A mixture of two of these, taken from Dow's list of top 16 candidates, was ultimately used by Tate conservators to remove the graffiti from the work, titled Black on Maroon.
"It was an honor to be able to help," said Keefe, a chemist now stationed at Dow's research labs in Collegeville.
Bronwyn Ormsby, senior conservation scientist at Tate, praised Keefe and her colleagues for their skill and insight.
"What Dow contributed was key to minimizing any experiments on the painting itself," Ormsby said.
The 1958 canvas was part of a series initially intended for the walls of the Four Seasons restaurant in New York. But the artist decided they were too dark for that setting and withdrew from the commission, eventually donating the works to Tate in 1970.
Black on Maroon was marked with black ink in October 2012 by a man named Vladimir Umanets, who described himself as a fellow artist making a statement. He spent a year and a half in prison and later apologized for his acts, calling them "selfish" in a commentary he wrote in the Guardian.
Meanwhile, scientists set to work undoing the damage.
Keefe already had worked with Ormsby in a long-term partnership to develop better methods for cleaning modern art. Along with colleagues at the Getty Conservation Institute in Los Angeles, she and Ormsby tailored techniques for removing dirt from acrylic paint.
Keefe, who earned an undergraduate degree in chemistry from Pennsylvania State University and a Ph.D. in that subject from Northwestern, works in Dow's coatings unit. The company allows her to work on the art conservation partnership as part of her day job.
The Rothko presented an unusual challenge.
As with any attempt to clean a painting, the Tate conservators needed to dissolve or otherwise remove the offending substance while causing as little harm as possible to the painting itself.
But as in many of his other works, Rothko had used a variety of media in Black on Maroon, including oils, resins, egg-based coatings, and animal glue. Each one of those made the job that much trickier. Was there something that could dissolve substance A but not B? And not X, Y, or Z?
Tate officials ordered more of the ink that Umanets had used and sent some of it to Keefe. Dow conducted tests on the ink and ran the results through software to come up with a list of several dozen candidate solvents.
Keefe and colleagues, including Felipe Donate in Dow's oxygenated solvents unit, then spent several days trying more than 20 of the solvents on test canvases. They submitted a list of 16 to Tate, which then conducted additional tests before selecting the final two: benzyl alcohol and ethyl lactate.
The former is a type of alcohol, as the name indicates; the latter is from a class of compounds called esters, and is considered biodegradable.
"It's fairly new to conservation," Keefe said. "It was something that they don't typically use."
Tate conservator Rachel Barker spent months working on the actual painting, cleaning it and gently retouching areas where the solvent, despite her best efforts, removed some of the original paint.
Though Keefe's main contribution to the 18-month project lasted just a few days, she remained in contact throughout, and went to the gallery in May to see the painting when it was rehung.
In the dim lighting of the gallery, the retouched area looks very much as it did originally, though traces of the repair work can be seen in bright light, Ormsby said.
"That's going to be part of the painting's history," Ormsby said. "It's going to be with us forever now."