287946 Theoretical and Experimental Approach to Nitrogen Selective Membranes

Monday, October 29, 2012: 3:35 PM
403 (Convention Center )
Ekin Ozdogan, Panithita Rochana and Jennifer Wilcox, Department of Energy Resources Engineering, Stanford University, Stanford, CA

The main objective of the proposed research project is to develop N2-selective catalytic membrane technology with potential sustainability benefits of indirect CO2 capture and ammonia synthesis. The N2-selective membrane technology benefits from the driving force of N2 flue gas (73 wt.%) stream for indirect CO2 capture as it provides atomic nitrogen on the permeate side of the membrane during separation. With these properties if the H2 is provided as a sweep gas alongside this novel technology, additional benefits, such as formation of ammonia as a byproduct of the N2 separation process would be observed.

Within this study, N2 in particular, has been investigated to determine the mechanism of surface adsorption, dissociation and subsequent atomic diffusion into the bulk crystal structure. Theoretical calculations have been carried out using Vienna ab initio Simulation Package (VASP)[i],[ii]. The electron exchange correlation effects are described by a generalized-gradient approximation (GGA) using the Perdew-Burke-Ernzerhof functional[iii] with a plane-wave expansion with a cutoff of 600 eV. The surface Brillouin zone integration has been calculated using a 7x7x7 Monkhorst-Pack mesh[iv]. For absorption energy estimates, binding energy calculations of atomic N in the BCC structure of bulk V and V-Ru alloys have been simulated using a unit cell of 16 atoms with an optimized lattice constant of 2.98 (3.03 experiment). Atomic N was found to be stable in the octahedral, face-octahedral and tetrahedral interstitial crystal sites (Figure 1.). It was determined that the octahedral site is the most favorable binding site for N within bulk V, with a binding energy of -2.132 eV. N binding in V is nearly two orders of magnitude stronger compared to the well-known H binding[v] case (-0.076 eV). These strong N binding energies indicate that N may have difficulty diffusing through the material. We found through alloying V with Ru, that this binding energy could be tuned so that it approaches that of H. Alloying V with 6.25 at.% Ru yields a weaker N binding energy of -0.889. It is anticipated that increasing the content of Ru in the alloy will lead to a further reduction in the stability of N within the crystal lattice. Optimal N permeability includes a careful balance between N solubility and diffusivity within a given material.

(a)    O-site

(b)   Face-O site

(c) T-site

Figure 1 Schematic representation of atomic nitrogen in the interstitial crystal sites

(a) octahedral, (b) face-octahedral and (c) tetrahedral.

Binding energy calculations are carried at different interstitial sites of Vanadium bulk structure. Nitrogen was found to be stable primarily at O-sites within the bulk V lattice. Bader charge analysis has been carried out to investigate the mechanism associated with bulk absorption phenomena. Additionally, alloys of ruthenium (Ru)-V have been studied indicating that Ru can be used as a dopant to tune the electronic structure of the bulk to enhance atomic diffusion. Surface calculations on the 3 low-index surfaces, V(110), V(100) and V(111) based on DFT indicated that the V(111) surface binds N2 the strongest at the fcc site, with an adsorption energy of 1.4 eV.

Experimental efforts involve pure and mixed gas permeation to determine the N2 and CO2 permeabilities of membrane foils comprised of Group V metals and their different composition alloys with Ru over a range of temperatures from 500 to 1000 C. The schematic of the permeation test set-up is given in Figure 2. In this study, linear regression analyses of the flux isotherms were performed to obtain Sieverts' Constant (n) and it was shown to be 0.5 for N2 permeation through V and Nb membranes. Unlike the lighter gases such as H2 and N2 permeating through the metallic membranes by solution-diffusion mechanism, CO2 was shown to diffuse only through the grain boundaries or defects on the dense metal surfaces. As the CO2 fluxes were compared with N2 fluxes for the pure foil membranes, N2 fluxes were found to be 2 to 3 orders of magnitudes higher than the CO2 fluxes. Characterization of the foils before and after the permeation tests was performed by XRD, XPS and SEM techniques to investigate the surface and subsurface structural changes in the membranes. In total, characterization and flux measurements were compared alongside electronic structure predictions of DFT to determine the performance of Group V metals and their alloys for N2-selective membrane application.


Figure 2.Schematic of Experimental Set-Up


[i] G. Kresse and J. Hafner, Phys. Rev. B 48 13115 (1933).

[ii] G. Kresse and J. Furthmuller, Comput. Mater. Sci. 6 15 (1996).

[iii] P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77 3865 (1996).

[iv] H.J. Monkhorst and J.D. Pack, Phys. Rev. B 13 5188 (1976).

[v] S. Aboud and J. Wilcox, J. Phys. Chem. C 114 10978-10985.

Extended Abstract: File Not Uploaded
See more of this Session: Membranes for Gas Separations I
See more of this Group/Topical: Separations Division