logo

Home>Conservation> Oils > On the Yellowing of Oils

On the Yellowing of Oils

28 thoughts on “On the Yellowing of Oils”

    • Thanks Jean! Definitely some curious results that ran counter to some of our own expectations. Will be interesting to see where this ends up at the 5 and 10-year mark.

      Reply
    • Hi Avery. We did not purposively ignore Lead White as much as simply needing to start somewhere and Titanium White is unarguably the most broadly used white for paints and grounds. Plus it has some unique issues, like its tendency to push oil to the surface. Our hope is to include lead white in future rounds, as well as lithopone (PW 5), and to explore more fully the impact of functional solids as well as stabilizers like aluminum stearate that are used broadly by artist paint manufacturers, although we have chosen to use small amounts of beeswax ourselves. So a lot to look forward to!

      Reply
  1. Dear Sarah,

    Thanks for this valuable work.

    I recently had some issues with yellowing with cold pressed linseed oil which i’d purchased in a raw state and then cleaned myself and allowed to lighten in the sun for 12 months for use with hand made paint and mediums; It mainly seems to be an issue with yellowing in the dark – but it is much more pronounced then when in the past I made mediums with stand oil and damar resin.

    To avoid this problem i’ve switched to hand refining cold pressed walnut oil – this has has made a cleaner and brighter paint but the handling qualities of paint/medium made with this oil is very different than oil paint/mediums made with linseed oil. The walnut oil paint has far less resin like qualities compared with the hand refined linseed which is something I miss.

    Kind regards,
    Darren

    Reply
    • Hi Darren – Thanks for the comment and sharing your experience. In some future iteration, we hope to include more walnut oil, and especially a water-washed one, just as another variation to look at. It is, along with linseed, one of those classic oils stretching back to the Renaissance, but is often talked about has having a thinner, slippery-like feel which some love, but others will miss – as you seem to do – the more bodied feel of linseed oil. In any case, hand-refining oils is always a great way to learn about them and gain some insight into the past. And if we can ever help further, just ask!

      Reply
  2. Dear Sarah,

    This is a very thorough examination of the darkening of oils. I am very appreciative of Golden for providing artists with these results to think about. Interesting (figure 5) how bees wax (sample 4) reduces surface oil to inhibit yellowing, and so does low cobalt drier (sample 7) but not high cobalt drier (sample 6). I was under the misunderstanding bees wax increases yellowing. This is an area palette knife painters could consider, as knife work has had the reputation to increase yellowing by drawing oil to the surface. What were the concentrations of cobalt drier and bees wax used? I am interested to see where this study heads to in about 10 years time, especially with stand oil and bees wax. Thank you once again.

    Kind Regards

    Sam

    Reply
    • Hi Sam –

      Thanks for the compliments about the thoroughness of our testing and the sharing of the results. Feedback like yours means a lot to us. Also, know that the spirit of open-ended inquiry and a desire to make as much information available as we can are cornerstones of our culture here. In terms of your questions, you are correct that beeswax has that reputation but we think it is tied more to the fact that, at higher percentages, it increases transparency enough that the inherent yellowing of the oil is not masked. Given the opacity of titanium white, the very small percentage of beeswax in the mix (less than 3%), and its role as simply helping bind the oil and stabilize the paint, it has a net positive effect overall. And at substantially higher levels, say 10-20% or more, you will start to produce much softer paint films that are more susceptible to solvents, slower drying, and, yes, more yellowing! The notion that palette knife painting by itself causes oil to come to the surface is intriguing and I will need to look into our testing around thicker applications to see if we find a similar issue, or at least to add that variable to the mix. As for the cobalt drier levels, the differences in the thinner films was very slight and it’s hard to capture those type of subtle temperature shifts in an image, so I think the difference you see is a touch larger than you would sense in real life. But it is true that in the thick disks one did see a clearer difference. My guess here – and it is something that would need to be proved in further rounds – is that the higher level of cobalt drier causes the top of the thicker application to skin over faster and that this skin in turns acts as a diffusion barrier to oxygen and delays the crosslinking lower down, thus allowing more unbound free oil to be mobile for longer. But just a theory. Clearly, too, you can see that the cobalt drier at both levels increases surface wrinkling in thicker applications for similar reasons – the quick skinning-over caused principally by cobalt (a surface drier) happens when the paint is at its original full volume, then as it loses volume from off-gassing and shrinks, you get those tell-tale puckered and wrinkled surfaces seen in thick oil paint applications. As for the exact percentages of the drier, that is something we do need to keep proprietary as it takes a long time through trial and error to start to locate those ranges. We hope you understand. But certainly, in general, you should find roughly similar results if you push levels towards the lower and upper range recommended by whatever cobalt/manganese drier you happen to be using. Just keep in mind those percentages would still not be the same as ours since we are working with extremely concentrated pastes that are not available to painters directly. For reasons of safety and ease of control, the driers you find in art stores are provided in a more diluted liquid to be added dropwise.

      Hope that helps!

      Reply
  3. Excellent information, thanks Sarah & Golden. What % volume addition of barium sulfate or calcium carbonate to TW would you have used? Wondering how to best balance the risk of yellowing caused by increasing transparency against TW alone, while acknowledging it’s not definitive yet

    Your standard tube formulations don’t usually contain beeswax or other stabilisers? (does Williamsburg TW including unbleached titanium normally contain beeswax)?

    Reply
    • Hi Bob –

      Thanks for the questions. Because much of the testing we shared was based on and working towards our own optimizations, we are not able to share the percentages as that is proprietary. It becomes part of the art of paintmaking and things gained through multiple trials tested over the years. And keep in mind that beyond just percentages, the type of calcium carbonate or barium sulfate would also impact the results. So it’s a bit complex. The main point we wanted to show is that these inert additives have an impact on the results which can often be positive and that frequently there is a synergy among different components that is greater than any one of them alone. Not completely unlike the fact that Titanium-Zinc is whiter than either Titanium or Zinc separately. As for our standard tube formualtions, we do use small amounts of beeswax not only in Titanium White and Unbleached Titanium but in general across the line at extremely minute amounts (generally in the 1-2% range) representing the bare minimum needed to prevent separation over longterm storage and, as the testing also showed, there are times when it also contributes to other qualities, such as less yellowing. Hope that helps!

      Reply
  4. Thanks very much Sarah, helpful as always. I also wonder, does Williamsburg add a small amount of bodied linseed oil to pigments like raw umber to help prevent sinking in? Seems like a good idea as sinking in can happen even unexpectedly on a well prepared support. I’d prefer that to me adding extra oil to the painting

    Reply
    • Hello Bob,
      Thank you for the comment. Sarah is on sabbatical, so I’m happy to answer your question. We do not add bodied oil to any of our Williamsburg paints. Although it is an interesting idea, it may be a bit of a slippery slope trying to predict what characteristics to accentuate through formulation. In the end, we prefer to keep our ingredients as simple as possible and leave the rest to the creativity and technique of the artist.
      We have also seen sinking in and matting of oil surfaces even on well prepared and on non-absorbent test surfaces. The thickness of the stroke or application, the pigment and/or the binding oil can all have an impact on the surface sheen of the dry oil paint. You have likely seen this, but just in case, here is a link to Sarah’s article on oiling out and dead spots: https://justpaint.org/oiling-out-of-dead-colors-in-oil-paintings/
      Take care,
      Greg Watson

      Reply
    • In our testing with straight oil and whites made with different types of oil, the samples tend to settle into a degree of yellowing withing a year or two. whether or not you will be able to see yellowing in a painting depends on the dominant colors in the work and how well they can hide the yellowing of the oil. Some colors are more vulnerable than others to yellowing – whites and blues in particular.
      Best Regards,
      Greg Watson

      Reply
  5. Thank you Sarah, and great to read this.

    I wondered if you had any advice regarding how to deal with a yellowed area of a painting (created with 50% walnut oil and 50% Williamsburg titanium white)?

    I have a painting that I had to put in dark storage,for about a month, just a few weeks after it was painted, and it yellowed noticeably. Do you know whether this is reversible and if not, whether a thicker (more pigment heavy- straight out of tube) white paint, could be used to paint over the yellowed area (without having the same thing happen again?)
    I wondered if there might have been an adverse reaction between the alkali processed linseed oil titanium white and the walnut oil medium?

    Thanks for your help!
    best,
    Daniele

    Reply
    • Hello Daniele,
      We would suggest putting the painting in direct sunlight and then checking periodically to see if the yellowing wanes with exposure. It is likely to take more than a couple hours of sun, though.
      Warm Regards,
      Cathy

      Reply
  6. Thanks so much for making your research available to artists, it’s so valuable. Could you outline the two different water washing methods used? Also curious which “historical texts” these washing methods come from?

    Reply
    • Hi Christopher –

      Thanks for the comment and we are so glad that you found this helpful. In terms of the water washing, the two oils were from outside sources and not prepared by us. One is from a commercial company that specializes in water-washed oils based on ‘historical texts’. Unfortunately, they do not share what those texts are or the exact process that they follow. Obviously there is a range of hand-washing methods that can all claim to be historical, but the current field of handwashing is also small enough and well-researched enough that everyone is generally aware of the same information. While certainly possible that an exotic or rarely used method is being employed, I have tended to believe that it is some variation of the water-washing methods described in de Mayerne’s “Lost Secrets of Flemish Painting” and/or the variation on it laid out in “Methods and Materials” by Charles Eastlake. The second hand-washed sample was supplied to us by an independent scholar and specialist using the general sand and water method developed by Tad Spurgeon, who is well regarded in this area and also draws from the above texts, as well as his own research. On a personal note, I have repeatedly followed Tad’s sand and water process myself at different times with results that at least appeared similar, although I think even Tad would acknowledge that there is always some variation involved. In the long run, the inclusion of just two samples does not make for a definitive conclusion about water-washed oils but we hope they represent at least what is typically being used by many painters interested in these things. Hopefully one day we can drill down further and generate a range of oils ourselves so we can more fully speak to the exact methods and processes used.

      Hope that helps.

      Reply
      • Thanks Sarah, that’s all very helpful, I am familiar with these texts and with Tad’s extensive work on the subject so I’ll look back at them. I am always looking for practical and proven ways to improve the durability and look of paint so it’s great to see your team approach these topics so rigorously, despite doing many of my own tests over the years it’s hard to trust any of the results without repeated confirmation from others, and despite my best efforts I doubt I can control as many variables as your lab.

        Related to yellowing or darkening with age, I’ve never been able to find any information on the quantity of manganese in umber pigments and how it compares to the quantity in the drier additives. Is there concern for the manganese in umber pigments to darken or yellow on their own or in mixtures with white? I’ve always considered the iron oxides to be the gold standard of stability, expecting that the increase in transparency with aging might cancel out any mild darkening in darker umber passages. Is there concern for significant changes/yellowing with age among siennas and umbers due to aluminum soaps, manganese, etc? Thanks so much.

        Reply
        • Hi Chris –

          There is a good article analyzing manganese-containing pigments used in art, which includes chemical analyses and Mn % of various commercial oil paints: Study of the chemical composition and the mechanical behaviour of 20th century commercial artists’ oil paints containing manganese-based pigments As you see, in Table 3, the Mn % in Umbers ranged from 3-11%. I would not get too locked in to which brand had what – both of these are naturally mined and the percentages of all the components will invariably be, well, variable! But it at least gives a ballpark. Trying to compare this to the percentages in commercial driers is complicated. Commercial Mn driers will have concentration ranges from 6-10%, but will be added into the paint at a very, very small fraction well below 1% by weight. Then how available the Mn is in the pigment, versus the metallic salts, and the complex range of other metals that the umbers bring along as well, and I am not sure a direct comparison is practical without a lot of analysis. In terms of yellowing, however, I have not seen evidence that the Mn in umbers would contribute to it. And of course, as we shared in these tests, despite the often voiced concern about driers causing yellowing in whites, at least in terms of Titanium White we found the opposite – that the judicial use of driers actually aided in preventing yellowing. Finally, we have never seen reports raising concerns about the yellowing of umbers or oxides. By far the greatest concerns with these natural earths has to do with hydrolysis, where interaction with humidity causes a weakening of the paint film. You can read more about that in this excellent article by Marion Mecklenburg: The Influence of Pigments and Ion Migration on the Durability of Drying Oil and Alkyd Paints

          Hope that helps!

          Reply
  7. Hi, reading.this again, thanks again for.the information. Assuming the drawdowns are done with a palette knife, this causes more oil to rise? Or are tests swatches applied with a brush?

    Reply
    • Hi Bob –

      The drawdowns are done with a drawdown bar, which you can think of as just a very controlled ‘palette knife’ like application done at a very specific, uniform thickness. So definitely not brushed on, which is harder to standardize from person to person, and more difficult for a spectrometer to read due to the surface texture. I am curious, though, about your comment of a palette knife causing more oil to rise to the surface. It is not something I had heard of before, so looked for some reference to the phenomenon but came up empty. Is this something you have seen or is there a source you can point us to?

      Reply
  8. I am wondering what version of linseed oil yellows least…Regular Linseed Oil, Cold Pressed Linseed Oil, Refined Linseed Oil or Sun Thickened Linseed Oil. Would you please answer this for me? Thanks so much!

    Reply
    • Hi Mary – Thanks for the question. In our testing and experience, Cold-Pressed Linseed Oils, while they can sometimes do quite well, are just very variable as a product and can differ not just from manufacturer to manufacturer, but even batch to batch. Sun-Thickened Linseed will yellow the most among these options, and really its main attraction and use are that it dries quickly, levels, and is viscous. If wanting something with a similar feel, but don’t mind that it dries slowly, then Stand Oil is one of the least yellowing options among linseed oils. Not sure what you are referring to by Regular Linseed Oil since nearly all of the linseed oils you will find for artists’ use, which is not labeled specifically as Cold-Pressed, is Alkali Refined and is what most people would think of as being the regular and most common choice. And really that is also the one we would recommend as being the most consistent and reliable. Hope that helps!

      Reply
  9. Dear Sarah,
    you seemed curious about palette knife causing oil to rise to the surface. From my experience, the oil is somehow drawn to the top when a palette knife is used, compared to brush work. I wonder if it is a capillary action or the pigment sinking. It seems to happen over a few days/weeks, just something I notice with my work as the paintings appear glossier with time. I am not sure if Vincent Van Gogh encountered the same problem, as he asked Theo to wash down his paintings after they had dried to remove the excess oil. I have also read about palette knife use and rising oil somewhere but cannot recall where.

    Reply
    • Hi Sam –

      It is true that titanium white by itself does not bind the oil very strongly within the paint film, leading to oil being able to exude out, although one can improve its performance through the use of other materials in a fuller formulation. We even show that in the article – see image 5, especially the dramatic examples on the thicker 60 ml disks on the bottom row. But as to why a palette knife would influence or cause it remains a mystery. As for Vincent Van Gough, he would not be painting with titanium dioxide, which did not come into general use until the 1920s. Anyway, if you do find anything that speaks to the phenomenon let me know, and in the meantime, we can add this to our list of curious things to test.

      Reply

Leave a Comment

*

css.php