The ozone layer was supposed to be one of the great environmental comeback stories. Under the 1987 Montreal Protocol, nearly every nation on Earth agreed to phase out chlorofluorocarbons (CFCs), the synthetic chemicals that had torn a hole in the atmospheric shield protecting life from the sun’s harshest ultraviolet rays. Decades later, monitoring data confirmed the strategy was working. The ozone layer was healing, and full recovery over mid-latitude regions was expected around the middle of this century.
Now, peer-reviewed research published in April 2026 in Nature Communications warns that recovery could come as many as seven years late, not because of rogue nations or black-market refrigerants, but because of a loophole hiding in plain sight inside the factories that make the replacement chemicals themselves.
The factories meant to fix the problem are part of it
The research centers on a counterintuitive problem: manufacturing hydrofluorocarbons (HFCs) and hydrofluoroolefins (HFOs), the newer refrigerants and propellants designed to replace CFCs, generates controlled CFCs as chemical by-products. Some quantity of those by-products escapes into the atmosphere during production.
For decades, regulators and scientists assumed that these “feedstock losses” amounted to roughly 0.5% of total feedstock volume, a figure small enough to be considered negligible. But the new atmospheric modeling study tested what happens when real-world leak rates exceed that benchmark. The results were stark: at higher, more realistic emission levels, mid-latitude ozone recovery slows significantly, with delays reaching up to seven years unless additional measures are taken to curb the leaks.
“The feedstock loophole is the single biggest uncontrolled source of CFC emissions remaining under the Montreal Protocol,” said Luke Western, an atmospheric scientist at the University of Bristol and a co-author of the study. “If we do not address it, we are looking at a measurable delay in ozone recovery that was entirely avoidable.”
A separate study, also published in Nature Communications, used Bayesian statistical analysis to trace unexpected CFC concentrations in the atmosphere back to modern HFC and HFO production. That finding shifts the narrative. The problem is not just old refrigerators and crumbling building foam slowly releasing chemicals from a bygone era. It is active, ongoing manufacturing on modern production lines.
What the air itself is saying
Atmospheric measurements back up the modeling. The National Oceanic and Atmospheric Administration’s Global Monitoring Laboratory tracks concentrations of ozone-depleting gases through a network of stations worldwide, feeding data into NOAA’s Annual Greenhouse Gas Index and Ozone Depleting Gas Index. Those records show that certain regulated CFCs, including CFC-113, are declining more slowly than Montreal Protocol phaseout schedules predicted.
That sluggish decline is consistent with ongoing industrial releases rather than a clean break from production. If the only remaining CFC sources were old equipment and insulation gradually breaking down, atmospheric scientists would expect a smoother, more predictable decay curve. The monitoring data suggests something else is adding to the signal.
A regulatory gray zone
Part of the explanation lies in how international and domestic rules treat these emissions. The Montreal Protocol effectively banned direct CFC production for most commercial uses, but it carved out exemptions for CFCs used as “feedstocks,” meaning chemicals consumed during a reaction to produce something else, and for limited “process agent” applications.
The U.S. Environmental Protection Agency maintains this distinction in its own regulations. A Federal Register notice published in October 2024 updated rules related to process agent uses. In theory, feedstock CFCs are destroyed during manufacturing and never reach the atmosphere. In practice, if a reaction does not fully consume the feedstock, the leftover CFC can leak out, and those emissions fall outside the tighter controls applied to direct production.
This creates a gap. The legal framework assumes near-complete destruction. The atmosphere suggests otherwise.
What scientists still cannot pin down
For all the strength of the atmospheric evidence, significant uncertainties remain. No publicly available dataset tracks real-time CFC by-product emissions from individual HFC or HFO factories. The Nature Communications studies work from the top down, inferring industrial leak rates from what monitoring stations detect in the air and testing emission scenarios until they match observed trends. That approach is scientifically rigorous for identifying global patterns, but it cannot assign precise blame to specific plants, companies, or countries.
Chemical manufacturers have not publicly disclosed their feedstock emission rates, and EPA enforcement data tied to specific violations at fluorochemical facilities has not surfaced in available compliance databases. Without facility-level measurements, the gap between the assumed 0.5% loss rate and actual emissions remains a well-supported estimate rather than a directly measured fact.
Complicating matters further, legacy CFC “banks” still exist. Old refrigeration units and building insulation installed decades ago continue to release CFCs as they degrade. Earlier research estimated these aging chemical stocks could delay ozone recovery by multiple years on their own. Separating the atmospheric fingerprint of legacy emissions from active industrial leaks is technically difficult, and the two sources overlap in monitoring data. Determining how much each contributes remains an open question, one that directly affects where regulators should focus their efforts.
Global compliance with process agent exemptions under the Montreal Protocol is also poorly documented. Countries report aggregate production and consumption figures, but independent verification of what happens inside individual factories is limited. In fast-growing chemical sectors, unreported or underreported emissions could widen the gap between official inventories and what the atmosphere actually contains.
What closing the loophole would take
The Montreal Protocol is often called the most successful environmental treaty ever negotiated. The World Meteorological Organization’s most recent scientific assessment, published in 2022, projected that the ozone layer over mid-latitudes would return to 1980 levels by approximately 2045, with the Antarctic ozone hole closing around 2066. A seven-year delay would not undo those gains entirely, but it would extend the period of elevated ultraviolet exposure linked to skin cancer, cataracts, and damage to crops and marine ecosystems.
Researchers behind the new modeling argue that the delay is avoidable. Closing the gap would require tighter international audits of fluorochemical manufacturing, better integration of atmospheric monitoring with facility-level reporting, and potentially new rules that treat feedstock losses with the same urgency as direct CFC production. Concrete steps could include mandating continuous emissions monitoring at large plants, requiring independent verification of destruction technologies, and linking international trade in fluorochemicals to transparent reporting of by-product controls.
Multilateral reviews under the Montreal Protocol could then compare this bottom-up industrial data with top-down atmospheric trends, flagging discrepancies for investigation before they accumulate into years of lost progress.
A treaty tested by its own supply chains
The treaty has not failed. But the science published this spring makes clear that its success depends on accounting for emissions the original agreement never fully anticipated: by-products quietly leaking from the very supply chains built to replace the chemicals the world agreed to stop using.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
