Document 1gE2YdzLBv1pEoBJNGKr8bnVd

FABES Forschungs-GmbH fr Analytik und Bewertung von Stoffbergngen Investigation Report Modelling on BPAF migration from FKM materials The results of the present investigation report are property of the contracting body. The partial use of the results is only permitted with a prior written approval of FABES Forschungs-GmbH. Contracting body: Bluefrog Scientific on behalf of Bisphenol AF consortium LLP 23 Old Fishmarket Close 190 High Street Edinburgh RH1 1AE UNITED KINGDOM Assignment date: 05.06.2023 Commission number: 8735-23_A Sample receipt: /; Information receipt: 31.05.2023, 12.06.2023, 14.06.2023 Test period: 12.06.2023 - 19.06.2023 Date of report: 20.06.2023 Number of pages: 14 current page: 1 Sample: No. Specification - None FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 2 of 14 1. Introduction In this report, the migration of Bisphenol AF (BPAF; CAS-No. 1478-61-1) from FKM materials into rain water will be estimated by using a theoretical modelling approach. Two different scenarios will be considered: 1. Articles made of fluoroelastomer materials (FKM) are deposited in a landfill. These FKM articles are stored in form of bales, which can be imagined as stacks of FKM sheets one-ontop of the other. The size of such a bale is assumed to be a cube with 1 m sides (Vt = 1 m, H = 1 m, W = 1 m, L = 1 m). This cube is filled to 65 % with FKM sheets. The bales are stored under the open sky so that during rainfalls rainwater wets these FKM articles and, in a first approximation, forms a uniform rainwater film on the surface area of the sheets. 2. The outer surface of an O-ring made of an FKM material is exposed to rainwater under the open sky. Due to the geometric complexity of this system, the O-ring will be treated in a first approximation as a flat surface with a uniform rain water layer in one-sided contact with it. It is known that these FKM articles contain a certain residual amount of the crosslinking agent Bisphenol AF. It is documented in the literature, that such compounds are not covalently bound to the matrix of the polymer and because of that have a certain mobility (diffusion property) in the polymer. Because of this mobility and due to the BPAF concentration gradient between the FKM articles and rainwater a migration process of BPAF from the polymer into the film of rainwater occurs (Figure 1). The rate of this process is controlled mainly by the diffusion coefficient of BPAF in the matrix of the FKM samples. On the other hand, the BPAF migration process from FKM into rainwater is limited by the partitioning process of BPAF between FKM and (rain)water. In this process the equilibrium solubility of BPAF in the FKM material and in (rain)water plays a central role. In Central Europe the average precipitation is 700 mm/year [1]. Of course, this total amount of rainwater is not distributed uniformly during the year but periods with more intense rainfalls (especially in spring and autumn) alternate with periods with less rainfalls (in winter and summer). Therefore, to simulate, as realistically as possible, migration of a substance from the FKM material into rainwater the exact rainfall scenario over a year should be considered at the site of the investigated scenario. But such information is difficult (if not impossible) to obtain. Therefore, in this report, in a first approximation, a much simpler rainfall scenario over the duration of one year will be assumed. This consists of a 24 hours continuous rainfall period (the "wet" period) followed by 48 hours without rainfall (the "dry" period) for the first 6 months. In the second 6 months, a rainfall period of 24 hours will be followed by a dry period of 120 hours. Thus, in one year there are 91 wet periods (61 in the first half and 30 in the second half of the year). A temperature of 25 C for all periods will be assumed. The objective of this report is to get an estimate of the amounts of BPAF leaching into rainwater in such a migration scenario. However, to do such calculations additional information and assumptions are necessary. These will be presented in the followings. FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 3 of 14 2. Calculation/estimation of migration The estimation of migration of BPAF from a FKM sample in water can be done by solving Fick's 2nd mass transport equation [2]. In practice, because the migration from a FKM sample into a film of rainwater takes place preferentially in a normal direction (90) to the surface of the samples, the mathematical task of solving Fick's 2nd equation reduces from the complex task of solving a real three-dimensional (3D) problem to a one-dimensional (1D) problem with the appropriate mass transport Equation (1): = (1) where: Ck (in g/cm) is a local concentration and Dk (in cm/s) a constant diffusion coefficient in medium k, x is the one-dimensional (1D) spatial coordinate of the polymerwater system, t the migration time, k = P the index for the FKM material and k = W the index for water (Figure 1 for the 1D migration). Equation (1) can be used - with appropriate input parameters, for mono-layer polymers as well as for multi-layer structures. In this report the FKM samples are all considered to be mono-layer polymers with a homogeneous structure and a homogeneous initial distribution of the migrants in the matrix of the polymer. Figure 1: One-dimensional migration of a substance from a FKM sample into a film of rainwater. The specific problem of the migration process envisaged in this work is the fact that it must be assumed realistically that the FKM samples are covered with a rainwater film only for a limited time during the rainfall event and after that for a certain time they are dry. Therefore, during the wet periods there is a leaching of BPAF from the FKM samples into the water film while during the dry periods there is no leaching of BPAF from the FKM at all. However, from a physical-chemical point of view, it is correct to assume that during the dry periods there is an internal redistribution of the BPAF in the FKM sample. This process is driven by the following facts. After a migration process of BPAF (or any other substance contained by the FKM sample) there is an additive/substance concentration gradient between the core of the FKM sample and its boundary layer (which was in contact with the rainwater). While in the core of the FKM sample, the concentration of the additive/substance is FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 4 of 14 practically unchanged at its initial level, in the boundary layer this concentration is lower as result of the substance migration into water. So, across the FKM sample there is an additive/substance concentration gradient which acts as a driving force for an internal diffusion from the core towards the boundary of the FKM sample. The longer the dry period is, the higher is the increase of BPAF concentration in the boundary layer. This BPAF concentration increase in the boundary layer of the FKM sample has then an influence on the migration level of the subsequent wet period. Therefore, to estimate correctly the repeated migration of BPAF from the FKM samples during a succession of rainfall events it is necessary to estimate also the redistribution of BPAF across the matrix of the polymer during the dry periods between the wet ones. The software MIGRATEST-RU-2022 developed by FABES Forschungs-GmbH can be used to solve Equation (1) for such a migration scenario [3]. This software solves Equation (1) with finite differences (FD) numerical methods [3,4]. The software also contains a batchprocessing routine, MIGRATESTRunner2Phases [5], with which the succession of wet and dry events during the assumed storage time can be processed automatically. It is far beyond the scope of this short report to give any details on the actual FD-procedures used in this software but such information can be found for example in [2] and [4]. FABES Forschungs-GmbH assures that the numerical calculations with MIGRATEST-RU-2022 are mathematically correct. This was verified and validated by comprehensive testing and can be demonstrated on demand. However, how close to reality the computed results are depends on the accuracy with which the conceived migration model and the specified input data reproduce what happens in reality. Because of that, when running the migration calculations with MIGRATEST-RU-2022 the correct specification of the input parameters, initial and boundary conditions for the above migration scenarios must be done as close to reality as possible. 3. Input parameters for the calculations Annotations: a) The data/parameters marked with (+ are assumptions made by FABES Forschungs-GmbH in absence of specifications from the commissioner of the report and/or data in the available literature. b) The data/parameters marked with (++ are information from the commissioner of the project. FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 5 of 14 3.1 Physical & geometrical data for the FKM sheet-rain water system The physical and geometrical data for the FKM sheet - rain water system is summarized in Table 1. Table 1: Physical data for the system. FKM sheet Volume of cube at the recycling facility Volume (VP) of FKM sheets in the cube Density of FKM (P) Value 1 0.65 1.8 Unit m3 (++ m3 (++ g/cm3 (++ Annotation Thickness of FKM sheet (dP) Total surface area (Atot) of FKM sheets FKM O-ring Total surface area of O-ring (Atot) Surface area of O-ring exposed to rain water (AO) Density of FKM (P) Mass of O-ring (mP) Rain water Volume of rain water (VW) per m2 per raining event Thickness of water layer in contact with FKM sheet during one raining event (dW,S) Thickness of water layer in contact with O-ring during one raining event (dW,O) Density of water (W) 25 C 5 260 108.57 54.29 1.8 48.85 7.7 29.6 7700 0.997 mm (++ m2 (+ Atot = 2 VP / dP cm2 (++ cm2 (++ g/cm3 (++ g (++ dm m (+ [1] (+ dW,S = VW / Atot m (+dW,O = VW / 10000 cm2 g/cm [6] FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 6 of 14 3.2 Chemical data The following chemical data for the migrant BPAF was available (Table 2): Table 2: Chemical data for the system. Migrant Value Unit Annotation Bisphenol AF (BPAF) (CAS-No. 1478-61-1) Typical molecular formula Molecular weight of migrant (Mw) Octanol-water partition coefficient of migrant (log KOW) at 20 C Solubility of migrant in water (SW) 20 C Initial concentration of migrant in the FKM material (CP0) Initial concentration of migrant in rain water (CW0) (++ C15H10F6O2 336.24 g/mol 2.79 - 222 mg/Litre 100 mg/kg 0 g/Litre (+ [7] substance with low polarity (+ [7] poorly soluble (++ (+ FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 7 of 14 3.3 Migration data For the theoretical estimation of a migration process like the one shown schematically in Figure 1, the information about the mobility of the migrant/s in the matrix of the FKM sheet material and in rain water (diffusion coefficients DP and DW) and of the partitioning at equilibrium of the migrant/s in the P-W system (partition coefficient KPW) are indispensable and very important information. For this project, kinetic data of this system have been obtained at three different temperatures. From these data, it is possible to extract the diffusion and partition coefficients with the knowledge of the model system parameters which are listed in Table 3. Table 3: Parameters for the model system (kinetic measurements). FKM sample Length (L) of one FKM sample Value 3 Unit cm (++ Annotation Width (W) of one FKM sample 1.5 cm (++ Thickness (dp) of FKM sample 0.22 Volume of one FKM sample 0.99 Surface area (A) of one FKM 9 sample Density of FKM (P) 1.8 cm cm3 cm2 g/cm3 (++ (++ (+ A = 2 L W the area of the edges will be neglected in a first approximation (++ Thickness of FKM sample (dP) Number of FCM samples in test Acceptor water Volume of acceptor water (AW) One-sided thickness of water layer (dW) Density of water (W) 25 C 2.2 2 36.6 2.03 0.997 mm (++ - (++ cm cm g/cm (++ (+ dW = VW / (2 A) [6] The procedure applied to model theoretically the experimental results obtained with the laminate films was to find a "best fit" between the discrete experimental results - time dependent concentrations of migrant in the acceptor medium, Wexp (), and the continuous curve, Wth (), of the calculated diffusion/permeation/migration process. For this, first a FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 8 of 14 number of strictly necessary input data were fed to the MIGRATESTExp software (last update 2022) [8]. This data starts with the so-called "physical parameters" given in Table 3. Further necessary parameters are the diffusion and partition coefficients ("chemical parameters"). These parameters were adjusted alternately and step-by-step until a satisfactory agreement between the theoretical and experimental curves was observed. This agreement is shown in Figure 2 for the data points taken at 25 C. migrated amount [% of initial amount] 0,04 0,02 calculated migration curve measured values 0,00 0 200 400 600 time [h] Figure 2: Agreement between theoretical (continuous line) and experimental data points. The obtained parameters were DP = 1 10-15 cm2/s and KP/W = 5.9 104 (g/cm3)/(g/cm3). In this report it was decided to assume a slightly underestimated partition coefficient of 5.0 104 (g/cm3)/(g/cm3) which leads to a slight overestimation of migration. This coefficient can be further refined when more data points have been collected. FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 9 of 14 4. Initial and boundary conditions for the calculations Annotation: The following initial and boundary conditions for the calculations were considered: 1) Initially the migrant (BPAF) is uniformly distributed in the whole mass of the FKM samples. 2) There is no decomposition of the migrant during the diffusion/migration process. 3) There is no evaporation of the BPAF from the FKM material during the dry cycles. 4) During the wet cycles the FKM samples are completely wet on both sides (sheets)/one side (O-ring) with a uniformly distributed layer of rain water. 5) The migration process is not hindered by a boundary resistance at the polymerrainwater interface where a partitioning of the migrant occurs. 6) The diffusion coefficients defined above are independent of the local migrant concentrations in the polymer and respectively rainwater. 7) The migrant concentration profiles calculated for each wet or dry cycle are saved and used as initial conditions for the diffusion/migration in the subsequent cycle. FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A 5. Estimated results 5.1 Results for the FKM-sheets/rainwater scenario In Figure 3, the results for the first year are shown. page 10 of 14 migrated amount per wet period [mg] 0,0275 0,0270 0 50 100 150 200 number of wet period Figure 3: BPAF concentration in rain water after each wet period in the first year of consecutive wet and dry periods in the FKM-sheets/rainwater scenario at 25 C. The estimated amount of BPAF migrated in one year is 2.5 mg. This corresponds to about 0.0021 % of the initial amount of BPAF in the FKM sheets. FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 11 of 14 In Figure 4, the results for 20 years are shown. 0,028 migrated amount per wet period [mg] 0,027 0,026 0,025 0,024 0 5 10 15 20 time [years] Figure 4: BPAF concentration in rain water after each wet period at 25 C for 20 years in the FKMsheets/rainwater scenario. The total estimated amount of BPAF migrated in 10 years is 23.2 mg. This corresponds to about 0.020 % of the initial amount of BPAF in the FKM sheets. The total estimated amount of BPAF migrated in 20 years is 45.6 mg. This corresponds to about 0.039 % of the initial amount of BPAF in the FKM sheets. FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A 5.2 Results for the O-ring/rain water scenario In Figure 5, the results for the first year are shown. page 12 of 14 0,07 migrated amount per wet period [g] 0,06 0,05 0,04 0,03 0,02 0,01 0 50 100 150 200 number of wet period Figure 5: BPAF concentration in rain water after each wet period at 25 C in the first year in the O-ring/rainwater scenario. The estimated amount of BPAF migrated in one year is 1.6 g. This corresponds to about 0.033 % of the initial amount of BPAF in the FKM O-ring. FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A In Figure 6, the results for 20 years are shown. page 13 of 14 0,06 migrated amount per wet period [g] 0,05 0,04 0,03 0,02 0,01 0,00 0 10 20 time [years] Figure 6: BPAF concentration in rain water after each wet period at 25 C for 20 years in the Oring/rainwater scenario. The total estimated amount of BPAF migrated in 10 years is 5.8 g. This corresponds to about 0.12 % of the initial amount of BPAF in the FKM O-ring. The total estimated amount of BPAF migrated in 20 years is 8.3 g. This corresponds to about 0.17 % of the initial amount of BPAF in the FKM O-ring. FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019 Investigation report 8735-23_A page 14 of 14 6. References 1. Vangheluwe, M., Eliat, M., Oorts, K., Report "Final report risk assessment storage PVC waste - Risk assessment of lead migration during storage of PVC waste", ARCHE Consulting, Gent, 31st of October 2016. 2. Crank, J., "The Mathematics of Diffusion", Chapter 8 "Numerical methods", Clarendon Press, Oxford, 1975, pp.137. 3. Mercea, P., Tosa, V., Mercea, P.C., "MIGRATEST-RU-2022-Software", FABES Forschungs-GmbH, Munich, 2022. 4. Tosa, V., Kovacs, K., Mercea, P., Piringer, O., "A finite Difference Method for Modeling Migration of Impurities in Multilayer Systems", AIP-Conf.Proc., Vol. 1048 (2008) 802. 5. Mercea, Paul, Mercea, P.V., "MIGRATESTRunner2Phases" - A batch processing software, Stuttgart, 2023. 6. The Engineering Tool Box, "Water-Density, Specific Weight and Thermal Expansion Coefficient", https://www.engineeringtoolbox.com/water-density-specific-weight-d_595.html. 7. REACH registered substance factsheet for Bisphenol AF, https://echa.europa.eu/de/registration-dossier/-/registered-dossier/23236/4/8. 8. Mercea, P., Piringer, O., Petrescu, L., Toa, V., "MIGRATEST Exp (Updated in October 2022)-Software for estimation of migration in multilayer systems", FABES Forschungs-GmbH, Munich, 2022. Munich, 20.06.2023 FABES Forschungs-GmbH This document was transmitted electronically and is therefore unsigned. Christoph Losher (project leader) Dr. Regina Wyrwich (scientist-migration modelling) FABES Forschungs-GmbH, Schragenhofstr. 35, 80992 Munich, phone: +49-(0)89-149009-50, www.fabes-online.de 09.03L00601 / 18.09.2019