ACS PRF | ACS
All e-Annual Reports
41313-AC6
The Influence of Confinement on Conformational Dynamics in Polar and Nonpolar Solvents
We have investigated how conformational equilibria are modified in solvents confined on the nanometer length scale. Conformational equilibria and dynamics are important in a number of practical applications, e.g., the design of porous catalysts, the development of chemical sensors based on microporous materials, protein folding in the presence of macromolecular crowding, and single-molecule studies in confined fluids. Despite this, there is little understanding about how the properties of the confining framework (e.g., size, shape, flexibility, and interactions with the solvent and/or reactants) should be modified to accelerate or decelerate a particular reaction. This work aimed at addressing this fundamental issue by providing new insight into the effect of nanoscale confinement and the cavity/pore characteristics on conformational properties.
In a first study we calculated the conformational free energies of 1,2-dichloroethane (DCE) in nanoconfined methanol. Specifically, we used Monte Carlo simulations to construct two degree-of-freedom free energy surfaces of 1,2-dichloroethane dissolved in methanol confined in hydrophobic spherical cavities of varying size (1.0, 1.2, and 1.5 nm radius) and solution density (0.6-0.79 g/cm3). The free energy surfaces were calculated as a function of the (center-of-mass) distance from the cavity wall of DCE and the Cl-C-C-Cl dihedral angle. The free energy surfaces are somewhat complex with several local minima. However, we showed that their structure is strongly influenced by the methanol solvent layering in the nanocavity which is determined by both packing and hydrogen-bonding effects. The gauche and trans conformers both had minima corresponding the molecule lying within a solvent layer parallel to the cavity wall; the trans conformer also had minima for the molecule perpendicular to the wall, spanning two solvent layers. The latter orientation is due primarily to packing effects with only slight contributions from electrostatics. Taken as a whole, this work showed that the most stable conformer depends on the molecule position in the nanocavity.
To study more realistic confining frameworks, we developed an approach for modeling roughly cylindrical nanoscale pores in amorphous silica with well-controlled surface chemistry (i.e., hydrophilicity or hydrophobicity). We used these models to investigate the effect of confinement and surface chemistry on the conformational equilibrium of ethylene glycol (EG), a system that has been the subject of previous experimental studies. In particular, we have simulated EG in hydrophilic amorphous silica pores (approximately 12 Å in radius) and examined the trans-gauche equilibrium. We find that, in agreement with experimental results, the fraction of trans conformers is greater in the nanoconfined EG than in the bulk liquid. (Preliminary simulations in hydrophobic pores also have indicated that there is little change in the fraction of trans conformers as has been found experimentally.) In addition, we observed interesting behavior in terms of the heterogeneity between pores and a two-state model frequently used to interpret such equilibria.
Specifically, we found that there was significant heterogeneity in the results for different model pores prepared using the same algorithm but different random starting configurations. Simulation of a single pore would have led to an interpretation that depends on which pore was chosen. Statistically significant differences in the fraction of trans conformers are observed between many of the pores. Thus, this work indicates that simulations of a single amorphous silica pore should be interpreted with great care, and the conclusions accompanied by a caveat. Indeed, we expect that the heterogeneity of our model pores is less than that of pores in the sol-gels probed in experiments.
A widely applied framework for explaining the differences between the bulk and confined behavior of liquids is the two-state model. This model is based on two fundamental assumptions. First, that the liquid can be divided into two distinct subensembles: one consisting of molecules at the “surface,” the other comprised of molecules in the “interior” of the pores. Second, that the interior liquid possess nearly identical properties to that of the bulk liquid. Together, these assumptions determine the physical picture of the two-state model: the deviation of an observed property from its bulk value must be due only to the liquid molecules interacting directly with the surface. However, our results do not support such a two-state model. The simulation data indicate that the two-state model is too restrictive to be fully correct for this system. In particular, a distinct difference in the fraction of trans conformers between the surface and interior regions is not observed, the trans conformer population is not dramatically increased in the surface layer and is not equal to the bulk fraction in the interior. The two-state model may represent a more accurate description of other measurable properties or different nanoconfined liquid systems, but is less likely to be applicable for properties, such as conformational equilibrium, that are influenced by longer-range interactions.
