Effect of Mixed-Solvent Environments on the Selectivity of Acid-Catalyzed Dehydration Reactions

The composition of the liquid phase can alter the rates of individual reaction steps and thus alter the selectivity of acid-catalyzed reactions, but these solvent effects are difficult to anticipate for design purposes. Herein, we report the kinetics and selectivity of Brønsted acid-catalyzed 1,2-propanediol dehydration in pure water and in aqueous mixtures of the polar aprotic cosolvents γ-valerolactone, 1,4-dioxane, tetrahydrofuran, N-methyl-2-pyrrolidone, tetramethylene sulfoxide, and dimethyl sulfoxide at 433 K. We find that the major product of 1,2-propanediol dehydration is propanal in most mixed-solvent environments with selectivities between 1 and 68 mol %. In contrast, 1,2-propanediol dehydration in aqueous mixtures of dimethyl sulfoxide affords acetone as the major product with up to 48% selectivity with minimal propanal formation. We use classical molecular dynamics simulations to probe these solvent effects by computing the difference between the solvation free energies of 1,2-propanediol and propanal in aqueous mixtures of polar aprotic cosolvents and in pure water. We find that the difference in the solvation free energies is correlated with the rates of propanal formation in all mixed-solvent environments, indicating that the solvent-mediated stabilization of the product state relative to the reactant state translates to increased selectivity toward the same product. Similar agreement between simulated solvation free energies and experimental reaction rates/selectivities is established for the acid-catalyzed dehydration of cis- and trans-1,2-cyclohexanediol and 1,3-cyclohexanediol. Finally, analysis of the solvation environment around 1,2-propanediol shows that dimethyl sulfoxide uniquely competes against water to solvate reactive hydroxyl groups, which causes a change in reaction mechanism in this solvent system that leads to the formation of acetone rather than propanal. These results represent a step toward the computationally efficient screening of solvent systems for acid-catalyzed, liquid-phase processes.液相组成可改变各反应步骤的反应速率，进而改变酸催化反应的选择性，但这类溶剂效应在工艺设计中难以预判。本研究报道了433 K下，布朗斯特酸（Brønsted acid）催化1,2-丙二醇（1,2-propanediol）在纯水以及极性非质子共溶剂（polar aprotic cosolvents）γ-戊内酯（γ-valerolactone）、1,4-二氧六环（1,4-dioxane）、四氢呋喃（tetrahydrofuran）、N-甲基-2-吡咯烷酮（N-methyl-2-pyrrolidone）、四亚甲基亚砜（tetramethylene sulfoxide）与二甲基亚砜（dimethyl sulfoxide）的水溶液混合体系中的脱水反应动力学与选择性。研究发现，在多数混合溶剂体系中，1,2-丙二醇脱水的主要产物为丙醛（propanal），选择性介于1%至68 mol%之间。与之相反，在二甲基亚砜水溶液混合体系中，1,2-丙二醇脱水的主要产物为丙酮（acetone），最高选择性可达48%，且几乎不生成丙醛。本研究通过经典分子动力学（classical molecular dynamics）模拟，计算1,2-丙二醇与丙醛在极性非质子共溶剂水溶液混合体系及纯水中的溶剂化自由能（solvation free energy）差值，以此探究上述溶剂效应。研究发现，在所有混合溶剂体系中，溶剂化自由能差值与丙醛的生成速率呈相关性，这表明溶剂介导的产物态相对于反应物态的稳定作用，可提升对应产物的生成选择性。对于顺式、反式1,2-环己二醇（cis- and trans-1,2-cyclohexanediol）以及1,3-环己二醇（1,3-cyclohexanediol）的酸催化脱水反应，模拟得到的溶剂化自由能与实验反应速率/选择性之间也呈现出类似的一致性。最后，对1,2-丙二醇周围溶剂化环境的分析表明，二甲基亚砜可独特地与水竞争溶剂化反应性羟基基团，进而改变该溶剂体系中的反应机理，最终生成丙酮而非丙醛。本研究结果为酸催化液相反应体系的溶剂体系高效计算筛选提供了可行路径。