Registration Dossier

Diss Factsheets

Environmental fate & pathways

Bioaccumulation: aquatic / sediment

Currently viewing:

Administrative data

Link to relevant study record(s)

Referenceopen allclose all

bioaccumulation in aquatic species: fish
Type of information:
Adequacy of study:
supporting study
Study period:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
results derived from a valid (Q)SAR model and falling into its applicability domain, with adequate and reliable documentation / justification
Justification for type of information:
QSAR prediction
other: REACH guidance on QSARs R.6, May 2008
Principles of method if other than guideline:
Calculated with BCF Program BCFBAF v.3.01 included in the Estimation Programs Interface (EPI)-Suite. The estimation methodology is based on the chemical structure of an organic compound and its log octanol-water partition coefficient (log Kow). Depending on chemical structure, structural correction factors are applied.
GLP compliance:
Test organisms (species):
other: none, estimated by calculation
794 L/kg
other: calculation

Any decomposition of the substance in water is not considered by the program.

Validity of model:

- Defined endpoint: bioconcentration of a substance in biota

- Unambiguous algorithm: linear regression QSAR. Because of the deviation from rectilinearity, different models were developed for different log Kow ranges. Metals (tin and mercury), long chain alkyls and aromatic azo compounds are specially treated.

- Applicability domain: the model is applicable to ionic as well as non-ionic compounds. It is applicable to substances with a logKow in the following range: -6.50 to 7.86 (ionic compounds) and -1.37 to 11.26 (non-ionic compounds). Applicable to substances with a molecular weight in the following range: 102.13 to 991.80 g/mole (ionic substances) and 68.08 and 959.17 g/mole (non-ionic compounds).Model predictions may be highly uncertain for chemicals that have estimated logKow values > 9. The model is not recommended at this time for chemicals that appreciably ionize, for pigments and dyes, or for perfluorinated substances.

- Statistical characteristics:

number in dataset = 527

correlation coef (r2) = 0.833

standard deviation = 0.502

- Mechanistic interpretation: The BCF is an inherent property used to describe the accumulation of a substance dissolved in water by an aquatic organism based on the lipophilicity of the compound.

Adequacy of prediction: Ditolyl ether falls within the applicability domain described above and, therefore, the predicted value can be considered reliable taking into account that the standard deviation error of prediction of the external test set is 0.59 (logBCF). Considering that error, the predicted value is not above or close to the criterion to consider a substance as potential bioaccumulative.

The bioaccumulation factor of ditolylether was estimated to be 2.9 using the BCFBAF model.
Executive summary:

The bioaccumulation factor of ditolyl ether was estimated to be 2.9 using the BCFBAF model included in the EPI-Suite Programm concluding that the substance has a low potential to bioaccumulate in biota. Within the scope of the Persistency-Bioaccumulation-Toxicity (PBT)-Assessment, the substance does not fulfill the B-criterion. Ditolyl ether falls within the applicability domain described above and, therefore, the predicted value can be considered reliable.

bioaccumulation in aquatic species: fish
Data waiving:
other justification
Justification for data waiving:

Description of key information

In the recent OECD 309 guideline study, the impurity 2,2'-Dimethyldiphenylether has been proven very persistent (vP) according to the Annex XIII criteria, with a DT50 of 63d in freshwater. Therefore, endpoint 5.3.1 Bioaccumulation aquatic/sediment will - as part of a forthcoming update - be revised with an experimental study based on ECHA decision number SEV-D-2114341466-49-01/F.



Status as of 2019

The endpoint 5.3.1 Bioaccumulation aquatic/sediment will - as part of a forthcoming update - be revised

a) either with an experimental study in case one or more of the ditolylether isomers  is considered persistent (P) or very persistent (vP) according to the Annex XIII criteria, based on ECHA decision number SEV-D-2114341466-49-01/F


b)  the existent section 5.3.1 will be revisited and updated taking into account ECHA decision number SEV-D-2114341466-49-01/F (which would include reworking the read-across approach to diphenyl ether, CAS 101-84-8) in case none of the ditolylether isomers  is considered persistent (P) or very persistent (vP) according to the Annex XIII criteria.


Further information including timelines for a potential bioaccumulation study is provided in the attachment of the waiver entry for the OECD 305 study in this section as well as in the dossier header.


Status as of 2015:

The BCF of ditolyl ether was calculated from diphenyl ether using the read across approach and the Dow Chemical Company study, 1973.

Key value for chemical safety assessment

BCF (aquatic species):
982 dimensionless

Additional information

Currently, there is only a QSAR estimation of the bioaccumulation potential of ditolyl ether which was calculated taking into account the experimentally measured log Kow of 4.9 and which estimates the BCF to be 794 L/kg ww. This BCF was estimated using a valid and recommended software tool (BCF Program BCFBAF v.3.01 included in the Estimation Programs Interface (EPI)-Suite) and, thus, the result is assumed to be reliable.

As there is no experimental BCF of ditolyl ether the bioaccumulation potential of the substance could be derived based on the screening criterion log Kow of 4.9. However, a definite conclusion based on screening criterion is not possible. Furthermore, the QSAR estimation indicates that the screening criterion for B / vB is not met.

As an experimental study on bioaccumulation is not available the read across / grouping approach might be used to fulfil the requirements, and to avoid an additional vertebrate study (if possible), using experimental data available from diphenyl ether (CAS 101-84-8).

The approach of using data from other substances to fulfil data requirements is introduced and described in REACH Regulation (EC) No 1907/2006, Annex XI, section 1.5: “Grouping of substances and read-across approach”. The legal text elucidates that data requirements might be fulfilled by this approach if the read across substance and the substance of interest show similarities.

In the case of ditolyl ether the bioaccumulation potential (BCF) shall, thus, be read across from diphenyl ether. Both substances are mainly structural similar. The structure of ditolyl ether is additionally methylated, with one methyl group at each of the two aromatic rings. Thus, both molecules have comparable chemical structures, molecular sizes, and molecular weights. Both substances are not hydrophilic but only moderate hydrophobic. However, without methyl groups diphenyl ether is less lipophilic and can be assumed not to accumulate as much as would be expected for ditolyl ether and, thus, a correction of its BCF is needed.

ECHA Guidance Document R.7c (v1.1, November 2012, p. 23) provides support on how to proceed in this specific case: “The BCF value of a substance is generally positively correlated with its hydrophobicity. Therefore, if the substance to be evaluated has a higher log Kow than an analogue substance for which a BCF is available, the BCF value has to be corrected. The use of the same factor of difference as for Kow will be a reasonable worst-case estimate, because generally the relationship between BCF and Kow is slightly less than unity. For example, if the substance to be evaluated has one methyl group more than the compound for which a BCF value is available, the log Kow will be 0.5 higher and the estimated BCF from read-across is derived from the known BCF multiplied by a factor of 10E0.5. In principle, this correction should give reasonable estimates as long as the difference in log Kow is limited.”

Thus, the guidance document clearly highlights that in the case of ditolyl ether the BCF may be estimated based on the BCF of diphenyl ether as both log Kow values are similar (ditolyl ether = 4.9; diphenyl ether = 4.21). However, the experimental BCF of diphenyl ether may not be used directly but has to be corrected for the methyl groups. The Guidance Document does not clearly state if for the second and further methyl groups the read across BCF has to be refined using n times 0.5. This approach could be appropriate for highly lipophilic compounds. Bioaccumulation of substances, however, depends on lipophilicity, shape, size and partition coefficient. Ditolyl ether and diphenyl ether appear to be molecules of moderate lipophilic character, both molecules are compact and of low molecular weight. Thus, estimating ditoly ether`s BCF using two times 0.5 in the calculation would be overly conservative and not a reasonable and justified realistic worst case.

Consequently, such a BCF would neither be appropriate for the PBT assessment, nor for the purpose of classification and labelling. However, REACh Regulation, Annex XI, section 1.2, clearly gives that “If the group concept is applied … in all cases results should be adequate for the purpose of classification and labeling and / or for risk assessment”.

Therefore, instead of a factor 2 * 0.5 = 1 a factor 0.7 should be sufficiently conservative and allow the calculation of a realistic worst case BCF.

Now that the approach to refine the read across BCF is discussed there is a need to identify this BCF from the available studies. Currently, there are two experimental studies on bioaccumulation behaviour of diphenyl ether which could be used for read across:

a) The DOW Chemical Company, 1973, and

b) Chemicals Inspection & Testing Institute (CERI), 1992.

Both studies show short-comings and test documentation is incomplete and may not answer all questions on validity and reliability which is elucidated in the following sections.

While the DOW study, 1973, was non guideline it followed the test procedure of Branson et al., 1973 (A Bioconcentration Test: Steady-State Concentrations of 2, 2', 4, 4'-Tetrachlorobiphenyl in Trout, DOW Report NCL-73011, March 1973.). Rainbow trout (O. mykiss) were used as test organisms and the test item was radioactively labelled, which ensures high precision and reliability. Test concentrations were 2.8 and 0.4 µg/L (analytically confirmed) and 40 fish were tested per concentration and in the control. Uptake and depuration phase were 4 days each and BCF at steady state was calculated from rate constants of uptake and depuration. Whole fish was analysed for bioaccumulation. Lipid content of fish is given as 1-1.5% and was considered in the final BCF. Depuration DT50 is given as 23 hours. Feasibility of this short cut test was confirmed with 14C-diphenyl oxide and exposure over 14, 28 and 42 days at 0.28 and 1.7 µg/L. Measured and predicted concentrations from this comparative approach and, thus, the reliability of the short cut test with ditolyl ether is “validated”. The final BCF was corrected for lipid content and is reported as 196.

Although the reported data is incomplete there are no obvious short-comings and the study may be used for read across purposes.

CERI, 1992, determined the BCF in accordance with CERI guidelines MOE No.5, MHW No.615 and MOL No.392 (according to bio-concentration test in fishery body). A flow through test approach was used. In a preliminary test acute toxicity was determined (48 hour-LC50 = 4.6 mg/L) and the main study test concentrations were derived from it (0.3 and 0.03 mg/L). The pre-test was conducted with O. latipes while the main test used C. carpio. The test substance was mixed with heavy castor oil as solubiliser. Uptake phase lasted 8 weeks while depuration took 7 days. The study report provides data on measured concentrations. Initial lipid content is reported as 2-6% but the final BCF was not normalized for it. The BCF with a test concentration of 0.3 mg/L was determined as 112-583 and that of 0.03 mg/L as 49-594.

The method description of the CERI study is more complete than that of the DOW study but raises some questions which make the results questionable: Different fish species were used for range finding and main study which does not allow to correctly determine the main test concentration; test concentrations might have been in the range of the acute toxicity. The current OECD guideline 305 gives that the highest test concentration should be a factor 100 below the acute toxicity. In the CERI study the acute toxicity was determined at LC50 = 4.6 mg/L but the highest test concentration (0.3 mg/L) was only about a factor 10 below. Varying fish species in pre- and main test and a main test concentration too close to acute toxicity raise questions on the study reliability. Furthermore, the use of a lipophilic agent (heavy castor oil) to solubilize a lipophilic substance is strange and unusual. OECD 305 lists different solubilizers to be used. The BCF was not corrected for the lipid content and there is no lipid content of the fish at the end of the experiment. However, the lipid content is important and the BCF should be normalized to be comparable with that of other studies. Furthermore, after 8 weeks the fish might have started sexual differentiation which impacts the study result as well. Finally, at least some of the reported measured concentrations show high deviations and the finally calculated BCF might be based on outlier concentrations which would not be acceptable. The reported not lipid normalized BCF were 49-594 L/kg ww (0.03 mg/L test concentration) and 112-583 L/kg ww (0.003 mg/L test concentration).

Thus, there are some major uncertainties with the CERI study and the reported BCF seems not to be reliable enough to be used for read across in the risk assessment of ditolyl ether. The registrant contacted the Japanese CETI to clarify these aspects but, unfortunately, received the information that the Japanese ministry METI does not have the original report available anymore. Thus, the uncertainties from the limited study data which is still available today may not be clarified and the study should not be used for the risk assessment of ditolyl ether.

The BCF of diphenyl ether was also estimated using the valid and recommended software tool (BCF Program BCFBAF v.3.01 included in the Estimation Programs Interface (EPI)-Suite). The reported BCF, as derived with the model, is 278. This modeled BCF is closer to the result of the DOW study (BCF = 198) than to the CERI study ( BCF <= 594) which gives an additional hint that the discussed short-comings of the CERI study might have had a significant impact on the reported BCF range.

Thus, the QSAR BCF estimation indicates that the DOW study result should be preferred over the CERI study.

Concluding, the DOW study as well as the CERI study show short-comings but as there are major uncertainties with the CERI study only the DOW study may be used for read across. Thus, based on the mathematical approach provided by ECHA Guidance Document R.7c (as detailed above) and the BCF of 196 reported in the DOW study the BCF of ditoly ether is calculated as 196*E0,7 = 982.

The DOW study is in the range of the valid QSAR estimation of ditolyl ether with a BCF = 794, and both, QSAR as well as read across with diphenyl ether, indicate that its BCF is clearly below the B and vB criterion of 2000 and 5000, respectively.