Ceramic materials and wares, chemicals

Long-term toxicity to aquatic invertebrates

Description of key information

Based on the justification of both main components of the test substance: 
For the aluminium compounds NOECs and EC10s ranged from 0.076 to 4.9 mg Al/L and 0.021 to 0.997 mg Al/L, respectively. Water quality data suggest a direct relationship between toxicity and pH, hardness, and DOC. For studies that experimentally manipulated water quality (e.g., CIMM 2009 and 2010a), toxicity decreased with increasing pH, hardness, and DOC. 
In the environment, lime substances rapidly dissociate or react with water. From these reactions it is clear that the effect of calcium oxide will be caused either by calcium or hydroxyl ions. Since calcium is abundantly present in the environment and since the effect concentrations are within the same order of magnitude of its natural concentration, it can be assumed that the adverse effects are mainly caused by the pH increase caused by the hydroxyl ions.

Additional information

There are no studies available for “Reaction product of thermal process between 1000°C and 2000°C of mainly aluminium oxide and calcium oxide based raw materials with at least CaO+Al2O3 >80% , in which aluminium oxide and calcium oxide in varying amounts are combined in various proportions into a multiphase crystalline matrix”. As this substance is an UVCB substance with aluminium oxide (AL2O3) and calcium oxide (CaO) as main constituents, data based on both main components were taken into account by read across following a structural analogue approach.

 

Aluminium-compounds:

Literature Review for aluminium compounds: Six long-term chronic toxicity studies to two species of aquatic invertebrates(Ceriodaphnia dubia and Daphnia magna) were identified as acceptable studies. EC_r10s were calculated using raw data provided from each study using the statistical program Toxicity Relationship Analysis Program (TRAP) version 1.10 from the US EPA National Health an Environmental Effects Research Laboratory (NHEERL). All other endpoints were as reported in each study. NOECs and EC10s ranged from 0.076 to 4.9 mg Al/L and 0.021 to 0.997 mg Al/L, respectively. Water quality data for these studies suggest a direct relationship between toxicity and pH, hardness, and DOC. For studies that experimentally manipulated water quality (e.g., CIMM 2009 and 2010a), toxicity decreased with increasing pH, hardness, and DOC.

Recent studies conducted by the Chilean Mining and Metallurgy Research Center (CIMM) tested aluminium toxicity to C. dubia and D. magna (one data point) across a range of pH, DOC, and hardness values. These results demonstrated that increasing DOC concentration has a protective effect on aluminium LC50s for invertebrates. Increasing water hardness also had a protective effect. Aluminium toxicity was reduced at high pH, but a larger reduction was observed when changing pH from 6 to 7 than from 7 to 8. 

The acute fish BLM developed for S. salar was applied to the chronic invertebrate data (CIMM 2009, CIMM 2010; Figure 7.1.1.2.2.-1, see attachment) by developing a critical accumulation value appropriate for this organism. In addition, the chronic invertebrate data suggested that overall fit would be improved with a small increase in the Ca binding parameter (i.e. the log K for Ca binding at the biotic ligand was increased from 4.2 to 4.8), which is the same adjusted value used in the chronic fish model. After application of the modified Al BLM, the variability in the response curve data substantially decreased (Figure 7.1.1.2.2.-2, see attachment). These data were subsequently used to establish the CA10 (i.e. the critical accumulation level that results in a 10% reduction in reproduction), and likewise, the CA50. The CA10 and CA50 values can then be used to predict EC10 values and EC50 values in various water types.  

Figures 7.1.1.2.2.-3 and 7.1.1.2.2.-4 (see attachment) provide an evaluation of the ability of the chronic invertebrate Al BLM to predict EC50 and EC10 values. All of the EC50 values are predicted within 2-fold of the reported EC50 values. Most of the EC10 values are predicted within 2-fold of the reported EC10 values, and all of the predicted EC10 values are within 4-fold of the reported values. These results indicate that the chronic Al BLM performs reasonably well for predicting sublethal effects of Al on invertebrates. It should be noted that in both the fish and the invertebrate tests, saturation index calculations suggested that the majority of the toxicity values exceed Al(OH)₃solubility. However, bioavailability factors (i. e. pH, DOC, and hardness) still are consistent with the trends predicted by the Al BLM.

Two additional LC50 values that are not included in this comparison were reported for pH 7 and pH 8 in filtered test media (i. e., filtered before organisms were exposed). The filtered test media were approximately 5-fold less toxic, meaning that their LC50s were approximately 5-fold higher than the results from exposure to unfiltered media. Therefore, toxicity was largely a function of exposure to aluminium hydroxides, which are removed by filtration through these types of filters.

Calcium-compounds:

One long-term chronic toxicity study for calcium dihydroxide with saltwater invertebrates (Crangon septemspinosa) is available. The duration in this study (Locke et al., 2008) was 14 d and this test resulted in a NOEC of 32 mg/L (nominal).

In the environment, lime substances rapidly dissociate or react with water. These reactions, together with the equivalent amount of hydroxyl ions set free when considering 100mg of the lime compound (hypothetic example), are illustrated below:

Ca(OH)2 <-> Ca2+ + 2OH-

100 mg Ca(OH)2 or 1.35 mmol sets free 2.70 mmol

CaO + H2O <-> Ca2+ + 2OH-

100 mg CaO or 1.78 mmol sets free 3.56 mmol

From these reactions it is clear that the effect of calcium oxide will be caused either by calcium or hydroxyl ions. Since calcium is abundantly present in the environment and since the effect concentrations are within the same order of magnitude of its natural concentration, it can be assumed that the adverse effects are mainly caused by the pH increase caused by the hydroxyl ions. Furthermore, the above mentioned calculations show that the base equivalents are within a factor 2 for calcium oxide and calcium hydroxide. As such, it can be reasonably expected that the effect on pH of calcium oxide is comparable to calcium hydroxide for a same application on a weight basis. Consequently, read-across from calcium hydroxide to calcium oxide is justified.