Speaker Interview with OFI – Austrian Research Institute

There are unfortunately still a lot of challenges, and the 2030 goal of a minimum of 10% recycled plastic in any type of food packaging will be extremely difficult to reach.

From a safety perspective, our results indicate that classic consumer misuse (e.g. someone storing a pesticide in a bottle) is not the main safety concern. The most critical contaminants that we detect in recycled plastic actually originate from the food packaging itself.

Some types of adhesives and printing inks can form critical DNA-reactive degradation products under heat stress, and there are even warnings from the producers that high-temperature treatments have to be avoided. But in a recycling process it is impossible to avoid high temperatures, and we need solutions to prevent these packaging components from entering the input streams if we want to achieve safe circular recycling.

But I have to stress: for PET bottle recycling, which has already been approved by EFSA, we are not seeing any safety concerns.

For virgin plastic, the traditional approach to safety assessment focuses on intentionally used substances. At least in theory, one should know what was used in production, and can then test for these substances with targeted analysis.

For recycled plastic, this formulation-based approach is impossible, as you can never know what types of contaminants to expect. There are nearly endless possibilities: many of them completely harmless compounds from previously stored food, but also potentially hazardous substances. It is impossible to identify and risk-assess all these substances for each individual batch, so the regulatory focus is on the recycling process rather than on the individual article.

To prove that a recycling process is working, you need to know the worst-case level of input contaminants, the decontamination efficiency, and a toxicological threshold for residual contaminants in the final recycled plastic. Currently, as a worst-case assumption, all contaminants have to be assumed to be DNA-reactive mutagens. DNA-reactive mutagens have extremely low thresholds in EFSA’s Threshold of Toxicological Concern concept: by a factor of 120 lower than any other substance group, and also much lower than the thresholds for classic poisons like cyanide or arsenic. This makes it very difficult to demonstrate that a recycling process is safe.

If we look at current regulatory developments for virgin plastic, the differences between the safety assessment and testing of virgin and recycled plastic will shrink. The newest amendment of the plastics regulation (EU) No 10/2011 also requires the worst-case assumption that all impurities in virgin plastics are mutagens. And the plans for a new European framework regulation on food contact materials are moving away from the traditional formulation-based approach, aiming to treat non-intentionally added substances in the same way as intentionally used substances.

It is not easy to answer this in general for all types of packaging, as it depends heavily on the packaging type and waste stream. We have seen very promising results for all three technologies, but there will certainly still be challenges in establishing them at scale.

We also need to find the right balance. Measures like de-inking can help improve safety and reduce the risk of possible mutagenic impurities in the recycled plastic, but they also result in higher prices and higher energy consumption. So we always have to consider whether this improved safety is really needed. Measures like de-inking make sense for critical applications like food and cosmetic packaging, but we should not worry too much about trace amounts of mutagenic contaminants in other applications.

We will never be able to avoid mutagens 100%, and we are always exposed to trace amounts of mutagenic substances via food and the environment. If we use recycled plastic from printed input streams to produce a flower pot, there will be no significant health risk for the consumer, even if tomatoes are planted in this pot.

From 2030, we will have a mandatory recycled content of at least 10% in all plastic food packaging. I don’t think this can be achieved exclusively through better sorting, de-inking and recycling processes without compromising safety. We will also have to rethink design for recycling. Current design-for-recycling approaches do not yet consider the safety aspects of recycling back into food packaging. If food packaging is to be recycled in a circular way, we need to consider these safety aspects already at the design stage.

We already have one perfect example of good “design for safe circular recycling of food packaging”: the PET bottle. It is never directly printed, but has labels that can be easily separated by sink-float separation, and adhesives that can be washed off during the hot wash. This is one of the reasons why we have never seen any safety concerns in our testing of recycled PET bottles.

However, the screw caps were originally never recycled back into food contact, and here there is still room for improvement. Ironically, it can be exactly the “Recycle Me!” print on the caps that makes safe circular recycling more difficult. Some of the colours currently used in screw caps are also not sufficiently stable to withstand the strongly increased heat stress of recycling processes and can form critical degradation products.

For screw caps, it would therefore help to avoid printing the caps, or to use inks that can be washed off in a caustic hot wash. And for colourants that cannot be removed, it would be important to select colours that remain safe after the increased heat stress of the recycling process, which often reaches much higher temperatures than the original HDPE processing.