Background  Diacetyl (DC) is widely used in the food flavoring industry and excessive occupational exposure to DC can cause serious respiratory diseases. However, there is no corresponding national standard method for the determination of DC in the air of workplace.

Objective  To establish a method for the determination of DC in workplace air by high performance liquid chromatography using 4-nitro-o-phenylenediamine (NPDA) as precolumn derivatization.

Methods  DC in the air of workplace was collected by solution absorption method. This experiment used NPDA as the derivatization reagent. By adjusting acidity of solution and optimizing concentration ration of DC/NPDA, derivatization temperature, and time, a method for the determination of DC in workplace air was proposed, and its performance indexes such as linearity, detection limit, and lower limit of quantification were obtained. Sampling efficiency was evaluated by relative comparison method, and sample stability was evaluated by sample preservation test. Accuracy and precision of the method were evaluated by standard addition recovery test with blank samples, and an interference test was carried out by adding standard samples. The established method was applied to actual samples to evaluate its adaptability.

Results  A combination of 60 °C for 2 h was selected for derivatization because a higher derivatization reaction temperature and a longer reaction time associated with a higher derivatization efficiency. The solution was separated by SB-C18 column (250 mm×4.6 mm, 5 μm) at 30 ℃, using a mixture of methanol and water (v/v, 65%/35%) as mobile phase with an elution flow rate of 1.0 mL·min⁻¹, and was detected with a variable wavelength detector (λ_max=257 nm) by qualitative analysis based on retention time and quantitative analysis based on external standard method. In terms of the proposed method, the linear range of detection was from 5 μg·L⁻¹ to 2000 μg·L⁻¹, with a correlation coefficient of 0.9999, and a detection limit of 1.3 μg·L⁻¹, the quantitative detection of the lower limit was 4.3 μg·L⁻¹, with a sampling volume V₀ of 3.0 L, the minimum detection concentration was 4.3 μg·m⁻³, and the minimum quantitative concentration was 14.3 μg·m⁻³. The recovery rate was 99.1%-100.8%, the intra-batch precision was 0.5%-3.0%, and the inter-batch precision was 1.2%-2.0%. The average sampling efficiency of this method was 94.5%, and the sample could be stored at 4 °C for at least 14 d. The coexisting components in the air of the workplace did not interfere with the determination of DC. The DC content in the air of a flavor workplace was 5.86-8.85 mg·m⁻³.

Conclusion  A determination method for DC in workplace air by high performance liquid chromatography using NPDA as precolumn derivatization after being collected by 1.0% phosphoric acid absorbent is proposed and has the advantages of simple operation, high sensitivity, and good accuracy. With no DC loss and degradation, the method may satisfy the request for DC determination in the air of workplace.