The Abraham solvation parameter model is a good approach for analyzing and predicting biological activities and partitioning coefficients. The general solvation equation has been used to predict the solute property (SP) behavior of drug compounds between biological barriers. Gas chromatography (GC) retention time can be used to predict molecular descriptors, such as E, S, A, B & L for existing and newly developed drug compounds. In this research, six columns of different stationary phases were used to predict the Abraham molecular descriptors more accurately. The six stationary phases used were 5% phenylmethyl polysiloxane, 6% cyanopropylphenyl 94% dimethylpolysiloxane, 5% diphenyl 95% dimethylpolysiloxane, 100% dimethylpolysiloxane, polyethylene glycol and 35% diphenyl 65% dimethylpolysiloxane. Retention times (RT) of 75 compounds have been measured and logarithm of experimental average retention time Ln(RTexp) are calculated. The Abraham solvation model is then applied to predict the process coefficients of these compounds using the literature values of the molecular descriptors (Acree Compilation descriptors). Six correlation equations are built up as a training set for each of the six columns. The six equations are then used to predict the molecular descriptors of the illegal drugs as a test set. This work shows the ability to extract molecular information from a new compound by utilizing commonly used GC columns available with the desired stationary phases. One can simply run the new compound in GC using these columns to get the retention time. Plugging in the retention time into the developed equations for each of the column will predict the molecular descriptors for the test compound and will give some information about the properties of the compound.

The experiment successfully established the mathematical correlations between the logarithm of retention time of illegal drugs with GC system and the solute descriptor L from the Abraham model. the experiment used the method of Gas Chromatography to analyze the samples of illegal drugs and obtain the retention time of each one. Using the Abraham model to calculate and analyze the sorption coefficient of illegal drugs is an effective way to estimate the drugs. Comparison of the experimental data and calculated data shows that the Abraham linear free energy relationship (LFER) model predicts retention behavior reasonably well for most compounds. It can calculate the solute descriptors of illegal drugs from the retention time of GC system. However, the illegal drugs chosen for this experiment were not all ideal for GC analysis. HPLC is the optimal instrument and will be used for future work. HPLC analysis of the illegal drug compounds will allow for the determination of all the solute descriptors allowing one to predict the illegal drugs behavior in various Abraham biological and medical equations. the results can be applied to predict the properties in biological and medical research which the data is difficult to measure. the Abraham model will predict more accurate results by increasing the samples with effective functional groups.

Experimental data were obtained for the computation of mole fraction solubilities of three dichloronitrobenzenes in organic solvents at 25oC, and solubility ratios were obtained from this data. Abraham model equations were developed for solutes in tributyl phosphate that describe experimental values to within 0.15 log units, and correlations were made to describe solute partitioning in systems that contain either "wet" or "dry" tributyl phosphate. Abraham model correlations have also been developed for solute transfer into anhydrous diisopropyl ether, and these correlations fit in well with those for other ethers. Abraham correlations for the solvation of enthalpy have been derived from experimental and literature data for mesitylene, p-xylene, chlorobenzene, and 1,2-dichlorobenzene at 298.15 K. In addition, the enthalpy contribution of hydrogen bonding between these solutes and acidic solvents were predicted by these correlations and were in agreement with an established method. Residual plots corresponding to Abraham models developed in all of these studies were analyzed for trends in error between experimental and calculated values.