*2.2. Thermogravimetric Experiment*

Combustion experiments were conducted using the HCT-4 thermal analyzer. The usage of anthracite, coke powder and MBF was 1.2 ± 0.1 mg for each group, and 30 ± 0.1 mg for QPF and SDM respectively. Each sample was loaded into a crucible for thermal analysis with a height of 4 mm and diameter of 5 mm. The samples were heated from room temperature (25 ◦C) to 1000 ◦C at heating rates of 5.0, 10.0, 15.0, and 20.0 ◦C/min, respectively. The rate of heating is expressed as β. At the same time, the air was injected at a flow rate of 100 mL/min during the heating to provide an oxidizing atmosphere for the heating. To ensure the accuracy of all experimental results, each experiment was repeated at least three times under the same conditions.

The conversion of the sample (α) was calculated with the mass loss data collected during the combustion

$$\alpha = \frac{m\_0 - m\_t}{m\_0 - m\_\infty} \tag{1}$$

where *m*<sup>0</sup> is the original mass of the sample; *m*<sup>t</sup> is the mass at time t; *m*<sup>∞</sup> is the final mass of the sample after the reaction.

In the thermogravimetric combustion experiment, the parameters of the sample can be determined by using the thermal analysis curve (TG-DTG), including ignition temperature (*T*i), combustion temperature (*T*j), peak temperature (*T*p), combustion reaction time (*t*), flammability index (*C*) and combustion characteristic index (*S*). The determination method of characteristic parameters is shown in Figure 2.

**Figure 2.** Schematic diagram of characteristic parameter determination method of the thermogravimetric curve.

The flammability index reflects the ability of the sample to react at the beginning of combustion. This index can measure the ignition stability of the sample during combustion,

$$C = v\_{\mathbb{P}} / T\_i^2 \tag{2}$$

where *v*<sup>p</sup> is the maximum combustion reaction rate in min<sup>−</sup>1.

The combustion characteristic index reflects a combined characteristic of the ignition and combustion of the sample. If the value of *S* is larger, the combustion performance of the sample is better,

$$S = v\_{\rm p} \times v\_{\rm m} / (T\_{\rm i}^2 \times T\_{\rm j}) \tag{3}$$

where *v*<sup>m</sup> is the average burning rate of the sample from *T*<sup>i</sup> to *T*<sup>j</sup> in min<sup>−</sup>1.
