*3.4. Mineralogical Assemblage and Carbonation*

As the hydroxyl groups escape the serpentine octahedral sheets, the remaining atoms of Mg and Si are reorganized through the amorphous components. As the temperature increases, serpentine is transformed into amorphous phases, then recrystallizes into forsterite.

The best mineralogical assemblage is shown to be a mixture of amorphous phases, as observed in sample F with the highest proportion of Mg2<sup>+</sup> leached after both two batches and successive batches carbonation tests. McKelvy et al. [21] observed significantly higher carbonation reaction rates in the presence of amorphous meta-serpentine, formed above 610 ◦C, than in the presence of α-meta-serpentine formed below 580 ◦C and identified as intermediate amorphous components here. When structurally stable Mg-bearing phases are present in the assemblage, such as serpentine or forsterite, material reactivity decreases. In serpentine, Mg atoms are bonded to the hydroxyls groups whereas in forsterite, they are ionically bonded to silica tetrahedron [54]. The differentiation of the two amorphous phases based on the proposed calculation in this paper were revealed to be accurate because the sample showing the highest value of metaserpentine appeared to be the one demonstrating the highest proportion of Mg2<sup>+</sup> leached, in accordance with observations proposed in the literature.

Moreover, reactivity appears to be affected more by mineralogy than by surface area. Larachi et al. [15] provide surface area measurements on calcined samples, from 300 ◦C to 1200 ◦C, showing that it tends to decrease with increasing temperature as dehydroxylation occurs.

Furthermore, observations made here are in accordance to those made by other authors regarding the mineralogical transformations occurring during thermal activation and dehydroxylation [21,22]. Nevertheless, a shift of ideal temperature is observed, as observations made by others suggest that reactions at 650 ◦C were more likely to occur, rather than at 750 ◦C as occurred in the present study. Such

a change can be attributed to numerous factors such as the initial material mineralogy, the experimental set up, and the methodology used to evaluate activation efficiency. For instance, Li et al., [22] used hydrochloric acid to perform lixiviation tests, which is far from neutral pH and weak acid conditions used in the present study. On the other hand, McKelvy et al.'s [21] work set the basis of serpentine dehydroxilation understanding using TGA/DTA and XRD. Conversely, their carbonation conditions used high temperature and supercritical CO2, which is again far from the conditions tested here. Based on their results, past ideal activation conditions were shown to be effective, but not necessarily optimal. Therefore, the present results highlight the importance of considering the mineralogical assemblage alongside the thermal treatment parameters (temperature and residence time). Such an approach will allow us to take into account the effect of the initial material composition and potential specificity of the activation conditions/technique. Indeed, conditions in a rotary kiln will be very different from a furnace or a fluidized bed. As results, a study of the mineralogical assemblage can lead to an accurate optimization of heat activation operating conditions in accordance with the material activated and the equipment used.

#### **4. Conclusions**

In this study, a novel approach of amorphous phase quantification, resulting from serpentine thermal activation, is introduced. It enables a better understanding of their implications in serpentine dissolution using carbonic acid as a lixiviant, in similar conditions to those used in the direct flue gas mineral carbonation process developed at INRS. The following conclusions can be made from this study:


**Author Contributions:** Methodology, C.D.B., L.C.-P., G.D., J.-F.B., M.C.I., G.M.; formal analysis, C.D.B.; writing—original draft preparation, C.D.B., L.C.-P., G.D., J.-F.B., M.C.I. and G.M.; writing—review and editing, C.D.B., L.C.-P.; supervision, L.C.-P., G.D., J.-F.B., M.C.I. and G.M.; project administration, G.M.; funding acquisition, L.C.-P., G.D., J.-F.B., M.C.I. and G.M.

**Funding:** This research was funded by Fond de Recherche Quebecois en Nature et Technology, projet de recherche en équipe 2015–2016.

**Acknowledgments:** This research was funded by 'projet de recherche en équipe' grant from FRQNT. The authors would like to thank Matti Raudsepp. Kate Carroll. Ian Power from the University of British Columbia (Vancouver. Canada) and Connor Turvey from Monash University (Melbourne. Australia) for their advice and help on the XRD and Rietveld refinement application to the serpentine minerals.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
