Development of correlations for describing solute transfer into acyclic alcohol solvents based on the Abraham model and fragment-specific equation coefficients

Abstract

Gas-to-alcohol partition coefficients have been compiled for 1880 different solute-alcohol combinations, which comprised 23 acyclic alcohols. These partition coefficients were converted into water-to-alcohol partition coefficients using the corresponding gas-to-water partition coefficients. Both sets of partition coefficients were analyzed using the Abraham solvation parameter model with fragment-specific equation coefficients. The derived equations correlated the experimental gas-to-alcohol and water-to-alcohol partition coefficient data to within 0.14 and 0.15 log units, respectively. The fragment-specific equation coefficients that have been calculated for the CH 3 , CH 2 , CH, C and OH fragment groups can be combined to yield expressions capable of predicting the partition coefficients of solutes in other anhydrous alcohol solvents.2

Comment on “Development of Correlations for Describing Solute Transfer into Acyclic Alcohol Solvents Based on the Abraham Model and Fragment-Specific Equation Coefficients”

“…If 6 more fitting parameters are added to account for the slight dependency of the partitioning coefficients on the number of carbon atoms of alcohols, the accuracy improves to 0.15-0.16, comparable to the performance by the 30-parameter model proposed by Sprunger et al [1]. …”

“…In other words, a fragment increment in the structure of alcohol has little influence on log K. Obviously, in this situation it is not possible to accurately calibrate a fragment contribution (if any) to the LSER system parameters. The apparent good fitting reported by Sprunger et al [1] is not surprising, because their models appear to contain more parameters than necessary for their data set. This point is illustrated below.…”

“…Sprunger et al [1] recently reported that partitioning coefficients from air into dry acyclic aliphatic alcohol solvents (K) and from water into these solvents (P) can be described by fragment-specific Abraham linear solvation energy relationship (LSER) models that use CH 3 , CH 2 , CH, C, and OH as fragments. Based on the good agreement between experimental and calculated values of log K and log P (standard deviations (SD) = 0.14-0.15), they concluded that the "methodology of splitting the Abraham model equation coefficients into fragment specific values (.…”

“…For instance, in Table S1 of Ref. [1], 153 solute compounds can be identified for which experimental log K values are available in four alcohol solvents or more. Standard deviations of the log K values over different alcohols are <0.2 for 129 compounds (84%) and <0.3 for 148 compounds (97%).…”

“…Considering the fact that the log K and log P values do not vary much across different alcohol solvents, it may be possible to set up a simpler model with fewer fitting parameters. First, we fitted the LSER model in the original form (6 fitting parameters) to all 1803 experimental log K values and all 1804 log P values listed in [1] for different alcohols (again, ethylene glycol is not considered):…”