Chemical Kinetics Parameter Estimation for Ammonia Borane Hydrolysis

Onboard hydrogen storage is an enabling factor in the development of fuel cell powered passenger cars. Ammonia borane (AB) hydrolysis is one of the potential technologies for onboard hydrogen storage. In this study, kinetics of catalyzed ammonia borane hydrolysis using ruthenium-supported-on-carbon has been measured. For reacting flows, chemical kinetics determines the rates of heat generation and species production or consumption in the overall energy and mass balances respectively. Kinetic measurements under isothermal conditions provide critical data for the design of hydrolysis reactors. It is, however, not always possible to eliminate the effects of internal diffusion in a heterogeneous chemical reaction. In such cases, the reaction efficiency (η), which depends on the effective liquid phase diffusivity (D_eff) in the catalyst medium, should be determined. Determination of intrinsic kinetic parameters using apparent kinetics data is, thus, a challenge. In this study, the change in AB concentration (C_AB) with reaction time (t) has been directly measured. It was observed that the AB hydrolysis reaction had orders between zero and one in a temperature range of 26°C to 55°C. A unified Langmuir-Hinshelwood (LH) model has been adopted to describe the reaction kinetics. The intrinsic kinetic parameters (A, E_a, ΔH_ads, K₀) as well as D_eff need to be estimated by inverse analysis of the measured C_AB vs t data. Conventionally, kinetic parameters are determined using linear fitting. Sometimes, however, it is impossible to converge to a unique value by using the linear fitting approach as there are several values providing regression coefficients greater than 0.99. In this study, the multiple-variable inverse problem has been solved using a nonlinear fitting algorithm based on Powell’s conjugate-gradient error minimization. This algorithm minimizes errors without using derivatives. As a result, the uncertainties in the kinetic parameter estimation have been significantly reduced by the new approach.