Application of HS-SPME-GC-MS combined with electronic nose technology in the odor recognition of pseudostellariae radix

Abstract Correct identification of the substance basis of Pseudostellariae Radix (PR) odor is important not only for the quality control of the products, but also for the safety of the consumers. PR is often described with a special smell, such as strange, moldy or earthy. Electronic nose-based technology coupled with headspace solid phase microextraction gas chromatography-mass spectrometry (HS-SPME-GC-MS) was used to investigate the volatile components in PR from 47 germplasms cultivated in traditional fields. A total of 48 common compounds were identified based on HS- SPME-GC-MS technology, and 25 of them with aroma characteristics were found based on Alpha soft 13.4. The 1-Octen-3-ol, geosim, (E)-2-nonenal and 1- methylnaphthalene as contributing marker compounds of the ‘specific smell’ of PR were identified. The odor recognition mode, with demonstrated excellent accuracy in recognition abilities, enabled the correct identification of commercial samples including complex mixtures. Graphical Abstract


Introduction
moistening lungs, and producing fluids (Chinese Pharmacopeia 2020). Cultivated PR in different fields (Guizhou, Fujian, Anhui and Shandong province) are listed (Table S1). Modern phytochemical studies showed that extracts and constituents of PR exhibited a broad range of biological activities, such as polysaccharides (Chen et al. 2016), amino acids (Lin et al. 2009), cyclic peptides (Morita et al. 1995), volatile components (Hua et al. 2016), etc. Through market investigation of TCM, PR has a special odor characteristic, which is often described as strange, moldy or earthy. Therefore, the quality is also questioned by some users (Qian et al. 2010). The PR odor has been recorded as slight qi and sweet taste in Pharmacopoeia of People's Republic of China (Chinese Pharmacopeia 2020). But The Processing Standard of Traditional Chinese Medicine in Zhejiang Province clearly describes the odor of PR as 'qi specificity' (Zhejiang Food and Drug Administration 2015). A traditional experience in China shows that distinguishing the authenticity and quality of Chinese herbal medicine according to their aroma characteristics is important (Xu et al. 2011). The smell of TCM products is usually influenced by different people's senses. Thus, it is difficult to implement commodity standardization of medicinal materials and to ensure the effectiveness and safety of its clinical application.
In our research effort, aimed at the marker compounds that do not contribute to the therapeutic action but serve only for analytical purposes of PR and its quality formation mechanism. The electronic nose is a very useful technique applied to complex vegetal matrices according to odor evaluation standards. This odor fingerprint analyser can simulate people's olfactory function and identify plant species, harvest season, storage, and geographic origins (Zou et al. 2012;Hou et al. 2015;Ma et al. 2015;Wu et al. 2015). The recent advent of the headspace solid phase microextraction gas chromatography-mass spectrometry (HS-SPME-GC-MS) expanded the possibilities of detecting trace volatile components (Hoffman et al. 1972) in a complex matrix. This approach appears to be a viable alternative in traditional GC-MS analysis according to the sample pretreatment method close to the electronic nose. It is also gradually applied to the field of TCM products (Guo et al. 2007(Guo et al. , 2008Wu et al. 2016;Bai et al. 2017). In the present study, electronic nose-based fingerprint analyser coupled with HS-SPME-GC-MS qualitative determination was performed as odor recognition mode to identify potential trace analytical markers and to quality control in complex matrices such as PR from different cultivated fields.

Evaluation of the response value by electronic nose
The data of the electronic nose determination is carried out on the best conditions of PR (Huang et al. 2020) have exhibited the results of twelve sensors of electronic nose which obey the normal distribution, and the detailed information is listed (Table S2).

HS-SPME-GC-MS data analysis
A total of 48 compounds were quantitatively analyzed using the HS-SPME-GC-MS. This method which was developed, optimized and validated has been applied for the simultaneous determination of 47 batches of PR. 25 compounds with aroma characteristics were identified based on Alpha soft 13.4 (Table S3). Among the common peaks, peak 2, 5,6,7,9,11,14,15,16,17,25,28 and 30 were subject to normal distribution (Table S4).

Conclusion
The methodology used in this work proved to be effective in the analysis of odor profiles in complicated matrices. The use of odor recognition techniques, such as HS-SPME-GC-MS, combined with electronic nose technology, allowed the detection of specific smells of PR even in low concentrations, demonstrating the sensitivity and robustness of the technique in question.

Disclosure statement
No potential conflict of interest was reported by the authors.