Supervisors:
Dr. J. Dyer
Dr. D. Tamburini
Examiner:
Dhr. Prof. Dr. ing.
M. R. van Bommel
MSc Conservation and Restoration
of Cultural Heritage
Conservation Science
Master Thesis
HISTORICAL FORMULATIONS OF
LAC LAKE PIGMENTS AND DYES:
A STUDY OF THE
COMPOSITIONAL VARIABILITY
By
Sanne Berbers
10193030
May 2018
Department of Scientific Research at
the British Museum in London
Samenvatting: Historische recepten van lac pigmenten en
kleurstoffen, een studie naar de chemische samenstelling
mei 2018Sanne Berbers
Lac is een organische roze paarse kleurstof die onder andere gebruikt kan worden als textiel en leer verf of neergeslagen op een substraat als pigment. De belangrijkste kleurstof moleculen zijn laccaci acid A en B. De kleurstof wordt gewonnen uit de harsachtige secreties (stoklak) van de zogenaamde lakschildluis, (Kerria lacca kerr) een luisachtige die leeft op bomen in India, Cambodia en Thaiand. Van deze secreties wordt ook de hars schellak gemaakt. De stoklak bestaat uit kleurstoffen (4-8%) en schellak componenten (was 6-7% en hars 70-80%) om deze verschillende componenten van elkaar te scheiden wordt gebruikt gemaakt van de wateroplosbaarheid van de kleurstoffen en de onoplosbaarheid van de schellak componenten. Als gekeken wordt naar historische productie methodes is het aannemelijk dat een deel van de schellak componenten mee kwam in de kleurstof matrix tijdens de productie. Historische recepten geven aanwijzingen dat dit een bekend en ongewenst probleem was tijdens de productie van lak voor het verven van textiel. Er zijn echter ook aanwijzingen dat dit een gewenst fenomeen was bij de productie van pigmenten aangezien het kan bijdragen aan verfeigenschappen.
Om de mate van inclusie van schellak componenten in lak pigmenten en kleurstoffen te onderzoeken zijn test-monsters gemaakt. Hiervoor zijn eerst historische recepten afkomstig uit Europa bestudeerd, gedateerd van af het vroegst bekende recept uit de 9de eeuw tot aan recepten uit de 18de eeuw, waaruit de parameters met de grootste invloed op het productie proces werden gehaald, de pH en temperatuur. Aan de hand van deze informatie werden drie groepen monsters geproduceerd: i) pigmenten met kleurstoffen geëxtraheerd uit stoklak ii) textiel geverfd met kleurstoffen geëxtraheerd uit stoklak, en iii) pigmenten gemaakt door de extractie van textiel geverfd met lac. Het extraheren van kleurstoffen uit textiel voor de productie van pigmenten was veelvoorkomend in Europa van de 14de tot de 17de eeuw, en is daarom meegenomen binnen dit onderzoek.
Voor de analyse van de monsters is gebruik gemaakt van multispectrale fotografie en de analytische techniek HPLC-DAD-ESI-Q-ToF die recentelijk ook gebruikt is voor de chemische karakterisering van schellak. Aan de hand van deze resultaten kon geconcludeerd worden dat in het productie proces van lac altijd schellak componenten meekomen, zelfs als wordt gekeken naar de meest milde extractie condities namelijk kamer temperatuur en neutrale pH. De pigmenten die direct gemaakt waren van stoklak hadden de hoogste concentratie kleurstof moleculen en schellak componenten, variërend van vrije zuren tot penta-esters. De textielmonsters bevatten minder polyesters maar een grote verscheidenheid aan vrij zuren en diesters afkomstig uit schellak. Deze componenten werden voor het overgrote gedeelte overgedragen naar de pigmenten gemaakt uit de textiel. Hieruit werd geconcludeerd dat het onvermijdelijk is om tijdens de pigment of kleurstof productie schellak componenten mee te extraheren uit stoklak, tenzij gebruikt wordt gemaakt van moderne zuiveringstechnieken. Het voordeel van HPLC-MS als onderzoek methode ten opzichte van HPLC-DAD is met dit onderzoek geïllustreerd door de mogelijkheid van MS om ook niet of zwak gekleurde moleculen te kunnen detecteren die geen karakteristiek DAD spectrum hebben. De resultaten van dit onderzoek kunnen gebruikt worden bij de interpretatie van de analyse van een roze lake pigment gevonden op een hellenistisch terracotta oinochoe beeld (3de eeuw v. Chr.), waarin en laccaic acid A en schellak componenten waren gevonden.
Abstract: Historical formulations of lac lake pigments and dyes,
a study of the compositional variability
May 2018Sanne Berbers
Lac is an organic insect-derived pink-purple colourant which can be used as a dye for textiles and leather or precipitated onto an inert substrate as a lake pigment. The main colourant molecules are Laccaic Acids A and B. It originates from the resinous secretion of the oriental insect Kerria lacca kerr which is called sticklac, the same material from which shellac is made. A separation between the colourants (4-8%) and the shellac components (wax 6-7% and resin 70-80%) is generally obtained by an aqueous extraction of the water soluble laccaic acids while assuming the non-solubility of shellac in water. However, considering cruder historical production methods, it is plausible that shellac components are transferred into the colourant matrix, when attempting the separation. Historical recipes indicate that the inclusion of resinous components was a known unwelcome phenomenon in dyeing practices, whereas by contrast it might have been a desirable feature in pigments for manuscript illumination, as this afforded improved binding properties.
To study the significance of shellac inclusion within lac lake pigments and dyes, mock-up samples were made by reproducing a range of different historical recipes from Europe dating from 9th to 18th century. Particular attention was paid to some key parameters in pigment and dye production, such as temperature and pH. These parameters were used to create three different sample groups: i) pigments made with colourants extracted from sticklac, ii) textiles dyed with colourants extracted from sticklac, and iii) pigments made by extracting colourants from the dyed textiles. This last sample group was included as the production of lake pigments from textile clippings was a common practice in Europe from the 14th to the 17th century.
Multispectral imaging was carried out on all the samples which were also analysed by HPLC-DAD-ESI-Q-ToF, the latter being recently applied to the chemical characterisation of shellac. From these results it was concluded that shellac components were always present to some degree in the colourant matrix, even when considering the mildest colourant extraction conditions (room temperature and neutral pH). The pigments made directly from sticklac showed the highest amount of colourant molecules and the presence of shellac components, ranging from free acids to penta-esters. The textiles contained less polyesters but a wide range of free acids and diesters, which were transferred to a large extent into the pigments made from these textiles, due to the relative harsh extraction methods commonly used to extract pigments from textiles. From this it was concluded that the inclusion of shellac components within any lac pigment or dyed textile can be considered unavoidable unless more advanced refinement techniques are used. HPLC-MS has also shown its potential in comparison to HPLC-DAD due to its capability to detect non- or weakly coloured molecules that have no characteristic DAD spectrum. These data are finally useful to interpret the results obtained by analysing the pink lake pigment on a Hellenistic terracotta oinochoe (3rd century BC), where evidence of both laccaic acid A and shellac components was found.
Table of Contents
Samenvatting ... 2
Abstract ... 3
1. Introduction ... 5
2. Current scientific knowledge ... 8
3. Research Methodology ... 10
Historical recipes ... 10
Mock-up samples ... 13
Technical examination samples ... 15
4. Results and discussion ... 21
5. Concluding remarks ... 34
6. Acknowledgements ... 34
Bibliography ... 35
Appendix I: Historical recipes ... 39
Lac lake pigments ... 40
Dyeing with lac ... 44
Extracting lac from dyed textiles ... 48
Miscellaneous other useful recipes and references ... 51
Appendix II: Experimental information ... 52
Appendix III: Multispectral Images ... 55
Appendix IV: SEM images ... 62
1. Introduction
A recent study into pigment use on Hellenistic terracotta’s at the British museum led to the examination of a decorative oinochoe wine jug, on which, among others, lac pigment was identified.1 This is a remarkable find considering that there are no contemporary uses of this pigment known to this date within Europe, the earliest currently reported European find of lac can be attributed to 7th-8th century lac dyed Merovingian textiles.2 One of the colourant molecules (laccaic acid A) and shellac components were found using HPLC-DAD-ESI-Q-ToF. Conventional organic pigment analysis using DAD as a detection method is limited because of its capability of detecting only coloured molecules. The addition of the mass spectrometer allowed for identification of the non- or weakly coloured components, such as the ones that originate from shellac, in the colour matrix. This can be specifically crucial for a pigment such as lac where the colorant molecules are prone to degradation but other- non colourant- residual components from the original material might remain detectable. This is dependent on the degree in which shellac components are transferred into the colour matrix during the pigment or dye production. More insights into pigment preparation, degradation and the molecular composition could aid in the further identification and knowledge about lac use throughout history.
Sticklac is the resinous secretion of the oriental insect Kerria lacca kerr, indigenous to India, Thailand and Cambodia, from which lac and shellac are produced. Lac is comprised of a group of anthraquinone colourant molecules: laccaic A and B in major quanites, and laccaic acids C, D and E in minor quantities. The hydroxyanthraquinonees erythrolaccin and deoxyerythrolaccin are colorant molecules which are also present in sticklac, but are considered part of the shellac matrix as they have a reduced solubility in water. Shellac can be characterised as a complex mixture of monoesters, polyesters of hydroxyl aliphatic and sesquiterpenoid acids. The exact composition of the resin is dependent on environmental factors such as the insect’s life, tree species it was growing on, the tree’s environment. A separation between the colourants (4-8%) and the shellac components (wax 6-7% and resin 70-80%) is generally obtained by an aqueous extraction of the water soluble laccaic acids while assuming the non-solubility of shellac in water.3,4 However, considering cruder historical production methods, it is plausible that shellac components are transferred into the colourant matrix when attempting the separation. Historical recipes indicate that the inclusion of resinous components was a known unwelcome phenomenon in dyeing practices, whereas by contrast it might have been a desirable feature in pigments for manuscript illumination, as this afforded improved binding properties.5 The possible presence of shellac pigments has also been noted in several studies on organic lake pigments.6
This master thesis will study the degree in compositional variability of lac as a result of changes in key parameters in the production process. Through an extensive survey of historical lac dye and pigment recipes, ranging from the earliest recipe found dating from the 9th century to the beginning of industrialisation in the 18th century in Europe, fundamental recipes were derived which allowed the study of different production parameters. This was utilised to produce three different mock-up sample groups: i) pigments made with colourants extracted from sticklac, ii) textiles dyed with colourants extracted from sticklac, and iii) pigments made by extracting colourants from the dyed textiles. Analysis with, among other techniques, multispectral imaging and HPLC-DAD-ESI-Q-ToF was employed to investigate the relationship between the production method and the molecular composition of the resulting lac pigments and dyes.
The oinochoe vase which served as the starting point of this research is made from terracotta, has polychrome paint layer which is largely preserved and originates from the Hellenistic Greek period (323 – 31 BC) in Canosa di Puglia. A photographic reproduction of the vase can be found in Figure 1. Parts of the pink-purple paint layer remain present on part so the dress of the female figure standing on the top part of the vase. Analysis was done by J. Dyer and D. Tamburini at the British Museum on five samples taken from different areas of the dress.1 Microscopic analysis indicated a very underbound mixture of Egyptian blue and a pink lake physically mixed together, which was a common practice of the period.7 Analysis of the pink lake by HPLC-DAD-ESI-Q-ToF led to the identification of a mixture containing madder, coccid dye, shellac and tannin markers. From this the main colorant of the lake pigment was be identified as madder (Rubia spp.), the other markers led to the identification of insect-derived colourants from polish cochineal (Porphyrophora spp.) and lac (Kerria Lacca Kerr).
The markers for shellac included the colourant molecules erythrolaccin and deoxyerythrolaccin, and resinous components such as free acids and esters. The these colourant molecules have reduced solubility in water and their presence in lac pigment is considered to be indicative of a alkaline extraction of raw sticklac.6,8–10 The presence of resinous shellac components due to use as a consolidant during a conservation treatment was excluded by microscopic examination of the paint surface of the object. The markers for tannin can be indicative of a colourant which has been extracted from coloured textiles.6 The production of red lake pigments from coloured textiles was common in the 14th to 17th century throughout Europe, but has not been documented for Antiquity. The high value of colourants as a commodity does make it plausible that an effort might have been made to reuse colourants from clippings or worn down textiles. The presence of both shellac resin and tannins might be explained by the alkaline extraction of a textile source of lac, if the resinous component are transferred on a lac-dyed textile in the dyeing process and are then sequentially extracted with the colourant to form a pigment. It is known that resinous components can be transferred to textiles, historical recipes give warnings about the difficulty of dyeing with lac due to the high chances of resin transfer onto the textile.
It is often considered that lac was widely introduced into Europe with the invasion of the Iberian peninsula in the 8th century by the Umayyad Caliphate.10,11 This reinstated direct trading routes with the middle east which had been unreliable due to the turbulent period in Europe after the fall of the Roman Empire. An increase in archaeological evidence of extensive trading in recent years has created a shift towards the notion that these geographical limitations on trading were far less strict, and widespread trading with the middle east was continuous throughout history. Only limited number of texts survive from antiquity and information on craft practices was usually not deemed important enough to write about, instead it was disseminated through oral practice. However, circumstantial evidence can be found which supports the likely importation of lac from India to Greece during the Hellenistic period. The earliest written mention of lac as a dye can be found in India in the late Vedi period which corresponds to 1100-500 BC.12 Lac dye has been reported to be an export article from the city of Barygaza (Bharuch) in India which is mentioned by Pliny in the second century AD to be a source of pearl for well to do Roman ladies.13,14 This city is described in Periplus of the Erythraean Sea, a Greek – Roman text ascribed to the 1st century AD which describes navigation and trading opportunities of the time. The earliest references to lac found in ancient text are thought to be written by Aelian (ca. 175-235 AD), a Roman writer who wrote in Greek. His ‘de natura animalium’ is a collection of texts about animals where he describes, among others, the lac insect and the high value garments coloured with lac have. According to the text they are worn even by the Persian King.15 Although the precise date and origin of the text is
difficult to determine due to the fragmented early Christian copies that survive today, it illustrates the likeliness that lac was known as a colourant in ancient times.16
The overall structure of this thesis takes the form of four chapters, including first a brief overview of previous findings of lac pigment and dye within literature. Then the research methodology will be discussed which is comprised of three sections, starting with a discussion on the historical recipes used and how they can be interpreted. Followed by presenting the mock-up samples and how they were made, and concluding with an introduction into the analytical techniques and why they were chosen to examine the mock-up samples. The next chapter is comprised of the results of the analytical examination of the mock-up samples and a discussion into the meaning of these results. The last chapter will conclude by reviewing the findings made in this research and how this could aid the future detection of lac on historical samples.
Figure 2 Molecular structures of the colourant molecules present in sticklac
OHOH N H HO HO HO O O O O OH O OH Laccaic acid A OHOH HO HO HO O O O O OH OH OH OH OH HO HO HO O O O O OH OH O OH NH2 OH HO HO CH3 O O O OH OH OH HO HO HO O O O O OH OH NH2 O O OH OH OH Laccaic acid B
Laccaic acid C Laccaic acid D
Laccaic acid E Erythrolaccin O O OH OH Deoxyerythrolaccin HO HO
2. Current scientific knowledge
This chapter will discuss different findings of Lac in historical pigments and dyes which have been reported by literature. Lac appears to remain an elusive pigment, although lac is mentioned as a colourant in historical texts from earlier dates, when consulting literature it appears to be found on objects in Europe ranging from the 8th century onwards.17,18 A discrepancy can also be seen between the occurrence of historical findings of lac as a pigment and lac as a textile dye. The earliest European textile finds which contain lac can be attributed to the 7th century whereas until recently the earliest known uses of lac pigment in Europe can be attributed to 11th century Portuguese illuminated manuscripts. The historical recipes used to manufacture mock-up samples which are discussed in the next chapter further complicate the lac timeline, the earliest recipe found by the author describing dyeing with lac originates from the seventeenth century opposed by the first recipe for a lac pigment dating from the 11th century.
There are two possible explanations for why the history of Lac remains complicated. First of all, the colourant had to be imported from India to Europe, the scale of this remains largely unknown and has not been studied in detail. It is possible that pigments and textiles made their way to Europe much earlier than the unprocessed stick lac which could have started at a later date, in this case it might even be possible the exact source of the colourant was unknown to the user. A second reason might be found in the detection methods for lac pigments. Laccaic acids have proven to be very vulnerable to acid hydrolysis, which occurs during the commonly used dye extraction method with hydrogen chloride.2,19 More recently a two-step extraction was introduced a soft extraction followed by a slightly more aggressive extraction with a small concentration of oxalic acid in order to obtain all possible colourants without causing degradation.20–23 The first step extracts all direct and vat dyes, and the second step the mordant dyes. This ensures that all possible colourant components are extracted intact to be analysed and possible glycoside decomposition is avoided.24 In conjunction with this development analytical methods have become more sensitive allowing better detection of small concentrations of compounds from very small samples. This creates possibilities for the detection of lac where it might have been missed in the past.
The earliest known samples of lac in Europe to date which can be found consist of historic textiles. Wouters et al. found lac on a Merovingian silk textile dated to the 7th-8th century where it is mixed with madder.2 And more recently Lac was identified on Coptic textiles dating from the 30 BC to 7th century together with madder.25 Lac was also documented to be found on Coptic textiles dating around 900 A.D. where it was identified as the single colourant on wool.18 Pfister and Verhecken suggest that Indian lac was introduced in Egypt after it was conquered by the Arabs in the 7th century, where it was then used as a rival to madder in red colouring.26 However recent believe is that trade was much more fluent than original thought and only an increase in trade occurred due to easier trading routs, making the east more accessible and therefore cheaper, which is why it is found more after the 7th century. Eastwood found Roman textiles at Quseir al Qadim , double dyes blue and red – 1st/2nd AD indigotin + madder, which considering the mixture of dyes might be a suspect for undetected lac pigment.27 From the same site, later 13th – 16th century samples containing lac AD on wool were also found. Most lac dyed textiles appear to be silk, this could be due to the expensive nature of the dye which would most commonly be used on expensive materials, especially when these textiles would be exported.28,29
Lac is often not found as the single textile dye but in combination with madder.2,18,25,30 Madder was possibly used to obtain a certain hue together with lac, either as a mixture in the same dye bath or by overdyeing in sequential dyebaths. Madder could also have been used as an adulterant, to supplement the more expensive and exotic lac dye. Another explanation might be that in these historical textiles the warp which is not visible was dyed with madder and only the visible weft was dyed with the more expensive lac. Thought could be given to the revaluation of some historical textiles on which madder was found during old analysis to see if lac could be found with the new standards of analysis available now. Research by Petroviciu et al indicates that the combination of lac and madder was the primary red dye in the supporting silk fabric for 15th and 16th century Romanian religious textiles.31 They propose an oriental origin for these materials because of the apparent absence of large scale lac dyeing in Europe. Together with the growing understanding that lac is often found in combination with madder gives reason to suspect that there might be more evidence of the use of lac dye on dyed textiles which have been identified as madder dyed. Interestingly this apparently common practice of dyeing lac together with madder is not reported in any historical recipes for dyeing with lac.
Outside of Europe Gleba et al report the identification of lac pigment in combination with munjeet on a wool sample originating from the 5th-7th century in Nepal.32 On this sample resin components were identified which could indicate the presence of shellac. Since it is known that lac was used for dyeing in India since around 1500 BC it would be expected to find historical findings of lac dyed textiles. However this is not reflected by literature as no scientific analysis of ancient lac dyed Indian textiles could be found, if this is due to linguistic barriers remains unclear.
The first known European finds of lac lake pigments seems to be originating in the Iberian Peninsula. Melo identified lac paint in the illustrations of ‘the book of birds’ 1183-1184, no5, ff. 6 and 73v from Lorvão manuscripts from Portugal.33 They were made in St Mamede of Lorvãdo (São Mamede do Lorvão) currently at the Arquivo Nacional da Trore do Tombo (ANTT). Portugal was then under Islamic rule, lac came to the Mediterranean in the 11th century through Arab commercial trading routes with local Jewish partners.11,34,35 Lac was also found in the red paint layers on two different icons, both red lake paints bound in egg, from different cities originating from around the 12th century Cyprus orthodox tradition.36 On one Icon the lac was found as a pure pigment in a paint layer and in the other icon it was found together with redwood. These finds might reflect trade routes of the Arabic world into Europe but the authors note that more research need to be done into icons from this early period for confirmation. Kirby et all describe lac lake being used in a number of European easel paintings, with the earliest example being found in a false enamel decoration on the Westminster Retable (London, Westminster Abbey) dated to around 1260-80.6 These and other samples in later paintings all contain traces of shellac components, indicating an alkaline extraction of the pigment from sticklac.
Table 1 Common names for lac throughout history
3. Research Methodology
The methodological approach taken in this thesis research is a combination of historical recipe research, the creation of mock-up samples based on these recipes, and the technical examination of these samples using several different analytical techniques. These three different research strategies will be discussed separately below.
Historical recipes
Careful consideration has to be taken when reviewing historical recipes for lac pigments or dyes. When organic colorants are precipitated on a substrate to create pigments they are called lake pigments, the terminology regarding red lake pigments presumably has its origin in Sanskrit word for lac, laksha, which was simultaneously used for both lac, shellac and sticklac. The distinction between a red lake pigment made from other sources (such as the plant material madder, and the insect sources Kermes and cochineal) and a lake pigment made from lac is often difficult to discern in historical sources because the word lac and lak are used intermingled, some of which can be found in Table 1.
The earliest reference known today which alludes to the use of lac can be found in India, where it is mentioned for its medical purposes. India had been the centre of lac production and use since the Vedic period which is 1500-600 BC.12 None of these early Indian manuscripts describe dyeing recipes, nor are there surviving recipes from this time period from other parts of the world. The recipes used in this thesis come from various sources, such as manuscripts and dyers
manuals, with dates ranging from around the 9th century to the 18th century. These generally have complicated origins as the surviving texts are often known to be copies of earlier texts which are now lost. An overview of recipes and their text used can be found in Appendix I. A summary of the recipes and their most important parameters is listed in Table 2.
There are different historical recipes that mention the production of a pigment/dye from sticklac. They can be considered in three categories: the lac colorants are extracted and precipitated on Al3+ to form a lake pigment, the lac colorant is extracted from the sticklac source and left in solution as a colorant or the lac colorant is extracted and used to dye textiles. Although the use of textile clippings for the production of lake pigments was common practice between the 14th and the 17th century in Europe, it seems to be used primarily for kermes, cochineals and madder. It is thought that lac and brazilwood lakes were more often made from the raw dyestuff source.22 This is corroborated by the fact that none of the recipes for making lake pigment from textile clippings specifically mention lac dyed textiles. Cennini even advised against the use of lac lake made from textile clippings, this is an indication that lac pigment was being made from textile clippings but considered as bad practice.40
Common
names22,27,37–39 Dutch English Laklak Lac dye, lac lake Finish Lakka
French Gomme Laque, lac lac German Lic-lac
Italian Lacca Spanish Laca Persian Lak Hindu Lakh, Lakha Sanskrit lākshā (������) Synonyms Bengal lac, cake lac, caked lac,
gummilack, grained lac, kadi lakh, seed-lac, rangbatti,
Table 2 An overview of the different recipes, their main ingredients and important observations as recorded in the recipes. The complete recipes and their sources can be found in Appendix I
Recipe Type Date Urine Ley Colorant
extraction Precipitation Notes
J. Le Begue
§12 Lake pigment 1431 boiled Urine or ley Boiled Roche alum and Sal gem with ley boiling
After
precipitation 15 days in urine/ley
J. Le Begue
§36 Lake free colourant 1431 Very old, then boiled - With raw lac 4:1 And alumina zuccarino
J. Le Begue
§37 Lake free colourant 1431 3 days old - Keep lac in urine 3 days then boil Roman vitriol – sulphuric acid Bolognese §129 Lake pigment 15 th Week old
then boiled Create ley from that urine Warm Roche alum cold
Bolognese
§130 Lake pigment 15
th 20 days old
then boiled - In boiling old urine with roche alum Precipitation direct during extraction Natural colorants and dyes Lake
pigment Modern adoptation - Potassium carbonate solution
Simmering Potash, 50 °C Ley is replaced by potassium carbonate
Ibn Badis Lac
colourant free in solution c. 1025 - Borax and sodium carbonate
Boiled Used free in solution
Mappae
Clavicual Lake pigment 12
th or 9th Urine boiled
to reduce Boiled Alum
Ziegler Dyed with
brazil wood
1677 - Tartar and
alum Boiled Yarn in bath for 30 min when still hot bath
Plichto Dyeing silk 1548
italian - - Warm water, extract 2 times Tarter, alum mordanted silk, boil for 1 hour
Boil silk with black soap, roche alum 2 days
Cardon india Dyeing silk - - Overnight cold
water Alum dyeing, cold/hot?
Hellot Dyeing
wool 1789 - - In linen sack, boiled Boiling Bad because it includes the resins through the boiling
Hellot Dyeing
wool 1789 - - Boil, cool Let particles sink to bottom
After concentration dye with it, boiling with alum and tartar
Hellot Dyeing
wool 1789 - - - - Several methods to incorporate less resin, which use other types of ingredients
J Le begue
§11 Lake from clippings 1431 - Ley from ashes Boil ley with clippings More ley, with alum, add liquid Not specifically for lac , scarlet rubea de grana
J Le begue
§13 Lake from clippings 1431 Urine (no time specified)
Clippings in urine for some days
No alum, from scarlet of rubea de grana
Nuremburg
Kunstbuch Lake from clippings 15
th - Ley from
ashes, boiling Boiling ley with clippings Grind with alum Boil agian Paris red from red clippings
Bolognese
§110 Lake from clippings 15
th - Ley of ashes
strong Boil clippings with ley Adding roche alum until precipitated, stirring until cool
Rosato clippings
Bolognese
§111 Lake from clippings 15
th - Ley from ashes
and quicklime (CaO)
Boil clippings
When using recipes from historical manuscripts it is necessary to keep in mind that the recipes are usually very subjective to interpretation. Often translations of the original text are used, and it is often known that the original text is also a copy from an earlier text. The recipes vary in detail; most do not give quantities or time scales and assume things are common knowledge which they are no longer. Further considerations have to be made regarding the materials referred to in the recipes and how they relate to modern materials. This can also be related to the fact that these recipes were often not written by craftsmen but by scholars observing the practices of craftsmen. It is known that these types of recipes often contain some omission of essential parameters, which are either unknown to the author of the recipe, a secret or thought to be common knowledge and therefore omitted. Especially in dyeing practises the recipes were carefully guarded, craft secrets, only transferred verbally and through people directly associated with a dyeing workshop. There are two ways to approach the reconstructions resulting from the recipes as discussed above, one is to exactly follow the historical recipes and see how they result in different types of pigments. But these recipes allow for a large amount of interpretation on important parameters such as amounts and times. The other option is to use the important parameters in each recipe and from this create more analytical reproducible recipes with controlled parameters, as was introduced by Kirby et all in the book ‘Natural Colourants’.22 In the three sections below these important parameters for each type of recipe will be identified.
Lac pigment
Most recipes which describe the production of lake pigments directly from the sticklac pigment sources consist of two essential steps which can be considered general for all lake pigments. The first step is the addition of an alkaline solution to extract the colourant from the resin (often accompanied by heating) followed in a second step by the addition of alum to precipitate the pigment.
Both these steps require an alkaline solution, although most of the colourant molecules are water soluble, the extraction process is much more effective and faster in an alkaline environment. The precipitation of the lac colorant molecules on Al3+ by the addition of alum to form an insoluble metal–dye complex needs an neutral to alkaline environment to achieve good precipitation. Laccaic acid has three pKa values at 5.6, 7.0 and 9.8 which are corresponding to different shades of lac pigment so the pH can also be adjusted to the requirements of the resulting pigment.5 The use of urine to create an alkaline environment is a common practice throughout history. By heating the urine together with ashes a very alkaline solution called ley could be obtained. It is unknown if there is an added value to the other additives which are present in urine. It was used in dyeing practices to scour the wool, remove the grease and clean the wool as preparation for the dyeing process.41 In this process the ammonia, which is formed when heating the urine, plays an important role. When the solution is too alkaline it can damage the wool fibres.
Dyeing with lac
Dyeing with lac was done in a three stage process where the colourants were extracted from the sticklac, the textile was mordanted and the textile was consequently dyed with the extracted colourants. There are three factors which vary between the different recipes; extraction temperature, the dyeing time (which is a variable parameter for every dyestuff), and the addition of potash to the dyeing pot. Often the colorants are extracted in a neutral environment, with cold water over a longer period of time. Extraction of the dye under higher temperatures contributes to the inclusion of resinous components in the dyebath. The sub-sequential dyeing is then done with alum mordanted textiles in several dye baths, sometimes the textiles are mordanted repeatedly in
between the dyebaths. The recipes do not indicate alkaline solutions were being used for the extraction of the colourant. This could have several reasons, it might be the different colours which can be obtained from lac at different pH are not maintained due to the pH environment on the textile. Or perhaps dyeing with lac at high pH is not feasible, there seems to be no indication that dyeing with slight alkaline pH is damaging to the silk fibres, whereas it is damaging for wool.42 In other dye recipes potash is often added to the dyebath which increases the pH and sometime the colour of the resulting dyed fabric. Many dyes show a shift in colour when the pH of the dyebath is changed, but this does not always result in a different colour on the textile, as might be the case for lac. There is one recipe by Plichto (§118 see Appendix I) which mentions the addition of cream of tartar to the dye bath; changes the pH of the solution but was added to remove calcium from water which is not necessary when using demineralised water.22,39 This is general for all the recipes, some note to use the cleanest water possible, or only from a fast streaming river. It is assumed that the use of soft water with a minimal amount of minerals was considered common practise.
Lac pigment from textile clippings
Recipes for the production of pigment from textile clippings consists of two steps: the extraction of colourants from the clippings, and the addition of alum to precipitate the colourant molecules. The recipes to extract lake pigment from textile clippings used here are meant for the extraction of madder or cochineal. These were used because the lack of recipes specifically for the extraction of lac pigment from clippings. The textile extractions are usually done in an highly alkaline environment, with variations in temperature and time. Dyed silk requires less aggressive extraction methods than dyed wool because the silk binds less strongly to the dyestuff.22 In some recipes the fibres are disintegrated by the alkaline extraction and material is incorporated into the final pigment. This is more likely to occur when using wool clippings which contain a higher amount of sulphur that dissolved causing the protein structure of the wool to disintegrate.6
Mock-up samples
Following the considerations described above, three different mock-up sample groups were made: i) pigments made with colourants extracted from sticklac, ii) textiles dyed with colourants extracted from sticklac, and iii) pigments made by extracting colourants from the dyed textiles.
There are two genera lac-dye insects, Kerria and Paratachardina, but research by Santos et al has shown that only the Kerria type contains sufficient red colorants molecules to be used.8 Within Kerria lacca several different species can be identified which are used as dye: laccifer lacca, Carteria
lacca, Tachardia lacca and Lakshadia lacca.43 According to the supplier, Kremer Pigmente GmbH &
Co. (Aichstetten, Germany), the sticklac used in this research originates from the species Tachardia lacca.
Research has shown that there are three important factors that determine the solubility of shellac, and thus the amount of resin which is incorporated into the final pigment: the temperature and pH used during manufacturing and the age of the raw sticklac. These are also factors which are often left open to interpretation in historical recipes. The resin fraction becomes more soluble as the pH and the temperature during the extraction of the colourants go up. Heating the sticklac to above 60⁰C when it starts melting, and a pH above 8 for extraction promotes the incorporation of resin.10,44 Research by Farag et al has shown that there is a significant decrease in the solubility of shellac resin when one year old resin is compared to 5 year old resin due to the polymerisation of the resin network.45 The raw sticklac available at the British Museum was 10 years old, for that reason new sticklac was ordered from Kremer Pigmente GmbH & Co. (Aichstetten, Germany). They report the acquirement of the batch of sticklac currently sold in March 2017, which allows for the assumption
that the sticklac used is around 1 year old. The increased solubility of the shellac fraction is likely to contribute to an increase in the amount of colourant molecules extracted from the sticklac and releases them from the polymer network in which they might been encapsulated. When the polymer network is not broken down the extraction of the colourant molecules from the polymer network is dependent on the solubility and diffusion of these molecules through the shellac polymer network. Extraction then becomes dependent on the age of the shellac and the extraction time, this is reflected in the historical recipes where cold colourant extractions are often done over long periods of time.
It is known that lac does not bind well to plant fibres, excluding cotton and linen for substrate fibres, but ancient examples can be found of silk and wool dyed with lac.38 There is a high likelihood that only expensive fabrics such as silk were transported from India to Greece in the Hellenistic period. The presence of tannins in the paint samples from the terracotta oinochoe vase is an indication that silk clippings were used as source for the pigment.1,6 Silk was often weighted using tannins which, if silk clippings are used as a source for colourants, are consequently extracted together with the colourant molecules and end up in the resulting pigment.
In view of these consideration and an attempt to limit the scope of this research it was decided to only focus on silk as textile clipping source and exclude wool. Assuming that the silk is degummed no further pre-treatment of the silk is necessary except for the mordanting which is discussed further onwards. Although the presence of tannins influenced this research in the direction of silks and pigments made from clippings, the silks used in the mock samples are not weighted. There is little information known about the weighting of silks, research by Hacke shows that the earliest areference of weighted silks in Europe can be found in court records of 1606.46 There is no documentation to be found on earlier historic weighing practices or oriental weighting practices. In conjunction to the weighting of silk, tannins could also be used as a type of mordanting.38 Sources of tannins include a wide variety of materials: sumach, gambier, chesnut, alder bark, gall nuts, catechu, myrabolams, valonia, acacias and divi divi.38,46 Stick lac naturally contains pieces of bark from the tree where it originates from as the material is deposited on the bark, which can also be seen as a possible source of tannins. Within this research is was decided to use unweighted silks to limited the scope of the research and to evaluate the possibility of the introduction of tannins through pieces of tree bark in the sticklac. The silk was dyed as a plain woven textile, as this was the material readily available within the British Museum.
Lac is a mordant dye, and the mordant used in each of the historical recipes is alum. It is more common to uses alum in European dyeing practices for insect based dyes; however in Asia iron was also often used. Iron mordanting was often used in combination with tannin weighting for dark colours through which even larger amounts weighting could be achieved.38,46 Because the scope of this research is constricted to Europe it was chosen to solely use an alum based mordanted textile. The mordant changes the colour of the dye, aluminium gives a crimson colour and iron a more purple colour (tin gives a scarlet colour but was not used until the 17th century).38 Silk should never be in a bath with a temperature above 70°C which was taken into consideration when dyeing47.
Taking into consideration the important experimental factors determined from the historical recipes, standard procedures for mock-up samples were made, by adapting the standard recipes for organic lake pigments by Kirby at all as published in their book Natural Colorants for Dyeing and Lake Pigments.22 This resulted experimental procedures which can be found in Appendix II and the resulting samples of which a list as can be seen in Table 3 and 4 . Extractions at high temperature indicate boiling point for 10 min, whereas room temperature signifies an extraction of three days. A distinction is made between the samples which are presented in the two tables, Table 3 shows the
main sample group which were made by varying only the temperature and the pH of extraction. Table 4 represents other samples which were made during the course of this research and investigate other experimental parameters of the manufacturing process. The pigments made from textile clippings are called textile pigments, whereas the pigments made directly from the sticklac are called direct pigments.
Table 3 List of samples and names, the abbreviation in the sample names stand for: P) pigment ;DP) direct pigment made from sticklac; TP) textile pigment, a pigment made through extraction of colourants from a dyed textile; Tu) dyed textile unaged; Ta) dyed textile aged; H) hot extraction of colourants at boiling point; C) cold extraction of colourants at room temperature; N) at a neutral pH of 7; A) at an alkaline pH of 11; 1) textile dyed once; 3) textile dyed 3 times with mordanting in between.
Sample Name Description of the main experimental parameters in the production of the sample P Lac dye, Kremer pigmente, 36020 pure colourant
DP_H_N High T, pH 7 extraction DP_C_A Room T, pH 11 extraction DP_H_A High T, pH 11 extraction
TP_C1_C Pigment made from textile Tu_C1 by room T and high pH extraction TP_C1_H Pigment made from textile Tu_C1 by high T and high pH extraction TP_C3_C Pigment made from textile Tu_C3 by room T and high pH extraction TP_C3_H Pigment made from textile Tu_C3 by high T and high pH extraction TP_H1_C Pigment made from textile Tu_H1 by room T extraction and high pH TP_H1_H Pigment made from textile Tu_H1 by high T and high pH extraction TP_H3_C Pigment made from textile Tu_H3 by room T extraction and high pH TP_H3_H Pigment made from textile Tu_H3 by high T and high pH extraction Tu Undyed silk
Tu_C1 Room T extraction, dyed once
Tu_C3 Room T extraction, after 3rd mordanting and 3rd dyebath
Tu_H1 High T extraction, dyed once
Tu_H3 High T extraction, after 3rd mordanting and 3rd dyebath
Ta Undyed silk Tu aged one month Ta_C1 Dyed silk Tu_C1 aged one month Ta_C3 Dyed silk Tu_C3 aged one month Ta_H1 Dyed silk Tu_H1 aged one month Ta_H3 Dyed silk Tu_H3 aged one month
Table 4 List of samples and names, the abbreviation in de sample names stand for: P) pigment ;DP) direct pigment made from sticklac; TP) textile pigment, a pigment made through extraction of colourants from a dyed textile; Tu) dyed textile unaged; Ta) dyed textile aged; H) hot extraction of colourants at boiling point; C) cold extraction of colourants at room temperature; N) neutral pH; A) alkaline pH; 3.1/3.2/3.3) the same dye bath in which three consecutive pieces of silk are dyed.
Sample Name Description of the main experimental parameters in the production of the sample DP_H_A_T Extracted pigment, after unsuccessful dyeing of silk TU_H_A at high pH and high T
DP_H_U Urine, high T extraction
TP_H3.1_C Pigment made from textile TU_H3.1 by room T extraction and high pH
TP_H3.1_H Pigment made from textile TU_H3.1 by high T and high pH extraction
TP_H3.2_C Pigment made from textile TU_H3.2 by room T extraction and high pH
TP_H3.2_C Pigment made from textile TU_H3.2 by high T and high pH extraction
Tu_H3.1 High T extraction, 1st silk dyed in dyebath
Tu_H3.2 High T extraction, 2nd silk dyed in same dyebath
Tu_H3.3 High T extraction, 3rd silk dyed in same dyebath
Tu_H_A High T and high pH extraction
Ta_H3.1 Dyed silk Tu_H3.1 aged one month
Ta_H3.2 Dyed silk Tu_H3.2 aged one month
Many different techniques were used to examine the samples after they were made. This section will briefly introduce those techniques and explain why they were used and to what means. The specific technical details of each technique can be found in the experimental section of Appendix II. Care must be taken when comparing the different samples to each other with each of the described techniques. Analysis of the textiles should be viewed separately from the pigments because of the medium difference. Although care was taken the same quantities were used to produce each sample within each sample group, this is was impossible between the sample groups. This creates difficulties when comparing the textile pigments with the direct pigments. It is impossible to estimate the concentration of colourant molecules extracted from the textiles and how this relates to the concentration of colourants extracted to create the direct pigments. However the relative ratios of compounds can be used as a comparative feature as this remain the same when the total concentration is altered.
Multispectral imaging
Multispectral imaging is the term used to describe the employment of different sections of the electromagnetic spectrum to visualise the characteristics of materials under these types of light. In this case a modified camera, filters and different light sources were used to capture this. By using standardised photography settings, colour calibration and post processing these images can be used to deduce scientific results. In this research the settings standardised by the CHARISMA working group have been used.48 Often false colour images are used to highlight these differences and make the information more comprehensible. The visible reflectance images (VIS) are made by blocking all the ultraviolet (UV) and infrared (IR) light from entering the camera detector, resulting in an image that records the wavelengths between 400-700 nm. The ultraviolet reflectance (UVR) is made by using UV lamps resulting in the capturing of the reflected UV radiation (200-400 nm) by blocking the VIS and IR parts of the spectrum . The resulting image captures the ability of the subject material to reflect the UV light. Ultraviolet induced luminescence (UVL) uses UV radiation and records the resulting VIS light (400-700 nm) induced luminescence. Reflectance infrared (IIR) uses a filter to block all UV and VIS light and visualises the IR light which is reflected (700-1100 nm). Visible induced visible luminescence (VIVL) records the emission of light in the visible region (500-700 nm) induced by a different part of the visible light spectrum (400-500 nm).
Colourimetry
When studying samples of which colour is a parameter of interest, a way needs to be found to quantify and describe the way the colours are perceived by humans. This is called colourimetry and can be done in multiple ways, usually by recording the colour spectrum and translating this to values in a colour space. There are special spectrophotometers which have a day light lamp that emit a specific known colour spectrum. From this the calibrated colours in a specific colour space can be calculated. The downside of this is that the colour needs to be on a homogeneous flat surface, making it impossible to measure loose pigments. Another way to calculate colour values is to usespace colour calibrated images and average the colour values using a processing tool. These values are obtained under less controlled conditions and less accurate because they do not take into account the irregularities in the surface of the subject. But this feature is also what makes it possible to get an indication for the colour values of samples such as loose pigments. The colour space used was the CIE L*a*b* (CIELAB) introduced by the international commission on illumination. It uses three different coordinates to describe colour: L* represents lightness with 0 being black and 100 diffuse white, a* green for negative values and magenta for positive values, b* blue for negative values and yellow for positive values. To evaluate the colours irrespective of the lightness only the a* and b* values are often used in two-dimensional model. The lightness values can be neglected
because they are determined by the colourant concentration and can thus only be viewed per sample group.
HPLC-DAD-ESI-Q-ToF
High-performance liquid chromatography – diode-array detection - electrospray ionisation – quadrupole – time-of-flight (HPLC-DAD-ESI-Q-ToF) comprises of three different components. First is the chromatographic separation technique (HPLC) which separates a liquid sample according to the molecular structure of the different analytes present in the sample. Then two sequential detection techniques, DAD is a non-destructive method that uses light to measure the absorption of the eluent coming from the HPLC. Followed by a mass spectrometer (MS) which couples quadrupole magnets with a time-of-flight set-up to detect the mass to charge ratio (m/z) of each molecular species present in the eluent as it enters the machine. This can be translated to an accurate mass for the molecules. The ionisation method used is electro spray which is considered a soft ionisation mode that allows the molecules to be ionised without breaking the molecular structures. The MS has a double function with two parts, it can be used with both parts functioning as an MS which is called MS/MS mode. The quadrupole then works as a selective ion filter after which the selected molecules are detected by the ToF. This function is useful when dealing with unknown compounds, the molecules break down from which the original molecular structure can be deduced. However thanks to the work done on the molecular composition of shellac by Tamburini et al. this proved unnecessary.49 The quadrupole was used to focus the molecular ions and the ToF was used to record the m/z values.
DAD can be considered as a quantitative technique due to the linear relationship between the intensity of the signal and the concentration of analytes responsible for the signal over a range of concentrations, as proven by the law of Lambert-Beer.50 It has its limitations in the fact that it can only detect coloured molecules in non-absorbent mobile phases and the sensitivity for low concentration of analytes is problematic as deviations from linearity occur.
The DAD measurement was recorded at 491nm as this corresponds to a high intensity area for laccaic acid A and B. The peaks were integrated and adjusted to their possible dilution factor and the sample weight used in the extraction. The difference in substrate unfortunately makes it impossible to normalise the textile samples to the pigment samples. The average of the normalised integrated peak area of the measurements in triplo was used in the final results. For some measurements no significant DAD signal was obtained. This can be attributed to so called hyphenation problem of coupling a MS detector to a DAD detector. Both detectors have different requirements of the HPLC system and analytes. The DAD detector requires relative high concentrations of analytes and volumes (which can be translated to high flow rates). MS systems require the opposite, low concentrations of analytes as they are very sensitive, and low flow rates to obtain complete ionisation. The ESI ionisation technique is one of the few which can be used in combination with a DAD detector as it can function with higher flow rates. However all analytical conditions from the sample preparation to the system were optimised for MS functionality and not DAD.
Mass spectrometry, as used during this research, is not a quantitative technique due to ionisation discrimination. Not all molecules are ionized by the ESI and some molecules ionise more easily than others which results in those molecules being detected at a higher intensity than others. Quantification can only be achieved when know reference standards can be used of each analyte of interest, which are not available for all molecules of interest in this research. Although the intensity of the signal is not quantitative, it can be consistent across a sample and be used as an indication of the relative ratios within compounds. The systematic error caused by the instrument was evaluated
by measuring each extracted sample in triplo. If the peak areas remain within statistical agreeable limits this can confirm that although the ionisation is discriminative, it is consistent across the samples and can be compared for this system. The random error caused by possible inhomogeneity within the mock-up samples and the experimental conditions was not evaluated. The triplo HPLC measurements were all done from an extract taken from one sample taken from the corresponding mock-up sample. Although this can create uncertainty within the results, this is consistent with general museum practise. Due to the value of historical objects and their function inside museums samples are only taken in a very limited manner, it is very rare to have enough samples to obtain statistically sound results. Each mock-up sample was only made once, making it impossible to evaluate the variation causes by the production method.
In order to effectively compare the MS spectra of the different samples a selection of sixteen molecular species was made that represents the different free acids, esters and polyesters present in shellac as well as the colourant molecules Laccaic Acid A and B, Erythrolaccin and Deoxyerythrolaccin. They were chosen by considering the molecular composition of shellac and their abundance within the sticklac matrix. A list of these molecules and their exact m/z value can be found in Table 5. The separate MS spectra of each sample were evaluated separately as there are a lot more different molecules present in the sample matrix than the sixteen presented in Table 5. However a complete characterisation of these spectra did not proof to be feasible in the scope of this research as it did not aid in answering the research questions. From previous research the elution time of the molecules of interest was known, which was used in combination with the accurate m/z value to determine the molecular ion peak within the total ion current (TIC).49 The integrated peak areas of all detected molecules per sample were summed from which they were normalised to 100%. Each sample was measured three times, the percentages were calculated for the separate measurements, after which an average of the relative peak area was taken. The standard deviation was calculated over the three relative peak areas calculated per measurement. The total standard deviation was calculated by averaging the standard deviations over all samples per molecule of interest.
PCA
Principle component analysis was used to interpret the MS data. It uses a statistical approach to create a different visualisation of the data which makes it easier to identify the variations and similarities that exist within the dataset. It does this by converting the data into so called principle components, directions in which the data shows the most variance. Then, these become the new axis on which the data are projected instead of the x and y axis. This is mathematically done by deconstructing the data set into eigenvalues and eigenvectors. These represent the respective degree of variance and the direction of the variance, the eigenvector with the highest eigenvalue becomes the first principle component – the x axis. The number of eigenvalues – vectors corresponds to the amount of variables exist in the original dataset. But the PCA identifies and quantifies the variables which contain the most variation, and thus the most interesting information. This becomes especially useful when the dataset contains a large amount of variables, as is the case in the dataset obtained from the HPLC-DAD-ESI-Q-ToF measurements. The PCA calculations were done using EXSTAT extension of Microsoft excel, information on the calculations can be found in the experimental section of Appendix II. An important last step in the PCA is identifying which variables are represented in principle components and what kind of variation (positive or negative) they represent. To visualise the difference between colourants from lac (Laccaic Acid A and B) and the often shellac associated colourants (Erythrolaccin and Deoxyerythrolaccin) these were summed for the PCA analysis. It was also decided to sum all the polyesters to create a more comprehensible PCA. The tannin source Ellagic Acid was excluded from
the PCA as it distorted the PCA due to being uncorrelated to the other variables and clear to interpret.
FTIR
Fourier transform infrared spectroscopy was used as an preliminary research tool. It is a technique that measures the vibrational spectrum of molecular bonds by the characteristic absorption of IR radiation. This can give information about the molecules and their molecular environment. When looking at more composite materials such as the samples in this research this can become more complicated, as some materials such as the silk textile might give a very high intensity signal causing other signals to become more difficult to see and interpret.
Scanning electron microscopy
Scanning electron microscopy (SEM) was employed to evaluate the condition of the textile fibres before and after the dyeing process and the extraction of colourants. By looking at the fibres under high magnification and high resolution any damage that might occur during the production process can be seen.
Ageing
Ageing experiments were done on the textile samples. Because the textile surface does not require binding medium, it allows for a preliminary indication of the behaviour of the lac colourants and any shellac components which might be present and the possible difference in degradation between them. The original samples from the Hellenistic terracotta sculpture were so under bound that it was not possible to determine the binding medium. Evaluating the pigments in binding media would have required more research and unfortunately proved to be out of the scope of this research. Ageing was done under visible light, without UV, at temperature of 40°C and a relative humidity of 50%.
Table 5 The molecular species investigated using HPLC- DAD-ESI-Q-ToF and their accurate m/z value.
M/z Name Component type PCA data manipulation
243.1996 Butolic acid Free acid from shellac 263.1298 Laccijalaric acid Free acid from shellac
269.0455 Deoxyerythrolaccin Colorant from shellac Summed with Erythrolaccin for PCA 279.1238 Jalaric acid Free acid from shellac
285.0405 Erythrolaccin Colorant from shellac Summed with Deoxyerythrolaccin for PCA
300.99 Ellagic acid Tannin Excluded from PCA
303.2177 Aleuritic acid Free acid from shellac
495.0569 Laccaic B Colorant from lac Summed for PCA
536.0834 Laccaic A Colourant from lac 549.3433 Aleuritic-Laccijalaric Ester from shellac 565.3352 Aleuritic-Jalaric Ester from shellac
791.5315 Jalaric-Aleuritic-Butolic Diester from shellac Summed for PCA 827.4587 Jalaric-Aleuritic-Jalaric Diester from shellac
1113.673 Jalaric-Aleuritic-Aleuritic-Jalaric Triester from shellac 1339.866 Jalaric-Aleuritic-Butolic-Aleuritic-Jalaric Tetraester from shellac 1662.008 Jalaric - Jalaric - Jalaric – Aleuritic-
Figure 3 Visible light image of all pigments mock-up samples and their names. Also shown here are alum, an orange shellac and a contemporary industrially made rose madder lake from kremer pigmente.
4. Results and discussion
As discussed in the last chapter a total of 33 samples was made which will be discussed here divided into two sample groups. The main sample group can be found in Table 3 and contains all the samples which were made by variations in the temperature and pH of extraction. These samples will be the main focus of this discussion and result section as they represent the fundamental experimental parameters in lac production. The other samples which were produced, as shown in Table 4, contain samples that represent variations in different parameters, this group will be discussed separately at the end of this chapter. After discussing some observations which were made during the production of the mock up samples the main sample group will be discussed per analysis technique.
The main sample group contains three direct pigments, eight textile pigments and four textiles (unaged and aged). There originally were supposed to be four direct pigments, but lac colourants will only precipitate on alum under alkaline conditions. When the colourants were extracted under cold and neutral conditions no pigment was obtained, under neutral and hot conditions some pigment precipitated although not in comparison to the pigments precipitated in alkaline conditions. All colourants for dyeing were extracted at neutral pH, dyeing is unsuccessful at a high pH, which gives an indication of the amount of shellac that can possibly be extracted under cold and neutral conditions.
From the sticklac residue after the colourant extraction an initial evaluation of the possible degree of shellac inclusion in the final product can be made. Under cold and neutral conditions the sticklac appears the same after extraction. When heat is applied the residue melts and forms a sticky lump. When the colourants are extracted in alkaline condition at room temperature, some of the shellac appears to dissolve but a residue remains. When alkaline conditions and heat is applied, the sticklac dissolves completely and only a few bits of wood remain after the extraction. The colourants extracted for dyeing appear the same, and result in similarly coloured silks. Some are unevenly dyed which is caused by insufficient stirring, and the silk texture has changed to slightly more rough after dyeing. This could be caused by a shellac coating deposited on the silk fibres. From the textile clippings from which the textile pigments were made a clear distinction can be seen between the hot and cold extraction methods (both under alkaline conditions). When heat is applied the clippings are back to their undyed colour, whereas when the extraction is done at room temperature a pink colour remains on the silk after the extraction. The resulting pigments show a lot more variation in colour than their original textiles, indicating that perhaps the colour saturation of the silk was reached during dyeing. One last interesting observation about the production of these samples is that some parts of the historical recipes are more logical after having reproduced them. Recipes for direct pigments often mention that a clean bowl should be used after extraction to precipitate the pigments in, which is probably related to the melted shellac residues which stick to the vessel after extraction.
Imaging
A set of multispectral images was made of the pigments and dyed textiles (unaged and aged). The visible light image of the pigments can be found in Figure 3, the other images can be found in Appendix II. These images can provide a lot of insights which can be corroborated by other techniques which will be discussed further on. In the visible light image of the pigments twenty one samples are presented. On the top alum and shellac can be seen and the top left the Indian lac obtained from Kremer Pigmente, this colourant was proven to be pure laccaic acid A and B (not precipitated on a substrate) by FTIR, DAD-ESI-Q-ToF and the fact that it readily dissolves in water.
Together these three form a reference to determine their individual influence on the mock-up samples. Rose madder lake is included as an example to show the properties of a commonly used lake pigment. There is a lot of difference in colour between the pigments. The most striking difference is between the Kremer pigments and the pigments made during this research. This difference can be attributed to the three pka values of the laccaic acids. From this it can be concluded that the Kremer colourant was produced in an acid environment, and when dissolved in water it becomes a similar shade as other pigments. The difference in colour of the mock-up pigments might be attributed to a number of factors such as: pH, concentration of colourants, inclusion of shellac or other production method related factors. The two direct pigments made in an alkaline environment are darker in colour than the majority of the textile pigments, which are somewhat more pink in tone.
All pigments show a degree of infrared reflectance. Some of the bright pink pigments (TP_C3_C; TP_H1_C; TP_H3_C) show a reflectance similar to the reference Kremer pigment. The darker pigments (DP_C_A; TP_H1_H) show the least reflectance. Note that the reflectance of shellac is difficult to measure because of the flaky characteristics of this sample it is reflecting on its own, non IR induced reflectance. In the UV reflectance image the reference pigment shows no reflectance under UV while alum is highly reflecting (shellac difficult because own reflectance). The darker pigments show no reflectance, they absorb similarly to the reference pigment. The light pink pigment (DP_H_N) shows a higher reflectance, similar to the alum. This shows that UVR is mainly due to the alum present, while reflectance is reduced by a high concentration of pigment which absorbs the UV. The UV luminescence shows that the alum and reference pigment do not luminescence under UV, but the shellac does, indicating that any luminescence in the pigments must come from shellac. The rose madder is very different under UV. Also to be noted is that the 2nd column and 4th column of pigments in Figure 3 luminesce more than the other two. These correspond to the pigments made through a hot extraction, in the case of the textile pigments hot extracted from the silk disregarding the original extraction method of the dye for the silk. This could be an indication for a higher concentration of shellac.
There is little difference to be observed between the textiles in the multispectral images. Inhomogeneity in the dye can be seen throughout the images on most of the samples indicating that the dyebath was not stirred enough. The aged silks show little change in the multispectral images except for a change in colour intensity corresponding to a probable fading and thus degradation of the colourant molecules.
