LL2-22 digest — Ch22 Nanotechnology: early lessons from early warnings#

EEA Report No 1/2013, pp. 530–560 (text pp. 530–551). Authors: Hansen, Maynard, Baun, Tickner, Bowman. The chapter updates their 2008 Nature Nanotechnology article. It has no panels. Andrew Maynard, the project lead, is a co-author.

Core argument#

A prospective case, from the report’s “Emerging issues” part. The historical chapters look back at harms that happened. This one audits a young technology, while it was still being developed, against the 12 lessons of the 2001 volume. The authors’ verdict: stakeholders are “doing some things right, but we are still in danger of repeating old, and potentially costly, mistakes” (p. 542). Nanotechnology was unusual in that risk discussion accompanied its early development (p. 542). Efforts were “better than” for earlier technologies (p. 543). But coordinated action was slow, and twenty years after the first toxicological signals “many governments still call for more information as a substitute for action” (p. 547). The question, they conclude, is not whether the lessons have been learnt but whether they are being applied (p. 548).

Key evidence#

• Early warnings, described as “preliminary” (p. 536):
• Ultrafine TiO2 was far more inflammatory than larger particles in rat lungs (1990), and surface area was a better dose measure than mass (1992). NIOSH (2011) called ultrafine TiO2 a potential occupational carcinogen, based on one rat tumour study (p. 534).
• SWCNT granulomas (2004). Long MWCNTs gave asbestos-like responses (Poland 2008). Industry inhalation studies confirmed inflammation (p. 535).
• Nanosilver was widespread in consumer products, in unknown volumes. Ecotoxic effects appeared at µg/L, against modelled environmental levels in the low ng/L range. The REACH freshwater PNEC for silver (0.04 µg/L) sits only about an order of magnitude above those levels (my comparison). It was unresolved whether the nano form is worse than ionic silver (pp. 535–536).
• Contrary studies are presented (p. 534).
• Regulation (pp. 537–542):
• Chemicals are identified by CAS number, so nano and bulk forms look the same (TSCA, REACH).
• Tonnage thresholds are unsuited to nano volumes.
• Under TSCA, the regulator must show risk before it can demand data: “a classic regulatory paradox” (p. 537).
• Voluntary reporting failed: 13 UK submissions, 31 US organisations. The schemes were also “somewhat onerous” (p. 537).
• Registrants split on whether CNTs are new substances or graphite (p. 538).
• The 2009 Cosmetics Regulation (labelling, safety assessment) is the chapter’s main example of EU law that expressly differentiates nanomaterials. The text also credits the Biocides Regulation with nano data requirements (p. 547). Elsewhere, measures were incremental: US SNURs for some CNTs and Australia’s NICNAS changes (pp. 537, 541). The Novel Foods recast failed (p. 539).
• The authors feared nano would be “overlooked” in the 2012 REACH review, with the next revision not until 2019 (pp. 539–540).
• Assessment science (pp. 541–542):
• Mass-based dose measures, monotonic dose–response and the persistence/bioaccumulation (PBT) criteria may not transfer to nanomaterials.
• It has been claimed that there are up to 50,000 SWCNT variants, which calls into question whether case-by-case assessment is practical.
• Table 22.1 (p. 548) scores 11 expert reports and 4 EU regulations against the 12 lessons:
• Knowledge lessons are widely noted.
• Governance lessons are rarely noted, and alternatives never.
• No “+++” (strategy) appears anywhere. But section 22.8 describes strategic recommendations from several of the same bodies, so this reflects an unexplained coding threshold or an inconsistency, not an absence of proposals.

Authors’ lessons and recommendations#

Diagnostic: Legacy regimes “inherited” nanomaterials, and governments kept the status quo (p. 541). Several things block action (pp. 546–548): - the NNI both promotes the technology and oversees its risks; - disciplinary worldviews, such as the EPA’s “rooted in chemistry” (p. 543); - developers’ belief that knowledge diffuses one way, from research to application (p. 545); - limited public engagement; - bureaucratic inertia.

There is also a structural lag between products and regulators’ use of safety data, which the authors say “will not be less than two decades” (p. 549).

Advocacy: - act now on existing knowledge (p. 547); - separate promotion from EHS management (implicit, via approving quotes from NAS 2012 and Denison; pp. 546–547); - revise REACH now, at least for its “most obvious and short-term limitations”, rather than wait for 2019 (p. 540); - fund interdisciplinary infrastructure and the critical EHS research gaps (p. 543); - assess alternatives by function (“could” versus “should”, p. 546); - design criteria and green nanotechnology, with safety treated “equal to economic and performance considerations” (pp. 550–551); - public engagement (p. 545).

Precaution is framed as enabling “safe as possible” products (p. 547).

Main mechanisms#

1. Legacy categories: regulatory inheritance and identity keys that make new variants invisible.
2. Invisibility: voluntary disclosure fails and producers hold the data.
3. Burden of proof: the burden-of-proof paradox, and how ambiguity undermines REACH’s reversal of the burden.
4. Paradigm strain: the assessment toolkit doesn’t fit the new materials.
5. Combinatorial variety: too many variants to assess one by one.
6. Timing: products outpace safety data, and review cycles are slow.
7. Promoter–overseer conflation: it shapes research agendas.
8. Solution-push: hype, “nano-fever”, crowds out alternatives.
9. Lock-in: the design window closes as interests become entrenched.
10. Contested definitions: including a “competitiveness” ground for moving the threshold (p. 540).
11. Talk without action: anticipatory discussion substitutes for action.

Transferable insights (technology-neutral)#

Main caveats#

• Prospective case: there is no realised harm against which to test the lessons, and false positives aren’t considered.
• Hazard versus risk: the evidence is mostly about hazard, from animal studies with high doses and instillation. The language escalates from “preliminary” (p. 536) to “rapidly increasing evidence of risks” (p. 539).
• Few dissenting voices: there are no panels or industry or regulator responses. The only counter-voice is a brief quote from the Commission’s JRC (Stamm) on the definition (p. 540). Developers’ views are summarised from one survey and dismissed (p. 545).
• Proximity: about a quarter of the references are the authors’ own work. The authors co-produced some of the warnings (e.g. Poland 2008) and benchmarks.
• Table 22.1 method is undocumented: “10 reports” but 11 columns; no “+++” despite strategies described in 22.8; Biocides is scored blank on independence although the text credits it.
• Missing references and errors:
• key sources (Powell 2007; Lawless 1977; Ludlow et al. 2007) are missing from the reference list;
• probable unit errors (“50 mg” in Poland et al.; the seawater PNEC), both confirmed as printed;
• “not less than two decades” is presented as a floor, but its sources say “one or two decades” (Lawless) and “decades” (EEA 2001), and the chapter itself calls them “worst-case” (p. 549).
• Balance: the chapter is fair in places. It concedes that the regulatory lag is “not surprising”, credits progress, gives both sides in the definition dispute, and asks for lay knowledge to be critically appraised.
• Hindsight, to verify: REACH nano annexes applied from 2020; TSCA reform came in 2016 with nano reporting in 2017; the TiO2 and CNT classifications evolved after 2013.