Transport on a nanoscale; quasi-elastic neutron scattering and molecular dynamic studies

Abstract

The objective of this thesis was to investigate how hydrogen is transported in the polymer electrolyte membrane fuel cell (PEMFC) electrodes. We wanted to investigate if the graphitic catalyst support material can contribute to the transport of hydrogen to the platinum (Pt) catalyst particles in the catalyst layer. Knowledge of the transport mechanism in the PEM fuel cell electrodes can help design better PEMFCs.

 

To perform such an investigation we need methods that can discriminate between hydrogen molecules adsorbed on the graphite surface and hydrogen molecules in the gas phase. One method that is able to do that is quasi-elastic neutron scattering (QENS), described in Chapter 3.  At  low temperatures QENS gives information on the movement of the hydrogen molecules adsorbed on graphitic material. In our experiments we adsorbed hydrogen molecules adsorbed on graphitic materials. In our experiments we adsorbed hydrogen on XC-72, a graphitic material that is commonly used in PEMFC as catalyst support material. In Chapter 5 we report QENS data in the 40 to 70 K range for a hydrogen loading equal to a full monolayer at 2 K.  The data show that the surface self diffusion coefficient on XC-72 is the same as on single wall nanotubes and oriented graphite. We found that the diffusivity is in the 1 to 4 x10-8/m2/s range for the investigated temperatures. In this experiment the surface concentration was not measured during the QENS measurements. To examine the effect of the surface concentration on surface self diffusion we preformed a second QENS experiment. In chapter 7 we report QENS data in the 40 to 90 K range for a hydrogen loading equal to half a monoplayer at 2 K. In this experiment the surface concentration was measured during the QENS measurements. We found that the surface self diffusion increased in this experiment due to the lower hydrogen loading compared to the first QENS experiment. This can be seen from the change in the temperature dependence presented in Chapter 7.

 

To get additional data and better understanding of the system we did equilibrium molecular dynamics (EMD) simulations of hydrogen in contact with a graphite surface. The system is described in Chapter 6. From EMD simulations the motion and energies of all molecules are known and this makes it possible to obtain dynamic and thermodynamic properties of the simulated system. In Chapter 6 we report thermodynamic and kinetic properties of the simulated hydrogen-graphite system. We simulated isotherms from 70 to 370 K and used them to calculate the equilibrium constant of adsorption. We found that the density close to the graphite surface always was higher than in the gas phase. The isotherms were used to investigate the adsorption enthalpy and entropy as function of surface concentration. From the adsorption and desorption rates we proposed a new set of rate equations that give the same values for the equilibrium constant of adsorption as we get from the isotherms.

 

The simulated data was also used to find the surface self diffusion coefficient of hydrogen from 70 to 350 K at different surface concentrations, see Chapter 7. When using the temperature dependence of the surface self diffusion found from QENS, to calculate the diffusivity at 350 K, the values are 4 to 10 times lower than the values from EMD. From both the QENS and the EMD data we found that the diffusion activation energy was 3 times lower than the adsorption energy. The EMD values of Ds was linear with cs-1 for the high density regime. From the simulated data we also calculated the average time between adsorption and desorption events for the hydrogen molecules. This data followed an exponential trend that decreased with temperature. With the values of Ds and the average time we calculated the mean displacement of the molecules. This data enabled us to add to the main objective, to investigate the surface transport of hydrogen along the graphite surface to a catalyst particle in a PEMFC. By investigating an ideal system where we considered the catalyst as as hemisphere, described in Chapter 7, we found that the graphite support will transport at least 50% of the hydrogen reactant to the catalyst particle compared to direct adsorption from the gas phase.