3M

A technician kept working after he was told to stop

Between masking tape's 1925 introduction and 3M's reported completion of its PFAS-manufacturing exit at the end of 2025, the company repeatedly faced the same problem from opposite directions.¹² How could a large industrial corporation keep an uncertain technical observation alive long enough to learn what it meant—and how could it stop a successful technology once evidence of harm appeared? 3M became known for answering the first question. Its fluorochemical history shows that the second required different routes for evidence, authority, and public challenge.

3M's corporate chronology describes a failed 1902 attempt to mine corundum and the company's subsequent survival as an abrasives manufacturer.¹ In the early 1920s, laboratory technician Richard Drew delivered sandpaper samples to an automobile body shop and saw painters struggle to make clean borders on two-tone finishes. The Smithsonian object catalog says Drew obtained permission to develop a masking material, was ordered to return to his regular work after weeks without success, continued the experiment, and brought masking tape to market in 1925.³

The account also places the original problem outside the laboratory: painters showed Drew what the existing materials could not do. Turning the experiment into a marketed product necessarily added manufacturing and commercial work to the laboratory result. The organizational inference is narrower than the company legend: at least in this documented case, continued work outside an assignment produced a useful result that formal direction would have ended.

Management turned an exception into an institution

William McKnight, who rose from assistant bookkeeper to president and later chairman, made initiative part of a broader philosophy of delegated technical work. A company-commissioned history reproduces his 1944 argument that expanding organizations must delegate responsibility and tolerate honest mistakes. The same history says Richard Carlton established a central research laboratory in 1937 to supplement divisional laboratories and investigate longer-range problems.⁴ These records establish the stated design and formal investment, not how evenly employees received discretion or which unsuccessful projects disappeared.

3M says that since about 1948 it has encouraged employees to spend roughly 15 percent of working time pursuing ideas, gathering resources, and forming teams.⁵ Its corporate history also describes the Technical Forum, organized in 1951, as a voluntary peer network for exchanging problems and ideas across divisions.⁶ Central research, divisional laboratories, discretionary time, and a peer network offered different routes by which an observation could survive and find another use. The available company records describe those mechanisms but do not measure portfolio-wide access to them or the rate at which they caused a commercial success.

A weak adhesive needed a network, not only slack

The Post-it Note makes that movement visible, although public institutional accounts disagree on the first date. The National Inventors Hall of Fame dates Spencer Silver's discovery of the removable acrylic-microsphere adhesive to 1968; a Smithsonian account based on Art Fry's testimony dates it to 1970.⁷⁸ Both support the more important sequence: Silver continued explaining an adhesive without an obvious application, and Fry later connected it to another problem.

Fry recalled that in 1974 he wanted a bookmark that would stay in his church hymnal without damaging the page. He tested Silver's adhesive, and coworkers began using the prototypes as notes and requesting more. Fry's Smithsonian account also describes the manufacturing, quality, packaging, and sales work needed to make the combination into a product.⁸ The Hall of Fame profile dates the national launch to 1980 after years of production and design work.⁷

The case is therefore more than a parable about an accident. Silver needed a way to preserve and circulate a result with no application; Fry's colleagues became early users; and several functions had to make a new category reliable and demonstrable. Protected exploration preserved the possibility. A technical network and repeated use made it legible.

Six Sigma put search through a defect-reduction gate

That community met a different operating logic when James McNerney arrived from General Electric as chief executive around the turn of 2001. 3M's 2001 annual report says the company introduced Six Sigma in February, trained about 4,000 employees that year, and had more than 1,000 projects under way. It reports manufacturing-yield and cost examples as program benefits, but the same filing shows that aggregate operating income and net income fell amid weak demand, currency effects, and restructuring.⁹ The filing therefore supports the program's scale and management's reported process gains, not a simple claim that Six Sigma caused company-wide profitability.

A longitudinal peer-reviewed study describes Six Sigma's low initial fit with 3M's innovation culture, employee resistance, and the subsequent adaptation of both the practice and the culture.¹⁰ Contemporary reporting adds the research-level detail: staged reviews and forecasts pressed scientists to identify customers, returns, and schedules while projects were still uncertain.¹¹ That report said the share of sales from products introduced within five years fell from roughly one-third to one-quarter during McNerney's tenure and that successor George Buckley reduced Six Sigma's reach in research while retaining process discipline elsewhere.¹¹ The timing and participant reports show a serious conflict; they do not isolate Six Sigma as the sole cause of the sales-share change.

McNerney moved to Boeing in 2005.¹¹ The links to GE and Boeing provide comparative context for the executive's career. They do not establish identical engineering cultures or prove that a specific practice produced the same effect in all three companies.

The correction preserved two different kinds of work inside one corporation. Where a known production process should yield the same safe result repeatedly, variation can be a defect. Where researchers are still discovering the relationship between a material and a use, variation is often the evidence. 3M's new-product measure gave executives one view of the commercial portfolio. It did not show every cost created by a technically successful product.

PFOS showed where technical evidence lost authority

Fluorochemicals gave fabrics, food packaging, and firefighting foams valuable resistance to water, grease, and heat. They also joined two activities that 3M's celebrated innovation story tends to separate: invention and toxicology. By 2000, EPA described PFOS as used in those product categories and said 3M's data showed persistence, tissue accumulation, and potential long-term risk.¹² Minnesota's official exhibit index for its litigation against 3M lists company records on fluorochemicals in blood, 1978–79 rat and monkey studies, meetings with toxicologist Harold Hodge in 1979, and a 1998 Kris Hansen laboratory notebook.¹³ The index establishes the dates and existence of those records; Sharon Lerner's document- and interview-based investigation supplies the account of what the studies found and how people inside 3M responded.¹⁴

In 1997, 3M chemist Kris Hansen was assigned to determine why a laboratory kept finding PFOS in blood supplied by the American Red Cross. She and junior scientists cleaned instruments, tested materials, repeated the work on other machines, and obtained blood from several locations. The signal persisted. Lerner reports that archived pre-commercial blood tested negative, that Hansen later found earlier internal evidence, and that managers challenged her methods and restricted parts of the work while senior medical staff reassured her about exposed workers.¹⁵ This is an independent reconstruction based substantially on internal documents and retrospective interviews, not a judicial finding about any individual's intent.

Hansen's experience reverses the Drew and Silver stories. She had the anomaly, the instruments, persistence, and a peer team. What she lacked was a reliable path by which evidence distributed across specialties could challenge the people responsible for an established product line. Her dependence on her job, her narrower expertise in chemical detection, and senior medical reassurance all shaped what she could contest. Technical autonomy was real, but it was not the same as authority over commercial use or public exposure.

Communities and regulators forced a second accounting

After negotiations with the Environmental Protection Agency, 3M announced in May 2000 that it would phase out the affected PFOS chemistry. EPA's contemporary announcement said 3M's data showed persistence, accumulation in human and animal tissue, and potential long-term risk; the agency also said it was still evaluating health and environmental effects.¹² A production phaseout could reduce new releases from those chemistries but could not by itself remove persistent material already released.

Residents and public officials in communities east of the Twin Cities encountered that durability as a drinking-water problem. Minnesota's settlement program says the state sued in 2010 alleging that 3M's PFAS production and disposal damaged drinking water and natural resources. The 2018 settlement provided $850 million, with about $720 million after expenses directed to drinking-water and natural-resource projects, including treatment systems and private-well testing.¹⁶ The official program establishes the state's allegations, the settlement terms, and the funded response; settlement did not produce an adjudication of every underlying liability or health claim.

The accounting widened again when a federal court approved 3M's 2024 settlement with U.S. public water suppliers. The approval order states that 3M agreed to pay $10.5 billion to $12.5 billion on a thirteen-year schedule for a class that included roughly 12,000 public water systems facing testing, treatment, and compliance costs.¹⁷ The order also says the agreement is not an admission of liability. It resolves specified water-system claims rather than measuring all health effects, responsibility, or remediation costs.

3M says it completed its exit from PFAS manufacturing at the end of 2025. Its current account also says historical treatment and remediation continue and that third-party PFAS inputs and some product transitions or uses may continue after that date.² The statement is a company report rather than an independent audit, and a manufacturing exit is not the same as removal of past releases.

The concept fingerprint joins invention to answerability

Knowledge, expertise, and professional autonomy and learning, quality, and reliability explain how technical anomalies could survive long enough to be tested. Innovation, entrepreneurship, and renewal captures the institutional routes that moved selected anomalies into products. Governance, stewardship, and accountability and culture, informal organization, trust, and voice mark the harder boundary: whether workers, scientists, regulators, and communities could make a profitable anomaly answerable to evidence of harm.

Evidence still needed

• First-person records from production workers and affected communities would broaden a record now weighted toward company histories, government and court records, and Hansen's experience.
• Independent verification of the manufacturing exit and an inventory of residual third-party inputs, uses, releases, and legacy sites would test the scope of the company's 2025 statement.
• Project- and portfolio-level data would be needed to estimate how often discretionary time, technical forums, or Six Sigma gates changed innovation outcomes rather than merely appearing in prominent examples.
• Settlements identify large drinking-water and natural-resource responses but do not provide a complete distribution of health, ecological, worker, and public costs.

Across this century, the memorable devices—Drew's body-shop observation, Silver's weak adhesive, Fry's hymnal bookmark, Hansen's mass spectrometer—each began as a difference someone could have dismissed. What changed was who could carry that difference, what kind of consequence counted, and who could compel a response. 3M's history leaves the inquiry open: can the routes that let an inventor preserve a promising anomaly also let workers, scientists, regulators, and communities stop a profitable one before its costs become an inheritance?