The Stockholm Convention at a Crossroads: Questionable Nominations and Inadequate Compliance Threaten Its Acceptance and Utility

Abstract

Twenty years since coming into force, the Stockholm Convention has become a “living” global agreement that has allowed for the addition of substances that are likely, as a result of their long-range environmental transport (LRET), to lead to significant adverse effects. The recent listing of the phenolic benzotriazole UV-328 in Annex A and a draft nomination of three cyclic volatile methylsiloxanes (cVMS) for Annex B draw attention to the fact that many chemicals are subject to LRET and that this can lead to questionable nominations. The nomination of UV-328 and the draft nomination of cVMS also raise the spectre of regrettable substitutions. At the same time, atmospheric monitoring across the globe reveals that environmental releases of several unintentionally produced POPs listed in Annex C, such as hexachlorobenzene and hexachlorobutadiene, are continuing unabated, highlighting shortcomings in the enforcement of the minimum measures required under Article 5. There is also no evidence of efforts to substitute a chemical whose use has been known for three decades to unintentionally produce polychlorinated biphenyls. These developments need to be rectified to safeguard the long-term viability and acceptance of a global treaty of undeniable importance.

Short abstract

Recent nominations to the Convention expand the range of substances perceived as causing significant adverse effects as a result of long-range transport, while releases of hexachlorobenzene listed in 2004 continue or even increase.

The Stockholm Convention on Persistent Organic Pollutants (POPs)¹ was negotiated through the late 1990s and early 2000s, adopted in May 22, 2001, and entered into force on May 17, 2004. As two decades have passed since it became a globally binding international agreement aimed at preventing and reducing the release of chemicals that are likely as a result of their “long-range environmental transport to lead to significant adverse human health and/or environmental effects such that global action is warranted”, it is an opportune moment to reflect on the Convention’s current role and potential future from the perspective of researchers with a long-standing interest in POPs and their global control but no stake in the treaty’s implementation. In particular, we observe two developments that over time could threaten the acceptance and utility of the Convention. One relates to recent nominations to the Convention that highlight the great range of chemicals subject to long-range environmental transport (LRET) and the resulting risk of questionable nominations; the other relates to an apparent failure to ensure compliance regarding the release of unintentionally produced POPs. While the first appears to suffer from overambitiousness, the second could benefit from more ambition.

Recent Nominations Stretch the Notion of What Constitutes a POP

LRET is a key justification for the existence of the Stockholm Convention. Persistent, bioaccumulative, and toxic (PBT) substances leading to significant adverse human health and/or environmental effects can and often should be subject to national regulation but do not necessarily warrant global action. A country should have the right to regulate the use of a chemical that does not cause effects outside of its borders. Global action is warranted if the effects are likely to occur in humans and wildlife populations in jurisdictions different from those where the chemical is used and released, because regulating such releases is then outside of the purview of the affected national jurisdictions. This is the domain of the Stockholm Convention.

One of the undisputable strengths of the Convention is that it is a “living” agreement that in Article 8 lays out a process for listing chemicals in its Annexes A, B, and C. This process, which is initiated by the nomination of a chemical by a Party, has proven effective in adding chemicals to the Convention so that they become subject to the measures aimed at reducing or eliminating environmental releases globally. The process proved sufficiently flexible to lead to the listing of compounds such as perfluorooctyl sulfonate, perfluorooctanoate, and perfluorohexylsulfonate, which differ considerably from previously listed POPs, e.g., in terms of their mode of LRET, the importance of formation from precursor compounds, and the mechanistic basis of their bioaccumulation.

The nomination of the UV-stabilizing phenolic benzotriazole UV-328 and the draft nomination of the cyclic volatile methyl siloxanes (cVMS) expand the property profile of chemicals that can be considered for inclusion in the Stockholm Convention. UV-328 was added to Annex A during the 11th Conference of the Parties (COP) held in May 2023.² The nominators³ made the case for the potential for LRET of UV-328 on the basis of (i) relatively high values for the characteristic travel distance (CTD) and transfer efficiency (TE) calculated with the Organisation for Economic Co-operation and Development (OECD)’s P_ov and LRTP Screening Tool⁴ that indicate that UV-328 can reach remote regions with atmospheric particles and (ii) the presence of UV-328 in preen gland oil of seabirds from remote waters.⁵ It was hypothesized that seabirds can take up UV-328 from ingested plastic debris that had reached the remote region by floating in the ocean. According to these arguments, any chemical found in floating debris can be argued to potentially undergo LRET in the oceans. Similarly, the OECD P_ov and LRTP Screening Tool will calculate the same relatively high CTD and TE values for any chemical with a low volatility, i.e., a logarithmic octanol–air partitioning ratio K_OA above ∼11.

On June 14, 2023, the European Chemicals Agency released for public consultation a draft European Union Proposal to list the cVMS octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5), and dodecamethylcyclohexasiloxane (D6) in Annex B.⁶ cVMS are distinctly different from UV-328. They are very volatile and will not associate with atmospheric particles to any significant extent.⁷ It is clearly established that cVMS can be widely dispersed as vapors in the atmosphere, especially in winter.⁸ Should the cVMS also become a precedent, then not only involatile chemicals but any chemical with an atmospheric half-life longer than 2 days and less volatile than D4 would have to be considered to fulfill the LRET screening criterion (Figure 1).

Figure 1.

Volatility scale based on the octanol–air equilibrium partitioning ratio K_OA or the saturation vapor pressure of the (subcooled) liquid P_L, ranging from very volatile chemicals on the left to very involatile chemicals to the right. If UV-328 is deemed capable of LRET on the basis of results obtained with the OECD P_ov and LRTP Screening Tool, any chemical with volatility low enough to be substantially particle bound in air in that model would automatically fulfill the LRET criterion. If D4 is deemed capable of sufficient atmospheric deposition to be able to cause adverse effects in remote regions, any chemical with a volatility lower than D4 would fulfill the LRET criterion unless it is degraded in air with a half-life shorter than the 2-day threshold stipulated in Annex D. The properties for UV-328 were estimated with polyparameter linear free energy relationships;⁵² those for the cVMS are from Xu and Kropscott.⁵³

In practice, detection in remote regions is often an additional requisite for establishing LRET in a successful nomination to the Convention. While this will constrain the number of chemicals eligible for nomination, this may constitute merely a small obstacle. The number of substances found in samples from remote areas is large and increasing rapidly because of constantly improving analytical capabilities.⁹ Nontarget and suspect screening of Arctic air samples detected “over 700 compounds of interest in the particle phase and over 1200 compounds in the gaseous phase”.¹⁰ Similarly, hundreds of unrecognized halogenated contaminants were reported in polar bear serum.¹¹

It is not surprising that chemicals with diverse properties are subject to LRET. Modeling studies have shown that organic chemicals covering a large expanse of the chemical partitioning space have a strong potential to accumulate in the Arctic as a result of LRET and to even bioaccumulate there, provided that they are emitted to air and are relatively persistent.¹²⁻¹⁴ Furthermore, in a recent evaluation of the LRET of 12 615 high-production-volume chemicals conducted using the model in the OECD P_ov and LRTP Screening Tool, 24% were identified as having an LRET greater than that of existing POPs.¹⁵

Consequently, many PBT chemicals will fulfill the LRET screening requirement. This could eventually lead to very large numbers of chemicals being eligible for nomination and listing in the Convention. For example, ECHA’s PBT assessment list currently includes 229 unique substances, although chemicals on the list may ultimately be assessed not to have PBT properties.¹⁶ Three hundred ninety-seven substances on the Canadian Domestic Substances List were assessed to meet categorization criteria for persistence, bioaccumulation, and inherent toxicity.¹⁷ The current SIN (“Substitute It Now”) list of the International Chemical Secretariate includes 86 chemicals designated as PBT,¹⁸ all of which have a volatility lower than D4. While there are differences in the definition of P, B, and T between different regulations, most of the substances on the named lists would fulfill the PBT screening criteria in the Stockholm Convention.

A large increase in the number of chemicals that fulfill the screening requirements could lead to a considerable expansion in the number of chemicals being eligible for nomination and listing. The current rate of adding chemicals to the Annexes of the Convention is already straining or even exceeding the capacity for implementation of many Parties to the Convention. Wang et al.¹⁹ write that “as of March 2021, over 70 Parties have not submitted their updated National Implementation Plans (NIPs) addressing chemicals listed since 2009.” An increase in the pace of nominations would inevitably exacerbate these issues. Furthermore, greatly expanding the number of chemicals nominated and listed could well jeopardize the acceptance of the Convention. If the LRET criterion can no longer serve to discriminate chemicals requiring global action from those that do not, signatories could view the Convention as a threat to their sovereign right to formulate their own chemical policy. Without broad acceptance, the Convention will become irrelevant.

One possible consequence of the large number of chemicals eligible for nomination is that political considerations take on a larger role in the nomination process. Such a situation would be rife with potential conflicts of interest, compounded by manifest power imbalances between the different parties to the Convention. Notably and maybe not surprisingly, almost all nominations so far originated in the Western European and Other States Group (WEOG).

Rather than this scenario, we believe that the ongoing success of the Convention depends upon strengthening the role of science in the nomination and listing process. To do so, we should first accept that there are many chemicals that will satisfy the four screening criteria in the Convention (P, B, T, and LRET), without necessarily fulfilling the requirement for being listed in one of the Annexes. The screening stage was conceived under the premise that few chemicals would possess the specified POP properties. Now that we realize that many chemicals possess these properties, the screening approach is obsolete. We suggest that the screening stage should be adapted to create a prioritization stage, with the purpose of identifying chemicals for the second stage of the assessment, the risk profile. In addition to the current screening criteria, the prioritization process would incorporate other relevant criteria that are not formally enshrined in the Convention but that may have been considered by member states when deciding whether to prepare a nomination, such as the chemical’s production volume and release pattern. A nomination dossier would motivate why the assessment of the nominated compound is prioritized over other potential POPs. As a consequence of the prioritization, the available resources could then be concentrated on assembling the risk profile with great care, establishing “whether the chemical is likely, as a result of its long-range environmental transport, to lead to significant adverse human health and/or environmental effects”.¹

Worryingly, the listing of UV328 and the draft nomination of cVMS suggest that the nomination and assessment process is currently not working well. UV-328 was listed in Annex A on the basis of a purported risk of secondary poisoning. However, the purported risk was derived from a comparison of a predicted no-effect concentration for chronic effects with an exposure outlier.²⁰ Exposure was based on the highest concentration ever reported in seabirds, not considering that the concentration in seabirds is extremely variable. Looking at more appropriate measures of chronic exposure, the median measured concentration in the seabirds was a factor 300 lower than the maximum, which would lower the risk quotient from 0.1 to 0.0003. Using the median of all studied seabirds would lower the risk quotient by at least another order of magnitude, but it would be undefined because UV-328 was only detected in 21% of the sampled seabirds.⁵

We want to clarify that we do not exclude the possibility that plastic additives may travel over considerable distances in the ocean with floating debris and that such transport could fall within the Convention’s definition of LRET.²¹ However, we maintain that the mere presence of UV-328 in floating marine plastic²² and the possibility of some floating debris to cover large distances²³ are insufficient to demonstrate that this is a significant LRET process for UV-328. We further stress that the presence of UV-328 in birds sampled on Marion and Gough Island,⁵ while indeed implicating leaching of UV-328 from ingested plastic as a source, is not evidence that the plastic debris reached these very remote islands by LRET. In fact, the lack of correlation of UV-328 concentration, or the likelihood of encountering a bird with a high UV-328 concentration, with proximity to source areas,⁵ is difficult to reconcile with systematic LRET but implicates the possibility of local plastic sources, e.g., from fishing equipment. We note that the seas around Marion and Gough Island are within or close to areas where the Japanese and Chinese off-shore fishing fleet is active,²⁴ and fishing gear has been shown to be a significant source of marine plastic in the open ocean²⁵ and to contain UV-stabilizing benzotriazoles.²⁶ Furthermore, the large-scale surface ocean currents in these areas do not indicate the existence of a flow path originating in highly populated coastal areas.²⁷ In summary, we believe that this risk profile demonstrates neither that UV-328 causes significant adverse effects as a result of LRET nor that the few measured cases of high exposure are the result of LRET.

cVMS are proposed for nomination although there is considerable evidence that airborne cVMS cannot be deposited from the atmosphere in notable amounts.⁸ The absence of notable deposition makes it difficult to establish the likelihood of adverse effects as a result of LRET, as the potential for human and wildlife exposure to atmospherically transported cVMS is miniscule.²⁸ In the draft nomination dossier for the cVMS, it is argued “that because of the PBT properties of D4, D5, and D6, the atmospheric redeposition does not need to be a significant source of these compounds to cause concerns and to require minimisation of the emissions into the atmosphere”.⁶ In other words, significant exposure of humans or wildlife in the remote region is not required, which stands in direct contradiction to the Convention’s criterion. These two examples show that increasing the range of chemicals nominated to the Convention can increase the risk of questionable nominations. To cope with this, more rigorous application of science is required, particularly when assembling the risk profile.

Recent Nominations Continue to Pay Insufficient Attention to the Possibility of Regrettable Substitution

Regulatory restriction can lead to regrettable substitutions i.e., the restricted chemical is replaced with a chemical with similar or worse properties.^29,30 The Stockholm Convention is not immune to this issue, as short-chain chlorinated paraffins, listed in 2017, were often replaced with medium-chain chlorinated paraffins, which now themselves are under consideration for listing under the Convention. Other examples include the replacement of polybrominated diphenyl ethers with related substances that share many of the same undesirable attributes. It appears that the risk of regrettable substitutions is still not appropriately addressed. This can be illustrated using the nominations discussed above.

A clear rationale for singling out UV-328 from a wide range of other phenolic benzotriazoles (UV-BTs) is lacking. Other involatile UV-BTs have the same predicted CTD and TE as UV-328, and more volatile UV-BTs, such as UV-P, arguably may have a higher potential for atmospheric dispersal.³¹ Other UV-BTs are as, if not more, bioaccumulative than UV-328.^32,33 Similarly, there is little evidence that would suggest that UV-328 is more toxic³⁴ and more persistent than its structural relatives. Indeed, the SIN list of PBT-designated substances includes seven phenolic benzotriazoles other than UV-328 (UV-360, UV-350, UV-320, UV-327, UV-329, UV-234, and UV-928).¹⁸ In light of this, it is remarkable that the Risk Management Evaluation of the Convention mentions “other phenolic benzotriazoles” as “widely available alternatives” for the substitution of UV-328.³⁵ It is very likely that UV-328 is and will be substituted with related chemicals of comparable functionality without any indication that these substitutions will have a lower likelihood to lead to significant adverse human health and/or environmental effects as a result of LRET. While there is wording in the risk management evaluation of UV-328 advising of the danger of regrettable substitution, there is really no mechanism in place that could prevent this from occurring. Without such a mechanism, there is little reason to believe that the regrettable substitutions that arose from earlier listings of POPs (e.g., medium-chain chlorinated paraffins, polybrominated diphenyl ethers) would not materialize again. What then has been accomplished with the listing of UV-328 as a POP?

Substitution of cVMS would likely be far more challenging, considering the rather unusual combination of their partitioning properties (highly hydrophobic yet volatile), which is essential to their functionality. The cVMS share these unusual partitioning properties with certain polyfluoroalkyl substances (PFASs). For example, D5 has very similar partitioning properties as 1H,1H,2H,2H-perfluorododecyl acrylate (10:2-FTAC) and 1H,1H,2H,2H-perfluorododecyl methacrylate (10:2-FTMAC).³⁶ While we do not know whether PFASs would be considered viable cVMS replacements, they are potent greenhouse gases and known to degrade to bioaccumulative, toxic, and extremely persistent perfluoroalkyl acids.³⁷ There is every reason to believe that they would have a higher likelihood to lead to significant adverse human health and/or environmental effects as a result of LRET than cVMS. If cVMS were to be replaced with PFASs, that may give a whole new ring to what constitutes regrettable substitution.

While decisions to regulate chemicals that are clearly deserving of global restrictions should not be influenced unduly by the risk of regrettable substitutions, safe substitution should be a consideration when nominating chemicals to the Convention. In particular, if substitution with substances with very similar or worse characteristics can be anticipated, it may be appropriate to nominate those likely substitutes at the same time. Above we questioned whether UV-328 really meets the threshold for a POP. If, however, it is deemed to meet the threshold, it is counterintuitive to tolerate substitution with related substances that clearly would meet that same threshold.

Past experiences and the substances currently being listed as POPs illustrate the persistent and seemingly intractable problem of regrettable substitution. While Wang et al.¹⁹ see the primary research need in this context in the identification of safer alternatives, we believe that the search for, and implementation of, safer alternatives should be the responsibility of industry. We rather see a primary research need in the design of policies that would incentivize industry to avoid regrettable substitutions. Interestingly, the Convention already includes text that points in this direction: Provisions in Article 3, paragraphs 3 and 4 oblige Parties to the Convention with assessment schemes for chemicals and pesticides to “take measures to regulate with the aim of preventing the production and use of new pesticides or new industrial chemicals which, taking into consideration the criteria in paragraph 1 of Annex D, exhibit the characteristics of persistent organic pollutants”.

Currently, in most jurisdictions whatever chemical is not explicitly restricted can presumably be made and used. In some jurisdictions, industry may incur liabilities if they knowingly release chemicals in quantities likely to cause adverse effects. Is it possible to transfer the responsibility of not producing and releasing chemicals which exhibit the characteristics of POPs from states to emitters? Efforts should be devoted to devising chemical safety evaluation procedures that implement the provisions of Article 3 and which could then by adopted by the Parties to the Convention.

Apparent Complacency in Response to Increasing Releases of Unintentionally Produced POPs

Hexachlorobutadiene (HCBD) was listed as a POP under Annex A in 2015 for elimination and under Annex C as an unintentional product in 2017. In 2017, monitoring data from Okinawa indicated high and rapidly increasing atmospheric concentrations in East Asia.³⁸ More recent results from the Japanese monitoring efforts document that greatly elevated concentrations of HCBD prevailed until at least 2021, whereby average concentrations in nationwide surveys continue to exceed 2000 pg m^–3.³⁹ Air concentrations of HCBD reported for the Chinese atmosphere are many orders of magnitude higher than these levels.⁴⁰ Shunthirasingham et al.⁴¹ demonstrated that increases in HCBD concentrations in the atmosphere are a global phenomenon and did not start in 2017 but had in fact been ongoing since at least 2008. HCBD is now by far the most prevalent POP in the global atmosphere. At the same time, Wang et al.⁴² reported massive and rapidly increasing releases of HCBD in China, mostly associated with the chlorinated solvent industry.⁴³

Hexachlorobenzene (HCB) was among the first batch of POPs regulated under the Convention. Based on air monitoring across Europe conducted in 2006 and 2016, Halvorsen et al.⁴⁴ concluded that HCB was not only the compound detected in highest concentrations among the investigated POPs but also the only POP whose concentrations had significantly increased between the two study years. This trend was confirmed by Fiedler et al.⁴⁵ who found ∼50% higher air concentrations of HCB globally in 2017–2019 compared to 2010/11. More than 10 years of global monitoring revealed that after a drop in HCB air concentrations between 2005 and 2008, levels have been steadily and significantly rising between 2008 and 2016.⁴¹

While the deliberate production of polychlorinated biphenyls (PCBs) is restricted by being listed in Annex A of the Convention, they are also listed in Annex C, meaning that they count among the unintentionally produced POPs. The unintentional production of several different PCB congeners from the use of 2,4-dichlorobenzoyl peroxide as an initiator in silicone and polyester production has been known since the early 1990s.⁴⁶ Twenty-five years later, studies revealed that the use of this chemical continues to be a notable and widespread source of emissions of PCBs to the indoor⁴⁷ and outdoor atmosphere.^48,49 Many other sources of unintentionally produced PCBs have been identified. Because most monitoring efforts do not quantify the relevant congeners, no knowledge of temporal trends in their release rates currently exists.⁵⁰

Collectively, there is thus ample evidence that releases of several unintentionally produced POPs not only continue but in many cases are increasing rapidly. Article 5 of the Convention lists the minimum measures that Parties shall take to reduce releases of the unintentionally produced POPs listed in Annex C, specifically related to action and implementation plans. It is obvious that at least some Parties do not live up to these requirements and have been failing to develop and implement plans that would curb the unintentional production of HCB and HCBD. Furthermore, there is no record of any discussion of, let alone response to, the issue of increasing releases of unintentionally produced HCB and HCBD in the documents produced during the most recent COPs.

There are further indications of complacency in dealing with unintentionally produced POPs. For example, we are not aware of any Party having implemented laws or policies related to the management of releases connected to the use of 2,4-dichlorobenzoyl peroxide, even though paragraph (5c) asks for Parties to “require the use of substitute or modified materials, products and processes to prevent the formation and release of the chemicals listed in Annex C.″¹ We did not find any evidence of the existence of guidelines adopted by decision of the COP (as mentioned in Article 5) related to the use of 2,4-dichlorobenzoyl peroxide.

It is possible to produce chlorinated solvents such as tri- and perchloroethylene on a large scale without the release of massive quantities of HCBD. The UNEP Guidance on preparing inventories for hexachlorobutadiene observes that in facilities with best available technologies (BATs) “HCBD is recycled [···] together with other high-boiling byproducts or is destroyed by high temperature incineration.″⁵¹ Similarly, chlorine-free substitutes for 2,4-dichlorobenzoyl peroxide that do not result in the production of PCBs clearly exist.⁴⁸ The problem of unintentional production of POPs thus does not appear to be one of technical feasibility but rather failure to adopt BATs. The Parties to the Convention may consider developing mandatory technical guidance documents, where timelines for the adoption of BATs are stipulated in instances where failure to do so leads to the Convention’s ineffectiveness in reducing releases globally.

In conclusion, we caution against nominations that are motivated primarily by fulfilling the screening criteria instead of the likelihood of significant adverse effects as a result of LRET, because this could threaten the long-term viability of the agreement and thus ultimately weaken, rather than strengthen, global chemical safety. This is particularly so, as the Parties to the Convention appear disengaged in assuring compliance with its requirements regarding POPs already listed in Annex C. We urge all actors to redirect some of the zeal with which some questionable nominations are being pursued toward reducing releases of chemicals already listed in the Convention.

Biography

Frank Wania is Professor of Environmental Chemistry in the Department of Physical and Environmental Sciences at the University of Toronto at Scarborough since 1999. He has wide-ranging research interests related to environmental contaminant transport and distribution, with a focus on gaining a mechanistic understanding of contaminant enrichment processes through a combination of fieldwork, laboratory experimentation, and model simulations. He studied environmental science at the University Bayreuth in Germany and received his Doctorate in Chemical Engineering and Applied Chemistry from the University of Toronto in 1995.

The authors declare no competing financial interest.