Rapid quantification of analog complex using partial least squares regression on mass spectrum
- 40 Downloads
Gas chromatograph combining mass spectrometry (GC–MS) has been widely applied to separate complicated complex and quantify specific components. But it was usually time and energy consuming, plus indispensably needed establishing the quantitation curves. To provide an alternative, a new application of the accumulation mass spectra (MS) combined with partial least squares (PLS) regression was used to determine the compositions of analog complex. An analog system with five organic reagents in random proportion and a complex system with 63 botanic leaf samples were detected by GC–MS. After accumulating mass spectra of the retention time on each sample, a serious of tri-dimensional data were created. Based on the chemometrical treatment of mass spectrum, satisfactory results of five chemicals and the nicotine concentration were obtained as the coefficients of determination (R2) of 93.59, 95.99, 98.87, 96.06, 96.02 and 89.16%, and the residual prediction deviation (RPD) of 5.25, 3.91, 4.42, 6.43, 3.35 and 3.26, respectively. Moreover, it was found that low concentration component can be quantified with high sensitivity by specifying the suitable mass-to-charge ratio range from the data matrix. In this study, the proposed combination method was proved to be a feasible, convenient and potential tool for the determination of complex in the future.
KeywordsMass spectra Partial least squares Chemometrics Characteristic peak
This research was supported by the Fundamental Research Funds for the Central Universities of China (Nos: 2652015164; 3142017100), Langfang Technology Research and Development Program (No: 2017013131), and Key Laboratory of Mine Geological Hazards Mechanism and Control Project (KF2017-13).
- Lavine B, Workman J (2004) Chemometrics. Anal Chem 82:4137–4145Google Scholar
- Massart DL, Vandeginste BM, Buydens LC, Jong S, Lewi PJ, Smeyers-Verbeke J (1992) Handbook of chemometrics and qualimetrics part A. Elsevier, LondonGoogle Scholar
- Shariati-Rad M, Hasani M (2010) Selection of individual variables versus intervals of variables in PLSR. J Chemometr 24:45–56Google Scholar