
[Deutsche Version des gleichen Artikels HIER]
We’re celebrating World Pigment Day 2022 and as a special treat from a fellow pigment nerd, I wanted to go on a little excursion into the topic of lightfastness with you. Today I’m showing you the results of a small lightfastness study I made.
If you’re not familiar with the word lightfastness, I’ve already written a (German) post about it here. In short, the term lightfastness describes, how much a pigment (or rather the paint made with it) changes over time when it is exposed to UV light. This change can mean that it fades, grays, changes in hue or undertone, and even a change in texture of the swatch can occur. Usually, lightfastness is tested with the help of the wool scale or by the ASTM in standardized procedures – both of which is not the case here, of course.
But before we go straight to the results, here’s the setting of my little test, for all of you who are interested in the more technical side of things: I’ve made 15 Swatches of watercolors on pure white cotton paper. All swatches were done as uniformly as possible, with a gradient from masston on top of the swatch to a middle tone in the middle to a very watered down area right at the bottom of the swatch. Those quare swatches were then cut in half and labelled on the back for reference. One half I put in a closed manila envelope, which was stored in a folder, right between other papers. That way they were shielded from light, like they would be when they’re in a sketch book. To see what UV light can do to the paints, I hung the other half in the south-west facing window of my studio on the first floor (for our American friends: the second floor). To be exact: From the 1stof March 2021 until to the 20th of March 2022 they were exposed to light. All swatches were attached with clingfilm top and bottom, right behind the double glassed window. This means maximum UV exposure with minimum exposure to other weather.
For most of them I, used the right half of the swatch to expose to light, unfortunately I wasn’t consistent with this for 2 of the swatches. Those I will point out to you in the text below, although it is pretty obvious which is which, since these were two of the less lightfast colors showing greater losses of colour than the others.
Four of these colors were made by big companies. Since I was only interested in their colours with the lesser lightfast pigments, I want to stress that the results of my tests are in no way representative of their whole range. So please, do not project the results of this teeny tiny study of mine to their whole range of paints – I specifically picked the more delicate cases out of sheer curiosity:
For one, I wanted to test the lightfastness of Alizarin Crimson, since I like the pigment and wanted to see how the paint I made from it compares to that of bigger companies. And secondly I picked Aureolin (cobalt yellow), because it is said to be graying over time and exposure to light. Since I try to work with lightfast pigments most of the time, I didn’t buy any of this pigment, so there’s no paint of mine to compare to these two swatches.
All the paints for the other 10 Swatches were handmade by me. 9 of the pigments were labelled excellent in lightfastness – well, we’ll put that to the test. One paint contains a pigment that is extinct – I had heard due to poor lightfastness. Well, let’s test that, too.
But now we’ve heard enough about the experimental setup, let’s dive into the results. We begin with our children of sorrow and go on to the more delightful results from there, which don’t need that much words:



- Alizarin Crimson (PR83): On the big picture (overview) you see three Alizarin swatches: second last row to the left, there’s Schmincke’s Horadam Alizarin Crimson (light exposed side of this swatch is left). On it’s right side, there’s my handmade paint made of Alizarin Crimson pigment and in the last row, right below the Schmincke swatch, there’s the same color made by Daniel Smith. Alizarin Crimson is a deep, cool red. Think about biting into a fresh, juicy cherry – if the fresh juice drops onto white paper, you’ve got pretty spot on the color of Alizarin Crimson. Since this is a lake pigment, it’s lightfastness is not as high as we’re used from other synthetic organic pigments.
My sample test shows very well that all of these swatches loose some of their cool undertone when exposed to light. All test swatches have become significantly warmer in hue and the swatch looks less smooth, more frizzy/grainy, like a color photography with poor lighting. There’s a visible change in color strength of these UV exposed swatches. Mainly the swatches from Schmincke and Daniel Smith loose much of their color strength. The parts of the lightest tone in the swatch are no longer visible, they have faded completely. All that is left in these areas is the white of the paper. The midtones have become way lighter and more patchy. The most color is left in the masstone, as expected, although it is also much lighter and more patchy. The color swatch of my homemade paint fares a little bit better than the others. It has also faded the most in the lightest part, but a light pink remnant shows where the color used to be. The midtone is still visible, though more patchy and more grainy, like I described above. Same goes for the masstone. A result that’s very much to my paintmaker’s heart 😉


- Aureolin (PY40): A light, bright mid yellow that tends towards the cool side. This pigment is said to be graying over time. I had to wait quite some time for this effect to show, even after half a year in direct light, the swatches had only faded a little bit, but not changed in hue or grayed. On the big picture (overview), you see the swatch of Winsor and Newton (W&N) Professional Aureolin in the middle of the last row – this swatch shows the light exposed side on the left. Second last row to the right is Daniel Smith’s Aureolin. If you look at both swatches, the first thing you’ll see is that the color that used to be a bright mid-yellow, leaning to the cool side, has changed into a beigy color, almost like Buff Titanium. The DS swatch has a hint of yellow left in the masstone (if you look closely enough), the W&N shows a little bit more of it left in the masstone until to the mid tone. Both swatches are only hardly visible in the watered down swatch area, a fact that is more pronounced on the W&N swatch where the fading goes way into the mid tone. Tinting strength is a bit better in DS’s swatch, the gradient in three stages is still visible in this one, just in beige instead of yellow. For me, the result is that it was a wise choice not to buy this pigment, even though it the color itself is beautiful when it’s not exposed to light.
Now we come to the more positive side of this little experiment. All following colors are handmade by me:

- Sternennacht-dark blue: Last row to the right you’ll see the swatch of a blue I use for my color called “Sternennacht” (starry night). Normally it contains mica pigments as well, but since I wanted to test the blue, I used it without mica. All in all it behaves well, if you look closely, you’ll notice a slight, subtle change in hue from cool to a warmer tone. Masstone and mid tone have barely changed, the more it’s watered down, the more the color faded. But even that area is still visible well enough – compared to the Alizarin swatches. If you follow me for some time, you know that Sternennacht is one of my babys I don’t share the pigmentcode, that’s why you’ll find none here as well.
From now on we‘ll go through the rows from top to bottom, left to right, since there are no significant changes in lightfastness amongst these colors anymore:

- (Some nondescript experiment made of) Disazo Yellow and Phthalo Green, blue shade (PY?? and PG 7): This color comes quite close to Schmincke’s “May Green” and I’ve tested it because I wanted to know more about my mysterious Yellow pigment. I bought it at the beginning of my paintmaking-carreer, still in the experimental stage, and it contains no pigment code (a mistake I’ve learned from!). There’s a less and one more lightfast version of Disazo Yellow – this seems to be the latter, since it faded only slightly. Thank goodness! And I was interested if Phthalo Green is as lightfast as they say – in short: yes, there’s no change in the green, although there was only a little bit of it used here. All in all there’s just a very slight change in hue of the color (if you look closely enough) which attests to the yellow fading. Strength of color and smoothness have not changed.

- Havana (yellow iron oxide and cobalt green): Even if you look closely, there is no change of color whatsoever in this swatch. The excellent lightfastness of iron oxide (at least this one) and cobalts seems to be proven here.

- Green Earth (PG23): At first glance, this color seems not to have changed at all, however, if you look more closely, there’s a slight, slight change in hue towards the warmer side on the light exposed side of the swatch and the watered down swatch seems to have a little bit more strength on the “dark” side. But these differences seem to be so small that one can call them insignificant.

- Quinacridone Rose (PR122): Just the same as with the Green Earth, the changes are very subtle, leaning towards the warmer side when exposed to light. If you now suspect that this might be a result of the paper yellowing in the sun – no, that’s not the case, the color of the paper stayed the same throughout all the swatches, no change of color at all visible.

- Nickel Azo Yellow (PY150): Only the slightest change in color in the watered down area, but no changes in hue, strength or any patchyness.

- Phthalo Blue, yellow shade (PB15:3): On first sight, it looks like there’s no change at all in this swatch. If you look more closely, you’ll see that the watered down area of the light exposed side is slightly lighter than it’s light protected twin, who also looks slightly – very slightly – cooler in hue than the light exposed side.

- Elfenleuchten (Elf’s light, PB27, PG8): This color didn’t measure up to my expectations: My little experiment cannot confirm the poor lightfastness of the extinct green pigment. There has been no change in color at all in this swatch. It’s a shame that this wonderful, deep pine colored PG8 isn’t produced anymore.

- Venetian Red (PR 102): Again, the change is so subtle that you need to look very closely and twice if you want to find it: The undertone of the light exposed side has changed ever so slightely towards the warm side that it’s almost invisible. Strength of color and evenness have not changed.

- Kornblume (corn flower, PB29, PW4): no changes at all
So, we’ve almost reached the end of my teeny tiny little experiment. If you want to see some of these colors after three months – there’s a post on my IG (@lelles_colors, 3rd June 2021).
And if you’re wondering why the watered down parts of the swatches fade more than the masstone – one of the reasons for that is that the pigments there are not layered as much. In the masstone they can protect each other by providing the one underneath it some “shade” (very much simplified!). Furthermore, the Gum Arabic, which protects the pigments, doesn’t surround the pigment as much in the watered down swatch anymore. These are two of the factors at work which lead up to more fading in watered down areas. A pretty good argument for using (the not so loved) white paint in watercolour, if you ask me. Even if you can make a color look pastelly by using water and letting the white of the paper shine through, you risk more fading of these areas.
If you have questions about lightfastness or my experiment – don’t hesitate to reach out to me, I’d love to nerd about pigments with you 😊
Greetings from Germany,
Nicola