Decamethylcyclopentasiloxane

PBT assessment

PBT assessment: overall result

PBT status:

the substance is not PBT / vPvB

Justification:

Persistence

D5 does not meet the definitive criteria for persistence in the aquatic or soil compartments; it is not P or vP. The criteria for persistence (P/vP) in the sediment compartment are met.

Bioaccumulation

D5 meets the current Annex XIII assessment criteria for B and vB based on laboratory BCF data. However, in the view of the Registrants, a weight-of-evidence approach (Bridges and Solomon, 2016), considering laboratory and field biomagnification data, indicates that D5 is unlikely to biomagnify in the food chain and therefore should not be considered as B/vB.

Toxicity

The substance does not meet the definitive criteria for toxicity; it is not T.

Weight of evidence

The EU Member State Committee concluded that D5 meets the regulatory criteria for vPvB, and D5 has been included on the Candidate List of Substances of Very High Concern (SVHC) (20June 2018).

REACH guidance on the use of weight-of-evidence for PBT/vPvB assessment is limited; this is an area of science still changing, and areas of development and uncertainty are still being discussed amongst technical leaders in the field. When using the weight of evidence assessment (Bridges and Solomon, 2016), the Registrants are convinced that D5 is not vPvB.

The purpose of the following sections is to summarise the Registrants’ approach to weight of evidence concerning persistence and bioaccumulation, and to acknowledge that scientific understanding of persistency and bioaccumulation is growing and changing. REACH guidance on the use of weight-of-evidence for PBT/vPvB assessment is limited; this is an area of science still changing, and areas of development and uncertainty are still being discussed amongst technical leaders in the field.

Registrants’ opinion concerning weight of evidence on persistence

Persistence is not intrinsic to a substance; it will depend instead on the experimental conditions at the time and place of testing. Due to the physical properties of D5, traditional laboratory tests for persistence, especially sealed systems, are not appropriate for extrapolation in the complex environment. For substances such as D5, it is P_{overall (}P_ov)or P in the final sink (air) that matters in the global context.Consequently, chemical persistence should be assessed with respect to a compound’s P_ovin an evaluative multimedia regional or global environment.The overall persistence (P_ov) is more important for D5 than for other classes of chemicals, such as the classical PBTs/POPs. Since D5 partitions readily to atmosphere (major compartment) where it is degraded more rapidly than in other matrices, D5’s presence in the global environment is much shorter (months) than the classical POPs where global lifetimes are much longer (several years). The predicted P_ovfor D5 is 16 days indicating very different persistence characteristics compared to classical POPs which have half lives in years.

There are no guidelines for using P_ovto classify chemicals as P and vP; however, expert judgment of the results of global modeling shows that global half-lives in air are short and, if use were to cease, partitioning into and then rapid degradation in the atmosphere would result in complete dissipation in a few years. As these processes are ongoing, this means that concentrations in the global environment are in a quasi-steady state at this time and will not increase with continued use and release (Bridges and Solomon, 2016).

D5 tends to be distributed mainly in air (>95%). In air, it is well established that D5 readily degrades by interaction with OH radicals (Atkinson, 1991; Latimeret al., 1998, Sommerladeet al. 1993). D5 is mainly released from the urban centers where the OH radical concentrations are much higher than the global average OH radical concentration used to estimate their current half-lives (Suzukiet al., 1984; Nunnermackeret al.1998; Dillonet al., 2002; Renet al.2002; 2003); Hjorth 1984; Schadeet al.2002). Very recent work using actual air monitoring data demonstrates the real life degradation of D5 (and other VMS) in air may be much faster than what is currently estimated (Xu and Warneret al., 2017) and involve other mechanisms beyond OH radical degradation. This work suggests a calculated real-life half-life (empiricaltvalue) of 2.0 days in the air for D5. Discussions with experts from Norway, Stockholm University, and Canada at the 2016 SETAC conference provided support for this hypothesis and an ongoing collaboration with these experts is now underway to further explore this hypothesis (Kimet al., 2017).

In summary, the Registrants acknowledge that D5 meets the individual laboratory criteria for persistent in sediment. However, the available lines of evidence show consistently that various forms of degradation of D5 are possible (including in sediment) and that it is necessary to take into account the evidence concerning rates of reaction in assessing if D5 would be truly persistent in the environment, including the aquatic environment. In addition, the potential for transfer between compartments also needs to be considered. The combination of degradation and potential for transfer between compartments is reflected in the P_ovassessment and P_ovis most appropriate for assessing persistence in the global context. 

Lastly, by the very nature of D5 (volatile, strongly bound to organic carbon, subject to degradation), the presence of D5 in sediment will be constrained by these properties and the maximum absorptive capacity of the sediment (Bridges and Solomon 2016). Therefore, sediment does not produce an unlimited potential for uptake of D5 into biota, and the presence (or persistence) of D5 in sediment does not lead to an increased uncertainty in the estimation of risk to this compartment. A quantitative risk assessment can be conducted near emission points to ensure there is no risk from D5 to this compartment. 

 

Registrants’ opinion concerning weight of evidence on bioaccumulation

The purpose of this section is to summarise the Registrants’ approach to weight of evidence concerning bioaccumulation. That does not ignore the conclusion of the Member States Committee (MSC) for REACH, which has its own responsibility to interpret the REACH Regulation and Guidance as they stand. However, scientific understanding of bioaccumulation is growing and changing.

The Registrants consider that D5 is not bioaccumulative in the sense that D5 will not biomagnify to increasing unpredictable concentrations in the food web. Although water is an exposure route for lower trophic level organisms based on the BCF, a concern for bioaccumulation would require presence in water, high persistence in water, low potential for elimination from biota at the lower trophic levels, little metabolism or excretion by higher species, and the potential for toxicity expression in these organisms. This is not the case for D5 because D5 is volatile, poorly soluble, and not persistent in natural waters. D5’s presence in surface water is low to non-existent and laboratory and field data support metabolism of D5 in aquatic species. In addition, in terrestrial and other air breathing organisms, both metabolism and elimination of D5 via exhalation prevent the accumulation of D5. D5 will not therefore biomagnify to increasing unpredictable concentrations in the food web and could never accumulate and pose a threat to top predators and humans through biomagnification. This is supported by field data assessing trophic magnification.

Factors that control the accumulation of D5 in biota, laboratory and field    

Mackayet al., 2015 examines the role of physical-chemical properties important in evaluating the potential for accumulation from “superhydrophobic” compounds such as D5. The term superhydrophobic is applicable to compounds with log K_owgreater than about 7, which includes D5. The authors compiled a series of conventional uptake equations and a simple accumulation model for aquatic organisms. The model was applied to D5 simulating conditions in standard toxicity tests. The authors then discussed the reasons for the apparent lack of accumulation to concentrations of toxicity with these substances, and the potential to accumulate to toxic levels under field conditions in which diet is likely the principal source of lipophilic chemical loading to aquatic organisms (notably fish) that occupy intermediate and higher trophic level positions in aquatic food webs. Among the important factors that were examined were the roles of hydrophobicity and metabolism.

Hydrophobicity is described by log K_owvalues that include D5 (i.e., 8.07 at 25°C). Hydrophobic substances are predicted to substantially accumulate in aquatic organisms in the absence of metabolism or growth dilution. Hydrophobicity serves to constrain the ability of a compound to accumulate by limiting the aqueous solubility. Metabolism plays a key role in the accumulation and toxicity of hydrophobic substances. As metabolic rates increase, greater amounts of a compound must be accumulated in order to cause toxicity. From the model calculations by Mackayet al.(2015), even modest rates of metabolism can substantially affect the accumulation and toxicity of D5. The authors concluded that the extent of possible biomagnification of a persistent substance is controlled by the chemical’s hydrophobicity, dietary assimilation efficiency, elimination rates, and the nature of the diet. If there is even slow biotransformation such that k_Mapproaches or exceeds k_2,biomagnification is unlikely to be significant. They further indicate that the preferred approach for elucidating the possibility of biomagnification is to obtain monitoring data for representative food webs and convert the concentrations to fugacities or chemical activities to reveal their relative equilibrium status. Additional research is underway with independent experts assessing the biotransformation capacities of D5 in fish and benthic organisms. These studies will provide detailed knowledge of biotransformation rates (k_M) and uptake efficiencies (E_D) on D5 and other cVMS.

The MSC gives greater weight to BCF. However, this is in reality a laboratory screening metric; substances of high hydrophobicity inevitably have a high BCF, however fast they metabolise, and so for such substances BCF should not be used as a definitive criterion.

Activity or Fugacity Analysis to Assess Biomagnification Potential of D5

Gobaset al., 2015b have assembled concentration data for environmental matrices such as water, effluent and sediment, as well as concentration data for biota. The biota included different several taxa including plankton, invertebrates, fish, birds, and terrestrial and marine mammals.

Dimensionless chemical activities are related to chemical fugacities (units of Pa), which are often described as the escaping tendency of a chemical from one matrix to another. Chemical activities are simply the ratio of a concentration and its solubility. This allows expression of all data being examined to range from 0 to 1. Chemical activities are easy to calculate and allow the comparison of concentration data that are of differing units.

Chemical activity of aqueous concentrations is calculated as the ratio of the concentration with the solubility of D5 (17 µg/l) in water. Activities of sediment and soil concentrations are the ratio of organic carbon (OC) based concentrations with the OC-based ‘maximum sorptive capacity’. The value for D5 is 2514 mg/kg-OC. Activities of concentrations in biota are the ratio of the lipid-based concentration and the apparent solubility of the chemical in lipid, which is approximated by the compound’s K_owand aqueous solubility values.

The results of the analysis for D5, Gobas et al., 2015b, is shown in Figure 8.1.1 of the CSR. The figure shows the ranges of chemical activities for the several taxa for which D5 data are available. Fugacity and chemical activity ratios of D5 derived from several taxa indicate that D5 does not biomagnify in food-webs, likely due to biotransformation.

For plankton and invertebrates, activities ranged from 10^-7to 10^-3, while chemical activities for fish ranged from 10^-8to 10^-4. Chemical activities for birds, terrestrial and marine mammals ranged even lower at 10^-9to 10^-7.D5 is a very hydrophobic substance (logK_ow= 8.09) with a low affinity (relative to itsK_ow) for organic carbon (logK_oc= 5.17). Also, D5 has a sufficiently high affinity for air (logK_oa= 4.93) that allows for a significant rate of respiratory loss in air breathing organisms.Due to the highK_owof D5, the primary route of uptake of D5 in aquatic and terrestrial food-webs is by dietary uptake. In aquatic and terrestrial biota, D5 is metabolised at a rate that is sufficiently high relative to its rate of uptake to prevent the occurrence of dietary biomagnification. As a result, D5 is subject to trophic dilution (i.e. the opposite of biomagnification) in food-webs, causing fugacities and activities to decline with increasing trophic level.

D5 chemical activities in WWTP effluents, ambient water, sediment, plankton, invertebrates, fish, birds, and terrestrial and marine mammals (used with permission from Gobaset al., 2015b) from different North American locations, in relation to the maximum activity (a=1, red line) and various NOEC values in sediment and soil-dwelling invertebrate and plant species.

Figure8.1.1Figure on D5 activities (unitless)

Overall the collective chemical activity or fugacity data for D5 (Fairbrother and Woodburn 2016), and Gobaset al. (2015b), support the conclusions that D5 will notbiomagnify to increasing unpredictable concentrations in the aquatic food web.Fugacity or activity ratios of D5 derived from assembled concentration data for environmental biota including different several taxa (plankton, invertebrates, fish, birds, and terrestrial and marine mammals) indicate that D5 does not biomagnify in food-webs, likely due to biotransformation..

 

Relationship to determination of TMF

The discussion of activities allows an overview of many measurements. It provides a line of evidence regarding field measurements and helps discussion of weight of evidence as concluded by Bridges and Solomon. Bridges and Solomon (2016) conducted a weight of evidence (WOE) analysis for D5. The WOE analysis included studies that examined data on concentrations in various environmental matrices, persistence, bioaccumulation, and toxicity. TMF studies reported in the literature could be considered as particular subsets of measurements.

Field studies from locations in North America, and Asia have shown that D5 biodilutes as you move up the food chain. Trophic food web biodilution behaviour of D5 has been documented in systems varying from freshwater to marine and pelagic to benthicpelagic in food web structure.

The registrants acknowledge that there is a lack of full agreement of the TMFs across various study areas and this may reflect differences in TMF study design. TMFs may also be biased because of environmental conditions and sample collection. There is often a lack of common sampling areas for the species considered in the TMF calculation and the presence of point sources such as wastewater treatment plants can cause concentration gradients in the sampling area. These gradients can have a significant impact on study outcomes, as shown by Gobaset al., 2015, Kimet al., 2016 and Mackayet al., 2016. Therefore, the Industry continues to conduct additional work to add more certainty to these assessments.

Furthermore, as demonstrated by the Multibox-AQUAWEB model, TMF values above 1 can occur if sampling has been carried out with high spatial variation (see Annex 4 to the Exposure CSR for further details). Work is currently underway with external experts to update a Multibox-AQUAWEB model that takes into consideration these variables and confounding factors. As noted above,research is underway assessing the biotransformation capacities of D5. The additional understanding of biotransformation rates and uptake efficiencies in aquatic organisms will be used to further calibrate the Multibox-AQUAWEB model for assessing field TMFs. 

Overall weight of evidenceregarding bioaccumulation potential

The various documents mentioned above indicate and advocate an approach that identifies specific lines of evidence for specific areas and then brings these lines of evidence together to apply a weighting to each line in terms of its relative importance on the conclusion. The approach applied is qualitative only in that there was no agreed ‘scoring’ methodology in this context. In many cases the weighting is based on expert judgment and argument, rather than a quantitative weighting or scoring system. As indicated in these submissions a robust quantitative and transparent weight of evidence PBT assessment on D4 and D5 by academic experts (Bridges and Solomon, 2016), has been addressed.

The results of this robust quantitative weight-of-evidence (WoE) (Bridges and Solomon, 2016) as well as additional lines of evidence on biomagnification potential of D5 is presented.

It should be noted that application of a weight-of-evidence approach that considers all available information relating to the substances does not lead to the conclusion that D5 is bioaccumulative or very bioaccumulative (Bridges and Solomon, 2016). The evaluation concluded:

A transparent quantitative weight-of-evidence (WoE) evaluation of each study was needed to characterize their properties. Measurements of concentrations of cVMSs in the environment are challenging but, at this time, concentrations measured in robust studies are generally small and all less than thresholds of toxicity. The cVMSs are slightly persistent in air with half-lives ≤11 d. They are rapidly degraded in dry soils and partition from wet soils into the atmosphere. They are not classifiable as persistent in soils. Persistence in water and sediment is variable but greatest concentrations in the environment are observed in sediments. Based on their overall persistence in the environment, cVMSs should not be classified as persistent. The overall weight of evidence for the studies in food-webs supports a conclusion that the cVMSs clearly do not meet the criteria for biomagnification in the environment, a conclusion that is consistent with results of toxicokinetic studies. Toxicity was not observed at the solubility limit in water or the maximum sorption capacity in soils and sediments. Combining all of these lines of evidence shows that the traditional measures of persistence and biomagnification used to classify the legacy POPs are not suitable for the cVMSs. Refined approaches are needed and, when applied, these show that these materials are not classifiable as persistent, bioacumulative or toxic in the environment.

In addition, Gobaset al., (2015a) completed an in-depth independent review (Bioaccumulation of Decamethylcyclopentasiloxane (D5): A Review), involving re-analysis of original data according to current analytical practices. The findings of this study support the view that:

(i)            

(ii)          [1]

and

(iii)        

 

Review of the available information shows a high degree of internal consistency among findings from different studies and supports a broad comprehensive approach in bioaccumulation assessments that incorporates all available bioaccumulation measures in addition to the K_OWand BCF. Furthermore, this study supports the Siloxane (D5) Board of Review (2011) conclusion that D5 does not biomagnify in food-webs and is of low concern from a bioaccumulation perspective.

Gobaset al. (2015a) also examined the available field data on D5 and concluded that:

 

“Fugacity and activity ratios of D5 derived from bioaccumulation measures indicate that D5 does not biomagnify in food-webs, likely due to biotransformation.”

 

The registrants agree with the conclusions of these experts. The approach of the MSC is to give high weight to the laboratory data above other considerations.

 

Additional considerations regarding aquatic toxicity

D5 is non-toxic to aquatic organisms.At the highest concentrations that could be tested in water (limited by low water solubility), there were no observable effects in both short- and long-term studies with fish, daphnids, or algae.

Based on this ecotoxicity data it can be concluded that the criteria for T are not met.

As an alternative approach to the aquatic ‘T’, the possible back-calculation using the equilibrium partitioning (EqP) calculation of a sediment NOEC value to an aquatic NOEC value is set out in REACH Guidance R.11 (PBT/vPvB Assessment)[2]. It is not proposed there to use this method for PNEC calculation. The calculated aquatic NOEC may then be compared with the Annex XIII aquatic T criteria. The method converts the sediment NOEC to a dissolved water concentration (or aquatic NOEC).

It should be noted that the original purpose of the EqP method was to avoid unnecessary soil or sediment studies

The ‘T’ criteria defined in REACH Regulation Annex XIII and the REACH PBT guidance (R11) do not cover sediment or soil. The registrant considers that this gap in regulatory guidance should not be filled using the equilibrium partitioning calculation, which is not validated in this respect.

The equilibrium partitioning theory is intended to be used to calculate an initial PNEC_{sediment/soil}from a pelagic PNEC_waterwhen no sediment/soil studies are available. The R11 guidance allows for the use of equilibrium partitioning (EqP) theory to back-calculate a NOEC value of a sediment test to a pelagic NOEC value which can then be compared with the T criterion. The guidance suggests that this method be considered if it is technically not feasible to perform a test via the water phase (e.g. for substances with Log K_ow> 6, where the substance partitions out of solution). However, direct studies on pelagic organisms are available for D5 and these should be given more weight.

In addition, use of EqP theory to back-calculate a NOEC value of a sediment test to a pelagic NOEC value is based by implication on the assumption that the sediment toxicity is mediated through the pore water. However, the benthic studies include exposure via direct contact and ingestion of solid phase also, which is why they are a standard requirement at Annex X. Use of EqP methods necessitates that the method is shown to be valid for the particular substance. The absence of pelagic effects for D5 means that use of EqP for any purpose has not been validated.

Therefore, the registrant considers the use of EqP theory to back-calculate a NOEC value of a sediment test to a pelagic NOEC value to not be valid.

This conclusion is supported by the evidence presented in a submission to the RAC by Henriette Selck (Roskilde University, November 2014). The submission discussed various aspects of the aquatic toxicology of D4 and D5, and also reviewed the nature of the exposure of sediment organisms to adsorbed substances. The unamended extract is attached to the IUCLID EPS.

Therefore, long term sediment studies should not be used to assess ‘T’ for D5.

Furthermore, Selck and Forbes (2018) have set out a robust rationale explaining why the assessment of risks to the benthic compartment should be based on data for that compartment, and not the pelagic compartment. The corollary is clear, that risks to the pelagic compartment should not be based on data from the benthic compartment.

In a further publication in preparation, Selck and Windfeld[3]demonstrate that regulatory frameworks for assessment of persistence, bioaccumulation and toxicity based on water-exposure and microbial degradation only are not appropriate for hydrophobic organic chemicals (HOCs) such as cyclic volatile methyl siloxanes.

Overall Weight of evidence

The EU Member State Committee concluded that D5 meets the regulatory criteria for vPvB, and D5 has been included on the Candidate List of Substances of Very High Concern (SVHC) (20June 2018).

REACH guidance on the use of weight-of-evidence for PBT/vPvB assessment is limited; this is an area of science still changing, and areas of development and uncertainty are still being discussed amongst technical leaders in the field. When using the weight of evidence assessment (Bridges and Solomon, 2016), the Registrants are convinced that D5 is not vPvB.

The R11 guidance states:

PBT substances are substances that are persistent, bioaccumulative and toxic, while vPvB substances are characterised by a particular high persistency in combination with a high tendency to bio-accumulate, which may, based on experience from the past with such substances, lead to toxic effects and have an impact in a manner which is difficult to predict and prove by testing, regardless of whether there are specific effects already known or not. These properties are defined by the criteria laid down in Annex XIII of the Regulation…

Experience with PBT/vPvB substances has shown that they can give rise to specific concerns that may arise due to their potential to accumulate in parts of the environment and: that the effects of such accumulation are unpredictable in the long-term; such accumulation is practically difficult to reverse as cessation of emission will not necessarily result in a reduction in chemical concentration.

Furthermore, PBT or vPvB substances may have the potential to contaminate remote areas that should be protected from further contamination by hazardous substances resulting from human activity because the intrinsic value of pristine environments should be protected.

Some key conclusions can be reached:

• There is a high probability that B/vB is not met in the environment itself. Definitive methodology for assessment of trophic magnification factor (TMF) is still needed. Therefore, B/vB cannot be considered as being met with certainty, and therefore PBT/vPvB may not apply under consideration of weight-of-evidence. Fugacity ratio calculations support that biomagnification is not occurring.
• Any ecotoxicological or mammalian toxicological effects are understood and are not irreversible or feasible under realistic exposure conditions
• D5 is not found in remote regions except when associated with human activity, and the concentrations in the environment are not increasing such as to suggest a high level of overall persistence.
• The absolute concentrations of D5 measured in the environment are consistent with the local release pattern and do not indicate unexpectedly high persistence, i.e. there is no indication that half-lives in the environment are longer than the laboratory data suggest. In addition, analysis of available air monitoring data suggests the half-lives are shorter than the laboratory data suggest.
• D5 has uses in industry and in the personal care sector (declining sharply) which result in a predictable level of the substance in environmental compartments (see section 10). The fraction of D5 that is released that can ultimately be found in compartments such as fresh water and marine sediments is very small, and the tonnage released directly to the environment is low. The majority is released to, and will degrade in, air. Modelling data using standard and advanced methods shows that the scope for movement of the substance by long-range mechanisms is limited by degradation in air and a very low potential for deposition to soil or water from air. This is strongly supported by the monitoring data obtained to date.
• PEC values are low in absolute terms. The environments in which D5 is found are protected by use of PEC/PNEC methodology. The exposures and risk characterisation ratios are discussed in detail in Sections 9 and 10. Monitoring data available also supports that the environments in which D5 is found are protected.

Industry fully acknowledges the MSCconclusion (20 June 2018)that D5 meets the current regulatory criteria (as defined in REACH Annex XIII) for vPvB whilst recognizing that there are differences in global scientific opinions on how to use weight of evidence to evaluate vPvB potential.Industry will support and work with downstream users to successfully implement the targeted restriction and continue to minimise emissions.

The silicones industry continues to work proactively to address regulators’ concerns. Our monitoring activities in particular, both to water/sediment compartment and air aim at achieving a common understanding of the fate of D5 in the environment, if and when the substance is emitted. In particular, we have a five-year monitoring programme in place in different regions of the world that is evaluating concentration trends of cVMS in sediment and biota near point source discharges.

 

Registrants continue to work with the most knowledgeable scientific partners from academia to:

-                     address remaining uncertainties e.g. on the presence and behaviour of D5 in air;

-                     review data and studies pertaining to vPvB assessment such as new biomagnification information and biotransformation/elimination/uptake efficiency of D5 in aquatic organisms;

-                     improve modelling and apply probabilistic risk assessment methods to D5 to re-confirm the lack of risk in aquatic systems.

[1](the data supporting this statement are not currently available);

[2]It is applied without the factor of 10 adjustment to PNEC which would actually raise the calculated NOEC for pelagic species.

[3]Selck and Windfeld, 2019. Publication in preparation.