Final Report Summary - NANOFATE (Nanoparticle Fate Assessment and Toxicity in the Environment)
Executive Summary:
Executive Summary
Description of the work performed during the NanoFATE project and the main results achieved: All of the nine main NanoFATE S&T objectives, were successfully addressed at the end of the project, and can be summarised as follows:
I. Sourced, produced and fully characterised the required versions of commercial engineered nano particles (ENPs) (Ag, ZnO and CeO2) and developed and matched the tagged version of ZnO as closely as possible (eventually using Co doping). (Main Obj. 1)
II. Established particle behaviour in the pure ecotox media to be used and at higher than environmental concentrations. Modified the ecotox media and protocols to optimise ENP behaviour and organism performance to allow undertaking of hazard assessment studies under most realistic exposure conditions possible (Main Obj. 3, 4 & 5)
III. Established acute and chronic toxicity that enabled selection of relevant doses for the progression of the work into environmentally relevant media conditions. (Main Obj. 5, 6, 7 & 8)
IV. Developed initial simple assumption based fate models and estimated worst case environmental concentrations. Refined these model assumptions as real data and understanding (NanoFATE and externally derived) became available and identified parameters of most influence and importance for ENP specific fate and risk assessments(Main Obj. 2 & 8)
V. Train all staff cross discipline, but especially the 14 PhDs and Post Docs directly employed on NanoFATE, and disseminate our findings to other EU NMP projects and stakeholders (Main Obj. 9)
Delivery was divided in the 3 components below covering the 6 RTD and the dissemination WPs.
Particle chemistry, behaviour and fate component:
Work package 1 (Characterisation & tracking of ENPs) provided high quality well characterised commercial particles for the remaining project partners. Subsequent work focused on characterisation of these ENPs in exposure media & natural systems showing that the dissolution rates of ENPs (ZnO, Co-doped ZnO & Ag) not only depended heavily on the water and media involved, but particularly for zinc oxide on the particle morphology & coating. Work package 2 (ENP environmental behaviour & fate modelling) initially identified & prioritised specific properties to be considered during the development, adaptation & validation of environmental fate models for ENPs. Initial fate models were developed to predict CeO2 deposition in soils & nano ZnO & Ag influent to and discharge from sewage treatment works. WP2 then parameterized the ENP environmental fate modelling with information on dissolution, agglomeration (homo- & hetero-agg.), sedimentation & aging (transformations) properties. Kinetic models building on colloid theory have been developed to deliver fate predictions at the EU & regional scale.
Ecotoxicology and bioavailability component:
Work Package 3 (ENP Ecotoxicology) developed improved standard ecotoxicological exposure protocols, that ensures relevant & homogenous presentation of ENPs during toxicity testing. Employing these improved protocols, the exposures needed for the hazard assessment was completed. This included both acute and chronic testing, plus specific work to deliver data on bioavailability drivers for WP4. Samples have been archived for use in WP5 (some retained for future collaborative explorations) and sensitive endpoint data was collated for use in PNEC assessment (WP6). As for CeO2 little evidence of toxicity was seen across the tested organisms extra work was included to look at long term (6 months) accumulation and also co-exposure of CeO2 with organic contaminants, while for ZnO in freshwater a relavant confounding factor was seen as UV radiation meaning combined ZnO x UV exposures were prioritised. Sufficient data was produced to increase by 50% the species data points available for use in Species Sensitivity Distribution derivation in WP6. Work package 4 (ENP bioavailability) has produced and published critical reviews of all available information addressing data quality & identifying which environmental factors have the greatest proven effect on the bioavailability & toxicity of ENPs to organisms. This identified pH, organic matter & cation effects to be the most important for bioavailability and trials testing these were consequently implemented within WP3. For the non inert ENPs of Ag and ZnO in waters dynamics seemed either fast with changes mainly occurring within hours or so slow (changes over weeks to months) that test protocols would yield appropriate hazard data, as test are either short (1-3 days) or requires media exchanges at the same timescale of a few days. Contrastingly for soils fate processes appeared slow in comparison to the 2-8 week duration of the tests. Hence for soils additionally long-term exposures (6-12months) addressing ageing were conducted. The main findings included establishing that Zn toxicity in all cases was highest at lower soil pH; that coated ZnO ENPs were more toxic to springtails than uncoated ones; that ZnO ENPs & non-nano ZnO particles in soil released Zn ions very slowly (still going on after 12 months), with this release being highest at intermediate exposure levels; that ageing of ZnO ENPs increased Zn concentrations in pore water, but reduced the toxicity for both nano ZnO, non-nano ZnO & ZnCl2; & that release of Zn into soil pore water was much slower for coated than uncoated ZnO. The longterm aging for Ag ENPs in soil produced increasing toxicity over the 12 month period eventually reaching, but not exceeding, the toxicity of the ionic Ag exposures (on a weight for weight basis). Work package 5 (ENP toxicokinetics & toxicodynamics) developed sample handling protocols & preservation systems enabling successful tissue banking. A tiered approach was developed for tracking of ENPs in tissues that focused activities in a systematic manner exploiting high throughput methods in initial sample screening, and samples only passing to complex & costly analyses for the most detailed studies if ENP presence was likely. Samples were analysed for biological markers of ENPs to develop the knowledge of signatures of possible early ENP tissue damage. An agreed data structure was developed with a deliberately simple and flexible format to enable data entry at the labbench and export to the Cluster database when established. Toxicokinetic studies were completed for ENPs in soil & aquatic invertebrates and in situ characterisation of ENP uptake has been undertaken.
Risk assessment and knowledge transfer component:
Work package 6 (Integrated risk assessment) has assessed ENP production & product incorporation estimates based upon a review of the peer-reviewed as well as grey literature on production volumes of the three NanoFATE ENPs. The initial fate & distribution models (from “local” to “EU region” scale) developed produced very informative initial conservative (high) estimates of predicted environmental concentration (PEC) maps. Improved fate models and updated production estimates has refined these models and have been combined with protective environmental Predicted No Effect Concentrations (PNEC) based on species sensitivity distributions developed in NanoFATE.
The final refined modelling suggests that at current predicted levels of nano Ag and ZnO are unlikely to exceed thresholds of concern for aquatic or soil species. However, the hazard data is very sparse (especially for soils) and in details the conclusions look as follows:
In soils two quite separate prediction methods gave similar soil PECs for the UK (proven to represent the EU worst case in terms of land application of WWTP sludge). For nano silver the highest predicted value over a 12 yr period in UK soil scenario was 0.009 mg/kg whilst the lowest effect level of concern (EC50) for nano Ag in soil was 4.8 mg/kg (2-orders of magnitude higher). However, due to the paucity of toxicity data and SSD approach was not possible and hence the ECHA guidance document for deriving PNECs is followed which means using an assessment factor of 1000 on the most sensitive species, and then the nano Ag PNEC would be 0.0048 mg/kg and the NanoFATE PECs exceed this. A recent soil PNEC proposed for silver REACH registration proposed 1.24 mg/kg (W/W) based on read across from silver ions to which nanosilver is presumed to correspond. This would err towards the NanoFATE predicted nanosilver soil concentrations being of low risk. But more nanosilver soil toxicity data is needed.
For nano ZnO in soils the highest predicted value for nano ZnO over a 12 yr period in UK soil scenario was 0.939 mg/kg. The lowest effect level of concern for a single species for nano ZnO in soil was 119 mg/kg (2-orders of magnitude higher). However, again the paucity of toxicity data leads to inclusion of an assessment factor of 1000 and the resulting nano ZnO PNEC should be 0.119 mg/kg which the NanoFATE predictions exceed. Therefore, more nano ZnO soil toxicity data is needed.
In waters, the predicted starting effluent concentration for nano ZnO is closer to its lowest PNEC than for nanosilver suggesting that at the point of discharge to the river environment nano ZnO is the greater concern. If the current best estimate of a rapid loss mechanism for nanosilver in rivers is accepted, then for Europe around 50% of rivers would be predicted to have concentrations of 0.0001 ng/L or less and 95% of rivers have 0.1 ng/L or less of nanosilver in the expected (not worst case) scenario. Thus, most European rivers would have nanosilver levels 2-orders of magnitude below the Ag PNEC of 29 ng/L. For nano ZnO around 50% of rivers would be predicted to have concentrations of 1 ng/L or less and 95% of rivers have 500 ng/L or less. This is only 1 order of magnitude below the PNEC of 2200 ng/L. This analysis puts nano ZnO as being of greater concern to the aquatic environment than nanosilver. This is because of the apparent much higher use of nano ZnO and in this simulation the high water column settlement rate chosen for nanosilver (an effective half-life of as little as 45 min for nanosilver compared to 15 d for nano ZnO).
Work package 7 (Dissemination and Training activities) developed the Project Website and a series of Electronic Newsletters (circulated to more than 800 contacts) for the external profile and communication of results and designed to be readable & interesting for a broad lay audience. Two summary articles were published in “International Innovation” with hardcopies delivered to hundreds of European decision makers. The website was overhauled to accompany these publications and yielded excellent statistics: a cumulative total of 51,052 unique visitors to the site representing 894,159 page visits. In this way a significant audience was informed of eight NanoFATE training workshops & modules organised jointly with other NSC projects, 33 contributions to NanoSafety Cluster events, & 29 conference presentations. On the publication side, results are flowing: 21 peer reviewed ISI articles, 7 currently submitted, 4 mass media publications and 46ISI papers under preparation. Regulators have been engaged through targeted emails, website updates, distribution of leaflets at the Commission’s 2nd Regulatory Review on Nanomaterials Workshop, NanoFATE members have been invited to OECD WPMN working group meeting by national regulators to act as scientific experts and have delivered policy advice also at various national meetings and to sit on the scientific committee for a forthcoming ECHA workshop on nanomaterials.
Project Context and Objectives:
Project Context and Objectives
NanoFATE was conceived to fill knowledge and methodological gaps currently impeding sound assessment of environmental risks posed by engineered nanoparticles (ENPs). Our vision was to assess environmental ENP fate and risk in for example high-volume products for which recycling is not an option, namely; fuel additives, personal care products and antibacterial products. To represent these products two commercial ENPs of CeO2, ZnO and Ag (of varying size, surface and core chemistries) were followed through their post-production life cycles, i.e. from environmental entry as “spent product”, through waste treatment to their final fates and potential toxic effects. The applicability of current fate and risk assessment methods and identify improvements required for assessment of ENPs at an early stage were tested.
Objectives: Delivery of a systematic study of the environmental fate and toxicity of selected ENPs will entail addressing nine S&T objectives:
1. Design, tagging and manufacture of ENPs
2. Analysis of ENP interactions with abiotic and biotic entities
3. Generating predictive models for ENP exposure in waters and sludge-amended soils
4. Studying the fate and behaviour of ENPs through wastewater treatment
5. Determining acute and chronic ecotoxicity
6. Assessing effects of physico-chemical properties on ENP bioavailability
7. Defining mechanisms of uptake, internal trafficking, and toxicity
8. Developing spatial RA model(s)
9. Improving understanding of ENP risks
Background
To support the responsible development of the nanotechnology sector, it must be recognised that the development of environmental risk assessment methods should not lag too far behind those for human health. Past experiences highlight a number of other environmental issues such as organochlorine pesticide usage, endocrine disruptions, secondary effects of pharmaceuticals on wildlife, and genetic modification, where environmental impacts rather that direct effects on human health, emerged as the major area of concern. In each of these cases, the unexpected nature of these effects had a profound affect on public confidence in new technologies. This required that rapid regulatory action was put in place to control and mitigate risks. By ignoring effects on the environment, nanotechnology runs the risk that similar damaging and costly effects could occur.
Because of the initial and wholly understandable focus on direct risk to human health, knowledge of fundamental aspects of the environmental risks associated with ENPs is low in several key areas. These include:
- the post-production fate of ENPs from entry into the environment to final residence;
- how ENP-ENP and environmental interactions affect the biotic availability of ENPs and how different ENP properties (size, surface) affect exposure/uptake;
- how crucial ENP properties such as size distribution, surface chemistry, shape and optical properties influence toxicity;
- chronic aspects of ecotoxicity, which to date has mainly been assessed at environmentally unrealistic concentrations or in inconclusive studies where it was uncertain whether the co-solvent used for dispersal, impurities, or the ENP itself resulted in the observed toxic effect;
- the mechanisms of toxicity of ENPs when compared to the bulk chemical or free metal ion and how observed effects of ENPs on the expression of genes or proteins associated with particular pathways (e.g. such as oxidative stress in cell lines) relate to higher level in vivo effects;
- the fitness for purpose of existing risk assessment approaches designed for standard chemicals for use with ENPs and the modifications needed to allow existing frameworks and policies to be used in future for the risk assessments of nanotechnology products.
By studying the fate and behaviour of the selected ENPs and their effects on biota, NanoFATE has gone beyond the superficial initial assessments that have been possible so far, thereby enabling a scientifically rigorous analysis in relation to each of the above aspects. The data gained in meeting each of the nine NanoFATE objectives will allow us to go beyond the current state-of-the-art.
Project Results:
Scientific and technological challenges
The potential human health effects of ENPs are of obvious importance and a review of European research and national programs indicates that a number of ongoing projects are already addressing this issue (e.g. NANOTOX, CELLNANOTOX, IMPART, NANOSH, NanoReTox). In distinct contrast, there are as yet few studies that have focused on developing and refining methods to assess the fate of ENPs in ecosystems (e.g. soils and natural waters) and any resulting ecotoxicological effects. For this reason NanoFATE intends to focus on these neglected aspects and their integration.
To support the responsible development of the nanotechnology sector, it must be recognised that the development of environmental risk assessment methods should not lag too far behind those for human health. Past experiences highlight a number of other environmental issues such as organochlorine pesticide usage (Newton and Wyllie 1992; Newton et al. 1999; Sibly et al. 2000), endocrine disruptions (Jobling et al. 1998; Tyler et al. 1998), secondary effects of pharmaceuticals on wildlife (Oaks et al. 2004), and genetic modification (Haughton et al. 2003; Heard et al. 2003), where environmental impacts rather that direct effects on human health, emerged as the major area of concern. In each of these cases, the unexpected nature of these effects had a profound affect on public confidence in new technologies. This required that rapid regulatory action was put in place to control and mitigate risks. By ignoring effects on the environment, nanotechnology runs the risk that similar damaging and costly effects could occur.
Because of the initial and wholly understandable focus on direct risk to human health, knowledge of fundamental aspects of the environmental risks associated with ENPs is low in several key areas. These include:
- the post-production fate of ENPs from entry into the environment to final residence;
- how ENP-ENP and environmental interactions affect the biotic availability of ENPs and how different ENP properties (size, surface) affect exposure/uptake;
- how crucial ENP properties such as size distribution, surface chemistry, shape and optical properties influence toxicity;
- chronic aspects of ecotoxicity, which to date has mainly been assessed at environmentally unrealistic concentrations or in inconclusive studies where it was uncertain whether the co-solvent used for dispersal, impurities, or the ENP itself resulted in the observed toxic effect;
- the mechanisms of toxicity of ENPs when compared to the bulk chemical or free metal ion and how observed effects of ENPs on the expression of genes or proteins associated with particular pathways (e.g. such as oxidative stress in cell lines) relate to higher level in vivo effects;
- the fitness for purpose of existing risk assessment approaches designed for standard chemicals for use with ENPs and the modifications needed to allow existing frameworks and policies to be used in future for the risk assessments of nanotechnology products.
By studying the fate and behaviour of the selected ENPs and their effects on biota, NanoFATE will go beyond the superficial initial assessments that have been possible so far, thereby enabling a scientifically rigorous analysis in relation to each of the above aspects. The data gained in meeting each of the nine NanoFATE objectives will allow us to go beyond the current state-of-the-art as set out in the section below.
Scientific & Technological results
WP 1. Characterisation and tracking of ENPs during processes involved in fate and toxicity.
Objective 1.1. Manufacture, procurement and detailed characterisation of the two versions of each ENP (CeO2, ZnO and Ag) available from the industrial partners and selected as the focus for risk characterisation, fate, and effects experiments and risk assessment. This characterisation will build on manufacturers and existing data where possible and consider carefully the forms in which ENPs will be delivered (e.g. powders or suspensions, including issues of carriers and stabilisers). [UOXF.DJ NT, AXME ]
This objective has been achieved through the successful completion of D1.1 D1.2 and D1.3 ensuring also M1.1 M1.2 and M1.3. Beneficiaries contributed by ensuring the commercial ENP starting materials for the project were suitable, available and fully characterised. UOXF.DJ (P3) characterised the initially selected ENPs discovering the proposed ZnO ENP material was unsuitable for the project. This had an insufficiently controlled particle size and shape and containing both spherical, cylindrical and high aspect ratio particles, with the latter dissolving too rapidly for the proposed work. Material supplied from Microniser was subsequently validated and found suitable for use. AMEPOX (P10) initially supplied small (3-8 nm) Ag ENP with a well-controlled narrow size distribution, but stable only in organic solvent. AMEPOX developed procedures for preparing and dispersing the same ENPs directly in water for use in WP 2-5 and undertook tests for this water-based colloid to ensure adequate stability with time. NT (P6) supplied Ag ENP powder (50 nm). NT collaborated with AMEPOX (P10) developing a way of dispersing and stabilising their material in water for use in WP 2-5. UOXF.DJ (P3) undertook quality control checks on all material supplied for WP 2-5, as well as periodic checks on stock material, using a multi technique approach. This demonstrated the importance of TEM imaging of all samples.
Objective 1.2. Production of ZnO ENPs tagged with rare earth elements in very low natural concentrations (e.g. Ytterbium or Luthetium) that match as closely as possible the ZnO nanoparticles common in commercial products. This is needed to allow tracking of the ZnO ENPs by ICP-based element analysis in the experiments to measure fate and effects. (For Ce and Ag based ENPs the natural background is so low they can be tracked by their core elements, while Zn has too high natural background for this to work if not tagged). An alternative approach, although significantly more expensive, is to synthesize the ZnO nanoparticles from a stable enriched isotope, e.g. 67Zn or 70Zn. [IHPP]
This objective has been achieved through the successful completion of D1.4 ensuring also M1.4. IHPP (P8) developed methods of producing doped ZnO ENPs initially using Al as proof of method. Subsequently they refined this method to produce material with Gd and Lu dopants, however because of the mismatch in atomic scale of rare earth elements compared with zinc it was not possible to incorporate rare earth dopant atoms into the ZnO lattice. UOXF.DJ (P3) demonstrated with TEM that the dopants consisted of discrete material on the surface of the ZnO ENPs in the form of oxide. Because of the similarity in atomic scale the consortium decided to try Co doped ZnO ENPs and IHPP (P8) successfully manufactured such material. The natural Co background level in soil meant a high dopant concentration was required and a 10 wt% dopant material was produced for the tracking experiments performed in WP 2-4. IHPP (P8) and UOXF.DJ (P3) characterised this material for particle size distribution (30 – 60 nm) and lattice parameter proving it was comparable to the undoped ZnO.
Objective 1.3. Characterisation of the ENPs to be used initially in the standard exposure medias of WP3. For ZnO this will be done for both the tagged and untagged for ZnO to check how the behaviour of tagged particles match that of untagged particles. (IHPP, UOXF.DJ UGOT]
This objective has been achieved through the successful completion of D1.5 D 1.6 and D1.7. UOXF.DJ (P3) and UGOT (P11) carried out experiments using AMEPOX Ag ENPs dispersed in water and also the exposure media employed in the ecotoxicological tests in WP3. This investigation enabled a better understanding of the form of the ENPs as presented to the organisms being dosed and the dynamics and rates of change in these ENP forms during exposures. IHPP (P8) similarly provided this information on the behaviour of ZnO and Co doped ZnO in ecotoxicological exposure media.
Objective 1.4. Validation of element based analysis for gross tracking of ENPs. Method development and validation for the detailed tracking and localisation of ENPs in exposure and tissues relevant in WP2, WP3 and WP4 to refine methods for detecting particle in environmental media and in cells and tissue of exposed organisms. [IHPP, UOXF.DJ UGOT]
This objective has been achieved through the successful completion of D1.8 ensuring also M1.5. IHPP (P8) supplied 10 wt% Co doped ZnO ENPs characterised by IHPP (P8), UOXF.DJ (P3) and UGOT (P11) with instructions for use in fate experiments by NERC (P1) Lumbricus rubellus (1000 and 4000 µg Zn/g ENPs dosed in soil and food and compared with ionic dosing), VUA (P2) Folsomia candida (400 and 1000 µ g Zn/g ENPs dosed in soil) and UAVR (P4) Daphnia magna (100 µg Zn/L dosed). We have demonstrated that the Co doped ZnO nanoparticles manufactured by IHPP are fit for the purpose they were designed for and can be used in tracking experiments, if the dose is sufficiently high, to determine whether the nanoparticles are taken up by organisms as intact particles or as ions. Successful detection methods in these trials were digestion and bulk metal analysis such as AAS, OES, and ICP-MS. High resolution X-ray Fluorescence imaging of sections of Lumbricus rubellus showed concentrations of Zn, but the Co was below the limit of detection for imaging. From the Lumbricus rubellus experiments undertaken by NERC (P1) we now have evidence suggesting that metals in particle form may be uptaken as intact particles or at least through different routes than metal ions. Co is taken up via the gut, but not the skin when presented as Co doped ZnO nanoparticles, while in contrast Co ions when mixed with Zn ions are not taken up by the gut, although there possibly is minor ion uptake via the skin.
Objective 1.5. Identification, quantification and characterisation of ENPs in samples provided from WPs 2, 3 and 4 to provide: (i) qualitative and quantitative data on environment-ENP interactions, and (ii) the association of ENPs with both matrix and organism components in different environmental systems. [IHPP, UOXF.DJ UGOT, NERC, VUA, CU, UAVR]
This objective has been achieved by all beneficiaries working together to produce a protocol for the detection of ENPs in complex and dilute organic systems (media and organisms). Samples of dosed media and organisms were produced from ecotox experiments performed by UGOT, NERC, VUA and UAVR (WP2-4) along with bulk chemical analysis data. In samples where elemental or other simple high throughput analysis indicated presence of the ENP material, CU (P9) performed a series of specialised imaging experiments with differing spatial resolution and limits of detection e.g. CARS and Synchrotron Micro focused X-ray imaging (XRF and CAS) and applying expert sample preparation techniques for subsequent HRTEM performed by UOXF.DJ on areas of high concentration of ENP material as determined by the other more rapid, but less detailed imaging methods. This multi method, tiered approach to detection of ENPs in environmental samples has been submitted for publication. Most of this work has involved the use of AMEPOX Ag ENPs and has been reported in WP 2-4. We have HRTEM evidence for indistinguishable particle formation in organisms after dosing with both Ag ENP and AgNO3 suggesting dissolution of Ag ENP to an ionic form and re-precipitation though it has not yet been determined at what point in this occurs.
WP 2. ENP environmental behaviour and fate modelling
Objective 2.1: Identification of critical parameters governing the fate of ENPs [NERC, F+B, UGOT].
Initially efforts focused on the estimations of product usage volumes and worst case release amounts. For ZnO and Ag, estimates of production volumes were therefore taken from literature and calculated into per capita down the drain usage and environmental entry through waste water. For CeO2 fuel additive worst case usage calculations were based on all diesel cars in the UK using CeO2 additives and all deposition restricted to within 20m of UK A and M road network only (Johnson, 2012). An important assumption nano CeO2 emissions is the effectiveness, or otherwise, of fine particulate filters in exhaust systems. Although commercial (PSA Citroen and Cummins Diesel, plus filter manufactures Johnson Matthey) and US EPA references to diesel particulate filter exist quoting performances ranging from 85% to 99% efficiency, only one published study could be found, but also giving 95-99% removal (Constantini, 2011). A concern here is that the perfoamances may refere to particle sizes higher than those potentially carrying nano CeO2 through the exhaust system. Recent data from studies carried out using a test engine (Zhang et. al, 2103) that do not incorporate any exhaust filters indicates that only 40% of the cerium added to the fuel is emitted, with the remainder lost in the pipes, unburned residues and tars (pers. Comms. Zhang). However, to maintain worst case estimates to give the most conservative of potential risk predictions, the NanoFATE fate modelling include no removal in filters. This still indicated a difference in the order of 3-5 orders of magnitude between worst case soil exposure and effect levels, and hence any further precision in the modelling was not considered critical.
The big focus for NanoFATE was the identification and prioritization of specific properties of the environment and materials that need principal consideration during the development, adaptation and validation of environmental fate models for ENPs were. Ionic strength and composition, natural organic matter quantity and pH were concluded to be the dominant properties of natural waters, although not solely important. It was concluded that the core composition, and surface functionalization were the two most critical material factors, although many others could play a role in the fate processes. In terrestrial systems, the soil solids composition and grain size will be important in addition to the parameters listed above for water.
The importance of these parameters in the initial fate models developed (with WP6) to predict nano ZnO and Ag discharge (effluent and sludge) from sewage treatment works and further environmental fate have been assessed and increasingly better parameterised through the course of the NanoFATE project. This has been delivered through the WP2 experimental fate testing framework for fate of ENPs in natural waters. This has developed the first preliminary data gathered for such models to be developed further. Kinetic models building on colloid theory are especially promising and have been developed to deliver fate predictions at the EU and regional scale (see Objective 6.5 in the work package 6 section of this report).
insert Cerium removal here:-
Objective 2.2: Development and application of an experimental framework for fate testing of ENP in natural waters
The full objective was to develop experimental methods to study ENP-environment interactions and their consequences for ENPs fate and behaviour through wastewater, sewage treatment processes, and in receiving natural waters and sludge [UGOT, NERC, UOXF.DJ IHPP].
In order to mimic the natural environment in the laboratory or in field microcosms decisions about complexity and relevance versus simplicity and mechanistic understanding have to be taken. Studies of real environmental matrices are highly relevant, but it is very difficult to probe the nanomaterials and how they interact with living and dead natural matter in the full complexity. Furthermore, real natural environmental matrices are time and space specific so the capability to extrapolate to wider areas is very limited. On the other hand, building up knowledge on environmental relevance by simple electrolyte solutions of increasing relevance will never lead to applicable knowledge on the risk for the environment. Therefore in NanoFATE we selected a framework that built on model natural waters which was standardized and slightly more simplified than real natural waters. The model waters were based on a multivariate classification of all Europes surface waters. In addition to electrolytes, also model natural organic matter and model inorganic colloids was developed and implemented in these 6 “standard European model waters”. The experimental platforms for fate studies was built around the model waters using the method time-resolved dynamic light scattering (DLS) for homoagglomeration studies and a custom designed settling cone for studies of homo- plus heteroagglomeration and their combined influence on sedimentation. Suitable analytical techniques had to be developed to analyze the particles in the settling cone test platform.
In order to conduct a proof of principle study that investigated agglomeration-sedimentation processes in isolation from particle dissolution meant using gold particles as a substitute for the Nanofate commercial silver particles. These gold particles have many similarities with silver, but do not oxidize and dissolve. The main conclusions on the fate of the tested ENP under the realistic concentrations was that they did not settle out as a result on agglomeration during the two months study under most of the model waters, but in waters with high ionic strength the removal was significant and directly dependant on the ionic strength. The homo and heteroagglomeration also depend on if it was the gold nanoparticle or the model inorganic colloid (illite) that was destabilized by the specific model natural water.
Heteroagglomeration which has recently been claimed to be a dominant driving fate process in natural waters, did not under the tested conditions become such an efficient removal process.
Objectives 2.3 and 2.5 Apply refined experimental approach to investigate the effects of particle properties on ENP fate and behavior through waste water treatment processes and associated effluents and interatively incorporate avaoialable data on ENP-environment interactions to refine predicted environmental concentrations for waters and soils [UGOT, NERC, UOXF.DJ IHPP]
In order to estimate the removal rate in waste water treatment plants before the single particle ICPMS method had been fully developed, a study of ultrafiltration which captured the nanoparticulate and colloidal silver was used in influent and effluent samples from a range of different types of waste water treatment plants (WWTP). The average silver removal rate was found to be 49%. In addition, the most useful reference found for ZnO was that of Lombi et al. (2012) who found the partition coefficients (Kd) for the partitioning between supernatants and sewage sludge of three different ZnO NPs to be 1258, 1667, and 915. The average Kd that they found for nano ZnO was 1280. If we assume that most sorption of nano ZnO in WWTPs takes place in activated sludge tanks with a sludge concentration of about 0.0035 kg/L than that would mean that the ratio of sorbed over non-sorbed nano ZnO would be (1280*0.0035=) 4.48. Thus, the WWTP removal of nano ZnO would be about (100*4.48/(4.48+1)=) 81.8%. For nano Ag, assuming a per capita consumption of 0.055 mg/cap/d and the 49% removal discussed above, the per capita discharge would be 0.028 mg/cap/d. An average per capita discharge of 160 L/d would result in a potential effluent concentration of 175 ng/L. For nano ZnO a larger consumption rate of 0.55 mg/cap/d and a sewage removal of 82%, would give a per capita discharge of 0.099 mg/cap/d and a potential effluent concentration of 618 ng/L. It should be noted these values are likely to be a considerable over-estimates for the following reasons: (i) a proportion of ENP production in a continent may be exported, and (ii) many ENP products are unlikely to release their ENPs to water.
Objective 2.4: Track effects of particle properties and environmental physicochemical properties on the rates of dissolution of ENPs during wastewater treatment and in a range of natural soils and water. [UGOT, UOXF.DJ IHPP].
This objective was principally achieved through the development and application of advanced spectroscopic tools and microcosms to study the fate of silver nanoparticles in waste water treatment processes
Small angle Neutron Scattering (SANS) was evaluated to probe settling of commercial silver nanoparticles under real sewage conditions compared to pure water. The method is promising, but requires access to synchrotron beam time.
Single particle ICPMS (spICPMS) is a method capable of probing low concentration of metallic nanoparticles in complex samples such as natural waters of sewage. Although a very useful method it has limitations in terms of the smallest detectable size that can be discriminated from the background of dissolved and colloidal ions of the same element. Therefore we have improved the method with two different approaches. Firstly, a statistical method to describe the combined nanoparticulate and background signals in spICPMS was developed, which allowed us to deconvolute the nanoparticle contribution from the combined signal (Cornelis and Hassellöv, 2013). This allowed a significant improvement in smallest detectable size. Secondly, a totally new method of data acquisition, using true realtime signal collection (dwell time 0.1ms compared to 10 ms) allowed us to follow the whole nanoparticle signal peak instead of just discriminating outlier spikes in the signal (Tuoriniemi et al 2014). Both these method improvements could be combined to allow detection of nanoparticles down to the physical limits (around 5-7 nm for silver and 4-5 nm for gold). This method was used to track the silver nanoparticle fate transformations during the WWTP microcosm study (Cornelis et al., 2014).
WP 3. ENP Ecotoxicology
Objective 3.1: Collate information on the suitability of existing ecotoxicological tests systems for ENP testing and subsequent optimisation of the NanoFATE biotest systems. [NERC, VUA, UOXF.DJ UAVR, UNIPMN, CU, AXME, UGOT]
This objective was achieved through the successful completion of D3.1. The contributions of individual beneficiaries were carried out in dialogue to ensure protocols were optimised with regard to both the aspects of presenting the ENP to the organisms in the most realistic form possible and that organism conditions were as close to optimal as possible. Data on the toxicity of ENPs to different test organisms were collected from the literature to assess the suitability of currently available test guidelines for assessing ENP toxicity to the organisms selected for the work in NanoFATE. An on-line database was developed to summarize and critically reviewing literature toxicity data in collaboration with WP4, serving for all partners and WPs including as the main hazard data source used in WP6 for risk assessment
Partners from AMEPOX, IHPP, UOXF.DJ and UGOT worked together to provide well characterised nanoparticles suitable for performing ecotoxicity tests. Both UGOT and IHPP provided essential information regarding agglomeration behaviour of Ag and ZnO ENPs in the different aquatic media so that the most suitable could be identified and determine if any adaptations of the standard protocols or media were needed, and explaining some of the observed cases of toxicity differences.
NERC contributed with improvements to the nematode test media and, in collaboration with CU to the dosing protocols needed for delivering the ENPs into the earthworm test soils. The standard OECD and ISO protocols were adapted to provide stable and reproducible exposures for the experiments with the ENPs to be tested in the project. For both organisms, the modifications were chiefly based around ensuring optimisation of the realism of the presentation and exposure form of the ENPs through modifying exposure media and dosing protocols. VUA optimised exposures with the soil arthropod Folsomia candida (Collembola), using the standard test guideline prescribed by ISO (1999) as a basis. Rather than using an artificial soil, a natural standard soil Lufa 2.2 was recommended, since this is better suited to some type of analysis for particle detection in the soil media (in particular the high kaolinite content of the OECD soil proved a difficulty). Tests were run with ZnO ENP, including non-nano ZnO and ZnCl2 for comparison. UAVR optimised a protocol for the terrestrial isopod Porcellionides pruinosus, including optimisation of dosing to provide both soil and food exposure routes for ZnO and Ag nanoparticles. Comparisons of different methods of spiking the soil with ENPs demonstrated that the best method of spiking soil with nanoparticles will depend on the ENPs properties and delivery form.
For the aquatic compartment, UAVR checked the standardized procedures for testing chemicals with the cladoceran Daphnia magna for its suitability to conduct the toxicological assessment with ENPs. To test the effects of ENP agglomeration on toxicity, results based on changing media after 24 and 48 hrs were compared. Additionally for chronic testing, one issue that emerged was the use of supplementary seaweed extract usually used as a source of micronutrients, especially selenium, which could act as a chelating substance that might decrease ENP bioavailability in the test media. Test carried out without adding seaweed extract were conducted, but these resulted in high mortality of juveniles. As a compromise, it is suggested that the test media is maintained close to the standard composition, but should be regularly renewed during chronic tests to maintain worst case ENP presentation for the hazard assessment. Snails (Physa acuta) were also tested and protocols adapted mainly on water composition, where two media types were used: Artificial Pond Water (APW) and APW-adapted without CaCl2. This approach was also valuable for WP4 in terms of relating silver bioavailability with toxicity. UNIPMN developed modifications on the protocol for acute toxicity evaluation of ENPs in marine mussels using artificial sea water as test medium.
UGOT evaluated the protocol to test the effects of ENPs on growth and reproduction of the green algae Chlamydomonas reinhardtii and Pseudokirchneriella subcapitata, the cyanobacterium Synechoccocus leopoldensis and the diatom Cyclotella meninghiana. Test procedures followed the corresponding OECD guideline 201, and adaptations were carried out to improve methodology for testing the ecotoxicity of ENPs. The Woods Hole MBL media showed to be one of the best media to provide more accurate toxicity endpoints.
Objective 3.2. Ecotoxicological hazard characterisation of pure ENPs using the optimised media to establish chronic thresholds of concern (no observed effect concentrations, benchmark concentrations) for population relevant endpoints (e.g. reproduction) for selected aquatic and terrestrial test species. [NERC, VUA, UOXF.DJ UAVR, UNIPMN, CU, AXME, UGOT]
The work conducted in NanoFATE represents one of the largest bodies of bespoke cross species ecotoxicity testing of the chronic effects of commercial ENPs on environmental receptors. The data will, thus, be an important contribution to the risk assessment literature.
Using the artificial test media developed by NERC the toxicity of ionic Zn and Ag and multiple NP forms of Ag, ZnO and CeO2 were assessed in Caenorhabditis elegans. In collaboration with CU, exposures to CeO2 ENPs in nematodes and earthworms were undertaken, as well as exposures of the earthworm Eisenia andrei to dissolved Zn (as ZnCl2), NanoSun ZnO ENP (30nm), the two Z-cote ZnO ENPs (Z-cote and Z-cote HP1), and a non-nano ZnO (Sigma non-nano). Tissues from exposures were banked for M3.2 and use in WP5.
In addition, NERC developed a high throughput screening method for assessing the toxicity of ENPs to multiple bacterial species.VUA determined the toxicity of ZnO (both coated and uncoated ENP and non-nano forms) to the collembolan Folsomia candida in Lufa 2.2 natural soil, focusing on effects on survival and reproduction after 28 days of exposure. The toxicity of two different Ag ENPs to Folsomia candida in Lufa 2.2 natural soil was also tested. The Ag ENPs were obtained from AMEPOX and included the 3-8 nm ENPs dissolved either in hexane or in water.
Methodological improvements were developed and implemented to enable quantification of Ag in the test soils. Previously used digestion methods had to be modified to increase recovery of Ag from the test soil. This resulted in good recoveries in case of soil spiked with AgNO3 or Ag ENPs in water. This was also crucial to determine accurate body burdens (WP5), focusing on Ag uptake kinetics and possible Ag ENP uptake.
Chronic ENP toxicity to Porcellionides pruinosus was evaluated using Ag ENPs and ZnO ENPs by UAVR through testing avoidance behaviour and feeding inhibition. In addition, samples from the hepatopancreas and gut tissues of isopods exposed to Ag ENPs (8nm) via food were collected and supplied for use in WP5. In this approach, the avoidance behaviour test showed to be a very powerful tool, detecting 50% of avoidance behaviour in organisms at concentration of 18mg Ag ENP/Kg soil. VUA and UAVR showed that CeO2 ENPs presented no adverse effects on springtails and isopods at 1000 mg CeO2 ENP/Kg soil. Earthworm exposures for CeO2 ENPs were undertaken also showing low toxicity at concentrations up to 10000 mg CeO2 ENP/Kg soil.
UGOT studied the effects of silver and zinc nanoparticles in single species assays with selected species of algae and bacteria and a community-level bioassay with natural biofilms (so-called periphyton). An extended particle range outside the NanoFATE commercial ENPs were investigated in order to cover differing in coating, size and presence of stabilizers. Specific information on the physiological status (vitality) of the exposed organisms were determined by using fluorescence dyes and flow cytometry. Additionally, the formation of reactive oxygen species (ROS) as a consequence of exposure to nanoparticles was investigated. Using the MOPS buffer increased the sensitivity of the assay, most likely due to increased concentrations of free silver ions. Natural microbial communities (algae and bacteria) were also assessed and the effects on the bacterial part of the community were determined using Ecolog plates, which provides an indication of the ability of the community to degrade a set of 31 different carbon sources and hence data on the ability of the communities to drive carbon cycling under toxicant stress. The carbon utilization pattern also gives an indication of changes in species composition (biodiversity, toxicant-induced succession).
Effects on changes in the biofilm algae community were studied by HPLC-based pigment profiling. Algal growth was determined as chlorophyll α content, species composition and interferences with specific physiological processes were estimated as changes in pigment patterns.
In addition, the following particles were tested by UGOT in the growth inhibition assays with the gram-negative bacterium Pseudomonas putida, performed according to ISO guideline 10712: AG7 (Amepox, Poland), PL-Ag-S10 (Plasmachem AG, Germany), NM-300K (OECD WPMN program, JRC, Ispra, Italy), PELCO® NanoXact (Tedpella, US), Silver colloid (British biocell, UK), AG6 (Nanotrade, Czech Republic). Silver nitrate was used as a reference compound. For each particle type a full concentration-response curve was recorded in at least two independent experiments. Particle size distribution and number were determined in all stock suspensions, using Nanosight nanoparticle tracking analysis (NTA) and transmission electron microscopy (TEM).
Based on protocol adaptations, UAVR conducted chronic assays with Daphnia magna, with the Ag ENP from AMEPOX (3-8 nm), and AgNO3 as ionic comparator, and also compared results from ZnO particles’ exposures (30nm, 80-100nm and 200nm, Nanosun) and ZnCl2. UNIPMN carried out a comparative assessment of the AG7 and AG6 silver ENPs (8 nm, 50 nm size, respectively) and silver nitrate in a range of concentrations from 0.001 to 10 mg/L to the marine mussel Mytilus galloprovinciali. In addition, two sizes of the Nanosun ZnO ENPs (30nm and 80-100nm), ZnO 200nm and ZnCl2 were also tested. Acute (i.e. mortality during post-exposure removal from water to air), and sub-lethal effects on byssus synthesis were evaluated in M. galloprovinciallis as a function of time and concentration within four days exposure.
Objective 3.3. Determination of the ecotoxicology of the selected ENP types in environmentally realistic situations for aquatic and terrestrial systems, including the assessment of interactive effects with physiochemical components of different exposure media and interactions with UV and organic chemicals as potential co-stressors. [NERC, VUA, UOXF.DJ UAVR, UNIPMN, IHPP, CU, UGOT]
The information to derive confounding factors (soil and water chemistry) on the ecotoxicology of ENPs has been gathered from studies on ZnO, Ag and CeO2 ENPs. Within this framework, IHPP studied the interactions of the developed Co doped ZnO ENPs to enable easy detection, and stability, in different test media including the ASTM medium, deionized water and Lufa 2.2 soil extract.
CU, NERC, VU and UAVR carried out large scale tests to study how soil pH affects ZnO ENP toxicity using modified field soils for earthworms, collembola and isopods. Samples were analyzed and characterized for ENP dissolution and behaviour kinetics, and tissue samples from earthworms were prepared and preserved for molecular analysis and ultra histology. The results of all these trials were used to parameterise models developed in WP4 modelling and provide input for WP5.
The toxicity of ZnO nanoparticles to F. candida and P. pruinosus in soils ranging in organic matter content was also evaluated by VUA and UAVR. Four soils ranging in organic matter content from low (1.64 %) to high (16.7%) were used to study the effect on springtail survival and reproduction after 28-d exposure and isopods feeding inhibition and biomass loss in a 14-d exposure. VUA and UAVR carried out chronic toxicity tests with collembolan and isopods to understand the effects of CeO2 ENPs on the toxicity of phenenthrene through possible co-transport.
UNIPMN had undertaken studies on the effects of realistic ENP forms on Mytilus spp in natural waters. Microcosm long term exposures were carried out, where mussels were exposed for 4 weeks in natural seawater to AG7 ENPs (3-8 nm, AMEPOX).
Finally UAVR conducted several assays with D. magna were effects from mixture exposures were analysed. For that combinations of Ag forms (Ag ENP x AgNO3), Zn forms (ZnO ENP x ZnCl2) and also Ag x Zn (both ENPs or ionic) forms were carried out using immobilization and feeding inhibition tests. Conceptual models for mixture toxicity were used to derive additivity patterns or interactions between forms inside the organisms.
Objective 3.4. Identification of the major drivers for the ecotoxicological effects of ENPs, such as the contribution of the ENP itself and the metal ions they liberate, the coating of the ENPs after passage through WWTPs, the dependence on the exposed species and considered exposure route. Compilation of the experimental information and existing information from the literature and other R&D projects. Knowledge transfer to WP 4 (bioavailability studies) and WP6 (risk assessment). [NERC, VUA, UOXF.DJ UAVR, UNIPMN, CU, UGOT]
From the ecotoxicity testing carried out during the project, several issues were identified as drivers for effects induced by ENPs, regarding both aquatic or soil systems. The behaviour of nanoparticles govern their toxicity and was shown to change greatly with time at scales often much shorter or longer than standard exposures. In time, ENPs change their characteristics, which have to be described in order to understand their deleterious effects, understanding simultaneously if effects are nanoparticle-related or not.
One example was the study on the exposure of D. magna to different sized ZnO ENPs (Nanotrade). Their initial size was confirmed to have no effect on ENPs toxicity mainly because after 48h they ended up all having agglomerated to very similar size and no longer in the nano-range (Lopes et al. 2014). Therefore the first hypothesis raised that initial size could influence toxicity showed not to be easy to test nor entirely relevant in realistic exposure terms, but their size upon initial agglomeration in the exposure cannot be excluded from discussion of important parameters. In addition, ZnO particle size showed no influence on soil sorption, and little differences were observed in terms of toxicity when F. candida was exposed to ZnO ENPs of different sizes (30nm and 200 nm ZnO, Nanotrade) (Waalewijn-Kool et al. 2013).
Dissolution rate proved a characteristic that has to be taken into consideration when looking at ENPs toxicity, again making exposure time a key factor. With time, some metal ENPs will release ions and those ions will also be more or less responsible for their toxicity. This was observed in the toxicity tests with D. Magna, but also with algae; upon Ag ENPs (AMEPOX) exposures, algae toxicity was mainly due to the internalization of dissolved or ionic Ag (Ribeiro et al. 2014).
The most sensitive organisms tested were the water microbial communities. Results from the standard tests with algae and bacteria revealed that the prokaryotic organisms reacted more sensitive than the eukaryotic. The bacterium P. putida reacted most sensitive with EC50 values of 0.16 µg/L for AgNO3, resp. between 0.25 and 13.5 μg/L for AgNPs. However, no particle specific effects could be detected indicating that the silver ions are the toxicity causing factor; neither specific coatings nor certain sizes could be linked to increasing or decreasing toxicity for all tested particles and organisms and no clear toxicity patterns could be detected in terms of species specific attributes. The characterization of particle behavior (and for a certain proportion for the total and dissolved silver content) in the test media during the exposures proved to be impossible for most exposure situations due to the high sensitivities of the test organisms.
For soils, the dissolution process and the way ENPs interact with soil pore water was found to be of major relevance particularly regarding collembolan and earthworms. This was mainly justified by their exposure routes, where collembolans are exposed through the soil pore water and earthworms can be exposed both dermally and through soil ingestion. In exposures of earthworms to Zn (using ZnO ENPs, ZnCl2 and micron scale ZnO) in soils of varying pH, it was observed that pH highly influenced Zn dissolution (Heggelund et al., 2013). Indeed, the WP3 experiments has shown pH to play a major role in bioavailability, as for ENP exposures the competition between H+ and metallic ions is not as straightforward as it was for the well-studied case of ionic metal exposures (see WP4 below for details on bioavailability considerations). This information was crucial to determine the relevant bioavailable fraction and which were the factors affecting bioavailability (WP4), but also providing relevant input data needed for risk assessment procedures (WP6).
WP4. ENP bioavailability – relations between soil and water chemistry and particle properties.
Objective 4.1. Review and utilise existing data-holdings and where necessary primary research articles that contain information on ENP, and media properties and their relationship with acute and chronic toxicity in aquatic and terrestrial species. [NERC, VUA, UAVR, UNIPMN, CU, UGOT]
This objective has been achieved through the collation, data-basing and mining of existing and newly available data. WP4 partners (NERC, VUA, UAVR, UNIPMN, CU, UGOT) jointly developed and contributed to creating a database, bringing together relevant literature data. IHPP and UOXF.DJ provided relevant information on ENP properties that assisted in properly interpreting the available literature data. The database, supported the work within WP3 and WP4, and served as the basis for collating the toxicity data generated from NanoFATE and the partners other nanosafety projects and for constructing species-sensitivity distributions for use within spatial risk assessment models developed in WP6.
Results from this exercise also provided basic input regarding the evaluation of confounding factors to be included in assessing the bioavailability and toxicity of ENPs to the different test organisms in soil and water. Organic matter content and pH were identified as major factors affecting ENP fate and bioavailability in both soil and water. For aquatic media, also ionic strength and salinity are important factors. These findings lead to the prioritisation and planning of experimental work under a realistic range of environmental conditions and modified media conditions within WP3 and WP4, respectively.
The critical review of the literature performed to identify factors affecting ENP bioavailability was further elaborated by UAVR, VUA and NERC into a paper focusing on soil. A similar manuscript dealing with aqueous exposure media is under development by NERC.
Objective 4.2. Establish and model how environment physicochemical properties govern ENP stability, soil–solution partitioning and transformation (e.g. dissolution) in natural waters and soils utilising time dependent kinetic modelling; chemical speciation modelling using established tools such as the Windermere Humic Acid Model and adapted version of the Biotic Ligand Model that include both particle dependent and dissolved metal ion toxicity. [NERC, VUA, UAVR, CU, UGOT]
This objective (and objectives 4.3 & 4.4) was achieved through completion of detailed experiments on how ENP and media properties interact to affect ENP presentation in environmental media and their availability for uptake and integrated assessment and modelling of ENP uptake and toxicity across soils and waters of varying chemical properties.
In close collaboration with the other WP4 partners and WP1, IHPP developed cobalt-tagged ZnO ENPs to enable proper assessment of ENP fate in test systems as well as possible uptake in test organisms. IHPP also determined the stability of the cobalt-tagged ZnO ENPs and other ZnO ENP and non-nano forms in different aquatic test media used for both aquatic and soil organisms. The results supported the experimental work preformed to fulfil Task 4.2 (Detailed experiments on how ENP and media properties interact in time to affect ENP presentation in environmental media and their availability for uptake), and have been reported in Deliverable 4.2.
NERC assessed and modelled the interaction of ENPs with media properties in aquatic and soil test media. In tests using the NanoFATE developed simulated soil pore water, NERC assessed how the changes in the concentrations of DOC (fulvic acids) and major cations (H+, Ca2+, Na+, Mg2+) affected ENP availability. These data have indicated the extent to which competition effects need to be accounted for during the development of biotic ligand models for ionic silver and silver ENPs. In terrestrial tests, NERC, VUA and UAVR studied the effect of soil pH on the comparative uptake and toxicity of ionic Zn, ZnO ENP and non-nano sized ZnO by earthworms, collembolans and isopods. These experiments indicated that ZnO dissolution was faster at low soil pH values, but for the same total soil concentration the soil porewater concentrations of Zn was much lower in ZnO dosed soils that in ZnCl2 dosed ones. For all species the pore water based toxicity was higher for ZnO then ZnCl2 based exposures, suggestion that not the full effect could be explained by the dissolved Zn and that the particulate ZnO contributed to toxicity. The normal “protective effects from protons” that reduces toxicity of porewater Zn with decreasing pH for ionic zinc was much less evident for the porewater in the ENP based exposures. For earthworms more zinc was accumulated in worms exposed to ZnO ENPs than those exposed to ionic zinc at the same effect levels, suggestion that uptake routes and internal handling differ between the two Zn forms. .
To assess the timescale of ENP stability dynamics and its consequences for ENP toxicity in soil, VUA performed ageing studies to determine long-term fate of triethoxyoctylsilane-coated and non-coated ZnO ENPs, non-nano ZnO and ionic Zn in soil and possible changes in toxicity to Collembola. Coated ZnO ENPs were more toxic than non-coated ones, and only after aging the soils for one year of equilibration was the toxicity of coated ZnO ENPs reduced to the initial (time zero) level of non-coated particles, suggesting the degradation of the coating has a 1 year timescale in soils. Both ZnO ENPs and non-nano ZnO slowly released Zn ions, with Zn ion release being highest at intermediate exposure levels and reduced at high exposure levels. This probably is due to stronger agglomeration at increased soil concentrations, which was also pH dependent. In spite of increasing Zn concentrations in pore water, ageing reduced the toxicity of ZnO ENPs, but also of non-nano ZnO and of ZnCl2, which could be explained from different effects on soil pH resulting from the addition of the different Zn forms. This study showed that soil pH is a major factor affecting ZnO ENP fate, availability and toxicity to the test organisms, and that a coating although reducing dissolution of ZnO ENPs may leave them more toxic. A paper describing this work was included in Deliverable 4.3. A similar longterm soils aging study was undertaken for Ag ENP toxicity to earthworms at NERC, showing that again the dynamics of the ENP behaviours are at the timescales of a year. Here ENP toxicity based on total soils Ag was initially much lower than for the ionic based exposure. However, while the toxicity of the ionic exposure reduces over time as the soils were aged, the ENP toxicity increased (presumably as the ENPs dissolved) and after a year the ENP toxicity had reached the level of toxicity seen for the freshly spiked ionic exposures at the start of the aging experiment. Again suggesting that ENP dynamics in soils are slow and that results from short standard tests (even 56day earthworm tests) may give a slightly false sense that ENP forms of metals are far less toxic than the ionic forms, when long term they may become as toxic.
Objective 4.3. Undertake specifically designed studies to investigate interactive effects among factors (ENP and media characteristics) that influence availability and uptake of ENPs in key species. [NERC, VUA, UAVR, UNIPMN, CU, UGOT]
In close collaboration, NERC, UAVR and VUA performed a study on the toxicity and bioaccumulation of ZnO ENPs in relation to soil properties. The design properties and media modification parameters were developed under WP4 based on the critical review of the current literature and the exposures run under WP3. To achieve the full range of natural soil pH (4-8) in a soil of otherwise closely matched properties NERC collected an acidic field soil (pH 4) and adjusted it to different pH levels. The zeta potential of ZnO ENPs in soil extracts related to pH in opposite direction to that seen in deionized water, showing that fate of ENPs in soil cannot be predicted from that in pure water. All three soils have been tested for toxicity of ZnO ENPs using earthworms (CU & NERC), Collembola (VUA) and isopods (UAVR). In addition, VUA also determined the effects of other soil properties on ZnO ENP toxicity to Collembola. Toxicity of ZnO ENPs was always higher at lower soil pH, but at higher pH levels results were not consistent across species: for earthworms and springtails toxicity decreased with increasing pH while for isopods toxicity was lowest at intermediate pH levels. Similarly NERC included other natural soils that covered the normal range in soil organic matter too, but within normal soil ranges the effect of pH is much stronger than that of organic matter content. Results of these studies have been detailed in four published research manuscripts resulting from this work (Tourinho et al., 2012; Tourinho et al., 2013; Waalewijn-Kool et al., 2013a; Waalewijn-Kool et al., 2013b; Heggelund et al., 2014).
For the aquatic compartment, tests on the toxicity of Ag ENPs to Daphnia magna were carried out in ASTM media with different pH values (5.3 7.8 9.0). In addition, the effect of hardness on ENP toxicity was determined. This study confirmed that also for the aquatic environment, pH is an important factor determining toxicity of Ag ENPs.
Using the snail Physa acuta, UAVR studied Ag ENP toxicity in test media with different composition. Two different media were used, taking into account the knowledge that silver ions bind to chlorides reducing the toxicity of silver: Artificial Pond Water (APW) and APW-adapted without CaCl2. This study showed the importance of complexation between Ag and Cl, which leads to the formation of AgCl colloids and lowers toxicity of silver.
UNIPMN determined the toxicity to marine organisms (molluscs) of different Ag ENPs as a function of their stability in water and solubilisation. Results suggest that the toxicity of Ag ENPs is due not only to the release of silver ions, but also due to the behaviour of ENPs in the media, which can vary widely with concentration and media properties. Results are reported in a manuscript that is included in Deliverable 4.4.
Objective 4.4: Develop kinetic models that describe environmental effects on particle dissolution rates and incorporate both free ion and particle associated toxicity to address the relationship between the relative contribution of ENP- and free metal ion- derived toxicity for exposed species [NERC, VUA, UOXF.DJ UAVR, UNIPMN, IHPP, CU, UGOT]
NERC integrated results from the literature and those obtained within NanoFATE to explore through speciation modelling the relative contributions of ENP and free metal ions to toxicity for the exposed species in aquatic media. NERC also provided other partners (VUA and UAVR) with calculations supporting assessments of the relative contribution of free metal ions to the toxicity of ZnO ENPs in different soils. Similarly, NERC supported UAVR and UNIPMN with speciation calculations to interpret ENP toxicity in aquatic media. These activities support the experimental work performed to fulfil objectives 4.2 & 4.3.
Deliverable 4.4 was finalized to include research manuscripts on the effect of soil pH on the toxicity of ZnO ENPs to earthworms (NERC), Collembola (VUA) and isopods (UAVR). In addition, the deliverable for this work includes submitted manuscripts on the effect of soil properties on ZnO ENP toxicity to Collembola (VUA) and the effects of media properties on the toxicity of Ag ENPs to molluscs (UNIMPN).
During a workshop held in Wallingford on 26th September 2013, an inventory was made of all data on the influence of media properties on ENP bioavailability generated within NanoFATE. The workshop identified some data gaps and resulted in agreements on a limited number of further measurements to obtain all data needed for the development of such modelling framework. Together with the outcome of the workshop and the results of the additional measurements, the work described in Deliverable 4.4 provides the starting point for two review papers, describing and applying a modelling framework for predicting metal-based ENP bioavailability in soils and waters currently being written up by NERC.
Overall the scientific findings of WP4 (based on the WP3 experiments) that may have specific consequences for (long-term) risk assessment of ENPs in the environment are:
- Soil pH is a major factor governing metal-based ENP fate and effects in soils, and the effect of pH seems to overrule that of other factors like organic matter content.
- For waters in addition to pH, ionic strength and/or salinity are major factors affecting ENP fate and effects to aquatic organisms.
- The presence of a coating inhibit their dissolution rate, but may increase toxicity of metal-based ENPs .
- While aquatic ENP behaviour dynamics are fast and operate on shorter timescales tan the standard experiments this is NOT true for soils, where the fate dynamics may have timescales of many months of years and will not be covered by standard test protocols.
- Long-term equilibration of metal-based ENPs in soils may lead to a slow dissolution of metal ions over time, which may still continue after 1 year and is concentration and pH dependent.
- Long-term equilibration in soils may decreased toxicity of ZnO ENPs, but did cause and increased toxicity of Ag ENPs as was shown in a NanoFATE collaborative study performed by NERC.
WP5 - ENP toxicokinetics and toxicodynamics.
Objective 5.1: Identify the internal fate of trackable particles in exposed organisms by combining time series exposures with detailed tissue and cellular morphology using qualitative and quantitative particle detection methods. [UNIPMN, CU, UOXF.DJ IHPP].
This objective was initially addressed through a comprehensive review of fixation and visualization techniques along with methodologies and guidelines for determining trace metals in biological tissues with particular emphasis on silver. Completion of this output by CU and UNIPM and IHPP provided a basis from which the consortium has been able to develop the research activities to approach further objectives within this work package. For research purposes, and to make sure complex techniques are only applied when we have evidence from more routine methods that the atomic elements or high electron density signatures for the ENPs are present. We proposed a tiered approach to the tracking of ENPs in tissues by a logical order of moving a sample in an incremental manner through the various analytical and characterization methods. Regarding the ability of tracking nano- objects within cells and tissues, much has been done within the NanoFATE framework. Studies have shown that ENPs undergo a range of physical and chemical transformations in the environment to the extent that exposures to pristine materials will occur only rarely in nature. Methods to track assimilation and internal distributions must, therefore, be capable of detecting these modified forms. Techniques able to provide a strong chemical signature, such as X-ray absorption spectroscopy and EM, allowed tracking ENPs into animal tissues. A striking evidence of that was provided by quantitative modelling of X-ray Absorption Near Edge Spectroscopy (XANES) profiles obtained in ZnO ENP exposed earthworms obtained at the Diamond light source synchrotron in the UK (CU, UOXF.DJ NERC). This analysis indicated that the Zn atomic centres within the matrix of the focal deposits imaged in epidermis and nephridium of earthworms are associated with oxide, a chemical phase distinguishable from the phosphate-associated Zn pool in the chloragogenous tissue. Similar indications of differential metal speciation for Ag in tissues of both woodlice and earthworms depending on whether they were exposed to Ag ENPs or ions, where also found when doing some technically XANES work were undertaken at Diamond to explore if working at the lower energy Ag k-edge was possible. Preliminary analysis of further XANES data of zinc speciation in nematodes exposed to ZnO or artificially aged (phosphatised) ZnO indicate that the main form within nematode tissues was a zinc phosphate regardless of exposure form. All-in-all suggesting that organisms of different size and physiology may be affected by and experience ENP exposure in different ways in terms of internalised metal species. Extensive EM work was carried out also in other organisms such as mussel tissues, document that here uptake involves dissolution and reconstitution of metal (Ag) environmental/biogenic nanoparticle into gills (UNIPMN, UOXF.DJ CU).
Objective 5.2: Investigate the link between internalisation and toxicological effects in time series exposures, and use model prediction of toxicokinetic parameters to study the internal tissue distribution and form of ENPs (using analytical electron microscopy) to attribute causal effects to intact ENPs and dissolved free metal ions. [VUA, NERC, UGOT, CU, UOXF.DJ]
This objective bridges Tasks 5.1 “Modelling and measurement of ENP toxicodynamics and toxicokinetics” and 5.2 “Comparative mechanisms of action of dissolved, bulk, and nanoparticulate forms” by linking the temporal effects of ENPs and their eventual biological fate with the mechanisms of toxicity observed. The findings of the kinetic and bioaccumulation studies of NanoFATE has been summarise in D5.5. Bioconcentration Factor (BCF) values, uptake (k1) and excretion/release (k2) constants were determined according to various mathematical approach including the one-compartment modelling. These studies were carried out on several and distinct biological systems that differ from each other for both structural/morphological features and/or their ecological niche: unicellular algae, springtails, daphnids, terrestrial detritivorous (Eisenia spp) and marine filter feeding bivalves (Mytilus spp). The emerging picture has been rather heterogeneous depending on either model species and/or chemicals. In springtails, higher ionic BCFs has been explained by greater k2 values for silver ENP (5 nm Amepox, Ag7) exposures (VUA). In algae treated with same chemicals, conversely, greater BCF values for ions appeared to depend on k1, whilst k2 did not differ significantly (UAVR). A higher BCF value for the ionic form has consistently been observed in earthworms (Eisenia foetida) exposed to cerium salt or different uncoated CeO2 ENPs at 256 mg/kg in Lufa 2.2 soil during standard 28 day exposures (NERC). Finally, in bivalve molluscs, taking carefully into account the actual silver dose in seawater column, BCF values of ionic silver and two different Ag ENP differing in size (5 and 50 nm) types were very similar (UNIPMN).
The Dynamic Energy Budget (DEBtox) model was considered to allow further comparisons of toxicokinetics and toxicodynamics between ENPs and ions (D5.3). The experiments with C. elegans (NERC and VUA) and D. magna (UAVR and VUA) were analysed with a DEBtox model that assumes chemicals have to be taken up to cause effects. The kinetics of uptake determines how fast chemical concentrations in the body of the test organism will develop with time at different exposure concentrations. By following development of toxicity with time upon exposure to a test chemical, DEBtox is able to test assumptions on its mode of action. Experiments with M. galloprovincialis did not allow application of the DEBtox model. These experiments included observation of ecophysiological parameters upon exposure to Ag ENPs and were analysed with a normal DEB model. This enabled making predictions of the long-term effects of Ag ENPs on these test organisms.
For the toxicity of ZnO ENPs (30 nm and 80-100 nm) and Ag ENPs (5 nm, Amepox Ag7) to D. magna, DEBtox analysis showed increased metabolic costs for growth and reproduction to be the best fitting mode of action. Non-nano ZnO had a different mode of action compared to ZnO ENPs, leading to increased metabolic costs for daphnid growth. Daphnids were more tolerant to non-nano ZnO, which could be attributed to the higher elimination rate on non-nano ZnO compared to ZnO ENPs and ZnCl2. Ag ENPs (5-8 nm) and AgNO3 had the same metabolic mode of action in D. magna (increased metabolic costs for growth and reproduction). Elimination rates constants for Ag obtained using DEBtox were higher than values from uptake-elimination kinetics experiments determining Ag levels in the daphnids, suggesting that at least part of the Ag taken up is sequestered and not involved in causing toxicity to the animals. The DEBtox analysis of data on the toxicity of Ag ENPs to C. elegans concluded that “hazard to the embryo” was the most likely mode of action, which links well with the much higher sensitivities found for the juvenile lifestages in the full lifecycle exposures underpinning the DEBtox analysis.
DEB analysis of M. galloprovincialis data showed that ecophysiological effects on respiration, food assimilation efficiency and growth would cause effects upon life-long exposure with No Observed Effect Concentrations (NOECs) for ionic Ag and Ag ENPs of 0.2 and 2 μg/L, respectively (corresponding to about 100 ng/L/h in both cases when the actual dose was derived by means of a kinetic model).
Objective 5.3: Investigate the suitability of existing biomarkers for major modes of action known to be linked to ENP toxicity (e.g. reactive oxygen species (ROS) generation, lysosomal responses) for identification of ENP exposure and early-stage effects. [UNIPMN, NERC, VUA, CU]
This objective has been achieved within Task 5.2 by the completion of deliverable D5.4. UNIPMN carried out a comprehensive study to assess a role for oxyl radicals in Ag ENP toxicity on marine mussels. These results showed that in the digestive gland low doses of 5 nm silver nanoparticles (Amepox, Ag7) (as low as 10 ng/L/h for 28 days) are able to modulate relative mRNA abundances of antioxidant responsive genes such as Super Oxide Dismutase, Glutathione Peroxidase and Glutathione Transferase and to inhibit catalase enzymatic activity. Clear effects of an oxidative stress syndrome, however, were evident only at much higher ENPs exposure levels (1000 ng/L/h) for which lipid peroxidation of biological membrane occurred along with a drop in bioenergetics performances as demonstrated by a scope for growth assessment.
UAVR and NERC tested the effects of zinc oxide nanoparticles. In isopods zinc nanoparticles did not elicit a clear molecular/biochemical antioxidant response. In isopods exposed to 30 nm ZnO (Synthetic surfactant-free P99/30 nm NanoSun), at 500 and 750 mg/kg lipid peroxidation products were even lower than in control conditions, confirming the antioxidant properties of zinc. A set of biochemical biomarkers (GST and CAT activity, along with lipid peroxidation byproduct accumulation) was studied in the three life-stages of Porcellionides pruinosus (juveniles, pre- adults and adults) but no zinc effect in either ionic or nano form was observed (UAVR).
NERC investigated the roles of metal chelating proteins (metallotioneins and phytochelatins) in protection against ZnO nanoparticle (Synthetic surfactant-free P99/30 nm NanoSun) effects using a metal sensitive triple knockout (mtl-1;mtl-2;pcs-1(zs2)) nematode strain. The results suggested that these metalloproteins are instrumental in the protection against oxyl radical formation and further cytotoxic damage. Moreover, ZnO ENP exposure was shown to decrease growth and development, reproductive capacity and lifespan, effects that were amplified in a triple knockout strain.
Objective 5.4: Utilise a toxicogenomic approach to investigate novel modes of action associated with ENP exposure when compared to effects attributable to the free metal ions in samples from organisms exposed at ion levels close to concentrations known to cause life-history effects. [UNIPMN, CU, VUA]
The project aimed to evaluate toxicogenomic responses to ENPs and related ionic forms by investigating the global transcript responses in three target organisms, the nematode Caenorhabditis elegans, the filter-feeding marine mussel Mytilus galloprovincialis and the epigeic earthworm Eisenia fetida. These activities were informed by WP3 (ENP ecotoxicology) to establish exposure conditions, but also required the creation of appropriate data structures that will allow the consortium to link; i) the ENP physical/chemical parameters, ii) the routes/time of exposure, and iii) observed life-history responses to the impact observed within the global transcriptome.
UNIPMN generated normalized cDNA libraries and obtained about 90 millions of short reads (100 paired ends, 9 giga bases in total) by means of RNA sequencing on an Illumina HiSeq platform. This allowed the construction of a 15 K high-resolution oligonucleotide array that has been used for transcriptomic studies in Mytilus spp exposed to either 5 nm Ag ENPs (Ag7, Amepox) or ionic silver. Results obtained from short-term exposures at a nominal dose of 100 µg/L showed that apart from a few common features (hsp70s, proton transporters, peptidases and ubiquitine-related genes), then, distinct biological processes and cellular components occurred for the effects of ions compared to effects from ENPs. In both gills and the digestive gland of ENP exposed specimens, microarray analysis strongly suggested the involvement and impact of the endocytic pathway. Membrane docking, secretion, lysosomes, exocyst were over-represented GO terms in ENPs samples. Ionic silver appeared to evoke even more deleterious processes such as DNA repair and oxidative stress (the latter in gills), however, 2D proteomic analysis evidenced a higher rate of oxidized –carbonylated- proteins in ENP treated specimens. The systems toxicology approach, thus, suggested that oxyl radicals represent a common toxicity initiator (or mediator), but different Adverse Outcome Pathways are occurring for the effects of the two physical form of silver tested.
Very similar results to those seen in mussels were obtained in nematodes for the effects of Ag7 ENP. NERC with an award from FP7 Quality Nano, investigated transcriptomic profiles using high-resolution arrays (Agilent). These results again showed few common features between ionic and ENP silver exposed worms. For nematodes exposed to ionic or ENP forms of silver genes associated with neuropeptide signaling pathways, the endocytic pathway and metalo-peptidase activity. NERC, moreover, investigated the role of coating and size finding the distinct molecular traits above in different ways, with size differentially affecting genes from gene ontologies integral to membrane function, and charge affecting gene ontologies relating to ion binding.
Finally, RNA sequencing has been used to carry out a comprehensive transcriptomic survey in the earthworm Eisenia foetida exposed to either silver ions (49 mg/Kg, 2.2. Lufa soil) or PVP coated silver nanoparticles (456 mg/kg) representing equitoxic conditions (EC50 for reproduction, 28 days). This work required getting prior information on earthworm genotypes, sequencing, assembly and mapping of millions of short DNA reads. These results showed a consistent overlap between differentially expressed genes obtained for the effects of Ag ENP or ionic silver. Over-represented features (Gene Ontology terms) associated to the common group were represented by ribosome, adenylate cyclase, sugar and aminoacid metabolism. However, specific processes associated to microtubules, projection of membrane structures, basolateral plasma membrane were identified when a direct statistical comparison between Ag ENP and Ag gene associated ions GO terms was undertaken.
These results does suggest that early and low dose effects from ENP exposure are different from those seen during exposure to metal ion based exposures, with a broad picture across the species of membranes, cell structure and metal trafficking.
WP6: Integrated environmental risk assessment
Objective 6.1: Conduct literature review and a stakeholder consultation to define current and potential future uses of ENPs and determine the driving parameters (including associated uncertainties) that define plausible future release scenarios. Collate consumption/discharge information and fate information (WP2) for use in exposure modelling. [F+B, NERC, NT, AXME, SYMLOG]
This objective was led by F+B with the aim to obtain information on the size of the ENP market and assess market developments. Considerable effort was invested for communicating the layout, content and dissemination strategy of the questionnaire with stakeholders from industry, with particular help from NT and AXME being received. An important stage was to compile the market information acquired from the questionnaires sent to more than 500 companies with the collaboration of several industry bodies, into an overview of market consumption of ENPs. However, this overview could not be completed as originally intended, due to the extremely poor questionnaire response from industry. Thus, in order to achieve its objectives, the study had to focus on reviewing the available evidence, by compiling and evaluating production volumes and modelled environmental concentrations from the scientific and grey literature for the initial exposure predictions and risk maps. The topic of consumption and discharge was re-examined by NERC and F&B towards the end of the NanoFATE project and it was decided to rely on the studies carried out at ETH Switzerland (Piccinno et al., 2012; Sun et al., 2014). The updated information and was used in the revised exposure and risk maps. The modelling assumed that for Europe 32 tons of nano Ag and 1580 tons of nano ZnO were consumed per year. The estimates for nano CeO2 discharge and soil accumulation focused on use in fuel additives in the UK and was reported in D2.2 (Johnson and Park, 2012). For soils, the highest predicted contamination level was 0.016 mg/kg within 20m of a road following seven years of continuous deposition. This value would represent 0.027% of reported natural background cerium. River water contamination considered direct aerial deposition and indirect contamination via runoff in the water and entrained soil sediment, with the highest level of 0.02 ng/L predicted. The highest predicted water concentration of 300 ng/L was associated with water draining from a road surface, assuming a restricted deposition spread. These predictions are well below most toxicological levels of concern.
Objective 6.2: Conduct initial estimates of the concentrations of nanoparticles in different environmental compartments (air, sludge/soil, water and sediments) using the standard EU EUSES modelling approach based on the scenarios developed in Objective 1.[NERC, NT]
This objective was completed by NERC with help from NT. The NanoFATE scenario for airborne entry of commercial use ENPs into the environment was CeO2 used as fuel additives, where possible resulting exposure levels in soils and surface waters were found to be low (D2.2 published in Johnson and Park, 2012). The likely distribution of ZnO and Ag concentrations in European surface waters based on the assumptions that the total consumption of these ENP materials are used equally by all EU citizens and disposed of “down the drain”. The actual geographical modelling was undertaken using the GWAVA (Global Water AVailability Assessment) model (see further below), mapping of WWTP emission points across the EU and knowledge from WP2 on WWTP retention efficiencies for ENPs. For soils environmental release estimation represented a conservative tier 1 approach to predicting ENPs in the soil compartment either from aerial deposition and/or the deposition of sewage sludge containing bound particles. As the majority of ENPs which go down the drain are considered to associate with sludge this is perhaps the most important route, by weight, through which ENPs are released to the natural environment. This deliverable meets the requirements of the Work Package 6 objectives 6.2 and 6.4. These initial estimates informed the designs of realistic ecotoxicology studies in WP3 and WP4. In later revisions of the risk mapping the original soil estimate concentration maps were refined by highlighting agricultural land likely to receive sewage sludge. Research into national differences in sewage sludge disposal also informed these soil concentration maps.
Objective 6.3: Generate robust Predicted No Effect Concentrations (PNECs) by integration of the experimental hazard data from WP3, the bioavailability data from WP4 and ecotoxicology data from the literature with the exposure assessments to develop a sound understanding of the environmental hazards of ENPs and how environmental media (water and soil) affect this. [VUA, UAVR, NERC, UC, UNIPMN]
This was a separate and distinct task from the chemical risk assessment work in WP6. This objective was fulfilled firstly through a report on phylogenetic and trait based analysis of effects across species and the range of ENPs used including a discussion on applicability of the species sensitivity distribution (SSD) approach. Secondly, through the generation of SSD distributions; this included ZnO and Ag in both dissolved metal ion and ENP forms, in both water and soil. The value of applying either assessment (safety) factor based considerations from the ECHA guidance document, or the SSD approach to the available species toxicity information for obtaining appropriate predicted no effect concentrations (PNECs) was reviewed. It was considered that using SSDs was the most robust approach and sufficient data existed (with the exception of nano CeO2) for this to be attempted. However, it was acknowledged that reviewing nanoparticle literature remains problematic due to problems of the characterisation and measurement of the nanoparticle exposures being inconsistent. SSDs were prepared for nano Ag, ionic Ag, nano ZnO and ionic Zn based on combining the available literature with NanoFATE’s own data. In both cases the most sensitive species in water were bacteria and planktivores. The evidence from one specific case of feeding inhibition of in Daphnia indicated that nano ZnO was slightly more toxic (weight for weight) than the ionic species in water (Lopes et al., 2014). For Ag there was only one case of ENP toxicity exceeding (on a weight for weight basis) that of ionic Ag, and again in water, namely for the bacterial part of the microbial freshwater communities (biofilms), based on the total biomass.. In all other cases the ion form was more toxic weight for weight than the nano form.
Objective 6.4: Incorporate information on PNECs into multi-media (EUSES) and GIS based models for soil (using ADMS) and receiving waters (using LF2000-WQX) to allow risk visualisation analyses for using generic and taxa specific PNEC and other endpoint data as appropriate. [F+B, NERC].
This objective led by NERC produced colour-coded maps of European catchments available for all partners, showing predicted river water ENP concentrations. These were prepared using the (Global Water AVailability Assessment) GWAVA model. Due to scientific developments it was considered that GWAVA was more suitable for the project as it provided pan-European predictions as opposed to the Low Flows 2000 model (LF2000-WQX), which is UK-based only. Modelling efforts were refined, partly through the findings of WP 2, on removal of ENPs in sewage. In the expected scenario removal in sewage treatment of 93% nano Ag and 85% nano ZnO was selected and in the best case scenario 99% nano Ag and 88% nano ZnO removal from effluent to sludge. The final modelling included the fate process of loss of particles from the water column by settling within the river. Initial methods were developed allowing future mapping to take into account the effects of the EU regional differences in water chemistry, by developing 6 overall EU water classes that could be applied to the EU Water Framework Directive catchments. The overall results showed a small fraction of Europe’s rivers could exceed the NanoFATE generated PNECs.
For soils predicted concentrations in agricultural soils were carried out for nano Ag and ZnO based on the use of sewage sludge as a fertiliser, taking into account country specific policies for land applications of such biosolids. Whilst the predicted values were 2 orders of magnitude below the PNECs for the ionic forms of Ag and ZnO, they could exceed in certain circumstances the nano form PNEC, which had been generated with a 1000-fold assessment factor due to the scarcity of species data.
Using an entirely separate soil prediction method based on soil nitrogen fertilisation requirement related application rates, similar values for the UK were generated. Soil concentrations for nano CeO2 were predicted based on the emission and deposition from vehicle exhausts of fuel additive CeO2. The highest value was considered to represent only 0.027% addition to the natural background ceria.
Objective 6.5 Undertake appraisal of the current and potential future environmental risks of CeO2, ZnO and Ag ENPs, based on the work conducted in WPs 2, 3, 4, 5, including gap analysis relating to the uncertainties in hazard, fate and exposure assessment as well as the use and emission scenario definitions. [NERC, VUA, F+B, UGOT, UAVR, CU].
For the final risk assessment analysis somewhat higher nano Ag and ZnO discharge into the environment was assumed than in the previous analysis. It was encouraging to note that the predicted expected and best case nano Ag WWTP effluent concentrations of 3-19 ng/L was similar to the measured value of 5-6 ng/L in effluent measured by NanoFATE and reported by others (Li et al., 2013; Johnson et al., 2014b). Following the final analysis of SSD the NanoFATE PNECs were at a challenging low level. Thus, whilst most water bodies would not exceed these PNECs the possibility for some exceedance near effluent discharge now existed. Exceedance of soil PNECs were now considered possible for UK soils, but the soil nanoparticle PNECs were generated using a 1000-fold safety factor. Hence more data on the refinement of the PNEC would possibly lower these and re-establish the separation between the PNEC and the predicted environmental concentrations (PECs).
Remaining gaps in our knowledge begin with our continuing ignorance of true nanoparticle consumption and discharge levels in Europe. The detailed knowledge of how ENP transformations affect the fate and behaviour of nanoparticles in sewage, sludge, soil and water remains weak. The ecotoxicity of nanoparticles to different soil organisms is still limited. Further information on the ecotoxicity of weathered nanoparticles (such as sulphide coatings) would be welcome in future. From the fairly comprehensive studies of ecotoxicity to water species the most vulnerable appear to be bacteria and planktivores.
Objective 6.6: Develop appropriate project material to disseminate results and applied methodologies to relevant stakeholders, in order to improve understanding of ENP risks. [NERC, VUA, F+B, UGOT, UAVR, CU, SYMLOG]
This objective was achieved with the publication of risk maps with accompanying guidance published on the NanoFATE website, an effort lead by risk communicators SYMLOG.
As far as we know NanoFATE is the first project to model risk from nanoparticles in Europe in a truely geographic sense. While a first for soils this is still only possible at a basic level due to the lack of geographical linkage between location of biosolid production (for which ENP concentration estimation is possible) and the eventual land application location. In contrast, due to the piped and linked flows, this geographical approach is particularly advantageous in the field of river water concentrations as the temporal and spatial variation in dilution is included. Previous models and risk assessments have used only crude box models to predict soil and river concentrations. For the purposes of risk assessment new low PNECs were used mainly based on a comprehensive assessment by NanoFATE using SSD based approaches generated in the project. The spatio-temporal variability of concentrations in surface waters and soils across Europe was modelled using the model GWAVA including population demographics and sewage effluent discharge point maps. Loading of nano Ag and nano ZnO from sewage to rivers was modelled by accounting for connectivity to sewerage and sewage treatment efficiency. The resulting nano Ag and nano ZnO concentrations in rivers were modelled by considering the effect of dilution, water abstraction, residence time, and particle settling. Temporal variability in particle concentrations as caused by temporal (seasonal) climate variation was simulated using climate data for the 1979-2000 period. Model scenarios also compared the influence of different reported sewage removal rates on potential river concentrations (Johnson et al., 2014a).
References
Cornelis, G., Juergens, M., Tuoriniemi, J., Hassellov, M., 2014. Research report on ZnO and Ag ENP fate in STW: Measurement of nano ZnO and Ag partitioning and removal during primary and secondary sewage treatment stages., NanoFATE Deliverable D2.7 Wallingford, p. 29.
Costantini M.G. 2011. Evaluation of human health risk from cerium added to diesel fuel. Health Effects Institute Communication 9. Boston, MA, USA.Heggelund L.R. Diez-Ortiz, M., Lofts, S., Lahive, E., Jurkschat, K., Wojnarowicz, J., Cedergreen, N., Spurgeon, D., Svendsen, C., 2014. Soil pH effects on the comparative toxicity of dissolved zinc, non-nano and nano ZnO to the earthworm Eisenia fetida. Nanotoxicology 8, 559-572.
Johnson, A.C. Dumont, E., Keller, V., Williams, R.J. 2014a. Report on ENP risk assessment in the different usage scenarios; this will include gap analysis, critical appraisal of available hazard and exposure assessment methodologies and the identification of vulnerable species and environmental compartments, NanoFATE Report - Deliverable 6.7. Natural Environment Research Council, Wallingford, p. 33.
Johnson, A.C. Juergens, M.D. Lawlor, A.J. Cisowska, I., Williams, R.J. 2014b. Particulate and collidal silver in sewage effluent and sludge discharged from British wastewater treatment plants. Chemosphere in press, 19.
Johnson, A.C. Park, B., 2012. Predicting contamination by the fuel additive cerium oxide engineered nanoparticles within the United Kingdom and the associated risks. Environmental Toxicology and Chemistry 31, 2582-2587.
Li, L., Hartmann, G., Doblinger, M., Schuster, M., 2013. Quantification of nanoscale silver particles removal and release from municipal wastewater treatment plants in Germany. Environmental Science & Technology 47, 7317-7323.
Lopes, S., Ribeiro, F., Wojnarowicz, J., Lojkowski, W., Jurkschat, K., Crossley, A., Soares, A.M. Loureiro, S., 2014. Zinc oxide nanoparticles toxicity to Daphnia magna: size-dependent effects and dissolution. Environ Toxicol Chem 33, 190-198.
Piccinno, F., Gottschalk, F., Seeger, S., Nowack, B., 2012. Industrial production quantities and uses of ten engineered nanomaterials in Europe and the world. Journal of Nanoparticle Research 14.
Sun, T.Y. Gottschalk, F., Hungerbuhler, K., Nowack, B., 2014. Comprehensive probabilistic modelling of environmental emissions of engineered nanomaterials. Environmental Pollution 185, 69-76.
Tourinho, P.S. van Gestel, C.A. Lofts, S., Soares, A.M. Loureiro, S., 2013. Influence of soil pH on the toxicity of zinc oxide nanoparticles to the terrestrial isopod Porcellionides pruinosus. Environ Toxicol Chem 32, 2808-2815.
Tourinho, P.S. van Gestel, C.A. Lofts, S., Svendsen, C., Soares, A.M. Loureiro, S., 2012. Metal-based nanoparticles in soil: fate, behavior, and effects on soil invertebrates. Environ Toxicol Chem 31, 1679-1692.
Waalewijn-Kool, P.L. Diez Ortiz, M., van Straalen, N.M. van Gestel, C.A. 2013a. Sorption, dissolution and pH determine the long-term equilibration and toxicity of coated and uncoated ZnO nanoparticles in soil. Environ Pollut 178, 59-64.
Waalewijn-Kool, P.L. Ortiz, M.D. Lofts, S., van Gestel, C.A. 2013b. The effect of pH on the toxicity of zinc oxide nanoparticles to Folsomia candida in amended field soil. Environ Toxicol Chem 32, 2349-2355.
Zhang, J., Nazarenko, Y., et al., 2013. Impacts of a Nanosized Ceria Additive on Diesel Engine Emissions of Particulate and Gaseous Pollutants. Environ. Sci. Technol. 47, 13077−13085
Potential Impact:
Potential Impact
WP1: Characterisation and tracking of ENPs during processes involved in fate and toxicity
WP1 Underpinned much of the work of the other work packages within the project, however, it also contributed significantly to commercialisation aspects of the project with new commercial products being developed. As part of their role within the project, NanoTRADE (NT; http://www.nanotrade.cz/) and Amepox Ltd.(AMEPOX; http://www.amepox.com.pl/) were required to develop their products to facilitate their use in environmental fate and toxicity testing. As a result of NT in collaboration with AMEPOX have launched a new product approved for veterinary use. Further more the Insitute of High Presure Physics - Polish Academy of Sciences (IHPP), who were tasked with developing and producing cobalt-doped zinc oxide particles, now have a joint venture with an industrial company for scale up of production. Patents have been filed for materials developed during the project. Further as well as these direct commercialisation opportunities, the involvement in NanoFATE of partners NT, AMEPOX and IHPP has also provided each with valuable information and knowledge of the risk assessment issues for their productsas an contribution to imp[roving their sustainable production and use of ENPs. This information is generally beneficial to SMEs involved in the development and commercialisation of nanotechnology products. As direct outcomes of these activities,NT has initiated a Memorandum in collaboration with other manufacturers for NANOSAFETY in the Czech Republic. IHPP have made recommendations for safety regulations in Poland based on conclusions from results of nano-ZnO strong aggregation in test waters, behaviour in humid conditions and solubility studies. IHPP have contributed to a Book of recommendations for the European Commission for nano-regulations along with the NANOFORCE project (http://www.nanoforceproject.eu/) .
NanoFATE has also led to improvement in knowledge based and quality assured analytical and consultancy services at University of Oxford (UOXF.DJ) and IHPP. This has provided a focus for research to support the safety assessment and sustainable production of nanotechnology through the national centres of excellence. Dedicated training courses have been held at UOXF.DJ to increase the level of understanding in the physical characterisation of materials for participants from outside materials science field and to increase the utilisation of the academic public funded equipment base.
WP2: ENP environmental behaviour and fate modelling
WP2 is responsible for the nanomaterial fate, behaviour and transport studies to enable predictions of environmental concentrations have developed both cutting edge experimental and modelling approaches. In the first deliverable it was identified which processes and properties will be most important to determine experimentally in order to mechanistically parameterize fate processes for modelling. The experimental designs are targeting such key parameters.
The highest novelty in WP2 together with WP6 is the EU-wide spatially distributed emissions, dilutions, and fate processes that are built on GIS-based hydrology, population density and water chemistry models.In the production of these new risk assessment products, Nanofate has gone way beyond the state-of-the-art in terms of emission and fate modelling. The results has been used by regulatory authorities in informal discussions and formalised reporting of the current risk assessment status of ENPs nationally and in Europe (e.g. The Swedish Chemical Inspection and the UK Department for Environment, Food and Rural Affairs (Defra). Representatives from these organisation have also highlighted these results in paper and statement made through their role as have representatives to the OECD working party on manufactured nanomaterials.
WP3 – ENP Ecotoxicology
NanoFATE has contributed significantly to the improvement of existing exposure protocols to test ENPs in both aquatic media and soil. This was attained based on full collaboration amongst partners. One highlight on improved methodology was the development of a Synthetic Soil Pore Water that allows toxicity trials to be carried out with nematodes in liquid culture. This contrast with previous high ionic strength media that resulted in extensive agglomeration. For soil exposure, a homogeneous introduction of ENPs in soil is crucial and was achieved by using a suspension spiking method or by mixing dry ENP powder directly in dry soil followed by adjusting moisture content to the desired level. For food contamination, alder leaves were submerged in stock solutions of nanoparticles and then air dried, prior to be used. These methodologies have already been followed in laboratories in the UK, Ireland, Norway, France, Italy, Portugal Belgium, Germany and the USA, who are adopting the NanoFATE methods for routine testing in support of standardised toxicity testing.
NanoFATE research has shown that existing methodologies for testing chemical effects to aquatic organisms are usually suitable for ENPs, although some modifications may be needed depending on the nanomaterial in study. As an example, guidance for media renewal, i.e. the choice for the period of exposure media renewal, should be based on results from agglomeration/aggregation experiments. These studies are relevant to both the acute and chronic tests.
WP3 ecotoxicity tests were also carry out to study ENPs bioavailability in test media, assessment of ENP uptake and provide material to tissue bank in order to evaluate potential internalization of nanoparticles in WPs 4 and 5.
The work conducted in WP3 has made a major contribution to the ENP toxicity database for Ag, ZnO and CeO2. The data covered multiple taxonomic groups making the data particulay valuable for risk assessment (aquatic: crustaceans, fish, molluscs, fish, algae, bacteria; soil: earthworms, springtails, isopods, plants, bacterial) The overall results from WP3 has, thus, considerably improve data available for nanoparticles’ toxicity, and were used in WP6 along with bibliography data to derive more robust PNEC values, and estimate risk from the ENPs studied. The fact that all studies have been conducted using protocols optimised for testing the toxicity of ENPs means that the generated data is of high quality. There is an acknowledged gap in the literature in relation to the availability of high quality effect studies on the chronic toxicity of ENPs for environmentally relevant receptors. The body of work conducted to date in NanoFATE is certain to be a major contribution to the literature in this area. Indeed both the nanoparticles-specific and ionic form silver ecotoxicity data, generated through NanoFATE, has been incorporated directly into the Chemical Safety Report Dossier for REACH registration for silver.
WP4 - ENP bioavailability – relations between soil and water chemistry and particle properties.
WP4 studies on the bioavailability of ENPs in test media were used for ecotoxicity testing and assessment of ENP uptake in WPs 3 and 5, respectively. Results of WP4 are essential for a proper understanding of the impact of media and particle properties on ENP bioavailability, uptake and effects in soil and water and provided input for spatially specific risk assessment efforts undertaken both in WP6 and outside NanoFATE , including by regulatory agencies.
WP4 partners have contributed to training activities to improve the EU skill base. Results of the project were presented at several meetings, including annual meetings of SETAC and the NanoSafety Cluster meetings. This assisted in communicating the results and their significance for standard testing protocol development and data interpretation to the wider community including industry and regulators. WP4 partners, through the NanoFATE coordinator, contributed to drafting text for the Horizon 2020 programme and the NanoSafety Vision 2015-2020.
At the national level, WP4 partners contributed to developing strategy and status documents regarding the environmental and health assessment of ENPs. Through national coordinators, WP4 partners also contributed to activities of the OECD Working Party for Manufactured Nanomaterials (WPMN) by providing input for meetings and commenting on proposals for modifying existing or developing new test guidelines for testing ENPs. The knowledgebase generated through NanoFATE has been vital to these activities.
WP4 partners have actively contributed to the organization of sessions on the fate and effects of nanomaterials at join scientific/industry/regulator meetings, e.g. the annual meetings of the European branch of the Society of Environmental Toxicology and Chemistry (SETAC). In addition, WP4 partners contributed to activities of the SETAC Advisory Group on Nanomaterials that bring together academic, industry and regulator perspective on environmental nanosafety.
WP4 has also sought to make all results freely available in publications in the peer-reviewed literature (especially open access) and though other written report. For that purpose, results of the critical review of ENP bioavailability (Deliverable 4.1) have been turned into three review papers, of which one has already been published. The continuous update of this data overview, which is a joint effort of WPs 3, 4 and 5, has provided the ecotoxicological database (EC50s, EC10s etc.) for the derivation of Predicted No-Effect Concentrations (PNECs) by WP6, and can be used in collaboration by other EU projects.
The scientific findings of WP4 that may have specific consequences of (long-term) risk assessment of ENPs in the environment are:
- Soil pH is a major factor governing metal-based ENP fate and effects in soils, and the effect of pH seems to overrule that of other factors like organic matter content.
- In addition to pH, ionic strength and/or salinity is a major factor affecting ENP fate and effects to aquatic organisms.
- The presence of a coating increase toxicity of metal-based ENPs but may inhibit their dissolution rate.
- Long-term equilibration of metal-based ENPs soils may lead to a slow dissolution of metal ions over time, which may still continue after 1 year and is concentration and pH dependent.
- Long-term equilibration may decreased toxicity of ZnO ENPs, but did cause and increased toxicity of Ag ENPs as was shown in a NanoFATE related study performed by NERC.
WP5: ENP toxicokinetics and toxicodynamics
WP5 underpinned our understanding of the mechanistic effects of ENPs in biological systems employed by consortium members undertaking ecotoxicology experiments in the entire NanoFate project. This work allowed a better comprehension of modes of action of metal based ENPs to project partners and help to widen the societal acceptance for the newly developed technologies to stakeholders by reducing the uncertainty associated to risk of nanomaterials. WP5 in collaboration with ENNSATOX also developed the architecture and structure of a relational database that allowed data sharing across NanoSafety Cluster projects and the option to submit data to ICPC NanoNet database.
WP5 partners have contributed to the scientific content of booklet “Nanosafety in Europe 2015-2025: Towards Safe and Sustainable Nanomaterials and Nanotechnology Innovations”. WP5 partners have contributed to training activities to improve the EU skill base being part of the NanoFATE PhD workshops in Aveiro, Oxford and Birmingham during different FP7-project meeting/workshops.
WP5 partners have organized an Open Workshop on Mechanistic Toxicology of Nanomaterials (QualityNano NanoFATE & NanoMILE Joint Meeting, Birmingham, UK, March 5th -6th 2014) illustrating in front of a wide international audience recent advances obtained from toxicogenomics and proteomics studies on ENPs. WP5 scientists have also organized national and international meeting sessions such as at SETAC in order to promote mechanistic studies on nanomaterials as a comprehensive tool to reduce risk assessment uncertainty.
Data obtained within the WP5 framework have been presented at several EU-US meetings, discussed with a heterogeneous audience (scientist, policy makers, regulators) in specific EU workshops and roundtables, debated within the EU Nanosafety Cluster meetings. The aforementioned networking activities also promoted the establishment of new scientific collaboration at national and international level.
WP5, beyond the primary NanoFATE scopes, realized innovative research tools for the transcriptomic analysis in invertebrate species such as mussels and earthworms. These tools are commercially exploitable. This includes not only providing information on mode of action to support better risk assessment and risk communication, but also through the development of potential tools for identifying ENP exposure in organisms taken from the field.
WP6 ENP Risk Assessment
As far as we know NanoFate is the first project to model risk from nanoparticles in Europe in a truly geographic and mechanistic sense. This is particularly advantageous in the field of river water concentrations as the temporal and spatial variation in dilution is included. Further the inclusion of parameterised values for specific fate process, both in wastewater treatment and in the river channel itself, provides a unique view of the state of the art that has not been so widely captured in other initiative. Previous models and risk assessments have used only crude box models to predict soil and river concentrations. For the purposes of risk assessment new low PNECs were used mainly based on a comprehensive assessment by NanoFate using species sensitivity distributions generated in the project. The spatio-temporal variability of concentrations in surface waters and soils across Europe was modelled using the model GWAVA usage population and sewage effluent discharge point maps. Loading of nano Ag and nano ZnO from sewage to rivers was modelled by accounting for connectivity to sewerage and sewage treatment efficiency. The resulting nano Ag and nano ZnO concentrations in rivers were modelled by considering the effect of dilution, water abstraction, residence time, and particle settling. Temporal variability in particle concentrations as caused by temporal climate variation was simulated using climate data for the 1979-2000 period. Model scenarios compared the influence of different reported sewage removal rates.
For Europe around 50% of rivers would be predicted to have nano Ag concentrations of 0.0001 ng/L or less and 95% of rivers have 0.1 ng/L or more of nano Ag. Thus, most European rivers would have nano Ag levels 2-orders of magnitude below the PNEC. For Europe around 50% of rivers would be predicted to have nano ZnO concentrations of 1 ng/L or less and 95% of rivers have 500 ng/L or less of nano ZnO. The latter value is only a factor 5 below the PNEC. This analysis puts nano ZnO as being of greater concern to the aquatic environment than nano Ag. This is because of the apparent much higher use of nano ZnO and in this simulation the particularly high water column settlement rate chosen for nano Ag. Although no data are presently available for toxicity to river benthic invertebrates, predictions were also made for nanoparticle accumulation in river bed sediments which indicated that sediment loading of nano Ag of between <0.5 and >0.5 mg/m/y with the higher loading rates being more apparent in northern Europe.
Soil concentrations for agricultural land were predicted for countries across Europe based on 1) national sludge disposal practices and their populations; and, 2) based on soil nitrogen fertilisation recommendations. The soil predictions assumed all the nano Ag and nano ZnO removed from sewage effluent were available for application to agricultural land via sewage sludge. Model scenarios compared different transfer rates to sewage sludge, different proportions of land receiving the national sludge (1 or 57%) and different soil ploughing (mixing) depths of 12 cm and 25 cm. These scenarios largely followed that recommended by EUSES the simple model used in risk assessment for chemicals in the EU. The first approach facilitated soil predictions on a geographic basis something which has not been attempted before to our knowledge.
The highest predicted soil value for nano Ag was 8.8 µg/kg dry weight for agricultural soils whilst the lowest value for nano Ag toxicity was 4.8 mg/kg. However, if a 1000x safety factor is invoked to derive a PNEC of 4.8 µg/kg then we would exceed this. A safety factor of 1000 is common applied when using acute toxicity data to chemical risk assessments. Similarly, the highest nano ZnO soil concentration was 939 µg/kg dry whilst the most conservative value for nano ZnO toxicity was 119 mg/kg. However, if a 1000x safety factor is invoked to derive a PNEC of 120 µg/kg then we would exceed this. This soil prediction approach used here contrasts with some other authors who do not consider the route of sludge being transferred to agriculture.
To summarise, NanoFate research indicates that assuming no change in form, nano Ag and ZnO concentrations in sewage effluent could reach or exceed the PNECs, although this is heavily dependant upon the large safety factor applied. However, given dilution and settling within river bodies most rivers will be well below PNECs. Soil nano Ag and ZnO concentrations could exceed the lowest PNEC particularly in the UK given its disposal policies and limited agricultural land area with respect to its population size.
Engagement with regulatory and standardisation bodies.
NanoFATE partners have played a very active role at international conferences such as the annual SETAC conferences (Europe and North America) which provide an opportunity for academia and government and industry to engage with each other regarding the sustainable use of ENPs. The NanoFATE consortia has participated in the form of chairing sessions, platform presentations, posters, and contributing to the SETAC Advisory Group on Nanomaterials.
The regulatory outreach of the project has been significantly ramped up over the course of the project as the main messages of the project have been more clearly defined. These have responded the following phases of the disseimination strategy: A) better understanding the knowledge demands of regulators; B) identifying individuals and collecting their contact information to ensure accurate targeting and enhance communication flow; C) fostering two-way communication to build a good fit between regulatory needs and science outputs and D) specific dissemination actions. Actions undertaken under this increasing regulatory outreach included advice and direct input by NanoFATE partners to the OECD Working Party on Manufactured Nanomaterials (WPMN). NanoFATE researchers attended expert meetings or provided information on demand from national delegates to: (i) OECD- Testing and Assessment Workshop (Paris, December 2013; (ii) OECD WPMN Steering Group 5/6 "Risk Assessment and Regulatory Programmes" (Paris, Dec. 2013), (iii) Expert meeting, Ecotoxicology of nanomaterials (Berlin, February 2013). More meetings will be attended after the close of the project. A broader regulatory audience was addressed through participation by SYMLOG at the ‘EU 2nd Regulatory Review on nanomaterials’ (30 January 2013, Brussels). During this meeting the big questions, issues and needs by regulators were captured and a flyer highlighting the work of NanoFATE was developed for direct distribution at this workshop. The big questions are presented on the NanoFATE website along with advice notes developed through the NanoFATE project that respond to these issues. The Advice Notes, largely disseminated by NanoFATE and NanoSafety Cluster newsletters, are useful for researchers but also for regulators. They point the way to further research, and flag issues that regulators should attend to when setting assessment guidelines and regulatory practice. The NanoFATE Advice Notes are downloadable two-page documents containing links to the supporting NanoFATE research in deliverables and peer-reviewed papers (which are also accessible via our online library).
NanoFATE has fostered and maintained an active dialogue with representatives from the European Chemicals Agency (ECHA). These discussions have centred around the questions and issues that ECHA have regarding the assessment of nanomaterials under REACH regulations. Broad areas covered have included characterisation and hazard assessment of ENPs as well as in what circumstances ENPs can be grouped for assessment. As a result of these discussions the NanoFATE project coordinator has been invited by ECHA to participate Topical Scientific Workshop on Nanomaterials in October 2014 as a member of the Scientific Committee for this workshop.
Within the 48 months of the project, a wealth of planned and opportunistic dissemination actions have been implemented including regular website updates, an extensive online library of Advice Notes, deliverable summaries, hi-res image bank available to non-project scientists, training workshop reports, newsletters, a social sciences survey, peer-reviewed journal articles, conferences, posters, invited lectures, summer or winter school teaching, training workshops, expert advice and knowledge sharing, and some mass media appearances.
The public website www.nanofate.eu is the hub of all our dissemination actions: ALL NanoFATE information transits from or transits towards the website and is constantly made available for all our audiences in dedicated sections to best fit their needs. The website statistics are therefore an important means of measuring the impact of our dissemination and communication strategy. Google statistics were implemented as of February 2011 to track the traffic and visits on our project website. The figures below show encouraging statistics (very positive compared to other FP7 project websites known to the WPL). The number of unique visitors continued to grow and at mid February 2014 had reached a cumulative total of 51,052 unique visitors (some 70% increase from 36 months) and 894,159 page views (some 35% increase). NanoFATE still continues to recruit new visitors (around 30-35 per month) as well as enjoying a consistently high rate of return visits (twice as many as new visitors).
NanoFATE scientists have played a full and active role in NanoSAFETY Cluster (NSF) activities. Consortia members have participated in Cluster working groups (WGs) concerned with materials characterisiation (WG1), hazard assessment (WG2&3), exposure assessment (WG4), Databasing (WG5), risk assessment (WG6), modelling (WG7) and dissemination (WG7). In addition NanoFATE members have contributed to a wide range of NSF events co-organising 6 workshops/training events, participating in a further 15 meetings, and direct collaborative activities with 11 other NSF sister projects.
Adding to our knowledge base and training the nano-scientists of the future
To date, 21 ISI papers have been published based on NanoFATE results (see publications list for details), with a further 5 papers submitted and 41 planned. All seven RTD work packages have contributed to the preparation of papers. Most publications present the work of more than one work package reflecting the multi-disciplinary approach within the project. In addition to the academic publications we have focussed our five newsletters, aspects of our web content and two International Innovation articles at a broader audience to engage the full spectrum of stakeholders relating to the sustainable use of ENPs.
At the inception of the project it was recognised that one of the key limitations to the advancement of our understanding of the fate and toxicity of ENPs was a lack of expertise within the research community. Therefore there has been a significant training component to the NanoFATE project. In addition to the two open PhD workshops organised by NanoFATE consortia members has provided training in a further seven training events and workshops. NanoFATE has directly provided a research platform for the studies of three MSc students, nine PhD and DPhil candidates and six early-career-stage post doctoral researchers.
Work package 7 (Dissemination and Training activities) developed the Project Website and a series of Electronic Newsletters (circulated to more than 800 contacts) for the external profile and communication of results and designed to be readable & interesting for a broad lay audience. Two summary articles were published in “International Innovation” with hardcopies delivered to hundreds of European decision makers. The website was overhauled to accompany these publications and yielded excellent statistics: a cumulative total of 51,052 unique visitors to the site representing 894,159 page visits. In this way a significant audience was informed of six NanoFATE training workshops & modules organised jointly with other NSC projects, 33 contributions to NanoSafety Cluster events, & 29 conference presentations. On the publication side, results are flowing: 21 peer reviewed ISI articles, 7 currently submitted, 4 mass media publications and 46ISI papers under preparation. Regulators have been engaged through targeted emails, website updates, distribution of leaflets at the Commission’s 2nd Regulatory Review on Nanomaterials Workshop, NanoFATE members have been invited to OECD WPMN working group meeting by national regulators to act as scientific experts and have delivered policy advice also at various national meetings and to sit on the scientific committee for a forthcoming ECHA workshop on nanomaterials.
List of Websites:
http://www.nanofate.eu
Executive Summary
Description of the work performed during the NanoFATE project and the main results achieved: All of the nine main NanoFATE S&T objectives, were successfully addressed at the end of the project, and can be summarised as follows:
I. Sourced, produced and fully characterised the required versions of commercial engineered nano particles (ENPs) (Ag, ZnO and CeO2) and developed and matched the tagged version of ZnO as closely as possible (eventually using Co doping). (Main Obj. 1)
II. Established particle behaviour in the pure ecotox media to be used and at higher than environmental concentrations. Modified the ecotox media and protocols to optimise ENP behaviour and organism performance to allow undertaking of hazard assessment studies under most realistic exposure conditions possible (Main Obj. 3, 4 & 5)
III. Established acute and chronic toxicity that enabled selection of relevant doses for the progression of the work into environmentally relevant media conditions. (Main Obj. 5, 6, 7 & 8)
IV. Developed initial simple assumption based fate models and estimated worst case environmental concentrations. Refined these model assumptions as real data and understanding (NanoFATE and externally derived) became available and identified parameters of most influence and importance for ENP specific fate and risk assessments(Main Obj. 2 & 8)
V. Train all staff cross discipline, but especially the 14 PhDs and Post Docs directly employed on NanoFATE, and disseminate our findings to other EU NMP projects and stakeholders (Main Obj. 9)
Delivery was divided in the 3 components below covering the 6 RTD and the dissemination WPs.
Particle chemistry, behaviour and fate component:
Work package 1 (Characterisation & tracking of ENPs) provided high quality well characterised commercial particles for the remaining project partners. Subsequent work focused on characterisation of these ENPs in exposure media & natural systems showing that the dissolution rates of ENPs (ZnO, Co-doped ZnO & Ag) not only depended heavily on the water and media involved, but particularly for zinc oxide on the particle morphology & coating. Work package 2 (ENP environmental behaviour & fate modelling) initially identified & prioritised specific properties to be considered during the development, adaptation & validation of environmental fate models for ENPs. Initial fate models were developed to predict CeO2 deposition in soils & nano ZnO & Ag influent to and discharge from sewage treatment works. WP2 then parameterized the ENP environmental fate modelling with information on dissolution, agglomeration (homo- & hetero-agg.), sedimentation & aging (transformations) properties. Kinetic models building on colloid theory have been developed to deliver fate predictions at the EU & regional scale.
Ecotoxicology and bioavailability component:
Work Package 3 (ENP Ecotoxicology) developed improved standard ecotoxicological exposure protocols, that ensures relevant & homogenous presentation of ENPs during toxicity testing. Employing these improved protocols, the exposures needed for the hazard assessment was completed. This included both acute and chronic testing, plus specific work to deliver data on bioavailability drivers for WP4. Samples have been archived for use in WP5 (some retained for future collaborative explorations) and sensitive endpoint data was collated for use in PNEC assessment (WP6). As for CeO2 little evidence of toxicity was seen across the tested organisms extra work was included to look at long term (6 months) accumulation and also co-exposure of CeO2 with organic contaminants, while for ZnO in freshwater a relavant confounding factor was seen as UV radiation meaning combined ZnO x UV exposures were prioritised. Sufficient data was produced to increase by 50% the species data points available for use in Species Sensitivity Distribution derivation in WP6. Work package 4 (ENP bioavailability) has produced and published critical reviews of all available information addressing data quality & identifying which environmental factors have the greatest proven effect on the bioavailability & toxicity of ENPs to organisms. This identified pH, organic matter & cation effects to be the most important for bioavailability and trials testing these were consequently implemented within WP3. For the non inert ENPs of Ag and ZnO in waters dynamics seemed either fast with changes mainly occurring within hours or so slow (changes over weeks to months) that test protocols would yield appropriate hazard data, as test are either short (1-3 days) or requires media exchanges at the same timescale of a few days. Contrastingly for soils fate processes appeared slow in comparison to the 2-8 week duration of the tests. Hence for soils additionally long-term exposures (6-12months) addressing ageing were conducted. The main findings included establishing that Zn toxicity in all cases was highest at lower soil pH; that coated ZnO ENPs were more toxic to springtails than uncoated ones; that ZnO ENPs & non-nano ZnO particles in soil released Zn ions very slowly (still going on after 12 months), with this release being highest at intermediate exposure levels; that ageing of ZnO ENPs increased Zn concentrations in pore water, but reduced the toxicity for both nano ZnO, non-nano ZnO & ZnCl2; & that release of Zn into soil pore water was much slower for coated than uncoated ZnO. The longterm aging for Ag ENPs in soil produced increasing toxicity over the 12 month period eventually reaching, but not exceeding, the toxicity of the ionic Ag exposures (on a weight for weight basis). Work package 5 (ENP toxicokinetics & toxicodynamics) developed sample handling protocols & preservation systems enabling successful tissue banking. A tiered approach was developed for tracking of ENPs in tissues that focused activities in a systematic manner exploiting high throughput methods in initial sample screening, and samples only passing to complex & costly analyses for the most detailed studies if ENP presence was likely. Samples were analysed for biological markers of ENPs to develop the knowledge of signatures of possible early ENP tissue damage. An agreed data structure was developed with a deliberately simple and flexible format to enable data entry at the labbench and export to the Cluster database when established. Toxicokinetic studies were completed for ENPs in soil & aquatic invertebrates and in situ characterisation of ENP uptake has been undertaken.
Risk assessment and knowledge transfer component:
Work package 6 (Integrated risk assessment) has assessed ENP production & product incorporation estimates based upon a review of the peer-reviewed as well as grey literature on production volumes of the three NanoFATE ENPs. The initial fate & distribution models (from “local” to “EU region” scale) developed produced very informative initial conservative (high) estimates of predicted environmental concentration (PEC) maps. Improved fate models and updated production estimates has refined these models and have been combined with protective environmental Predicted No Effect Concentrations (PNEC) based on species sensitivity distributions developed in NanoFATE.
The final refined modelling suggests that at current predicted levels of nano Ag and ZnO are unlikely to exceed thresholds of concern for aquatic or soil species. However, the hazard data is very sparse (especially for soils) and in details the conclusions look as follows:
In soils two quite separate prediction methods gave similar soil PECs for the UK (proven to represent the EU worst case in terms of land application of WWTP sludge). For nano silver the highest predicted value over a 12 yr period in UK soil scenario was 0.009 mg/kg whilst the lowest effect level of concern (EC50) for nano Ag in soil was 4.8 mg/kg (2-orders of magnitude higher). However, due to the paucity of toxicity data and SSD approach was not possible and hence the ECHA guidance document for deriving PNECs is followed which means using an assessment factor of 1000 on the most sensitive species, and then the nano Ag PNEC would be 0.0048 mg/kg and the NanoFATE PECs exceed this. A recent soil PNEC proposed for silver REACH registration proposed 1.24 mg/kg (W/W) based on read across from silver ions to which nanosilver is presumed to correspond. This would err towards the NanoFATE predicted nanosilver soil concentrations being of low risk. But more nanosilver soil toxicity data is needed.
For nano ZnO in soils the highest predicted value for nano ZnO over a 12 yr period in UK soil scenario was 0.939 mg/kg. The lowest effect level of concern for a single species for nano ZnO in soil was 119 mg/kg (2-orders of magnitude higher). However, again the paucity of toxicity data leads to inclusion of an assessment factor of 1000 and the resulting nano ZnO PNEC should be 0.119 mg/kg which the NanoFATE predictions exceed. Therefore, more nano ZnO soil toxicity data is needed.
In waters, the predicted starting effluent concentration for nano ZnO is closer to its lowest PNEC than for nanosilver suggesting that at the point of discharge to the river environment nano ZnO is the greater concern. If the current best estimate of a rapid loss mechanism for nanosilver in rivers is accepted, then for Europe around 50% of rivers would be predicted to have concentrations of 0.0001 ng/L or less and 95% of rivers have 0.1 ng/L or less of nanosilver in the expected (not worst case) scenario. Thus, most European rivers would have nanosilver levels 2-orders of magnitude below the Ag PNEC of 29 ng/L. For nano ZnO around 50% of rivers would be predicted to have concentrations of 1 ng/L or less and 95% of rivers have 500 ng/L or less. This is only 1 order of magnitude below the PNEC of 2200 ng/L. This analysis puts nano ZnO as being of greater concern to the aquatic environment than nanosilver. This is because of the apparent much higher use of nano ZnO and in this simulation the high water column settlement rate chosen for nanosilver (an effective half-life of as little as 45 min for nanosilver compared to 15 d for nano ZnO).
Work package 7 (Dissemination and Training activities) developed the Project Website and a series of Electronic Newsletters (circulated to more than 800 contacts) for the external profile and communication of results and designed to be readable & interesting for a broad lay audience. Two summary articles were published in “International Innovation” with hardcopies delivered to hundreds of European decision makers. The website was overhauled to accompany these publications and yielded excellent statistics: a cumulative total of 51,052 unique visitors to the site representing 894,159 page visits. In this way a significant audience was informed of eight NanoFATE training workshops & modules organised jointly with other NSC projects, 33 contributions to NanoSafety Cluster events, & 29 conference presentations. On the publication side, results are flowing: 21 peer reviewed ISI articles, 7 currently submitted, 4 mass media publications and 46ISI papers under preparation. Regulators have been engaged through targeted emails, website updates, distribution of leaflets at the Commission’s 2nd Regulatory Review on Nanomaterials Workshop, NanoFATE members have been invited to OECD WPMN working group meeting by national regulators to act as scientific experts and have delivered policy advice also at various national meetings and to sit on the scientific committee for a forthcoming ECHA workshop on nanomaterials.
Project Context and Objectives:
Project Context and Objectives
NanoFATE was conceived to fill knowledge and methodological gaps currently impeding sound assessment of environmental risks posed by engineered nanoparticles (ENPs). Our vision was to assess environmental ENP fate and risk in for example high-volume products for which recycling is not an option, namely; fuel additives, personal care products and antibacterial products. To represent these products two commercial ENPs of CeO2, ZnO and Ag (of varying size, surface and core chemistries) were followed through their post-production life cycles, i.e. from environmental entry as “spent product”, through waste treatment to their final fates and potential toxic effects. The applicability of current fate and risk assessment methods and identify improvements required for assessment of ENPs at an early stage were tested.
Objectives: Delivery of a systematic study of the environmental fate and toxicity of selected ENPs will entail addressing nine S&T objectives:
1. Design, tagging and manufacture of ENPs
2. Analysis of ENP interactions with abiotic and biotic entities
3. Generating predictive models for ENP exposure in waters and sludge-amended soils
4. Studying the fate and behaviour of ENPs through wastewater treatment
5. Determining acute and chronic ecotoxicity
6. Assessing effects of physico-chemical properties on ENP bioavailability
7. Defining mechanisms of uptake, internal trafficking, and toxicity
8. Developing spatial RA model(s)
9. Improving understanding of ENP risks
Background
To support the responsible development of the nanotechnology sector, it must be recognised that the development of environmental risk assessment methods should not lag too far behind those for human health. Past experiences highlight a number of other environmental issues such as organochlorine pesticide usage, endocrine disruptions, secondary effects of pharmaceuticals on wildlife, and genetic modification, where environmental impacts rather that direct effects on human health, emerged as the major area of concern. In each of these cases, the unexpected nature of these effects had a profound affect on public confidence in new technologies. This required that rapid regulatory action was put in place to control and mitigate risks. By ignoring effects on the environment, nanotechnology runs the risk that similar damaging and costly effects could occur.
Because of the initial and wholly understandable focus on direct risk to human health, knowledge of fundamental aspects of the environmental risks associated with ENPs is low in several key areas. These include:
- the post-production fate of ENPs from entry into the environment to final residence;
- how ENP-ENP and environmental interactions affect the biotic availability of ENPs and how different ENP properties (size, surface) affect exposure/uptake;
- how crucial ENP properties such as size distribution, surface chemistry, shape and optical properties influence toxicity;
- chronic aspects of ecotoxicity, which to date has mainly been assessed at environmentally unrealistic concentrations or in inconclusive studies where it was uncertain whether the co-solvent used for dispersal, impurities, or the ENP itself resulted in the observed toxic effect;
- the mechanisms of toxicity of ENPs when compared to the bulk chemical or free metal ion and how observed effects of ENPs on the expression of genes or proteins associated with particular pathways (e.g. such as oxidative stress in cell lines) relate to higher level in vivo effects;
- the fitness for purpose of existing risk assessment approaches designed for standard chemicals for use with ENPs and the modifications needed to allow existing frameworks and policies to be used in future for the risk assessments of nanotechnology products.
By studying the fate and behaviour of the selected ENPs and their effects on biota, NanoFATE has gone beyond the superficial initial assessments that have been possible so far, thereby enabling a scientifically rigorous analysis in relation to each of the above aspects. The data gained in meeting each of the nine NanoFATE objectives will allow us to go beyond the current state-of-the-art.
Project Results:
Scientific and technological challenges
The potential human health effects of ENPs are of obvious importance and a review of European research and national programs indicates that a number of ongoing projects are already addressing this issue (e.g. NANOTOX, CELLNANOTOX, IMPART, NANOSH, NanoReTox). In distinct contrast, there are as yet few studies that have focused on developing and refining methods to assess the fate of ENPs in ecosystems (e.g. soils and natural waters) and any resulting ecotoxicological effects. For this reason NanoFATE intends to focus on these neglected aspects and their integration.
To support the responsible development of the nanotechnology sector, it must be recognised that the development of environmental risk assessment methods should not lag too far behind those for human health. Past experiences highlight a number of other environmental issues such as organochlorine pesticide usage (Newton and Wyllie 1992; Newton et al. 1999; Sibly et al. 2000), endocrine disruptions (Jobling et al. 1998; Tyler et al. 1998), secondary effects of pharmaceuticals on wildlife (Oaks et al. 2004), and genetic modification (Haughton et al. 2003; Heard et al. 2003), where environmental impacts rather that direct effects on human health, emerged as the major area of concern. In each of these cases, the unexpected nature of these effects had a profound affect on public confidence in new technologies. This required that rapid regulatory action was put in place to control and mitigate risks. By ignoring effects on the environment, nanotechnology runs the risk that similar damaging and costly effects could occur.
Because of the initial and wholly understandable focus on direct risk to human health, knowledge of fundamental aspects of the environmental risks associated with ENPs is low in several key areas. These include:
- the post-production fate of ENPs from entry into the environment to final residence;
- how ENP-ENP and environmental interactions affect the biotic availability of ENPs and how different ENP properties (size, surface) affect exposure/uptake;
- how crucial ENP properties such as size distribution, surface chemistry, shape and optical properties influence toxicity;
- chronic aspects of ecotoxicity, which to date has mainly been assessed at environmentally unrealistic concentrations or in inconclusive studies where it was uncertain whether the co-solvent used for dispersal, impurities, or the ENP itself resulted in the observed toxic effect;
- the mechanisms of toxicity of ENPs when compared to the bulk chemical or free metal ion and how observed effects of ENPs on the expression of genes or proteins associated with particular pathways (e.g. such as oxidative stress in cell lines) relate to higher level in vivo effects;
- the fitness for purpose of existing risk assessment approaches designed for standard chemicals for use with ENPs and the modifications needed to allow existing frameworks and policies to be used in future for the risk assessments of nanotechnology products.
By studying the fate and behaviour of the selected ENPs and their effects on biota, NanoFATE will go beyond the superficial initial assessments that have been possible so far, thereby enabling a scientifically rigorous analysis in relation to each of the above aspects. The data gained in meeting each of the nine NanoFATE objectives will allow us to go beyond the current state-of-the-art as set out in the section below.
Scientific & Technological results
WP 1. Characterisation and tracking of ENPs during processes involved in fate and toxicity.
Objective 1.1. Manufacture, procurement and detailed characterisation of the two versions of each ENP (CeO2, ZnO and Ag) available from the industrial partners and selected as the focus for risk characterisation, fate, and effects experiments and risk assessment. This characterisation will build on manufacturers and existing data where possible and consider carefully the forms in which ENPs will be delivered (e.g. powders or suspensions, including issues of carriers and stabilisers). [UOXF.DJ NT, AXME ]
This objective has been achieved through the successful completion of D1.1 D1.2 and D1.3 ensuring also M1.1 M1.2 and M1.3. Beneficiaries contributed by ensuring the commercial ENP starting materials for the project were suitable, available and fully characterised. UOXF.DJ (P3) characterised the initially selected ENPs discovering the proposed ZnO ENP material was unsuitable for the project. This had an insufficiently controlled particle size and shape and containing both spherical, cylindrical and high aspect ratio particles, with the latter dissolving too rapidly for the proposed work. Material supplied from Microniser was subsequently validated and found suitable for use. AMEPOX (P10) initially supplied small (3-8 nm) Ag ENP with a well-controlled narrow size distribution, but stable only in organic solvent. AMEPOX developed procedures for preparing and dispersing the same ENPs directly in water for use in WP 2-5 and undertook tests for this water-based colloid to ensure adequate stability with time. NT (P6) supplied Ag ENP powder (50 nm). NT collaborated with AMEPOX (P10) developing a way of dispersing and stabilising their material in water for use in WP 2-5. UOXF.DJ (P3) undertook quality control checks on all material supplied for WP 2-5, as well as periodic checks on stock material, using a multi technique approach. This demonstrated the importance of TEM imaging of all samples.
Objective 1.2. Production of ZnO ENPs tagged with rare earth elements in very low natural concentrations (e.g. Ytterbium or Luthetium) that match as closely as possible the ZnO nanoparticles common in commercial products. This is needed to allow tracking of the ZnO ENPs by ICP-based element analysis in the experiments to measure fate and effects. (For Ce and Ag based ENPs the natural background is so low they can be tracked by their core elements, while Zn has too high natural background for this to work if not tagged). An alternative approach, although significantly more expensive, is to synthesize the ZnO nanoparticles from a stable enriched isotope, e.g. 67Zn or 70Zn. [IHPP]
This objective has been achieved through the successful completion of D1.4 ensuring also M1.4. IHPP (P8) developed methods of producing doped ZnO ENPs initially using Al as proof of method. Subsequently they refined this method to produce material with Gd and Lu dopants, however because of the mismatch in atomic scale of rare earth elements compared with zinc it was not possible to incorporate rare earth dopant atoms into the ZnO lattice. UOXF.DJ (P3) demonstrated with TEM that the dopants consisted of discrete material on the surface of the ZnO ENPs in the form of oxide. Because of the similarity in atomic scale the consortium decided to try Co doped ZnO ENPs and IHPP (P8) successfully manufactured such material. The natural Co background level in soil meant a high dopant concentration was required and a 10 wt% dopant material was produced for the tracking experiments performed in WP 2-4. IHPP (P8) and UOXF.DJ (P3) characterised this material for particle size distribution (30 – 60 nm) and lattice parameter proving it was comparable to the undoped ZnO.
Objective 1.3. Characterisation of the ENPs to be used initially in the standard exposure medias of WP3. For ZnO this will be done for both the tagged and untagged for ZnO to check how the behaviour of tagged particles match that of untagged particles. (IHPP, UOXF.DJ UGOT]
This objective has been achieved through the successful completion of D1.5 D 1.6 and D1.7. UOXF.DJ (P3) and UGOT (P11) carried out experiments using AMEPOX Ag ENPs dispersed in water and also the exposure media employed in the ecotoxicological tests in WP3. This investigation enabled a better understanding of the form of the ENPs as presented to the organisms being dosed and the dynamics and rates of change in these ENP forms during exposures. IHPP (P8) similarly provided this information on the behaviour of ZnO and Co doped ZnO in ecotoxicological exposure media.
Objective 1.4. Validation of element based analysis for gross tracking of ENPs. Method development and validation for the detailed tracking and localisation of ENPs in exposure and tissues relevant in WP2, WP3 and WP4 to refine methods for detecting particle in environmental media and in cells and tissue of exposed organisms. [IHPP, UOXF.DJ UGOT]
This objective has been achieved through the successful completion of D1.8 ensuring also M1.5. IHPP (P8) supplied 10 wt% Co doped ZnO ENPs characterised by IHPP (P8), UOXF.DJ (P3) and UGOT (P11) with instructions for use in fate experiments by NERC (P1) Lumbricus rubellus (1000 and 4000 µg Zn/g ENPs dosed in soil and food and compared with ionic dosing), VUA (P2) Folsomia candida (400 and 1000 µ g Zn/g ENPs dosed in soil) and UAVR (P4) Daphnia magna (100 µg Zn/L dosed). We have demonstrated that the Co doped ZnO nanoparticles manufactured by IHPP are fit for the purpose they were designed for and can be used in tracking experiments, if the dose is sufficiently high, to determine whether the nanoparticles are taken up by organisms as intact particles or as ions. Successful detection methods in these trials were digestion and bulk metal analysis such as AAS, OES, and ICP-MS. High resolution X-ray Fluorescence imaging of sections of Lumbricus rubellus showed concentrations of Zn, but the Co was below the limit of detection for imaging. From the Lumbricus rubellus experiments undertaken by NERC (P1) we now have evidence suggesting that metals in particle form may be uptaken as intact particles or at least through different routes than metal ions. Co is taken up via the gut, but not the skin when presented as Co doped ZnO nanoparticles, while in contrast Co ions when mixed with Zn ions are not taken up by the gut, although there possibly is minor ion uptake via the skin.
Objective 1.5. Identification, quantification and characterisation of ENPs in samples provided from WPs 2, 3 and 4 to provide: (i) qualitative and quantitative data on environment-ENP interactions, and (ii) the association of ENPs with both matrix and organism components in different environmental systems. [IHPP, UOXF.DJ UGOT, NERC, VUA, CU, UAVR]
This objective has been achieved by all beneficiaries working together to produce a protocol for the detection of ENPs in complex and dilute organic systems (media and organisms). Samples of dosed media and organisms were produced from ecotox experiments performed by UGOT, NERC, VUA and UAVR (WP2-4) along with bulk chemical analysis data. In samples where elemental or other simple high throughput analysis indicated presence of the ENP material, CU (P9) performed a series of specialised imaging experiments with differing spatial resolution and limits of detection e.g. CARS and Synchrotron Micro focused X-ray imaging (XRF and CAS) and applying expert sample preparation techniques for subsequent HRTEM performed by UOXF.DJ on areas of high concentration of ENP material as determined by the other more rapid, but less detailed imaging methods. This multi method, tiered approach to detection of ENPs in environmental samples has been submitted for publication. Most of this work has involved the use of AMEPOX Ag ENPs and has been reported in WP 2-4. We have HRTEM evidence for indistinguishable particle formation in organisms after dosing with both Ag ENP and AgNO3 suggesting dissolution of Ag ENP to an ionic form and re-precipitation though it has not yet been determined at what point in this occurs.
WP 2. ENP environmental behaviour and fate modelling
Objective 2.1: Identification of critical parameters governing the fate of ENPs [NERC, F+B, UGOT].
Initially efforts focused on the estimations of product usage volumes and worst case release amounts. For ZnO and Ag, estimates of production volumes were therefore taken from literature and calculated into per capita down the drain usage and environmental entry through waste water. For CeO2 fuel additive worst case usage calculations were based on all diesel cars in the UK using CeO2 additives and all deposition restricted to within 20m of UK A and M road network only (Johnson, 2012). An important assumption nano CeO2 emissions is the effectiveness, or otherwise, of fine particulate filters in exhaust systems. Although commercial (PSA Citroen and Cummins Diesel, plus filter manufactures Johnson Matthey) and US EPA references to diesel particulate filter exist quoting performances ranging from 85% to 99% efficiency, only one published study could be found, but also giving 95-99% removal (Constantini, 2011). A concern here is that the perfoamances may refere to particle sizes higher than those potentially carrying nano CeO2 through the exhaust system. Recent data from studies carried out using a test engine (Zhang et. al, 2103) that do not incorporate any exhaust filters indicates that only 40% of the cerium added to the fuel is emitted, with the remainder lost in the pipes, unburned residues and tars (pers. Comms. Zhang). However, to maintain worst case estimates to give the most conservative of potential risk predictions, the NanoFATE fate modelling include no removal in filters. This still indicated a difference in the order of 3-5 orders of magnitude between worst case soil exposure and effect levels, and hence any further precision in the modelling was not considered critical.
The big focus for NanoFATE was the identification and prioritization of specific properties of the environment and materials that need principal consideration during the development, adaptation and validation of environmental fate models for ENPs were. Ionic strength and composition, natural organic matter quantity and pH were concluded to be the dominant properties of natural waters, although not solely important. It was concluded that the core composition, and surface functionalization were the two most critical material factors, although many others could play a role in the fate processes. In terrestrial systems, the soil solids composition and grain size will be important in addition to the parameters listed above for water.
The importance of these parameters in the initial fate models developed (with WP6) to predict nano ZnO and Ag discharge (effluent and sludge) from sewage treatment works and further environmental fate have been assessed and increasingly better parameterised through the course of the NanoFATE project. This has been delivered through the WP2 experimental fate testing framework for fate of ENPs in natural waters. This has developed the first preliminary data gathered for such models to be developed further. Kinetic models building on colloid theory are especially promising and have been developed to deliver fate predictions at the EU and regional scale (see Objective 6.5 in the work package 6 section of this report).
insert Cerium removal here:-
Objective 2.2: Development and application of an experimental framework for fate testing of ENP in natural waters
The full objective was to develop experimental methods to study ENP-environment interactions and their consequences for ENPs fate and behaviour through wastewater, sewage treatment processes, and in receiving natural waters and sludge [UGOT, NERC, UOXF.DJ IHPP].
In order to mimic the natural environment in the laboratory or in field microcosms decisions about complexity and relevance versus simplicity and mechanistic understanding have to be taken. Studies of real environmental matrices are highly relevant, but it is very difficult to probe the nanomaterials and how they interact with living and dead natural matter in the full complexity. Furthermore, real natural environmental matrices are time and space specific so the capability to extrapolate to wider areas is very limited. On the other hand, building up knowledge on environmental relevance by simple electrolyte solutions of increasing relevance will never lead to applicable knowledge on the risk for the environment. Therefore in NanoFATE we selected a framework that built on model natural waters which was standardized and slightly more simplified than real natural waters. The model waters were based on a multivariate classification of all Europes surface waters. In addition to electrolytes, also model natural organic matter and model inorganic colloids was developed and implemented in these 6 “standard European model waters”. The experimental platforms for fate studies was built around the model waters using the method time-resolved dynamic light scattering (DLS) for homoagglomeration studies and a custom designed settling cone for studies of homo- plus heteroagglomeration and their combined influence on sedimentation. Suitable analytical techniques had to be developed to analyze the particles in the settling cone test platform.
In order to conduct a proof of principle study that investigated agglomeration-sedimentation processes in isolation from particle dissolution meant using gold particles as a substitute for the Nanofate commercial silver particles. These gold particles have many similarities with silver, but do not oxidize and dissolve. The main conclusions on the fate of the tested ENP under the realistic concentrations was that they did not settle out as a result on agglomeration during the two months study under most of the model waters, but in waters with high ionic strength the removal was significant and directly dependant on the ionic strength. The homo and heteroagglomeration also depend on if it was the gold nanoparticle or the model inorganic colloid (illite) that was destabilized by the specific model natural water.
Heteroagglomeration which has recently been claimed to be a dominant driving fate process in natural waters, did not under the tested conditions become such an efficient removal process.
Objectives 2.3 and 2.5 Apply refined experimental approach to investigate the effects of particle properties on ENP fate and behavior through waste water treatment processes and associated effluents and interatively incorporate avaoialable data on ENP-environment interactions to refine predicted environmental concentrations for waters and soils [UGOT, NERC, UOXF.DJ IHPP]
In order to estimate the removal rate in waste water treatment plants before the single particle ICPMS method had been fully developed, a study of ultrafiltration which captured the nanoparticulate and colloidal silver was used in influent and effluent samples from a range of different types of waste water treatment plants (WWTP). The average silver removal rate was found to be 49%. In addition, the most useful reference found for ZnO was that of Lombi et al. (2012) who found the partition coefficients (Kd) for the partitioning between supernatants and sewage sludge of three different ZnO NPs to be 1258, 1667, and 915. The average Kd that they found for nano ZnO was 1280. If we assume that most sorption of nano ZnO in WWTPs takes place in activated sludge tanks with a sludge concentration of about 0.0035 kg/L than that would mean that the ratio of sorbed over non-sorbed nano ZnO would be (1280*0.0035=) 4.48. Thus, the WWTP removal of nano ZnO would be about (100*4.48/(4.48+1)=) 81.8%. For nano Ag, assuming a per capita consumption of 0.055 mg/cap/d and the 49% removal discussed above, the per capita discharge would be 0.028 mg/cap/d. An average per capita discharge of 160 L/d would result in a potential effluent concentration of 175 ng/L. For nano ZnO a larger consumption rate of 0.55 mg/cap/d and a sewage removal of 82%, would give a per capita discharge of 0.099 mg/cap/d and a potential effluent concentration of 618 ng/L. It should be noted these values are likely to be a considerable over-estimates for the following reasons: (i) a proportion of ENP production in a continent may be exported, and (ii) many ENP products are unlikely to release their ENPs to water.
Objective 2.4: Track effects of particle properties and environmental physicochemical properties on the rates of dissolution of ENPs during wastewater treatment and in a range of natural soils and water. [UGOT, UOXF.DJ IHPP].
This objective was principally achieved through the development and application of advanced spectroscopic tools and microcosms to study the fate of silver nanoparticles in waste water treatment processes
Small angle Neutron Scattering (SANS) was evaluated to probe settling of commercial silver nanoparticles under real sewage conditions compared to pure water. The method is promising, but requires access to synchrotron beam time.
Single particle ICPMS (spICPMS) is a method capable of probing low concentration of metallic nanoparticles in complex samples such as natural waters of sewage. Although a very useful method it has limitations in terms of the smallest detectable size that can be discriminated from the background of dissolved and colloidal ions of the same element. Therefore we have improved the method with two different approaches. Firstly, a statistical method to describe the combined nanoparticulate and background signals in spICPMS was developed, which allowed us to deconvolute the nanoparticle contribution from the combined signal (Cornelis and Hassellöv, 2013). This allowed a significant improvement in smallest detectable size. Secondly, a totally new method of data acquisition, using true realtime signal collection (dwell time 0.1ms compared to 10 ms) allowed us to follow the whole nanoparticle signal peak instead of just discriminating outlier spikes in the signal (Tuoriniemi et al 2014). Both these method improvements could be combined to allow detection of nanoparticles down to the physical limits (around 5-7 nm for silver and 4-5 nm for gold). This method was used to track the silver nanoparticle fate transformations during the WWTP microcosm study (Cornelis et al., 2014).
WP 3. ENP Ecotoxicology
Objective 3.1: Collate information on the suitability of existing ecotoxicological tests systems for ENP testing and subsequent optimisation of the NanoFATE biotest systems. [NERC, VUA, UOXF.DJ UAVR, UNIPMN, CU, AXME, UGOT]
This objective was achieved through the successful completion of D3.1. The contributions of individual beneficiaries were carried out in dialogue to ensure protocols were optimised with regard to both the aspects of presenting the ENP to the organisms in the most realistic form possible and that organism conditions were as close to optimal as possible. Data on the toxicity of ENPs to different test organisms were collected from the literature to assess the suitability of currently available test guidelines for assessing ENP toxicity to the organisms selected for the work in NanoFATE. An on-line database was developed to summarize and critically reviewing literature toxicity data in collaboration with WP4, serving for all partners and WPs including as the main hazard data source used in WP6 for risk assessment
Partners from AMEPOX, IHPP, UOXF.DJ and UGOT worked together to provide well characterised nanoparticles suitable for performing ecotoxicity tests. Both UGOT and IHPP provided essential information regarding agglomeration behaviour of Ag and ZnO ENPs in the different aquatic media so that the most suitable could be identified and determine if any adaptations of the standard protocols or media were needed, and explaining some of the observed cases of toxicity differences.
NERC contributed with improvements to the nematode test media and, in collaboration with CU to the dosing protocols needed for delivering the ENPs into the earthworm test soils. The standard OECD and ISO protocols were adapted to provide stable and reproducible exposures for the experiments with the ENPs to be tested in the project. For both organisms, the modifications were chiefly based around ensuring optimisation of the realism of the presentation and exposure form of the ENPs through modifying exposure media and dosing protocols. VUA optimised exposures with the soil arthropod Folsomia candida (Collembola), using the standard test guideline prescribed by ISO (1999) as a basis. Rather than using an artificial soil, a natural standard soil Lufa 2.2 was recommended, since this is better suited to some type of analysis for particle detection in the soil media (in particular the high kaolinite content of the OECD soil proved a difficulty). Tests were run with ZnO ENP, including non-nano ZnO and ZnCl2 for comparison. UAVR optimised a protocol for the terrestrial isopod Porcellionides pruinosus, including optimisation of dosing to provide both soil and food exposure routes for ZnO and Ag nanoparticles. Comparisons of different methods of spiking the soil with ENPs demonstrated that the best method of spiking soil with nanoparticles will depend on the ENPs properties and delivery form.
For the aquatic compartment, UAVR checked the standardized procedures for testing chemicals with the cladoceran Daphnia magna for its suitability to conduct the toxicological assessment with ENPs. To test the effects of ENP agglomeration on toxicity, results based on changing media after 24 and 48 hrs were compared. Additionally for chronic testing, one issue that emerged was the use of supplementary seaweed extract usually used as a source of micronutrients, especially selenium, which could act as a chelating substance that might decrease ENP bioavailability in the test media. Test carried out without adding seaweed extract were conducted, but these resulted in high mortality of juveniles. As a compromise, it is suggested that the test media is maintained close to the standard composition, but should be regularly renewed during chronic tests to maintain worst case ENP presentation for the hazard assessment. Snails (Physa acuta) were also tested and protocols adapted mainly on water composition, where two media types were used: Artificial Pond Water (APW) and APW-adapted without CaCl2. This approach was also valuable for WP4 in terms of relating silver bioavailability with toxicity. UNIPMN developed modifications on the protocol for acute toxicity evaluation of ENPs in marine mussels using artificial sea water as test medium.
UGOT evaluated the protocol to test the effects of ENPs on growth and reproduction of the green algae Chlamydomonas reinhardtii and Pseudokirchneriella subcapitata, the cyanobacterium Synechoccocus leopoldensis and the diatom Cyclotella meninghiana. Test procedures followed the corresponding OECD guideline 201, and adaptations were carried out to improve methodology for testing the ecotoxicity of ENPs. The Woods Hole MBL media showed to be one of the best media to provide more accurate toxicity endpoints.
Objective 3.2. Ecotoxicological hazard characterisation of pure ENPs using the optimised media to establish chronic thresholds of concern (no observed effect concentrations, benchmark concentrations) for population relevant endpoints (e.g. reproduction) for selected aquatic and terrestrial test species. [NERC, VUA, UOXF.DJ UAVR, UNIPMN, CU, AXME, UGOT]
The work conducted in NanoFATE represents one of the largest bodies of bespoke cross species ecotoxicity testing of the chronic effects of commercial ENPs on environmental receptors. The data will, thus, be an important contribution to the risk assessment literature.
Using the artificial test media developed by NERC the toxicity of ionic Zn and Ag and multiple NP forms of Ag, ZnO and CeO2 were assessed in Caenorhabditis elegans. In collaboration with CU, exposures to CeO2 ENPs in nematodes and earthworms were undertaken, as well as exposures of the earthworm Eisenia andrei to dissolved Zn (as ZnCl2), NanoSun ZnO ENP (30nm), the two Z-cote ZnO ENPs (Z-cote and Z-cote HP1), and a non-nano ZnO (Sigma non-nano). Tissues from exposures were banked for M3.2 and use in WP5.
In addition, NERC developed a high throughput screening method for assessing the toxicity of ENPs to multiple bacterial species.VUA determined the toxicity of ZnO (both coated and uncoated ENP and non-nano forms) to the collembolan Folsomia candida in Lufa 2.2 natural soil, focusing on effects on survival and reproduction after 28 days of exposure. The toxicity of two different Ag ENPs to Folsomia candida in Lufa 2.2 natural soil was also tested. The Ag ENPs were obtained from AMEPOX and included the 3-8 nm ENPs dissolved either in hexane or in water.
Methodological improvements were developed and implemented to enable quantification of Ag in the test soils. Previously used digestion methods had to be modified to increase recovery of Ag from the test soil. This resulted in good recoveries in case of soil spiked with AgNO3 or Ag ENPs in water. This was also crucial to determine accurate body burdens (WP5), focusing on Ag uptake kinetics and possible Ag ENP uptake.
Chronic ENP toxicity to Porcellionides pruinosus was evaluated using Ag ENPs and ZnO ENPs by UAVR through testing avoidance behaviour and feeding inhibition. In addition, samples from the hepatopancreas and gut tissues of isopods exposed to Ag ENPs (8nm) via food were collected and supplied for use in WP5. In this approach, the avoidance behaviour test showed to be a very powerful tool, detecting 50% of avoidance behaviour in organisms at concentration of 18mg Ag ENP/Kg soil. VUA and UAVR showed that CeO2 ENPs presented no adverse effects on springtails and isopods at 1000 mg CeO2 ENP/Kg soil. Earthworm exposures for CeO2 ENPs were undertaken also showing low toxicity at concentrations up to 10000 mg CeO2 ENP/Kg soil.
UGOT studied the effects of silver and zinc nanoparticles in single species assays with selected species of algae and bacteria and a community-level bioassay with natural biofilms (so-called periphyton). An extended particle range outside the NanoFATE commercial ENPs were investigated in order to cover differing in coating, size and presence of stabilizers. Specific information on the physiological status (vitality) of the exposed organisms were determined by using fluorescence dyes and flow cytometry. Additionally, the formation of reactive oxygen species (ROS) as a consequence of exposure to nanoparticles was investigated. Using the MOPS buffer increased the sensitivity of the assay, most likely due to increased concentrations of free silver ions. Natural microbial communities (algae and bacteria) were also assessed and the effects on the bacterial part of the community were determined using Ecolog plates, which provides an indication of the ability of the community to degrade a set of 31 different carbon sources and hence data on the ability of the communities to drive carbon cycling under toxicant stress. The carbon utilization pattern also gives an indication of changes in species composition (biodiversity, toxicant-induced succession).
Effects on changes in the biofilm algae community were studied by HPLC-based pigment profiling. Algal growth was determined as chlorophyll α content, species composition and interferences with specific physiological processes were estimated as changes in pigment patterns.
In addition, the following particles were tested by UGOT in the growth inhibition assays with the gram-negative bacterium Pseudomonas putida, performed according to ISO guideline 10712: AG7 (Amepox, Poland), PL-Ag-S10 (Plasmachem AG, Germany), NM-300K (OECD WPMN program, JRC, Ispra, Italy), PELCO® NanoXact (Tedpella, US), Silver colloid (British biocell, UK), AG6 (Nanotrade, Czech Republic). Silver nitrate was used as a reference compound. For each particle type a full concentration-response curve was recorded in at least two independent experiments. Particle size distribution and number were determined in all stock suspensions, using Nanosight nanoparticle tracking analysis (NTA) and transmission electron microscopy (TEM).
Based on protocol adaptations, UAVR conducted chronic assays with Daphnia magna, with the Ag ENP from AMEPOX (3-8 nm), and AgNO3 as ionic comparator, and also compared results from ZnO particles’ exposures (30nm, 80-100nm and 200nm, Nanosun) and ZnCl2. UNIPMN carried out a comparative assessment of the AG7 and AG6 silver ENPs (8 nm, 50 nm size, respectively) and silver nitrate in a range of concentrations from 0.001 to 10 mg/L to the marine mussel Mytilus galloprovinciali. In addition, two sizes of the Nanosun ZnO ENPs (30nm and 80-100nm), ZnO 200nm and ZnCl2 were also tested. Acute (i.e. mortality during post-exposure removal from water to air), and sub-lethal effects on byssus synthesis were evaluated in M. galloprovinciallis as a function of time and concentration within four days exposure.
Objective 3.3. Determination of the ecotoxicology of the selected ENP types in environmentally realistic situations for aquatic and terrestrial systems, including the assessment of interactive effects with physiochemical components of different exposure media and interactions with UV and organic chemicals as potential co-stressors. [NERC, VUA, UOXF.DJ UAVR, UNIPMN, IHPP, CU, UGOT]
The information to derive confounding factors (soil and water chemistry) on the ecotoxicology of ENPs has been gathered from studies on ZnO, Ag and CeO2 ENPs. Within this framework, IHPP studied the interactions of the developed Co doped ZnO ENPs to enable easy detection, and stability, in different test media including the ASTM medium, deionized water and Lufa 2.2 soil extract.
CU, NERC, VU and UAVR carried out large scale tests to study how soil pH affects ZnO ENP toxicity using modified field soils for earthworms, collembola and isopods. Samples were analyzed and characterized for ENP dissolution and behaviour kinetics, and tissue samples from earthworms were prepared and preserved for molecular analysis and ultra histology. The results of all these trials were used to parameterise models developed in WP4 modelling and provide input for WP5.
The toxicity of ZnO nanoparticles to F. candida and P. pruinosus in soils ranging in organic matter content was also evaluated by VUA and UAVR. Four soils ranging in organic matter content from low (1.64 %) to high (16.7%) were used to study the effect on springtail survival and reproduction after 28-d exposure and isopods feeding inhibition and biomass loss in a 14-d exposure. VUA and UAVR carried out chronic toxicity tests with collembolan and isopods to understand the effects of CeO2 ENPs on the toxicity of phenenthrene through possible co-transport.
UNIPMN had undertaken studies on the effects of realistic ENP forms on Mytilus spp in natural waters. Microcosm long term exposures were carried out, where mussels were exposed for 4 weeks in natural seawater to AG7 ENPs (3-8 nm, AMEPOX).
Finally UAVR conducted several assays with D. magna were effects from mixture exposures were analysed. For that combinations of Ag forms (Ag ENP x AgNO3), Zn forms (ZnO ENP x ZnCl2) and also Ag x Zn (both ENPs or ionic) forms were carried out using immobilization and feeding inhibition tests. Conceptual models for mixture toxicity were used to derive additivity patterns or interactions between forms inside the organisms.
Objective 3.4. Identification of the major drivers for the ecotoxicological effects of ENPs, such as the contribution of the ENP itself and the metal ions they liberate, the coating of the ENPs after passage through WWTPs, the dependence on the exposed species and considered exposure route. Compilation of the experimental information and existing information from the literature and other R&D projects. Knowledge transfer to WP 4 (bioavailability studies) and WP6 (risk assessment). [NERC, VUA, UOXF.DJ UAVR, UNIPMN, CU, UGOT]
From the ecotoxicity testing carried out during the project, several issues were identified as drivers for effects induced by ENPs, regarding both aquatic or soil systems. The behaviour of nanoparticles govern their toxicity and was shown to change greatly with time at scales often much shorter or longer than standard exposures. In time, ENPs change their characteristics, which have to be described in order to understand their deleterious effects, understanding simultaneously if effects are nanoparticle-related or not.
One example was the study on the exposure of D. magna to different sized ZnO ENPs (Nanotrade). Their initial size was confirmed to have no effect on ENPs toxicity mainly because after 48h they ended up all having agglomerated to very similar size and no longer in the nano-range (Lopes et al. 2014). Therefore the first hypothesis raised that initial size could influence toxicity showed not to be easy to test nor entirely relevant in realistic exposure terms, but their size upon initial agglomeration in the exposure cannot be excluded from discussion of important parameters. In addition, ZnO particle size showed no influence on soil sorption, and little differences were observed in terms of toxicity when F. candida was exposed to ZnO ENPs of different sizes (30nm and 200 nm ZnO, Nanotrade) (Waalewijn-Kool et al. 2013).
Dissolution rate proved a characteristic that has to be taken into consideration when looking at ENPs toxicity, again making exposure time a key factor. With time, some metal ENPs will release ions and those ions will also be more or less responsible for their toxicity. This was observed in the toxicity tests with D. Magna, but also with algae; upon Ag ENPs (AMEPOX) exposures, algae toxicity was mainly due to the internalization of dissolved or ionic Ag (Ribeiro et al. 2014).
The most sensitive organisms tested were the water microbial communities. Results from the standard tests with algae and bacteria revealed that the prokaryotic organisms reacted more sensitive than the eukaryotic. The bacterium P. putida reacted most sensitive with EC50 values of 0.16 µg/L for AgNO3, resp. between 0.25 and 13.5 μg/L for AgNPs. However, no particle specific effects could be detected indicating that the silver ions are the toxicity causing factor; neither specific coatings nor certain sizes could be linked to increasing or decreasing toxicity for all tested particles and organisms and no clear toxicity patterns could be detected in terms of species specific attributes. The characterization of particle behavior (and for a certain proportion for the total and dissolved silver content) in the test media during the exposures proved to be impossible for most exposure situations due to the high sensitivities of the test organisms.
For soils, the dissolution process and the way ENPs interact with soil pore water was found to be of major relevance particularly regarding collembolan and earthworms. This was mainly justified by their exposure routes, where collembolans are exposed through the soil pore water and earthworms can be exposed both dermally and through soil ingestion. In exposures of earthworms to Zn (using ZnO ENPs, ZnCl2 and micron scale ZnO) in soils of varying pH, it was observed that pH highly influenced Zn dissolution (Heggelund et al., 2013). Indeed, the WP3 experiments has shown pH to play a major role in bioavailability, as for ENP exposures the competition between H+ and metallic ions is not as straightforward as it was for the well-studied case of ionic metal exposures (see WP4 below for details on bioavailability considerations). This information was crucial to determine the relevant bioavailable fraction and which were the factors affecting bioavailability (WP4), but also providing relevant input data needed for risk assessment procedures (WP6).
WP4. ENP bioavailability – relations between soil and water chemistry and particle properties.
Objective 4.1. Review and utilise existing data-holdings and where necessary primary research articles that contain information on ENP, and media properties and their relationship with acute and chronic toxicity in aquatic and terrestrial species. [NERC, VUA, UAVR, UNIPMN, CU, UGOT]
This objective has been achieved through the collation, data-basing and mining of existing and newly available data. WP4 partners (NERC, VUA, UAVR, UNIPMN, CU, UGOT) jointly developed and contributed to creating a database, bringing together relevant literature data. IHPP and UOXF.DJ provided relevant information on ENP properties that assisted in properly interpreting the available literature data. The database, supported the work within WP3 and WP4, and served as the basis for collating the toxicity data generated from NanoFATE and the partners other nanosafety projects and for constructing species-sensitivity distributions for use within spatial risk assessment models developed in WP6.
Results from this exercise also provided basic input regarding the evaluation of confounding factors to be included in assessing the bioavailability and toxicity of ENPs to the different test organisms in soil and water. Organic matter content and pH were identified as major factors affecting ENP fate and bioavailability in both soil and water. For aquatic media, also ionic strength and salinity are important factors. These findings lead to the prioritisation and planning of experimental work under a realistic range of environmental conditions and modified media conditions within WP3 and WP4, respectively.
The critical review of the literature performed to identify factors affecting ENP bioavailability was further elaborated by UAVR, VUA and NERC into a paper focusing on soil. A similar manuscript dealing with aqueous exposure media is under development by NERC.
Objective 4.2. Establish and model how environment physicochemical properties govern ENP stability, soil–solution partitioning and transformation (e.g. dissolution) in natural waters and soils utilising time dependent kinetic modelling; chemical speciation modelling using established tools such as the Windermere Humic Acid Model and adapted version of the Biotic Ligand Model that include both particle dependent and dissolved metal ion toxicity. [NERC, VUA, UAVR, CU, UGOT]
This objective (and objectives 4.3 & 4.4) was achieved through completion of detailed experiments on how ENP and media properties interact to affect ENP presentation in environmental media and their availability for uptake and integrated assessment and modelling of ENP uptake and toxicity across soils and waters of varying chemical properties.
In close collaboration with the other WP4 partners and WP1, IHPP developed cobalt-tagged ZnO ENPs to enable proper assessment of ENP fate in test systems as well as possible uptake in test organisms. IHPP also determined the stability of the cobalt-tagged ZnO ENPs and other ZnO ENP and non-nano forms in different aquatic test media used for both aquatic and soil organisms. The results supported the experimental work preformed to fulfil Task 4.2 (Detailed experiments on how ENP and media properties interact in time to affect ENP presentation in environmental media and their availability for uptake), and have been reported in Deliverable 4.2.
NERC assessed and modelled the interaction of ENPs with media properties in aquatic and soil test media. In tests using the NanoFATE developed simulated soil pore water, NERC assessed how the changes in the concentrations of DOC (fulvic acids) and major cations (H+, Ca2+, Na+, Mg2+) affected ENP availability. These data have indicated the extent to which competition effects need to be accounted for during the development of biotic ligand models for ionic silver and silver ENPs. In terrestrial tests, NERC, VUA and UAVR studied the effect of soil pH on the comparative uptake and toxicity of ionic Zn, ZnO ENP and non-nano sized ZnO by earthworms, collembolans and isopods. These experiments indicated that ZnO dissolution was faster at low soil pH values, but for the same total soil concentration the soil porewater concentrations of Zn was much lower in ZnO dosed soils that in ZnCl2 dosed ones. For all species the pore water based toxicity was higher for ZnO then ZnCl2 based exposures, suggestion that not the full effect could be explained by the dissolved Zn and that the particulate ZnO contributed to toxicity. The normal “protective effects from protons” that reduces toxicity of porewater Zn with decreasing pH for ionic zinc was much less evident for the porewater in the ENP based exposures. For earthworms more zinc was accumulated in worms exposed to ZnO ENPs than those exposed to ionic zinc at the same effect levels, suggestion that uptake routes and internal handling differ between the two Zn forms. .
To assess the timescale of ENP stability dynamics and its consequences for ENP toxicity in soil, VUA performed ageing studies to determine long-term fate of triethoxyoctylsilane-coated and non-coated ZnO ENPs, non-nano ZnO and ionic Zn in soil and possible changes in toxicity to Collembola. Coated ZnO ENPs were more toxic than non-coated ones, and only after aging the soils for one year of equilibration was the toxicity of coated ZnO ENPs reduced to the initial (time zero) level of non-coated particles, suggesting the degradation of the coating has a 1 year timescale in soils. Both ZnO ENPs and non-nano ZnO slowly released Zn ions, with Zn ion release being highest at intermediate exposure levels and reduced at high exposure levels. This probably is due to stronger agglomeration at increased soil concentrations, which was also pH dependent. In spite of increasing Zn concentrations in pore water, ageing reduced the toxicity of ZnO ENPs, but also of non-nano ZnO and of ZnCl2, which could be explained from different effects on soil pH resulting from the addition of the different Zn forms. This study showed that soil pH is a major factor affecting ZnO ENP fate, availability and toxicity to the test organisms, and that a coating although reducing dissolution of ZnO ENPs may leave them more toxic. A paper describing this work was included in Deliverable 4.3. A similar longterm soils aging study was undertaken for Ag ENP toxicity to earthworms at NERC, showing that again the dynamics of the ENP behaviours are at the timescales of a year. Here ENP toxicity based on total soils Ag was initially much lower than for the ionic based exposure. However, while the toxicity of the ionic exposure reduces over time as the soils were aged, the ENP toxicity increased (presumably as the ENPs dissolved) and after a year the ENP toxicity had reached the level of toxicity seen for the freshly spiked ionic exposures at the start of the aging experiment. Again suggesting that ENP dynamics in soils are slow and that results from short standard tests (even 56day earthworm tests) may give a slightly false sense that ENP forms of metals are far less toxic than the ionic forms, when long term they may become as toxic.
Objective 4.3. Undertake specifically designed studies to investigate interactive effects among factors (ENP and media characteristics) that influence availability and uptake of ENPs in key species. [NERC, VUA, UAVR, UNIPMN, CU, UGOT]
In close collaboration, NERC, UAVR and VUA performed a study on the toxicity and bioaccumulation of ZnO ENPs in relation to soil properties. The design properties and media modification parameters were developed under WP4 based on the critical review of the current literature and the exposures run under WP3. To achieve the full range of natural soil pH (4-8) in a soil of otherwise closely matched properties NERC collected an acidic field soil (pH 4) and adjusted it to different pH levels. The zeta potential of ZnO ENPs in soil extracts related to pH in opposite direction to that seen in deionized water, showing that fate of ENPs in soil cannot be predicted from that in pure water. All three soils have been tested for toxicity of ZnO ENPs using earthworms (CU & NERC), Collembola (VUA) and isopods (UAVR). In addition, VUA also determined the effects of other soil properties on ZnO ENP toxicity to Collembola. Toxicity of ZnO ENPs was always higher at lower soil pH, but at higher pH levels results were not consistent across species: for earthworms and springtails toxicity decreased with increasing pH while for isopods toxicity was lowest at intermediate pH levels. Similarly NERC included other natural soils that covered the normal range in soil organic matter too, but within normal soil ranges the effect of pH is much stronger than that of organic matter content. Results of these studies have been detailed in four published research manuscripts resulting from this work (Tourinho et al., 2012; Tourinho et al., 2013; Waalewijn-Kool et al., 2013a; Waalewijn-Kool et al., 2013b; Heggelund et al., 2014).
For the aquatic compartment, tests on the toxicity of Ag ENPs to Daphnia magna were carried out in ASTM media with different pH values (5.3 7.8 9.0). In addition, the effect of hardness on ENP toxicity was determined. This study confirmed that also for the aquatic environment, pH is an important factor determining toxicity of Ag ENPs.
Using the snail Physa acuta, UAVR studied Ag ENP toxicity in test media with different composition. Two different media were used, taking into account the knowledge that silver ions bind to chlorides reducing the toxicity of silver: Artificial Pond Water (APW) and APW-adapted without CaCl2. This study showed the importance of complexation between Ag and Cl, which leads to the formation of AgCl colloids and lowers toxicity of silver.
UNIPMN determined the toxicity to marine organisms (molluscs) of different Ag ENPs as a function of their stability in water and solubilisation. Results suggest that the toxicity of Ag ENPs is due not only to the release of silver ions, but also due to the behaviour of ENPs in the media, which can vary widely with concentration and media properties. Results are reported in a manuscript that is included in Deliverable 4.4.
Objective 4.4: Develop kinetic models that describe environmental effects on particle dissolution rates and incorporate both free ion and particle associated toxicity to address the relationship between the relative contribution of ENP- and free metal ion- derived toxicity for exposed species [NERC, VUA, UOXF.DJ UAVR, UNIPMN, IHPP, CU, UGOT]
NERC integrated results from the literature and those obtained within NanoFATE to explore through speciation modelling the relative contributions of ENP and free metal ions to toxicity for the exposed species in aquatic media. NERC also provided other partners (VUA and UAVR) with calculations supporting assessments of the relative contribution of free metal ions to the toxicity of ZnO ENPs in different soils. Similarly, NERC supported UAVR and UNIPMN with speciation calculations to interpret ENP toxicity in aquatic media. These activities support the experimental work performed to fulfil objectives 4.2 & 4.3.
Deliverable 4.4 was finalized to include research manuscripts on the effect of soil pH on the toxicity of ZnO ENPs to earthworms (NERC), Collembola (VUA) and isopods (UAVR). In addition, the deliverable for this work includes submitted manuscripts on the effect of soil properties on ZnO ENP toxicity to Collembola (VUA) and the effects of media properties on the toxicity of Ag ENPs to molluscs (UNIMPN).
During a workshop held in Wallingford on 26th September 2013, an inventory was made of all data on the influence of media properties on ENP bioavailability generated within NanoFATE. The workshop identified some data gaps and resulted in agreements on a limited number of further measurements to obtain all data needed for the development of such modelling framework. Together with the outcome of the workshop and the results of the additional measurements, the work described in Deliverable 4.4 provides the starting point for two review papers, describing and applying a modelling framework for predicting metal-based ENP bioavailability in soils and waters currently being written up by NERC.
Overall the scientific findings of WP4 (based on the WP3 experiments) that may have specific consequences for (long-term) risk assessment of ENPs in the environment are:
- Soil pH is a major factor governing metal-based ENP fate and effects in soils, and the effect of pH seems to overrule that of other factors like organic matter content.
- For waters in addition to pH, ionic strength and/or salinity are major factors affecting ENP fate and effects to aquatic organisms.
- The presence of a coating inhibit their dissolution rate, but may increase toxicity of metal-based ENPs .
- While aquatic ENP behaviour dynamics are fast and operate on shorter timescales tan the standard experiments this is NOT true for soils, where the fate dynamics may have timescales of many months of years and will not be covered by standard test protocols.
- Long-term equilibration of metal-based ENPs in soils may lead to a slow dissolution of metal ions over time, which may still continue after 1 year and is concentration and pH dependent.
- Long-term equilibration in soils may decreased toxicity of ZnO ENPs, but did cause and increased toxicity of Ag ENPs as was shown in a NanoFATE collaborative study performed by NERC.
WP5 - ENP toxicokinetics and toxicodynamics.
Objective 5.1: Identify the internal fate of trackable particles in exposed organisms by combining time series exposures with detailed tissue and cellular morphology using qualitative and quantitative particle detection methods. [UNIPMN, CU, UOXF.DJ IHPP].
This objective was initially addressed through a comprehensive review of fixation and visualization techniques along with methodologies and guidelines for determining trace metals in biological tissues with particular emphasis on silver. Completion of this output by CU and UNIPM and IHPP provided a basis from which the consortium has been able to develop the research activities to approach further objectives within this work package. For research purposes, and to make sure complex techniques are only applied when we have evidence from more routine methods that the atomic elements or high electron density signatures for the ENPs are present. We proposed a tiered approach to the tracking of ENPs in tissues by a logical order of moving a sample in an incremental manner through the various analytical and characterization methods. Regarding the ability of tracking nano- objects within cells and tissues, much has been done within the NanoFATE framework. Studies have shown that ENPs undergo a range of physical and chemical transformations in the environment to the extent that exposures to pristine materials will occur only rarely in nature. Methods to track assimilation and internal distributions must, therefore, be capable of detecting these modified forms. Techniques able to provide a strong chemical signature, such as X-ray absorption spectroscopy and EM, allowed tracking ENPs into animal tissues. A striking evidence of that was provided by quantitative modelling of X-ray Absorption Near Edge Spectroscopy (XANES) profiles obtained in ZnO ENP exposed earthworms obtained at the Diamond light source synchrotron in the UK (CU, UOXF.DJ NERC). This analysis indicated that the Zn atomic centres within the matrix of the focal deposits imaged in epidermis and nephridium of earthworms are associated with oxide, a chemical phase distinguishable from the phosphate-associated Zn pool in the chloragogenous tissue. Similar indications of differential metal speciation for Ag in tissues of both woodlice and earthworms depending on whether they were exposed to Ag ENPs or ions, where also found when doing some technically XANES work were undertaken at Diamond to explore if working at the lower energy Ag k-edge was possible. Preliminary analysis of further XANES data of zinc speciation in nematodes exposed to ZnO or artificially aged (phosphatised) ZnO indicate that the main form within nematode tissues was a zinc phosphate regardless of exposure form. All-in-all suggesting that organisms of different size and physiology may be affected by and experience ENP exposure in different ways in terms of internalised metal species. Extensive EM work was carried out also in other organisms such as mussel tissues, document that here uptake involves dissolution and reconstitution of metal (Ag) environmental/biogenic nanoparticle into gills (UNIPMN, UOXF.DJ CU).
Objective 5.2: Investigate the link between internalisation and toxicological effects in time series exposures, and use model prediction of toxicokinetic parameters to study the internal tissue distribution and form of ENPs (using analytical electron microscopy) to attribute causal effects to intact ENPs and dissolved free metal ions. [VUA, NERC, UGOT, CU, UOXF.DJ]
This objective bridges Tasks 5.1 “Modelling and measurement of ENP toxicodynamics and toxicokinetics” and 5.2 “Comparative mechanisms of action of dissolved, bulk, and nanoparticulate forms” by linking the temporal effects of ENPs and their eventual biological fate with the mechanisms of toxicity observed. The findings of the kinetic and bioaccumulation studies of NanoFATE has been summarise in D5.5. Bioconcentration Factor (BCF) values, uptake (k1) and excretion/release (k2) constants were determined according to various mathematical approach including the one-compartment modelling. These studies were carried out on several and distinct biological systems that differ from each other for both structural/morphological features and/or their ecological niche: unicellular algae, springtails, daphnids, terrestrial detritivorous (Eisenia spp) and marine filter feeding bivalves (Mytilus spp). The emerging picture has been rather heterogeneous depending on either model species and/or chemicals. In springtails, higher ionic BCFs has been explained by greater k2 values for silver ENP (5 nm Amepox, Ag7) exposures (VUA). In algae treated with same chemicals, conversely, greater BCF values for ions appeared to depend on k1, whilst k2 did not differ significantly (UAVR). A higher BCF value for the ionic form has consistently been observed in earthworms (Eisenia foetida) exposed to cerium salt or different uncoated CeO2 ENPs at 256 mg/kg in Lufa 2.2 soil during standard 28 day exposures (NERC). Finally, in bivalve molluscs, taking carefully into account the actual silver dose in seawater column, BCF values of ionic silver and two different Ag ENP differing in size (5 and 50 nm) types were very similar (UNIPMN).
The Dynamic Energy Budget (DEBtox) model was considered to allow further comparisons of toxicokinetics and toxicodynamics between ENPs and ions (D5.3). The experiments with C. elegans (NERC and VUA) and D. magna (UAVR and VUA) were analysed with a DEBtox model that assumes chemicals have to be taken up to cause effects. The kinetics of uptake determines how fast chemical concentrations in the body of the test organism will develop with time at different exposure concentrations. By following development of toxicity with time upon exposure to a test chemical, DEBtox is able to test assumptions on its mode of action. Experiments with M. galloprovincialis did not allow application of the DEBtox model. These experiments included observation of ecophysiological parameters upon exposure to Ag ENPs and were analysed with a normal DEB model. This enabled making predictions of the long-term effects of Ag ENPs on these test organisms.
For the toxicity of ZnO ENPs (30 nm and 80-100 nm) and Ag ENPs (5 nm, Amepox Ag7) to D. magna, DEBtox analysis showed increased metabolic costs for growth and reproduction to be the best fitting mode of action. Non-nano ZnO had a different mode of action compared to ZnO ENPs, leading to increased metabolic costs for daphnid growth. Daphnids were more tolerant to non-nano ZnO, which could be attributed to the higher elimination rate on non-nano ZnO compared to ZnO ENPs and ZnCl2. Ag ENPs (5-8 nm) and AgNO3 had the same metabolic mode of action in D. magna (increased metabolic costs for growth and reproduction). Elimination rates constants for Ag obtained using DEBtox were higher than values from uptake-elimination kinetics experiments determining Ag levels in the daphnids, suggesting that at least part of the Ag taken up is sequestered and not involved in causing toxicity to the animals. The DEBtox analysis of data on the toxicity of Ag ENPs to C. elegans concluded that “hazard to the embryo” was the most likely mode of action, which links well with the much higher sensitivities found for the juvenile lifestages in the full lifecycle exposures underpinning the DEBtox analysis.
DEB analysis of M. galloprovincialis data showed that ecophysiological effects on respiration, food assimilation efficiency and growth would cause effects upon life-long exposure with No Observed Effect Concentrations (NOECs) for ionic Ag and Ag ENPs of 0.2 and 2 μg/L, respectively (corresponding to about 100 ng/L/h in both cases when the actual dose was derived by means of a kinetic model).
Objective 5.3: Investigate the suitability of existing biomarkers for major modes of action known to be linked to ENP toxicity (e.g. reactive oxygen species (ROS) generation, lysosomal responses) for identification of ENP exposure and early-stage effects. [UNIPMN, NERC, VUA, CU]
This objective has been achieved within Task 5.2 by the completion of deliverable D5.4. UNIPMN carried out a comprehensive study to assess a role for oxyl radicals in Ag ENP toxicity on marine mussels. These results showed that in the digestive gland low doses of 5 nm silver nanoparticles (Amepox, Ag7) (as low as 10 ng/L/h for 28 days) are able to modulate relative mRNA abundances of antioxidant responsive genes such as Super Oxide Dismutase, Glutathione Peroxidase and Glutathione Transferase and to inhibit catalase enzymatic activity. Clear effects of an oxidative stress syndrome, however, were evident only at much higher ENPs exposure levels (1000 ng/L/h) for which lipid peroxidation of biological membrane occurred along with a drop in bioenergetics performances as demonstrated by a scope for growth assessment.
UAVR and NERC tested the effects of zinc oxide nanoparticles. In isopods zinc nanoparticles did not elicit a clear molecular/biochemical antioxidant response. In isopods exposed to 30 nm ZnO (Synthetic surfactant-free P99/30 nm NanoSun), at 500 and 750 mg/kg lipid peroxidation products were even lower than in control conditions, confirming the antioxidant properties of zinc. A set of biochemical biomarkers (GST and CAT activity, along with lipid peroxidation byproduct accumulation) was studied in the three life-stages of Porcellionides pruinosus (juveniles, pre- adults and adults) but no zinc effect in either ionic or nano form was observed (UAVR).
NERC investigated the roles of metal chelating proteins (metallotioneins and phytochelatins) in protection against ZnO nanoparticle (Synthetic surfactant-free P99/30 nm NanoSun) effects using a metal sensitive triple knockout (mtl-1;mtl-2;pcs-1(zs2)) nematode strain. The results suggested that these metalloproteins are instrumental in the protection against oxyl radical formation and further cytotoxic damage. Moreover, ZnO ENP exposure was shown to decrease growth and development, reproductive capacity and lifespan, effects that were amplified in a triple knockout strain.
Objective 5.4: Utilise a toxicogenomic approach to investigate novel modes of action associated with ENP exposure when compared to effects attributable to the free metal ions in samples from organisms exposed at ion levels close to concentrations known to cause life-history effects. [UNIPMN, CU, VUA]
The project aimed to evaluate toxicogenomic responses to ENPs and related ionic forms by investigating the global transcript responses in three target organisms, the nematode Caenorhabditis elegans, the filter-feeding marine mussel Mytilus galloprovincialis and the epigeic earthworm Eisenia fetida. These activities were informed by WP3 (ENP ecotoxicology) to establish exposure conditions, but also required the creation of appropriate data structures that will allow the consortium to link; i) the ENP physical/chemical parameters, ii) the routes/time of exposure, and iii) observed life-history responses to the impact observed within the global transcriptome.
UNIPMN generated normalized cDNA libraries and obtained about 90 millions of short reads (100 paired ends, 9 giga bases in total) by means of RNA sequencing on an Illumina HiSeq platform. This allowed the construction of a 15 K high-resolution oligonucleotide array that has been used for transcriptomic studies in Mytilus spp exposed to either 5 nm Ag ENPs (Ag7, Amepox) or ionic silver. Results obtained from short-term exposures at a nominal dose of 100 µg/L showed that apart from a few common features (hsp70s, proton transporters, peptidases and ubiquitine-related genes), then, distinct biological processes and cellular components occurred for the effects of ions compared to effects from ENPs. In both gills and the digestive gland of ENP exposed specimens, microarray analysis strongly suggested the involvement and impact of the endocytic pathway. Membrane docking, secretion, lysosomes, exocyst were over-represented GO terms in ENPs samples. Ionic silver appeared to evoke even more deleterious processes such as DNA repair and oxidative stress (the latter in gills), however, 2D proteomic analysis evidenced a higher rate of oxidized –carbonylated- proteins in ENP treated specimens. The systems toxicology approach, thus, suggested that oxyl radicals represent a common toxicity initiator (or mediator), but different Adverse Outcome Pathways are occurring for the effects of the two physical form of silver tested.
Very similar results to those seen in mussels were obtained in nematodes for the effects of Ag7 ENP. NERC with an award from FP7 Quality Nano, investigated transcriptomic profiles using high-resolution arrays (Agilent). These results again showed few common features between ionic and ENP silver exposed worms. For nematodes exposed to ionic or ENP forms of silver genes associated with neuropeptide signaling pathways, the endocytic pathway and metalo-peptidase activity. NERC, moreover, investigated the role of coating and size finding the distinct molecular traits above in different ways, with size differentially affecting genes from gene ontologies integral to membrane function, and charge affecting gene ontologies relating to ion binding.
Finally, RNA sequencing has been used to carry out a comprehensive transcriptomic survey in the earthworm Eisenia foetida exposed to either silver ions (49 mg/Kg, 2.2. Lufa soil) or PVP coated silver nanoparticles (456 mg/kg) representing equitoxic conditions (EC50 for reproduction, 28 days). This work required getting prior information on earthworm genotypes, sequencing, assembly and mapping of millions of short DNA reads. These results showed a consistent overlap between differentially expressed genes obtained for the effects of Ag ENP or ionic silver. Over-represented features (Gene Ontology terms) associated to the common group were represented by ribosome, adenylate cyclase, sugar and aminoacid metabolism. However, specific processes associated to microtubules, projection of membrane structures, basolateral plasma membrane were identified when a direct statistical comparison between Ag ENP and Ag gene associated ions GO terms was undertaken.
These results does suggest that early and low dose effects from ENP exposure are different from those seen during exposure to metal ion based exposures, with a broad picture across the species of membranes, cell structure and metal trafficking.
WP6: Integrated environmental risk assessment
Objective 6.1: Conduct literature review and a stakeholder consultation to define current and potential future uses of ENPs and determine the driving parameters (including associated uncertainties) that define plausible future release scenarios. Collate consumption/discharge information and fate information (WP2) for use in exposure modelling. [F+B, NERC, NT, AXME, SYMLOG]
This objective was led by F+B with the aim to obtain information on the size of the ENP market and assess market developments. Considerable effort was invested for communicating the layout, content and dissemination strategy of the questionnaire with stakeholders from industry, with particular help from NT and AXME being received. An important stage was to compile the market information acquired from the questionnaires sent to more than 500 companies with the collaboration of several industry bodies, into an overview of market consumption of ENPs. However, this overview could not be completed as originally intended, due to the extremely poor questionnaire response from industry. Thus, in order to achieve its objectives, the study had to focus on reviewing the available evidence, by compiling and evaluating production volumes and modelled environmental concentrations from the scientific and grey literature for the initial exposure predictions and risk maps. The topic of consumption and discharge was re-examined by NERC and F&B towards the end of the NanoFATE project and it was decided to rely on the studies carried out at ETH Switzerland (Piccinno et al., 2012; Sun et al., 2014). The updated information and was used in the revised exposure and risk maps. The modelling assumed that for Europe 32 tons of nano Ag and 1580 tons of nano ZnO were consumed per year. The estimates for nano CeO2 discharge and soil accumulation focused on use in fuel additives in the UK and was reported in D2.2 (Johnson and Park, 2012). For soils, the highest predicted contamination level was 0.016 mg/kg within 20m of a road following seven years of continuous deposition. This value would represent 0.027% of reported natural background cerium. River water contamination considered direct aerial deposition and indirect contamination via runoff in the water and entrained soil sediment, with the highest level of 0.02 ng/L predicted. The highest predicted water concentration of 300 ng/L was associated with water draining from a road surface, assuming a restricted deposition spread. These predictions are well below most toxicological levels of concern.
Objective 6.2: Conduct initial estimates of the concentrations of nanoparticles in different environmental compartments (air, sludge/soil, water and sediments) using the standard EU EUSES modelling approach based on the scenarios developed in Objective 1.[NERC, NT]
This objective was completed by NERC with help from NT. The NanoFATE scenario for airborne entry of commercial use ENPs into the environment was CeO2 used as fuel additives, where possible resulting exposure levels in soils and surface waters were found to be low (D2.2 published in Johnson and Park, 2012). The likely distribution of ZnO and Ag concentrations in European surface waters based on the assumptions that the total consumption of these ENP materials are used equally by all EU citizens and disposed of “down the drain”. The actual geographical modelling was undertaken using the GWAVA (Global Water AVailability Assessment) model (see further below), mapping of WWTP emission points across the EU and knowledge from WP2 on WWTP retention efficiencies for ENPs. For soils environmental release estimation represented a conservative tier 1 approach to predicting ENPs in the soil compartment either from aerial deposition and/or the deposition of sewage sludge containing bound particles. As the majority of ENPs which go down the drain are considered to associate with sludge this is perhaps the most important route, by weight, through which ENPs are released to the natural environment. This deliverable meets the requirements of the Work Package 6 objectives 6.2 and 6.4. These initial estimates informed the designs of realistic ecotoxicology studies in WP3 and WP4. In later revisions of the risk mapping the original soil estimate concentration maps were refined by highlighting agricultural land likely to receive sewage sludge. Research into national differences in sewage sludge disposal also informed these soil concentration maps.
Objective 6.3: Generate robust Predicted No Effect Concentrations (PNECs) by integration of the experimental hazard data from WP3, the bioavailability data from WP4 and ecotoxicology data from the literature with the exposure assessments to develop a sound understanding of the environmental hazards of ENPs and how environmental media (water and soil) affect this. [VUA, UAVR, NERC, UC, UNIPMN]
This was a separate and distinct task from the chemical risk assessment work in WP6. This objective was fulfilled firstly through a report on phylogenetic and trait based analysis of effects across species and the range of ENPs used including a discussion on applicability of the species sensitivity distribution (SSD) approach. Secondly, through the generation of SSD distributions; this included ZnO and Ag in both dissolved metal ion and ENP forms, in both water and soil. The value of applying either assessment (safety) factor based considerations from the ECHA guidance document, or the SSD approach to the available species toxicity information for obtaining appropriate predicted no effect concentrations (PNECs) was reviewed. It was considered that using SSDs was the most robust approach and sufficient data existed (with the exception of nano CeO2) for this to be attempted. However, it was acknowledged that reviewing nanoparticle literature remains problematic due to problems of the characterisation and measurement of the nanoparticle exposures being inconsistent. SSDs were prepared for nano Ag, ionic Ag, nano ZnO and ionic Zn based on combining the available literature with NanoFATE’s own data. In both cases the most sensitive species in water were bacteria and planktivores. The evidence from one specific case of feeding inhibition of in Daphnia indicated that nano ZnO was slightly more toxic (weight for weight) than the ionic species in water (Lopes et al., 2014). For Ag there was only one case of ENP toxicity exceeding (on a weight for weight basis) that of ionic Ag, and again in water, namely for the bacterial part of the microbial freshwater communities (biofilms), based on the total biomass.. In all other cases the ion form was more toxic weight for weight than the nano form.
Objective 6.4: Incorporate information on PNECs into multi-media (EUSES) and GIS based models for soil (using ADMS) and receiving waters (using LF2000-WQX) to allow risk visualisation analyses for using generic and taxa specific PNEC and other endpoint data as appropriate. [F+B, NERC].
This objective led by NERC produced colour-coded maps of European catchments available for all partners, showing predicted river water ENP concentrations. These were prepared using the (Global Water AVailability Assessment) GWAVA model. Due to scientific developments it was considered that GWAVA was more suitable for the project as it provided pan-European predictions as opposed to the Low Flows 2000 model (LF2000-WQX), which is UK-based only. Modelling efforts were refined, partly through the findings of WP 2, on removal of ENPs in sewage. In the expected scenario removal in sewage treatment of 93% nano Ag and 85% nano ZnO was selected and in the best case scenario 99% nano Ag and 88% nano ZnO removal from effluent to sludge. The final modelling included the fate process of loss of particles from the water column by settling within the river. Initial methods were developed allowing future mapping to take into account the effects of the EU regional differences in water chemistry, by developing 6 overall EU water classes that could be applied to the EU Water Framework Directive catchments. The overall results showed a small fraction of Europe’s rivers could exceed the NanoFATE generated PNECs.
For soils predicted concentrations in agricultural soils were carried out for nano Ag and ZnO based on the use of sewage sludge as a fertiliser, taking into account country specific policies for land applications of such biosolids. Whilst the predicted values were 2 orders of magnitude below the PNECs for the ionic forms of Ag and ZnO, they could exceed in certain circumstances the nano form PNEC, which had been generated with a 1000-fold assessment factor due to the scarcity of species data.
Using an entirely separate soil prediction method based on soil nitrogen fertilisation requirement related application rates, similar values for the UK were generated. Soil concentrations for nano CeO2 were predicted based on the emission and deposition from vehicle exhausts of fuel additive CeO2. The highest value was considered to represent only 0.027% addition to the natural background ceria.
Objective 6.5 Undertake appraisal of the current and potential future environmental risks of CeO2, ZnO and Ag ENPs, based on the work conducted in WPs 2, 3, 4, 5, including gap analysis relating to the uncertainties in hazard, fate and exposure assessment as well as the use and emission scenario definitions. [NERC, VUA, F+B, UGOT, UAVR, CU].
For the final risk assessment analysis somewhat higher nano Ag and ZnO discharge into the environment was assumed than in the previous analysis. It was encouraging to note that the predicted expected and best case nano Ag WWTP effluent concentrations of 3-19 ng/L was similar to the measured value of 5-6 ng/L in effluent measured by NanoFATE and reported by others (Li et al., 2013; Johnson et al., 2014b). Following the final analysis of SSD the NanoFATE PNECs were at a challenging low level. Thus, whilst most water bodies would not exceed these PNECs the possibility for some exceedance near effluent discharge now existed. Exceedance of soil PNECs were now considered possible for UK soils, but the soil nanoparticle PNECs were generated using a 1000-fold safety factor. Hence more data on the refinement of the PNEC would possibly lower these and re-establish the separation between the PNEC and the predicted environmental concentrations (PECs).
Remaining gaps in our knowledge begin with our continuing ignorance of true nanoparticle consumption and discharge levels in Europe. The detailed knowledge of how ENP transformations affect the fate and behaviour of nanoparticles in sewage, sludge, soil and water remains weak. The ecotoxicity of nanoparticles to different soil organisms is still limited. Further information on the ecotoxicity of weathered nanoparticles (such as sulphide coatings) would be welcome in future. From the fairly comprehensive studies of ecotoxicity to water species the most vulnerable appear to be bacteria and planktivores.
Objective 6.6: Develop appropriate project material to disseminate results and applied methodologies to relevant stakeholders, in order to improve understanding of ENP risks. [NERC, VUA, F+B, UGOT, UAVR, CU, SYMLOG]
This objective was achieved with the publication of risk maps with accompanying guidance published on the NanoFATE website, an effort lead by risk communicators SYMLOG.
As far as we know NanoFATE is the first project to model risk from nanoparticles in Europe in a truely geographic sense. While a first for soils this is still only possible at a basic level due to the lack of geographical linkage between location of biosolid production (for which ENP concentration estimation is possible) and the eventual land application location. In contrast, due to the piped and linked flows, this geographical approach is particularly advantageous in the field of river water concentrations as the temporal and spatial variation in dilution is included. Previous models and risk assessments have used only crude box models to predict soil and river concentrations. For the purposes of risk assessment new low PNECs were used mainly based on a comprehensive assessment by NanoFATE using SSD based approaches generated in the project. The spatio-temporal variability of concentrations in surface waters and soils across Europe was modelled using the model GWAVA including population demographics and sewage effluent discharge point maps. Loading of nano Ag and nano ZnO from sewage to rivers was modelled by accounting for connectivity to sewerage and sewage treatment efficiency. The resulting nano Ag and nano ZnO concentrations in rivers were modelled by considering the effect of dilution, water abstraction, residence time, and particle settling. Temporal variability in particle concentrations as caused by temporal (seasonal) climate variation was simulated using climate data for the 1979-2000 period. Model scenarios also compared the influence of different reported sewage removal rates on potential river concentrations (Johnson et al., 2014a).
References
Cornelis, G., Juergens, M., Tuoriniemi, J., Hassellov, M., 2014. Research report on ZnO and Ag ENP fate in STW: Measurement of nano ZnO and Ag partitioning and removal during primary and secondary sewage treatment stages., NanoFATE Deliverable D2.7 Wallingford, p. 29.
Costantini M.G. 2011. Evaluation of human health risk from cerium added to diesel fuel. Health Effects Institute Communication 9. Boston, MA, USA.Heggelund L.R. Diez-Ortiz, M., Lofts, S., Lahive, E., Jurkschat, K., Wojnarowicz, J., Cedergreen, N., Spurgeon, D., Svendsen, C., 2014. Soil pH effects on the comparative toxicity of dissolved zinc, non-nano and nano ZnO to the earthworm Eisenia fetida. Nanotoxicology 8, 559-572.
Johnson, A.C. Dumont, E., Keller, V., Williams, R.J. 2014a. Report on ENP risk assessment in the different usage scenarios; this will include gap analysis, critical appraisal of available hazard and exposure assessment methodologies and the identification of vulnerable species and environmental compartments, NanoFATE Report - Deliverable 6.7. Natural Environment Research Council, Wallingford, p. 33.
Johnson, A.C. Juergens, M.D. Lawlor, A.J. Cisowska, I., Williams, R.J. 2014b. Particulate and collidal silver in sewage effluent and sludge discharged from British wastewater treatment plants. Chemosphere in press, 19.
Johnson, A.C. Park, B., 2012. Predicting contamination by the fuel additive cerium oxide engineered nanoparticles within the United Kingdom and the associated risks. Environmental Toxicology and Chemistry 31, 2582-2587.
Li, L., Hartmann, G., Doblinger, M., Schuster, M., 2013. Quantification of nanoscale silver particles removal and release from municipal wastewater treatment plants in Germany. Environmental Science & Technology 47, 7317-7323.
Lopes, S., Ribeiro, F., Wojnarowicz, J., Lojkowski, W., Jurkschat, K., Crossley, A., Soares, A.M. Loureiro, S., 2014. Zinc oxide nanoparticles toxicity to Daphnia magna: size-dependent effects and dissolution. Environ Toxicol Chem 33, 190-198.
Piccinno, F., Gottschalk, F., Seeger, S., Nowack, B., 2012. Industrial production quantities and uses of ten engineered nanomaterials in Europe and the world. Journal of Nanoparticle Research 14.
Sun, T.Y. Gottschalk, F., Hungerbuhler, K., Nowack, B., 2014. Comprehensive probabilistic modelling of environmental emissions of engineered nanomaterials. Environmental Pollution 185, 69-76.
Tourinho, P.S. van Gestel, C.A. Lofts, S., Soares, A.M. Loureiro, S., 2013. Influence of soil pH on the toxicity of zinc oxide nanoparticles to the terrestrial isopod Porcellionides pruinosus. Environ Toxicol Chem 32, 2808-2815.
Tourinho, P.S. van Gestel, C.A. Lofts, S., Svendsen, C., Soares, A.M. Loureiro, S., 2012. Metal-based nanoparticles in soil: fate, behavior, and effects on soil invertebrates. Environ Toxicol Chem 31, 1679-1692.
Waalewijn-Kool, P.L. Diez Ortiz, M., van Straalen, N.M. van Gestel, C.A. 2013a. Sorption, dissolution and pH determine the long-term equilibration and toxicity of coated and uncoated ZnO nanoparticles in soil. Environ Pollut 178, 59-64.
Waalewijn-Kool, P.L. Ortiz, M.D. Lofts, S., van Gestel, C.A. 2013b. The effect of pH on the toxicity of zinc oxide nanoparticles to Folsomia candida in amended field soil. Environ Toxicol Chem 32, 2349-2355.
Zhang, J., Nazarenko, Y., et al., 2013. Impacts of a Nanosized Ceria Additive on Diesel Engine Emissions of Particulate and Gaseous Pollutants. Environ. Sci. Technol. 47, 13077−13085
Potential Impact:
Potential Impact
WP1: Characterisation and tracking of ENPs during processes involved in fate and toxicity
WP1 Underpinned much of the work of the other work packages within the project, however, it also contributed significantly to commercialisation aspects of the project with new commercial products being developed. As part of their role within the project, NanoTRADE (NT; http://www.nanotrade.cz/) and Amepox Ltd.(AMEPOX; http://www.amepox.com.pl/) were required to develop their products to facilitate their use in environmental fate and toxicity testing. As a result of NT in collaboration with AMEPOX have launched a new product approved for veterinary use. Further more the Insitute of High Presure Physics - Polish Academy of Sciences (IHPP), who were tasked with developing and producing cobalt-doped zinc oxide particles, now have a joint venture with an industrial company for scale up of production. Patents have been filed for materials developed during the project. Further as well as these direct commercialisation opportunities, the involvement in NanoFATE of partners NT, AMEPOX and IHPP has also provided each with valuable information and knowledge of the risk assessment issues for their productsas an contribution to imp[roving their sustainable production and use of ENPs. This information is generally beneficial to SMEs involved in the development and commercialisation of nanotechnology products. As direct outcomes of these activities,NT has initiated a Memorandum in collaboration with other manufacturers for NANOSAFETY in the Czech Republic. IHPP have made recommendations for safety regulations in Poland based on conclusions from results of nano-ZnO strong aggregation in test waters, behaviour in humid conditions and solubility studies. IHPP have contributed to a Book of recommendations for the European Commission for nano-regulations along with the NANOFORCE project (http://www.nanoforceproject.eu/) .
NanoFATE has also led to improvement in knowledge based and quality assured analytical and consultancy services at University of Oxford (UOXF.DJ) and IHPP. This has provided a focus for research to support the safety assessment and sustainable production of nanotechnology through the national centres of excellence. Dedicated training courses have been held at UOXF.DJ to increase the level of understanding in the physical characterisation of materials for participants from outside materials science field and to increase the utilisation of the academic public funded equipment base.
WP2: ENP environmental behaviour and fate modelling
WP2 is responsible for the nanomaterial fate, behaviour and transport studies to enable predictions of environmental concentrations have developed both cutting edge experimental and modelling approaches. In the first deliverable it was identified which processes and properties will be most important to determine experimentally in order to mechanistically parameterize fate processes for modelling. The experimental designs are targeting such key parameters.
The highest novelty in WP2 together with WP6 is the EU-wide spatially distributed emissions, dilutions, and fate processes that are built on GIS-based hydrology, population density and water chemistry models.In the production of these new risk assessment products, Nanofate has gone way beyond the state-of-the-art in terms of emission and fate modelling. The results has been used by regulatory authorities in informal discussions and formalised reporting of the current risk assessment status of ENPs nationally and in Europe (e.g. The Swedish Chemical Inspection and the UK Department for Environment, Food and Rural Affairs (Defra). Representatives from these organisation have also highlighted these results in paper and statement made through their role as have representatives to the OECD working party on manufactured nanomaterials.
WP3 – ENP Ecotoxicology
NanoFATE has contributed significantly to the improvement of existing exposure protocols to test ENPs in both aquatic media and soil. This was attained based on full collaboration amongst partners. One highlight on improved methodology was the development of a Synthetic Soil Pore Water that allows toxicity trials to be carried out with nematodes in liquid culture. This contrast with previous high ionic strength media that resulted in extensive agglomeration. For soil exposure, a homogeneous introduction of ENPs in soil is crucial and was achieved by using a suspension spiking method or by mixing dry ENP powder directly in dry soil followed by adjusting moisture content to the desired level. For food contamination, alder leaves were submerged in stock solutions of nanoparticles and then air dried, prior to be used. These methodologies have already been followed in laboratories in the UK, Ireland, Norway, France, Italy, Portugal Belgium, Germany and the USA, who are adopting the NanoFATE methods for routine testing in support of standardised toxicity testing.
NanoFATE research has shown that existing methodologies for testing chemical effects to aquatic organisms are usually suitable for ENPs, although some modifications may be needed depending on the nanomaterial in study. As an example, guidance for media renewal, i.e. the choice for the period of exposure media renewal, should be based on results from agglomeration/aggregation experiments. These studies are relevant to both the acute and chronic tests.
WP3 ecotoxicity tests were also carry out to study ENPs bioavailability in test media, assessment of ENP uptake and provide material to tissue bank in order to evaluate potential internalization of nanoparticles in WPs 4 and 5.
The work conducted in WP3 has made a major contribution to the ENP toxicity database for Ag, ZnO and CeO2. The data covered multiple taxonomic groups making the data particulay valuable for risk assessment (aquatic: crustaceans, fish, molluscs, fish, algae, bacteria; soil: earthworms, springtails, isopods, plants, bacterial) The overall results from WP3 has, thus, considerably improve data available for nanoparticles’ toxicity, and were used in WP6 along with bibliography data to derive more robust PNEC values, and estimate risk from the ENPs studied. The fact that all studies have been conducted using protocols optimised for testing the toxicity of ENPs means that the generated data is of high quality. There is an acknowledged gap in the literature in relation to the availability of high quality effect studies on the chronic toxicity of ENPs for environmentally relevant receptors. The body of work conducted to date in NanoFATE is certain to be a major contribution to the literature in this area. Indeed both the nanoparticles-specific and ionic form silver ecotoxicity data, generated through NanoFATE, has been incorporated directly into the Chemical Safety Report Dossier for REACH registration for silver.
WP4 - ENP bioavailability – relations between soil and water chemistry and particle properties.
WP4 studies on the bioavailability of ENPs in test media were used for ecotoxicity testing and assessment of ENP uptake in WPs 3 and 5, respectively. Results of WP4 are essential for a proper understanding of the impact of media and particle properties on ENP bioavailability, uptake and effects in soil and water and provided input for spatially specific risk assessment efforts undertaken both in WP6 and outside NanoFATE , including by regulatory agencies.
WP4 partners have contributed to training activities to improve the EU skill base. Results of the project were presented at several meetings, including annual meetings of SETAC and the NanoSafety Cluster meetings. This assisted in communicating the results and their significance for standard testing protocol development and data interpretation to the wider community including industry and regulators. WP4 partners, through the NanoFATE coordinator, contributed to drafting text for the Horizon 2020 programme and the NanoSafety Vision 2015-2020.
At the national level, WP4 partners contributed to developing strategy and status documents regarding the environmental and health assessment of ENPs. Through national coordinators, WP4 partners also contributed to activities of the OECD Working Party for Manufactured Nanomaterials (WPMN) by providing input for meetings and commenting on proposals for modifying existing or developing new test guidelines for testing ENPs. The knowledgebase generated through NanoFATE has been vital to these activities.
WP4 partners have actively contributed to the organization of sessions on the fate and effects of nanomaterials at join scientific/industry/regulator meetings, e.g. the annual meetings of the European branch of the Society of Environmental Toxicology and Chemistry (SETAC). In addition, WP4 partners contributed to activities of the SETAC Advisory Group on Nanomaterials that bring together academic, industry and regulator perspective on environmental nanosafety.
WP4 has also sought to make all results freely available in publications in the peer-reviewed literature (especially open access) and though other written report. For that purpose, results of the critical review of ENP bioavailability (Deliverable 4.1) have been turned into three review papers, of which one has already been published. The continuous update of this data overview, which is a joint effort of WPs 3, 4 and 5, has provided the ecotoxicological database (EC50s, EC10s etc.) for the derivation of Predicted No-Effect Concentrations (PNECs) by WP6, and can be used in collaboration by other EU projects.
The scientific findings of WP4 that may have specific consequences of (long-term) risk assessment of ENPs in the environment are:
- Soil pH is a major factor governing metal-based ENP fate and effects in soils, and the effect of pH seems to overrule that of other factors like organic matter content.
- In addition to pH, ionic strength and/or salinity is a major factor affecting ENP fate and effects to aquatic organisms.
- The presence of a coating increase toxicity of metal-based ENPs but may inhibit their dissolution rate.
- Long-term equilibration of metal-based ENPs soils may lead to a slow dissolution of metal ions over time, which may still continue after 1 year and is concentration and pH dependent.
- Long-term equilibration may decreased toxicity of ZnO ENPs, but did cause and increased toxicity of Ag ENPs as was shown in a NanoFATE related study performed by NERC.
WP5: ENP toxicokinetics and toxicodynamics
WP5 underpinned our understanding of the mechanistic effects of ENPs in biological systems employed by consortium members undertaking ecotoxicology experiments in the entire NanoFate project. This work allowed a better comprehension of modes of action of metal based ENPs to project partners and help to widen the societal acceptance for the newly developed technologies to stakeholders by reducing the uncertainty associated to risk of nanomaterials. WP5 in collaboration with ENNSATOX also developed the architecture and structure of a relational database that allowed data sharing across NanoSafety Cluster projects and the option to submit data to ICPC NanoNet database.
WP5 partners have contributed to the scientific content of booklet “Nanosafety in Europe 2015-2025: Towards Safe and Sustainable Nanomaterials and Nanotechnology Innovations”. WP5 partners have contributed to training activities to improve the EU skill base being part of the NanoFATE PhD workshops in Aveiro, Oxford and Birmingham during different FP7-project meeting/workshops.
WP5 partners have organized an Open Workshop on Mechanistic Toxicology of Nanomaterials (QualityNano NanoFATE & NanoMILE Joint Meeting, Birmingham, UK, March 5th -6th 2014) illustrating in front of a wide international audience recent advances obtained from toxicogenomics and proteomics studies on ENPs. WP5 scientists have also organized national and international meeting sessions such as at SETAC in order to promote mechanistic studies on nanomaterials as a comprehensive tool to reduce risk assessment uncertainty.
Data obtained within the WP5 framework have been presented at several EU-US meetings, discussed with a heterogeneous audience (scientist, policy makers, regulators) in specific EU workshops and roundtables, debated within the EU Nanosafety Cluster meetings. The aforementioned networking activities also promoted the establishment of new scientific collaboration at national and international level.
WP5, beyond the primary NanoFATE scopes, realized innovative research tools for the transcriptomic analysis in invertebrate species such as mussels and earthworms. These tools are commercially exploitable. This includes not only providing information on mode of action to support better risk assessment and risk communication, but also through the development of potential tools for identifying ENP exposure in organisms taken from the field.
WP6 ENP Risk Assessment
As far as we know NanoFate is the first project to model risk from nanoparticles in Europe in a truly geographic and mechanistic sense. This is particularly advantageous in the field of river water concentrations as the temporal and spatial variation in dilution is included. Further the inclusion of parameterised values for specific fate process, both in wastewater treatment and in the river channel itself, provides a unique view of the state of the art that has not been so widely captured in other initiative. Previous models and risk assessments have used only crude box models to predict soil and river concentrations. For the purposes of risk assessment new low PNECs were used mainly based on a comprehensive assessment by NanoFate using species sensitivity distributions generated in the project. The spatio-temporal variability of concentrations in surface waters and soils across Europe was modelled using the model GWAVA usage population and sewage effluent discharge point maps. Loading of nano Ag and nano ZnO from sewage to rivers was modelled by accounting for connectivity to sewerage and sewage treatment efficiency. The resulting nano Ag and nano ZnO concentrations in rivers were modelled by considering the effect of dilution, water abstraction, residence time, and particle settling. Temporal variability in particle concentrations as caused by temporal climate variation was simulated using climate data for the 1979-2000 period. Model scenarios compared the influence of different reported sewage removal rates.
For Europe around 50% of rivers would be predicted to have nano Ag concentrations of 0.0001 ng/L or less and 95% of rivers have 0.1 ng/L or more of nano Ag. Thus, most European rivers would have nano Ag levels 2-orders of magnitude below the PNEC. For Europe around 50% of rivers would be predicted to have nano ZnO concentrations of 1 ng/L or less and 95% of rivers have 500 ng/L or less of nano ZnO. The latter value is only a factor 5 below the PNEC. This analysis puts nano ZnO as being of greater concern to the aquatic environment than nano Ag. This is because of the apparent much higher use of nano ZnO and in this simulation the particularly high water column settlement rate chosen for nano Ag. Although no data are presently available for toxicity to river benthic invertebrates, predictions were also made for nanoparticle accumulation in river bed sediments which indicated that sediment loading of nano Ag of between <0.5 and >0.5 mg/m/y with the higher loading rates being more apparent in northern Europe.
Soil concentrations for agricultural land were predicted for countries across Europe based on 1) national sludge disposal practices and their populations; and, 2) based on soil nitrogen fertilisation recommendations. The soil predictions assumed all the nano Ag and nano ZnO removed from sewage effluent were available for application to agricultural land via sewage sludge. Model scenarios compared different transfer rates to sewage sludge, different proportions of land receiving the national sludge (1 or 57%) and different soil ploughing (mixing) depths of 12 cm and 25 cm. These scenarios largely followed that recommended by EUSES the simple model used in risk assessment for chemicals in the EU. The first approach facilitated soil predictions on a geographic basis something which has not been attempted before to our knowledge.
The highest predicted soil value for nano Ag was 8.8 µg/kg dry weight for agricultural soils whilst the lowest value for nano Ag toxicity was 4.8 mg/kg. However, if a 1000x safety factor is invoked to derive a PNEC of 4.8 µg/kg then we would exceed this. A safety factor of 1000 is common applied when using acute toxicity data to chemical risk assessments. Similarly, the highest nano ZnO soil concentration was 939 µg/kg dry whilst the most conservative value for nano ZnO toxicity was 119 mg/kg. However, if a 1000x safety factor is invoked to derive a PNEC of 120 µg/kg then we would exceed this. This soil prediction approach used here contrasts with some other authors who do not consider the route of sludge being transferred to agriculture.
To summarise, NanoFate research indicates that assuming no change in form, nano Ag and ZnO concentrations in sewage effluent could reach or exceed the PNECs, although this is heavily dependant upon the large safety factor applied. However, given dilution and settling within river bodies most rivers will be well below PNECs. Soil nano Ag and ZnO concentrations could exceed the lowest PNEC particularly in the UK given its disposal policies and limited agricultural land area with respect to its population size.
Engagement with regulatory and standardisation bodies.
NanoFATE partners have played a very active role at international conferences such as the annual SETAC conferences (Europe and North America) which provide an opportunity for academia and government and industry to engage with each other regarding the sustainable use of ENPs. The NanoFATE consortia has participated in the form of chairing sessions, platform presentations, posters, and contributing to the SETAC Advisory Group on Nanomaterials.
The regulatory outreach of the project has been significantly ramped up over the course of the project as the main messages of the project have been more clearly defined. These have responded the following phases of the disseimination strategy: A) better understanding the knowledge demands of regulators; B) identifying individuals and collecting their contact information to ensure accurate targeting and enhance communication flow; C) fostering two-way communication to build a good fit between regulatory needs and science outputs and D) specific dissemination actions. Actions undertaken under this increasing regulatory outreach included advice and direct input by NanoFATE partners to the OECD Working Party on Manufactured Nanomaterials (WPMN). NanoFATE researchers attended expert meetings or provided information on demand from national delegates to: (i) OECD- Testing and Assessment Workshop (Paris, December 2013; (ii) OECD WPMN Steering Group 5/6 "Risk Assessment and Regulatory Programmes" (Paris, Dec. 2013), (iii) Expert meeting, Ecotoxicology of nanomaterials (Berlin, February 2013). More meetings will be attended after the close of the project. A broader regulatory audience was addressed through participation by SYMLOG at the ‘EU 2nd Regulatory Review on nanomaterials’ (30 January 2013, Brussels). During this meeting the big questions, issues and needs by regulators were captured and a flyer highlighting the work of NanoFATE was developed for direct distribution at this workshop. The big questions are presented on the NanoFATE website along with advice notes developed through the NanoFATE project that respond to these issues. The Advice Notes, largely disseminated by NanoFATE and NanoSafety Cluster newsletters, are useful for researchers but also for regulators. They point the way to further research, and flag issues that regulators should attend to when setting assessment guidelines and regulatory practice. The NanoFATE Advice Notes are downloadable two-page documents containing links to the supporting NanoFATE research in deliverables and peer-reviewed papers (which are also accessible via our online library).
NanoFATE has fostered and maintained an active dialogue with representatives from the European Chemicals Agency (ECHA). These discussions have centred around the questions and issues that ECHA have regarding the assessment of nanomaterials under REACH regulations. Broad areas covered have included characterisation and hazard assessment of ENPs as well as in what circumstances ENPs can be grouped for assessment. As a result of these discussions the NanoFATE project coordinator has been invited by ECHA to participate Topical Scientific Workshop on Nanomaterials in October 2014 as a member of the Scientific Committee for this workshop.
Within the 48 months of the project, a wealth of planned and opportunistic dissemination actions have been implemented including regular website updates, an extensive online library of Advice Notes, deliverable summaries, hi-res image bank available to non-project scientists, training workshop reports, newsletters, a social sciences survey, peer-reviewed journal articles, conferences, posters, invited lectures, summer or winter school teaching, training workshops, expert advice and knowledge sharing, and some mass media appearances.
The public website www.nanofate.eu is the hub of all our dissemination actions: ALL NanoFATE information transits from or transits towards the website and is constantly made available for all our audiences in dedicated sections to best fit their needs. The website statistics are therefore an important means of measuring the impact of our dissemination and communication strategy. Google statistics were implemented as of February 2011 to track the traffic and visits on our project website. The figures below show encouraging statistics (very positive compared to other FP7 project websites known to the WPL). The number of unique visitors continued to grow and at mid February 2014 had reached a cumulative total of 51,052 unique visitors (some 70% increase from 36 months) and 894,159 page views (some 35% increase). NanoFATE still continues to recruit new visitors (around 30-35 per month) as well as enjoying a consistently high rate of return visits (twice as many as new visitors).
NanoFATE scientists have played a full and active role in NanoSAFETY Cluster (NSF) activities. Consortia members have participated in Cluster working groups (WGs) concerned with materials characterisiation (WG1), hazard assessment (WG2&3), exposure assessment (WG4), Databasing (WG5), risk assessment (WG6), modelling (WG7) and dissemination (WG7). In addition NanoFATE members have contributed to a wide range of NSF events co-organising 6 workshops/training events, participating in a further 15 meetings, and direct collaborative activities with 11 other NSF sister projects.
Adding to our knowledge base and training the nano-scientists of the future
To date, 21 ISI papers have been published based on NanoFATE results (see publications list for details), with a further 5 papers submitted and 41 planned. All seven RTD work packages have contributed to the preparation of papers. Most publications present the work of more than one work package reflecting the multi-disciplinary approach within the project. In addition to the academic publications we have focussed our five newsletters, aspects of our web content and two International Innovation articles at a broader audience to engage the full spectrum of stakeholders relating to the sustainable use of ENPs.
At the inception of the project it was recognised that one of the key limitations to the advancement of our understanding of the fate and toxicity of ENPs was a lack of expertise within the research community. Therefore there has been a significant training component to the NanoFATE project. In addition to the two open PhD workshops organised by NanoFATE consortia members has provided training in a further seven training events and workshops. NanoFATE has directly provided a research platform for the studies of three MSc students, nine PhD and DPhil candidates and six early-career-stage post doctoral researchers.
Work package 7 (Dissemination and Training activities) developed the Project Website and a series of Electronic Newsletters (circulated to more than 800 contacts) for the external profile and communication of results and designed to be readable & interesting for a broad lay audience. Two summary articles were published in “International Innovation” with hardcopies delivered to hundreds of European decision makers. The website was overhauled to accompany these publications and yielded excellent statistics: a cumulative total of 51,052 unique visitors to the site representing 894,159 page visits. In this way a significant audience was informed of six NanoFATE training workshops & modules organised jointly with other NSC projects, 33 contributions to NanoSafety Cluster events, & 29 conference presentations. On the publication side, results are flowing: 21 peer reviewed ISI articles, 7 currently submitted, 4 mass media publications and 46ISI papers under preparation. Regulators have been engaged through targeted emails, website updates, distribution of leaflets at the Commission’s 2nd Regulatory Review on Nanomaterials Workshop, NanoFATE members have been invited to OECD WPMN working group meeting by national regulators to act as scientific experts and have delivered policy advice also at various national meetings and to sit on the scientific committee for a forthcoming ECHA workshop on nanomaterials.
List of Websites:
http://www.nanofate.eu
