Non-target screening method for VOCs analysis based on GC-QTOF

Volatile organic compounds (VOCs) are an important family of pollutants in the atmosphere, which will produce a series of ecological and environmental effects as well as human health effects. China has issued a series of documents to strengthen the control of VOCs. The monitoring and analysis of VOCs is the premise and foundation of its management and control. Although target analysis has the advantage of clear compound list, it also has the problem of pollutant missing because VOCs are a great variety in structure. To date, non-target screening has become a research hotspot in the field of environmental analysis. In this study, a non-target VOCs screening method was established based on qualitative screening of environmental air samples by GC-QTOF. To optimize mass spectrum signal, the effects of ion source temperature and electron ionization energy on molecular ion peaks were investigated. The results show that the ion source temperature in the range from 200 to 270°C has little influence on the abundance of molecular ion peaks. However, the response of molecular ion peaks for different VOCs varies significantly with the electron ionization energies from 15 to 70 eV, which results from the difference in its reaction ability with electrons because of diverse structure. According to the variation trend of molecular ion peaks, the tested VOCs can be divided into three categories: first, the molecular ion peaks show little change with the increase of electron ionization energy, such as benzene, chlorobenzene and phenol. Second, the molecular ion peak shows an upward trend with the increase of electron ionization energy, as dichloromethane, dibromodichloromethane, etc. Third, molecular ion peak decreases with the increase of electron ionization energy, and even molecular ion peak loss occurs such as phthalates, straight chain alkanes, and some haloalkanes under 70 eV. Therefore, ionization at 70 eV was convenient to match with the NIST mass spectrum library, and Ionization at 15 eV was helpful to obtain the molecular ion signal. Based on the boiling point, a prediction model of VOCs retention time was established. The final non-target screening process includes: (1) Firstly analyze VOCs in the sample under 70 eV, compare the obtained mass spectrogram with the spectrogram library of the instrument to screen out possible candidates; (2) the molecular ion peak was obtained under the condition of 15 eV, and the compound was identified according to the accuracy error of the candidate molecular ion 5 ppm as the mass threshold value; (3) isotopic pattern analysis if it involves; (4) further identification of compounds using the retention time prediction model; (5) standard samples can be used to confirm the identified substances finally. Preliminary testing of 108 standard substances as unknown VOCs in the non-target screening process demonstrated that 94 species could be correctly identified with 87% accuracy. The constructed non-target screening method was used to analyze VOCs in the environmental air samples collected from the workshop and boundary of one factory in Dalian, and 30 VOCs were screened out, among which 23 compounds were confirmed by standard samples. The sample analysis indicates that the non-target screening method constructed in this study can be used for the qualitative identification of VOCs in environment air samples.