen in plates 8a and 8b; the FTIR spectrum is shown in FIG. 1.
In one pigment, used for a deep red flower in The Painter’s Garden at Saint-Privé (NG 1358) painted by Henri-Joseph Harpignies in 1886, a different extender is present. Here the cochineal pigment, which appears from IR analysis to be a carmine, is mixed with colourless particles of chalk. Chalk was also found associated with the cochineal pigment in Monet’s Lavacourt under Snow (NG 3262), painted about 1878–81. The unusual large rounded particle form suggests that the chalk was artificially precipitated and dyestuff can be seen clinging to the colourless particles.
|Indian Science and Technology in the Eighteenth Century||An ordinance of spring lake heights borough, county of monmouth, state of new jersey enacting chapter 22-530 "stormwater control"," to the borough of spring lake heights land use regulations|
|Edith and I were born in the prior century-the nineteenth century. Things were quite a bit different than they are today. Kerosene lamps lit our homes. Our heat||The Restoration and the Eighteenth Century|
|Eighteenth century states||Nineteenth Century Legal Treatises|
|Nineteenth Century Legal Treatises||Nineteenth Century Legal Treatises|
|Archaeology of an Eighteenth-Century Indian Village in North Carolina||Archaeology of an Eighteenth-Century Indian Village in North Carolina|