59th Annual Report on Research 2014

Introduction Industrially, hydrogen is produced via the water-gas-shift reaction at temperatures near 820°C. Palladium (Pd)-based crystalline membranes are the benchmark metallic membrane for hydrogen separations. Amorphous metallic membranes, based on non-precious metal alloys, are more cost efficient than these Pd-based membranes. However, the glass transition temperature (T_g) is a key property of amorphous metals that is related to their thermal stability. Long annealing times of amorphous metals at temperatures near T_g and in the super cooled liquid region (SCLR – the region between T_g and the crystallization temperature) can cause crystallization of the amorphous materials.  Thus, amorphous metals with high T_g (relative to the permeation temperature) and wider SCLR are preferable for hydrogen separation applications.  Elements which can bind hydrogen easily may help improve hydrogen permeability; however elements which bind hydrogen easily may also readily embrittle and fail. Zirconium (Zr) has a large negative enthalpy of hydride formation, so Zr containing amorphous metallic membranes are potential candidates as hydrogen separation membranes. In 2003, S. Hara et al. reported that uncoated Zr₃₆Ni₆₄ has hydrogen permeability only one order magnitude lower than Pd membrane at 673K. Our research goal is to investigate the relationship between SCLR, hydride formation and hydrogen transport properties in amorphous metallic glass membranes. Specifically, this year, we studied the impact of Zr-content on hydrogen permeability of Zr-based amorphous metallic membranes in the binary Cu-Zr glass forming system.

Experimental

 We weighed and arc-melted high purity raw metals to yield master ingots of the desired amorphous compositions. Then we used splat quenching to synthesize ~ 40 micrometer thick membranes of selected compositions. Splat quenching is a technique that rapidly cools a molten metal at rates up to 10^6 K/s. To promote the hydrogen dissociation on the surface, we sputtered thin layers of Pd onto both sides of the splat quenched membranes. After sputtering, we annealed the membranes in vacuum oven at 523K to promote the adhesion of Pd to the surface. We used Rutherford backscattering spectrometry (RBS) to estimate the thickness of the sputtered Pd layer.

Figure 1 RBS plot of Zr-based amorphous metallic membranes with Pd coating on top.   We used X-ray diffraction (XRD) to verify the structure of the as-cast membranes and after permeation testing. We used differential scanning calorimetry (DSC) to measure the thermal properties (including the SCLR). We used our custom-built system to test the hydrogen permeability of membranes. We sealed the un-supported membranes into our stainless steel permeation module with annealed copper gaskets. The stainless steel module is heated in muffle furnace with thermocouple to monitor the temperature near the membranes. We performed permeation testing at temperatures below the measured T_g of the alloy.

Results and future work After Pd sputtering and vacuum annealing, the membranes are still amorphous. We found that the width of the SCLR decreased with increasing Zr content in our binary amorphous metals. We collected the hydrogen permeability of Zr-based amorphous membranes both uncoated and coated with Pd. After six hours of hydrogen permeability testing, all of the membranes crystallized even though we tested the membranes at temperatures lower than the measured T_g (Figure 2).

Figure 2 XRD pattern of as-cast membranes and after hydrogen permeability test. According to our permeability data, we found that there is no relationship between the Zr concentration and the hydrogen permeability. In 2011 S. Hao et al. used first principles to calculate the hydrogen permeability of one of the compositions we experimentally tested. Our experiments gave a lower permeability than what S. Hao et al. found in their simulation. We hypothesize that the lower permeability we observed experimentally compared to the simulation may be the result of an oxidation layer on the surface which causes increased resistance of hydrogen transport into the membranes. The oxidation formation is composition dependent. So these compositions may not be suitable for hydrogen separation. Our future work focuses on the hydrogen permeability of Ni-Nb-based membranes. The compositions in this amorphous alloy system have higher T_g than the Zr-system we studied. We will investigate the influence of doping elements on the T_g, hydrogen permeability and hydrogen embrittlement. We will also try to anneal the Ni-Nb-based membranes to different enthalpy states to establish the relationship between enthalpy states and hydrogen permeability.

Impact This fund was the first, externally funded project the PI received as an assistant professor. It expands the PI’s expertise to gas separation. This fund also supports her first graduate PhD student, Tianmiao Lai. PhD student Tianmiao Lai gave talks at the 2014 North America Membrane Society (NAMS) annual meeting in Houston, TX and the 10th International Congress on Membranes and Membrane Process (ICOM) in Suzhou, China. Currently, a publication on Pd-Ag crystalline membranes is under review and we are preparing publications of our results on amorphous alloys for submission. Tianmiao earned her M.S. in Materials Science in August 2013 and she will defend her Ph.D. thesis in May 2015.