Aluminium sulphate

Toxicity to aquatic algae and cyanobacteria

Description of key information

Acute and chronic toxicity to algae was observed at pHs between 6 to 8. EC50s for growth rate based on Total aluminium varied from 0.12 to 3.2 mg/L. EC10s for growth rate were based on measured concentrations of dissolved aluminium and varied from  0.005 to 0.7 mg/L. The differences in endpoint values depended upon the pH, DOC content and hardness of the test medium. The relative importance of these parameters was judged by the reviewer to be pH>DOC>water hardness.

Additional information

Six studies are available for five aluminium salts of which two are strcitly valid following REACH GLP obligations:

Aluminium sulphate ( anhydrous) (10043 -01 -3)

One study was performed on aluminium sulphate (although this is is considered supporting information), using a new algal assay technique to determine the effect of the chemical form of aluminium on toxicity in a synthetic hard water. Maximum toxicity occurred at pH 5.8 to 6.2, where as little as 5 ug/L of labile aluminium significantly inhibited algal growth. The EC50 was determined to be 22.4 mg/L at pH 5.2 and 6.8. Algal assay and computer modelling results implicate Al(OH)2+ as the aluminium species most toxic to algae.

Polyaluminium chloride hydroxide sulphate (CAS 39290 -78 -3)

A valid study was performed with the fresh water green algae Pseudokirchneriella subcapitata following OECD 201 and EC C3 under GLP. The test substance contained the aluminium salt in aqueous solution such that the nominal aluminium content was 4.6%.

Tested concentrations were 0, 0.0463, 0.1463, 0.148, 0.463, 1.48 and 4.63 mg/L as nominal total aluminium.

Measured concentrations: The initial concentrations of aluminium were generally in agreement with the Total aluminium calculations (72-100%) decreasing over the test period reducing to between 15 and 70% of the Total aluminium concentration at the end of the study. From 24 hours of exposure onwards undissolved material was observed in all test concentrations. The pH was maintained between 7.1 and 8.4 over the study period.

ErC50 was 0.644 mg/L based on Total aluminium; 0.24 mg/L based on TWA of measured aluminium

ErC10 was 0.14 mg/L based on Total aluminium; 0.051 mg/L based on TWA of measured aluminium

While measured concentrations were called "dissolved concentrations" in the test report, the actual nature of the aluminium during the test is unclear as no attempts were made to centrifuge the solutions and thus some of the aluminium deamed dissolved may have been dispersed particularly at the beginning of the study.

Aluminum chloride

One valid study with the fresh water green algae Selenastrum capricornutum (previous name) with aluminum chloride is available.

96 hr exposure of Selenastrum capricornutum to aluminium chloride concentrations up to 5.8 mg/L resulted in EC50s for reduction in biomass at pHs of 7.6 of 0.57 mg/L and at pH 8.2 of 0.46 mg/L. Specific growth rates were not determined.

One supporting study with the fresh water green algae Chlorella Pyrenoidosa with aluminum chloride is available. The author determined EC30 values, which cannot be used in the assessment. The EC10 and EC50 (at pH 5) were determined by linear regression by a previous reviewer based on nominal concentrations, available in the article. At pH 5 the EC10 is 0.084 mg/L and the EC50 is 0.179 mg/L (dissolved Al).

 

Aluminium nitrate (13473-90-0)

Two studies (CIMM 2009, 2010) were available including multiple tests on Pseudokirchneriella subcapitata at varying degrees of pH (6, 7, 8), hardness (24.3, 60, 120 mg/L) and DOC (0, 2, 4 mg/L) in the first study and at just two pHs (7.7 and 8) in the second. In the first study EC50s for growth rate were reported for dissolved aluminium and were 16.9, 19 and 20 µg/L based on measured concentrations at pH7 and 266, 321 and 350 µg/L at pH8, for 0 mg/L DOC and increasing hardness of 24.3, 60 and 120 mg/L, respectively. While EC10s were based on estimated Total Al concentrations and ranged from 0.123-0.382 mg/L for studies performed at pH7 and 0.249 to 0.709 mg/L at pH 8. The EC10s were influenced more by pH than by hardness. Unfortunately, data were not provided for dissolved values at the ErC10s or estimated values at the ErC50s, nor for values in the pH6 experiment.

When DOC was increased to 2 mg/L, EC50s for growth rate for dissolved aluminium and increased to 42.8, 38.7 and 55.3 µg/L based on measured concentrations at pH7 and 430, 435 and 451 µg/L at pH8 and increasing hardness of 24.3, 60 and 120 mg/L, respectively. While EC10s were also reported for pH 6 and based on estimated Al concentrations ranging from 0.293 to 1.389 mg/L at pH 6, from 1.969 -2.432 mg/L for studies performed at pH7 and from 0.673 to 1.57 mg/L at pH 8. Again the EC10s were influenced more by pH than by hardness.

At a DOC of 4 mg/L, EC50s for growth rate for dissolved aluminium and increased further to 118.7, 136.6 and 45.5 µg/L based on measured concentrations at pH7 and 476, 478 and 700 µg/L at pH8 and increasing hardness of 24.3, 60 and 120 mg/L, respectively. While EC10s were also reported for pH 6 and based on estimated Al concentrations ranging from 0.582 to 1.258 mg/L at pH 6, from 2.478 -3.155 mg/L for studies performed at pH7 and from 1.086 to 1.509 mg/L at pH 8. Once again the EC10s were influenced more by pH than by hardness.

In the second study EC50s for growth rate at pH7.7 and pH8 of 0.23 and 0.515 mg/L dissolved Al, respectively, were found corresponding to nominal (Total aluminium) ErC50 values of 1.418 and 1.477 mg/L, respectively. Growth rate NOECs were calculated based on elemental concentrations of 0.4 mg/L at pH7.7 and pH8.

While these studies can be considered technically valid they may not be considered valid under REACH as they were performed non-GLP after 2008.

Aluminium nitrate. 9H2O

CIMM performed one study in the hydrated nitrate at just one pH, 6 and three hardnesses, 24.3, 60 and 120 mg/L. EC10s for growth rate based on the element were reported as 0.153, 0.203 and 0.403 mg/L. While DOC was not reported it is assumed to be the same in all cases. Thus in this case hardness appears to play some role in attenuating toxicity.

Conclusion

Acute and chronic toxicity to algae was observed at pHs between 6 to 8. EC50s for growth rate based on Total aluminium varied from 0.12 to 3.2 mg/L. EC10s for growth rate were based on measured concentrations of dissolved aluminium and varied from 0.005 to 0.7 mg/L. The differences in endpoint values depended upon the pH (5.8 to 8), DOC content (0 to 4 mg/L) and hardness of the test medium (23 to 120 mg/L). The relative importance of these parameters was judged by the reviewer to be pH>DOC>water hardness.