Final Report Summary - ENFIRO (Life Cycle Assessment of Environment-compatible Flame Retardants ( Prototypical case study)) Executive summary:Some brominated flame retardants (BFRs) have unintended negative effects on the environment and human health. Less toxic alternatives appear to be available already but comprehensive information on their possible toxicological effects are lacking. The European Commission-funded project ENFIRO investigated the substitution options for some BFRs and compared the hazard, exposure, fire, and application performances. Based on these results risk and impact assessments were carried out. In total 14 halogen-free flame retardants (HFFRs) as alternatives for decaBDE, TBBP-A, and brominated polystyrenes were selected. These flame retardants were studied in five applications - printed circuit boards (PCBs), electronic components, injection moulded products, textile coatings and intumescent paint.Project Context and Objectives:ENFIRO follows a prototypical case study approach, in which the alternative FRs are evaluated regarding their flame retardant properties, their influence on the function of products once incorporated, and their environmental and toxicological properties. The main objectives are:- To deliver a comprehensive dataset on viability of production and application, environmental safety, and a life cycle assessment of the alternative FRs. - To recommend certain flame retardant/product combinations for future study based on risk and impact assessment studies.ENFIRO evaluated viable substitution options for a number of BFRs (decaBDE, TBBP-A and brominated polystyrene). There are several non-brominated FRs existing on the market. However, there is limited information available about their environmental and toxicological impact. Furthermore, the alternatives should not be applied before tests have shown that they do not adversely affect the quality of consumer products. ENFIRO approachENFIRO follows a practical approach in which HFFRs are evaluated and compared to BFRs regarding their flame retardant properties, their influence on the function of products once incorporated, and their environmental and toxicological properties. This is achieved by performing screening and case studies, which will gather a comprehensive set of information on environmental behaviour and toxicological impact, as well as an assessment of the performance of the FR in a specific application. The case studies will give recommendations for industrial and governmental stakeholders for the replacement of BFRs and viable alternative FRs.The ENFIRO approach developed follows a chemical substitution cycle anchored in four major elements. In the first element the alternative HFFRs are prioritized and the most viable alternatives are selected. These flame retardants were studied in five applications - printed circuit boards (PCBs), electronic components, injection moulded products, textile coatings and intumescent paint. The second major element focussed on the technical performance (fire and application), hazard and exposure assessment of the selected HFFRs. Finally, the collected information was analysed in a comparative hazard and risk assessments (third element), and in combination with information on costing and socio-economics of the HFFR/products the outcome was digested in impact assessment studies using life-cycle assessments (LCA) (fourth element). This finally resulted in a recommendation of certain HFFR/product combinations. ENFIRO used a unique approach to assess the data at three different levels: the chemical (flame retardant), material and the product.In conclusion, ENFIRO used the full cycle of the chemical alternative chain and developed a novel three level assessment approach based on the flame retardant, material, and product, by the comparison of the alternative FRs with the BFRs:i) Flame retardant: hazard characterization, exposure and risk assessmentii) Material: fire performance and application studiesiii) Product: impact assessment (life cycle assessment (LCA), market study, social life cycle assessmentENFIRO had the following objectives:1. Collect information on the availability of alternative FRs, their characteristics in relation to fire safety regulations, environmental behaviour, possible toxic effects, economic aspects, compatibility with polymer production, and impact on the function and reliability of end products. 2. Select substitution options for specific BFRs based on this pre-study and prioritize FRs for further study in a small number of case studies.3. Technical assessment studies on application requirements regarding production properties and application functions.4. Technical assessment on five alternative FR/product combinations; printed circuit boards, electronic components, injection moulded products, textile coatings and intumesent paints.5. Determine the toxicological effects and environmental behaviour of the selected FRs.6. Determine the technical qualities of the FRs and their behaviour and possible effects when incorporated in products 7. Perform a risk assessment based on all environmental and human hazard information from the toxicological and environmental, exposure, fate and modelling studies of the alternative FRs.8. Determine and predict the social and economic effects of replacing the specific BFRs by the selected alternative FRs. 9. Perform life-cycle assessment (LCA) analysis to advice on the safe production and use of one or more of the alternative FRs studied. 10. Recommend certain FR/product combinations for future study based on LCA, LCC and risk assessment studies.11. Disseminate the knowledge to stakeholders (producers, formulators, users), environmental organisations and policy representatives.These objectives are the backbone of the project, and ENFIRO is organised in 9 work packages. The scientific WPs focus on the prioritization and selection of alternative FRs (WP2), the hazard characterization (WP3), exposure, fate and modeling (WP4), flame retardant capability studies (WP5), application studies (WP6), risk assessment (WP7), and impact assessment (WP8). WP 1 is dedicated to management and WP9 to dissemination. Project Results:Prioritization and selection (WP2)As the first phase of ENFIRO a prioritization and selection of alternative flame retardants was carried out. The main objective was to select and prioritize a range of non-brominated FRs that are viable alternatives to specific commercial BFRs on the market through literature and other reliable scientific sources based on how they affect the material's characteristics of the polymers that are flame retarded. Such characteristics included compatibility, electrical properties, and various ageing properties and was based on already available data on toxicity, exposure risks and environmental fate. This results in the assessment of viability criteria for specific FR applications that consist of flame retarded marketable polymers.Fire performance (WP 5)The objective of the fire performance studies was to quantify the severity of the toxicity, smoke and heat flux of alternative HFFRs against BFRs in fire (smoldering and flaming) incidents.The measurements selected to assess the flammability and toxicity of BFR substitutes are: tendency to dripping, solid degradation in mg scale, gaseous products in mg scale, cone calorimeter in standard atmosphere, special calorimeter in controlled atmosphere, species production in modified ISO TS19700 tube furnace. These properties have led to the assessment of alternative FRs and comparison with BFRs by quantitatively assessing the following parameters: tendency to dripping, (low) heat release rate, late ignition, strength of char, (low) smoke yield and production rate, (low) toxicity and corrosion. Based on these properties the global effects of these materials in fire have been addressed by quantifying their behaviour in standard tests (UL94, LOI), their behaviour in large fires, impact on life and property safety and damage.A large number of FR/polymer materials were compounded based on the selected HFFRs. The selection of BFRs is done according to their dominance in the market for each of the base polymers selected i.e. PBT/GF, PC/ABS, epoxy resin, PA66/GF, EVA and PPE/HIPS. Alternative FRs included phosphorus FRs (e.g. metal phosphinates), inorganic tin-based FRs, nanoclays and combination of nanoclays with phosphinates. Details of formulations were compiled and used to select the prototype base polymers and FR for compounding. Polymer compounding and UL94 characterisation has been done utilising twin-screw extrusion and injection moulding (for ABS formulations) and high-shear mixers / curing in RTV silicone moulds (for epoxies). The FRs involved in this study includes BFRs (benchmark), phosphorus FRs (phosphinates), inorganic tin-based FRs, nanoclays and combination of nanoclays with phosphinates. A large number of FR/polymer combinations have been compounded (PA6.6 PBT, PPE/HIPS, PC/ABS, epoxy resin encapsulates, EVA), for the FR/polymer combinations. Polymer formulations have been optimized using UL94 test as regulated for industrial applications. A thickness of 3.2mm for UL94 tests was first chosen but then replaced by 0.8mm to address more demanding industrial applications. Large batches of compounded materials (pellets and moulded plates) were prepared for the fire performance, application, leaching, and air emission studies. Assessment of flammability and toxicityThe flammability and toxicity of thermoplastic retardant polymers and thermoset materials were assessed. In addition, the intumescent coating prepared by one of the partners were tested for flammability and application (weathering, adhesion) properties. All formulations were investigated in TGA/FTIR/DSC/ATR and cone calorimeter. Based on these results and analysis, a method was developed to characterize both the UL94 and the fire behaviour of materials using parameters related to fire growth and smoke production deduced from data obtained from cone calorimeter experiments. This method is outlined for PBT +GF and similar figures for all materials were made.The use of nanoclay (nano-MMT) combined with Alpi (PG3B) reduces the characteristic fire growth by 60% in comparison to the formulation containing only Alpi (PG4A). The nano-MMT formulation also provides a characteristic fire spread growth in the same order of magnitude as the halogenated formulation (brominated polystyrene, PG2), but also yields around 15% lower smoke. It is noted that all formulations have smoke yield higher than the desirable 0.05g/g. The efficiency of combustion provides useful information to assess the production of toxic species. Thus, it is considered that the formulation with the BFR (PG2), with an efficiency of combustion about 0.45 produced much more unburned species than the HFFR formulations (PG3B and PG4A) having an efficiency of 0.85. In conclusion, although the Alpi formulation (PG4A) could be a possible alternative to the brominated formulation (PG2), the formulation of Alpi with nanoclay (PG3B) is the best alternative (even superior to brominated formulation PG2) regarding the characteristic fire spread growth and smoke yield.Severity of the toxicity, smoke and heat flux of BFRs against alternative FRs in realistic fire (smoldering and flaming) incidentsMajor results are outlined in the following summary with the focus on PBT with glass fibre (GF) as an example of the results. In addition, a discussion of the tube furnace results is also included in this summary.Intumescent coatingPure HIPS are difficult to flame retard to V(0) without halogen. A novel intumescent coating has been developed by one of the partners and was investigated for flammability and toxicity by applying it to HIPS. The materials was either waterborne (better when possible) or solvent-borne coatings, white (or anyway opaque layers) or transparent thin films. In parallel, significant issues of weathering (e.g according to ISO EN ISO 4892-3) and adhesion were addressed. The developed intumescent coating system have proven efficient and makes HIPS fulfill V(0) and fulfill the glow wire test. This could have significant commercial potential. General conclusions for toxicity based on results from the tube furnaceA simple method was developed for characterising the toxicity of polymers using the effective heat of combustion. This parameter assessed the inefficiency of combustion by comparing unburned hydrocarbon compounds and possibly, their toxicity. Namely, we compared the actual heat of combustion (obtained from cone calorimeter) divided by the theoretical heat of combustion and the results was subtracted from one. The developed methodology was used to compare results from Cone calorimeter with the results from the tube furnace. Due to the complexity of the analysis, only PG (PBT + GF) formulations were studied in tube furnace in great detail. All major permanent gases evolved by means of FTIR coupled to the tube furnace were identified. We have shown the existence of highly toxic hydrogen bromine evolved from formulations containing BFRs. The toxicity parameter was compared with concentrations of carbon monoxide and methane between various formulations. It was shown that (especially for methane) the findings based on the inefficiency of combustion in Cone calorimeter were valid for overventilated conditions but were not so obvious during underventilated conditions and pyrolysis under nitrogen. This will require further studies.Summary fire performanceThere is no single drop in replacement for BFRs by HFFRs available for the polymer systems. We have investigated primarily thermoplastics, but also thermosets (epoxy) and elastomers (EVA). Important input was received from the Stakeholder forum on the formulations. The used formulations were optimized only for UL94 performance and not for other properties.Applications studies (WP6)The overall objective was to perform technical assessments of the use of alternative FRs in various applications by comparison with traditional FR systems. The applications were printed circuit boards, electronic components, injection moulded products, textile coatings, and intumescent paint.Printed Circuit Boards (PCBs) The objectives were to identify and assess the reliability risks with selected HFFRs for use in printed circuit boards using a physic-of-failure approach. That is, possible failure mechanisms caused by the HFFRs were identified based on how the products were manufactured and used for some typical applications.The main failure mechanisms in PCBs identified that may be affected by the flame retardant used are:- Cracking of metal platings in plated through holes (PTH barrels) and internal interconnect failures between PTH barrels and conductors in inner layers- Fracturing of the resin and adhesion failures between the resin and other materials- Current leakage due to formation of conductive anodic filaments- Current leakage due to decreased surface insulation resistance and electrochemical migrationOne of the objectives with ENFIRO was to evaluate the viability of some specific HFFRs in various products. To fulfil this objective, it was necessary to get information from the laminate manufactures of the HFFRs used in their laminates. However, the type of HFFR used in commercially available laminates is in most cases considered proprietary information and it was only possible to get information of the HFFR used from one laminate producer. This laminate contained a mixture of DOPO, ATH, boehmite and phosphazene. In addition to the laminates from this laminate producer, laminates with HFFRs from three other laminate manufacturers were also evaluated. A laminate with TBBPA was used as reference. Four test methods were chosen for the technical assessment of the laminates with HFFRs. These were:- Interconnect stress test (IST) for testing the impact on the reliability of PTH barrels and interconnects between PTH barrels and inner layers.- DELAM test for testing the impact on fracturing in the resin and adhesion failures between the resin and other materials- CAF test for testing the impact on formation of Conductive Anodic Filament (CAF)- ECM test for testing the impact on Surface Insulation Resistance (SIR) and Electrochemical Migration (ECM)The results from the reliability assessments of the laminates containing HFFRs were generally as good as for the reference laminate containing TBBPA or better. The laminates with HFFRs showed especially good resistance against formation of CAF, the failure mechanism that is perhaps most likely to be affected by the flame retardant used.Electronic Components The objectives were to identify and assess the reliability risks with selected HFFRs for use in encapsulated electronic components using a physic-of-failure approach. A complicating factor that affected the work on electronic components was that the aim of ENFIRO was to study commercially available HFFR alternatives to brominated polymer systems for IC encapsulation. It turned out after the initial search that the brominated systems had been replaced with highly cross linked epoxy systems without flame retardants. Due to reliability concerns with halogen-free flame retardants, especially phosphorus-based, component manufacturers are very reluctant to use these in encapsulation materials.Thus, there was really no substitution case to study. The contents of this deliverable therefore were changed from the original plan and additionally more effort was directed to tasks on printed circuit boards.Injection Moulded Products Within the ENFIRO project alternatives to traditional brominated flame retarding systems for thermoplastics were developed and evaluated. Some of these alternate systems call for quite substantial amounts of additives to the polymer matrix to be effective. Such large contents, up to 30 percent per weight, will affect the properties of the products as well as the processability of the material. A total of 13 UL-94 optimised (V0 pass) products of alternative flame retarding systems have been compared to their unaltered base plastics concerning their mechanical properties. The tensile and impact behaviour as well as processability of the injected moulded materials has been tested.Textile Coatings The objective was to evaluate the HFFRs for textile coating applications where currently BFR are used. The efficiency of flame retardants is dependent on the textile polymer systems. Therefore, two different fibre types were investigated with two different coating polymers frequently used on the market. Polyamide (PA) weave and polyethylene terephtalate (PET) weave (also referred to as polyester) are used as filament plain weaves. The coating polymers were water based emulsion systems without cross linkers. The polymers in the two emulsions are acrylic respectively polyurethane. A reference was also made for comparing the studied systems with best practise. This best practise was composed with decaBDE/antimony trioxide (ATO) system. Dispersions with alternative FRs (APP, MPP, PER), coating of substrates, and fire testing of the coatings were performed. Test vehicles were tested for fire retardant behaviour, peel adhesion strength between weave and coating, tensile properties of pure coating and friction. Representative textiles were used in this study and test vehicles were fire tested according to appropriate fire standards, required for the specific use. Dispersions of acrylic and polyurethanes were used for coatings on PET and PA weave.Results showed that for suitable flame retardancy for PUR on PET weave 30% of MPP is needed. The combination with APP and PER is not more effective. A formulation of three HFFRs (MPP+APP+PER) gives improved extinguishing compared to decaBDE/ATO. This HFFR combination is suitable for PUR on PA weave. The minimum amount of HFFR needed is 20% in solid coating, and the effectiveness is similar to decaBDE.ATO. Acrylics on PET weave can be flame proofed with 30% APP, but the combination with MPP and PER is not more effective. In this case also the combination of MPP+APP+PER gives similar extinguishing compared to decaBDE/ATO systems. For acrylics on PA weave none of the tested HFFRs seem to be effective.Tensile tests were performed on the coating according to modified SS-EN ISO 13934-1:1999. Bromine containing formulations show high tensile strength and maintained or improved elongation at break for both PUR and acrylic. The HFFR formulations are good for acrylic but poor elongation at break for PUR. The test also showed that the bromine formulations make a more flexible coating which is an advantage in many cases. The peel tests showed that optimised HFFR formulations (MPP+APP+PER) with PUR on PET weave had a 57% drop compared to decaBDE/ATO systems. The PUR on PA weave system performed better with the HFFRs and showed only a 17% drop of peeling compared to the decaBDE/ATO. Interestingly, the acrylic coating on PET and PA weave gave no adhesion with the decaBDE/ATO system but the MPP, APP, PER system gave adhesion. Summary application performanceFor textiles, the developed HFFR formulations have slightly lower coefficient of friction than the bromine containing formulations both for polyurethanes and for acrylics. In many applications reduced friction is a positive factor at use. The combination with MPP+APP+PER is good for PUR on PA weave, minimum is 20% in solid coating. Tensile test mainly shows a drop in elongation at break compared to decaBDE/ATO. For all flame retardant plastic formulations (both HFFR and BFR) equal or better performance on processability for injected moulding products was found. The tensile and impact properties are roughly similar for brominated and halogen-free flame retardant formulations. The results of HFFRs in the printed circuit boards showed as good or better results compared to the BFRs.Hazard characterisation (WP3)The objective of the hazard characterisation was to perform ecotoxicological studies of selected HFFRs using water and sediment toxicity tests and to perform a health hazard characterisation of the HFFRs on a molecular and cellular level, with emphasis on geno-, endocrine-, and neuro-toxicity using in vitro studies and a limited number of ex vivo validation studies. It was found that a great number of the selected HFFRs have a very poor solubility in water and in organic solvents. As a result considerable efforts have been made to test different methods to dissolve the FRs in water or organic solvents. For these FRs so called water-accommodated fractions were prepared resulting in maximum water soluble concentration that was used for the toxicity tests.Acute toxicity to Daphnia magnaSeveral compounds were not acutely toxic, five of them (MPP, Mg(OH)2, RDP, ZHS and ZS) showing no effect at their water solubility (Sw) (EC50 greater than Sw) and two of them (ATH and BDP) showing 25-26% effect at Sw. For the FRs that were toxic below their water solubility, clear dose-response relationships were observed. Alpi, APP and DOPO showed a low acute toxicity. ATO was classified as moderate toxic. TPP and TBBP-A exerted a high toxicity within 48 hours to Daphnia magna (less than1 mg/L). All concentrations, except of TBBPA, are measured with either ICP-AES or HPLC-MS/MS. TBBPA is tested nominally. The classification is based on the REACH criteria (European Union, 2006; European Union, 2008).Some specific remarks must be added with respect to the results. Firstly, the nanoclay Cloisite was tested at a nominal concentration of 100 mg L-1, which caused 100% immobility. Since the exact chemical composition of this material is unknown, it was not possible to determine the actual exposure concentration, thereby preventing classification of this HFFR. For seven compounds ATH, BDP, RDP, MPP, MgOH2, ZS, and ZHS the EC50 is higher than their water solubility and therefore could not be established. CytotoxicityCytotoxicity of organic and inorganic HFFRs was studied in rat liver cells using MTT and LDH leakage assays. Cells were exposed to 10 µM of the organic FRs and compounds with low solubility were tested at the maximum solubility in water. The results showed that no effects were observed for the FRs on rat liver cell viability. Cell respiration assays were performed to assess effects of FRs on bacterial viability and cellular respiration processes. The results indicate that at high concentrations (100 µM) TBBPA, DOPO and to a lesser extent TPP, RDP and BDP induce a limited reduction in cell respiration. At up to 10 µM, none of the organic FRs tested had any effect on cell respiration, with the exception of TBBPA that already inhibited cell respiration at 1 µM. Endocrine disruption and mutagenicityAs some brominated FRs are powerful disruptors of the thyroid hormone axis, the organic HFFRs were tested for their potency to displace T4 from the T4 carrier protein transthyretin (TTR). Of the BFRs that were able to displace T4, TBBPA being more potent (IC50 = 26 nM) than DecaBDE (IC50 greater than 25 µM), but all organic HFFRs had no effect. Neurotoxicity For risk assessment purposes, chemicals are often classified according to the different categories of potential harm that they may cause. An example of a widely used CLP classification classify existing data on specific endpoints in relation to potential risks during their use by the European Union (European Union, 2006; European Union, 2008), which is based on LD50 (half maximal lethal dose) or EC50 (half maximal effective concentration) values. However, as complete concentration-response curves are often absent for in vitro tests used in ENFIRO, e.g. because of low solubility of the test compounds of many HFFRs, a rank ordering of the tested HFFRs based on the lowest observed effect concentration (LOEC, potency) and maximal effect size per tested endpoint was made. These combined results were used to make an overall assessment of the in vitro neurotoxic potencies of the selected HFFRs.Summary hazard characterisationBased on literature information, databases, and the ENFIRO hazard assessment seven of the selected HFFRs showed to have less issues of toxicity concern (APP, Alpi, ATH, MPP, DOPO, ZS, ZHS), with the remark that Alpi showed moderate chronic aquatic toxicity, than some BFRs. Two HFFRs (RDP and BDP) are of some concern as these show varying results between aquatic toxicity studies in the literature (moderate-low and high-low toxicity, respectively). This variation may be due to the amount of TPP present in the technical products; TPP is a by-product and known to be very toxic for aquatic organims (classified as H400, H410 (M = 1) by nearly all notifiers). In addition, BDP is a persistent compound. Another compound that is of concern and needs further study is the nanoclay (nano-MMT) that showed a strong in vitro neurotoxicity effect. Also the fate (leaching) of nanoclay from polymers needs further study.Exposure, fate and modelling (WP4)The major objectives were i) to collect physical-chemical property information for the selected HFFRs, ii) perform modelling of these to produce initial and updated environmental exposure assessments, iii) identify the most important knowledge gaps, carry out experiments to fill these gaps (determine water solubilities of the organic HFFRs, study leaching of HFFRs from polymers to water and emissions from polymers to air), iv) to perform persistency tests, and v) to collect samples for the field monitoring study and analyse these. Modelling and exposure assessmentsThe literature was first mined for information relevant to the sources, physical-chemical properties, degradation rates, environmental occurrence and environmental behaviour of six selected organic HFFRs being considered for the case-study. Gaps in data for key physical-chemical properties and degradation half-lives were estimated using available structure-property relationships. The information for the organic compounds was then synthesized using existing multimedia fate modelling tools to produce an initial environmental exposure assessment. In a next step, physical-chemical properties of the three organic HFFRs included in the case study (DOPO, RDP, BDP) were updated and their environmental fates were reassessed. Two fugacity-based multimedia models were adopted, i.e. the equilibrium criterion (EQC) model (version 2.80.1) TaPL3 model (version 3.00) and Low Resolution Multi-Species (LoResMS; version 1.0) model for the reassessment. Model predictions indicate that the emission mode (environmental media receiving emission) has a large influence on fate and distribution in the environment. Comparison of model results from the previous and updated study suggested that the updated changes in physical-chemical properties (particularly predicted water solubility) largely altered the predicted environmental fate and distribution of two of the three selected organophosphorus FRs, i.e. BDP and DOPO. New evidence suggested that DOPO is a weak acid and ionisable under environmental conditions. Persistency testsIn a first step, a ready biodegradability experiment was performed based on the OECD 301 guideline. An automated respirometer was used to study the mineralisation of Alpi, BDP, DOPO, RDP, MPP, TPP and TBBPA. Full degradation was monitored by measuring CO2 with the respirometer hourly for 28 days. The compounds were added to diluted sewage sludge at concentrations of about 100 mg L-1. None of these compounds fulfilled the ready biodegradability criteria (greater than 60% degradation in 10 days after start of mineralisation) at these concentrations. In the vitality control, activity of the sludge microorganisms was confirmed by rapid mineralisation (ready biodegradability) of glucose. However, toxicity controls revealed toxicity to the microorganisms at 100 mg L-1, as limited CO2 production was observed in these vessels. This should have been the same rapid mineralisation as in the vitality controls. Due to a limitation of the set-up, the flame retardant responsible for the observed toxicity could not be elucidated. Therefore the conclusion of these experiments was that Alpi, BDP, DOPO, RDP, MPP, TPP and TBBPA are not ready biodegradable or one of them is/they are all toxic at 100 mg L-1 to the microorganisms of the waste water treatment plant. Air emission and leaching to waterThe results of the initial assessment identified the need for measurement of chemical release to indoor air from the application (FR-treated plastics or textiles) at elevated temperatures to be expected from electronic equipment in use or in a car sitting in sunlight. Due to widely varying water solubilities for DOPO, RDP and BDP predicted from the modelling studies, it was determined that these should be determined experimentally. There was also a need to measure the leaching of FRs from materials to the outdoor environment (water) as it is difficult to identify and trace back the inorganic FRs in the environment due to the non-specific character of metals (e.g. Al, Zn or Sn can have many different sources). For air emissions, the flame-retarded polymer to be studied was placed in a glass petri dish that was then placed at the bottom of a metal can. A pre-cleaned polyurethane foam (PUF) disc was suspended in the opening of the can using a collar of aluminium foil and the lid of the can replaced to form an airtight seal. The can was then placed either in a laboratory at room temperature or in a constant-temperature oven at a precise temperature for different lengths of time. At the end of each time period, the cans were removed from the oven and the PUFs were extracted and analysed. In a first step, the flame-retarded plastics were tested at a high temperature (80 oC) to see if there was any emission at all of FRs. After 7 d at 80 oC, measurable amounts of DOPO, RDP and decaBDE were found in the PUFs. The emission rates for the HFFRs and decaBDE were similar, ranging from less than0.5 to 2.0 pg/cm2/d. To simulate a more realistic situation, polymer plates were then tested at 40 oC for 3, 10, 15, 29 and 58 d, and at room temperature (22 oC) for 58 d. No DOPO, RDP or BDP was detected in PUFs after 58 d at 40 oC or 22 oC. However, TBBPA was detected in PUFs in the experiments performed at 40 oC (emission rate of 1.2-2.2 pg/cm2/d) as well as at 22 oC (emission rate of 0.03 pg/cm2/d). Thus, the HFFRs present lower risk for emissions to air from polymers compared to TBBPA.Field monitoring of HFFRsBased on the results from the updated environmental exposure assessment, key environmental media to be sampled for the field monitoring campaign were identified for the three organic HFFRs (RDP, BDP, DOPO). Due to their water solubilities and predicted behaviour in the modelling exercise, these included sewage sludge, sediments, STP effluent water, house dust and indoor air. Monitoring of the inorganic HFFRs in the outdoor environment was not carried out due to the difficulty to identify the source of the non-specific character of the metals. Based on this information sampling campaigns in several European countries were carried out for the organic HFFRs only. Risk assessment (WP7)An environmental (ERA) and a human risk assessment (RA) were carried out for a selected number of HFFRs. The risk assessment consisted of four major steps, i) hazard identification, ii) effect assessment, iii) exposure assessment and iv) risk characterization. The hazard identification and effect assessment were based on (eco)toxicity data from the literature, ECHA, and the ENFIRO hazard characterization. For the ERA Predicted No Effect Concentrations (PNEC) for water (PNECaqua) and sediment (PNECsediment) were taken from ECHA, but also calculated for water using the latest ecotoxicity information resulting in an ENFIRO PNECaqua. For the exposure assessment the Predicted Environmental Concentration (PEC) was calculated based on i) the leaching data of HFFRs to water, and ii) measured concentrations of RDP and BDP in sediment and STP effluent. The sediment and water data is limited to organic HFFRs only, but the leaching data is available for all selected HFFRs. PECs were calculated for water and sediment at the local and regional scale. For the local scale it is assumed that the discharge of effluent from STPs is the main source for environmental exposure of HFFRs. It is assumed that HFRRs will mainly enter the environment by leaching and volatilization from plastics during the end-user, product use or waste phase and finally entering the environment via STPs effluent or by air. It is assumed that the emissions at the production phase are low compared to the other life cycle phases. For the regional scale a worst case scenario was used assuming that all polymers with HFFRs occurring in the STP effluent will reach the fresh and marine environment.Impact Assessment Studies (WP8)The impact assessment studies comprise an environmental, social and economical component. The environmental component is most developed while methods for economical and social impact assessment studies are in general not yet well developed. The ENFIRO project contributed substantially to the development of the Social LCA along the UNEP/SETAC guidelines. The Env-LCA (LCA) has been conducted according to the procedures described in ISO14044. Each Life-Cycle Impact assessment study is divided into 4 phases:1. The goal and scope phase, in which the purpose of the study is stated, the level of detail and study boundaries are defined, and methodological choices are made.2. The life cycle inventory phase (LCI phase) results in an inventory of input/output data with regard to the system being studied. It involves the collection of the data necessary to meet the goals of the defined study.3. The life cycle impact assessment phase (LCIA) has the purpose to transform the large body of data on inputs and outputs from the LCI into a limited number of environmental effect scores.4. In the Life cycle interpretation phase, the results of the LCIA are summarized and discussed as a basis for conclusions and recommendations, in accordance with the goal and scope definition. In addition, the robustness and validity of the data and results are checked by performing sensitivity analyses.Environmental impact assessmentThe Env LCA findings have lead to the following conclusions.- In most phases of the life cycle of FRs, fossil energy use related impact categories dominate the LCA score: Climate change, Fossil depletion and Particulate matter formation.- The life cycle phases in which human toxicity and ecotoxicity play the largest role are: - Export of WEEE followed by improper waste treatment. In this waste scenario, the formation of dioxins during improper incineration of BFR containing plastics has the largest contribution to human toxicity.- Emissions of FRs during volatilization (or wearing/abrasion) in the use phase have LCA scores only in the toxicity impact categories. Emission factors of FRs are considered to be low.- During accidental fire, emission of ATO to air has a relatively high score for terrestrial ecotoxicity. However, when considering the complete life cycle, only a small fraction of the laptops will end up in an accidental fire, and therefore accidental fire has only a small contribution to the total score.- For the waste scenarios MSWI incineration and landfill, contributions to the toxicity impact categories come from both FR related (ATO, bromine, ZHS, ATH and Alpi) and non-FR related emissions (heavy metals).- The environmental impact in the production phase of FRs (cradle-to-gate, per kg) varies considerably. The highest impacts are found for ZS, ZHS and ATO. Lower total impacts are found for decaBDE, RDP, BDP, DOPO and Alpi, with differences in total scores within 20%. Then followed by TBBPA, MPP and BPS, and the lowest impact is found for the production of ATH. For the three FRs with the highest environmental impact (ZHS, ZS, ATO), the raw material mining phase has a high contribution to the total score.- For the production of flame retarded polymers (cradle-to-gate, per kg), differences in environmental impact between BFR and HFFR are not very large, with maximum differences of 16%. - Emission of FRs in the use phase through volatilization, wearing or abrasion cause the highest impact on human toxicity and freshwater ecotoxicity for the BFR scenario, and the highest impact on terrestrial and marine ecotoxicity for the HFFR scenario. In the overall score, the BFR scenario has the highest score, through the contribution of human toxicity.The emissions of FRs in this phase have only a small contribution to the overall impact over the complete life cycle, but this phase is still likely to be the most important exposure route for humans.- In the case of accidental fire, the BFR scenario has a higher overall impact than the HFFR scenario due to a higher rate of smoke formation and a higher terrestrial ecotoxicity score. In the HFFR scenario, the score for Climate change is higher than in the BFR scenario due to higher CO2 emissions (more complete combustion).- Of the four End-of-Life scenarios for WEEE, the option 'export followed by improper treatment' has the highest environmental impact for both the BFR and HFFR scenarios. In the BFR scenario, this high impact is mainly caused by the formation of (brominated) dioxins during improper WEEE incineration. The high LCA score for improper WEEE treatment in the HFFR as well as the BFR scenario shows that even when BFRs are substituted by HFFRs, these practices are still quite harmful, as there is a range of toxic emissions during improper treatment, including lead, arsenic, hydrogen fluoride, (chlorinated) dioxins and PAHs.- The main differences between the two full life cycle scenarios of the laptop with BFRs and with HFFRs are found in the scores for the impact categories human toxicity, freshwater, marine and terrestrial ecotoxicity and metal depletion. For these five impact categories, the impact is lower in the HFFR scenario. For the other impact categories, the scores of both full life cycle scenarios are almost equal.Social impact assessmentNo fully fledged out S-LCA methods are available yet and the field is still in its infancy. In this project the UNEP/SETAC guidelines were followed and the method developed in a study carried out by Ciroth and Franze (2011) was substantially improved by a further developing of the social life-cycle data inventory and the impact assessment methodology. Because of these challenges and the limitation of the study undertaken, results presented here should be cautiously interpreted. But although much work remains, the present research does show that by using the UNEP/SETAC S-LCA guidelines, social hotspots for a complete life cycle can in principle be determined and compared between different product alternatives. The primary aim of this study was to compare social hotspots between brominated and halogen free flame retardants over the complete life cycle. The results suggest that for both laptop alternatives concerns regarding social issues are present in all life cycle stages and affect all of the included stakeholders (workers, local communities and society). However, in both the laptop containing HFFRs and the BFR laptop the social hotspots were predominantly found for the extraction of raw materials and the improper treatment of e-waste in developing countries. In general the well-being of workers and the local community seems to be adversely affected by the highest number of social hotspots. The lowest number of social hotspots was identified for the HFFR production phase. When taking the complete life cycle into account, findings show no clear differences in the total number of social hotspots found between the two scenarios.An analysis of the characteristics of social hotspots influencing Economical Impact assessmentThe economical impact is mainly governed by the market price of the raw materials. The market for FRs keeps on growing mainly due to increased fire safety regulations. However, the growth is unevenly distributed. The market of brominated flame retardants in the EU is declining due to both restrictions, economical and social factors. Consequently the HFFR market will mainly benefit from this growth. For almost all polymer applications used in E&E products marketed HFFR options are available and some of them are already cheaper to apply than their BFR alternative. The price of some FRs is highly volatile, some of them depend on the availability of resources and some of them may have international trade restrictions. For example the declining market relevance of ATO can be explained by its volatile price changes and current restrictions on the export from the main raw material producer China. Its price has increased in Europe from $2/kg in 2002 to $5-6.50 in 2007 and to around $15 in 2011. Such a vast increase in price will increase the price of FR systems where ATO is used. This may drive substitution processes for ATO alone or for complete new FR systems. Differences in EU and LCA impact assessment studiesIn the ENFIRO impact assessment on BFR substitution, several assessment methods were used: a qualitative assessment using questionnaires (based on the EU guidelines for impact assessment), an environmental LCA, a social LCA and a market study. The impact assessment methods differed in methodology, system boundaries and level of detail. The EU guidelines are developed to assess the impact of policy changes on the sectors involved. The EU guidelines consist of large questionnaires divided into categories. However, most of these categories are also found in the LCA-approach. The questionnaires were sent to members of the ENFIRO Stakeholder Forum at the start of the project. The outcomes were then compared to the results of the LCA assessment methodology. Summary ENFIRO Science and Technology (S&T) resultsThe Project Approach- ENFIRO showed that viable alternative flame retardants are available with similar fire performance and technical application capabilities as some BFRs, and confirmed that some are environmentally friendly alternatives to brominated fire retardants and pose less risk for the environment and humans. - ENFIRO followed a practical approach in which HFFRs were evaluated and compared to BFRs regarding their flame retardant properties, their influence on the function of products once incorporated, and their environmental and toxicological properties. - The ENFIRO approach is based on the chemical substitution cycle which consists of four major elements: i) prioritization and selection of alternatives, ii) technical, toxicological, and exposure assessment, iii) risk assessment, and iv) impact assessment. - ENFIRO showed that it is important to follow the completed substitution chain based on the four above elements, in contrast to most substitution programs which focus on the first two elements only. - ENFIRO followed a unique approach to assess data at three levels, i) the flame retardants (hazard, exposure, risk), ii) the material (fire performance, technical applicability, leaching and air emissions), and iii) the product (impact assessment including LCA). - ENFIRO showed that all of the selected alternative halogen free flame retardants do fulfil the regulatory fire test. Results for Target Applications - A simple method was developed for characterizing the fire performance and fire toxicity of FR/polymers and to compare alternative FRs with BFRs. - An important finding was that halogen free systems have clear benefits as demonstrated, e.g. less visible smoke, in some cases lower peak heat release rate with halogen free products, and less toxic components in smoke. - There is no single drop-in replacement for BFRs by HFFRs available for the polymer systems. However, for all polymer systems investigated a HFFR option was found that was commercially available and fulfils the fire requirements. - For the polymer blends (PC/ABS, PPE/HIPS) the HFFRs do meet the fire performance requirements but performs less well than the material with BFRs for which it has been formulated, and there are some concerns on the environmental hazard. Additional work is needed to search for other alternative FRs for the polymer blend systems.- Both polymers with brominated and nonbrominated FR showed similar loss in mechanical properties compared to the polymer alone. - All formulations (both HFFR and BFR) showed equal or better performance regarding processability for injection moulding. - An important part of the project was the input received from the Stakeholder forum on formulations. - The results for the PCBs with the HFFRs where as good as or better compared to the reference PCBs produced using BFRs. - A novel intumescent coating system was developed for pure HIPS, which is difficult to flame retard without halogen, that was efficient and makes HIPS fulfill the fire regulatory tests as well as the industry fire standards relevant to the electronics industry. Hazard and Risk Assessments - From the initial selection of 14 alternative flame retardants seven (APP, Alpi, ATH, DOPO, MPP, ZS, ZHS) were found to be of less concern. - Bioaccumulation of the inorganics HFFRs is probably not a concern, but BDP is persistent. - Environmental fate models predicted that the organic HFFRs would be found primarily in soils, sediments and dust and to a lesser extent in water and air. - Controlled air emission experiments showed that all organic HFFRs emitted from polymers at elevated temperature but not at lower temperatures. - The used leaching methods are suitable to provide a measure for the leaching of FRs to the outdoor environment.- Leaching experiments showed that HFFRs and BFRs can leach to water. For some polymers no differences in leaching behaviour between BFRs and HFFRs were found, but for some the HFFRs systems had higher leaching properties than the BFR (e.g. polymeric based FRs). The type of polymer and the porosity are the main parameters determining the leaching behaviour.- Analysis of dust samples from microenvironments where these organic HFFRs might be used showed highest concentrations on and around electronic equipment, such as flat-screen television sets. Lower, but measurable concentrations were also found around other electronics, furniture, car seats and in apartments. - Some of the organic HFFRs were also found in the environment (STP effluents, sewage sludge and sediment). - Based on these results, it is clear that humans can be exposed to HFFRs via dust ingestion and organisms via sediment and STP emissions.- The risk assessments showed that some HFFRs show less risk for the environment and human health. The lower risk is mainly due to the lower hazards of the HFFRs, and probably not due to a lower exposure. - Leaching of flame retardants from polymers should be further reduced to reduce the human and environmental exposure.Life Cycle Assessments and Socio-economic considerations- The Env-LCA study showed that for improvements of the life cycle environmental performance of FRs, the waste treatment phase is critical. Export and improper treatment of WEEE has the highest impact of all waste treatment options for both the BFRs and HFFR scenarios, and efforts should continue (or be intensified) to reduce the amount of European WEEE ending up in this scenario. - The social hotspots were predominantly found for the extraction of raw materials (health and safety issues) and the improper treatment of e-waste in developing countries (fair salary, social security, health and safety issues).- For the overall severity of social impacts between HFFRs and BFRs small differences in impact scores were found, mainly for the extraction and production phase. - From an economic viewpoint price differences between polymer systems with BFRs or HFFRs are small and should be no obstacle.- The most viable alternative HFFR/polymer combinations from a fire performance.- For the substitution of chemicals a complete substitution cycle is needed: technical/application performances, hazard, exposure, and impact assessments. Such an assessment can only be performed with a group of experts from different disciplines (material experts, fire safety researchers, toxicologist, chemist, social scientist, life-cycle experts etc).Overall it can be said, that the approach adopted by ENFIRO was very successful and can be used for similar substitution studies, e.g. REACH.ReferencesCiroth. A., Franze, J. (2011). LCA of an ecolabeled notebook: consideration of social and environmental impacts along the entire life cycle. GreenDeltaTC GmbH, Berlin.European Union (2008) Regulation (EC) No 1272/2008 of the European Parliament and of the Council on classification, labelling and packaging of substances and mixtures. no1272/2008:1355.European Union (2006) Regulation (EC) No 1907/2006 of the European Parliament and of the Council concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency. no1907/2006:849. Hendriks, H. S., van Kleef, R. G., van den Berg, M., and Westerink, R. H. (2012a). Multiple novel modes of action involved in the in vitro neurotoxic effects of tetrabromobisphenol-A. Toxicol. Sci. 128, 235-246. Hendriks, H. S., van Kleef, R. G., and Westerink, R. H. (2012b). Modulation of human a4?2 nicotinic acetylcholine receptors by brominated and halogen-free flame retardants as a measure for in vitro neurotoxicity. Toxicol. Lett. 213, 266-274. Waaijers, S.L. Kong, D., Hendriks, H.S. de Wit, C.A. Cousins I.T. Westerink R.H.S. Leonards, P.E.G. Kraak M.H.S. Admiraal W., de Voogt P., and Parsons, J.R. 2013. Persistence, Bioaccumulation, and Toxicity of Halogen-Free Flame Retardants. Rev Environ Contam Toxicol. 222, 1-71.Potential Impact:DisseminationThe major activities of the ENFIRO dissemination was the launch of the Website, logo, flyer, newsletters, presentations, publications, organisation of the mid-term and final workshop, the establishment and meetings with the ENFIRO Stakeholder Forum (ESF), and the ENFIRO film. Press releases in English, Dutch and Swedish were released to inform the project start. A large number of presentations were given about ENFIRO in national and international symposia by several partners. For 2013 several presentations are planned to further disseminate the work at conference, workshops and publications. ENFIRO was presented at special sessions on chemical alternative assessments at conferences of SETAC North America 2010, BFR2010, SETAC Milano 2011, SETAC Berlin 2012, BFR2013, and SETAC Glasgow 2013. ENFIRO was also presented at the Going Green Care Innovation conference visited by a wide range of experts from industry, academia, NGOs and policy.The ESF was established consisting of 17 companies/institutes; 7 FR producers, 6 formulators, and 4 others (e.g. NGOs, waste recycling). Yearly ESF meetings with the ENFIRO consortium were established. The ESF functioned as a reference group for the identification, elaboration and evaluation of the drivers and barriers connected to the FR substitution process. Beside the drivers and barriers inventory, the input of the ESF in the ENFIRO project guaranteed a broad dissemination of the feasibility of successful substitution, and assures their active involvement and commitment to the required substitution process. At the first ESF meeting the ENFIRO approach, concepts, and the selected set of HFFRs was presented and discussed. The ESF members provided feedback on the selected HFFRs, i.e. 3 HFFRs were added, and additional information on physical-chemical properties and toxicity were provided. At the second meeting information of the HFFRs on fire performance, application, toxicity, and impact assessment was provided, and information for the impact assessment studies was exchanged. At the third meeting the main outcomes of the project were discussed and additional information for the impact studies were requested for specific flame retardants and polymers. The ESF was invited for the mid-term and final ENFIRO workshops. Economical and social impactThe economical impact is mainly governed by the market price of the raw materials. The market for FRs keeps on growing mainly due to increased fire safety regulations. However, the growth is unevenly distributed. The market of brominated flame retardants in the EU is declining due to both restrictions, economical and social factors. Consequently the HFFR market will mainly benefit from this growth. For almost all polymer applications used in E&E products marketed HFFR options are available and some of them are already cheaper to apply than their BFR alternative. Furthermore some of the HFFRs possess at least an equivalent market relevance as its BFR option. The price of some FRs is highly volatile, some of them depend on the availability of resources and some of them may have international trade restrictions. For example the declining market relevance of ATO can be explained by its volatile price changes and current restrictions on the export from the main raw material producer China. In most E&E products, price differences between polymer systems with BFRs and HFFRs are small or even absent. Therefore it is unlikely that any price differences of end-products where these FRs are used are caused by differences in the applied FR systems.Risk assessment and REACHThe ENFIRO project indicates that a full risk assessment (RA) for each HFFR-alternative is not feasible due to data gaps. However, a full RA may not be required since for substitution of the substance only a comparison between potential health and environmental risks must be made. This comparison can be based on the precautionary approach by establishing concern levels rather than hazard levels. In this case screening tests, as used in ENFIRO, are useful tools to establish quickly concern levels and defining the best screening methodology in a substitution process is then important. Screening tests were used in the ENFIRO project to quickly compare relevant hazard endpoints. The results do not lead to a chemical safety assessment (CSA) as outlined in the REACH but quickly indicates at least a health or environmental concern level of the alternatives in relation to the BFR used in the system. Therefore screening tests and assays can fulfill a crucial role in this comparison when dealing with substitution issues.Currently the RA methodology is hazard driven rather than exposure driven. However there is a tendency to switch to an exposure driven RA methodology. The exposure experiments conducted in ENFIRO shows that certain alternative FR-systems are already widely present in the environment and as such have a higher level of concern. Attention can then be given to assess whether these alternatives are indeed an improvement of the health and environmental profile of the substitute and could become a potential concern again. Therefore just substituting the undesired substance does not imply automatically a lower risk. The potential release for example by leaching is practically non-existent for reactive FR-systems. From a point of view of the risk assessment the FRs linked to the polymer chain thus have a lower risk than when an additive type of FR is used. However, while the RA methodology leads to a lower risk in this case, the Env-LCIA study shows that such a reactive system may lead to similar impacts in the waste phase when an additive BFR-system is substituted by its reactive one. Still the development of polymer-based or reactive FR systems should be stimulated to reduce the environmental and human exposure. From the viewpoint of waste proper waste treatment and recycling should be high on the agenda.The relevant data required in the RA of the alternative FR selected in the ENFIRO project indicates not only the lack of data for many relevant parameters but also that many relevant data cannot be estimated with reasonable accuracy. This is caused by the fact that most of these estimation methods are developed for organic substances and are not valid for inorganic or organometallic ones. It is recommended to develop estimations tools for exposure of inorganic compounds. In ENFIRO about 40% of the studied flame retardants were inorganic or organometallic based.Impact assessment studiesThe Life-Cycle impact assessment studies should be based on a realistic life-cycle of the product. ENFIRO included a number of processes in the Env-LCA that are often not addressed, such as illegal waste scenarios (improper WEEE treatment), which significantly contributed to the environmental perfomance of the products. Future LCA studies should increasingly include these issues, because ignoring them may result in an incomplete description of the full environmental impact of a product's life cycle. Other issues that were included in the Env-LCA study that are normally left out, are indoor human exposure via dust (local impact), human exposure to FRs in Asia (and not only in Europe) when burning electronics waste in open fires to recycle the metals (occupational health issue) and fire occurrence (incident). Therefore the different life-cycle assessment studies should be based on the real life-cycle including illegal and local aspects that are relevant in the impact assessment study. Env-LCA studies on a part of the complete life-cycle may obscure relevant impacts. It is recommended to broaden the scope of LCA studies.Alternative assessmentsCurrently the assessment of alternatives is often simply based on substitution of the undesired substance by another one without taking into consideration if the alternative possesses an improved health and/or environmental profile. This dilemma was the basis for the EU to establish a specific call to assess the health and environmental risk and impact of alternative FR-systems . ENFIRO developed an alternative strategy based on four elements of the chemical substitution cycle combined with a unique assessment approach of the data. The data was assessed at three levels the chemical (flame retardant), the material, and the product. This approach was very successful and could be used for similar substitution studies (e.g. REACH). Current chemicals policy does not require a pre-assessment of alternatives although REACH indirectly solves at least a part of this aspect if the substance must be registered under this regulation. In that case a summary of the registration report will be published and the different registration reports can be compared to each other.Alternative assessments can also include other ways to increase the safety of the product. The ENFIRO project did not pay attention for example to possible improvement of the fire retardancy by indicating new ways of design of products, which may not require flame retardants at all, as this was not the scope of the project.Gender issuesDespite substantial progress towards equality between women and men achieved by equal treatment legislation, gender mainstreaming, specific measures for the advancement of women, the social dialogue and the dialogue with the civil society, this has not led to the desired gender balance in the EU workforce. In the ENFIRO project, a platform was created for awareness raising and exchange of ideas through presentations by the gender officer and invited speakers that initiated and stimulated lively discussions within the project group. A valuable source of information on the current status of women representation in the EU was the 'She Figures 2009 - Statistics and Indicators on Gender Equality in Science', that revealed that the figures are encouraging but the gender imbalance is not self-correcting. The various aspects of the implementation of quotas for women representation in company boards was presented inspired by 'Women on Board The Norwegian Experience' (Aagoth Storvik and Mari Teigen, (Friedrich Ebert Stiftung), 2010). Guidelines for improvement of the quality of selection procedures for the appointment of professors with special emphasis on equal opportunities for women were highlighted, derived from the PhD thesis 'Behind the Scenes of Science Gender practices in the recruitment and selection of professors in the Netherlands (2010) ' by M. van den Brink. More general, world-wide information was derived from the 'Global Gender Gap Report 2010', a publication of the World Economic Forum. In addition, Johanna Andersson, Gender Coordinator at Chalmers University (Goteborg, Sweden) and Sara Hunter, Head of Equality and Diversity Services of the University of Ulster (Belfast, UK) were invited to present the initiatives and strategies implemented to address gender mainstreaming at these respective universities.List of Websites:http://www.enfiro.eu Related documents Final Report - ENFIRO (Life Cycle Assessment of Environment-compatible Flame Retardants ( Prototypical case study))
