From blue to red
April 27, 2018 5:04 AM   Subscribe

The Quest For The Next Billion-Dollar Color The world has never had a truly safe, stable and bright red pigment. The trail may start with YInMn, the first blue created in two centuries
posted by fearfulsymmetry (31 comments total) 30 users marked this as a favorite

I've never even heard of this element. What a country!
posted by thelonius at 5:06 AM on April 27, 2018

I've never even heard of this element. What a country!

There's a village in Sweden that gave its name to yttrium, erbium, terbium and ytterbium.
posted by sukeban at 5:28 AM on April 27, 2018 [8 favorites]

Just so long as they keep it out of Anish Kapoor’s hands.
posted by juv3nal at 5:29 AM on April 27, 2018 [18 favorites]

YInMn was the first blue pigment discovered in more than 200 years.

I was about to take exception, but International Klein Blue was an advance in resin, rather than pigment, science.
posted by logicpunk at 5:33 AM on April 27, 2018 [6 favorites]

It is the first blue discovered in centuries, but of course it is not the first new primary color in that time.
posted by delfin at 5:54 AM on April 27, 2018 [3 favorites]

There's a village in Sweden that gave its name to yttrium, erbium, terbium and ytterbium.

Then there's the combination of americium, californium, and berkelium, like someone's trying to give an address. Next discovery is milviastreetium?
posted by traveler_ at 6:43 AM on April 27, 2018 [5 favorites]

Ytterby also gets credit for scandium.

FWIW, stable means non-bio-degradable. Embrace decay, people.
posted by ocschwar at 6:48 AM on April 27, 2018 [2 favorites]

Oh, I have
posted by thelonius at 6:48 AM on April 27, 2018 [12 favorites]

FWIW, stable means non-bio-degradable.

As the article points out, iron oxide is stable. It's just not vibrant, but it is stable, safe, and non-toxic.

A stable bright red pigment that is also non-toxic is the holy grail. Right now we're at "stable, bright, non-toxic, pick two."
posted by explosion at 7:05 AM on April 27, 2018 [6 favorites]

I've always wondered (and asked a few material science folks) - can we predict or model any of this ? My original question to friends was "can we know if something is transparent/translucent a priori?" and this looks like a subset of that. (I always got hand-wavey answers about the bond length, lattice/crystal structure, wavelengths etc, that were not quite satisfactory - I'd get it, but not "OK, but can't we model or predict it?")
posted by k5.user at 7:12 AM on April 27, 2018

So interesting! thanks for posting
posted by k8bot at 7:54 AM on April 27, 2018

Great read, thanks for the find!
posted by like_neon at 8:26 AM on April 27, 2018

No lead or uranium? No thanks
posted by Existential Dread at 8:28 AM on April 27, 2018 [1 favorite]

"OK, but can't we model or predict it?"

I met a guy once whose job was to come up with new colors of laser. He had a crazy apparatus for reproducing the conditions deep inside the earth or in outer space or with 600,000 volts running through it, etc. He told me that while it was technically possible to predict if a given crystal (well first that it would even form a crystal) would lase, in practice it was much easier to just use trial and error methods, as the math is really that complicated and time consuming, even with computers, and the more atoms in the crystal (unit) or molecule, the more complex the equations get. I imagine this is true for any number of properties for a material (conductivity, transparency, etc), and I could easily see color being even more complex due to the interference of things like structural color and etc. (And then you have to figure out if it will work as a bind to other materials and remain stable etc. Which is why, historically, most pigments are discovered by 'surprise' (with sometimes additional colors soon after, from variations on the same chemicals))
posted by sexyrobot at 8:51 AM on April 27, 2018 [4 favorites]

One my old chemistry teachers had worked in industry on the chemistry of paint before becoming a teacher. I remember him saying how colour is actually pretty rare - the vast majority of things are either silvery metallic metals, white powders or clear colourless liquids.
posted by fearfulsymmetry at 9:25 AM on April 27, 2018 [2 favorites]

That is true. To have a color in the visible spectrum a molecule needs to be able to absorb photons in a reasonably narrow band in the visible spectrum, which is usually caused by a specific electron transition between molecular orbitals. Basically, a molecule absorbs the photon, and the absorbed energy causes an electron to transition from a low energy state to a higher energy state. Metallic complexes and conjugated organic molecules (like say the porphyrins in hemoglobin or chlorophyll) have these very specific electronic transitions, which correspond to vibrant colors. Most other compounds (ionic compounds, non-conjugated organics, and so on) do not have these transitions, although compounds like NOx or gold have their own internal physics that give rise to coloration.

Compounds can non-specifically absorb or scatter visible light, but that usually leads to opacity without defined color.
posted by Existential Dread at 3:40 PM on April 27, 2018 [1 favorite]

Chemistry post! I can comment! So, I don't work in dyes, pigments, or anything like that. But I know people that work in related fields, and I do work with emissive compounds, which are related.

You can try and predict the absorption spectra. Depending on where on the periodic table you are, you'll have more or less success. Organics, you can probably do pretty well. Organometallic complexes of transition metals? Pretty well. When you get into solid-state inorganic complexes, things get more hairy though, as you get more types of interactions to worry about. The hard part isn't calculating it though: it is designing the system.

Think of it this way: Your program will tell you what colour any chemical you draw will be. The hard part is figuring out what to draw, and then actually making the thing you've drawn.

So you start with some ideas, say you know this metal has a colour that is close to what you want. Then you work on adding things to it to adjust its current colour to be more like what you want. Sometimes this works, sometimes it doesn't.

What I'm not happy about is us wasting Indium on this: Indium is a really important metal for electronics, and as I recall, solar cells. It seems kind of wasteful using it on paint, you know? Not as bad as helium balloons (Should have been banned years ago.)

I am a bit surprised at them going through inorganics like this though. It is a really old fashioned approach (though I don't mind, as an inorganic chemist myself), it is just that organics are usually the main thing you look at when dealing with colour these days. Easier to tune the things you want, and you can get some pretty stable stuff (Indigo, pthalo blue and green)

Honestly, I'd be looking at organometallics. There is a lot more space there then pure organics, and some of them can be quite stable, see the above phthalocyanine compounds.
posted by Canageek at 4:25 PM on April 27, 2018 [3 favorites]

Hee! I worked in phthalocyanines for my grad work lo these ten years ago, and by far my favorite was titanyl phthalocyanine (the titanium ion at the center had an oxygen coordinated in the axial position). It formed long, rodlike crystals of a deep and vibrant violet color, but when crushed or smeared on a kimwipe or surface it would become a deep navy blue. I have no idea what kind of interactions led to those different pigmentations when physically deposited in different formats, but it was a really cool effect.

My least favorite compound by color was dysprosium phthalocyanine, which formed this weirdly pukey green color. I suspect it had some level of trapped solvent or contaminant that was influencing the coloration, though.

I stained a few shirts and shoes with rich blues from errant droplets of dissolved phthalocyanines (pendant organic groups rendered the molecules soluble in various organic solvents).
posted by Existential Dread at 4:41 PM on April 27, 2018 [1 favorite]

Hello, chemists.

Why did the article say that carbon dyes break down? Is it because of biological degradation, or because of the sun somehow? Or something else?

posted by clawsoon at 4:44 PM on April 27, 2018

Organic compounds are susceptible to photobleaching in sunlight, meaning their molecular structure gets degraded from long exposures to near UV and UV light. There might be other weathering and bleaching processes, as well.
posted by Existential Dread at 5:19 PM on April 27, 2018

Does it have something to do with the length of carbon bonds and the length of UV wavelengths?
posted by clawsoon at 5:21 PM on April 27, 2018

Yep! The shorter the wavelength, the higher energy the photon has. UV photons can be energetic enough to break carbon-carbon bonds, I believe, although I don't remember the exact energy levels off the top of my head.
posted by Existential Dread at 5:35 PM on April 27, 2018 [1 favorite]

but what about the chemical composition of the most important blue: metafilter blue

(I would totally buy a set of vials of elemental Metafilter-inspired pigments.)
posted by bonje at 7:11 PM on April 27, 2018

I'm only half-joking when I tell people I went into inorganic chemistry for the colours.

(Now off to actually read the article.)
posted by invokeuse at 4:28 PM on April 28, 2018

Red 254, aka Ferrari red, for example, is safe and popular, but it’s also carbon-based, leaving it susceptible to fading in the rain or the heat.

Aka PR254, aka Diketo-pyrrolo pyrrole.

The watercolor pigment resource:

TOP 40 PIGMENT [click for reflectance profile] Pyrrole red PR254 is a very lightfast, semiopaque, highly staining, dark valued, very intense red pigment; only three manufacturers offer it worldwide. Unrated (!) by the ASTM, industry and my own tests assign it an "excellent" (I) lightfastness. In watercolors PR254 shows a small drying shift, holding its lightness but losing 15% saturation. The average CIECAM J,a,b values for pyrrole red (PR254) are: 38, 85, 38, with chroma of 94 (estimated hue purity of 78) and a hue angle of 24. Most brands show a strong shift toward blue in the undertone, which creates a rather dull, bruised color in tints and a dulling of diluted color mixtures.

PR254 is becoming increasingly popular with paint manufacturers as the primary replacement for less lightfast middle red pigments, such as the naphthol reds, and the more polluting cadmium pigments. The pigment is very consistent in hue, saturation and texture across manufacturers; it blossoms readily and in most brands is extremely active wet in wet. I like Winsor & Newton winsor red for the depth and clarity of color, and because it is relatively less opaque and is uniquely inert wet in wet. The Daniel Smith pyrrol red is noticeably darker valued, more opaque, and less saturated in tints. The Rowney Artists permanent red lost much of its orange tone after five weeks of sunlight exposure. The Maimeri paint is surprisingly and unacceptably less lightfast. — A widely offered middle red that can is close to a scarlet hue (Schmincke sells it under the label "orange red"). It is a plausible substitute for cadmium red paints in masstone, but lacks cadmium's radiance in tints; it is a good mixing partner with synthetic organic yellows, but these mixtures seem less attractive to me than the equivalent mixture of cadmium red and yellow. Substitutions. A very good color match, with better transparency, can be mixed from quinacridone rose (PV19) and naphthol scarlet (PR188). See also the section on pyrrole pigments.
posted by sebastienbailard at 9:04 PM on April 28, 2018 [2 favorites]

Now I have a big crush on this watercolor pigment description page. So much passion and knowledge.
posted by away for regrooving at 10:07 PM on April 28, 2018 [1 favorite]

You'll like the characteristically thorough tutorial on laying a wash.
posted by sebastienbailard at 10:40 PM on April 28, 2018

Hmm. Going by this person's 4-year lightfastness test, PR254 isn't much more lightfast than a napthol red.
posted by tavella at 10:43 PM on April 28, 2018 [1 favorite]

Man, I'm such a sucker now for pigment research stuff; I remember hearing the initial hullaballoo back in 2009 or so about Subramanian's blue and thinking "that's interesting", but with the painting stuff I've been doing more recently I've started taking seriously both the history and the practical implications of pigment variations and their durability and composition and so I was nodding along to this like "yeah...yeah!"

If you're interested in an engaging (if at times a bit arch and/or salty) narrative of the history of pigment discovery and use, especially but not solely in the context of painting, I will heartily recommend the book Bright Earth by Philip Ball. I read it last year and enjoyed it thoroughly, and came away with a much improved sense of the overall arc of all the varyingly unstable, toxic, and just plain gross pigments and pigment-production methods behind classic art and modern manufacturing.
posted by cortex at 3:34 PM on April 29, 2018 [3 favorites]

Honestly, I'd be looking at organometallics. There is a lot more space there then pure organics, and some of them can be quite stable, see the above phthalocyanine compounds.

I definitely agree that organics are much more tunable, but I think organometallics are not the answer if you're looking for a stable pigment. A chemistry meme page I follow on Facebook posted a three-door meme recently:

Door 1) Restart life
Door 2) $5 billion
Door 3) An air- and water-stable organometallic complex

I suspect that even if a promising organometallic complex was air- and water-stable, it might still suffer from the same degradation processes that regular organic pigments are prone to. YMMV based on the ligand set, of course.

I would also point out that the phthalocyanin pigments are not technically organometallic; they coordinate to the metal through nitrogen, not carbon.
posted by invokeuse at 10:31 AM on April 30, 2018

Granted, I never took a watercolour studio but stuff like this article posted by sebastienbailard would have been incredibly useful in school.
posted by GoblinHoney at 2:52 PM on May 1, 2018

« Older Image inpainting for irregular holes using partial...   |   Women in Sonic Art Newer »

This thread has been archived and is closed to new comments