In a sparsely furnished back room at the Smithsonian’s Freer Gallery of Art, a trim, gray-haired chemist is hunched over a Japanese painting from the 19th century. After making minute adjustments to a fiber-optic probe suspended just millimeters above the painting, he flicks on the power.

A tiny dot of light beams from the probe’s tip, reflecting off of a blue pigment that was delicately laid down by the brush of Katsushika Hokusai (1760-1849) more than 200 years ago. The reflected light flows into a spectroscope that measures both visible and invisible ultraviolet light.

Titled “Miscellaneous Subjects,” the painted handscroll being examined is a stunning work in the Japanese ukiyo-e style. Yet Conservation Scientist John Winter of the Smithsonian’s Arthur M. Sackler Gallery and Freer Gallery of Art is looking for a European invader hidden in plain sight on the painting’s surface: a pigment called Prussian blue.

This synthetic paint made from iron salts and cyanide obtained from cattle blood was invented in Germany in 1704. Some 80 years later, it reached Japan through trade with the Dutch and Chinese. By the 1820s, after Prussian blue became readily available and less costly, its vividness and intensity sparked a fashion for blue among the producers of Japanese ukiyo-e woodblock prints, but not among painters, as Winter’s research reveals.

Woodblock prints were reproduced in large numbers by a guild of artists, block-carvers and printers. Perfectly suited for printmaking because of its intensity and fine grains, “Prussian blue quickly replaced indigo and other traditional blue pigments in Japanese prints,” says Marco Leona, the David H. Koch scientist-in-charge in the Department of Scientific Research at the Metropolitan Museum of Art in New York, who collaborated with Winter in the investigation.

But as Winter and Leona discovered in their work, Japanese painters were not as enamored with this new color as were Japanese printmakers. By compiling data from spectroscopic analysis of Japanese paintings in the Freer Gallery’s collection, Winter and Leona have documented the more conservative response to this novel European pigment by Japanese ukiyo-e painters.

Indigo vs. Prussian blue
For centuries, many different materials—such as clay, crushed oyster shell, gamboge (tree resin), mica, ground copper minerals and mercuric sulfide—have been used to create pigments for the palettes of Japanese artists. Many pigments and the knowledge of making them came to Japan through China and other areas of Asia. Pigments such as cinnabar, malachite and azurite, for example, came into Japan, along with the introduction of Buddhism, from mainland China.

Using microscopy and microchemical analysis, Freer Gallery of Art Research Associate Elisabeth West Fitzhugh recently published a comprehensive database identifying every pigment employed in the creation of some 500 Japanese ukiyo-e paintings, dating from the 16th to the 19th centuries and owned by the Freer. The database represents “a titanic work,” Leona says. It took Fitzhugh and others decades to complete.

Yet in specific areas on some paintings where the paint was exceedingly thin, Fitzhugh was unable to determine whether an artist had used Prussian blue or indigo, a traditional blue pigment made from plants.

Looking at a painting’s surface with the naked eye, it is impossible to tell Prussian blue and indigo apart, Winter says. Often, they appear in “very similar shades of dark blue. I don’t think anybody would claim to tell the difference just by looking at them through an optical microscope.”

One simple method to tell these two pigments apart is through chemical analysis. Both pigments react differently when exposed to specific chemicals, such as nitric acid, an oxidizing agent. Because each pigment is made up of particles that are quite tiny, the pigments are intense. “Painters did not need to apply much of either” to achieve the desired shade of blue, Winter explains. Layers of these pigments on Japanese paintings are often thin. Lifting even tiny samples for analysis may be impossible without damaging the paintings.

Ultraviolet analysis
“Using a spectroscope, however, is noninvasive,” Winter explains. “No sample is removed, and there is no contact between the probe and the work of art being examined.” Winter and Leona used a fiber-optic probe with the spectroscope that allowed them to collect reflected light—both visible and ultraviolet—from spots as small as .025 of an inch.

Armed with this equipment, Winter and Leona returned to inspect blue areas of some 200 paintings, areas that Fitzhugh had been unable to determine were painted with indigo or Prussian blue. “We built upon her work,” Leona says.

In the infrared range, invisible to the human eye, the differences between Prussian blue and indigo became immediately apparent. In visible light, both pigments absorb the color red and reflect blue. In the infrared, Winter explains, “Prussian blue continues absorbing red, while indigo does not. Indigo is faint, nearly invisible in the infrared. The difference is striking.”

Paintings and prints
Winter and Leona discerned the earliest use of Prussian blue in a painting dated with certainty to 1817. Of the 500 ukiyo-e paintings in the Freer collection, 139 contained either indigo or Prussian blue, alone or in mixtures.

One point of interest, Winter says, “is that a painter sometimes used indigo and Prussian blue for different design features in the same painting. This, I thought, was interesting in that it shows a remarkably subtle appreciation of color differences.”

Despite the availability of Prussian blue, indigo still remained popular with Japanese painters. Half of the 34 paintings made after 1830 that were selected for this study contained Prussian blue, while 31 of the 34 paintings contained indigo.

This was at a time that Prussian blue had replaced indigo in Japanese prints and was being used by printmakers in sizeable quantities for prints with large production runs, Leona points out. It stands to reason Japanese painters could easily have afforded Prussian blue during this time.

Overall, Prussian blue never achieved the popularity among Japanese painters in the 1820s and 1830s that it did among printmakers. “Probably most painters did not feel the need to add a new pigment to their repertoire,” Leona says, “particularly in the early years of its introduction, when the price was still quite high. They even had blue pigments other than indigo that could be used instead, such as azurite and smalt.” The coarse grains of azurite, a crushed mineral pigment, and smalt, a pigment made from deep-blue glass that has been pulverized, made them unsuited for woodblock prints.

“Printmakers had more incentive to change,” Leona continues. “The new ‘Technicolor’ hues of Prussian blue were in demand among the lower middle class who couldn’t afford paintings.”

On a number of paintings, Winter and Leona encountered Prussian blue and indigo mixed together as a single pigment. Did an artist combine the pigments to try to create a new color? Or was it an attempt to stretch the more costly Prussian blue by mixing it with indigo? Or as Winter questions, did a merchant cut his stock of Prussian blue with indigo before selling it to increase profits? These are a few questions awaiting answers.

“What we have done in this study is provide raw data on pigments contained in the Freer’s amazing collection of Japanese ukiyo-e paintings,” Leona says. “Now, it is the job of an art historian, economist or other scholar to try to apply this data to specific socio-economic questions.”