The Chemicals That Keep Moving

Stopping a process is not the same as ending a presence.

This is the unglamorous sequel to the DDT story, and to others like it. A compound is useful. It is stable—stability being another word for does not go away. It is used widely because it works. Years later the use ends, in a jurisdiction, on paper. The molecules do not read the paper.

They remain in sediment, in fat, in indoor dust, in the blood of people born after the factory closed. They move when we do not intend them to: on air masses, in ocean currents, in the bodies of fish and birds, in the slow rearrangement of a riverbed after a dam is pulled. The civic vocabulary of “ban” is a verb about us. Persistence is a property of them.

Persistence as a design feature

Many of the chemicals that later became famous as pollutants were not accidents of sloppy synthesis. They were made to last. A dielectric fluid that decomposes in a transformer is a bad dielectric fluid. A pesticide that vanishes in a week may be a poor one if the pest does not. A coating that sheds water and oil is, at the molecular level, a refusal to react.

Established: a subset of industrial and agricultural compounds resist photolysis, hydrolysis, and microbial attack on timescales of years to decades, sometimes longer. Organochlorine pesticides, polychlorinated biphenyls (PCBs), polychlorinated dibenzo-p-dioxins and furans, and many per- and polyfluoroalkyl substances (PFAS) belong, in different ways, to that subset. Their half-lives in soil, sediment, or the human body are not a single number. They depend on the congener, the matrix, the temperature, the oxygen. The direction is the point. Some substances leave. These do not, not on a human planning cycle.

Lipophilicity is the second design feature that becomes a problem. Compounds that dissolve in fat more readily than in water partition into living tissue. If the animal lives longer than the compound’s metabolic clearance, the body becomes a reservoir. If a predator eats many such bodies, the reservoir concentrates. Ecologists call this bioaccumulation when the organism’s concentration exceeds that of its surroundings, and biomagnification when concentrations rise with trophic level. The vocabulary is less important than the geography: the highest numbers often appear not at the factory but in an animal that never saw it.

Probable, and well supported for the classic organochlorines and for several PFAS: food-web concentration is not a rare pathology. It is what you should expect when a stable, fat-seeking molecule is released into a system that eats.

Transport without permission

Air, rivers, ocean currents, and migrating animals move residues across borders that legislation does not. Polar regions have received chemicals they did not manufacture. Downstream communities drink from catchments they did not spray. This pattern is established enough to be a principle: persistent substances have a geography larger than their market.

The physical idea has a name that sounds like a cartoon and is not. Semi-volatile compounds evaporate in warm conditions and condense in cold ones. Over seasons and latitudes they hop: soil to air, air to water or snow, snowmelt to soil again. Frank Wania and Donald Mackay, in papers from the early 1990s, formalized this as global distillation and global fractionation. More volatile members of a family travel farther and faster; less volatile ones linger nearer the source. E. D. Goldberg had used “global distillation” in 1975 for halogenated hydrocarbons. The Arctic evidence that followed was not a metaphor. It was residues in air, snow, seawater, and in the fat of animals that live where the compounds were never produced in quantity.

Established: long-range atmospheric and oceanic transport of persistent organic pollutants (POPs) to the Arctic and, to a lesser documented extent, the Antarctic. Established: cold trapping—deposition favored by low temperature—helps explain why a compound used in temperate agriculture or industry can appear in a polar food web. Plausible as a complete map, still under study in the details: how much of a given Arctic burden arrived in one atmospheric pulse versus many hops, how climate warming is remobilizing stores in ice and permafrost, how much of a local Arctic signal is now from local sources (airports, sewage, dumps) rather than from the south.

The Arctic Monitoring and Assessment Programme, operating since 1991 under what became the Arctic Council, has spent decades making that geography numerical. Its assessments underpinned the political case that POPs were not a domestic waste issue. They were a hemispheric one. Indigenous communities whose diets include marine mammals sit at the top of a food web that has been receiving other people’s chemicals for two generations. That is not a slogan. It is a measured exposure pathway.

You do not have to care about this because a campaign told you to. You can care because a molecule does not need your consent to travel, and because the people who receive it were not at the product meeting.

The river that remembered

A ban can stop a pipe. It cannot vacuum a sediment.

PCBs were manufactured in the United States, principally by Monsanto under Aroclor trade names, for use in transformers, capacitors, hydraulic fluids, and a long list of lesser applications that exploited their stability and low flammability. Production for most uses ended in the late 1970s under the Toxic Substances Control Act; EPA banned manufacturing in 1979. The Hudson River had already been receiving them. Between about 1947 and 1977, General Electric’s capacitor plants at Hudson Falls and Fort Edward discharged PCBs—EPA has cited estimates on the order of 1.3 million pounds—into the upper river. The compounds stuck to particles. They settled. When the Fort Edward Dam was removed in 1973, stored sediment moved downstream. New York restricted fishing. EPA listed some 200 miles of the Hudson as a Superfund site in 1984. Dredging of contaminated upper-river sediment, after years of argument about whether stirring the bed would do more harm than leaving it, removed on the order of 2.75 million cubic yards in the 2009–2015 seasons.

The numbers are local. The pattern is not. Estuaries, harbor muds, lake cores, and agricultural soils worldwide hold PCBs, DDT and its metabolites, and other organochlorines applied or spilled before the law caught up. In anoxic sediment the half-life can be long enough that a core is a history book. Fish advisories are the public-facing version of that book. They are not proof that the water is “dirty” in the nineteenth-century sense. They are proof that a reservoir remains biologically available.

DDT in soil and sediment is the same kind of unfinished sentence. Agricultural use ended in many countries decades ago. DDE still appears in raptor eggs, in human milk in some regions, in polar bears. The concentrations are generally lower than they were. Lower is not zero. Recovery of peregrines and eagles, treated at length elsewhere in this magazine, is evidence that reducing input works. It is also evidence that the clock of a persistent molecule is not the clock of a press cycle.

Established: legacy POPs remain in environmental reservoirs after production bans. Established: those reservoirs can continue to feed food webs. Unknown in any one harbor until it is measured: whether the remaining mass is declining fast enough, buried deep enough, or still cycling.

A treaty for things that outlast governments

The Stockholm Convention on Persistent Organic Pollutants was adopted on 22 May 2001 and entered into force on 17 May 2004. It began with twelve substances, the so-called dirty dozen: eight pesticides (aldrin, chlordane, DDT, dieldrin, endrin, heptachlor, mirex, toxaphene), two industrial chemicals (PCBs and hexachlorobenzene, the latter also a pesticide), and two classes of unintentional by-products (dioxins and furans). Parties agreed to eliminate or restrict production and use, with structured exemptions. DDT was placed on Annex B, restricted rather than immediately eliminated, with an acceptable purpose for disease-vector control. PCBs in existing equipment were given a long sunset. A review committee was created so that the list could grow.

It has grown. Later listings have included, among others, certain brominated flame retardants, perfluorooctane sulfonic acid (PFOS) and related substances, perfluorooctanoic acid (PFOA), and further organochlorines. The United States signed and has not ratified; it nonetheless regulates many of the same chemicals under domestic statutes. The treaty is not a magic wand. It is an admission, written into international law, that persistence plus long-range transport plus bioaccumulation is a class of problem no single parliament owns.

That class definition is the intellectual core. Toxicity at a given dose still has to be shown, compound by compound. Stockholm does not say that every persistent molecule is a poison at environmental levels. It says that if a substance is sufficiently persistent, bioaccumulative, and capable of long-range transport, the burden of proof and the duty to act are international. The alternative is to wait until a thinning shell, or an Arctic body burden, becomes undeniable in public. We have already run that experiment.

The current name

Public conversation now often turns to PFAS: per- and polyfluoroalkyl substances, used for water and stain resistance, firefighting foams, nonstick coatings, and industrial processes that need a carbon–fluorine bond because that bond does not break easily. EPA’s own primer is careful on a point the headlines often are not. There are thousands of PFAS. They are found in water, air, fish, soil, and in the blood of people and animals far from manufacturing sites. Some have been linked, in epidemiological and toxicological studies, to harmful effects. The class is not a single verdict.

PFOA and PFOS are the two with the longest paper trails. They persist. They occur in drinking water near manufacturing, military firefighting-foam sites, and some landfills. The C8 Science Panel, studying communities along the Ohio River exposed to PFOA from a DuPont plant, reported probable links to several clinical outcomes, including kidney and testicular cancer, elevated cholesterol, pregnancy-induced hypertension, thyroid disease, and ulcerative colitis. EPA has classified PFOA and PFOS as likely human carcinogens in its drinking-water work. Other PFAS are far less studied. Some are replacements for the ones that were restricted, which is an old pattern: solve for the molecule that is in the newspaper, discover its cousin in the product line.

Stockholm has listed PFOS, PFOA, and related substances under its procedures. That is not the same as a global cessation of all fluorinated chemistry. It is a recognition that the most documented members of the family meet the POP criteria.

Established: extreme persistence of many PFAS; widespread detection in the environment and in human blood; long-range transport of some members, including to the Arctic, by air (especially volatile precursors) and by ocean (especially perfluoroalkyl acids). Established for PFOA and PFOS: a substantial health literature, with agency reviews concluding risk at exposures that have occurred in contaminated communities and, at lower levels, in the general population—though the shape of the dose–response at the lowest environmental concentrations remains argued. Unknown, in the strict sense, for most of the thousands of structures: specific toxicity, specific transport, specific substitutes’ regrets.

The editorial method should not change with the acronym. Ask how long it lasts, where it goes, who is downstream, and what the evidence actually shows at the doses that exist.

Detection is not a diagnosis

Analytical chemistry has improved faster than toxicology. Instruments now report concentrations in water at parts per trillion, and in some methods lower. That is a public good. It is also a standing invitation to confusion. A number that would have been a non-detect in 1980 can be a headline in 2024 without any change in the water.

Established: presence can be measured at levels far below those at which a clinical effect has been demonstrated for many substances. Also established: for some substances, including lead and several POPs, effects have been found at doses once dismissed as trivial, and the history of underestimating low-dose harm is real. The two facts do not cancel. They require a distinction the news cycle hates.

A detection is a fact about a sample. Harm is a fact about a relationship between dose, duration, organism, and endpoint. Some PFAS detections in remote rainfall are evidence of planetary transport; they are not, by themselves, a body count. Some drinking-water detections near a foam-training site are in the range where agencies have set health-based values; those are a different sentence. Flying River will not collapse them into one.

The counterargument from industry, in every generation, is that the dose makes the poison and that modern levels are negligible. Sometimes that is true. The counterargument from advocacy, in every generation, is that persistence plus detection plus a bad cousin equals guilt. Sometimes that is a reasonable precaution. Neither slogan is a method. The method is to keep the chemistry specific, the exposure quantified, and the uncertainty labeled.

Probable: as detection limits fall, more of the planet will appear “contaminated” by compounds that have been circulating unnoticed. That is a change in knowledge, not always a change in risk. Plausible: some of those newly visible compounds will later join the list of substances we wish we had constrained earlier. The DDT and PCB records exist to make that plausibility expensive to ignore, not to make it automatic.

Replacement is a kind of delay

When a persistent chemical is restricted, the market looks for a substitute that does the same job. The substitute is often less studied, sometimes less persistent, sometimes merely persistent in a different way. Organophosphate insecticides replaced much agricultural DDT and brought their own acute toxicity. Hydrofluorocarbons replaced chlorofluorocarbons and became a climate problem. Short-chain PFAS have been offered as improvements on long-chain ones; whether they are improvements in the full life cycle is a research question, not a brochure.

This is not an argument against substitution. It is an argument for examining persistence and mobility before scale, rather than after the sediment has been named a Superfund unit. The Stockholm criteria—persistence, bioaccumulation, long-range transport, adverse effects—are a checklist that could have been applied, in spirit, to tetraethyl lead and to DDT. They were applied late. Applying them early is the only inexpensive version of this story.

Unknown: what the next abbreviation will be. The structural temptation is always the same. Make a bond that weather cannot break. Ship the product. Discover, in a core or a polar bear or a tap, that the bond did what it was hired to do.

What “banned” does not mean

Treat a prohibition as the start of a monitoring era, not the end of a story. The Hudson is still a PCB river in the sense that fish still carry a memory of Fort Edward. Arctic air still delivers compounds whose factories are in other time zones. PFAS in groundwater will outlast the press conference that announced a drinking-water standard.

We should not wait for a thinning shell every time. We also should not announce a verdict before the measurements. Those two restraints can occupy the same sentence.

Follow the connection far enough and the moral is almost dull. A useful molecule that does not degrade is a loan from the future. The interest is paid by whoever lives downstream, downwind, or later. You cannot see the debt in the product. That was never the requirement. The requirement is to stop confusing the end of a process with the end of a presence.

Sources

• Wania, Frank, and Donald Mackay. “Global Fractionation and Cold Condensation of Low Volatility Organochlorine Compounds in Polar Regions.” Ambio 22, no. 1 (1993): 10–18.
• Wania, Frank, and Donald Mackay. “Tracking the Distribution of Persistent Organic Pollutants.” Environmental Science & Technology 30, no. 9 (1996): 390A–396A.
• Goldberg, E. D. “Synthetic Organohalides in the Sea.” Proceedings of the Royal Society of London B 189 (1975): 277–289.
• Semeena, V. S., and Gerhard Lammel. “The Significance of the Grasshopper Effect on the Atmospheric Distribution of Persistent Organic Substances.” Geophysical Research Letters 32 (2005): L07804.
• Arctic Monitoring and Assessment Programme (AMAP). Assessments on persistent organic pollutants, mercury, and chemicals of emerging concern in the Arctic (from the 1998 assessment onward; later reports on climate influence and PFAS).
• Lohmann, Rainer, et al. Cross-cutting AMAP-related work on PFAS in the Arctic environment, wildlife, and humans (including 2024 synthesis on long-range transport).
• Stockholm Convention on Persistent Organic Pollutants. Adopted 22 May 2001, entered into force 17 May 2004. Text and listing decisions, including Annexes A–C and subsequent amendments (PFOS, PFOA, and others).
• U.S. EPA. Hudson River PCBs Superfund Site: site history, 2002 Record of Decision, and dredging summaries (discharge estimates on the order of 1.3 million pounds, 1947–1977; approximately 2.75 million cubic yards dredged in Phase 2 operations).
• U.S. EPA. “PFAS Explained.” Agency overview of the class, occurrence, and uncertainties.
• C8 Science Panel. Probable link reports on PFOA (C8) exposure in Mid-Ohio Valley communities.
• U.S. EPA. Drinking-water regulations and supporting documents for PFOA and PFOS (including 2024 National Primary Drinking Water Regulation materials).
• Toxic Substances Control Act restrictions on PCB manufacture; EPA ban on PCB manufacturing, 1979.
• Reiersen, Lars-Otto, Katrin Vorkamp, and Roland Kallenborn. “The Role of the Arctic Monitoring and Assessment Programme (AMAP) in Reducing Pollution of the Arctic and around the Globe.” Environmental Science and Ecotechnology (2024).