Previous: Ann Walsh's Mechanics of Color

Oil Refinery

Post #1826 • November 8, 2018, 9:57 AM • 2 Comments

[Image: ]

Earlier this year I was struggling with a painting. I was working in acrylic, and something about the pigment load was bothering me. I tried spiking the white with additional pigment. That helped, but by the time the surface looked how I wanted it to look, the dried paint was so underbound that I could crack it off with my fingers. It hit me: there's a type of paint that has a lot more pigment load than acrylics. They're called oils. I was using Golden acrylics, which as far as I'm concerned is the best stuff that money can buy. But the bottom line is that an acrylic polymer is ten times the size of an oil triglyceride and pigment disperses into each accordingly.

Me being me, that realization sent me down a rabbit hole of research. I hadn't painted in oils in several years, having extracted myself from what I had come to think of as a stale marriage with a palette knife. The kind of volumes I was interested in obliged me to work with an impasto medium, Gamblin's Galkyd Gel. Alkyds are one of those things that artists put up with because they have to. Both the working properties and the finish are mucilaginous. They're thinned with volatile solvents that are unpleasant and unhealthy to breathe. Galkyd's is petroleum naptha, which in my opinion makes turpentine smell like a bed of wildflowers in comparison.

There's a pile of newly available research into traditional methods that completely eliminate volatile solvents from the studio. Enormous credit for anthologizing and testing those methods goes to Tad Spurgeon, whose Living Craft is a treasure. I learned of Spurgeon and another figure in this world, Louis Velazquez, via Jane Pack at the Aegean Center. Jane has for a few years been delving into this material and has been pleased with the results in every way. Not only are they better aesthetically, but gone is the classroom with an open jar of turpentine for every student.

The basics of the setup are simple, but it's so wholly unlike how I was taught to use oil paint that I thought it worth posting about. You wash your brush out into safflower or walnut oil. Safflower is cheaper, but you have to be careful to wipe the brush clean before putting it back into your paint because it's a very slow drier. You thin your paint by adding linseed oil to it. You store your brushes downward in walnut oil. I rigged up a box full of empty cat food cans and put oil in them so I can lean the brushes against the edge of the box. That way they're not sitting right down on their points. Instead of getting harder and harder, as they do when you wash them in turpentine and let them dry, they get softer and softer. The procedure goes through significantly more rags than a turp-driven method, and oily rags have to disposed safely to avoid combustion, but the usual precautions apply in that regard.

There are many further refinements, but that is all that's necessary. One can finish a dried oil painting made this way with a coat of damar varnish. There are modern damar replacements available that seem to be better behaved. But after 500 years of conservation, we know so much about damar that it affords no surprises. Damon Lehrer, a friend of mine here in Massachusetts who is a fine realist painter, tells his classes that the most archival method is to make good art. If people value the painting, someone will figure out how to preserve it.

For a long time I used a medium gleaned from Ralph Mayer, equal parts of stand oil, damar varnish, and turpentine. This did nothing well. It didn't dry especially quickly, it didn't hold edges, it seemed to suck dust out of the air, and it darkened. It turns out that's way too much damar for a painting medium. Also, putting a gloss agent in an intermediate layer of paint is like wearing lingerie under a burqa. You may feel differently knowing it's there but it has no effect on the surface. As I learned from my residency at Golden, the glossy acrylic mediums are more transparent than the matte mediums, but you can put a final matte varnish over the top that will completely eliminate the gloss while preserving the transparency. Oil, in contrast, is plenty transparent even without damar. Historically, where damar appeared, it had been added by the drop.

Then there's this nonsense with the turpentine. "Solvent is unnecessary in a traditional oil painting process," Spurgeon explains.

The opening frotté, heavily diluted with solvent, is a convention which came about from a desire for alla prima brio, combined with an overemphasis on the lean principle. This overemphasis occurred because inferior mid-20th century commercial oils darkened readily. When overused, the lean principle can in fact become self-defeating in terms of film strength and the ability to to resist oxygen and humidity, even to refract light. Many alternatives to this approach exist in the older practice based on the stability of high quality oil as the basis of the system. As long as no resin spirit varnishes such as damar or the balsams are used, brushes can be kept on their sides in a slow drying oil such as walnut, wiped on rags before, after, or during use. Even small amounts of a balsam medium, or a hard resin varnish, can be incorporated into this system with no need for solvent as long as the brushes are cleaned in oil well.

This insight played into my review of the Edvard Munch show at the Met Breuer back in January. God as my witness, I had never heard of anyone storing brushes sideways in oil until I saw it in practice during a 2017 studio visit with Petey Brown.

The topic of oil quality deserves expansion. Pressed flax seeds produce an oil that is suitable for painting but for one defect, a gummy component that yellows all the way to a toasty brown. One filters the gum and conditions the rest of the fluid to make linseed oil. For much of the 20th century, even for oil destined for use by artists, the commercial methods of filtration and conditioning destroyed fatty acids that lend desirable working properties and finish. Even now this is often still the case.

One option is to insist on the best oil available. Golden is open about the process by which it offers its linseed oil under the venerable Williamsburg name. They degum with citric acid and heat under a vacuum, de-acidify with sodium hydroxide (basically—sorry—lye), bleach with bentonite clay, dewax with diatomaceous earth, and finally add back the fatty acids. You can read about this in detail.

Another option is to refine your own oil. This is not especially difficult. Cold-pressed flax oil preserved by refrigeration is available at the health food store. I tried a recipe from Living Craft using spring water and Smirnoff. This takes advantage of the fact that the mucilage is soluble in ethanol and the ethanol is soluble in water.

An equal volume of oil and forty percent (eighty proof) ethanol are placed in the jar, filling it one quarter to one third full. The jar is then shaken thoroughly. The mixture emulsifies readily. The jar is shaken repeatedly over the course of a day, the more shaking, the better, At the end of the day, an amount of spring or distilled water equal to at least twice the volume of the ethanol and oil mixture is added, and the jar is shaken again. The following morning, the water-ethanol mixture is an opaque white, the clear oil has risen to the top of the jar, and can be removed.

This can then be conditioned further with light, exposure to air, heat, et cetera. I left mine in a jar topped with cheesecloth on a windowsill for seven weeks. During that time neither the color nor the viscosity changed perceptibly, which is not to say that in seventy weeks they wouldn't either. It was once common practice to leave batches to cure for years. My results are pictured above, photographed in the sun while sitting on my woodpile. It was probably fine to paint with upon decanting.

Spurgeon has a half dozen other linseed oil recipes, each resulting in a product with particular advantages. Some dry faster, some are paler, some are thicker, some take less time, some (like the above) require less paraphernalia. The aforementioned Louis Velazquez has worked out a refinement method involving psyllium husks that he derived from Pacheco. This was the great Spanish humanist and artist who was the father-in-law of Diego Velazquez (to whom Louis is not related). The contemporary Velazquez doesn't divulge the details of the method on his site, but Pacheco's recipe is recorded in Artists' Techniques in Golden Age Spain: Six Treatises in Translation by Zahira Véliz, published in 1986 by Cambridge University Press. Fortunately this otherwise unobtainable book is in the reserve collection at the Boston Public Library. I checked it out yesterday. Pacheco's instructions for refining oil, in their entirety:

Here we will treat how to purify linseed oil and remove its yellowness so that it will serve for white and blue. Take a glass flask and a pound of clean, clear linseed oil, three ounces of agua ardiente que llaman de cabeza, and throw in two ounces of lavender seeds. Put this in the sun for fifteen days and stir it twice every day, and it will become clear and purified. Strain it into another glass flask. Then it can be used to advantage in whites, blues, and flesh. More or less than the quantity of one pound of oil can be prepared, adjusting the amount of the other things in respect of what has been said.

Véliz (now Dr. Véliz Bomford at the Museum of Fine Arts, Houston) interprets the phrase left in Spanish as "indicat[ing] the first spirits drawn from the still, suggesting that the alcoholic content would be relatively high." One surmises that twelve parts oil to three parts vodka to two parts lavender seeds by volume, stirred twice a day for two weeks in the sun and strained, won't be far off the mark. Psyllium husks undoubtedly will remove the mucilage and are easy to come by as a digestive aid. Nevertheless there may be a decided advantage to the lavender seeds. Soaking them in ethanol will distill a small amount of spike oil, which would speed drying time and tighten quickly as a glaze layer. Spike oil literally smells like wildflowers. Based on the proportions in Spurgeon's recipe I think there would be no harm in increasing the volume of alcohol.

White and blue come in for scrutiny because the yellow color of the linseed oil is most apparent in them. Also, Pacheco is clear that he's not talking about ultramarine, but a "Santo Domingo blue" that Véliz guesses was a naturally occurring copper blue in southern Spain. As a pigment it was prone to clouding, as Pacheco describes elsewhere. Modern titanium white is so icy that it might benefit from some warmth. Synthetic ultramarine is a stronger tinter than its lapis lazuli predecessors, to say nothing of whatever Santo Domingo blue might have been. (Their mileage may have varied. I once used a genuine lapis watercolor put out by Daniel Smith, and it was one of the weakest tinters I had ever handled. It was explained to me that a tube of modern lapis at maximum strength would cost hundreds of dollars.) The yellow of linseed oil is mitigated by storing the painting in indirect sun, though the material itself is known to yellow anyway. Williamsburg was moved to launch a line of colors ground in safflower oil, which is markedly paler.

One can test for mucilage in the oil by throwing it in the refrigerator for a few hours. If it clouds, it needs to be refined further. My batch of oil was clear after four hours in the refrigerator. Unconvinced, I put it in the freezer. Don't do this. It turned into a gel, though to its credit it still didn't cloud. A half hour at room temperature returned it to a fluid state.

Six Treatises has a note from Antonio Palomino y Velasco as well:

...linseed oil may be clarified by placing it in a flask and adding a portion of white lead powder. Mix this very well so that the oil appears to be white, and leave it in the sun and in the night air. Mix it again twenty-four hours later, and do the same thing three times more, and then use it, because with more repetitions, it will thicken.

This passage was indexed under "refining linseed oil," but I think it's refinement in the sense of enhancement. It's not going to degum flax. Nevertheless this seems like an interesting thing to try as a possible analogue of Golden's bleaching step with bentonite. (Powdered white lead is nothing to goof around with. Take the appropriate precautions.)

None of the homemade recipes call for lye, which strikes me as curious. Lye is easily obtainable (you can't make a decent bagel without it, though at pH 13 in pure form you don't want to give it to the baby to play with), and linseed oil is known to be acidic enough to merit the need for base. This may be something worth exploring as well. Spurgeon does go into chalk additions, and calcium carbonate is also alkaline. A pinch of chalk in any of these oils will do some good.

You can also add a lot more than a pinch. Anywhere from one to three parts of chalk or marble dust to one part of oil results in a putty with many virtues. The thicker formulations are good for impasto, the thinner ones for intermediate layers or glazing.

Lastly, this opens the possibility for mulling pigment into the oil, batching up your paint for the day as one would do with egg tempera. Indeed, this is the direction in which Jane is headed. Having made gouaches and watercolors and worked with egg tempera for a few years, I've learned that making paint is pretty easy. Storing paint is hard. Cobalt violet in store-bought gouache is uniformly terrible across brands. The pigment has tendency to fall out of solution and turn hard in the container, what manufacturers call a crash. To prevent crashes, they add dispersants, which have their own lousy working properties. They also displace pigment by volume. But if you mix up the cobalt violet you need and use it, it's beautiful and works as well as any other pigment.

That raises a question about the advisability of tubed oil. If we're serious about this, shouldn't we be optimizing our paint around painting, rather than storage? The answer is perhaps, but perhaps not. I would prefer for someone else to make my flake white, for one. But the point is to have a choice about the matter, and to make trade-offs in an informed fashion.

All of this verbiage might lead you to believe that I've been looking at someone like Jusepe de Ribera lately. In fact I've been looking mostly at Léger, in reproduction. I haven't seen that many Légers in person, to be honest. But the ones I have, such as Divers, Blue and Black at the Met, have indifferent surfaces. If Spurgeon is right about the general collapse of materials standards in the mid-20th century, it would explain a lot. But on the other hand, it doesn't matter. Divers has so much graphic power that it doesn't need a luscious surface to function.

The advantage of coming at this pursuit as a modernist is freedom from the misconception that some Lost Secret of the Old Masters™ will deliver us from artistic evil, or at least inadequacy. It would be easy to make fun of the neo-traditionalists on this point, but some similar error of belief plagues all kinds of artists, and that includes the conceptualists. For that matter it includes some of the modernists. The basic problem is, and shall always be, to identify a worthwhile project and pull it off. That said, I'm fond of Renoir's quip that being a good craftsman won't stop you from being a genius. While we try to establish what geniuses we are (spoiler alert: we're not), here's to not whiffing naptha all day.



Dana Gordon

November 8, 2018, 4:43 PM

Keep it simple: just use the best quality oil paint and don't add anything to it.

Don't bother cleaning brushes. It's a waste of time and wears them out. I wipe mine off, swish them in a little Turpenoid, then keep them in a little Turpenoid.



November 8, 2018, 7:24 PM

That's fine if it works for you. You're no better off touching or breathing Turpenoid than you are touching or breathing turpentine. Arguably it's worse because you can't smell the buildup of vapors.



Other Projects


Design and content ©2003-2018 Franklin Einspruch except where otherwise noted