*2.2. Activity Evaluation*

Figure 2 shows the catalytic performance in the CH4 dry reforming reaction with CO2 over the as-prepared catalysts. In the case of catalytic activity, the CH4 conversion over Ni/SiO2-0/0 exhibits a rapid drop from 78.3% to 53.0% in the early ten hours and then gradually becomes stable. In contrast, Ni/SiO2-0/1 gives a milder and continuous decrease in CH4 conversion until the end of the reaction. Surprisingly, Ni/SiO2-2/0 exhibits a stable and higher CH4 conversion over the whole reaction period of 50 hours. Furthermore, Ni/SiO2-2/1 displays a more stable and even higher CH4 conversion. The CO2 and CH4 conversions are similar for all Ni/SiO2 catalysts. However, in the corresponding

reaction period, the CO2 conversion is always slightly higher compared to the CH4 conversion. When the DRM reaction over Ni/SiO2-2/1 is stable, the conversion rates of CH4 and CO2 are 83.6% and 90.6%, respectively, which are slightly lower than their equilibrium conversion rates at 91% and 95% calculated by HSC chemistry 6.0 (Table S1). Also, in the case of the H2/CO molar ratio, it follows the same trend as that for CH4 conversion over all the Ni/SiO2 catalysts. Specifically, for the Ni/SiO2-2/1 catalyst, the desired H2/CO molar ratio at the value of 1/1 is obtained, which results from the efficiently suppressed reverse water gas shift (RWGS) reaction. How the Ni/SiO2 morphology affects the catalytic performance is discussed briefly in the following part.

**Figure 2.** CH4 conversion (**a**), CO2 conversion (**b**), and H2/CO molar ratio (**c**) as a function of time on stream for Ni/SiO2 catalysts prepared with the combustion method at different molar ratios of C2H5NO2 to NH4NO3 (black line: conventional wetness impregnation method; purple line: NH4NO3 only; blue line: C2H5NO2 only; red line: 2/1 ratio of C2H5NO2 to NH4NO3). The reaction was carried at 800 ◦C with 200 mg of catalyst and a molar ratio of CH4/CO2/N2 = 9/9/2 with 160 mL/min.

The DRM reaction is extremely endothermic. Equation (1) shows that the DRM process can produce a syngas with an H2/CO ratio of 1:1. During the DRM process, several reactions simultaneously occur, like CH4 dissociation (Equation (2)), reduction of CO2 to CO (Equation (3)), and the RWGS reaction (Equation (4)).

$$\text{CH}\_4 + \text{CO}\_2 = 2\text{CO} + 2\text{H}\_2 \text{ (}\Delta\text{H}\_{298\text{K}} = +247\text{ kJ mol}^{-1}\text{)}\tag{1}$$

$$\text{CH}\_4=\text{C(s)} + 2\text{H}\_2\text{ (}\Delta\text{H}\_{298\text{K}}=+75\text{ kJ mol}^{-1}\text{)}\tag{2}$$

$$\text{CO(s)} + \text{CO}\_2 = 2\text{CO (\Delta H}\_{298\text{K}} = +171\text{ kJ mol}^{-1})\tag{3}$$

$$\text{CO}\_2 + \text{H}\_2 = \text{CO} + \text{H}\_2\text{O} \text{ (}\Delta\text{H}\_{298\text{K}} = +41.2 \text{ kJ mol}^{-1}\text{)}\tag{4}$$

The driving force for Equations (2)–(4) strongly depends on the temperature, reactant partial pressure and catalyst structures. In the investigated Ni/SiO2 catalysts, both activation of CH4 and CO2 can occur on the active Ni surface since SiO2 support is inert material. It is believed that CH4 activation tends to form an intermediate, like CHx or a formyl group, but dissociates directly to C species and H2 at high temperature. Essentially, the DRM reaction of Ni catalysts might follow a dynamic redox type mechanism as the CO2 oxidizes Ni<sup>0</sup> to Ni+<sup>δ</sup> to give CO, and the oxidative state Ni+<sup>δ</sup> is reduced to Ni<sup>0</sup> by C species as a result of CH4 dissociation. As seen from the above reaction cycle, it is clear that the presence of O from CO2 helps the dissociation of CH4. To avoid the catalyst deactivation resulting from carbon accumulation, the C species from CH4 dissociation must react timely with CO2 to give CO. The reaction rate of this step is closely related to the Ni nanoparticle size, as the larger Ni surface favors the formation of multicarbon Cn species, which are potential precursors of carbon deposits such as coke. The smaller Ni nanoparticles allow a smaller amount of carbon species on the Ni nanoparticle surface. Thus, it is easier to keep the monoatomic C species isolated, and in time, they are oxidized by CO2 to CO. By minimizing the rate of C species combination, the carbon accumulation could be effectively suppressed. Indeed, as shown in Figure 3, Ni/SiO2-0/0, with an average nanoparticle size of 31.3 ± 13.5 nm, gives the highest amount of carbon deposits with 2.7 mg carbon deposits gCH4−<sup>1</sup>

as the BET surface area is decreased to the largest extent (Table S2). In contrast, the Ni/SiO2-2/1 with a smaller nanoparticle size of 6.1 ± 2.7 nm is significantly coke-resistant, as the amount of carbon deposits decreases to 0.9 mg carbon deposits gCH4−1. The above experimental results reflect that the smaller Ni nanoparticle size is favorable to lower carbon deposits and thereby improve the catalyst stability, as shown in Figure S2. It should be noted that, in spite of the significant decrease in carbon deposits over Ni/SiO2-2/1 catalyst, a considerable amount of coke is still formed during the DRM reaction of 50 hours. It can be deduced that most of the carbon deposits might not locate on the Ni nanoparticle surface but are located on the SiO2 support since the catalytic activity is quite stable. It is reasonable for us to imagine that the Ni nanoparticles are lying on the SiO2 support and not confined by porous layer material, which provides a chance for the carbon species to grow continuously along the SiO2 support surface initiated by the Ni nanoparticle and finally form strips of nanofiber.

**Figure 3.** TG patterns of spent Ni/SiO2 catalysts after the dry reforming (DRM) reaction of 50 hours. The catalytic results are shown in Figure 2.

As seen from Figure 2, the H2/CO molar ratio is highly dependent on the CO2 conversion. A lower CO2 conversion can cause a decrease in the molar ratio of H2/CO to a large extent as a result of the RWGS reaction, as the higher concentration of CO2 drives the reaction to the right side (Equation 4). At 800 ◦C, the standard free energy for the RWGS reaction (ΔG<sup>0</sup> = –8545 + 7.84T) and the reduction of CO2 to CO (ΔG<sup>0</sup> = 39810 − 40.87T) [13] is −132.68 kJ mol−<sup>1</sup> and −4043.51 kJ mol−1, respectively. It can be speculated that the reduction of CO2 to CO, C(s) + CO2 = 2CO, occurs more easily as a result of the lower ΔG. Comparing the value of ΔG in the RWGS reaction, the CO2 that oxidizes the C species to CO is more thermodynamically favored than its RWGS reaction. As the lower CO2 conversion corresponds to lower CH4 conversion, the C(s) species dissociated from CH4 is not sufficient for its reaction with CO2. Therefore, the CO2 reacting with H2 toward the RWGS reaction is promoted. In order to minimize the side reaction toward the RWGS, it is necessary to operate the DRM reaction with a high CO2 conversion rate.

#### **3. Materials and Methods**
