**2. Results and Discussion**

### *2.1. Catalytic Activity on the Ammoximation of MEK*

### 2.1.1. The Catalytic Activity of M-TS-1

Five kinds of transition metal (Fe, Co, Ni, Cu, Ce)-modified TS-1 catalysts were prepared. Moreover, the catalytic performances of M-TS-1 on the ammoximation of MEK were studied. The results are shown in Table 1. It is apparent that the various M-TS-1 catalysts revealed distinct different catalytic activity. The catalytic effect of Cu-TS-1 was the worst. The conversion of MEK was reduced from 98.9% of TS-1 to 74.1%. The selectivity of MEKO was 0, that is, MEK was all converted into by-products. Although the Cu can interact with Ti active sites [20], it has little catalytic e ffect in the ammoximation of MEK. This may be explained that copper may catalyze H2O2 to produce highly active free radicals [23], effectively oxidizing and decomposing MEK and MEKO. In the ammoximation of MEK catalyzed by Fe-TS-1 and Co-TS-1, the conversion of MEK was still lower than that of TS-1, which may be attributed to the oxides of iron or cobalt that can decompose H2O2, which is not conducive to the ammoximation of MEK. For the introduction of Fe, the selectivity of MEKO slightly improved from 92.8% of TS-1 to 95.6%. It confirmed that the electronic e ffect of Fe on Ti centers would activate TS-1 activity [21]. When the rare earth element cerium was introduced, the conversion of MEK was 97.3%, which was similar to TS-1, probably because the cerium's atomic radius was too large to recombine with TS-1. It was worth noting that Ni-TS-1 displayed the best catalytic activity in the conversion of MEK to MEKO. Relative to TS-1, the selectivity of MEKO (99.3%) was significantly improved while maintaining high conversion of MEK (99%). Thus, Ni-TS-1 was intensively studied in the following research.


**Table 1.** Comparisons of the catalytic performance for the ammoximation of methyl ethyl ketone (MEK) over various catalysts.

a Reaction conditions: MEK, 0.1 mol; t-butanol, 25 mL; catalyst, 1.00 g; temperature, 343 K; total reaction time, 2 h. The H2O2 and NH3·H2O were added at a constant rate for 1.5 h. b Total reaction time, 3.5 h. The H2O2 and NH3·H2O were added at a constant rate for 1.5 h. The other conditions were the same with a.

### 2.1.2. The Catalytic Activity of XNi-TS-1

In order to optimize the catalytic activity of Ni-TS-1, the e ffects of Ni dosage in Ni-TS-1 on the ammoximation of MEK was investigated. The results are shown in Table 2. The selectivity of MEKO went up and then down gradually with the increase of Ni dosage. When the dosage was 3 wt %, the catalytic activity was the best. This apparent change may be associated with the structure of the catalyst. When the Ni dosage was small, nickel species did not a ffect the Ti active sites. When the Ni dosage was too high, there were many Ni species to block the channels and cover Ti active sites, thus it was adverse to catalytic reactions.


**Table 2.** The e ffects of Ni dosage on the ammoximation of MEK over Ni-titanium silicalite-1 (TS-1).

Reaction conditions: MEK, 0.1 mol; t-butanol, 25 mL; catalyst, 1.00 g; reaction temperature, 343 K; total reaction time, 2 h. The H2O2 and NH3·H2O were added at a constant rate for 1.5 h.

### *2.2. Reusability Tests of 3% Ni-TS-1*

The main advantages of catalysts in heterogeneous catalytic systems are its good stability and recyclability. The 3 wt % Ni-TS-1 catalyst was reused and its service life was tested. As can be seen from Figure 1, the conversion of MEK and the selectivity of MEKO were not a ffected when used twice. When the catalyst was reused five times, the conversion of MEK remained above 90%, and

the selectivity of MEKO was over 74.1%. The decrease of catalytic activity may be attributed to the loss of a small number of catalysts, increase of surface acidity, and the obstruction of active sites by organic species. The catalyst after each reuse was dried. Then thermogravimetric analysis was carried out, and the weight loss rate was calculated (Table 3). It was found that the weight loss rate increased with the number of uses, and the blockage of pore channels by organic substances became more and more serious. This may be the main factor affecting the catalytic activity. After using five times, 3 wt % Ni-TS-1 was regenerated by calcination and applied to MEK ammoximation reaction (Figure 1). The catalytic activity of 3 wt % Ni-TS-1 was completely restored.

**Figure 1.** Repetition and regeneration of 3 wt % Ni-TS-1 on the ammoximation of MEK (Reaction conditions: MEK, 0.1 mol; t-butanol, 25 mL; catalyst, 1.00 g; reaction temperature, 343 K; total time, 2 h. The total time of other reactions were 3.5 h. The H2O2 and NH3·H2O were added at a constant rate for 1.5 h).


