PITTSBURGH—A soda can might seem unworthy of scientific research. But just ask chemist Tom Mallen. With his back to a laboratory table, he holds up a silvery disk and points to the spot where a metal pull tab is pinned to the middle. That spot is one of the toughest tests for the thin plastic layer coating the insides of cans. Enormous strain is placed on the lining as a machine pounds the metal lid to make a small bump where the pull tab attaches, and then squashes the bump to pin the tab in place. That makes it “the most difficult fabrication in this whole universe,” he declares.
Designing coatings that can withstand this rough treatment—and meet a host of other requirements—has defined much of Mallen’s 31-year career. Now, he and other employees of Sherwin-Williams, best known for selling paint, are trying a new way to develop such industrial chemicals.
The plastic lining of a can is just 2 micrometers thick, less than one-eighth the thickness of a human hair. A ubiquitous yet invisible part of the modern world, can coatings must stand up to the rigors of manufacturing and then last for years in baths as acidic as lemon juice while preserving a seamless barrier between the food or drink and the metal. Any crack in that layer can mean corrosion or a weird metallic taste. Good performance ensures a 3-year-old can of Coke will have the same fizz and flavor as one bought yesterday.
Those properties can come at a cost, however. The coating is often made from bisphenol A (BPA), a chemical that has gained notoriety because of evidence that it can disrupt the dance of hormones that influence growth and development. But replacing BPA isn’t simple. Companies produced more than 6 million tons in 2018, making it one of the world’s most common synthetic chemicals. It is cheap, durable, and flexible—key building block not just in can linings, but also in products as varied as automobile dashboards and sales receipts.
A decade ago, in a nondescript two-story building on a dead-end street in a scruffy industrial neighborhood here, Mallen and colleagues embarked on a quest to find a molecule that could do the near-miraculous things BPA can do without the downsides. In place of the usual corporate research initiative shrouded in secrecy, the company tried something almost unheard of among chemical companies: It sought scrutiny from some of the industry’s fiercest critics.
“I haven’t seen another company take that approach,” says Tom Neltner, chemicals policy director for the Environmental Defense Fund, who met twice with company officials.
As a result of the effort, the manufacturer and prominent university scientists known for criticizing BPA have joined forces to check for possible health effects of a candidate chemical. The endeavor has produced a promising molecule while earning tentative praise from environmental advocacy groups, and some point to it as a model for how companies might uncover safer chemicals. But although Sherwin-Williams’s molecule is already in production, it must also win over price-sensitive can manufacturers and exacting food companies. And no one knows whether it will prove truly benign.
Mallen hardly has the makings of an iconoclast. The 59-year-old has spent his career at the paint and chemical company Valspar, which Sherwin-Williams bought in 2017. Starting as a laboratory chemist, he rose to become a company vice president in charge of regulatory affairs for the division that makes can coatings. Baby-faced, with neatly trimmed blond hair, he speaks with the measured, matter-of-fact manner of the upstate New York farm boy that he is. But when company chemists approached him in 2009 with a plan to find a new can coating, he decided the company would need to break from tradition.
Starting in the late 1990s, scientists had sounded the alarm about BPA, particularly its ability to mimic the hormone estrogen. In animal and epidemiological studies, university researchers linked BPA to breast and prostate cancer, reduced fertility, diabetes, genital defects, and altered behavior. The Centers for Disease Control and Prevention found the chemical in the urine of 93% of U.S. adults.
Although the chemical industry and federal regulators have argued that the evidence doesn’t prove the compound poses health risks at the levels found in people, many consumers now shun BPA. Today, it is banned from baby bottles and toddler cups in Canada, Europe, and the United States. In 2015, France banned BPA in all food containers. California now requires warning labels or signs for food and drink cans with BPA. And makers have scrambled for alternatives, which quickly made their way into products sporting “BPA free” labels.
But the alternatives have flaws. Some substitutes—often related chemicals in the bisphenol family—appear to have similar hormone-mimicking properties. Can linings made from other plastics have a shorter shelf life, work only for particular foods and drinks, or require more coatings in the factory—a major drawback on assembly lines churning out 2000 cans per minute.
Valspar was in the thick of the controversy. The company’s can coating division relied heavily on BPA and struggled to find satisfactory replacements. Although many companies were running away from bisphenols, those compounds still promised to outperform other chemicals such as polyesters or acrylics. So Jeff Niederst, a chemist heading Valspar’s efforts to develop linings, posed a question: What if Valspar could find a bisphenol that wasn’t an endocrine disrupter?
The suggestion set off a blizzard of meetings. Valspar faced a multimillion-dollar, multiyear research investment marked with uncertainty. And if the company stuck with bisphenols, it would need to prove itself to a public skittish about that class of chemicals. To succeed, Mallen recommended the team open its work to outside inspection. “This whole idea of embracing transparency,” Mallen recalls, “we were going to have to get very comfortable with that.”
In searching for a coating that wouldn’t disturb the endocrine system, Valspar found itself in an emerging area of toxicology. Existing chemical safety tests used by companies and government regulators often rely on decades-old methods that look for glaring effects such as changes in organ weight. That situation has prompted researchers to look for faster, more sensitive ways to flag hazardous chemicals and pinpoint safer ones. U.S. federal agencies, including the National Institutes of Health’s National Toxicology Program and the Environmental Protection Agency (EPA), have spent the past decade developing ways to quickly check thousands of chemicals by dosing cells cultured in petri dishes, a project called Tox21. The European Commission, meanwhile, is pouring €50 million into research to improve ways to identify endocrine disrupters.
But much of that work is still experimental, and none has produced a definitive set of endocrine tests. “I’m not sure we’ll ever get there,” says Bernard Robaire, a pharmacologist at McGill University who is part of a Canadian chemical-screening initiative.
So Valspar decided to go beyond the usual regulatory testing, cobbling together its own screening on the advice of scientists inside and outside the company. It hired Mark Maier, a toxicologist who had worked for a drug company and a pesticide industry group. He used a computer program to search hundreds of bisphenol compounds for molecular structures that appeared unlikely to bind with estrogen receptors.
Some bisphenols were prohibitively expensive. Others didn’t react well to form a polymer, a long chain of molecules, that would make up the can lining. The top contenders were sent to a private lab for testing on yeast cells engineered to glow in response to chemicals that acted like estrogen or testosterone. The most promising was tetramethyl bisphenol F (TMBPF), a rarely used chemical previously tested as an insulator in electronic circuit boards.
Although it cost 50% more than BPA, TMBPF was comparable in price to other BPA substitutes, company officials say. And it shared some of BPA’s toughness. TMBPF had another apparent advantage. Manufacturers use two chemical steps to forge BPA into can linings. In the second stage, some BPA molecules don’t get incorporated into the polymer chains. Those strays are thought to account for most of the BPA that leaches from linings. Valspar chemists found a way to use TMBPF just once, early in the process, stanching the release of the unlinked molecules. The Food and Drug Administration and EPA gave Valspar the green light to use the new chemical in cans in 2014, after routine tests.
Before Valspar put the chemical on the market, executives wanted to be sure it would also pass muster with people sounding the alarm about BPA. “We couldn’t afford to spend a lot of time working on a wrong material,” Mallen says. The company decided “to go to the expert critics” so that they, too, could scrutinize it.
With the help of a public relations firm, Burson-Marsteller (now BCW), Valspar crafted a plan to ask prominent BPA skeptics to study the chemical for endocrine effects. If the results were promising, Valspar would court environmental activists and journalists with a message of transparency.
Maier began by contacting Maricel Maffini, a Frederick, Maryland–based biologist who consulted for environmental groups on toxic chemical issues. The two had met in 2012 on a bus in Parma, Italy, at a food safety meeting. Maffini’s reputation preceded her: “I had heard all these horrible stories,” Maier recalls. “She was vilified by industry.”
Maier, however, was impressed with her thoughtfulness, so he asked her to look at data on TMBPF. In 2014, the company hired her as a consultant. Maffini says she wrestled with how colleagues in the environmental world might view her work for industry. “I thought, ‘OK, these are people that are trying to do something different. Just screaming into the wind is not going to lead us very far.’”
Maffini became a bridge. She had been a research assistant professor at the Tufts University lab of Ana Soto, a reproductive endocrinologist and leading scientist raising concerns about BPA. In 2015, Valspar and Maffini approached Soto about analyzing the chemical, and she agreed.
The company was hands-off. Rather than sign a contract with Soto, Valspar made a donation to Tufts with no implied obligation: She would study the chemical however she saw fit. “They didn’t try to twist my arm,” Soto says.
She put the molecule through tests. It didn’t seep from coatings when soaked in acid or alcohol baths meant to simulate various foods. More important, it did not activate estrogen receptors in engineered breast cells or spur growth of breast cancer cells, she, Maier, Maffini, and colleagues reported in February 2017 in Environmental Science & Technology. The molecule also didn’t affect several genes that respond to estrogen. “As far as estrogenicity, we are pretty sure that there is no activity, and we are happy with that,” Soto says.
For the company, the finding was a coup—a well-known independent researcher had declared that the molecule showed no signs of sharing one of BPA’s biggest flaws. When Mallen learned the results, he says, “I had quite a few scotches that night.”
Estrogen mimicking wasn’t the only concern, so Valspar sought out other experts. Thomas Zoeller, an endocrinologist at the University of Massachusetts, Amherst, and an authority on how BPA affects the thyroid, reviewed the company’s test data and organ samples. The results suggest “it doesn’t seem to affect the thyroid system,” Zoeller says.
Outside researchers didn’t deliver uniformly good news, however. Valspar approached scientists at Baylor Medical School, who found that TMBPF blunted estrogen’s effect on test cells and that a polymer made from the molecule had a similar effect on testosterone, according to a 2017 paper in PLOS ONE. Adam Szafran, a molecular biologist who helped lead the research, says the findings weren’t conclusive and could be specific to the prostate cells they tested.
Mallen acknowledges that those results raise questions about the compound. But he says company-sponsored research showed that changes in test cells don’t translate into effects on an entire organism. That study, published online in Food and Chemical Toxicology in October 2019, showed no endocrine-related effects on rats fed TMBPF for 3 months.
The cautious praise Sherwin-Williams is earning from chemical safety advocates suggests to Maier that other companies might benefit from emulating its approach. He points to the agribusiness giant Monsanto, now owned by Bayer, which has been hit with hundreds of millions of dollars in legal judgments over claims that its glyphosate herbicide caused cancers. “Monsanto wouldn’t be in the trouble they are now if they had taken a transparent, proactive, collaborative approach,” says Maier, now working as a consultant in Albuquerque, New Mexico.
But some companies see downsides to the public spotlight, says Paul Anastas, a founder of the green chemistry movement at Yale University. Businesses quietly working to find safer replacement chemicals worry that attention could bring unwanted questions about the safety of current products. “The nail that sticks out gets the hammer,” he says.
Nor have critics wholeheartedly embraced Sherwin-Williams’s new coating. “It’s a step forward,” Neltner says. But, he adds, “We can’t make a statement that this product is absolutely safe.”
And Soto hesitates to go beyond her estrogen-related findings. “I can’t tell you there isn’t something else there because I haven’t tested for it.”
That statement highlights a challenge for Sherwin-Williams: How far does it need to go to prove itself? The endocrine system’s complexity stymies easy assessment. The variety of hormones, their fluctuations throughout life, and their effects on different parts of the body create a kaleidoscope of ways a chemical might trigger effects. “It’s very complicated,” says Terry Collins, a chemist at Carnegie Mellon University. “You are really asking what God knows about how to build life.”
Collins is part of a group of scientists who proposed a five-step approach in 2013 to look for endocrine disruption. The process starts with computer simulations and ends with extensive testing on rodents. TMBPF hasn’t passed two of those tests yet: Will minuscule amounts cause problems, even though the much bigger doses used in standard toxicology tests don’t? And will exposure for one generation of lab animals and their fetuses translate into illness for offspring?
To satisfy European regulators, Sherwin-Williams is planning tests that “will definitely answer at least 95% of the remaining questions,” Mallen says. The tests will expose rats to TMBPF starting in utero, and potential U.S. and European academic collaborators could recommend additional tests piggybacking on the study. That new round of experiments will be “what a lot of people are waiting for,” Mallen says.
Despite the remaining questions, the company has started to sell its coating in the United States, branded as valPure V70. It has been used in 22 billion cans since 2017. That’s a modest fraction of the estimated 350 billion aluminum beverage cans and 100 billion steel food cans produced each year worldwide.
Winning more of the market is daunting. A fraction of a penny per container could sway canmakers, and BPA is cheaper. Though now used in just 10% of steel cans in the United States, BPA is still in roughly half of all aluminum cans, Mallen says.
And consumers show little sign of demanding any particular non-BPA product. Cans often look alike under the gloss of brand labels. Even Mallen can’t tell which cans are treated with his company’s product. At a lunch break in a company conference room, he turns a Diet Coke in his hand, looking for the tiny insignia of the company that made the can. “I don’t think they’re using V70,” he says. “Maybe they are.”
Sherwin-Williams hopes its molecule might become a universal can liner—like BPA without the drawbacks. Getting there will take more than chemical safety testing. While Mallen scrutinizes his soda can, researchers in nearby rooms run elaborate torture tests to perfect formulas that will work with as many foods and drinks as possible—olives, alcoholic ciders, sauerkraut, dog food. All that for the humble can.
Read more from Science's special issue on chemistry for tomorrow's earth.