Chronic low-dose inhalation of arsenic may have long-term health effects. The toxicity of chronic inhalation of arsenic compound is related to its speciation and concentration. Therefore, it is necessary to establish an efficient, sensitive, and accurate method for the simultaneous determination of multiple arsenic species.

This methodological study aims to develop a high performance liquid chromatography-inductively coupled plasma mass spectrometry (HPLC-ICP-MS) method for the simultaneous determination of five arsenic species including arsenobetaine (AsB), dimethylarsinate (DMA), arseniteAs(Ⅲ), monomethylarsonate (MMA), and arsenateAs(Ⅴ) in PM_2.5 after optimizing chromatographic conditions and extraction conditions.

The extraction efficiency and stability of three kinds of extracts (pure water, 0.15 mol·L^-1 nitric acid, and 0.30 mol·L^-1 nitric acid) and four ultrasonic time durations (15, 30, 60, and 120 min) were compared, as well as the stability with and without ascorbic acid, to optimize the extraction method. By isocratic chromatography, the separation abilities of two chromatographic columns (reversedphase chromatographic column and anion exchange column) were compared, and the effects of four mobile phase concentrations (20, 30, 40, and 50 mmol·L^-1 ammonium carbonate) and four pH values (pH=9.3, 9.5, 9.7, and 9.9) on peak time, peak shape, and resolution were compared. The established method was applied to the detection of five arsenic species in 84 PM_2.5 filter membrane samples from Hongkou District, Shanghai, throughout the year of 2019.

Compared with pure water extract, acidic extract was more conducive to extraction stability. Adding an appropriate amount of nitric acid in the extraction solution significantly slowed down the oxidation of As(Ⅲ), while adding ascorbic acid further inhibited the oxidation. The optimal extraction parameters were adding 20 mL 0.15 mol·L^-1 nitric acid and 50 mg ascorbic acid, and performing ultrasonic extraction for only 30 min. Under these parameters, the As(Ⅲ) speciation was stable, the peak shape of As(Ⅴ) and the detection sensitivity were improved. The anion exchange column had a good separation effect on arsenic species, but its resolution and sensitivity were significantly affected by the ion concentration and pH value of mobile phase. The optimal mobile phase condition was 30 mmol·L^-1 ammonium carbonate (pH=9.5). The linear ranges for selected five arsenic species were between 0.50-100 μg·L^-1, and the regression equations showed good linear relationships, r > 0.999. Under standard conditions, when the sampling volume was 1 056 m³ and the constant volume was 20 mL, the limits of detection of the five arsenic species were 0.008-0.017 ng·m^-3, the limits of quantification were 0.027-0.058 ng·m^-3, the relative standard deviations were between 1.68%-6.28%, and the recovery rates of standard addition were 91.4%-107.9%. By applying the established method to the determination of the five arsenic species in the samples collected from Hongkou District, Shanghai in 2019, the median concentrations in descending order were As(Ⅴ), AsB, As(Ⅲ), DMA, and MMA.

The method established in this paper has a high extraction efficiency, stable sample speciation, good resolution, and high sensitivity, coupled with simple and quick operation and accurate determination. It is suitable for simultaneous determination of five arsenic species including AsB, DMA, As(Ⅲ), MMA, and As(Ⅴ) in PM_2.5.