| This study assesses the extent to which trends in behavior (reaction pathways, product distributions, etc.) for a liquid-phase process can be predicted without fitting any rate constants to process data. The model considers individual elementary steps rather than lumped reactions. A large number of plausible species and reactions are treated so that the importance of a given reaction pathway is an outcome of the calculations and not a starting assumption.; For neutral species, thermochemical quantifies were calculated by combining gas-phase results with a new procedure for evaluating solvation free energies. For charged species, pKa values were derived from data for analogous species. Solution free energies of activation were estimated using the Marcus equation. For ring-opening reactions, this equation was modified to account for the strain released in the transition state.; Aqueous-phase hydrolysis of ethylene oxide (EO) to ethylene glycol was selected as a case study. The model includes 73 species and 203 elementary reactions steps including 52 ring-opening steps. The impurities included are oligomers up to the tetramer, acetaldehyde and derivatives, and ringed structures.; The process was studied over the following range of conditions: pH 1 to 14, temperature 20 to 120°C, and initial EO mole fractions up to 0.2. It was found that, even though no parameters were fitted to process data, the model made reasonable predictions, e.g., published concentration versus time profiles were reproduced by the model. Observed trends such as reaction rate acceleration at low and high pH, increasing oligomer impurities with higher pH values, and low sensitivity of oligomer impurity distribution to temperature were also obtained.; The results indicate that, for the operating conditions considered, S N2 routes predominate over other plausible pathways for glycol and glycol oligomer formation. An SN1 mechanism is responsible for acetaldehyde and associated derivatives such as the 1,1-glycol. These species represent a thermodynamic sink in the chemistry, but the energy barrier for the rate limiting step is at least 9 kcal/mole higher than reaction to the desired glycol. This barrier difference is sufficient to prevent an appreciable EO yield loss to those impurities. |
