(329b) Study of Proton Transport Using Reactive Molecular Dynamics

            Proton transport in an aqueous media occurs through a combination of conventional diffusion (vehicular diffusion) and hopping mechanism (structural diffusion). The structural diffusion of a proton is represented by the following reaction:

Proton transport in the aqueous domains of a hydrated proton exchange membrane (PEM) is highly influenced by the acidic environment due to the sulfonic acid groups and confined environment due to the nano-channels. We have introduced a new reactive molecular dynamics (RMD) algorithm to study the proton transport as a function of acidity and/or confinement. The RMD algorithm takes input from both quantum mechanical (QM) and macroscopic models and attempts to describe a chemical reaction that is far less detailed than the macroscopic model and much more detailed than QM description. The algorithm incorporates the structural diffusion of a proton in a classical molecular dynamics simulation via three steps (i) satisfaction of triggers (ii) instantaneous reaction (iii) local equilibration. Information regarding the transition state is embedded on the triggers and the numerical values of the triggers are parameterized to macroscopic reaction rate. The first and the last steps of the algorithm ensure that the reaction has the correct starting and ending configuration without dynamically describing the transition state. Therefore, the RMD algorithm modeled for one system can be extended to other systems with similar transition states.

            We have modeled the structural diffusion of a proton in a bulk water system, which has the most reliable data. The determination of the functional form of the triggers are based on the structure of the predominant hydrated complexes (Zundel and Eigen cations) necessary for the structural diffusion of protons and are parameterized to reproduce the experimental rate constant and activation energy. A single hydrated proton in a system of 650 water molecules (infinite dilution) is studied as a function of temperature from 280 K ? 320 K. Since the reactivity is implemented by the algorithm rather than potentials, the model can predict the transport property of the proton once the reaction rate is fitted. The two components (structural and vehicular) of the total charge diffusion and water diffusivity are studied in detail. Proton diffusion in aqueous HCl solutions of concentrations ranging from 0.22 M ? 0.83 M is studied to investigate the pH dependence on proton transport. The presence of the chloride ion disrupts the environment of a proton both structurally and energetically leading to a reduced probability of reaction occurrence and is effectively captured by the triggers. Structural diffusion of a proton decreases with an increase in HCl concentration, resulting in a decrease in the total charge diffusion. The effect of confinement on structural and transport properties of a proton is analyzed by implementing the algorithm to model proton diffusion through water filled carbon nanotubes with different radii (5.42 Å ? 10.85 Å). The model shows that enhanced confinement drastically reduces structural diffusion by disrupting the energetic balance around the Zundel ion.

            Having independently examined the effect of acidity (HCl solution) and confinement (carbon nanotubes), we then proceed to study proton transport in a system that contains both confinement and acidity, namely hydrated PEM (Nafion 1144). We have identified three classes of reactions through which the structural diffusion of proton in hydrated membrane can take place.

                                       
                                   (1)

                                 
                             (2)

                                       
                                   (3)

The prevalence of each reaction is dependent on the degree of hydration of the membrane and location of the hydronium ion. Equation 1 is similar to proton transport in bulk water and will take place along the center of the aqueous channels and reactions shown in equations 2 and 3 will occur at the hydrophobic and hydrophilic interfacial regions. RMD algorithm can be modeled to accommodate all the above reactions and the triggers can register the presence of the sulfonic acid groups and other environmental factors like confinement (as shown in the earlier systems). These three reactions are individually implemented in different stages to better understand the contribution of each reaction to the structural diffusion of proton in the PEM at different hydration levels. Thus the algorithm will allow us to measure the diffusivities in the hydrated membranes and provide a molecular level understanding of how the environment in the nano-aqueous regions impact proton mobility as a function of hydration.