What’s All this About Van Gogh and LEDs?


There is a disconnect between recent claims that LED illumination is destroying great works of art and what the research behind the claims really involved.

Like van Gogh’s “Sunflowers”? 

What if museum lighting were destroying them?  What if your invention had enabled the lights that were doing the damage?

Last week, I received a curious press release from Soraa, The GaN-on-GaN LED company founded by Shuji Nakamura, the guy who invented a way to make blue and UV LEDs, which is the reason we can have high brightness (HB) white LEDs.  (The part of the spectral gamut that produces HB white light comes from “yellow” phosphors on the inside surface of the LED package.)

The release is passionate, insisting that there is no evidence that white LEDs cause pigment degradation and pointing out that the experiments that have been referenced did not use white or any other color) LEDs.  I’ll quote the release at the end of the article, with a diagram that compares the spectral content of light sources.

So I asked myself, where did this come from?  A little Googling turned up four dozen or so articles from the fine arts of popular press.

The primary source seems to be this one, in The Independent, a British cultural publication.  The tile is: “Sunflowers wilt: Van Gogh's masterpiece is slowly turning brown as a result of exposure to LED lighting” 

The gist is that some wavelengths of light accelerate the oxidation of certain chromium compounds that are part of the yellow pigments that van Gogh and other artists used.  The piece goes on to say that photons from LEDs do the greatest damage.

The first claim is supported by a series of scientific articles in Analytical Chemistry, a journal of the American Chemical Society.  The most recent was published last October.  The second is not supported by anything in the fourth paper, which is he one that describes experiments with a real (Xenon) light source.

The Analytical Chemistry articles share a common title: “Degradation Process of Lead Chromate in Paintings by Vincent van Gogh Studied by Means of Synchrotron X-ray Spectromicroscopy and Related Methods.” [] 

The abstract for the most recent paper says, in part: “The darkening of the original yellow areas painted with the chrome yellow pigment . . .is a phenomenon widely observed on several paintings by Vincent van Gogh, such as the famous different versions of Sunflowers. During our previous investigations on artificially aged model samples of lead chromate, we established . . . that darkening of chrome yellow is caused by reduction of PbCrO4 to Cr2O3·2H2O (viridian green), likely accompanied by the presence of another Cr(III) compound. . . .

“In the second part of this work, in order to demonstrate that this reduction phenomenon effectively takes place in real paintings, we study original paint samples from two paintings of V. van Gogh. . . .

[The article goes on to say that they employed] “high lateral resolution spectroscopic methods . . . [that used] synchrotron radiation (SR), such as microscopic X-ray absorption near edge (μ-XANES) and X-ray fluorescence spectrometry (μ-XRF. . . .

“Additionally, μ-Raman and mid-FTIR analyses were carried out. . . .  On both paint microsamples, the local presence of reduced Cr was demonstrated by . .  μ-XANES point measurements. The presence of Cr(III) was revealed in specific areas, in some cases correlated to the presence of Ba(sulfate) and/or to that of aluminum silicate compounds.”


XANES refers to ESERF, The European Synchrotron Radiation Facility in Grenoble, France.  ESERF’s Web page says: “Imagine a source of X-rays 10,000 billion times brighter than a hospital X-ray machine. This brilliant source is called a synchrotron, and the X-rays it produces reveal the structure of matter in all its beauty – and complexity.”

Okay, that’s written for a broad audience, but you get the idea.  They used this amazing synchrotron, but what did they actually do?  Where does the LED illumination come in?  (In fact it doesn’t.)

For $35, the ACS will let civilians like me read Analytical Chemistry papers.  Or, one can download them and read them for 48 hours before they disappear. 

I paid the freight and downloaded that October paper. 

The relevant paragraph says:

UVA-Visible Light Exposure.  All paint models have been aged for 800 h, employing a SOLARBOX 1500e System (CO.FO.ME.GRA, Milano, Italy.)  This chamber is equipped with a xenon lamp (550 W/m2) emitting between 290 and 800 nm (UVA-visible light) and operating at a temperature between 50 and 60 (degrees) C. A soda-lime glass UV filter between the light source and the samples allows simulating exposure to indoor light conditions.”

That’s it.  Just a reference to a spectral band obtained from a xenon lamp.

Where did a dozen publications get the LED story?

The ESERF site has a press release: “LED lights may be bad for Van Gogh paintings. []

The deck says: “Dozens of masterpieces are susceptible to light induced darkening of yellow areas.”

The body copy goes on: “An international team of scientists has used synchrotron X-rays to better understand why some bright yellow colours in Vincent Van Gogh’s paintings are turning brown with time, while others do not. The research focus was on chrome yellow, a colour favoured by Van Gogh to depict sunshine and light. Several types of this yellow were found to be very sensitive to green and blue light which causes a darkening of the painting. The scientists recommend that museums identify all paintings with this type of chrome yellow and protect them in particular from the increasingly popular LED lights which emit a large amount of blue.”

Oh?  But that latter assertion isn’t in the paper.

Where does that assertion come from?

The Independent attributes it to the man who runs the synchrotron:

“Claus Habfast, from the European Synchrotron Radiation Facility in France, where some of the artworks were studied, said: "LED lights appear to have many advantages but museums should carefully consider that paintings from the Van Gogh era could be affected by them.

"Paintings that have moderate darkening will find this accelerates in the coming years.

"Of course, it's not advisable to put these paintings in the dark because they are part of the cultural heritage of humankind and the public wants to see them. But museums have to strike the right balance," he says.

Since the January 13 date The Independent (apparently) broke the story until the search I just performed (7 days later), those quotes were picked up no fewer than 50 times, according to Google.

And I’m probably the first journalist to cough up the 35 bucks to read the actual paper.



By now, you probably want to know what the release that started my research said.  (It’s Soraa’s release; it doesn’t quote Mr. Nakamura himself.)  Here:

“The degradation studies [Monico et al., Anal. Chem., 2013, 85 (2), pp 860–867] in fact employed a high-intensity Xenon lamp (Cermax 175W) as the photo-excitation source. Xenon lamp spectra are vastly different than those of white-emitting LEDs for illumination, with the Xenon lamp including significant amounts of ultraviolet (UV) light in the UV-A, UV-B, and UV-C wavelength regimes. Degradation was observed under intense (more than 1000x conventional illumination) irradiation when the Xenon lamp was used in conjunction with a so-called “blue” filter passing wavelengths between 335 nm (UV-A) and 525 nm (green), which is in fact about five times broader than the blue wavelength regime. Degradation was more pronounced when a UV filter (240-400 nm) was used, or with no filter at all.

“Photo-oxidation effects are more probable as excitation wavelength gets shorter and it is very plausible that the degradation observed by the researchers at University of Antwerp is caused by UV-A light leaking through the “blue” filter they employed. In contrast, LEDs (including SORAA VIVID LED MR16 2700K, for example) emit essentially no UV-A emission – more than 100,000x less than the filtered Xenon lamp used in the University of Antwerp studies (see chart below).

Soraa’s chart shows the spectral content of Xenon and Halogen bulbs relative to Its VIVID LED product.  The LED UV spectrum peaks at wavelengths that are considerably shorter than those of the conventional lighting.  (Visible wavelengths are not an issue with respect to the studies of pigment degradation.)

The sunflowers in the Louvre may be getting browner, but it has not yet been demonstrated that it’s the fault of HB LEDs.

Discuss this Blog Entry 6

on Feb 19, 2013

I am surprised a museum would even consider using a blue+phosphor LED light source, considering its poor spectral content. However, this article does make sense as it is difficult to get any significant sub-400nm output from LEDs...

on Feb 20, 2013

Well Written.. it's a shame more 'journalists' don't take the time to do their research. I suspect the reason behind that varies depending on the subject matter (political research being bent to the will of the subjective researcher) while technical research I think often falls prey to non-technical authors who don't want to spend the time to research a subject properly. Instead, these 'journalists' just pump out sensationalist copy and bank a check. Thanks Don for the careful de-bunking!

on Feb 23, 2013

Maybe bright pigments need lower wavelengths for boost so they are bright to our eyes. Like a phosphor paint.

Put back the original light source. Does brown become yellow again over time? If so, i think the lower the spectral emission from the LED.

on Feb 23, 2013

Lower emission from the led may need restored similar to halogen.

on Feb 27, 2013

O.K. Quit whining...

Look you arty types just need to give in and accept a "modeling viewer" to replace actual Van Gogh works on display in the museums. You know...kind of like those "digital modeling amps" that supposedly emulate the sound of my real vacuum tube guitar amplifiers? Look, you can do anything with transistors and LEDs these days, right? Why waste a bunch of time on old fashioned oils and canvas and stuff? That's ANALOG, dude! Besides, lead's toxic and it is poisoning our environment. We want ROHS compliant Van Goghs and WE WANT THEM NOW.

Look, if you don't want to take the time to digitally model them for display, heck - I bet you could have a skilled copyist duplicate those Van Gogh works AND with ROHS compliant materials, too. If you did, I'd wager good money that 99% of the people who viewed them from 10 feet away would NOT be able to tell the difference anyway - just like folks who can't tell the difference between a digital modeling amp and a real 57 Strat through a real 63 Vibroverb.

Finally, hey - who decided that sunlight is "right" anyway? That "right and wrong" mindset is really out of date. Maybe I think they look BETTER in LED lighting. Maybe better in sodium vapor lamps. Who made you god, huh? Very judgmental to say they HAVE to be viewed in the same light Van Gogh used. It's light discrimination - and discrimination is just wrong folks.

So, to review;

1. Quit whining and wasting money on x-ray experiments. Just make a copy or a digitally modeled representative and display THAT. It's close enough, because...
2. Most people will never be able to tell the difference anyway. It's a picture of some flowers, o.k?
3. LEDs are not at fault, we get that, but it is wrong to discriminate the lighting types and insist that one kind is "wrong" and another is "right". That's discrimination.


Nice article, by the way. Dear to my heart when folks actually read the substance of the research, rather than cutting and pasting quotes from Reuters, most of which seem to be written by junior hacks who couldn't explain the difference between an avocado and Avogadro's number.


on Apr 5, 2013

Did someone fail chemistry? If you've seen a US Interstate Highway designator sign, they're red, white, and blue, the red part is usually faded out (esp if facing west). We see red because it absorbs blue-green. A photon of blue light has roughly twice the energy of red light; causing more chemical reactions. The phosphor is yellow because it absorbs blue incandescent light (note the LED diode emits blue-violet; the phosphor looks more orange under UV flashlight).
Chemistry. Oil paintings are made with - surprise - oil (like linseed aka flaxseed or walnut oil). Recommended to keep your edible oil out of direct light because it reacts with oxygen (to peroxides) which is accelerated by light. UV-A is an irrelevant arguement as the Yellow Lead Chromate blue light (yellow, like the phosphor). Also note that there is no "white" LED phosphor that does not degrade due to the "white" LED diode (blue-violet) light impinging on it. The known issue of LED color changing over time. Your spectrum shows much higher output at 405 than halogen. And although "warehouse stores", e.g. Costco, use high pressure mercury lamps with quartz envelopes, consumer bulbs like low pressure mercury filled flourescents have glass envelopes and phosphors that absorb UV. Xenon usually has a pinkish hue with lots of infrared tailing, looks more like an Argon spectra.

Please or Register to post comments.

What's Secondary Emissions?

Blogs from Electronic Design's Editor covering Analog and Power


Don Tuite

Don Tuite covers Analog and Power issues for Electronic Design’s magazine and website. He has a BSEE and an M.S in Technical Communication, and has worked for companies in aerospace,...
Commentaries and Blogs
Guest Blogs
Jan 26, 2017

An Amateur’s View on the P2 (Part 2): Slew Rate and the Oscillator 3

Justin Mamaradlo takes a further look into the P2 op amp and how it functions, analyzing the oscillation and slew-rate characteristics of the venerable component....More
Jul 15, 2016

Simple Yet Effective ESD Testing Methods for Higher Reliability 11

There are multiple ways to test for electrostatic discharge, ranging from implementing a human-body or machine model to....using a balloon and a comb?...More
Apr 8, 2016

Confabbing on the Fabless Fad 5

High capital and maintenance costs, and EDA advances along with abstractions to deal with chip complexity, have been leading contributors to the fabless migration....More

Sponsored Introduction Continue on to (or wait seconds) ×