Hydrogen storage in an iron-nickel cell: a DFT study

S. SIMONETTI, G. CANTO, C. LANZ, G. BRIZUELA, A. JUAN

Malaga

Simposio; 3rd International Symposium on Catalysis for Clean Energy and Sustainable Chemistry; 2016

SEC

Introduction Aresta et al. have investigated the dependence of the rate of the production of biogas upon the concentration of nickel, cobalt and iron at sub-toxic concentration and monitored its composition as amount of hydrogen, methane and carbon dioxide. The distribution of the added metals between the liquid and solid phase has also been monitored [1]. The results show that the addition of any of the listed metals to the sludge may cause the production of a higher amount of biogas and influence the methane or carbon dioxide percentage. Conversely, the effect on the hydrogen production depends upon the metal added, the age of the active sludge used, and its adaptation to the susbtrate. As a general feature, during the acidogenesis phase, nickel reduces, while iron increases, the percentage of dihydrogen in the biogas, while cobalt has no influence. On the hand, nickel and iron materials are sensitive to substances in the operating atmosphere such as hydrogen or water vapor, so that atmospheric degradation [2]. Extensive experimental studies have been conducted to clarify the process but the effect of hydrogen on the material behavior is very complicated; thus, so far no definitive H mechanism on iron-nickel alloys has been established.In this research we study the effect on the Me-Me bonding (Me = Fe, Ni) of one and two hydrogen atoms. The calculations were performed with a DFT based code using a Linear Combination of Atomic Orbitals and considering pseudopotentials for the core electrons as implemented in the SIESTA code [3-10]. A geometry optimization was performed applying relaxation calculations. We have calculated the electronic structure and bonding during the adsorption phenomena. We also analyze the possibility of forming an H2 molecule. Results and DiscussionHydrogens located near material vacancies. The system containing two hydrogens is a little more stable than the system containing one hydrogen, the energy difference is only 0.09 eV. That means that the hydrogen?s prefer to be associated forming a vacancy-H-H (VH2) cluster rather than alone (VH). Studies concerning H accumulation in vacancies showed that VH2 system is the most favourable for pure BCC Fe and FCC FeNi alloy [11, 12]. When two H are present, the H-vacancy distance is little bigger (0.04 Å) comparing with the system containing one hydrogen atom, and is in agreement with getting less repulsion between the hydrogens. On the other hand, the H atoms locate nearer Fe atoms than Ni atoms. The H-Fe overlap population (OP) is bigger than the H-Ni OP. For the final configuration, we found an H-H distance of 2.18 Å. Typical H2 bond is not formed in the vacancy because the calculated H-H distance is longer than in the free H2 molecule (0.74Å). A very little interaction arises between the hydrogen atoms because same antibonding states are filled resulting a small H-H OP. H atoms interact with nearest neighbors Ni and Fe atoms. The H-Fe and H-Ni interactions are developed while the Fe-Fe, Ni-Ni and Ni-Fe bond strength decrease. The metal-metal strength changes due to the local concentration of H. After one H atom adsorption, the strength of the nearest Fe-Fe, Fe-Ni and Ni-Ni bonds decreases to about 89 %, 15% and 1%. The adsorption of an additional H modified the metal-metal bond strength in a lesser percent. The Fe-Fe OP increases between 1% and 6%, the Ni-Ni OP decreases between 1% and 3% while the Ni-Fe OP changes between 1% and 13%. In general, the adsorption of an additional H does not increase the decohesion percent; therefore, more metallic bonds are affected. After one H adsorption, the more noticeable effect is observed in the Fe-Fe bonds; while after the second H location, the Fe-Ni bond is the most affected. We have obtained charge populations following two methodologies based on electronic charge densities. First, the Bader [13, 14] procedure was employed using the Tang implementation [15-17]. Second, the Density Derived Electrostatic and Chemical (DDEC) method developed by Manz [17, 19] was also employed in order to get charge populations that reproduce the electrostatic potential outside the electron distribution. Despite the magnitude of charge are different, the ionic character on each chemical species has the same sign. The Bader charges greater than the DDEC/c3 charges is a known behavior when a comparison between them is performed [18, 19]. In general both methods shows a negative charge for hydrogen mainly taken from iron and secondly from nickel nearest neighbor. However, both chemical species, Fe and Ni, retain the same net polarization as in FeNi bulk.