Abraham solvation parameter model: updated correlations for describing solute partitioning into plant cuticles from water and from air

Machine-Learning-Based Prediction of Plant Cuticle–Air Partition Coefficients for Organic Pollutants: Revealing Mechanisms from a Molecular Structure Perspective

“…The number of compounds simulated in the early studies was not more than a hundred; the types of compounds were concentrated, and the distribution behavior was more regular, resulting in a high fitting accuracy for their models [ 2 , 10 ]. Eddula, Xu, Jiang, Huang, Tirumala, Liu, Acree and Abraham [ 6 ] collected 215 measured K ca values and established pp-LFER models with excellent accuracy. However, the application of these models was limited by the number of descriptors.…”

“…It also serves as the main barrier to the interception of atmospheric pollutants [ 4 ]. Apart from the uptake and release of organic pollutants, the plant cuticle is also recognized as an accumulation chamber for organic pollutants [ 5 , 6 ]. The partitioning of organic compounds between the cuticle and air is of great interest due to airborne organic pollutants’ strong affinity to plant cuticles and their potential toxicity [ 7 ].…”

“…It is important to note that, at present, there is a paucity of experimentally measured K ca values, and the interaction mechanisms of organic pollutants between plant cuticles and air are still obscure [ 12 ]. To the best of our knowledge, only a few hundred K ca values for organic pollutants between a few plants and air have been measured experimentally [ 6 ]. The scarcity of these experimental K ca values is mainly attributed to the laborious and time-consuming nature of conducting experiments to measure them, as well as the challenges posed by the diversity of plant species and types of organic contaminants [ 13 , 14 ].…”

“…When few independent variables are included, it is difficult, or sometimes even impossible, for a model to adequately reflect the mechanisms of interest precisely due to the limited number of independent variables available, which are representative of the influencing factors [ 2 , 15 ]. Some studies have solved this problem by developing poly-parameter linear free-energy relationship (pp-LFER) models [ 6 , 10 ]. However, the application of these models has been limited by the scarcity of Abrahamic descriptors and the fact that Abrahamic descriptors are not applicable to certain non-polar compounds [ 16 ].…”