*3.3. Catalytic Tests*

The catalysts were tested in a tubular INCOLOY800HT reactor (Meccanica Padana, San Nicolò (PI), Italy) (length 500 mm; internal diameter 10 mm) placed in a furnace as reported elsewhere [28,48]. The bed temperature was controlled with a thermocouple. 1.0 g of catalyst (30–60 mesh) was loaded into the reactor where the pretreatment and eventual reduction of the active phase were conducted by fluxing a 500 mL flow of N2 or of an H2/N2 (10:90 v/v) gas mixture at 500 ◦C. Deionized water was fed by a HPLC pump (JASCO, Easton, MD, USA) and vaporized. The outlet gas (H2, CO, CO2, non-converted CH4 and vapor) was condensed in order to eliminate water. The dry gas mixture was analyzed by a 490 micro gas chromatograph (Agilent Technologies, Cernusco sul Naviglio (MI), Italy) with two di fferent columns. Hydrogen was separated through a MS5A 20 m long (carrier: N2), while CH4, CO, and CO2 were separated with a COx column 1 m long (carrier: He). Both modules were equipped with a Thermal Conductivity Detector TCD. The CEA-NASA software was used to calculate the outlet composition of the stream at the thermodynamic equilibrium. The software gave the molar gaseous outlet composition (non-converted CH4, non-converted H2O, CO, CO2, H2, and deposited

carbon if present), based on the feed composition in terms of molar percentage, reaction temperature, and pressure. Methane conversion was taken as reference in order to evaluate the catalytic activities of the tested samples and was compared with the one calculated at the thermodynamic equilibrium. Methane conversion accuracy was evaluated by calculating standard deviation. This resulted to be lower than ±0.9% for the tests at 750 ◦C and lower than ±0.7% in the tests at 500 ◦C; in a conservative way, we considered accuracy to be ±1% for the test at high temperature and ±0.8% for those at low temperature.
