NANOCLUSTER OF LIGHT METAL HYDRIDES FOR HYDROGEN STORAGE: A MULTI-SCALE SIMULATION STUDY (Orkoulas)
According to the DOE’s FreedomCAR target, hydrogen storage for transportation must operate within minimum volume and weight specifications and supply enough hydrogen for a 300-mile driving range. It also must uptake/release hydrogen near room temperature. Solid-state storage (hydrides) method holds considerable promise in meeting the DOE’s target, while other methods (compressed gas and liquid storage) are at a disadvantage because they require a much larger tank volume.
Comprehensive understanding of the interactions of hydrogen with metal hydrides and the properties of complex hydrides in hydrogen absorption and desorption are critical for future design and synthesis of new compounds to meet the DOE’s FreedomCAR targets. Due to the limitation of current technology, light metal hydrides have not been used widely as hydrogen storage devices, though they have many advantages over hydrogen gas tanks in reduced volume and improved safety.
3.1. Diffusion Controlled Absorption/Desorption in Hydride Nanoclusters
Recent experimental results indicate that the diffusion rate of hydrogen atoms through metal hydrides is slow, which implies absorption/desorption is largely controlled by diffusion in bulk metal hydrides. The “Ball Milling” technique [2] has been used to grind metals into micrometer or nanometer sized clusters, and due to the high surface-to-volume ratio of these small clusters, diffusion rate can be improved substantially. This process will be studied using a molecular dynamics simulation based on our previous study of Cu nanoclusters [3] with the embedded atom method (EAM) [4]. EAM is a semi-empirical approach to modeling inter-atomic forces. It is a very accurate approximation of the computationally expensive ab inito quantum mechanical calculations. Through variation of cluster sizes, catalysts, kinetics and recycling stability, this study will lead to better understanding of the diffusion process of hydrogen atoms through the different metal hydrides.
3.2. Binding Energy, Structure and Reversibility in Release/Uptake of Hydrogen
Previous first principle calculations [6-7] show that the electronic structure of solid NaAlH4 can be simply described as mixed ionic, i.e. Na+ , Al3+ , and 4H-. The ionic nature of solid NaAlH4 can be understood as a result of long-range Coulomb interactions in computer simulations. Consequently, melting involves the disruption of the H lattice and the loss of the long-range Coulomb stabilization of the H- ions [6]. In other words, the loss of ionic binding during melting will result in hydrogen release from the material. With this simplest possible ionic picture, it is feasible to utilize molecular dynamics/Monte Carlo simulation to study the behavior of complex hydrides in hydrogen absorption and desorption.
One of the faculty participants (G. Orkoulas) is a leading researcher in the field of simulation of ionic systems, and will be the investigator in charge of this part of the project. [8].
[1] B. Bogdanovic, and et al., “Metal-doped sodium aluminum hydrides as potential new hydrogen storage materials”, Journal of Alloys and Compounds 302, 36 (2000).
[2] A. Zaluska, and et al., “Nanocrystalline magnesium for hydrogen storage”, Journal of Alloys and Compounds 217, 288 (1999).
[4] Y. Mishin, and et al., Interatomic potentials for monoatomic metals from experimental data and ab initio calculations. Physical Review B, 59(15), 3393 (1999).
[5] N. Bazzanella, and et al., Catalytic effect on hydrogen desorption in Nb-doped microcrystalline MgH2. Applied Physics Letters 85 (22), 5212(2004).
[6] A. Aguayo and D. J. Singh, Electronics structure of the complex hydride NaAlH4 , Physical Review B 69, 155103, (2004)
[7] S. A. Bonev, and et al., A quantum fluid of metallic hydrogen suggested by first-principles calculations, Nature, 431, 669 (2004).
[8] Orkoulas G, S. K. Kumar, A. Z. Panagiotopoulos, Monte Carlo study of Coulombic criticality in polyelectrolytes, Physical Reivew Letters, 90(4), 048303, (2003).
[9] Ruichao Ren, and G. Orkoulas, Parallel Monte Carlo Simulation with Modified Markov Chain, Work in progress.