SEEING RED

SCROLL PHOTO JEAN FITZGERALD

What was it that brought us to the making of pigments? Was it fascination or just frustration? Like so many aspects of our trade, making pigments is another esoteric path one may follow when creating a fine instrument. It is by no means a necessary requirement to be a successful violin maker, although both the process and the end results are stimulating and intriguing. Not having a consistently reliable source of pigments was also becoming a personal frustration for us, and that became a driver to motivate our re-entry into the world of ‘making your own pigments from madder root’. Building on the work of others, in this article (which assumes a little prior knowledge) we try to explain some of the factors that influence the quality, intensity and colour hue of making lake pigments in small batches. We also share a useful method of keeping and comparing results. The majority of the experimental work took place at Oberlin College, OH, US, as part of the Violin Makers Workshop over the course of several years: 2016–19. Our thanks and recognition go to our colleagues there and around the world whose contributions have made this article possible.

Above left Raw madder root with samples of the finished pigments Above Tools, chemicals and further samples Below left Scroll varnished with the slow-cooked, acid-soaked madder root pigment in oil varnish (see box, page 45)

Like many others, we had made some pigments from madder root early in our careers. The results were variable and the process was by and large lengthy and haphazard. The colour hue was not always what we wanted. The intensity tended to be low and the pigment itself sometimes gritty and grainy. This meant it often felt like mulling gravel into your varnish: there were problems with opacity and unwanted texture. It was an easy decision for us to buy and use good-quality commercial pigments, such as those from Magister or Winsor & Newton, in our instrument making.

Fast-forward 10 to 15 years and these sources had become unavailable, while our own existing supplies were running low. So in 2016, during the ‘Obie 1’ violin project in Oberlin (see The Strad, November 2016) we decided to use our own pigments as well as a home-cooked varnish. Partly because we wanted to shorten the time required to create the pigments and also improve the consistency of the product, we used a modified version of Eero Haahti’s recipe outlined in the May and November 2012 issues of The Strad (republished in The Best of Trade Secrets 3). We used a coffee machine to ‘express’ the dyestuff from the raw materials and a Büchner funnel and vacuum system to add speed and consistency to the process. We decided against adding acetone as a finishing step, and just let the pigments dry naturally, assisted by a small fan. We also chose to use potassium aluminium sulphate (alum) and potassium carbonate (potash) as the acidic and alkaline chemicals in the making of our lake pigments.

The use of the coffee machine and the filtration system did result in a great increase in the speed of production and the end results were consistently usable. We used a mixture of madder root and buckthorn berries to create a fine orange pigment that combined easily with our cooked colophony resin and oil varnish. However, this ‘Madbuck Special’ pigment (as we called it) was not as intense as we wished, and tended more towards a yellow–orange rather than a deep orange–red hue. While we felt we followed our recipe quite closely, we realised we did not know enough of the ‘whys’ to change the recipe to create more intensity or vary the colour hue.

WE REALISED WE DID NOT KNOW ENOUGH OF THE ‘WHYS’ TO CHANGE THE RECIPE TO CREATE MORE INTENSITY OR VARY THE COLOUR HUE

GRAPH COURTESY KAE SATO-GOODSELL

The desire to know – what we would call ‘suffcient why’ – led us down the path of more comprehensive experimentation in 2018 and 2019. We say ‘suffcient why’ because neither of us is a trained organic chemist, nor do we have the intention of setting up commercial pigment production. We did not need to understand all the chemical processes, nor maximise yields while minimising input costs. We wanted to know enough to vary our recipes to make a repeatable, intense, transparent pigment of our desired hue in our workshops. As our main intention was to use the pigments in the manufacture of new instruments, even some variability of transparency and colour hue between batches was acceptable.

To this end we made 15 diTherent batches of pigments in 2018 and 2019, varying the pre-treatment of the raw madder root, the method of extraction of the dyestu., the mineral salts, and controlling the temperature and pH of the solutions. We compared the .nal pigments with two pigments we used and liked in our making: Magister Lacca Rubia no.1 Red, and Lacca Rubia no.2 Red Orange (both now sadly unavailable). We then quantiThed the results against the parameters of transparency, intensity and colour hue. A graphical representation of these results can be seen in the graph.

THE INNER MADDER ROOT TENDED MORE TOWARDS ORANGE, WHEREAS THE BARK TENDED MORE TOWARDS RED

Madder root contains a number of potential colouring dyestu.s in the form of sugar-like substances, such as alizarin and purpurin. These substances respond diTherently to heat, fermentation and acidity. Varying the pre-treatment of the raw madder root can help target certain dyestu.s and aid their transparency. Similarly the roots themselves have a ‘inner root’ and surrounding ‘bark’. By creating one batch using just the inner root, and another with.just the bark, and then comparing each of them with a ‘whole root’ batch, we could make conclusions about the eThects of the constituent parts. From our results we concluded that the.diTherent parts created only mild colour variations: the inner root tended more towards orange, whereas the bark tended more towards red. The diTherent parts are easily separated by soaking the roots in water for about an hour.

We also explored, on a small scale, methods of madder root preparations. For example, in a number of batches we presoaked the roots in dilute sulphuric acid and water (1% by volume), then rinsed the roots in clean water before extraction of the dyestu… This acid treatment consistently appeared to increase the transparency of the .nal pigments in varnish. We experimented with other forms of fermentation, both wild yeasts and baker’s yeast, with whole roots and water. In both cases this initially caused a strong yellow dyestu. to exude into the water solution. This yellow was likely a potential dyestu. for pigments that we did not pursue further in these investigations. However, neither form of fermentation appeared to move the resulting pigments towards greater transparency, nor greater red pigmentation, and they did not appear to diTher greatly in their end result.

Acid-treated madder root: slow-cook method

This recipe gives a fiery red semi-transparent pigment with excellent intensity

• Soak 120g of madder root in a 1% sulphuric acid solution for 20 hours in a glass container. Rinse the madder root under tap water for five minutes and then allow to soak in tap water for one hour.

• Pour away the soaking water and place the madder root into a stainless-steel pot with 3.2l of distilled water. Heat this mixture to a temperature between 70-80°C and maintain for one hour. Meanwhile, dissolve 120g of alum in 480ml of warm distilled water and add this solution to the pot. Maintain this madder root, water and alum mixture at 70-80°C for 6-7 hours.

• Strain the mixture while hot, first through a stainless-steel colander, then a stainless-steel strainer (sieve), fast-speed filter paper (oil filter), and finally through medium-speed filter paper in a Buchner funnel, to obtain a clear red liquid dye.

• While filtering the solutions, heat 480ml of distilled water and dissolve into it 120g of potash (potassium carbonate).

Place the dye into a large plastic bucket and slowly add the hot potash solution, while vigorously stirring the two together. They will react and foam while the pigment forms and precipitates out of solution. Then quench the reaction with a large amount of cold tap water.

• Allow the mixture to stand for one day. Siphon off most of the water and excess dye leaving about 3cm of liquid above the layer of pigment. Refill the bucket, allow to settle and siphon again. Repeat this process two more times. Each successive rinsing should take a shorter amount of time for the pigment to settle to the bottom. The final time, siphon off most of the water again and use a Buchner funnel, fast filter paper and a vacuum pump to remove as much of the remaining water as possible. With the pigment still on the filter paper, place aside to air-dry fully. Once dry, store in an airtight container. When needed, simply grind the pigment into varnish on a glass plate, ensuring all the pigment is adequately wetted and any lumps removed.

The methods of extraction did aThect the resulting dyestu and pigments. We used two main methods: the coThee-machine method mentioned above; and a modified version of a slow-cook method outlined by David Rubio (bit.ly/2oGedIU). The slow-cook method maintains a root, water and alum solution at.a.constant temperature of 70–80°C for 6–7 hours in a large pot (see box above). We believe the controlled temperature of the slow cooking allows for selective extraction of particular dyestufis from the madder. In both methods the dyes were filtered before combining them with other solutions to form the pigments.

The.slow-cook method also required extensive multi-step straining to remove sediment and bits of root from inclusion in the final pigments. The extraction method had the biggest correlation to intensity when potash was used as the alkaline solution. In both methods of extraction the solutions of dyestufis and mineral salts were kept hot when combined, resulting in smaller crystals of pigment being precipitated. We assumed for our experiments that pigment intensity was partly a function of crystal size. A larger number of smaller crystals will result in a more intense colour than a single large pigment-crystal of the same volume. Since the solutions were combined while hot, the chemical reaction was quicker and the crystals do not have time to grow in size as they fall out of solution. After precipitation and while the wet pigment dries, some of the crystals will clump together, but can then be easily reduced to their original size by grinding on a glass plate. However, the individual crystals cannot be ground smaller than their size at the moment of precipitation.

Therefore, how the pigment crystals are originally formed is an important factor in the intensity of the final pigment.

In our experiments we used four diTherent chemicals in combinations with the dyestufis to form the lake pigments: the aforementioned alum and potash, plus sodium hydroxide (lye or caustic soda) and stannous (tin) chloride. Of.the four, we did the least experimentation with the stannous chloride. In an initial test we found it didn’t produce a useful pigment for our purposes (opaque brown #11-WR) and recognised that how to use it successfully would require further research in the future.

The chemicals were added to heated distilled water to form solutions. The concentrations of these solutions and the overall amount of chemicals varied from batch to batch. In every one of them, we worked from the assumption that we wanted an excess of dyestu to available metal salts. In this manner we sought to avoid the formation of ‘blank’ uncoloured metal salts in the final pigment, as we felt this ‘filler and extender’ would reduce its intensity.

We varied the order of combination across certain batches. For example, adding the dyestu to the alkaline solution first, then adding the acid solution, and then vice versa. We determined that the eThect of pH on the combined solutions was more significant than the order they were combined. Chemist advisors told us that a certain pH di Therential is necessary to create and sustain the reaction. However, the level of pH at the point of precipitation does aThect the final colour, when the reaction is quenched by ooding with cold water. In general we found that a lower pH gives more orange; middle pH a clear red; and a higher pH a more purple red. However, our results did suggest that when lye was used as the alkaline solution (at the same pH as the potash solution) the pigments did tend to be more red and have the highest intensity. Some of this perceived intensity could also be due to a higher opacity. More opaque pigments have a better coverage, which we also see as intensity.

A Büchner funnel and vacuum pump are used to filter the liquid dye before the pigment is precipitated

Comparing different pigment samples in the glass slides for colour, transparency and intensity

Controlling the pH of the combined solutions was a challenge. In hindsight we could have achieved better control by adding more dilute solutions to the dyestufimixtures, or by delivering the more concentrated solutions using equipment such as pipettes or smaller syringes. Measuring the pH while adding and stirring the mixture was also part of the issue. We­used a variety of difierent litmus papers to measure the pH. Yet, as the solutions foam together in clouds of colour it is di.cult to determine colours accurately, particularly if one of the indicator colours is itself red. For more accuracy an electronic pH reader might have been more useful.

SLIDES MADE IN THE FIRST YEAR WERE RELATIVELY EASY TO STORE AND REMAINED UNCHANGED AFTER TWELVE MONTHS

One of the more immediate bene.ts of our experimentation was to pioneer a new way of presenting and recording the results of pigment tests. Assessing and comparing dry pigments ofiers little information. Pigments need to be thoroughly ‘wetted’ in a varnish medium to reveal their characteristic colour hue, transparency and intensity. Previously we had mixed pigments with varnish and smeared them on pieces of wood. The lack of homogeneity in the wood structure and surface preparation meant it was very dificult to compare samples. In particular, it was dificult to compare features such as transparency and intensity. As a more consistent alternative we used glass microscope slides, with abuilt-in cavity to hold the sample of the pigmented varnish.

To produce consistent samples of each pigment, our method was to use 0.1g of dry pigment crushed with a palette knife on to a glass plate. Initially wetted with mineral spirit, it was then combined with ­ve drops of Holtier Clear Violin Varnish (this is a commercial cooked-resin oil varnish, with no added solvents or driers). Approximately half was placed on a cavity microscope slide and covered with a .at microscope slide taped at either end and labelled.

Slides could be compared with each other and grouped for colour similarity, and placed on a colour range between orange (at1) and red (at 5). To assess transparency and intensity, the slides were compared in natural light, by holding them against a window facing the sun, and placed on a scale of 0 (lowest) to 8 (highest). We found the slides made in the first year were relatively easy to store and remained unchanged after 12 months – hence they could be compared with new examples in subsequent years. The varnish medium was stable in the cavity slide. There was some capillary creep of the drying varnish between the two slides, but the transparency and colour appeared unchanged. The glass cavities also allowed us to compare the e Thect of di Therent thicknesses of the same varnish sample: thinner nearer the edge, and thicker towards the centre. We could then hold the sample over diTherent wood preparations to get a feel for how the pigment and varnish might look on an instrument. This helps to make more informed decisions when varnishing instruments, by providing guidance on how thick a varnish and pigment film needs to be in order to achieve a desired colour saturation.

The ultimate test of pigments is using them on varnished instruments. Using some of these pigments on our own work, we typically use a lean, dark, cooked-resin oil varnish as the medium.

The varnish itself already provides a good amount of colour to the ­final instrument, ranging from an orange to red–brown. Pigments have been found in classic Italian varnishes by various studies on the subject. Our pigments are used primarily to modify the colour of the varnish and increase its intensity. The higher intensity allows the final varnish to be slightly thinner but still possess a dark enough hue. The pigments also add a subtle interest and liveliness to the varnish, which cooked-resin varnishes alone sometimes lack.

JEAN FITZGERALD

Acid-treated madder root bark: espresso method

A quick method for producing a very transparent pigment with an orange-red colour

• Select large pieces of madder root (about as thick as an adult’s little finger or larger) and soak in water for about 12 hours. Remove from the water and peel the bark from the inner part of the root by hand. Gloves may help to avoid staining.

• Soak 55g of madder root bark in a 1% sulphuric acid solution for 20 hours in a glass container.

Rinse the bark under tap water for five minutes and then allow to soak in tap water for one hour.

• Allow the bark to air-dry for two hours and grind the bark into small chips using a coffee grinder.

Fill the espresso machine’s portaflter with half of the prepared bark. Tamp down lightly. Fill the coffee machine with distilled water. Run the machine and pour the extract through filter paper into a container to make 300ml of red dye extract.

• Run the process again with the remaining bark to give a total of 600ml of red dye extract.

• While preparing the red dye extract, heat 400ml of distilled water in a glass container and dissolve into it 100g of alum. Stir until fully dissolved and set aside.

In a separate glass container heat another400ml of distilled water and this time dissolve 100g of potash. Stir until dissolved and set aside.

Now pour the red dye extract into a large plastic bucket and add the hot alum solution. Mix well. Then slowly add the hot potash solution while vigorously stirring. They will react and foam while the pigment forms and precipitates out of solution. Quench the reaction with a large amount of cold tap water.

Allow the mixture to stand for one day. Rinse the pigment using the same method described in the previous recipe. Once dry the pigment is ready to use in varnish or can be stored for future use.