The ozone layer — that thin, invisible shield protecting us from the sun's harshest ultraviolet rays — is on the mend. But scientists have just spotted a small leak in the rules that could hold its recovery back by about seven years.

A new study led by MIT and published in Nature Communications has, for the first time, put hard numbers on a quiet problem: industrial chemicals that are still legally allowed under the Montreal Protocol are escaping into the atmosphere far more than anyone realised.

The good news, the researchers stress, is that this is science doing exactly what it's supposed to do — catching a problem early enough to fix it.

What the ozone layer does, and why it matters

The ozone layer sits high in the stratosphere, roughly 15 to 35 kilometres above our heads. It absorbs most of the sun's ultraviolet radiation, the kind linked to skin cancer, cataracts and damage to crops and ecosystems.

In 1985, British Antarctic Survey scientists discovered a growing hole in it over Antarctica. The culprits — chlorofluorocarbons, or CFCs, then used in fridges, aerosols and air conditioning — were rapidly identified, and in 1987 the world's governments signed the Montreal Protocol to phase them out.

It worked. The Protocol is widely considered the most successful environmental treaty ever negotiated, and the ozone layer has been steadily healing. Some estimates have it returning to 1980 levels by around 2040.

The loophole

When the Protocol was drawn up, it carved out an exception for ozone-depleting chemicals used as "feedstocks" — raw ingredients converted into other materials, such as plastics and nonstick coatings. The thinking at the time was simple: industry wouldn't deliberately leak something it had paid to produce, and only about 0.5 per cent was thought to escape.

That assumption no longer holds. Monitoring data from the Advanced Global Atmospheric Gases Experiment (AGAGE), a global network co-founded at MIT, now puts the average leakage rate closer to 3.6 per cent — and higher still for some chemicals.

"We've realised in the last few years that these feedstock chemicals are a bug in the system," said Susan Solomon, a professor of environmental studies and chemistry at MIT, who was part of the original team that linked CFCs to the ozone hole back in the 1980s.

"Production of ozone-depleting substances has pretty much ceased around the world except for this one use, which is when you have a chemical you convert into something else."

Seven years lost — but recoverable

The international team modelled three scenarios out to the year 2100. If leakage stayed at the assumed 0.5 per cent, the ozone layer would recover to 1980 levels by 2066. With zero leakage, 2065. But at the real-world rate of 3.6 per cent, recovery slips to 2073 — a delay of about seven years.

"We've gotten to the point where, if we want the protocol to be as successful in the future as it has been in the past, the parties really need to think about how to tighten up the emissions of these industrial processes," said first author Stefan Reimann of the Swiss Federal Laboratories for Materials Science and Technology.

What happens next

Member countries of the Montreal Protocol meet every year, and feedstock emissions are already on the agenda for their working groups. The study's authors hope their findings will prompt tighter rules, better leak controls, or substitution with chemicals that don't damage ozone at all.

Professor Solomon is optimistic. "There are a lot of innovators in the chemical industry," she said. "They make new chemicals and improve chemicals for a living… There are thousands of other chemicals that could be used instead, so why not switch?"

As Reimann put it, the message is simple: "We could reduce the period of ozone depletion by years. It might not sound like a long time, but if you could count the skin cancer cases you'd avoid in that time, it would seem quite significant."

The ozone story remains, on the whole, a triumph. This is just the next chapter of getting it right.