Glycerides, C8-18

Toxicity to aquatic algae and cyanobacteria

Description of key information

ErL50 (72 h) > 320 mg/L (nominal, loading rate) 
ErL50 (72 h)= 49 mg/L (nominal, loading rate) 
ErL50 (72 h) = 13 mg/L (nominal, loading rate)
NOELR (72 h) = 32 mg/L (nominal, loading rate) 
NOErLR (72 h) = 4.1 mg/L (nominal, loading rate)
NOEC (72 h) = 1.19 mg/L (measured concentration)

Additional information

No experimental data evaluating the acute toxicity of Glycerides, C8-18 (CAS No. 85536-06-7) to algae are available. Therefore, toxicity data from structurally related category members (Glycerides, C14-18 and C16-18 unsatd. mono-, di- and tri-, CAS No. 91052-28-7, Glycerides, C8-C10, mono- and di- (CAS No. 85536-07-8) and Glycerides, C12-18 mono- and di- (CAS No. 91052-49-2) are used as read-across according to Regulation (EC) No. 1907/2006, Annex XI, 1.5. All four substances are esters formed from the combination of fatty acids and glycerol. The target substance contains fatty acids with C-chain lengths ranging from C8 to C18, and >80% triester content. The three read-across substances cover the fatty acid C-chain range of the target substance, being C8-10 for CAS No. 85536-07-8, C12-18 for CAS No. 91052-49-2 and C14-18 (unsaturated) for CAS No. 91052-28-7. The triester content of all three read-across substances is < 15%, being predominantly mono- and diesters. Due to differences on the degree of esterification of these substances with respect to Glycerides, C8-18 (CAS No. 85536-06-7), a higher bioavailability to aquatic organisms can be expected for the read-across substances. Generally, a higher degree of esterification will result in an increase of molecular size and weight of the substance. At higher molecular size and weight, the potential to cross biological membranes tends to decrease (Guidance on information requirements and chemical safety assessment, Chapter R.11 (ECHA, 2012). Furthermore, in a publication by Wu et al. (2006), it was demonstrated that free C18 unsaturated fatty acids were consistently more toxic than C18 saturated fatty acids to aquatic organisms. Considering the high content of C18 unsaturated fatty acids of Glycerides, C14-18 and C16-18 unsatd. mono-, di- and tri-, (CAS No. 91052-28-7) and the expected higher bioavailability of the three read-across substances, they represent a worst-case scenario for Glycerides, C8-18 (CAS No. 85536-06-7).

The toxicity of Glycerides, C8-C10, mono- and di- (CAS No. 85536-07-8) to algae was evaluated in a study by Hafner (2013). This test was conducted according to OECD 201, under GLP conditions. Desmodesmus subspicatus was exposed to the test substance for 72 hours at nominal loading rates ranging from 4.3 to 100 mg/L (WAF). Analytical measurement of test concentrations was performed via TOC and DOC analysis. After the exposure period, significant effects in growth rate were reported at the highest loading rates (45.5 and 100 mg/L, nominal). The resulting EL50 value (72 h) was determined to be 49 mg/L (based on growth rate, loading rate). The NOELR (72 h) was 20.7 mg/L (based on growth rate, loading rate) and the NOEC 1.19 mg/L (based on growth rate, measured final concentration). The test available for Glycerides, C12-18 mono- and di- (CAS No. 91052-49-2; Hafner, 2013) was also conducted according to OECD 201, under GLP conditions. Desmodesmus subspicatus was exposed to the test substance for 72 hours at nominal loading rates ranging from 4.1 to 99.4 mg/L (WAF). Analytical measurement of test concentrations was performed via TOC and DOC analysis at the start and at the end of the test. Measured concentrations ranged from 4.66 to 11.8 mg/L at a nominal loading rate of 99.4 mg/L and from < LOD to 1.52 mg/L for a nominal loading rate of 20.9 mg/L. Concentrations at the lowest nominal loading rate of 4.1 mg/L could not be detected (< LOD). After the exposure period, significant effects in growth rate were reported in all loading rates except for the lowest, 4.1 mg/L. The resulting EL50 value (72 h) was determined to be 13 mg/L (based on growth rate, loading rate). The NOELR (72 h) was 4.1 mg/L (based on growth rate, loading rate).

 

Nevertheless, the effects observed in these two tests might be caused by direct physical interference of test substance particles with algae cells, rather than intrinsic toxicity. For these tests, Water Accommodated Fractions (WAFs) were prepared by adding the test material into a defined volume of test medium, stirring for a period of 48 hours, followed by a sedimentation period of 1 hour. After the sedimentation period, all WAFs in the test conducted with CAS No. 85536-07-8 contained white flakes, which sedimented. In the test performed with CAS No. 91052-49-2, after the sedimentation period, the WAF with the highest loading rate (99.4 mg/L) was non-homogeneous and turbid. The next two lower loading rates (45.7 and 20.9 mg/L) were turbid, and also the second lowest loading rate (9.7 mg/L) was slightly turbid. The lowest loading rate (4.1 mg/L), the only one at which no effects were observed, was reported to be clear. The WAFs were not filtered for the final tests. According to the authors of the report for CAS No. 85536-07-8, at the highest loading rates (45.5 mg/L and 100 mg/L, for which effects were observed), algae were encased by small dispersed particles in the suspension. This was confirmed by microscopic observation. In the two lower loading rates (4.3 and 9.4 mg/L) no difference compared to the control vessels were observed. In the middle loading rate (20.7 mg/L nominal) cell count was lower and the cell shape was smaller compared to the control. In the second highest loading rate (45.5 mg/L) almost no algal cells could be observed anymore. The remaining cells formed piles which are typical for stress situations. In the highest loading rate (100 mg/L), no algal cells were observed but the whole microscope slide was coated with small oily drops with similar size to that of the algal cells. For CAS No. 91052-49-2, the authors report that at the highest loading rates (45.7 mg/L and 99.4 mg/L), algae cells were smaller and had a different shape (crumpled) compared to those in the control. Furthermore, at these two loading rates, a dense emulsion of oily drops was observed. This was confirmed by microscopic observation. At the middle loading rate of 9.7 mg/L, algae cells were larger than those in the control and small drops were reported. At a loading rate of 20.9 mg/L, the number of algae cells was significantly lower and the solution was turbid. In the lowest loading rate (4.1 mg/L) no difference compared to the control was observed. Based on the above information, mechanical disturbance of cells and cell growth cannot be excluded in these two tests due to emulsified test material, which is very likely to have caused the observed effects.

 

Scientific evidence showed that aquatic toxicity testing of this type of Glycerides is technically very difficult. In an article by Prajapati et al. (2012)(see IUCLID section 6.1.4), the phase behaviour of lipid/surfactant/water phases was investigated, where medium-chain (C8-10) mono-, di- and triglycerides represent the lipid. Phase boundaries between lipids (monoglycerides, diglycerides, triglycerides), surfactant (PEG-35 castor oil) and water were established by visual inspection after an equilibration period, and the results expressed in phase diagrams. Viscosity and particle size distribution were measured. The mixtures with monoglyceride displayed two predominant phases: microemulsion and emulsion phases, whereas di- and triglycerides showed additionally a gel phase. Mixtures of monoglycerides and diglycerides, and of monoglycerides and triglycerides seemed to promote an increase of the microemulsion phase (in the 4 phases equilibrium). Particle size in these mixtures was found to be much smaller than in the monoglyceride sample alone. Microemulsions are solutions with an average particle size < 0.2 µm. This particle size would not be intercepted by a standard filter used in an aquatic toxicity test (generally, pore size of 0.45 µm). Due to their small size, based on visual inspection, clear or translucent solutions might be observed even when these microemulsions are present. Glycerides, C8-10 mono- and di- contains mixed mono and diester C8-10 fatty acids, whereas Glycerides, C12-18 mono- and di- contains 40-70% C12 fatty acids. Therefore, formation of microemulsions in test solutions is possible for these substances.

 

Additionally, one study evaluating the toxicity of Glycerides,C14-18 and C16-18 unsaturated, mono-, di- and tri- (CAS No. 91052-28-7) to algae is available (Salinas, 2013). This test was conducted according to OECD Guideline 201, under GLP conditions. Desmodesmus subspicatus was exposed for 72 hours to the test substance at test concentrations of 3.2, 10, 32, 100 and 320 mg/L (nominal, loading rate). Chlorophyll-a fluorescence (pulsed excitation with light flashes having a wavelength of 430 nm) was the measured test parameter. Nevertheless, cell density was also determined in order to derive the linear correlation between both parameters. After 72 hours of exposure, effects on algal growth rate were observed at concentrations of 100 mg/L and 320 mg/L, leading to an ErL10 of 172 mg/L and a NOELR of 32 mg/L (nominal, loading rate). The ErL50 was determined to be > 320 mg/L (nominal, loading rate). The reported effects are well above the water solubility of the substance (water solubility < 0.05 mg/L) and therefore, they could be caused by physical interference or adsorption between algae and the test substance. Nevertheless, stock solutions were reported to be colourless and clear, and therefore, effects due to toxicity cannot be discarded.

 

Based on the above information, the observed effects in all three tests are expected to be caused by mechanical disturbance of the algae cells rather than due to intrinsic toxicity of the substance. No toxicity up to the highest attainable solubility (before microemulsion formation) is thus expected. Nevertheless, since the lowest NOEC value is within the water solubility range (2-3 mg/L) of Glycerides C8-18 (CAS No. 85536-06-7), 1.19 mg/L, this value is used for PNEC derivation as a worst-case approach.