European Commission logo
español español
CORDIS - Resultados de investigaciones de la UE

Development of sustainable solutions for nanotechnology-based products based on hazard characterization and LCA

Final Report Summary - NANOSUSTAIN (Development of sustainable solutions for nanotechnology-based products based on hazard characterization and LCA)

Executive Summary:

One main purpose of NanoSustain was to identify possible risks along the life-cycle of EN containing products that are close to or on the market and so help to improve their sustainability. Three EN manufactured worldwide in large quantities have been investigated (TiO2, ZnO and MWCNTs), and nanocellulose, which is relatively new but derived from sustainable sources and with an enormous potential for future applications. Four life-cycle phases have been studied: production of input EN, production of EN containing products, use of these products and their end-of-life recycling/disposal. Materials used and released at each stage were characterized, toxicity tested, and material + energy flows estimated. Two databases have been developed, a material and a literature database, to make the newly generated data and the knowledge developed by other projects on hazard and exposure of selected EN available to project partners. An online framework was created to systematically collect and document the new pc and biological data, and to allow their statistical treatment, correlation and evaluation. An enormous amount of new data has been generated on the toxicity of pure and lifecycle (LC) relevant materials and continuously transferred to the material database and to feed into Material Data Sheets (MDS) that have been prepared for the 4 selected nanomaterials (NM) together with safety relevant information on properties provided by manufacturers and the project. To ensure validation of applied methods and data consistency and comparability across all partner laboratories, standardized protocols (SOPs) have been established for NP characterization by inter-laboratory comparison exercised (round robins).
As our current knowledge is still scarce on how toxicity changes when NM are embedded in a product, specific test samples have been generated, including raw and LCA relevant materials and composites to study the toxicology of after-production handling, transport, reworking and disposal. Sanding, weathering, abrasion and leaching tests were used to simulate changes in exposure during these processes, and environmental stress. The source strength of dust emissions was measured from paints (± TiO2) and epoxy boards (± MWCNT) during sanding, breaking of nano-ZnO containing glass and tearing of nanocellulose containing paper, as a measure to assess real workplace exposure conditions and compared with conventional products. Both pure NP and NP embedded in products, and the collected dust, were characterized and in vivo tested in mice for inflammation and DNA damage. There was no increased emission of the tested nanoparticles (NP) during sanding of NP containing materials compared to reference samples not containing NP. Dose-response relationships of critical biological end-points were identified for dust and compared with results obtained from pure or embedded NM. There was no significantly increased toxicity when adding nano-TiO2 to paint or CNT to epoxy matrix, which suggests that toxicity may be masked in the product matrix. In contrast, the observed toxicity of nano-ZnO added to coated glass seemed to be conserved. Suitability tests of standard bioassays, such as the “Vibrio fischeri” test showed no acute or any nano-specific toxicity for the tested nano-TiO2 and nanocellulose.
Life-cycle process models were developed for the 4 selected NM and various types of environmental impacts assessed by Life Cycle Assessment (LCA) including LCI (Inventory) and LCIA (Impact Assessment). Also an exposure model was completed and generic data on material flow and predicted environmental concentrations (PEC) estimated. Based on the LCA, criteria + guiding principles have been established for a precautionary design and improved recyclability of NM. Results reveal a strong dependency of the environmental impact on the type of manufacturing and a high variation of impact factors compared with conventional materials. Also the influence of EN on the environmental impact of new applications was much depending on resource efficiencies but also on the lack of data, why still much uncertainty remains. However, none of the selected applications (TiO2 in paint, ZnO in glass coatings, MWCNT in epoxy material, nanocellulose as paper additive) showed any significant exposures to water, air and soil.
As ultimate goal, NanoSustain explored the potential of current disposal techniques for reuse/recycling, safe treatment and final disposal of NM containing products, to improve end-of-life processes. Nanocellulose was used to test the suitability of composting for organic recycling, a CNT containing epoxy composite to assess the feasibility of incineration as final treatment, and a nano-ZnO coated glass to test glass melting and land-filling as recycling and disposal option. Results showed good degradability and no eco-toxicity of paper containing nanocellulose. Also incineration gave a good energy recovery and combustion of CNT containing boards with no CNT found in bottom or fly ash. And melting of glass released NP independent of the type of coating, while leaching tests showed how Zn is released from ZnO-coated glass under landfill conditions.

Project Context and Objectives:
Since the production of engineered nanomaterials (EN) is increasing, the amount of products reaching the end of their life cycle may increase, and so exposure to man and the environment. We still do not know how and to what extent certain EN may be released from technical or consumer products and transported, transformed, dispersed or accumulate in man or natural systems, when used or after disposal. For this reason, there is a strong need to improve our knowledge on the impact and fate of products containing nanomaterials along the whole life cycle, in particular at end-of life phases. One of the main objectives of the NanoSustain project was to explore the applicability of current techniques for the safe reuse / recycling, final treatment and disposal of EN.
Due to their large surface area, nanoparticles (NP) behave quite differently from bulk materials, why a considerable amount of research is currently undertaken to commercially exploit their unique behavior for a wide range of applications. At the same time, there is a growing understanding that exactly the novel behavior of EN may give rise to negative effects on living systems, including man. Due to their small size (1-100 nm), EN may be more toxic than conventional materials with the same composition, and may cross or adsorb more easily to cell membranes or barriers.
NanoSustain has addressed all these questions and achieved main objectives, including (1) a compre-hensive hazard characterization of 4 commercially and environmentally relevant EN (nanocellulose, MWCNT, nano-TiO2 and nano-ZnO), (2) a preliminary life-cycle assessment (LCA) of these EN, and (3) assessment of their human and environmental impact, and (4) testing the applicability of current waste disposal techniques for their safe and sustainable recycling and final treatment. Sustainability was understood as the use of these new materials that matches the needs of future generations and takes all possible risks along their life-cycle duly into account, why in particular their recyclability, fate and disposal became crucial questions.
Another key challenge of the current research was also addressed, to assess if and to what extent ex-isting risk (RA) and life-cycle assessment (LCA) approaches need to be revised to take the particular properties and uncertainties of EN (e.g. regarding toxic dose levels or long-term behavior etc.) duly into account. For this different dose-specific parameters, such as particle size number instead of mass concentration, and different end-points have been applied. Most important material, hazard and exposure characteristics have been determined and critical dose-response relationships and life cycle stages identified, as well as no effect levels that may occur during handling, transport, use, recycling and disposal of EN. Also the state-of-the-art on RA methodologies was critically reviewed, including occupational + consumer risks from direct or indirect (environmental) exposure or from waste disposal.
A preliminary LCA approach was applied to identify potential environmental impacts of EN along the product value chain and opportunities to improve their environmental design and performance at an early stage. Both newly generated and literature data on exposure and toxicity of EN during critical life cycle stages have been used to develop criteria and guidelines for the improved recyclability, precautionary design and RA of selected EN. In addition, a new exposure model was developed to allow the prediction of environmental concentrations (PECs) and to evaluate the applicability of current LCA to EN.
One major limitation when assessing the risk of EN is the lack of reliable exposure and dose/response data. NanoSustain has generated new accurate exposure and toxicity data that take the specific properties of EN, such as size, aggregation or surface treatment, into account and their influence on biological end-points to reliably predict possible human and environmental impacts. For this reason, a set of LCA relevant test materials has been produced and investigated to simulate different life cycle stages and processes, and to characterize the hazard of NP that cover a wide range of properties, such as size, composition, morphology, crystal structure, solubility or surface layer composition, and to show how these properties and associated toxicities change during the life cycle. To assess uptake and effect of EN in the human body during critical LCA stages, such as handling, transport or disposal, NanoSustain used the inflammatory reaction and formation of ROS as well as genotoxic effects caused by DNA damage in mice. Also the suitability of current eco-toxicological tests was to assess the impact of EN on natural systems was examined. The vast amount of new data produced will be available after their peer publication to current standardization efforts on test and analytical procedures under OECD, ISO and CEN.
To explore the feasibility and sustainability of current waste disposal techniques for EN, lab-scale experiments were implemented including composting (organic recycling), melting (glass recycling), incineration (final treatment), and land-filling (final disposal). Organic recycling has been evaluated by testing the biodegradability of nanocellulose based products. The performance of glass recycling was assessed by measuring the release of NP during melting of ZnO-coated glass and the composition of the recycled glass. Also incineration was explored as a final treatment option for MWCNT containing epoxy composite products. Finally standard leaching tests were used to simulate the extractability of Zn from nano-ZnO coated glass and its mobility under landfill conditions. To monitor the release of NP from these treatment techniques, the performance of current analytical methods has been assessed to detect and quantify emissions of NP in the gas, ash or bottom samples or in the leaching solution. The produced new experimental, analytical and toxicological data will help to improve the efficiency of current waste disposal systems and their applicability to EN, and so ensure their safe recycling, final treatment and disposal. In addition, results will also assist the future design, fabrication and use of environmentally safe and sustainable products and the control of NP released at their end of life.

Project Results:
The ultimate goal of the NanoSustain project was to explore and examine new technical solutions for the sustainable design, use, re-use/recycling, safe treatment and/or final disposal of specific nanomaterials and associated products. End-of-life phases occurring after the production and use phase have been the final target of the implemented research and development work. The experimental part designed to find practical and innovative solutions for handling nanomaterials containing waste in a safe and sustainable was built on a comprehensive physicochemical and hazard characterization and exposure assessment of the following selected EN and associated products:
• nanocellulose based materials and products
• nano-TiO2 based materials and products,
• carbon nanotubes (CNTs) based materials and products
• nano-ZnO based materials and products.
The following main S/T objectives have been addressed:
1 Implementation of a comprehensive hazard characterization, exposure analysis and risk assessment of pure NM, associated composites and products, as well as dust derived from experiments that simulate handling, reworking and transport of NM as well as relevant LCA after-production phases that may occur during use or at the end-of-life of selected NM.
2 Establishing critical dose-response characteristics and no-effect levels to humans and the environment in particular for materials from after-production and relevant LCA phases (use, reuse/recycling, final treatment and disposal), also to evaluate the applicability of current risk assessment schemes to NM.
3 Performance of a preliminary life cycle assessment (LCA) for selected NM by applying leading edge methodology to identify potential environmental impacts throughout the whole life-cycle (from production, application and use phase to recycling and disposal), and further developing and improving the applicability of existing LCA methods (such as prospective LCA) to NM and their use for a more precautionary design and risk management of NM.
4 Assessing human health and environmental impacts of selected NM by producing a set of standardized nanomaterials that represent changing pc properties, LCA phases and processes (e.g. pure NM, NM embedded in a composite or product, in dust, ashes, compost, leaching water etc.) by using the well known inflammatory reaction that follows human exposure to nanomaterials in the lung, the comet assay to determine genotoxic effects as a measure of the reactive oxygen species (ROS) formation and for possible primary and secondary DNA damages in cells and animals, and by using eco-toxicity tests, such as the kinetic luminescent bacteria “Vibrio fischeri” test, to assess effects on aquatic systems and toxicity mechanisms that may arise directly after manufacturing, during use phase and/or recycling or disposal, but also indirectly from the environment, e.g. by accidental release or leakages.
5 Explore the performance of technical disposal solutions by laboratory experiments to elucidate the behavior and fate of nanomaterials during end-of-life phases, such as recycling (composting and glass melting), final treatment (incineration) or disposal (land-filling), to better design, engineer and fabricate more safer and sustainable NM. This specific goal was achieved by 1) testing and improving the biodegradability and suitability of nanocellulose based materials for organic recycling and the safety of resulting compost materials, (2) by measuring NP emissions during glass melting as a measure to recycle nano-ZnO-containing glass, (3) by exploring the suitability of incineration of CNT containing epoxy boards as a safe treatment option (if no recycling/reuse is possible), and (4) by leaching nanoparticles (NP) from products to simulate land-filling conditions and the fate of NM contained in waste. To monitor the efficiency and sustainability of these treatment techniques, the performance of current analytical methods was assessed and their capacity to detect and quantify the release, distribution and environmental fate of NM along their life cycle, but also to prevent severe shortages and economic risks for present and future applications. The generated new emission and exposure data associated with these final treatment techniques will create the needed scientific base for a more careful product design and sustainable use and management of selected NM but also for amending existing relevant legislation (e.g. on waste treatment or hazard classification).
Main S/T results achieved
A concise overview on main S/T results received by the NanoSustain project and generated within each work package (WP2-5) and task throughout the project and in line with Annex I of the Grant Agreement is documented in the various technical deliverable reports prepared by the project, and a condensed summary is available in D1.3_2. In addition, the main scientific outcome will be available through peer-reviewed publications already realized, and/or planned or in preparation, highlighting most significant S/T findings and their possible wider (societal) impact. A summary of main S/T results is given in the following for each technical work package (WP2-5) with reference to particular tasks and/or deliverables produced.
WP2: Data gathering, generation, evaluation and validation
WP2 has been responsible for developing and maintaining 2 project-specific databases, i.e. the (1) material (D2.1 + D2.2) and the (2) scientific literature database (D2.3). The technical (material) database was developed to create a framework appropriate to systematically collect, evaluate and document the new and complex scientific data produced during the whole project period (see also D2.5). For this, an Excel sheet was used and an online structure to allow easy data transfer and access to all partners, and a correlation of the physicochemical (pc) data with all biological endpoints determined at the toxicological, eco-toxicological, exposure and LCA level. In addition to the established technical database, also Material Data Sheets (MDS) were prepared for the 4 selected nanomaterials (i.e. nanocellulose, nano-TiO2, CNT and nano-ZnO), to present main material properties partly delivered by manufacturers and partly generated by the project, together with a presentation of methods and analyses performed (MDS are attached as Annex 1 to D2.5). The literature database (see also D2.4) was set up to regularly update the most recent literature on topics covered by the project and to make the most recent knowledge constantly available to all project partners within their field of research (such as hazard, exposure, RA and LCA, disposal of selected NM). A final task of this WP was the validation of the generated (pc and biological) data, partly through establishing internal advising expert groups (by the Project Internal Committee) specialized in material measurement and testing, toxicology, RA or LCA, to evaluate and discuss the quality and relevance of results achieved (D2.6) but in particular by implementing a comprehensive inter-laboratory comparison exercise using known reference standards, materials with unknown compositions, and materials that have been already tested within NanoSustain, to validate the reproducibility and accuracy of measurement methods used prior to toxicological testing (see D2.7).
A short description of tasks implemented and of the progress achieved in WP2 is given in the following (a more detailed description is available in the respective deliverable reports).
Task 2.1: Data Gathering. Technical data about the materials selected by the project was continuously updated by transferring information from new safety data sheets or from technical data sheets provided by producers Nanogate AG (for nano-ZnO) + UPM (for nanocellulose) (both members of the project), and Nanocyl (for MWCNTs) and Flügger/Denmark (for nano-TiO2). The database structure and content (D2.1) are discussed in D2.2 (both deliverables have been prepared during the 1st periodic report), where also a comparison with the experimental data is presented (as discussed and presented in D2.5).
The second part of Task 2.1 was to collect and evaluate knowledge from the most recent scientific literature on all aspects relevant for the project on the life cycle of the selected nanomaterials (i.e. TiO2, nanocellulose, ZnO, and CNT) from production to disposal. Several updates of the resulting deliverable report D2.3 have been produced during the project and a search function installed to allow the use of specific key words to search in recent papers by search engines such as “Scopus”. All partners have actively contributed to build up this literature database. One key requirement for a paper to become included into the literature database was the distinct documentation of a proper material characterization, where the actually measured data proved to be more relevant than data provided by the manufacturer or supplier of the test material (see D2.4).
However, some data were considered although they were lacking proper characterization, e.g. in cases where the relevance of the field of study or technical difficulties in characterization were indicated. All papers collected through this procedure were organized to summarize main findings relevant for the project, and results were analyzed and evaluated by simple statistics (e.g. frequency plots) and reported in a particular report (D2.3). There, a special table is highlighting main brand-new findings identified for each paper and the relevance for a particular task of the project (see D2.3 Table 3).
Task 2.2: Generation of missing data: The physicochemical and biological data that was missing to describe and explain the hazard characteristics and potential exposure to selected NM was newly generated in WP3 and WP5 and has been continuously transferred to the technical material database (see task 2.1 and deliverables D2.1 and D2.2) and immediately made available to all project partners for further evaluation, discussion and interpretation. Also, the knowledge collected and evaluated by the literature search (in D2.3) has been made available to project partners, to give them a continuous update on data and knowledge that is still missing in their respective field.
Task 2.3: Data Evaluation: A Project Internal Committee (PIC), subdivided into several expert groups, has been set up by Work Package Leader 2 (WPL2) to cover all areas of expertise available and required within the project and to ensure a high quality and the statistical treatment and evaluation of the collected and newly generated data and results (see D2.6 and D2.7). Another focus of this task was to design the results database (see D2.1) to compile and store all new data generated by the different partners. For each experimental activity, a separated (excel based) data template was drafted taking the structure and content suggestions from partners duly into account (data organized according to D2.4). The work done in the 2nd part of the project was to systematically collect and transfer the huge amount of continuously produced new data to the excel sheets. In D2.5 the data is briefly discussed by presenting what has been done (measurement endpoints), how (methods/protocols) and why (relevance for the project) it has been done. In addition to the excel version of the database, an online version was established at the project web site (see D6.1 and D6.2) and linked to the literature database (D2.3) and the Material Data Sheets (MDS).
Task 2.4: Data validation: As a first step of the data validation process (see D2.7) the Project Internal Committee (PIC) was discussing existing experimental and methodological protocols, and suggestions were given on the type of experimental set-up to be applied. In general, each partner already had an established set of accepted or even standardized method protocols. Another aspect of the data validation was the statistical treatment of received results, with the aim to measure the significance of observed data variations. This particular data validation step was implemented inside of each technical deliverable (within WP3 and WP5) that contains and presents newly generated scientific results.
Another step of the data validation was the performance of an internal inter-laboratory comparison study to assess the quality, consistency and accuracy of achieved results.. Methods used by the different partner laboratories to perform pc characterization of selected NP were validated, after the PIC and involved work package leaders agreed upon the necessary procedure and measurands. For this, 4-6 blind samples were purchased and distributed to the participating laboratories for comparison of their analytical results. All results obtained by the inter-laboratory comparison study are presented in D2.7.
National and international standardization committees, as well as several high level scientific journals, have begun to establish recommendations for adequate pc characterization data relevant for toxicology testing of NM. NanoSustain has complied with these high-level standards that were also emphasized by the results obtained from the scientific literature study (see D2.3). Consequentially, all newly generated toxicology data is based on a sound pc characterization of the test material. The established literature database was designed for 2 target groups: 1) the project partners, to make high-quality information relevant for the project continuously available, and 2) the wider scientific community, to offer a common database on research topics relevant for and covered by the project. The final database includes almost 200 peer-reviewed papers covering all phases of the life cycle of selected nanomaterials published by January 2013. One interesting outcome of the literature search is the fact that more and more recent studies document an increase in the performance of high-quality pc characterization of NP prior to any toxicity testing. This may be already a consequence of the increasing demand set up by the more important scientific journals on minimal reporting requirements that includes a sound scientific characterization of the NP to be tested. Most reviewed papers focus on ZnO and TiO2, but there is an increasing number of publications that addresses CNTs. For nanocellulose, however, the number of published papers is not high so far and no clear trend is in sight. Papers published during the last year (2012) mostly contain reviews on the synthesis and use of nanocellulose materials and on ongoing research, and only few papers deal with effects or exposure analysis, and none with fate and transport. Most frequently addressed pc properties include chemical composition, particle size and size distribution, morphology, surface area, and agglomeration/aggregation, which are increasingly measured also by using methods in combination. Core scientific findings of the reviewed papers have been extracted and summarized directly in the database and in descriptive tables of the D2.3 report.
A central core of the project is the project results database that has been developed by adapting a structure compatible with databases developed and used in several other Nanosafety Cluster (NSC) projects (see D2.1 D2.2 and D2.5). To allow direct and easy comparison with similar data available from the scientific literature on pc characterization of NM, the same structure and parameters were used as criteria to organize the data achieved during the project (see D2.4). The database was organized according to the following categories:
• physicochemical characteristics,
• toxicology,
• eco-toxicology,
• exposure,
• LCA.

For each category and for each material, the specific pc characterization is reported, if available (e.g. size distribution in the cell culture media), and the different measured endpoints given. It is possible to specify the methodology used in a category to implement the measurement. In addition, the horizontal organization of the database specifies the different experimental setups used and reported for the various tests, as they may impact the different endpoints measured (e.g. a varying aggregation may be seen of particles in different solvents). The results database will be used as a basis for other similar running or future projects and is available as an online version at the project website (for now in the restricted area of NanoSustain), reporting all data generated for each selected material. Also a link to the Material Data Sheets (MDS) (dossiers collecting the pc data, measurement and testing protocols) is given.
One critical step taken was the performance of an inter-laboratory comparison of methods that were used for pc characterization. The aim was to provide a measure on the quality and statistical confidence of the produced and reported scientific data and on the comparability and validation of the different methods and instruments used by the participating laboratories. Experts from the PIC agreed to compare data on the physical size of NP by TEM (2 partners) and on the hydrodynamic diameter by DLS (3 partners) with supporting data obtained by SEM and AFM as well as by the determination of the average crystallite size by XRD (2 partners). The participating laboratories included 5 different project partners. Test samples used included 3 NIST traceable PSL particle size standards, 3 TiO2 samples of which 2 were from the OECD WPMNM and 1 from NanoSustain itself, 2 ZnO samples of which 1 was again from the OECD WPMNM and 1 from NanoSustain itself. By including the NanoSustain powder samples, the generated analytical data, which was also included in the MDS, could be further used to generate “internal benchmark” values.
WP3: Hazard characterization and impact assessment
Task 3.1: Powders from selected nanomaterials have been generated for subsequent testing and measuring source strengths of dust emissions to simulate handling and reworking of selected EN and products, and for assessing toxicologically and eco-toxicologically relevant pc properties and concentrations of NP in pure and after-production materials (composites), and when released to the environment.
The pc characterization was done on all pure materials, and on dust produced for exposure assessment, and the results are presented and discussed in D3.2 and data of pure materials reported in the MDS (see D2.5) and in the excel file (of the results database), while the dust data are reported only in the online database (see D2.1). MDS data were used for the inter-laboratory comparison (see D2.7).
WPL3 coordinated the set-up of the material data sheets (MSDS) established for the pure EN (see D2.5 Annex 1) and also the inter-lab comparison exercise (see D2.7). The produced material data have been passed on via D3.6 to the database established in WP2.1 (see: D2.5). In addition to the materials prepared by partner NRCWE, UPM produced and delivered nanocellulose and various papers containing pulp (cellulose) and some % nanocellulose for testing in WP3. Also, the industrial partner Nanogate AG produced and delivered testing materials, like raw material, dispersions, uncoated glass as reference, and coated glass.
Task 3.2: The aim of this task was to produce data on human exposure to nanoparticles that may occur during handling of after-production materials.
Sanding dust was generated and collected for toxicological testing and the emission of dust characterized during:
• sanding of epoxy (CNT) and paint products (TiO2)
• breaking of coated glass sheets (ZnO)
• tearing (of paper) (nanocellulose)

Based on results received (see D3.2) a qualitative control banding and first order quantitative occupational exposure assessment model (Nanosafer) has been evolved and is under development by partner NRCWE and ready to be submitted for publication still this year (see D3.7). The model has been presented and discussed with the project LCA expert Michael Steinfeldt (Bremen University). Data from the exposure assessment, dustiness and leaching experiments (done in tasks WP3.2 & WP3.6) and generated by partners NRCWE + Veneto Nanotech (VN) have been forwarded to WP4 for preparing the LCA of selected nanomaterials (see D3.2 and D4.4 and D4.5).
Task 3.3: This task included the identification of dose-response relationships for critical end-points of human health effects for NP-containing after-production dust in mice and comparison with results obtained from pure materials (see D3.3).
For this, the following 2 animal experiments were performed at the lab facility of NRCWE (with ~800 mice tested in total):
• Experiment I: Exposure to 5 different kinds of sanding dusts from NM containing products.
• Experiment II: Exposure to glass treatment product with and without nano-ZnO.

At NRCWE the mice were analyzed for:
1) DNA damage in lung and liver tissue
2) BAL cell differentiation
3) Liver (all mice) and lung histology (mice exposed to ZnO and glass treatment products) (performed at the Danish Food Institute).

Partner VN performed the extraction of RNA from the tissues of treated mice, its retro-transcription to cDNA and the subsequent quantitative amplification via qPCR to determine the expression levels of biomarkers of different biological response pathways.
New data on the toxicities of NanoSustain pure and life cycle relevant products has been generated extensively documented in deliverable report D3.3 and uploaded to the results database (see D2.1) and made available to all project partners to be used in risk scaling.
Task 3.4: The objective of this task was to assess the eco-toxicity of nanoparticles and nanoparticle containing composites but also to evaluate the suitability of eco-toxicity assays, such as the “Vibrio fischeri” bacterial test for environmental hazard assessment of selected nanomaterials, namely CNT, ZnO, TiO2 and nanocellulose (nanofibrillar cellulose NFC) (see D3.8). The eco-toxicity of samples from waste management of NFC (composting) and nanoparticles containing glass (recycling) was also evaluated and results of this work reported in D3.3. Dose-response relationships of critical eco-toxicity endpoints have been established, e. g. for pure NM and contained in recycling materials obtained after composting (nanocellulose) or melting of glass (ZnO).
Task 3.5: Existing methods and tools from the chemicals sector have been identified (already delivered by D3.1 during the first project half) that are already used or have the potential to manage the risk to NM for man and the environment. The reviewed methodologies have been subjected to a critical analysis to study main features of each RA tool, to learn how they work and to identify the type of information required by each of them during all life cycle steps of a NM. Also the feasibility of these methodologies for potential users both inside and outside the manufacturing industry has been assessed. In addition, the uncertainties and gaps of the studied methodologies have also been determined. Of the methodologies identified, two of them are deemed to have more interesting features and to be more suitable for environmental RA of NM. These two methodologies include the Swiss Precautionary Matrix for Synthetic Nanomaterials and the Environmental Defence and DuPont Framework. Both of them have been applied to real cases, that is, the four commercially available nanomaterials studied in the project, and the results obtained are discussed in 2 reports D3.4 and D3.5.
Task 3.6: This task should develop and test an accurate and repeatable analytical method to detect and quantify one or more classes of NP in LCA relevant materials and environmental matrices and to apply methods to real samples. While NRCWE performed sanding experiments simulating the release of NP during handling, VN performed tests simulating environmental stress (by UV treatment, leaching and temperature stress) applied to boards and glass sheets treated with MN based products, namely TiO2 and ZnO. Quantification of their release under these stressing conditions was successfully achieved via ICP-MS (see D3.2 and D3.3).
Task 3.7: Here the type of prevalence and level of NP use in industry was investigated as well as the potential for exposure of human population to engineered nanoparticles. In general, there was no additive effect when adding nano-TiO2 to paints or CNT to epoxy boards and overall results suggest that the toxicity of TiO2 and CNT is rather masked when included in a product, such as paint or epoxy matrix, respectively. In contrast, the toxicity of ZnO was conserved when ZnO was included in a window-glass treatment product (see D3.3). The sanding experiments also showed that there was no increased emission of nanoparticles during sanding of NP doped materials compared to reference materials without NM (see D3.2). Of the tested products, nano-ZnO powder is the product which represents the highest risk because of high dustiness and high toxicity (see D3.3 and D3.8). Data has been uploaded to the database (see D2.1 and D2.5 in WP2).
Task 3.8: Standardized protocols have been established for nanoparticle measurement and pc characterization by a comprehensive inter-laboratory comparison to assure consistency and quality control across all participating laboratories including (see D2.7). By participating in the inter-laboratory comparison, partners contributed to the development of standardized test protocols. The internal data validation was carried out only for physical-chemical characterization data. The performed validation has to be seen as rather indicative, as involved labs were too few. Therefore, data quality was also checked by a data comparison, statistical treatment & analysis to determine their significance, precision, accuracy and reproducibility according to scientific best practices established at partner laboratories.
Summarizing human and eco-toxicity results: The human toxicity of a few pure nanomaterials has been tested and of sanding dust received from complex products and composites containing NM embedded (see: D3.3 and D3.2). The toxicity of dusts (and in parallel of the pure materials) was assessed in mice in terms of genotoxicity and inflammation. In general, there was no additive effect when adding nano-TiO2 to paints or CNT to epoxy. Results suggest that the toxicity of nanomaterials is masked when embedded in a product (paint or epoxy) matrix. In contrast, the toxicity of nano-ZnO was conserved when embedded in a complex window-glass treatment product (see D3.3).
Concerning eco-toxicity of the NPs analyzed, only ZnO powder showed a dose related response in acute toxicity tests by using “Vibrio fischeri” (see D3.8 and D3.3). No toxicity was observed for the 4 NFC (nanocellulose) samples: native NFC, slightly anionic, highly anionic and cationic and results indicate that surface modification applied to the test samples did not have any toxic effect.
On the basis of results received for the acute toxicity to “V. fischeri”, the tested nanoparticles can be ranked as follows: nano-ZnO: toxic; TiO2: not harmful; CNT: not harmful and NFC: not harmful. The latter is only applied to NFC samples tested in this project.
Concerning risk assessment methodologies, currently available schemes, methods and tools specific for nanomaterials have been reviewed and two of them, the “Swiss Precautionary Matrix” and the “DuPont Framework”, were selected for further study (D3.4 and D3.5). Their applicability to the NanoSustain test materials has been evaluated throughout the different life cycle stages. All materials studied are manufactured products containing nanoparticles. It was found that the “Swiss Precautionary Matrix” is a user friendly questionnaire, which might be very useful as a first approach to help establish the precautionary needs of the nanomaterials handled. In comparison with the “DuPont Nanorisk Framework”, it is relatively quick and easy to use, and also provides visual results.
WP4: Life cycle assessment (LCA) and preliminary assessment
The aim of Task 4.1 was to develop a specific process model for the application (pre-production, production, application manufacture) and use phase including all relevant material and energy flows of the 4 selected ENM. These process models were provided by using the Umberto LCA software tool and by a comprehensive literature search and a questionnaire that was sent to several manufacturing project partners. Through this, the present state-of-the-art of existing LCA studies on ENM has been reviewed, one the one hand, and quantitative benefits and specific application contexts of nanotechnology-based applications identified, on the other. Individual important life cycle steps were described and data sources investigated. The Umberto software tool is a very flexible and powerful software tool for modeling, calculation, visualization, and evaluation of material and energy flows. The results of this modeling work are reported in a specific D4.1 report. Detailed information on the methodology for LCA, on the Umberto software tool, and on the LCA database ecoinvent are presented there in more detail.
Developing a specific model for the end-of-life and recycling phases (re-use, recycling and/or final treatment and disposal) of nanoproducts was the aim of Task 4.2. For the after-use phase, re-use, recycling and/or final treatment and disposal of the products have been distinguished. The general problem is that for these particular process steps, there is still no detailed data or information available for the 4 selected nanomaterials and associated products. For the after-use phase, almost no data regarding environmental impact exists. For this reason, and to start with, only specific processes for the end-of-life and recycling phases have been examined. The individual important life cycle steps have been identified and described, data sources investigated, specific process models developed and pro-vided in the LCA software tool Umberto, and the results of this modeling work are presented in the D4.2 report (due in M12)
An essential step towards quantitative environmental risk assessment of new nanocompounds is to calculate their prospective environmental concentrations (see D4.7 and Task 4.3). By using the life cycle perspective, task 4.3 has developed a specific exposure model applicable for the selected nanoparticles. The necessary modeling was performed based on a Probabilistic Material Flow Analysis (PMFA) approach in combination with the LCA approach and the developed life cycle models (see D4.1 and D4.2). This mass balance and multi-compartment model allows to treat all parameters throughout the modeling as probability distributions. Thus, the model outcome represents a nanomaterial flow system, depicted by probability density distributions (D4.3).
Test of the guidelines and first life cycle data: As long as results from toxicological assessments are not sufficiently accurate to warrant special legal regulation of nanomaterials, their handling should be guided by a precautionary approach. But there is a need for a preliminary assessment and for a rational implementation of the ‘precautionary principle’ based on sound scientific data and knowledge indicating justifiable concern. Task 4.5 developed criteria and guiding principles for the precautionary design and for improved recyclability of engineered nanomaterials (D4.4).
Generation of Life-cycle data on nanomaterials and nanoproducts (see D4.5): NanoSustain has assessed the environmental impact of the following organic + inorganic nanomaterials (“cradle-to-gate”-LCA) and (prospective) associated products along their whole life cycle (“cradle-to-grave”-LCA):
• nanocellulose used as paper additive,
• nano TiO2 used in paint application,
• nano ZnO used in glass coatings, and
• MWCNT used in epoxy plates.

Based on a comprehensive literature search and a questionnaire sent to all manufacturing internal (UPM, NGAG) and external partners (Nanocyl, Flügger), specific process models for the application, use and end-of-life phases (recycling, treatment and disposal) have been developed as a first step (see D4.1 and D4.2).
These specific process models were the basis for the next steps of the LCA: the Life Cycle Inventory (LCI) and Life Cycle Impact Assessment (LCIA), including data interpretation (see task 4.4 and D4.5). The modeling, calculation, visualization, evaluation and analysis were carried out according to applicable environmental impacts by means of the LCA software Umberto and the Ecoinvent database. For evaluations of the life cycle inventory analysis the Cumulated Energy Demand (CED) has been used, among others. The life cycle impact assessment and evaluation of ecological considerations was based on the partial aggregation method developed by the Institute of Environmental Sciences (CML) at Leiden University, which allows a quantitative estimate of the various environmental impacts. This impact assessment approach is implemented in the Ecoinvent database and by Umberto as CML 2001 methodology (Frischknecht et al. 2007). The following impact categories were used for impact assessment: global warming potential, depletion of abiotic resources, acidification potential, eutrophication potential, summer smog potential, depletion of stratospheric ozone, human toxicity potential, and marine aquatic eco-toxicity potential.
(i) The environmental impact of the production of selected nanomaterials:
The final environmental impacts of the production of nanomaterials depend on the type of manufacturing processes including energy demand, demand of operating supplies, yield and purification rate.
In the performed case studies, a great range of factors of environmental impacts can be seen for the production of nanomaterials when compared with micro-sized materials. For example, the environmental impacts of the production of nanocellulose are greater than of conventional sulfite pulp by factors of 1.5 to 4. The reason is the additional need for energy and chemicals for the several production routes of nanocellulose. This is also the reason for the differences between the various production methods of nanocellulose. In contrast, the environmental impacts of the production of nano-TiO2 are in the same order of magnitude than conventional TiO2. In some environmental impact categories, nano-TiO2 is better than conventional TiO2. The environmental impacts of the production of nano-ZnO are much higher than of conventional TiO2. When compared with the modeled production process of pulsation reactor, the factor range is between 8 and 68. However, when compared with the modeled laboratory process of flame pyrolysis, the impact factor range is much higher.
(ii) Influence of selected nanomaterials on the environmental impact of new (prospective) applications:
The influence of the selected nanomaterials on the environmental impact of new (prospective) applications proved to be very different for the selected 4 case studies:
(1) The environmental impact of prospective paper applications with nanocellulose is primarily determined by the energy requirement of the kraft paper production and by the consumption of chemicals (see chapter 4.3 in D4.5). The production of nanocellulose has a differing influence on several environmental impact categories. For example, the cumulative energy demand may increase by 4.2% without the benefit of a reduction in weight. On the other side, the depletion of abiotic resources would only increase by 1.9% without the benefit of a possible reduction in weight.
(2) The environmental impact of nano-TiO2 containing paint applications is primarily determined by the preproduction of nano-TiO2 and of solvent chemicals (for preproduction of alkyd resin and white spirit) (see chapter 5.3 in D4.5). All other materials and processes have a very low impact between 1% and 2%. Also different benefits can be seen. For example, the global warming potential of the scenario “nano-TiO2 LC MC” is by 4.6% higher than the scenario “TiO2 white paint LC”. On the other side, for the acidification potential, the 2 nano-TiO2 scenarios are better than the conventional “TiO2 white paint LC” scenario. The acidification potential of the scenario “nano-TiO2 LC RC” is 10.2% lower than the scenario “TiO2 white paint LC”. The differences between the two nano-TiO2 scenarios illustrate the influence of the different emission estimations from the application and use phase.
(3) The environmental impact of nano-ZnO glass products is primarily determined by the preproduction of the glass with around 95%, the electricity demand of the coating production with 3.3 %, and the transport with around 1% (see chapter 6.3 in D4.5). The production of the new pro.Glass Barrier 401 coating (including the higher energy demand of the production of nano-ZnO) has almost no influence on the balance, which means that the increase of the possible product life time will generate the improvement of the cumulative energy demand. The portions of the entire balance are extremely small. A cause for this is the small thickness of the coating of 2 x 1.6 µm in relation to the 3 mm thick glass. For example, the acidification potential of the scenario “Conv. product LC1.25” is 24.93% higher than for the “Nano-ZnO product” scenario.
(4) In the case study on ‘Prospective MWCNT composite material - MWCNT in epoxy plates as rotor blades’, the environmental impact is primarily determined by the energy requirement for the production of the wind power plant and by the demand of conventional electricity (see chapter 7.3 in D4.5). In the comparison of scenario “WPP New0.15” and scenario “WPP old”, the improvement of the global warming potential is just around 4.5%. The absolute benefit results in this case (WPP New0.15 versus WPP old) in 72 t CO2-Eq/WWP. The preproduction of the MWCNT has a very low influence on the total balance, which means that the increase of the energy production efficiency of 0.15% is generating the improved global warming potential. The portions of the entire balance are very small partly due to the low content rate of 0.5% in the rotor blade in relation to the entire plant. Only 24.7 GJ-Eq is needed for manufacturing the 150 kg MWCNT.
(iii) Applications needed in the future to ensure high environmental (sustainable) benefits:
The potential and prospects for reducing the environmental load by nanotechnology based products and processes depends much on the type and level of innovation (nanotechnology generation, incremental vs. radical, end-of-pipe vs. integrated). Today most nanotechnology based applications provide incremental innovations, with many applications having a higher level of innovation still in the development stage. A varying potential for gains in resource efficiency could be shown and quantified in the performed case studies (also from a life cycle perspective). On the other hand, there is still a clear lack of data hampering development and innovation.
For future applications with high environmental (sustainable) benefits, a very good combination of following characteristics is needed:
• small content rate with better functionality,
• environmental benefit in the use phase (higher resource and/or energy efficiency),
• long-life (persistent) products, and
• nanomaterials integrated in the product matrix.

Establishing guidelines for improved recyclability and for precautionary design of engineered nanomaterials:
As nanotechnology is for the most part still in an early phase of development, there still exists, at least in principle, a large degree of freedom allowing research efforts to be steered towards sustainable development. There are great possibilities and opportunities to avoid errors and to avoid possible costs. However, as long as results from toxicological assessments are not sufficiently accurate to warrant special legal regulation of nanomaterials, their handling should be guided by a precautionary approach. There is a need for a preliminary assessment and for a rational implementation of the ‘precautionary principle’ based on sound scientific data and knowledge indicating justifiable concern. Task WP4.5 had the objective to develop criteria and guiding principles for a precautionary design and for the improved recyclability of engineered nanomaterials (D4.6).
For the precautionary design and for improved recyclability of engineered nanomaterials, a comprehensive and more holistic approach has been derived from presented approaches and related to environmental impact categories relevant for Life Cycle Assessment. This concept includes precautionary risk aspects, resource aspects and environmental impact categories.
Besides the description of the methodical approach, D4.6 also presents results for the four investigated nanomaterials and associated products (the 4 case studies):
• nanocellulose based materials and products (paper additive, industrial thickener, rheology difier);
• nano TiO2 based products (paint application), and
• nano ZnO based composites (glass coatings)
• MWCNT based products (epoxy plates; solar cells).

Based on these 4 case studies, it was shown that the developed approach allows for a differentiated consideration of precautionary design aspects, resource aspects, and environmental impact categories.
Generation of data on prospective environmental concentrations of engineered nanoparticles:
The executed computer simulations revealed that at present, when considered separately, none of the investigated applications (i.e. nano-TiO2 in paint coatings, nano-ZnO in glass coatings, MWCNT in epoxy material and cellulose) causes significant exposure to water, soil, or air systems. These findings apply even for model results computed for the maximal EN production and use scenarios. The highest nanoTiO2 PECs (2.2 ng/L) estimated for surface water represent an uncertainty factor of 500, which is smaller than the very conservative uncertainty factor of 1000 given by European guidelines (ECHA, 2008), and which reflects the predicted no effect concentrations (PNEC) used in a previous study (Gottschalk et al., 2009). An even clearer picture is obtained for nano-ZnO, where the near-zero water values (ng/L dimensions) could be compared to a conservative PNEC of 40ng/L. For the MWCNT application, the highest PECs (about 0.1ng/L) show an uncertainty factor that is 400,000 lower than the no effect level. Our PEC curves also include uncertainty factors for air and soils by orders of magnitude away from such rough eco-toxicological values (D4.7).
The received findings may exclude, with some certainty, potential risks caused by these specific EN applications for the environment. However, a general and final positive conclusions can still not be drawn since only single EN applications have been investigated that cover just a partial contribution to the total MN release and exposure to the environment (Gottschalk & Nowack, 2011). Consequently, for a general all-clear, exposure simulations are needed at a higher precision level, including levels for all relevant ENM applications. To do so, we need better data-application and specific knowledge that can be only provided by a closer cooperation with industries involved in the production (and disposal) and sales of nanomaterial-based products and applications. Ideally, such knowledge would illuminate the following:
• EN production processes and volumes;
• EN transformation when applied/embedded in products during product use, disposal and degradation;
• The form and amount of the NP release (environmental and/or workplace releases) during manufacturing and the use of nanoproducts (as powder, bounded in a matrix, free and loose on a product surface, etc.);
• Market dynamics for the specific applications of EN.

NanoSustain has generated completely new “Cradle-to-gate”-LCA data, (prospective) “Cradle-to-grave”- LCA data (D4.5) and prospective environmental concentrations (D4.7) for the 4 selected nanomaterials and associated products.
Today most nanotech-based applications are incremental innovations, with many applications having higher level of innovation still in the developmental phase. The performed LCA case studies could show a varying potential for gains in resource efficiency (also from a life cycle view). However, the current lack of real data is still hampering the LCA of EN.
For future applications with high environmental (sustainable) benefits, a very good combination of the following characteristics is highly needed:
• Small content rate with better functionality,
• Environmental benefit in the use phase (higher resource and/or energy efficiency),
• Long-life (persistent) product, and
• Nanomaterials integrated in the product matrix.

WP5: Development of innovative solutions for recycling and final treatment
Task 5.1 Laboratory studies to explore new solutions for recycling
The work in Task 5.1 (Laboratory studies to explore new solutions for recycling) comprised the following subtasks:
a) Production of nanocellulose-based material and associated end-products (see D5.4)
b) Organic recycling of nanocellulose materials (see D5.1)
c) Recycling of ZnO containing glass by melting (see D5.6)
a) Production of nanocellulose: A nanocellulose standard sample was produced and obtained from UPM (Partner 12) and characterized (see D5.4). This standard sample served as an internal standard for quality control, test validation, and reproducibility check for the planned laboratory experiments (degradation, composting etc.). The standard sample contained about 2% cellulose and 98% water. The sample was produced by high-shear friction grinding with a Masuko Sangyo’s Supermasscolloider. The grinding resulted in cellulose fibrils that formed an opaque and stable, well-dispersed, aqueous suspension with shear-thinning behavior. Optical microscopy showed that the suspension contained visible particles (fibrils) of different sizes, most of which were smaller than 20 µm in diameter. Truly individualized cellulose nanofibrils were not visible in the optical microscope. FESEM image analysis results showed that the most common fibril width was between 20 and 30 nm. The results of this sample are reported in D5.4 “Development of an internal nanocellulosic standard sample for quality control, test validation, and reproducibility check”. Also nanocellulose-based end-products i.e. papers, have been received from UPM for experiments on dustiness (in WP3) and for experiments on biodegradability and organic recycling (in WP5).
b) Organic recycling: To recycle packaging materials and plastics, e.g. by composting, the following requirements must be met:
• these materials must be biodegradable according to standards EN 13432 and EN 14995
• they need to be degradable during composting and
• there should not be any harmful influence on the composting process and compost quality.

The suitability of nanocellulose-based materials for composting was studied and evaluated and both nanocellulose films and paper products containing nanocellulose were tested (see D5.1). Biodegradability of the nanocellulose-based products was tested by the biodegradability test in controlled compost test according to EN 14046. During biodegradation, compost samples were taken to evaluate their eco-toxicity by the “V. fischeri” bioluminescence- test. Disintegration during composting was evaluated according to EN 14045. The obtained results suggest that all tested products can be considered as biodegradable under the used compost conditions and that they also disintegrate in the performed pilot-scale composting experiments. In addition, none of the compost samples was toxic towards “Vibrio fischeri”. In aquatic environment, nanofibrillar cellulose (NFC) gels showed a strong tendency to agglomerate, which in turn reduced the rate of biodegradability. However, it should be noted that the biodegradability test used was not optimal for gel-like products and that biodegradability should be evaluated by using also other tests.
c) Recycling of ZnO containing glass: The industrial partner Nanogate AG synthesized several batches of coated material, called “ barrier 401”, with and without ZnO-nanoparticles. A cleaning and pre-preparation step of glass sheets for coating was necessary to achieve a good optical coating quality. Glass sheets were treated in a clean room to take representative samples for project partners. In addition to the coatings, liquid samples and ZnO-powder were also delivered to the partners. Available data sheets and technical information were also provided for partners that are doing characterization and testing. At the end of the first period of the project, the outcome was discussed by all involved partners and recommendations given on the achieved deliverable regarding the recycling of coated glass (see D5.6).
Nanoparticles emitted during heating and melting of the window glass coated with “pro.Glass Barrier 401” provided by Nanogate were investigated at VTT (Partner 3) (see report to D5.6). In comparison, window glass coated only with sol-gel binder matrix was also investigated. As a reference, a plain window glass sample without any surface treatment was heated / melted. To determine the possible background particle concentration originating from the induction furnace material(s) used for heating/melting of the glass samples, an experiment with an “empty” furnace (i.e. no glass sample, only an empty crucible) was also carried out. The particle number and mass concentration as well as the number size distribution were measured during the heating/melting process mimicking the established recycling process of the product. The glass samples were cut to approx. 5x2-5 mm rectangles, and the mass of the heated/melted sample was mglass»7 g. The temperature range studied was 700-1500 °C. The main conclusions of this work were that particles are emitted during heating/melting of the glass. However, the number and mass concentration of emitted particles does not depend on whether a coating is applied or not, or on the type of coating. A notable increase in particle numbers began at temperatures > 1000 °C, and an increase in the particle mass began at temperatures > 1300 °C. It was further noted that also the size of the particles does not depend on whether coating is applied or not, or on the type of coating. The particle size began to increase at temperatures > 1200 °C. In the emitted particles, sodium (Na) was clearly enriched, whereas the amount of silicon (Si) was decreased in all of the particle samples. The amount of Zn was almost the same in the particles from uncoated as in the “pro.Glass Barrier 401” coated glass. The study, methods and results are reported in detail in D 5.6 (Recycling of ZnO nanoparticle containing glass), which is almost finalised. The first draft of the deliverable was ready by mid-October (M18) and was sent for review to Partner 11 (Nanogate), as requested. Nanogate made some additional valuable suggestions for further experiments. Including these into the report will improve the quality of the deliverable, making it a more useful document for future dissemination. A new update was ready on 21 December 2011 and the deliverable is expected to be finalised in January 2012.
Task 5.2 Laboratory studies to explore new solutions for final treatment
The work in Task 5.2 comprised the following subtasks:
a) Incineration of MWCNT (results presented in D5.2)
b) Release of nanoparticle under landfill conditions (results presented in D5.3/M33)
c) Modeling the transport of NP in environmental media (results presented in D5.5/M36)

(a) Incineration of MWCNT composites: Materials that cannot be recycled often end up in waste incinerators to generate heat and electric energy and so recover the energy contained in the waste. The number of incinerators has increased in Europe due to the ban of land-filling of untreated municipal waste. One such material that is not suitable for recycling is multi-walled carbon nanotube (MWCNT) containing epoxy composite. During the reporting period M19-37, a carbon nanotube (CNT) contain-ing composite was combusted together with wood chips as supporting fuel in a 40 kW solid fuel furnace with a step-grate burner to simulate energy material recycling. Three different fuel compositions were prepared and mixed: wood chips with 20 wt%, 5wt% and 0 wt% CNT containing composite. In addition, combustion conditions for the 5 wt% CNT containing composite were varied by adjusting the amount of air fed to the furnace: i.e. the lower amount of air simulated poor combustion conditions. The temperature of the feed air was much lower at poor combustion than under good combustion conditions, approximately 100 °C and 250-300 °C, respectively. The average combustion temperatures in the furnace were 700-800 °C during good combustion with 20 wt% of CNT containing composite and approximately 950-1050 °C for other conditions (5 wt% CNT).
Nanoparticles were observed in all combustion cases independent of the fuel composition according to the new EU nanoparticle definition (EU 2011) . However, the fraction of the nanopar-ticles of the measured ones varied depending on the composition of the fuel. Number concentration was highest for good combustion with 0 wt% CNT containing composite and mass concentration for poor combustion with 5wt% CNT containing composite. No indication of CNT like tubular structures was found in the particle samples analyzed by SEM/EDS. This was probably due to the low amount of the CNT containing composite in the fuel mixture, the low amount of CNT in the CNT containing composite, and the formation of a large and hard, highly sintered bottom ash deposit that may “bind” and immobilize the species in the CNT composite in a non-volatile matrix.
The Raman spectrum of the particles collected on filters during good and poor combustion of wood chips and 5 wt% CNT composite did not indicate the presence of CNTs. However, evidence of a nano-structured carbon that was not MWCNT, and neither pure graphite nor pure graphene, was found. The spectra of the samples resembled the spectrum of a disordered graphite that may be nanostructured. However, no graphite-like sheets were found in the SEM images. CNT like tubular structures were not either found in bottom ash samples analyzed by SEM/EDS. The Raman spectra of the bottom ash collected after good combustion of wood chips and 20 wt% CNT composite did not indicate the presence of CNTs. However, evidence of a nano-structured carbon that was not MWCNT, and neither pure graphite nor pure graphene, was found. The spectra of the sample re-sembled the spectrum of disordered graphite that may be nano-structured. However, the portion of CNT containing composite was too high for optimal combustion conditions, thus creating a large and hard deposit on the grate of the furnace. For the 5 wt% CNT composite, no indication of the presence of CNT was found. The spectrum indicated the presence of disordered amorphous car-bon and graphitic carbon atoms. The resulting sample of the 5 wt% CNT composite with poor com-bustion showed a too strong fluorescence for the analysis. No carbon structures (e.g. graphite, CNT, fullerene, etc.) were found in the bottom ash from composites without any CNT. This does not, however, mean that sample would not contain carbon at all. The chemical composition of particles and bottom ash indicated that Ca and Mg were enriched in the deposit, K was enriched in the particles but also found in the bottom ash, and Zn, Cu, S and Cl were enriched in the particles, but hardly any was found in the bottom ash (<< 1 wt%). Summarizing, first lab-scale incineration experiments indicated that CNTs may not be released in the combustion air or in the bottom ash after combustion of MWCNT containing epoxy composites, but degraded and transformed into various types of nano-structured and/or non-carbon particles, aggregates and compounds. None of the combustion cases presented evidence of CNT like tubular structures in the emitted particles (unfortunately, the particle sample for the 20 wt-% CNT case was missing because of the instability of the combustion process).
Results from these experiments can be summarized as follows:
According to the EU definition of a nanomaterial (1-100 nm) (EU 2011), nanoparticles were observed in all combustion cases independent of the fuel composition. However, the fraction of the measured nanoparticles varied depending on the composition of the fuel. Particle number concentration was highest for good combustion with 0 wt-% CNT containing composite, and highest mass concentration was found for poor combustion with 5 wt-% CNT containing composite. But no indication was found for CNT like tubular structures in the particle samples analysed by SEM/EDS, probably due to the low amount of CNT containing composite in the fuel mixture, the low amount of CNT in the CNT containing composite, and/or the formation of the large and hard, highly sintered bottom ash deposit that may bind and immobilize “CNT species” in the composite into a non-volatile matrix. Also, no CNT like tubular structures were found in bottom ash samples. But more research is needed to finally verify these findings.
Also the Raman spectrum of the particles collected on filters from the combustion gas during good and poor combustion with wood chips and 5 wt-% CNT composite confirmed that the absence of CNTs. However, there was evidence of nanostructured carbon that was however neither MWCNT nor pure graphite, or pure graphene. Raman spectra taken of the bottom ash collected after good combustion of wood chips and 20 wt-% CNT composite also confirmed the finding that CNTs did not survive the combustion process. But again evidence for a nanostructured carbon compound was found that was neither MWCNT nor pure graphite or pure graphene. The spectra of the sample resembled the spectrum of disordered graphite that may be nanostructured. Unfortunately, the por-tion of CNT containing composite was too high for optimal burning conditions, why a large and hard deposit was formed on the furnace grate. Also no indication for the presence of CNT was given for the 5 wt-% CNT composite. Again, the spectrum indicated the presence of disordered amorphous carbon and graphitic carbon atoms. During poor combustion the sample showed a too strong fluorescence to do the analysis. The resulting bottom ash did not show any CNT or other carbon structures (e.g. graphite, fullerene, etc.).
These primary experiments give first indications that incineration of MWCNT-containing products may be a promising approach that may allow the recovery of the energy contained in the waste and at the same time the elimination of possibly toxic CNT particles or other CNT-like compounds. But further research is needed to elucidate and optimize the process, also for other CNT containing products. The achieved laboratory results are very preliminary and need further experimental verification before any up-scaling or developing of a prototype.
(b) The Landfill Directive gives general principles for evaluating the waste disposal on landfills. The acceptance criteria established in Council Decision 2003/33/EC are primarily limit values for the release (leaching) of predominantly inorganic species from the waste and, to a limited extent, maximum compositional values for some other parameters, e.g. organic content. No limits have been given for total content of inorganic compounds (e.g. harmful metals), or for nanoparticles. NanoSustain tested the performance of standard leaching tests and the release behavior (leachability) of nano-ZnO powder mixed with glass beads under various laboratory conditions (see D5.3). The results were compared to tests carried out with micro-sized ZnO. Furthermore, additional tests were done with nano-ZnO coated glass. Testing of granular waste containing nano-ZnO was done by using modified standard methods, such as the percolation test CEN/TS 14405, batch test EN 12457 and the pH dependence test (e.g. CEN/TS 14997), developed to determine the leaching of soluble com-pounds from granular waste. The tank test procedure for surface leaching proved not suitable for nano-ZnO coated glass, because the release of Zn was strongly solubility controlled. For this reason, the leaching behavior was best determined by percolation tests. The results indicated that the release of Zn and the toxicity of the eluates decreased with increasing L/S ratio. Results in particular from the pH-dependent test showed that the release of Zn from nano-ZnO coated glass is strongly solubility controlled and influenced by the pH of the test solution. Tests with DOC additions resulted in a higher release compared to tests without DOC most probably due to organic complex binding of the dissolved Zn2+ by humic acid. Furthermore, results emphasize that the salt concentration of the leachate hampers the release of Zn most probably due to increased agglomeration processes, emphasizing again the fact that agglomeration of nanoparticles strongly affect the fate (e.g. mobility and retardation) of nanomaterials in landfills. More research is needed on the long-term behavior of the formed agglomerates. In total, the release of Zn from nano-ZnO coated glass was strongly pH controlled being lowest at pH 9.2 but was also affected by other environmental master variables that also steered the experimental conditions. Leaching of Zn from nano-ZnO decreased with salt concentration due to increasing NP agglomeration and deposition and increased with the addition of DOC, which enhanced the dissolution and mobility of Zn.
Eco-toxicological studies of the eluates obtained from the leaching experiments and using the acute toxicity Vibrio fischeri assay proved the close correlation between the toxicity of the eluates and the dissolved Zn concentration suggesting that the observed toxicity comes from Zn ions leached rather than from the nano-ZnO particles. However more tests are needed to evaluate the ecological relevance of long-term leaching of NP under various landfill conditions as well as their bioavailability and ground water migration, also in relation to dissolution and aggregation and agglomeration processes.
(c) Soil and water management and environmental protection require a profound knowledge of the evolution of water and solutes in the soil subsurface. For this a large number of models have been developed during the last decades to simulate water flow and contaminant transport in saturated and unsaturated soils. NanoSustain has used a first modeling approach to describe the release and transport of ZnO nanoparticles in soil and groundwater by analyzing two mechanisms of diffusion: (1) diffusion without pressure applied on the surface and (2) injection with a pressure of water solution containing nanoparticles. For both mechanisms, profiles of the downward distribution of ZnO nanoparticles have been evaluated and the preliminary results indicate that the flux and transport of ZnO NP in sandy and clayey soils is mainly controlled by solute sorption processes, the permeability of the soil and by the velocity of the moving groundwater (see D5.5).
From the performed experiments, the following main results and conclusions can be summarized:
(a) The biodegradability of nanocellulose based paper products was tested by composting according to EN 14046. During the biodegradation test, compost samples were taken and tested by the V. fischeri bioluminescence test to evaluate their acute eco-toxicity (see D5.1 and D5.4). The results obtained show that degradation of nanocellulose products by composting is feasible and that no harmful degradation products are formed. However, further testing and up-scaling is needed and further optimization of the composting/biodegradation process for the development of a final prototype.
(b) A carbon nanotube (CNT) containing composite sample was combusted together with wood chips as a supporting fuel in a solid fuel furnace with a step-grate burner to explore the performance of incineration as a final waste treatment option (see D5.2). The results obtained show that nanoparticles (1-100 nm) were observed in all combustion cases independent of the fuel composition. Particle number concentration was highest for good combustion with 0 wt-% CNT containing composite, and mass concentration for poor combustion with 5 wt-% CNT containing composite, respectively. There was no indication for CNT like tubular structures in any of the test samples collected and analysed by SEM/EDS and Raman spectroscopy. Particles collected on filters during good and poor combustion did not indicate the presence of the CNTs, although there was evidence of nanostructured carbon that was not MWCNT. Also, no CNT like tubular structures were found in bottom ash samples. The experimental and analytical results show that incineration may be a suitable nano-waste treatment technique both to recover the energy contained in CNT containing products and to eliminate the release and/or formation of possibly adverse compounds from the combustion of CNTs. However further up-scaling and testing is needed to optimize process conditions and further identification, characterization and testing of the compounds formed during combustion, and so to support the development of a prototype.
(c) The leachability of nano-ZnO under landfill conditions was evaluated from tests carried out with powders mixed with glass beads under various laboratory conditions. The results were compared with tests carried out with micro-size ZnO. Furthermore, additional tests were done with nano-ZnO coated glass (see D5.3). Results showed that the leachability of granular waste containing nano-ZnO can be done with a few modifications by existing standard test methods, such as the percolation test CEN/TS 14405, batch test EN 12457 and pH dependence test (e.g. CEN/TS 14997). The tank test procedure for surface leaching did not work, because the release of Zn is solubility controlled. Experimental results from the pH dependence test emphasize that the release of Zn from nano-ZnO coated glass is strongly pH and solubility controlled, but that also agglomeration (and aggregation) processes may influence the final destiny of Zn NP (e.g. their mobility, transport and retardation) in landfills. In particular the salt concentration had a significant effect on the particle size and large agglomerates were formed with particle size ranges of 500-2000 nm. Although pretreatment by sonication could break these agglomerates down to smaller particles (200-500 nm), a further breakdown was not possible as these particles are bound by stronger forces and would require more energy to shatter. The results received from the percolation test showed a decrease of the release and toxicity of Zn from nano-ZnO with increasing L/S ratio. Also tests with DOC resulted in a higher release compared to tests without DOC. But more knowledge is needed on the long term leaching behaviour of nanoparticles in products and of formed agglomerates in landfills. As the release of Zn from nano-ZnO coated glass was much pH dependent, disposal in contact with other waste materials having a low or high pH should be avoided.
(d) Results from the preliminary modeling approach used to describe the movement of NP in soils and groundwater indicate that the flux and transport of ZnO NP in sandy and clayey soils is mainly controlled by solute sorption processes, the permeability of the soil and by the velocity of the moving groundwater (for more details see D5.5).
The results obtained in WP5 could demonstrate that current principles and technologies designed and used for waste recycling, treatment and disposal, and for assessing the toxicity and release behavior of waste materials may be also applicable to NM containing products and for their safe and sustainable handling, recycling/reuse, final treatment and disposal. However, more research is still needed to elucidate the practical implications, but also to further adapt, modify and further develop the performance of these technologies, in particular to optimize the underlying operational and technical processes needed to control the release of NP and to safeguard the final fate of nanomaterials.
Potential Impact:
Nanomaterials are included in products because of their novel properties. However these novel properties mean that effects on human health and the environment may be different from the equivalent conventional materials. Correlating possible adverse effects with material physicochemical characteristics is essential to identify potential areas of risk within a given product’s life-cycle and to take appropriate action to minimize this (through reducing the hazard and/or exposure by modifying the material property that may cause such a risk). NanoSustain has performed pc characterization on nanomaterials representing relevant stages of their life-cycle. These included pure nanomaterials and nanomaterials incorporated into products and matrices, but also sanding dusts, combustion gases and ashes from incineration and melting; weathering and abrasion products, compost and leachates.
One main purpose of NanoSustain was to analyse the life-cycle of different products on or close to market that contain engineered nanomaterials (ENM), for the purpose of identifying areas of potential hazard and exposure, and to better understand the true sustainability profile of such products. NanoSustain investigated three prominent ENMs that are manufactured in large quantities across the globe (TiO2, ZnO and MWCNTs) and a fourth (nanocellulose) which is relatively new, but as it is derived from sustainable sources and has applications in paper, packaging and composites; has enormous potential for future product development. NanoSustain investigated four broad life-cycle phases: production of the input ENM, production of the product containing the ENM, normal use of such products, and end-of-life disposal or recycling of the product. At each stage, materials used and released were studied for toxicological and ecotoxicological effects, and estimated material and energy flows (input and output) were calculated.
Results and possible societal implications can be summarized as follows: In particular the release experiments gave new scientific evidence that nanoparticles (TiO2 and MWCNT) embedded in a product matrix may be less reactive than free nanoparticles, which however has to be proved on a case by case (as ZnO NP embedded in glass coating showed). In this context, we need to know more about different exposure scenarios throughout the lifecycle and we need to study biological effects of complex mixtures with NM before we can make valid predictions on safety of NM at all the different stages of a lifecycle. The LCA results show that the environmental impact from nanocellulose containing paper in terms of energy and material flows appears slightly greater than for conventional paper mainly due to more intensive manufacturing processes. A net benefit compared to conventional paper may be feasible if a 10% weight reduction of cellulose required for the resultant paper can be achieved. Also, nanocellulose was biodegradable and not eco-toxic (although human toxicity testing was not possible due to the gelling nature of nanocellulose). Risks to human and the environment seem to be low. Concerning nano-TiO2, environmental impacts seem to be quite similar with micro-sized TiO2 paint in terms of energy and material needed for production, use and disposal, and even less for certain nano-TiO2 containing paints. Dust from sanding of nano-TiO2 paint showed lower inflammatory responses than pure nano-TiO2, which may suggest that human risks are limited when pure nano-TiO2 is embedded in a product. However, the large volumes manufactured and the stability of TiO2 in nature may give rise for indirect (environmental) exposure. More research is needed to assess the long-term behavior and final fate of nano-TiO2 when released during use and disposal. Environmental impacts occurring along the product life cycle of nano-ZnO coated glass proved much higher than for conventional ZnO mainly due to manufacturing, but may be reduced due to the longer lifespan of nano-ZnO coated glass compared with conventionally surface treated glass. Sanding dust released from coated and uncoated glass did not show any significant difference in particle mass or number for nano-ZnO coated, not nano-ZnO coated or completely uncoated glass. However, the toxicity of ZnO was conserved when ZnO was included in a window-glass coating product. Also the well known relationship between Zn toxicity and dissolved Zn ions released from the nano-ZnO coated glass could be confirmed: the tested nano-ZnO containing glass coating produced similar toxic effects as known for Zn ions in aqueous solution. Leaching tests simulating landfill conditions showed a significant release of Zn from both nano and micro-sized ZnO particles at low and high pH, while increasing salt concentrations retarded and DOC increased the release indicating that the behavior of Zn in the nano-ZnO is solubility controlled. This means that higher human health risks may arise rather from manufacturing than during the use phase, provided that safe handling and disposal of glass is maintained. Epoxy resin boards were considered as product for the LCA of MWCNT and their use in wind-turbine blades. Although production of MWCNT is energy intensive, environmental impacts in terms of energy and material flows were slightly lower than for conventional composites with an estimated 0.15% increase in energy production over a 20 year period due to better strength and durability. However, the known human toxicity of certain (in particular thin and entangled) pure MWCNT forms was confirmed, although sanding dusts from epoxy composite boards containing MWCNT showed (as for nano-TiO2) no toxicity in mice. Incineration as a likely way of disposal of MWCNT containing epoxy boards showed no CNT-like structures in the resulting combustion gas, fly or bottom ash. In general LCA suggested that improving the resource and/or energy efficiency of manufacturing processes may enhance the sustainability of the use and disposal of EN.
Concerning practical implications and future RTD and innovation needs, the results obtained from the technical lab-scale experiments suggest that (1) composting may be an appropriate solution for the or-ganic recycling of nanocellulose and associated paper products as a good biodegradability was demonstrated and end-products were not (eco-) toxic. But further upscaling for the development of a prototype and additional toxicity testing are needed. Also (2) incineration proved to be a good measure for the energy recovery and final treatment of CNT containing waste, in particular as no CNT-like structures have been found in the resulting combustion gas, fly and bottom ash. But again, further RTD work is needed to verify these preliminary positive results and their application to other CNT containing products. Likewise, (3) melting appears to be a suitable technique for the safe recycling of glass coated with nano-ZnO barrier surface as resulting emissions did not differ from the melting of uncoated glass or when treated with a conventional coating. Also the composition of the recycled (melted) glass did not differ from the original glass product. Again, more research is needed to find out if the technology is also applicable to larger scales and to better understand and control the emission of NP at the high temperatures used (700-1500 oC). Finally, (4) the use of standard leaching tests provided new insight in what processes and factors may steer the behaviour of NP containing waste in landfills. Although the release of Zn from nano-ZnO coated glass was strongly pH and solubility controlled, also other environmental parameters, such as DOC, salt content and L/S ratio, affected the leaching behaviour of nano-ZnO. In demineralized (low salt content) water, less Zn was released from nano-ZnO coated glass than from micro-ZnO coated glass suggestion a strong influence of the salt concentration and hence of agglomeration processes on the final release of Zn. The strong influence of the salt concentration on NP agglomeration may hamper the release of dissolved Zn and Zn NP under landfill conditions. However, more research is needed on the long term behaviour of the formed agglomerates. Another practical implication comes from the strongly pH controlled leachability of Zn from nano-ZnO coatings, which means that their landfilling with wastes having a low or high pH should be avoided.
What did NanoSustain find?
NanoCellulose – test product was paper. Environmental impacts (in terms of energy requirements, material input and output) are greater for nanocellulose containing paper than conventional paper on an equivalent weight basis. This is mainly due to the more intensive processing required to produce nanocellulose in the first place. However, a net benefit compared to the conventional product will be realised if a 10% reduction in the weight of cellulose required for the resultant paper is achieved. With regards to hazard and exposure, nanocellulose was biodegradable and compostable, and no ecotoxicity was identified (toxicological testing could not be performed due to the gelling nature of nanocellulose). Thus risks to human health and the environment are low.
TiO2 – test product was paint. Environmental impacts (in terms of energy requirements, material input and output) are similar for paints containing nano- or micron-sized TiO2, and in some cases improved for the nanoscale TiO2 containing paints. In this case the higher energy and material requirements for the production of the nanoscale TiO2, compared with conventional TiO2, are negligible compared to the overall paint life-cycle. With regards to hazard and exposure, results indicated far lower inflammatory responses to paint sanding dusts compared with pure nanoscale TiO2, which will limit risks to human health. However, the relatively large volumes being manufactured and the stability of TiO2 in the environment must both be considered for environmental exposure.
ZnO – test product was coated glass. Environmental impacts (in terms of energy requirements, material input and output) are much higher for nanoscale ZnO than conventional ZnO due primarily to the manufacturing processes required. However, nanoscale ZnO coated glass would be expected to have a significantly longer lifespan than glass coated with current (organic) UV-blocking coatings, which would be expected to reduce the environmental impacts. With regards to hazard and exposure, Zn in ionic form demonstrates toxicity, and it has been observed that nanoscale ZnO rapidly dissolves in various aqueous solutions releasing Zn ions. Results from NanoSustain confirmed this toxicity and ecotoxicity (and that it correlates with the concentration of ionic Zn). The aqueous barrier coating, containing nanoscale ZnO, produced similar effects to solutions containing Zn ions. Sanding dusts were difficult to produce from the coated glass, however no significant difference was observed in the number or mass of nanoparticles emitted from coated vs uncoated glass during a typical melting cycle used for recycling. Experiments simulating landfill conditions indicated that low and high pH’s significantly increased the release of Zn from both nano and micron sized ZnO particles. Increasing salt concentration retarded, while the presence of dissolved organic carbon in the media, increased this release. This indicates that the greatest risk will occur during the manufacture of the coated glass rather than use, provided that safe handling and disposal of glass is followed.
MWCNT – test product was epoxy resin boards, and for life-cycle assessment (LCA) the use of such composites in wind-turbine blades was considered. Although the production of MWCNT is an energy intensive process, the overall environmental impacts (in terms of energy requirements, material input and output) when used in a wind turbine blade were calculated to be lower than conventional composites; if an estimated 0.15% increase in energy production was realised over a 20 year period (due to improvements in strength and durability). With regards to hazard and exposure, MWCNT in certain forms (in particular thin and entangled, but also with certain chemical modifications) have demonstrated toxicity in animal models. This has been confirmed in NanoSustain, however sanding dusts from MWCNT epoxy boards showed no toxicity in mice. Disposal of MWCNT epoxy resin boards is likely to be by incineration, and in tests using different weight percentages of MWCNT in epoxy resin boards under different combustion conditions, no CNT-like structures were identified in the combustion gases or ash.
In all cases the materials assessed were analysed using a variety of means to define their physicochemical characteristics. This had a dual purpose: to allow a greater understanding of those properties more critical to the effects seen, and which tools and techniques were more robust for the measurement of these properties.
In summary, when considering the environmental (and sustainability) benefits of including ENMs in products, the following aspects need to be considered:
• Using the minimum content of ENMs to achieve the desired functionality, as manufacturing ENMs generally requires higher production costs in terms of energy and materials than conventional materials.
• Improving the resource and/or energy efficiency of the product during the use phase, as this is where the largest environmental impacts are.
• Improving the useful life-span of the product and or decreasing the servicing requirements or failure rates.
• Embedding ENMs within a solid product matrix, as this masks potential toxic/inflammatory effects.
How did the Project Engage with Policymakers and other Key Stakeholders?
These results have been shared with key stakeholders including research organisations, industry and policymakers through: over 60 presentations by project beneficiaries, including many international conferences such as Nanotech Italy, EuroNanoForum and NanoSafety Cluster meetings; 6 peer-reviewed articles, 2 book chapters, 11 publications in conference proceedings and 4 articles in other press.
In addition to these activities, the project organised dedicated workshops to inform key stakeholders of the rationale behind the project and its results: Glasgow May 2011 (after the first year of the project), Venice November 2011 (at Nanotech Italy) and in Barcelona May 2013 (in a joint workshop with NanoPolyTox and NanoFate).
Policy and decision makers are exposed to vast amounts of information with limited time to process this, and so NanoSustain summarised its results in a more focused and condensed form (factsheets and case studies). These provide an entry point to the project and importantly signpost end-users to the relevant project partners. NanoSustain remained actively disseminating information to the wider community through its networks, and even in July 2013 (M39) there were still individuals downloading factsheets and case studies.
How will the results from NanoSustain be used?
The industrial partners NanoGate and UPM will incorporate the results of NanoSustain in their production processes particularly in terms of understanding the stages in their product’s life-cycle which must be carefully considered for risk (to human health and/or the environment), or which could be improved to decrease material and/or energy impacts (particularly in production of the ENM). Nanologica has gained useful experience in the characterisation of different ENMs which can be employed in the manufacture if ENMs within its product range that possess consistent characteristics. The academic partners have further increased their knowledge and expertise in the assessment of ENMs in different environment and at different life-cycle stages, all of which can be usefully employed to support the safe and responsible development of other nanomaterials and other products containing ENMs.

List of Websites:

Participant Main Contact Country Email
NOMI Rudolf Reuther SE
ION Mark Morrison UK
NRCWE Anne Thoustrup Saber DK
VTT Minna Vikman FI
UniHB Michael Steinfeldt DE;
VN Stefano Pozzi Mucelli IT
JRC David Rickerby IT
KTU Valentinas Snitka LT
IMT Monica Simion RO
NLAB Rambabu Atluri SE
NGAG Rolf Danzebrink DE
UPM Päivi Korhonen FI