Sulfur removal is important for a fuel cell that uses a hydrocarbon fuel, such as natural gas, liquefied petroleum gas, and gasoline, to prevent the downstream sulfur poisoning of catalysts in the fuel processor and in the fuel cell anode. Although most sulfur species are removed prior to reforming, the reducing environment of the reforming stage (such as autothermal reforming) converts residual sulfur to hydrogen sulfide (H2S). H2S in the reformate must be removed to ensure longevity of the catalysts in downstream processing and in the anode chamber of fuel cell systems. A unique modified ZnO sample with a different morphology has been prepared and comparatively studied together with a commercially available ZnO sample under various conditions. Extremely low H2S outlet concentrationsas low as 20 parts per billion by volume (ppbv)have been observed over the modified ZnO sample for extended periods of times. The sulfur-trap capacity (the amount of H2S trapped before breakthrough) also is dependent on space velocity, temperature, steam concentration, CO2 concentration, and particle size. Higher capacity is observed at higher H2S inlet concentration of 8 ppmv, compared to lower inlet concentrations of 1−4 ppmv. The trap capacity decreases monotonically as the temperature increases. Steam in the reformate inhibits the capture of H2S by ZnO; it seems to shift the equilibrium of the reaction ZnO(s) + H2S(g) ⇔ ZnS(s) + H2O(g) to the left, toward ZnO and H2S. The effect of steam seems to be reversible. Increasing the CO2 concentration in the feed up to 12 vol % decreases the capacity of ZnO for the capture of H2S.

No abstract

This work is part of our efforts to explore more-effective ways to remove hydrogen sulfide (H 2 S) for fuel cell applications. Various absorbents (ZnO, SnO 2 , coprecipitated NiO-MoO 3 , supported CuO-ZnO, V 2 O 5 -ZnO, and ZnO supported on γ-alumina) were tested for H 2 S removal. The absorbents that were wash-coated onto the monolith were compared with particulate traps in the inlet H 2 S concentration range of 0.5-8 parts per million by volume (ppmv). The monolith provides much-higher dynamic capacity (the amount of H 2 S trapped before breakthrough) under the same conditions. The ZnO-based monolith demonstrated the best performance. An extremely low H 2 S outlet concentration (as low as 20 parts per billion, by volume (20 ppbv)) was observed over ZnO-based monolith samples for extended periods of time, under various conditions relevant for the desulfurization of gas products from the autothermal reforming of hydrocarbon fuels for a proton-exchange membrane fuel cell. The capacity of the H 2 S trap is dependent on the monolith characteristics (active component loading per cubic inch, and the number of cells per square inch) and operating conditions (including inlet H 2 S concentration, space velocity, and temperature). Wash-coating of ZnO that was chemically modified by an ammonium carbonate treatment onto a monolith with 400 cells per square inch gave the best H 2 S absorbence with higher dynamic capacity.