Potassium propionate

Endpoint summary

Description of key information

Short-term toxicity to aquatic invertebrates:

-US EPA Database entry: the short-term toxicity to aquatic invertebrates was determined in Daphnia magna according to [72-2] Freshwater Invertebrate Acute using TGAI or TEP (FIFRA 158.490). Dahnids were exposed to several concentration of propionic acid for 48 h under static conditions. The LC50 value was determined at 22.7 mg/L.

-publication of Dowden and Bennett 1965: Daphnia magna were exposed to different concentrations of propionic acid for 48h and the median tolerance limit was determined to be 50 mg/L.

-publication of Bringmann and Kuehn 1977: The LC50 value either for 24h and for 48h was 2740 mg/L and 50 mg/L, respectively. The main parameter observed was mortality.

- OECD SIDS SIAM assessment 2007 for propionic acid. The acute toxicity to aquatic invertebrates is considered to be above 100 mg/L because in contrast to propionic acid, salts of the free acid, i.e. calcium propionate resulted in an EC50 of > 500 mg/L due to the absence of pH changes in the testing media.

- QSAR estimation with DEMETRA model 1.0.4, linear regression made on 16 molecular descriptors associate to Daphnia sp. toxicity. The regression coefficients have been calculated on the DEMETRA project original dataset, that contains 263 compounds extracted from various databases, split in 220 compounds for the training and 43 for the test set. Based on an experimental result in the training dataset the EC50 was estimated to be 68.39 mg/L not taking into account possible pH changes in the test media.

- QSAR estimation with the Estimation Programs Interface (EPI) SuiteTM v4.1 contains the ECOlogical Structure Activity Relationship (ECOSAR) software v1.11, the QSARs in ECOSAR for both neutral organics and classes with excess toxicity are based on a linear mathematical relationship between the measured log Kow values and the corresponding log of the measured toxicity values (mmol/L) for a suite of training set chemicals within each class of interest. The LC50 was estimated to be 5736.716 mg/L for the neutral organics-acid class of ECOSAR.

-OECD QSAR Toolbox v4.4.1, trendanalysis, based on structural similarity and the profiler Acute aquatic toxicity MOA by OASIS, moleculs were identified to predict the acute aquatic toxicity by linear regression. The estimated value based on 10 other substances was 705 mg/L.

Toxicity to aquatic algae and cyanobacteria:

In a 72 hour acute toxicity study, the cultures of Raphidocelis subcapitata were exposed to dipotassium propionate at nominal concentrations of {0, 12.5, 25, 50, 100, and 200 mg a.i./L under static conditions in accordance with OECD TG 201 (March 2006). The NOEC and EC50 values based on yield and growth rate were 100 and 200 mg a.i./L, respectively. 

Additional information

Short-term toxicity to aquatic invertebrates:

No data are available for potassium propionate, however, data are available for the structurally similar propionic acid and other propionates like calcium and sodium propionate.

A QSAR estimation was performed using the VEGA platform and the DEMETRA model (1.0.4). The DEMETRA model compares substances with data from the training dataset. Any structural or fragmental similarity is taken into account and evaluated based on the applicability domain index (ADI). Propionic acid is contained in the training dataset with an experimental LC50 value which was obtained from a study conducted equivalent or similar to OECD guideline 202. The parameter observed in this study was immobility. This 48h LC50 value was reported as 16.51 mg/L. Furthermore, the 48h LC50 value was also predicted on structural similarity and reported to be 68.39 mg/L. Differences are considered to result from pH changes that may occur during testing but are not accounted for in the estimation.

A second estimation was performed using the CAS number of potassium propionate as identifier in the ECOSAR QSAR model. However, ionic substances cannot be predicted with ECOSAR, thus, the prediction is based on propionic acid. The substance was classified as “neutral-organics-acid” based on its structure. The calculated logKow of potassium propionate (0.33) lies below the threshold (<5) as recommended by the model. Also the molecular weight of potassium propionate is small (112.17 g/mol), thus, the substance falls in the applicability domain of the model and the prediction is considered reliable. The 48 h LC50 value was determined at 5736.716 mg/L. Although this predicted value is high as compared to the experimental determined values, the model estimates the effect level without any pH changes.

In the publication of Dowden and Bennett (1965) the median tolerance limit of propionic acid in Daphnia magna was determined. The 48 h LC50 was reported to be 50 mg/L. In another arthopod species Culex sp. the 48h LC50 was determined at > 10000 mg/L. However, again the tests performed probably were conducted without any adjusted pH value, thus, the LC50 value for propionic acid is considered rather low as compared to propionate ions.

In another publication of Bringmann and Kuehn (1977) the 24h and 48h LC50 values for propionic acid in Daphnia magna were reported to be 2740 mg/L and 50 mg/L, respectively.

Also a database entry was found for propionic acid in the US EPA Pesticide Ecotoxicity Database. The 48h LC50 values were determined according to [72-2] Freshwater Invertebrate Acute using TGAI or TEP (FIFRA 158.490). The LC50 values was reported to be 22.7 mg/L and the respective NOEL was 10 mg/L.

In an OECD SIDS assessment for propionic acid (2007) the available data from calcium propionate was compared to respective data generated with propionic acid. Aquatic toxicity values for all three trophic levels with propionic acid were reported to be in a 96-hour static test with the fathead minnow (Pimephales promelas) 96-h LC 50 = 51.8 mg/L, in a static test with Daphnia magna, the 48-h EC 50 was 22.7 mg/L. In a test with green algae (Scenedesmus subspicatus), the 72-hr ErC 50 (growth rate) was calculated as 48.7 mg/L and the 72-hr EbC 50 (biomass) was calculated to be 43.3 mg/L.

These values, however, were considered to result from pH changes in the test media because in the tests with propionic acid the test solutions were not buffered and it is generally accepted that based on the pKa () of propionate ion at neutral pH it is the predominantly form whereas the unionized form will mainly occur in an acidic environment. Thus, acidification of the test medium will result in higher toxicity. Calcium propionate exposure was reported to result in the following effect levels for all three trophic levels: The 96-hr LC 50 in fish (Leuciscus idus) is >10,000 mg/L, the 48-hr EC 50 in D. magna is >500, and the 72-hr EC 50 (both growth rate and biomass) in algae (S. subspicatus) is >500 mg/L.

 

Based on the aforementioned available data the 48 h LC50 value for short-term toxicity to aquatic invertebrates is estimated at > 100 mg/L because this value is considered the most appropriate value determined for propionates. A Weight of Evidence approach was used in order to take into account the reliability of single studies as well as differences in the phsicochemical properties of the respective source substances. The exposure of arthropods like Daphnia magna to propionic acid is generally expected to result in significant pH changes, i.e. an acidification, which then results in an increased toxicity to aquatic invertebrates. The administration of propionates, i.e. potassium, calcium or sodium propionate was shown to not result in pH changes and thus are expected to exhibit a much lower toxicity to aquatic organisms. Hence, potassium propionate is not classified according to Regulation (EU) No. 1272/2008 (CLP) and the Globally Harmonized System for Classification and Labelling of Chemicals (GHS).

 

Toxicity to aquatic algae and cyanobacteria:

In a 72 hour acute toxicity study, the cultures of Raphidocelis subcapitata were exposed to dipotassium propionate at nominal concentrations of {0, 12.5, 25, 50, 100, and 200 mg a.i./L under static conditions in accordance with OECD TG 201 (March 2006). The NOEC and EC50 values based on yield and growth rate were 100 and 200 mg a.i./L, respectively.  The % growth inhibition in the treated algal culture as compared to the control ranged from 5.6 to 11.3 %.

 

There were no compound related phytotoxic effects.

This toxicity study is classified as acceptable and satisfies the guideline requirements for a Toxicity to aquatic algae and cyanobacteria study.

 

Results Synopsis

Test Organism: green alga (Raphidocelis subcapitata)

Test Type: Static

 

72 hr EC50: > 200.mg a.i./L}                95% C.I.: n.d.

72 hr NOEC: 100 mg a.i./L

 

Endpoint(s) Effected: yield and growth rate

 

The 96 –hr-acute toxicity of potassium Propionate to aquatic algae and cyanobacteria was estimated with The Estimation Programs Interface (EPI) SuiteTM v4.1 contains the ECOlogical Structure Activity Relationship (ECOSAR) software. Identifier used were CAS-number and SMILES code.

 

The QSARs in ECOSAR for both neutral organics and classes with excess toxicity are based on a linear mathematical relationship between the measured log Kow values and the corresponding log of the measured toxicity values (mmol/L) for a suite of training set chemicals within each class of interest. After collecting the training set information for each chemical including estimated log Kow and valid toxicity results, regression techniques are applied to the class-specific data sets to derive mathematical relationships between log Kow and toxicity (often called the resulting algorithm).  These resulting class-specific equations typically take the form of y = mx + b, where “y” represents the toxic effect concentration (i.e. log LC50 in mmol/L) and “x” represents the log Kow value.

 

The QSAR estimation conducted with the ECOSAR model the 96h EC50 value for short-term toxicity to aquatic algae and canyobacteria is 2483.01 mg/L at 25°C. The substance was classified as "Neutral-organics-acid" due to the carboxy moiety contained in the structure. The logKow of the substance (0.33) is below the logKow cut-off of the model, its molecular weight lies within the acceptable range for the prediction made with ECOSAR (< 1000) and the substance exhibits a good water solubility (1000 g/L at 25°C), thus the prediction is considered reliable. The value estimated with ECOSAR is far above the limit concentration of 100 mg/L which is recommended for testing short-term toxicity to aquatic alga and cyanobacteria. Hence the substance is not classified according to Regulation (EU) No 1272/2008 (CLP) and the Globally Harmonized System for Classification and Labelling of Chemicals (GHS) with respect to short-term toxicity to aquatic algae and cyanobacteria.