1. Molecular Modeling of Structure and Dynamics in Fuel Cell Membranes A. Roudgar, Sudha N.P. and M.H. Eikerling Department of Chemistry, Simon Fraser University, Burnaby, Canada,V5A 1S6 Morphology of Nafion I. Introduction Proper understanding of the relations between structure formation and mobility is critical for the development of highly performing proton conducting membranes for fuel cells. It is, however, impossible to study the complete scale of structural details in real membranes with quantum mechanical approaches (DFT and AIMD). Feasible routes are to utilize combinations of quantum mechanical and classical approaches or to consider small substructures of the membrane. Here we apply ab-initio approaches to simplified model systems. The objective is to understand co-operative phenomena in proton transport and explore effects of length, chemical structure and arrangements of polymeric side chains. II. Model System and Approaches Important characteristics of model system Length of sidechains. Distance between sidechains. Chemical structure of sidechains IV. Conclusion  We study effects of molecular structure on proton, solvent and polymer dynamics in PEMs.  Our model consists of a minimally hydrated 2-D array of sidechains with fixed end points.  We perform full quantum mechanical calculations using VASP.  Total energy calculation as a function of C-C distance was performed.  Upon increasing the C-C distance, a transition from dissociated to non-dissociated state occurs.  We have performed a molecular dynamics simulation for 3(CF 2 SO 3 H + H 2 O) at fixed C-C distance d=7.2Å. Our results show that a transition occurs at t=4.1ps and a new and more stable structure is formed at t=5ps. Nature of acid groups. Number of acid groups on sidechain. Water content. Computational details Two-dimensional hexagonal array with fixed positions of carbon atoms. 3 sidechains + 3 water molecules per unit cell Vienna Ab-initio Simulation Package (VASP) Only Γ point is considered in total energy calculation Projected Augmented Wave (PAW) pseudopotential with cut-off energy E cut =400 eV PW-91 Functional Binding energy as a function of sidechain - sidechain distance  A C-C distance of d=6.18Å corresponds to the largest binding energy - fully dissociated array.  The transition between fully dissociated and fully non- dissociated array occurs at d=7.2Å.  In similar calculations for CH 3 SO 3 H the transition between fully-dissociated and fully non-dissociated array occurs at d=6.7Å (weaker acid).  We expect a high probability of proton transfer in the region of d~7.2Å, where the difference in energies is small. III. Computational simulation of arrays of the simplest and shortest sidechain (CF 3 SO 3 H) PAN-g-macPSSA graft copolymers (32) (graft polymer). Conductivity 0.1 Scm -1 Nature of backbone Conductivity 0.01 Scm -1 (80°C) S-PBI butane S-PPBP Conductivity  0.001 Scm -1 Length of the side chain Partially sulfonated styrene ethylene. Conductivity 0.002 Scm -1 when x=9. Architecture of Membranes Part 2: Ab-initio Molecular Dynamics t=0 t=5.7 ps T t=2.1 ps Step 1: We consider a two- dimensional regular array of sidechains anchored to a substrate. Step 2: We remove the substrate and fix the positions of the endpoint atoms at their initial position. Part 1: Geometry Optimization Top view Fixed carbons Side view Computational details Two dimensional hexagonal arrays with C-C fixed distance d=7.2 3 sidechains + 3 water molecules per unit cell Constant temperature T=300K Nose-Hoover thermostat with Nose mass Q=0.05 PW-91 Functional In initial configuration (t=0) all acids groups are non-dissociated At t>0.5ps the acid head groups start to approach each. Local clusters are formed. A partially dissociated state develops.  At t>4.1ps the system evolves towards a transition state.  The potential energy drops.  Acid groups become fully dissociated  The energy of the new structure is 1eV lower than the initial (non-dissociated) configuration Acknowledgement We gratefully acknowledge the funding of this work by NSERC. The complexity and large number of involved atoms demand simple but reliable models for computational simulation of such a system. Compare the dynamics of the sidechains with and without the substrate (frequency spectra). Chemical architecture of the side chains PS-g-mac PSSA(21) (graft polymer) Conductivity 0.08 Scm -1 Distance between side chains The ionomer consists of an hydrophobic backbone with side chains that are terminated by acid groups. Good proton conductivity of the membrane is due a spontaneous “nanophase segregation” in the presence of water. Non dissociated acid Dissociated acid Top-view References Carmen Chuy, Jianfu Ding,Edward Swanson, Steven Holdcroft,Jackie Horsfall,and Keith V. Lovell, JECS,150(5) E271-E279(2003). M.Eikerling, A.A.Kornyshev, Journal of Electroanalytical Chemistry,502(2001),1-14. K.D.Kreuer, Journal of Membrane Science,185 (2001),29-39. E.Spohr, P.Commer, and A.A.Kornyshev, J.Phys.Chem.B 2002,106,10560-10569. M.Eikerling, A.A.Kornyshev, and U.Stimming, J.Phys.Chem.B 1997,101,10807-10820.