2.2.1. Thermal Activation

Heat treatments were performed in a muffle furnace, Furnatrol I33 (Thermolyne Subron Corporation). Twenty grams of samples were placed in a cast-iron skillet in a thin surface, then introduced into the furnace. Cast iron was chosen for its resistance to drastic temperature changes. No major iron contaminations were observed after thermal treatments according to chemical analysis. Tests were performed under isothermal conditions, meaning that the furnace was set to the targeted temperature beforehand to the test. At the end of the treatment, the furnace was open to help the skillet cool down, for five minutes. The skillet and the sample were weighted and the difference between initial and final masses corresponded to the mass lost induced by heat treatment. A series of seven tests were conducted at different residence times and temperatures, as shown in Table 3.


**Table 3.** Treatment conditions.

#### 2.2.2. Dissolution in Aqueous Carbonation Conditions

Dissolution reactions were performed using carbonation conditions in order to evaluate thermal treatment effects on the developing process. Carbonation reactions were conducted in a 300 mL stirred reactor, model 4561 of Parr Instrument Company (Moline, IL, USA) [29,31]). The tests were performed using a certified composition gas of 4.0 vol% O2 and 18.2 vol% CO2, balanced with N2, simulating cement plant flue gas. The pulp density was set to 15% (150 g·L<sup>−</sup>1), with a volume of 75 mL of water, 11.25 g of solid and a gas phase volume of 225 mL, as optimized by Mercier et al. [29]. A batch of 10.2 bar of gas was introduced into the stirred reactor and was allowed to react with the pulp for 15 min at room temperature 22 ◦C ± 3 ◦C with the agitation sets to 600 RPM (Figure 1). As the reaction happens, the pressure decreased between 0.1 and 3.5 bar, depending on the amount of gas initially introduced. At the end of each batch, the pulp was filtered to obtain the liquid for chemical composition analyses. The reactivity of the eight thermally treated samples were tested along with an untreated sample (U). They were only observed throughout the proportion of magnesium leached from the solid during the reaction, using Equation (4) [Mg]liq corresponds to the measured concentration of Mg2<sup>+</sup> at the end of

the reaction, V and m are the volume of the solution and the mass of solid, respectively, and CMg is the measured concentration of Mg in the post-thermal treatment solid. Carbonates were not precipitated from the solution as thermal activation impacts on the Mg leaching. Mg analysis was performed on liquid sample after reaction. The liquid fraction was obtained after filtration of the resulting pulp. Consistency in the procedure was validated by performing mass balance to highlight any precipitation occurring during manipulation or during the pressure release of the vessel.

Equation (4): Proportion of Mg2<sup>+</sup> leached

$$\% \text{Mg} = \frac{([\text{Mg}]\_{liq} \times V) \times 100}{\left(\mathbb{C}\_{\text{Mg}} \times m\right)}.\tag{4}$$

In a successive batches test, the solid was used for 12 batches of gas. Every two batches, the solid was filtered and reused with fresh liquid in the subsequent batches. After six batches, the solid was filtered, dried at 60 ◦C, and ground for 1 min at 700 RPM in a ring mill, to partially remove the silica layer formed around the grains and then re used for another series of 6 batches as described by Figure 2. The liquid phase was sampled and renewed every two batches to prevent saturation. Long term reactivity of samples D and F treated at 650 ◦C for 30 min and 750 ◦C for 15 min, respectively, were tested in a successive batches experiment.

**Figure 1.** Parr reactor experimental set up.

**Figure 2.** Batches dissolution experiments.

#### **3. Results and Discussion**

#### *3.1. Mass Loss*

The proportion of mass lost by each sample during thermal treatments has been registered and is presented in Table 4. As expected, the proportion of mass lost during treatment increased with the temperature. It reached a peak at 14.5% for sample F and was treated at 750 ◦C for 15 min. This value is in agreement with the expected one, between 12.0% and 14.0% [48,49].

**Table 4.** Mass lost during each thermal treatment, expressed as percent of initial mass of sample.

