Hydrogen purification technology uses a thin wall palladium – silver tube that is typically brazed into a manifold. The typical commercial hydrogen purifier uses relative short tubes and significant number of them to produce high purity hydrogen with little regard for the residual hydrogen in the raffinate. Our diffuser design requires both the product and raffinate streams to be as pure as possible, to extract as much hydrogen as possible with very low residual hydrogen in the raffinate. In the past, the diffuser vendor was responsible for both alloy selection and process development, that has left a knowledge gap. In order to fill this gap, this project was initiated to develop alloy process knowledge to compare alloys, to understand the processes, to better facilitate production of diffusers, and to own the technical requirements to produce diffusers that meet design goals and requirements. The diffuser operates between 350 and 430°C. The goal is to have a braze that 1) has with a melting temperature where the operating temperature is between 0.4 and 0.6 Th(abs), 2) is compatible with hydrogen, 3) can be vacuum furnace brazed, and 4) does not require flux.
Five commercially available braze alloys, Palniro-7, Nioro, Incusil, Palsil-10, and Palcusil-25, were selected for evaluation. 2 meters of 0.5 mm wire were acquired from Morgan Metals (Palniro-7, Nioro, Incusil, and Palsil-10) and Prince Izant (Palcusil-25). Sessile drop tests were conducted on each of the alloys using Type 304L stainless steel, Ni-200, and Pd-10 Ag. The tests were conducted by degreasing the surfaces with alcohol, cutting two pieces of the braze wire ~ 10 mm long, weighing the braze, melting the braze in a Centorr vacuum furnace by evacuating to a pressure of ~ 5 x 10-5 Torr or less, heating to the 25°C less than the solidus temperature at 10°C / min, hold for 10 minutes to equilibrate, heat to the braze temperature at 20°C/min hold for 15 minutes, furnace cool to less than 300°C then gas cool if needed for through put. The braze temperatures were 25, 50, and 75°C above the liquidus temperature. The wetting angle was measured using a Keyence VR3000 macroscope by taking a minimum of three images across the broadest section of the deposit. Metallographic cross sections were taken to get perpendicular to the deposit. The samples were mounted in epoxy, ground, polished, and then etched with appropriate solutions. The wetting angle was measured in cross section and the extent of braze erosion was evaluated using the etched microstructure and scanning electron microscopy. Second phases were characterized using energy dispersion spectroscopy.
Results and Discussion
The furnace profile and braze temperature resulted in braze wetting angles that increased, better wetting, with increased temperatures. Not all of the brazes wet out on the as prepared substrates. The Incusil braze did not wet the SS substrate at all, with the braze wire forming a non-wetting braze ball, i.e., wetting angle = 0. The nickel substrate exhibited the highest wetting angle and the lowest measured braze bead. The wetting angles measured to date indicate good wetting, angle > 90°, but additional materials combinations will be characterized. The surface of the brazes exhibit dendritic structures with some blushing on the surfaces.
The alloys were selected to have a melting point more than twice the operating temperature, wetted the nickel and palladium – silver alloy, and could be brazed at the lowest possible temperature. A series of sessile drop tests revealed that most of the tested combinations were successful except the stainless steel; it was not wet by all the braze alloys. Sessile drop and vertical capillary testing with macroscopic examination / characterization were effective at minimizing the number of component-braze trials. Actual diffuser components will be fabricated using the selected process and alloy.
Brazing, precious metals, characterization, sessile drop testing, wetting angles