2.3.1. Thermodynamic Investigation

FactSage software was used to study the thermodynamic equilibrium of the C2H4 CC and DR reactions at different conditions of temperature (450–850 ◦C) and molar ratios C2H4/CO2 (1/1-3/1) at atmospheric pressure. Equilibrium composition, heat, and enthalpy of the reaction, as well as the amount of deposited carbon, were studied during this research. This investigation allowed us to choose the experimental conditions. The results are shown in Figures 7–9.

**Figure 7.** Thermodynamic study of DR reaction at different temperature at ratio 1/1.

ΔH is negative for the CC reaction, which means that the reaction is exothermic, and it increases with the increase of temperature (Figure 9). The ΔH of the DR reaction is negative for temperatures below 600 ◦C for both ratios. This means that at temperatures higher than 600 ◦C the reaction is endothermic. The heat of this reaction increases with the increase of the amount of ethylene in the feedstock.

The decline of ΔG with temperature illustrates that the equilibrium is displaced toward the products. For the CC reaction, at temperatures below 700 ◦C, ΔG is positive and, therefore, the reaction is not taking place. For the DR reaction, ΔG is higher at higher C2H4/CO2 ratios, which means that conversion is favored.

**Figure 8.** Thermodynamic study of DR reaction at different temperature at ratio 3/1.

**Figure 9.** Thermodynamic study of CC reaction at different temperature.

For the CC reaction, H2 and carbon amounts at the equilibrium increase with temperature (Figure 9). However, for the DR reaction, the H2 amount increases with the temperature for both ratios, while the C amount decreases with the temperature at ratio 1/1 and reaches its maximum at 650 ◦C at a ratio of 3/1.

Based on these results, the three following temperatures have been chosen to be studied experimentally: 550 ◦C, 650 ◦C, and 750 ◦C. Since we want to maximize carbon and H2 production, a 3/1 ratio of C2H4/CO2 was chosen.

#### 2.3.2. Study of CC Reaction

The effect of temperature and Ni percentage on CNF and H2 yields, as well as carbon growth rate, are presented in Figure 10 and Table 5, respectively.

**Table 5.** Carbon growth rate for the CC reaction using Ni-UGSO with different Ni wt.% (5, 10, and 13) at T = 550 ◦C, 650 ◦C, and 750 ◦C for 2 h TOS.


Yu et al. [32], who used a bimetallic catalyst of Ni–Fe with different Ni loading levels, found that productivity is higher for higher T and higher Ni loading. This work shows that C and H2 yields

increase with temperature (Figure 10). This is in accordance with the previously reported literature [32]. Moreover, the carbon growth rate increases with T because the solubility and diffusion of the carbon in the solid metallic phases also increases with T [33]. This effect is even more pronounced at higher percentages of Ni in the catalyst.

**Figure 10.** Carbon and H2 yields for the CC reaction using Ni-UGSO with different Ni wt.% (5, 10, and 13) at T = 550 ◦C, 650 ◦C, and 750 ◦C for 2 h time-on-stream (TOS).

Since the highest carbon and H2 yields were observed at T = 650 ◦C and wt.% of Ni = 13%, the reaction results at these conditions are presented in detail in Table 6 and Figure 11.



**Figure 11.** Experimental results for CC reaction using Ni-UGSO 13% at 750 ◦C.

Overall, the highest carbon and H2 yields, YH2 = 74.46% and YC = 76.25%, respectively, are observed at T = 750 ◦C and 13% of Ni. Thus, the carbon turnover frequency (TOF) expressed per mass of catalysts was 2.8 gC·gcat<sup>−</sup>1·h−<sup>1</sup> at a flow rate of 30 mL/min (Table 6). When compared to the catalysts used in the literature, this catalyst has shown similar to better performance. Yu et al. [32] have produced 3 gC·gcat<sup>−</sup>1·h−<sup>1</sup> and 2.55 gC·gcat<sup>−</sup>1·h−<sup>1</sup> using bimetallic catalysts Ni–Fe(6-1) and Ni–Fe(5-5), respectively, with a feed of C2H4/CO/H2 (30/10/10). Diaz et al. [34] studied Ni-SiO2 catalyst for the catalytic decomposition of ethylene to produce carbon, between 600 ◦C and 700 ◦C. They obtained the maximum of carbon at 600 ◦C with 2 gC·gcat<sup>−</sup>1·h−<sup>1</sup> for 60 mL/min of C2H4.

Figure 11 shows that the steady-state has been reached very fast during the first 20 min of TOS. The conversion of C2H4 is nearly 100% and starts slightly decreasing in the last 20 min. The hydrogen yield is also constant around 80% for 120 min while the rate of carbon formation is also high and equal to 2.82 gC.gca<sup>−</sup>1.h−1. The observed high and constant rates of carbon and H2 formation are due to the high activity of the catalyst at the beginning of the reaction; moreover, even though carbon was formed, the catalyst did not show any deactivation during the TOS of operation. The latter can be explained by the type of carbon formed. Indeed, the carbon formed was analyzed by SEM and it has been proven that it was under the form of CNF (Figure 12), which was not affecting considerably the access of the reactants at the surface of the catalyst.

**Figure 12.** SEM analysis of CNF produced at 750 ◦C using Ni-UGSO 13% for CC reaction.

#### 2.3.3. Study of DR Reaction

The effect of temperature and Ni percentage on CNF and H2 yields, as well as carbon growth rate, are illustrated in Figure 13 and Table 7, respectively.

**Figure 13.** Carbon and H2 yield for the DR reaction using Ni-UGSO with different Ni wt.% (5, 10, and 13) at T = 550 ◦C, 650 ◦C, and 750 ◦C for 2 h TOS.


**Table 7.** Carbon growth rate for the DR reaction using Ni-UGSO with different Ni wt.% (5, 10, and 13) at T = 550 ◦C, 650 ◦C, and 750 ◦C for 2 h TOS.

Figure 13 shows that the carbon yield for all Ni contents has a maximum at 650 ◦C. Although it seems that the same applies to the H2 yield, the latter keeps increasing in the case of the 10% Ni content catalyst. The amount of Ni active sites is a parameter that plays a significant role in terms of catalytic activity. The BET results in Table 1 show that the specific surface and the average pore volume is not a function of the Ni content in the range between 5 and 13 wt.%. The small difference observed in the case of 10 wt.% Ni catalyst is within the experimental error and cannot be used as a differentiation argument. In light of the above, the difference in H2 yield observed in the case of the 10 wt.% Ni catalysts might be explained in the following way:


Since the target of the manuscript is the comparison of two regimes with a number of Ni-UGSO formulations, there are no available surface data to support further discussion. Our continuous efforts are now focusing namely on these aspects.

Both carbon and hydrogen yields are maximal at 650 ◦C. This behavior can be explained by carbon and H2 formation and consumption reactions. In fact, carbon is produced from C2H4 decomposition and Boudouard reaction and consumed by the gasification reaction, while H2 is produced by C2H4 decomposition and consumed by RWGS reaction. C2H4 decomposition and RWGS reactions are favored by high temperatures. Boudouard has a thermodynamic maximum of carbon formation around 550 ◦C. At T lower than 650 ◦C, the formation rate exceeds the consumption rate

Concerning the effects of Ni, the yield of carbon and H2 increases with the increase of the Ni weight percentage in the catalyst, and this is attributed to the higher catalytic activity at higher Ni loading levels and consequently faster reaction rates. We can notice an exception for Ni-UGSO 13% at T = 550 ◦C where the yields are very low. This could be explained by the fact that the catalyst at such a low T with such a high load of Ni has not reached its highest activity within 2 h.

Since the highest carbon and H2 yields were observed at T = 650 ◦C and wt.% of Ni = 13%, the reaction results at these conditions are presented in detail in Table 8 and Figure 14.

Ni-based catalysts have been used in the past for DR reactions, especially for methane and ethanol, but there are few studies on ethylene dry reforming. Jankhah et al. [15] examined in detail the dry reforming reaction of ethanol using activated stainless-steel strips as a catalyst (strip surface of 0.04 m2). Experiments have shown that the results that give the best yields of carbon and H2 are obtained at a temperature of 550 ◦C. They have obtained a carbon rate equal to 3.6 <sup>g</sup>·h−<sup>1</sup> and an H2 yield of 76.33%. Since this catalyst is 2D and not 3D, the equivalent carbon TOF is related to the catalyst surface and not to the weight and is equal to 90 <sup>g</sup>·h−1·m<sup>−</sup>2.

**Table 8.** General experimental results for the DR reaction at 650 ◦C and Ni-UGSO 13% for 2 h TOS.


**Figure14.** ExperimentalresultsforDRreactionusingNi-UGSO13%at650 ◦Cfor2hTOS.

We observed that, during the first 100 min, the conversion of C2H4 is near 100% and starts slightly decreasing during the last 20 min. Hydrogen yield is constant for 120 min and equal to 65% (Figure 14), and the rate of carbon formation is high and equal to 2.25 gC·gcat<sup>−</sup>1·h−1. These high and constant rates

of carbon and H2 formation are due to the high activity of the catalyst and can be explained by the following: the H2 formed contributes to the additional activation of the catalyst through the reduction of iron oxides. This is proven by the presence of Fe and Ni metal peaks and the disappearance of iron oxide peaks on used catalyst XRD (Figure 15). It has been demonstrated that the carbon under the form of catalytically induced CNF itself has catalytic properties [4]. Although the activity measured through carbon TOF per mass of CNF is lower, if the TOF is calculated per mass of carbides content of the CNF, it is shown to be higher. This explains, at least partially, why the catalytic activity remains high even when the catalyst surface is covered by CNF.

**Figure 15.** XRD analysis of Ni-UGSO 13% after the CC reaction at 750 ◦C and after the DR reaction at 650 ◦C for 2 h TOS.

The carbon formed was analyzed by SEM and it has been proven that it is mainly under the form of CNF (Figure 16).

**Figure 16.** SEM analysis of CNF produced at 650 ◦C using Ni-UGSO 13% for DR reaction for 2 h TOS.

#### *2.4. Characterization of CNF and Spent Catalyst*
