Predicting toxicities and deriving water quality criteria for antimony based on quantitative structure-activity...

Antimony (Sb) is widely present in the surface water of China, while lack of its water quality criteria (WQC) due to shortage of toxicity data of local organisms, makes it a big challenge for ecological risk assessment and pollution control. Based on the toxicity data of other heavy metals collected from literature, quantitative structure-activity relationship models were constructed using four algorithms, to predict the toxicities of Sb toward three representative native aquatic organisms in China. The multiple linear regression (MLR) model exhibited stable goodness-of-fit (R² > 0.80) across all three organisms, highlighting the close correlations between acute toxicity and physicochemical parameters such as softness index (σp) and electronegativity (X_m) of the heavy metals. Using species sensitivity distribution method, the general short-term (SWQC) and long-term WQC (LWQC) values for Sb were derived as 553 μg/L and 98 μg/L, respectively. After quantifying the effects of water chemistry parameters via stepwise MLR, the Sb toxicity end point values were positively correlated with the hardness and negatively correlated with pH of water. Based on this, the SWQC and LWQC were determined within the ranges of 321-11,200 μg/L and 45-1260 μg/L, respectively, taking into account water hardness levels of 50 to 200 mg/L CaCO₃ and a pH range of 7.0 to 8.0. Ecological risk assessment revealed higher Sb pollution risk in some sites of Yangtze and Pearl River basins. This study provides a novel perspective for developing WQC for pollutants with limited toxicity data and scientific support for the precise management of Sb-contaminated water environments.