A Database and Ontology Framework for Nanomaterials Design and Safety Assessment
4. To efficiently communicate to consumers, regulators and insurance communities that, by following the GUIDEnano Tool,
Figure 1. Target of the risk assessment in the GUIDEnano Tool
Specific goals of the project:
1. To develop methodologies to evaluate the risks of a wide diversity of nano-enabled products on human and environmental health, throughout their life cycle.
2. To develop innovative solutions to reduce the identified risks. A wide range of risk mitigation strategies and guidance on the selection of the most appropriate measures for each scenario associated to an unacceptable risk will be provided.
3. To integrate the risk evaluation and mitigation strategies into the GUIDEnano Tool and to carry out an iterative process of performance testing, feedback and improvement steps to validate its suitability and applicability to real-case NM-enabled products, including a detailed plan for the hosting and maintenance of the GUIDEnano Tool after the life time of the project.
4. To efficiently communicate to consumers, regulators and insurance communities that, by following the GUIDEnano Tool, risks associated with an NM-enabled product have been adequately identified, evaluated and mitigated across the whole of their life cycle. Thus, ensuring that workers, consumers and environmental health have been appropriately protected, and facilitating social acceptance, regulatory control, and insurance activities related to nanotechnologies.
4 Major Outcomes from project
GUIDEnano was structured into 11 work packages (Figure 2) arranged by four main blocks: the Coordination block (WP1 and WP2), the Knowledge block subdivided into different technological building sub-blocks (WP3, WP4, WP5, WP6, WP7 and WP8) that generated the scientific input to the GUIDEnano Tool, the Software Development and Demonstration block (WP9 and WP10) that created the Tool itself and validated it in real life case studies, and the Dissemination, Standardization, and IPR block (WP11).
The major outcomes achieved by each WP (the technical ones:
WP3-9) of the project is explained below.
productionNM
NM-enabled product manufacturing
Use Recycling End-of-life Nano-enabled
product X
Figure 2. Organization of the project in work packages
WP3: Release assessment
The main objective of this WP was to generate and validate strategies to identify and categorize the release processes that take place during the life cycle of a NM-enabled product, and assess potential release of NM during such processes. Apart from the case studies included in GUIDEnano, which included polymeric nanocomposites, antifouling paints, anti-slip and photocatalytic tiles, nano-enabled textiles, nanocellulose based coatings and bituminous road products (see Figure 3), this WP aimed at covering the most representative applications in the market as well as those nanomaterials with current higher production volumes.
Based on mass-flow diagrams describing the life cycle stages of nano-enabled products, WP3 identified the sources and release pathways of different nanomaterials (NMs, including TiO2, Ag, ZnO, MWCNT and Cu-based nanomaterials). These release pathways or release scenarios were the starting point of the GUIDEnano Tool aiming at delivering standardized formats and default NM release values from nano-enabled products into the different environmental compartments. Release scenarios are called Activity Cards (AC) in GUIDEnano Tool. An AC library with 160 activities has been created and implemented in GUIDEnano tool, in collaboration with WP4. From this list the user can select one or more AC and start the risk assessment process. This library covers release scenarios in different life cycle stages of nano-enabled products: 1) synthesis (e.g. flame spray pyrolysis), 2) manufacturing (e.g. dumping), 3) use (e.g. use of polymeric nanocomposites outdoors) and 4) end-of-life (e.g. incineration). These activities contain information relative to the release of NMs and waste containing NMs that is generated during different processes (default release values are expressed as mass fraction). This allows predicting amounts of NMs reaching the different environmental compartments (e.g. water), and link the release / exposure module with subsequent fate and toxicity in the GUIDEnano Tool. Such environmental release factors were defined from existing literature values and information coming from the industrial partners when possible. In absence of empirical data release was defined following the ECHA R.16 Guideline and also from expert judgment.
A series of experiments were planned to reduce key uncertainties that lead to the prediction of NMs release taking into account the life cycle stages that are most likely to result in the transformation and/or to result in the release of NM from different nano-enabled
products provided by the industrial partners in GUIDEnano. As a result, 7 nano-enabled products and 11 exposure scenarios were evaluated (see Figure 3) by LEITAT and CEA, investigating different parameters such as NM concentration, crystalline phase or coating. Experiments were performed to: 1) understand which are the processes promoting release, 2) how to reduce such release, 3) release kinetics, 4) release forms.
Figure 3. Release experiments performed in WP3.
No NMs release was found during abrasion of the antislip tiles containing Al2O3, paper film containing nanocellulose and polymeric composites with embedded MWCNT. Regarding the photocatalytic coatings applied on roads (TiO2), NMs were progressively removed from the surface, promoted by oxidation of bituminous compounds due to UV radiation. Interestingly, NMs release was successfully reduced by the application of safe-by-design strategies on three different case studies: photocatalytic tiles (TiO2), polymeric nanocomposites with TiO2 and textiles with Ag. These strategies consisted in modifying the interaction between matrix-nanomaterials, modifying the surface properties of the NMs and modifying the morphology of the NMs to improve adhesion on surfaces.
Based on criteria such as scalability, amount produced and functionalization properties, but also according to WP5 and WP6 requirements, WP3 supplied different nanomaterials to be tested in those WPs (fate and toxicity, respectively). These materials included TiO2, CeO2 and Ag with different coatings, release NMs (e.g. Ag2S), or materials collected from experimental simulations (e.g. waters collected during the leaching experiment with antifouling paints).
WP4: Exposure assessment
The main objective of WP4 was the development of models and guidance on (human) exposure assessment for the various stages of NM-enabled product value chains (life cycle). The human exposure assessment quantifies the indoor air concentration of Nano Objects and their Agglomerates and Aggregates (NOAA) by using available release/exposure measurement data, libraries and/or models for a wide variety of exposure scenarios, and guides the user of the GUIDEnano Tool to the available exposure data (in
libraries) or to the appropriate exposure assessment model in the absence of suitable measured data.
An activity card library, containing around 150 activities, was built with associated worse case material release rates. These release rates are then automatically processed in the dispersion model resulting into a worst case estimate of the exposure.
In addition to the activity card library, exposure scenario information from ongoing and finished FP7 projects (e.g.
Nanomicex, Sanowork, Nanodevice, MARINA, SUN, NanoReg, GUIDEnano) was collected using the MARINA exposure scenario template. The resulting GUIDEnano library contains over 200 exposure scenarios.
A quality, similarity and relevance scoring system were developed to rate the different information sources as to their analogy to the user’s scenario.
Because the GUIDEnano Tool aims to quantitatively assess human exposure (in respirable mass concentration), currently available qualitative risk/control banding tools cannot be used, and therefore for worker’s exposure it was decided to use the ART model (Advanced Reach Tool). For validation of ART for NM scenarios, high quality exposure scenarios were selected and run in ART to test the feasibility of the model. In addition, a model performance check was performed for the Advanced REACH Tool and the Stoffenmanager Nano 1.0 tool. This performance check included metric conversion methodology, analyses of nano-exposure measurements and finally the model performance check, which was designed to gain knowledge on the applicability of certain models for the use within nano-exposure assessment and thus for implementation in the GUIDEnano Tool.
Number count exposure data collected in standardized circumstances were converted to mass concentrations in order to be comparable to the model outputs. Statistical analyses were performed to examine correlation between model estimates and converted mass concentrations. Furthermore, model performance was evaluated based on the uncertainty given by the developers of the ART model.
Secondly, exposure measurement data were generated for the GUIDEnano case studies in order to refine the exposure scenarios mentioned above. Measurements were collected using the following direct-reading instruments: Condensation Particle Counter (CPC), NanoTracer, DiscMini (only Hempel and Servia Canto) and filter samples of airborne particles for SEM/EDX analysis. These measurement results gave good insight in the exposure levels during several activities in the value chain of the GUIDEnano case studies and were used to further validate the GUIDEnano model. Chamber experiments (under well controlled conditions) were also performed to simulate the scenarios of the case studies in the absence of workplace measurements.
WP5: Environmental Fate
The main objective of this WP was to generate strategies to understand how NMs behave in natural systems including the critical transformation reactions. Several key questions were addressed: (i) How do NM properties and their nanoscale features affect their behavior and interactions with other environmentally relevant parameters, i.e. which property-fate relationships are crucial for fate prediction? (ii) What transformations are likely to occur in natural systems? (iii) How do the transformations affect
the NM’s fate? These questions were quantitatively addressed to develop a conceptual fate model framework (Figure 4) focused on NM fate and behavior for implementation into the GUIDEnano Tool, and parameterized using available literature or by obtaining experimental results when data gaps were identified.
Figure 4. Schematic overview of the approach used in the model world that should take into account the following:
compartments/zones/timeframes/processes.
The integrated exposure model that was proposed for the prediction of the environmental fate of NMs in the various natural and man-made compartments had the kinetic nature of the NM fate processes as a central endpoint. A semi-mechanistic approach was used, with a minimum number of fate descriptors for the NM.
In general, environmental matrix interactions depend on NM properties and relevant environmental properties, such as organic and inorganic matter type and concentration, as well as ionic strength and the presence of mono- or divalent ions. The fate descriptors were detailed for the following compartments:
wastewater treatment plant (WWTP), water, soil and subsurface, and for the following transformations: heteroaggregation, sedimentation and sulfidation. The time-dependency of chemical species concentrations is an essential feature of this model, because it serves to link exposure scenarios with hazards that change over time.
WWTP microcosms using Ag NM showed that sedimentation depends on size and total suspended solids, whereas coating, shear forces or ionic strength had no effect. No aggregation was observed in these studies. The final model for the fate in WWTP predicts the mass and speciation of NM that ends up in soils AND freshwater as a function of size and Hamaker constants.
Dissolution, heteroaggregation and sedimentation experiments in the aquatic compartment were performed using several CuO, ZnO and Ag NMs with different manufactured coatings and sizes. One of the main conclusions from the experimental work was that NMs dissolution in relevant environmental matrices (i.e., in presence of organic matter) is not significant and independent of water composition, manufactured coating or size of the NM.
WP6: Hazard Assessment
The main work of this WP was developing a strategy for predicting the (eco)toxicological and human health hazard of the exposure-relevant NM forms released into exposure situations throughout the lifecycle of NM-enabled products.
The eco- and human toxicologists worked together to develop a hazard assessment strategy to estimate safety limit values making the fullest use of existing information while allowing prediction of
hazard on the basis of different levels of data availability (each with an associated level of uncertainty reflecting the richness and quality of data available). Existing PNEC, OEL or DNEL values for the material under evaluation will be used, if these are available.
Otherwise, generic highly conservative thresholds can be used.
When necessary, these can be refined based on available toxicity studies. The strategy mostly relies on information from studies following harmonized testing guidelines such as OECD and/or ISO, but is designed to also make use of ‘other non-standard’ studies and toxicity information. When assessing already existing individual toxicity studies from literature the hazard assessment strategy involves the establishment of scores to inform on quality (i.e. how good and reliable a study and its reporting is), relevance (i.e. how relevant the NM study is for the respective environmental compartment or human pathway/endpoint) and similarity (i.e. how well does the NM exposure in the study to be used reflect the exposure relevant form of nanomaterial which is being assessed).
These scores are used to select studies that can be included in the process to derive safety limit values. In addition, the similarity score is used to introduce a ‘dissimilarity’ uncertainty factor.
Hypotheses-driven experiments (different cores and coatings) were performed to evaluate the assumptions identified as most critical to reduce the most prominent uncertainties (e.g. from read across) in the process. To test the influence of the core material and the different coatings a base set of three different core materials (TiO2, CeO and Ag) each with three different coatings (citrate, PEG and the hydrophobic coating DDPA) was selected for testing in a wide range of in vivo and in vitro test systems. No consistent trends were observed for the effects of coatings. These effects depended on the core material and the experimental test.
In addition, experiments were performed to characterize the hazard profile of the release and exposure relevant materials of the GUIDEnano case studies and to evaluate the efficiency of safe-by-design modifications.
Figure 5. GUIDEnano hazard assessment strategy for a human and environmental risk assessment.
WP7: Risk assessment
The main goal of this work package was to develop a strategy for risk assessment of release- and exposure-relevant NMs in NM-enabled products throughout the various product life-cycle stages.
This risk assessment strategy was incorporated in the interactive
web-based GUIDEnano tool and was evaluated with hypothetical and real case studies within the project.
The initial decision flow for risk assessment incorporated in the GUIDEnano tool is presented in Figure 6. The safety limit value is comparable to the Derived No Effect Level (DNEL) for human health and Predicted No Effect Concentration (PNEC) for environment according to the REACH regulation.
Summary of the steps to derive a safety limit value:
1. Select hazard studies with associated effect levels (NOAEL, LOAEL, BMD, etc)
2. Determine modification/assessment factor for each effect level
3. Derive safety limit value by applying the assessment factors for each effect level (including a factor for dissimilarity)
4. Derive overall safety limit value per endpoint 5. Risk assessment: compare safety limit value with
corresponding exposure level
Figure 6. Updated GUIDEnano risk assessment decision flow The required factors (step 2) were listed and for each factor the purpose for application and the value are included. The risk assessment in GUIDEnano Tool v3 is deterministic, but some information is provided to the user to inform on the sources of uncertainty for the hazard limits derivation. For each assessment factor, it is indicated if it represents modification, uncertainty or variability. This information is needed to determine the possibility for reduction of the uncertainty of the safety limit value.
An output report has been designed and generated by WP7 together with the tool builders. This output report consists of a detailed background report with all information generated by the Tool based on the input of the user. This will be provided to the user after using the Tool, together with an executive summary of the risk assessment result (probability of risk) and proposed follow up actions (reduction of uncertainty, mitigation of risk by for instance safe by design strategies).
Case studies reported in the literature or hypothetical were used to evaluate the version 2 of the Tool. Feedback from this evaluation and feedback from stakeholders (e.g. industry, (re)insurance communities) obtained in dissemination events and via tailored questionnaires was used in the updates towards version 3. WP7 also monitored to what extent the industrial
partners of the project were able to correctly use the GUIDEnano Tool during the evaluation of the project case studies.
WP8: Risk Management
Work package 8 aimed to propose, develop and validate risks mitigation measures (RMM) to reduce the potential risks identified through the risk assessment provided by the web-based Guidance Tool GUIDEnano. In such a way the final user, will obtain suggestion on the most appropriate RMM allowing to control risks highlighted and decrease them to an acceptable level, making the whole process safer. Among possible RMM, safer-by-design (SbD), occupational exposure control and advanced waste management strategies, were selected and are being evaluated within the context of GUIDEnano case studies.
Safer-by-design strategies were proposed taking into account the case studies and the NM employed and focusing on relevant endpoints and effects/risks to mitigate. SbD strategies were intended to :
• re-design relevant physicochemical properties of NM to mitigate their hazardous potential, while maintaining their characteristic functionality within the NM enabled product,
• avoid or reduce the release of NM during different life cycle stages of the nano-enabled products by improving compatibility between NM and matrix, to lower the possibility of environmental and/or human exposure to NM,
• avoid or reduce the environmental and/or human exposure to NM by designing and synthesizing less reactive and/or less persistent NM.
A set of SbD NM were synthesized at lab-scale level with targeted modifications of relevant physico-chemical properties and were provided to the different WPs to be evaluated for specific endpoints/ functionalities. For selected case studies, the resulting SbD NMs were implemented within their real industrial setting.
The investigation of occupational exposure control measure was focused on the measurement of the effectiveness of the personal protective equipment (PPEs) commonly employed by the industrial partners involved in GUIDEnano, with the objective to obtain for each of them a nominal protection factor (NPF) against NMs. The NPF of different PPEs have been determined, modified and validated using standard protocols available for bulk materials.
Depending on the nature of the NM tested, two test protocols were adopted:
• a static test protocol were employed to evaluate the NPF of PPEs against metal oxide NMs, which are not yet classified as safe materials. These tests were performed under simulated conditions and using mannequins in substitution to individuals,
• a dynamic tests protocol were designed to evaluate the NPF of PPEs against NaCl nanoparticles, that being not considered hazardous materials, allowed that PPEs were worn by different individual, thus conducting the test under the more realistic dynamic conditions.
Several types of masks, suits and gloves were investigated under simulated conditions performing static tests using different NM and at different concentrations. Results demonstrated that suits and masks, especially respiratory filters, offered a quite good NPF, sometimes slightly lower than the default value. Data collected showed unlike performance of the glove depending on NM type and concentration employed, therefore further tests are needed.
Full and half masks were tested under the dynamic test protocol.
Results showed high variability on the performance of the masks due to the fitting to the facial geometry, that could cause leaks where NM penetrate. From results obtained, the half-masks
Results showed high variability on the performance of the masks due to the fitting to the facial geometry, that could cause leaks where NM penetrate. From results obtained, the half-masks