Multivariate relationship of methyl- and dimethyl-phenanthrenes to the maturity of organic matter in sedimentary sequences

At least thirteen parameters relating phenanthrene and methylated phenanthrenes to the maturity of organic matter have been defined in the literature and a mass of the data have been accumulated. However, these parameters are not always effective in many basins. To explore an effective approach for using these big data, this paper re-studies the relationship of methyl- and dimethyl-phenanthrenes to the maturity of organic matter, mainly using partial least square regression (PLSR) and geochemical stabilities of these phenanthrenes. The samples for this study were taken from the shale and mudstone in the Cretaceous Qingshankou Formation in the Songliao Basin, China. It was found that integrated use of methyl- and dimethyl-phenanthrenes can overcome the limitations inherent in using them separately, and thus can enhance the accuracy of maturity calculations. The multivariate regression equation achieved from PLSR utilizes 95% of the variance information contained in the relative abundances of 1-, 2-, 3- and 9-MPs, 1,7- and 2,7-DMPs, and (methyl-phenanthrene [MP]; dimethyl-phenanthrene [DMP]). The coefficients of these phenanthrenes in this equation are consistent with their geochemical stabilities. The square of the correlation coefficient ( R 2 = .94) of the multivariate regression equation is much higher than those (0.69–0.89) of the univariate regression equations derived from the previously-defined phenanthrene parameters. It is suggested that the multivariate regression approach presented in this paper substitute the previously-defined parameters and the corresponding univariate regression equations when they are not effective.