YInMn blue, accidentally created in 2009, was the first inorganic blue pigment in more than 200 years.


Read our COVID-19 research and news.

Meet the blue crew, scientists trying to give food, flowers, and more a color rarely found in nature

The accidental hue

Throughout history, making a blue pigment has taken hard work—or a stroke of luck

CORVALLIS, OREGON—Mas Subramanian's most celebrated discovery came out of the blue.

As a solid state chemist at the chemical giant Dupont, Subramanian had put his name on hundreds of publications and dozens of patents. He identified a new superconductor and found a more environmentally friendly route to produce the chemical fluorobenzene. When he left the company to work at Oregon State University here in 2006, he set out to develop a multiferroic, a material with a combination of electronic and magnetic properties that could lead to faster computers.

Following one of Subramanian's ideas, graduate student Andrew Smith one day mixed indium oxide, manganese oxide, and yttrium oxide and heated the mixture in the oven. The resulting material, it turned out, didn't have any special magnetic or electric properties. It was just very blue.

Subramanian's first thought was that Smith, who had recently switched from marine biology to chemistry, had made a mistake. His second thought was something that someone at Dupont had once told him: Blue is really hard to make.

It's so hard, in fact, that Subramanian's new color became a phenomenon. The New York Times called within days after his paper on YInMn blue, as he dubbed it, appeared in the Journal of the American Chemical Society. Shepherd Color Company in Cheltenham, Australia, licensed the new pigment, which art historian Simon Schama has called "the bluest blue to date," and marketed it as a paint for artists. The new hue has inspired a music festival, and chip company AMD is using it to dye the housing of a series of graphics processors. "There is something about the color blue that just fascinates people," Subramanian says.

Humans made pigments from red and yellow ochre and charcoal at least 100,000 years ago, but they didn't have blue. The Babylonians and Egyptians used bits of lapis lazuli, a blue semiprecious stone, in statuary and art, but the laborious process needed to turn it into the pigment ultramarine was only discovered in the sixth century B.C.E. (Recent evidence from a burial site in Turkey suggests people also ground the blue mineral azurite down to a fine powder 9000 years ago, possibly for cosmetics.)

"There is something about … blue that just fascinates people," says Mas Subramanian, discoverer of YInMn blue.


With natural blues scarce, people have tried to make their own. Ancient Egyptians mixed sand, plant ash, and copper to create Egyptian blue, the first synthetic pigment, about 5000 years ago. In the 19th century, chemists raced to create a synthetic ultramarine, and BASF spent an unprecedented 18 million gold marks, more than the company was worth at the time, to synthesize indigo, a deep blue dye from plants. These blues became some of the most sought after products of the booming chemical industry.

Yet blue pigments are still rare. Most blues in nature don't come from pigments that humans can co-opt. Animals such as the morpho butterfly and the blue jay appear blue not because of a pigment, but because their feathers or scales contain nanostructures that reflect light in a way that cancels out all but the blue wavelengths.

To appear blue, a dye or a pigment needs to absorb red light, which usually happens when red photons boost electrons in the pigment molecule from one energy level to the next. Because red light has the lowest energy of any visible light, those two energy levels need to be very close together—and such closely spaced energy rungs are found only in complicated molecules that are hard for organisms to make.

Plants have evolved many classes of pigments: Chlorophylls color leaves green; carotenoids come in orange (carrots), red (tomatoes), and yellow (maize); and betalains produce the red color of beetroot. But only one class of pigments is capable of producing blue: the anthocyanins. (The word literally means "blue flower.") And even most anthocyanins are not blue but red, because they naturally absorb blue light; only if the plant tacks on chemical groups can the molecule shift toward absorbing red.

In minerals, too, blue is a special case. Subramanian discovered that YInMn's color is created by a manganese ion surrounded by five oxygen atoms in a structure resembling two three-sided pyramids glued together at the bottom, a geometry rarely seen in natural minerals.

Designing materials from scratch to produce blue is difficult even today, Subramanian says. "So much chemistry has to come together," he says. Subtle changes in the arrangement of neighboring atoms can throw off the energy levels of an atom's electrons, altering the color it can absorb. The red of rubies and the green of emeralds both spring from chromium ions surrounded by six oxygen atoms; other atoms in the two stones cause the color difference by altering the chromium's energy levels. Such effects are very hard to predict, Subramanian says: "If rubies and emeralds did not exist in nature, no one would know how to create them."

But scientists have not given up hunting for new blues, continuing an age-old quest with 21st century tools. Although Subramanian's discovery came about by accident, other researchers are methodically using physics, chemistry, and genetics to find or create new blues for painters to dazzle with, edible colorants that make food more interesting, and blue flowers that, so far, only exist in artists' imaginations.


The impossible flower

Blue roses are the stuff of poems. Scientists are trying to make them reality

KYOTO, JAPAN—In 2004, Japanese researchers unveiled what they billed as the world's first blue rose. The only problem with the flower: It wasn't very blue.

Although its petals did produce a blue pigment, the overall appearance of the flower was more mauve. Even Yoshikazu Tanaka, the scientist behind the work, admits that his first thought on seeing the flower was: "could be bluer."

Fifteen years later, he is still trying to make that bluer rose. Tanaka works at the global research center of Japanese beverage giant Suntory, which grew out of Japan's first whisky distillery, opened in 1923. (The brand was made famous by the movie Lost in Translation, in which an aging actor played by Bill Murray shoots a whisky commercial in Tokyo.) The company decided to branch out into the cut-flower market in the 1980s after a tax hike made Japanese liquor more expensive. Company legend has it that the idea was to paint the English rose the Scottish national color, blue, as a kind of thank you for the invention of whisky, Tanaka says.

More likely, it just seemed a good business idea. After all, blue flowers are rare, including among cut flowers. Chrysanthemums, carnations, tulips—none of them naturally comes in blue. Blue orchids have usually been artificially dyed. Decades of breeding have yielded roses in every shade of yellow, pink, and red, but never blue ones.

In 2017, Japanese scientists announced the creation of blue chrysanthemums.


Artists have long noted this rarity. In German romanticism, the blue flower became a symbol of longing and the unattainable. Rudyard Kipling dedicated a poem to someone tasked by his lover to find her a blue rose: "Half the world I wandered through/Seeking where such flowers grew."

By the time he returns empty-handed, his love has died.

Scientists got their first glimpse of the complexity behind blue flowers in 1913, when German researcher Richard Willstätter announced he had isolated the blue pigment from cornflowers. It was an anthocyanin he named cyanidin. Two years later, when he isolated the pigment of red roses, it turned out to be the exact same molecule. Anthocyanins can change color depending on the acidity of a solution, so Willstätter proposed that roses had a different hue because the pH in their petals was lower than in cornflowers.

It was the first scientific theory about blue flowers. And it was wrong. Over the following decades, a different story emerged, one that was finally confirmed by x-ray crystallography in 2005. Cyanidin itself does not produce a stable blue color; instead, cornflowers combine six molecules of cyanidin with six molecules of a colorless copigment arranged around two metal ions—a huge molecular complex that stabilizes the cyanidin molecules and allows one electron to make the right energy transition. "Flowers are doing crazy chemistry to generate that blue," says Beverley Glover, a botanist at the University of Cambridge in the United Kingdom.

Several other blue flowers have hit on the same trick, but most produce a different anthocyanin, called delphinidin, that can more easily be coaxed to appear blue. The only difference between cyanidin and delphinidin is that the latter has an extra oxygen atom on one of its rings, put there by an enzyme called flavonoid 3',5'-hydroxylase. The entire family of roses, which includes apples and pears, lacks the enzyme, which means that delphinidin-producing roses can't be produced through traditional breeding.

Yoshikazu Tanaka with the "blue" roses developed in his lab. His search for a bluer version continues.


Tanaka is trying genetic engineering instead. By 1991, he and his colleagues had identified and patented the flavonoid 3',5'-hydroxylase gene in petunias. Transferring that gene into carnations coaxed them into producing delphinidin, turning them a purplish blue. But when the team shuttled the gene into roses, using the bacterium Agrobacterium tumefaciens as a courier, they didn't start to produce the blue pigment for some reason. It was the same gene from pansies that finally led to the delphinidinmaking—but not very blue—rose unveiled in 2004. Apparently, producing delphinidin alone wasn't enough. Scientists had to do some crazy chemistry themselves.

Since then, Tanaka's main strategy has been to transfer genes from bellflowers, pansies, and other blue flowers to "decorate" delphinidin chemically, hoping to hit a magic combination. Last year, he showed a visitor hundreds of tiny rose plants growing under fluorescent lights in his lab. "All of them are just to get a new blue color," he said.

In the meantime, however, a collaboration between Tanaka and a group led by Naonobu Noda at the Institute of Vegetable and Floriculture Science in Tsukuba, Japan, has led to an indisputably blue flower: a blue chrysanthemum. In a 2017 Science Advances paper, the researchers reported that inserting the flavonoid 3',5'-hydroxylase gene from bellflowers into red chrysanthemums, along with a gene that adds a glucose molecule, resulted in "the most blueshifted flowers" ever genetically engineered. Their idea was that the glucose would allow the flower's natural enzymes to attach further chemical groups to delphinidin, creating a stronger blue. To their surprise, added groups weren't necessary; instead, the glucose helped delphinidin assemble with copigments naturally produced in the flower, shifting the color to blue.

Using the exact same strategy has not worked in roses, probably because they don't have the same copigments and have a lower pH. But Tanaka has not given up. He has tried to add genes from gentian that modify the delphinidin and genes from the genus Torenia that produce copigments. In a nod to Willstätter, he is even trying to change the pH in the rose petals.

Tanaka is confident he will develop bluer roses before his retirement, only 5 years away, but almost 30 years of pursuing his quest have also taught him to be cautious: "It is hard to say how blue they will be."

Doughnuts to dye for

Scientists are looking for a natural pigment to turn food blue

NORWICH, U.K.—At first, Cathie Martin was interested in the nutritional value of food pigments. Then, she became obsessed with blue food for its own sake.

A decade ago, Martin, a scientist at the John Innes Centre here, genetically engineered tomatoes to produce anthocyanins in their fruits, so that other scientists could compare their dietary effects in humans with those of regular tomatoes. But the pigments also turned the vegetable a dark, purplish blue. And Martin began to wonder how to make other food blue.

Few foods are naturally blue, but the color has long been in demand as a food colorant. Synthetic ultramarine was once used to whiten cane sugar, which has a yellowish tinge. Blue food dyes are used to color candy, coatings, or drinks. They are also mixed with other colors. "We must have blue to make all the colors of the spectrum," says Richard van Breemen, a chemist who investigates natural products at Oregon State University in Corvallis.

Currently, there's not a lot to choose from. Two synthetic blue food dyes are approved in the United States: Brilliant blue, also named blue No. 1, was originally made from coal tar, like many synthetic dyes, and blue No. 2, or indigo karmine, is derived from synthetic indigo. Another synthetic blue colorant is available in the European Union: patent blue V, which gives blue curaçao liqueur its hue.

A bluish doughnut frosting developed in Cathie Martin's lab contains a mix of anthocyanins found in butterfly pea flowers.


Because consumers prefer natural ingredients, big companies such as Mars and Pepsi have invested in replacements for the synthetic colorants, with little success so far. "One of the big frustrations with the color blue is that it is very, very difficult to reproduce the colors that you see in nature with compounds that can be used in the same way for coloring food," Martin says.

The only natural blue colorant is a crude extract derived from spirulina algae that the U.S. Food and Drug Administration approved for use in confectionery and other food in 2014. But it is not very stable, Martin says—or very blue, for that matter. "It's a terrible blue," she says. "It's green really." And the color may change or disappear when foods are baked or boiled or exposed to light on grocery shelves.

Van Breemen has looked for better candidates in the microbial world. Reasoning that he was more likely to find stable blues in extreme conditions, he studied microbes from the hot springs at Yellowstone National Park, for instance; he also searched in marine bacteria. But none of the blue pigments he found was suitable. Many are chemical weapons, which the microbes release to fight other microbes, he says—making them more promising as antibiotics than as food colorants.

Plants may be a better bet, and they offer many compounds to choose from. Although most blue flowers create pigments based on delphinidin, they vary the molecule by adding different chemical groups, and many of the intermediates in the chemical pathway leading to delphinidin are blue as well. Martin is hoping she'll find a safe, stable food dye in the butterfly pea, whose beautiful blue flowers give the Malay rice dish nasi kerabu its blue color. (The flower's color, however, was not what most struck the men who gave the genus its Latin name, Clitoria.)

Martin initially bought Clitoria blossoms online, from Amazon, but stocks soon ran out; more recently, she received three bulging bags of blossoms from Saudi Arabia, where a scientist who had visited her lab asked people to collect them in the wild. A mix of butterfly pea anthocyanins has worked well for some food applications, Martin says. Researchers in her lab have used it to make bluish frosting for cupcakes and doughnuts as well as blue ice cream.

But these pigments, too, are fleeting. "Most blue anthocyanins have a half-life of about 24 hours. And we're talking about something that needs at least about 3 months," Martin says. So her quest continues.

The deepest blue

A mineral created under immense pressure inspired the search for a new pigment

LONDON—Geologist David Dobson of University College London (UCL) never realized that blue pigments are a big deal until he saw the excitement that a sample of the world's newest blue, Mas Subramanian's YInMn, created among colleagues at UCL's Slade School of Fine Art. "I thought: Hang on a minute," Dobson says. "I'm making blue all the time in my lab."

Dobson studies the transition zone, the part of Earth's mantle that stretches from about 410 to 660 kilometers beneath our feet. In his lab, he squeezes mineral samples in a machine called a multianvil cell to replicate the gigantic pressure at those depths—about 200,000 times that at Earth's surface. Under those circumstances, the four elements that make up olivine, the most common mineral in the mantle—iron, magnesium, silicon, and oxygen—form a different mineral called ringwoodite, whose physical and chemical properties Dobson is studying. The millimeter-size crumbs of ringwoodite also happen to be a deep blue.

A small piece of ringwoodite produced in David Dobson's lab. Dobson hopes to produce a new blue pigment that has a similar structure but is more stable.


Having seen the success of YInMn, Dobson decided to turn his deep-Earth blue into a new pigment. He expects it will find a market, if only because not everybody sees YInMn as the perfect blue. The element indium, one of its ingredients, makes it expensive; a 40-milliliter tube of the acrylic paint that made a splash at the Slade School, produced by a company named Derivan, sells for $130 or more. And it "was actually a bit soupy," says Jo Volley, a lecturer at the Slade School.

First, Dobson had to understand where ringwoodite's color comes from. "Everyone was used to it being blue, and no one had really considered that much why," he says. He found that the color arises not from an energy transition within one atom, but from the exchange of an electron between two types of iron ions, Fe2+ and Fe3+. (The same mechanism accounts for the color of Prussian blue, a pigment discovered by chance in 1706 when Berlin alchemists used contaminated potash in a recipe for a red pigment.) Ringwoodite's structure, with the iron ions surrounded by four oxygen atoms in a tetrahedral coordination, creates the right conditions for the electron swap to absorb red light. But that arrangement is stable only at the huge pressure in Earth's interior. At the surface, even simply grinding the mineral destroys the structure—and the color.

Dobson tried to create a similar structure that is stable at a pressure of 1 atmosphere by starting with zinc germanate, a mineral that also has metal ions—in this case zinc and germanium—surrounded by oxygen atoms. If enough iron replaces the zinc and germanium, the structure turns blue, Dobson says. He has already produced a sample of re-engineered zinc germanate in his lab, and it is blue—but he hopes to make the color richer by adding more iron.

Three centuries ago, Dutch painter Pieter van der Werff used newly discovered Prussian blue to color the sky and Mary's veil in a painting depicting the entombment of Christ. Subramanian's wife Rajeevi—a solid state chemist as well as an artist—was the first to use YInMn; it proved perfect for a painting of Crater Lake, not far from the couple's home, which is famous for its deep blue water.

Dobson hopes to develop a pigment that is similarly appealing. As the first blue pigment to be designed from scratch, rather than accidentally discovered or borrowed from nature, it would open a new chapter in humanity's love affair with blue.

*Correction, 13 June, 2 p.m.: An earlier version of this story misstated that indium is a rare earth element.