Important update on LED lighting!


In January 2013 a European research team issued a press release  making public the results of a multinational research project which succeeded in unravelling the chemistry responsible for the gradual darkening of certain chrome and cadmium yellows used by, among others, Vincent van Gogh. The work had been published in a series of reports in the journal Analytical Chemistry culminating in two final reports in December 2012.

The team also made a dramatic and highly publicised claim that their research showed LEDs may be unsafe for illuminating light-sensitive materials, including the paints in question. It read in part, “[t]he scientists recommend that museums identify all paintings with this type of chrome yellow, and protect them in particular from the increasingly popular LED lights as these emit a large amount of blue”, with one of the lead Italian authors remarking, “[w]e were surprised to note that already under conditions of illumination currently considered safe, some of our test samples started changing color quickly.” 

The news was quickly picked up in the international press and the issue went viral in the conservation community, fuelled no doubt by the prominence of the scientists and their institutions as well as the legitimate reservations that conservators have about any fundamental change in their professional landscape.

Significantly the researchers’ conclusions in relation to LEDs were not mentioned in the Analytical Chemistry articles themselves, where it is unlikely they would have withstood peer review. This is because they are based on a radical extrapolation of the result of a single very unrealistic accelerated light exposure test employing a UV rich illuminant, completely unlike LEDs used for general lighting which contain no UV at all and therefore do not need filtering as do the alternatives. In order to prosecute this shaky case they offered as evidence the “spectrum of a typical "white" LED, containing a substantial portion of harmful blue light”, which is in fact not “typical” of contemporary LEDs recommended for museum lighting.

Moreover they did not adequately explain that the cumulative light dose responsible for the dramatic and visually convincing darkening depicted in their photographs was equivalent to a century or more of exposure. Even ignoring the fact that the misleadingly named “blue” test light upon which they based their warning included UV-A from the xenon test lamp down to 335nm, this was not the sudden and unexpected event implied by the statement that they resulted from “conditions of illumination currently considered safe”. Had their evidence warranted the warning and were it consistent with the few published LED fading studies available (which reach the opposite conclusion), it should have been prefaced with a clear explanation this is an issue of incremental and relative harm, not a new factor associated with the introduction of LEDs. To render the colours of these paintings satisfactorily any illuminant must contain a proportion of “harmful blue light”.

Finally - and inexplicably given the resources at their disposal and the importance of the issue - they did not take the very simple step of repeating the test with an actual LED, not even the untypical high colour temperature example proffered as evidence of potential damage. This could have been achieved in a few weeks at virtually no cost. In the meantime, even if they were correct, the consequences for even the most fugitive pigments of delaying an announcement would be truly trivial. Fading is an important and sometimes pressing issue, but never an emergency.

We’ve been here before. There was something of a stir when Georgia O’Keeffe Museum conservator Dale Kronkright raised a question on Consdistlist in March 2010 about potential “hole burning” in situations where the same elevated peaks in the blue region of some LEDs of the kind referred to by the van Gogh team might selectively affect colourants whose activation spectra happen to coincide with them. The concern was legitimate and it was satisfactorily addressed on the AIC website by Steve Weintraub Although I wasn’t around at the time, I have been told that the introduction of fluorescent lighting caused a similar panic and no doubt there will be further outbreaks in future.

If you are involved in selecting LEDs for exhibition lighting, get hold of Jim Druzik (GCI) and Stefan Michalski’s (CCI) balanced and informative “Guidelines for Selecting Solid State Lighting for Museums” and for information on the likely comparative impact on light sensitive colourants of different brand-name lights, including those he does not list but you may be using if you have the means to measure spectra, follow the link to Joseph Padfield’s pages on the National Gallery of London’s website

The lesson is clear: look at spectra, read the literature and be sceptical of press releases and rumours, no matter what their source. Light from LEDs is just … light, not a new and unknown radiation out to destroy your paintings next week. Contemporary LEDs have spectral power distributions very similar to UV filtered incandescent sources and they increasingly offer excellent colour rendering which does not deteriorate with colour temperature when dimmed. They offer much greater flexibility in illumination, emit less heat, are much more power-efficient and last longer. There are better and worse LEDs from a damage perspective of course, which you can explore on Padfield’s website. But based on published LED fading data and the readily available technical data on contemporary LEDs, had the researchers tested a few their headline would have been a fizzer:

“Some badly chosen high colour temperature LEDs with awful colour rendering may cause marginally more fading of some pigments than UV-filtered halogens”.

Bruce Ford works as a microfading consultant and also researches fading at the National Museum of Australia. He has no connection with any LED lighting company and his views do not necessarily represent those of the NMA.

Bruce has participated in drafting a comprehensive response to this issue which is being coordinated by Jim Druzik at the Getty Conservation Institute.