When it comes to sunflowers, Vincent Van Gogh's series of paintings is burnt into our collective consciousness. Of course, his late 19th-century studies weren't completely faithful representations of flowers in a vase - the Dutch painter was, after all, a post-Impressionist artist unafraid to embellish reality. But, somehow, when we think of these vibrant blooms we think of French fields full of yellow heads arching towards the sun, and Van Gogh the artist.
And, in one peculiar way, Van Gogh's series was more realistic than perhaps he ever intended. Because, just as the genuine article begins to wither, lose its lustre and finally decay into a brown mush, the brilliant chrome yellow paint Van Gogh used also began to turn brown. For more than a century, the reasons for this decay have been the subject of much scientific debate - not least because not all Van Gogh's sunflowers painted with chrome yellow have suffered the same fate.
But, using samples of paint left over from tubes used by 19th century artists - astonishingly, some of the paint still exists - scientists at the University of Antwerp discovered the culprit: a combination of Van Gogh's own techniques and that perennially deadly enemy of art preservation: the sun. Sunlight, the Analytical Chemistry journal reported last month, begins a chemical reaction which first oxidises the oil in the paint and then turns the yellow pigments green. The mix of green paint with the oxidised oil ends up producing - you've guessed it - a brown colour.
Nevertheless, that doesn't fully excuse Van Gogh. For the chemical reaction takes place only in the paintings in which he used white pigments to lighten the yellow paint. The irony wasn't lost on the leader of the Antwerp research group, Koen Janssens. "Van Gogh intended to make a lighter yellow paint, but through this effect, nature darkens it," he said last month. "While he wanted to show a light, pale and delicate yellow, it instead becomes a darker, brownish yellow."
One might wonder why such research is important, but the findings do have a purpose. The conclusions - that such discolouration can be halted by reducing light levels and installing air conditioning units to keep paintings cool (heat accelerates the chemical reaction) - will be digested with interest by galleries around the world. After all, it's not just Van Gogh's sunflowers that are under threat - at the National Gallery in London, nearly every painting by Turner has discoloured, as has Bacchus's cloak in Titian's Renaissance masterpiece Bacchus and Ariadne.
Increasingly, art is becoming entwined with science in this way - not just in preserving valuable work as it was intended to be seen, but also in uncovering the history behind some of the great paintings. And if the techniques used to delve into the stories behind classic images sound like something out of a geeky science-fiction film - the National Gallery uses the magnificently titled gas chromatography-mass spectrometer to understand how paintings were made - then the results of such research are immensely crowd-pleasing. Last year, the National Gallery's Close Examination: Fakes, Mistakes and Discoveries was a hugely popular exhibition that used science to shed light on some of the art world's most disputed paintings.
It was no surprise that the National Gallery was at the heart of such research. Its scientific department was founded in 1934 and has become a world leader in the study of painting materials and techniques. So does the director of scientific research, Ashok Roy, go along with the idea that the change in the colour of Van Gogh's Sunflowers should be celebrated because it mimics the process of decay?
"This is an interesting point about the 'life history' of old master paintings," he says. "But although in this case the discolouration of the chrome yellow paints reflects the kind of colour change that a decaying sunflower might display, in most cases the type of colour change shows no such correlation. Our interest is to preserve paintings, so we would like to minimise the effect of changes in paint as far as possible."
Roy isn't being a killjoy. He goes on to explain that many Dutch still-life paintings now show blue-coloured foliage where the yellow glaze has faded from on top of a stable blue underpaint. But the role of director of scientific research isn't just about leading the charge to preserve paintings. As the Fakes, Mistakes and Discoveries exhibition proved, science can also help us to understand the history of art. One memorable section of the show, for example, highlighted how X-ray imaging proved a 17th-century painting by Pieter De Hooch entitled A Man With Dead Birds was originally painted with a wounded man at its centre, rather than the titular deceased birds. They were added later to appeal to more conservative 19th- century sensibilities.
"Part of the point of studying paintings by scientific means is to understand the history of painting technique," Roy says. "The more that is known about this, the more we are able to interpret the technical history of painting."
In the exhibition, the National Gallery showed off its infrared imaging work - which allowed us to appreciate the original intentions of Giovanni Bellini's 15th-century A Dominican With The Attributes of Saint Peter Martyr, before it was adapted. Elsewhere, electron microscopy - where an image is magnified far beyond what the human eye can see - enabled restorers to return an old Hans Holbein picture to its original glory. But while Roy is at home talking of X-radiography, infrared reflectography, and mass-spectrometry, he thinks the most important development in scientifically understanding paintings in his career has been the humble computer.
"That's perhaps surprising," he says. "But the application of digital technology to imaging paintings and imaging materials from paintings has probably had the greatest impact, partly because digital records are much easier to handle and compare, and partly because digital records can be shared with colleagues around the world with great speed.
"Of course, the technologies involved in analysing paintings have also expanded very considerably in the past 20 years. One major change is that the instruments used to analyse paint are very much more sensitive now, so smaller quantities of material are required for analysis. All this has accelerated advances in knowledge in our field enormously."
In fact, the University of Antwerp scientists are hardly the first to suggest sunlight is partly to blame for the browning of Van Gogh's Sunflowers; at the National Gallery, they've known for some time that light-related damage is caused by the ultraviolet component of daylight, and it has led to UV effectively being blocked from entering the gallery's exhibition space.
So, does all this mean that science can solve every artistic mystery and help preserve every painting in the future? Possibly. But the feeling that the Fakes, Mistakes exhibition imparted was actually one of awe, that the fakers and copyists had somehow thrived for so long. As the researcher Rachel Billinge said at the launch of the exhibition: "Sometimes the faker has gone to such lengths, you can respect their techniques."
As for Roy, he's more excited about the establishment of internationally accessible databases of the methods and materials of Old Master paintings. Which just goes to show, this work isn't just the preserve of geeks in lab coats. It's part of a quest for knowledge - and a desire to preserve the great paintings for many centuries to come.