**4. Conclusions**

Ni-UGSO was prepared from a mining residue UGSO and then used to produce CNF and H2 via CC and DR reactions. When mining residues are involved in formulations, there is a concern regarding the variability of its composition. Prior to the present study and the ones published previously [30], four di fferent batches of UGSO have been used to prepare the same catalytic formulations. The elemental analyses of the UGSO have shown that the variations were typically lower than 5% and the subsequent tests demonstrated that the observed conversion and yields deviations were lower than the overall experimental error and, consequently, not statistically significant.

Ni-UGSO 13% at 750 ◦C for the CC reaction and at 650 ◦C for the DR reaction exhibited the best performance in terms of H2 and CNF yields. The literature has already provided the first insight into the factors influencing the formation of CNF and the mechanism of their formation, but the current work reveals the complexity of the latter. Although it is widely thought that the diameter of the CNF depends on the size of the catalyst particles, a more careful literature review along with the results of this work proves that other factors are also important. It has also been proven that carbides are the precursors of CNF and that the CNF-DR have higher structural order than CNF-CC. The type of CNF is also di fferent. TEM images have shown that CNF-DR are fishbone shaped and CNF-CC form into tubular (MWNT) and stacked-cup structures. The results show that Fe is the main precursor of the CNF growth while Ni is more contributing to the split of C–C bonds. In terms of conversion and yield efficiencies, the performance of the catalytic formulations tested is proven at least equivalent to other Ni-based catalyst performances described by the literature. The experiments reported were conducted in a lab scale (g-lab) fixed-bed reactor and serve as a preliminary study. Ongoing work focuses on the production of CNF and H2 in a kg-lab scale fluidized bed reactor to prove the feasibility at a larger scale towards eventual process commercialization.

**Author Contributions:** A.A.: PhD student whose topic is directly related to the content of this manuscript. She has written the first draft of the manuscript. E.-H.B. and F.M.: Professors who co-supervise the scientific part of the PhD project. They have read and corrected the manuscript. M.C.: Post-doctoral fellow who has helped and provided guidance during the experimentation. F.G.: Professor who co-supervise the scientific part of the PhD project. He has read and commented the manuscript. N.A.: Professor who co-supervise the PhD work and he is the scientific and technical Director of the Research team and the overall project. He has defined the content of the manuscript, read, corrected and proofread all intermediate steps until final acceptance.

**Funding:** The authors are indebted to the Natural Sciences and Engineering Research Council (NSERC) of Canada (RDCPJ-500331-16), PRIMA-Quebec (R10-010), and to the industrial partners for providing project's funding. They would also like to thank Gilles L'Espérance, Raynald Gauvin, Nicolas Brodusch, and Jean-Philippe Masse for the assistance provided in terms of TEM analyses. Special thanks are due to Jasmin Blanchard for his scientific and technical contributions to this manuscript, and to all the instrumental specialists at the CCM of the Université de Sherbrooke for the TPR, SEM, and XRD analyses.

**Conflicts of Interest:** There are no conflicts of interest to declare.
