Several studies in the literature have reported that nickel particle size is closely linked with carbon deposition during hydrocarbon steam reforming. Previous work in our group using Ni/CZO catalysts for isooctane autothermal reforming has shown that high nickel loadings lead to high rates of carbon deposition. Catalysts with the smallest particle sizes, as measured by H₂ chemisorption, showed very low carbon deposition. Nickel particle size and loading were inseparable in the series of catalysts studied, however, and any effect of particle size could not be clearly determined.

The goal of this work is to better understand how variables such as the support preparation technique, calcination temperature, nickel precursor, and calcination atmosphere affect the final nickel dispersion. The effects of various synthetic parameters were studied through the use of fractional factorial designed experiments (DoEs), which allow for the screening of many variables in an efficient manner. The hope is that armed with this information we can synthesize a series of reforming catalysts wherein the only variable is nickel particle size to better examine its effect on carbon deposition.

Figure 1 shows the catalyst surface areas determined by N2 physisoprtion (BET single point method) and Figure 2 presents the nickel dispersions. Clearly the calcination atmosphere has a significant effect on the final surface area of catalysts. The nickel dispersion, however, is not linked to the final catalyst surface area. Those samples with the highest surface areas have some of the lowest nickel dispersions. In general, calcination under air yielded higher nickel dispersions than calcination under nitrogen or reducing atmospheres. This observation shows that sintering of nickel particles is effected by the gas phase. The nickel precursor had little effect on dispersion or catalyst surface area, as both nickel nitrate and nickel acetate gave similar results. Calcination of the support at high temperature prior to impregnation also gave generally higher dispersions. This seems to suggest that stabilizing the support structure prior to adding nickel aids in maintaining higher nickel dispersions. The nickel dispersion is not solely determined by support stabilization however, as some samples retained high surface areas, indicating limited structural changes, but showed low nickel dispersions.

Moving forward this study will focus on varying the nickel loading, calcination atmosphere, and calcination duration for a single support and nickel precursor system. Additional DoEs will be used to quantify the effect of these parameters on the nickel particle size. This should enable us to prepare appropriate catalysts to test for the effect of nickel particle size on reaction performance and catalyst durability.