Optimization of arsenic extraction in rice samples by Plackett–Burman design and response surface methodology

Statistical experimental designs were employed to optimize the extraction condition of arsenic species (As(III), As(V), monomethylarsonic acid (MMA) and dimethylarsonic acid (DMA)) in paddy rice by a simple solvent extraction using water as an extraction reagent. The effect of variables were estimated by a two-level Plackett–Burman factorial design. A five-level central composite design was subsequently employed to optimize the significant factors. The desirability parameters of the significant factors were confirmed to 60   min of shaking time and 85   °C of extraction temperature by compromising the experimental period and extraction efficiency. The analytical performances, such as linearity, method detection limits, relative standard deviation and recovery were examined, and these data exhibited broad linear range, high sensitivity and good precision. The proposed method was applied for real rice samples. The species of As(III), As(V) and DMA were detected in all the rice samples mostly in the order As(III)   >   As(V)   >   DMA.