2.3.1. Reaction Time

After applying fixed rotary speed and temperatures, the influence of different reaction times were assessed, as listed in Table 3. Figure 9 shows the resulting products exhibiting similar features as

described earlier. For the 10 min product (experiment A, Figure 9a), the oxide was just about to be reduced. Only the outer layer finished the oxide-to-sulphide conversion, whilst the inner core remained intact, which is in line with Tenne *et al.*'s TEM observation [26]. For the 50 min sample (Experiment B, Figure 9b), by analysing the intensity changes of the diffraction peaks, it is clear that the oxide particles have mostly converted to IF-WS2 and there is much less suboxide left in the core. Similarly, after an 80 min reaction, the WO*x* peaks at around 23–25 degrees continued to be reduced, and more and more IF-WS2 layers formed, suggesting an almost complete conversion. Further increase of the reaction time leads to no significant differences in the XRD profiles, and there is almost no peak detected for any tungsten oxide. Thus, a 110 min reaction time is believed to be sufficient for a thorough sulphidisation.

**Figure 9.** XRD profiles of samples from experiments A–F, demonstrating the effect of different reaction time at 800 °C from 10–170 min.

2.3.2. Reaction Temperatures

Of the experiments listed in Table 3, batches C and G used similar parameters, except for the temperature which was 900 °C for batch G but 800 °C for batch C. From 900 °C, the intensity of the resulting IF-WS2 (Figure 10) is much higher than that from 800 °C (Figure 10), confirming more WS2 layers formed at higher temperatures, under the same reaction time of 80 min.

2H-WS2 flakes were presented in both samples (Figure 11), however there was much more 2H-WS2 formed in batch G, possibly due to the higher temperature which made agglomeration more severe. The IF-WS2 particles are also considerably larger in batch G, some of which exhibited a diameter of 300–500 nm, in contrast to 100–200 nm in experiment C. This observation accounted for the extremely high relative intensity of the (002) peak for sample G.

**Figure 10.** XRD profiles of samples from batch C (800 °C, **red curve**) and G (900 °C, **black curve**) and their SEM images.

**Figure 11.** SEM pictures of samples from batch C (**a**–**c**) and batch G (**d**–**f**).

Actually, when compared with other experiments from A–G, the quality of products from batch G (the highest temperature used, 900 °C) are the worst of all, regardless of different reaction times. Thus, temperature is considered to be one of the most significant parameters directly linked with agglomeration in the present process.
