**Insights over Titanium Modified FeMgO***<sup>x</sup>* **Catalysts for Selective Catalytic Reduction of NO***<sup>x</sup>* **with NH3: Influence of Precursors and Crystalline Structures**

#### **Liting Xu <sup>1</sup> , Qilei Yang <sup>2</sup> , Lihua Hu <sup>1</sup> , Dong Wang 2,\*, Yue Peng <sup>2</sup> , Zheru Shao <sup>1</sup> , Chunmei Lu <sup>3</sup> and Junhua Li <sup>2</sup>**


Received: 3 May 2019; Accepted: 17 June 2019; Published: 24 June 2019

**Abstract:** Titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors were prepared by coprecipitation method with microwave thermal treatment. The iron precursor is a key factor affecting the surface active component. The catalyst using FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> as precursors exhibited enhanced catalytic activity from 225 to 400 ◦C, with a maximum NO*<sup>x</sup>* conversion of 100%. Iron oxides existed as γ-Fe2O<sup>3</sup> in this catalyst. They exhibited highly enriched surface active oxygen and surface acidity, which were favorable for low-temperature selective catalytic reduction (SCR) reaction. Besides, it showed advantage in surface area, spherical particle distribution and pores connectivity. Amorphous iron-magnesium-titanium mixed oxides were the main phase of the catalysts using Fe(NO3)<sup>3</sup> as a precursor. This catalyst exhibited a narrow T<sup>90</sup> of 200/250–350 ◦C. Side reactions occurred after 300 ◦C producing NO*x*, which reduced the NO*<sup>x</sup>* conversion. The strong acid sites inhibited the side reactions, and thus improved the catalytic performance above 300 ◦C. The weak acid sites appeared below 200 ◦C, and had a great impact on the low-temperature catalytic performance. Nevertheless, amorphous iron-magnesium-titanium mixed oxides blocked the absorption and activation between NH<sup>3</sup> and the surface strong acid sites, which was strengthened on the γ-Fe2O<sup>3</sup> surface.

**Keywords:** selective catalytic reduction (SCR); catalyst; precursor; NO*<sup>x</sup>* conversion

### **1. Introduction**

Nitrogen oxides (NO*x*) mainly come from fossil fuel combustion [1–3] and have caused a series of environmental problems such as nitric acid rain, photochemical smog, ozone layer depletion and fine particle pollution [4–7]. As the severe NO*<sup>x</sup>* emission situation and the rigorous emission legislation exhibit, many efforts have been made in NO*<sup>x</sup>* reduction [8]. SCR (selective catalytic reduction) of NO*<sup>x</sup>* with NH<sup>3</sup> [9–12] has been extensively proved to be the most efficient way for the removal of NO*<sup>x</sup>* from stationary sources. A catalyst is critical to create an efficient SCR reaction and operating cost [13]. Commercial catalysts, such as V2O5/TiO2, V2O5-WO3/TiO<sup>2</sup> and V2O5-MoO3/TiO<sup>2</sup> [8,14–18], were constrained for further development because of several inevitable drawbacks such as high cost, the bio-toxicity of the common vanadium compounds, etc. [19,20]. Furthermore, the commercial

vanadium-titanium catalysts have been managed as hazardous waste. The study of novel high-efficiency catalysts is of great significance.

Many efforts have been made on iron oxides based catalysts, such as Fe-Ti [21–23], Fe/Ce-Ti [24,25], Fe-Ce-W [26,27], Fe-Sn-Mn [28], Ce-Fe/WMH [29], WO3/Fe2O<sup>3</sup> [30], Fe/WO3-ZrO<sup>2</sup> [31], Mn-Fe/TiO<sup>2</sup> [32], FeMnTiO*<sup>x</sup>* [33], etc. Iron oxides catalysts have low prices and free secondary pollution. They showed good catalytic performance and high N<sup>2</sup> selectivity for SCR reaction [8,34,35]. NO*<sup>x</sup>* conversion of 60% was obtained over Fe2O3/TiO<sup>2</sup> prepared by Kato in 250–450 ◦C. In Fe/WO3-ZrO<sup>2</sup> [31], Fe-Mn-Ce/γ-Al2O<sup>3</sup> [36], and Fe-Er-V/TiO2-WO3-SiO<sup>2</sup> [37], the introduce of iron enlarged the surface area and pore volume, and meanwhile, it improved the Brönsted and Lewis acid sites. Zhu [38] studied Co-Fe/TiO<sup>2</sup> and Cu-Fe/TiO<sup>2</sup> and manifested that the reactants could be easily adsorbed on Co-Fe/TiO<sup>2</sup> because of its strong adsorption capacity, while Cu-Fe/TiO<sup>2</sup> showed better redox ability. The redox ability and surface acidity are the key factors that affect the catalytic performance. As far as we know, the type of precursors is the very first factor that could have great impact on the physicochemical properties of catalysts. It plays a decisive role on the surface active component, the surface area, the redox ability, the surface acidity, etc. Liu [39,40] et al prepared FeTiO*<sup>x</sup>* via coprecipitation with Fe(NO3)<sup>3</sup> and Ti(SO4)<sup>2</sup> as precursors. The catalyst exhibited good NH3-SCR activities and the NO*<sup>x</sup>* conversion exceeded 90% in 200–350 ◦C. They found that there was strong interaction between iron and titanium. Ma [41] prepared Fe2(SO4)3/TiO<sup>2</sup> by the impregnation method and found the NO*<sup>x</sup>* conversion was up to 98% in 350–450 ◦C. Heterogeneous agglomeration of iron oxides could be weakened and the Brönsted acid sites could be improved by using Fe2(SO4)<sup>3</sup> as a precursor. In our previous work [42–44], we found that the magnesium-based catalyst showed good SCR activity and sulfur tolerance. Furthermore we studied on titanium modified FeMgO*<sup>x</sup>* catalysts and found that titanium modified FeMgO*<sup>x</sup>* catalysts exhibited excellent catalytic performance in SCR reaction. However, to the best of our knowledge, the catalytic performance could be further improvement via modifying the precursors.

In this work, a series of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors were studied. The objective of this paper is to investigate the effect of precursor type on the physicochemical properties of titanium modified FeMgO*<sup>x</sup>* catalysts and to reveal the optimization mechanism of catalytic performance.

### **2. Experimental**

#### *2.1. Catalyst Preparation*

FeSO4·7H2O, FeCl2·7H2O and Fe(NO3)3·9H2O were used as iron precursors, Mg(NO3)2·6H2O and MgSO4·7H2O (analytical pure, Tianjin Kermel Chemical Reagent Co., Ltd, Tianjin, China) were used as magnesium precursors and NH3·H2O was used as precipitant in catalyst preparation. Titanium modified FeMgO*<sup>x</sup>* catalysts were prepared via coprecipitation method with microwave thermal treatment. A certain amount of iron precursor, magnesium precursor and TiSO<sup>4</sup> (analytical pure, Sinopharm Group Co., Ltd, Shanghai, China) were dissolved in 250 mL deionized water and sufficiently stirred for 1 h. NH3·H2O was titrated into the mixed solution with continuous stirring until the pH of the mixed solution was 9–10. Then the precipitate was filtered and washed by deionized water several times until neutral to remove the foreign ions. The precipitate was first impregnated by 1 mol/L Na2CO<sup>3</sup> solution and then disposed of by microwave thermal treatment. The impregnated precipitate was washed by deionized water to be neutral and then dried at 105 ◦C. After calcined at 400 ◦C for 5 h, the obtained sample was crushed and sieved into 40–60 mesh (0.28 nm–0.45 nm) for the test. The catalysts prepared were denoted as Ti modified FeMgO*<sup>x</sup>* with label SN, SS, CN, CS, NN and NS to represent different combination of precursors (S represents for sulfates, N represents for nitrates and C represents for chlorides), as seen in Table 1.


**Table 1.** Titanium modified FeMgO*x* catalysts with different precursors.

#### *2.2. Activity Test*

NH3-SCR activity was completed in a quartz fixed-bed tube reactor at atmosphere pressure. The simulated flue gas was provided with standard gases, including 0.1 Vol % NO, 0.1 Vol % NH3, 3.5 Vol % O<sup>2</sup> and balanced N2. The total flow rate of simulated gas was 2 L/min and the catalyst used in each experiment was 4 mL, thus the corresponding gas hourly space velocity (GHSV) was 30,000/h −1 . The concentration of NO and NO<sup>2</sup> was monitored and analyzed by the MGA5 Flue Gas Analyzer (MRU Instruments, Inc. Emission Monitoring Systems, Neckarsulm-Obereisesheim, Germany). Before entering the flue gas analyzer, the flue gas should be washed by phosphoric acid (100% pure) to absorb ammonia and avoid the impact of ammonia on the analyzer. Data were recorded every 25 ◦C from 100 ◦C to 400 ◦C. NO*<sup>x</sup>* conversion was calculated as follows:

$$\eta = \frac{\mathbb{C}[\text{NO}\_{\text{x}}(\text{inlet})] - \mathbb{C}[\text{NO}\_{\text{x}}(\text{outlet})]}{\mathbb{C}[\text{NO}\_{\text{x}}(\text{inlet})]} \times 100\% \tag{1}$$

where C[NO*x*(inlet)] and C[NO*x*(outlet)] meant the concentration of NO*<sup>x</sup>* in the inlet and outlet of the reactor, µL/L. NO*<sup>x</sup>* represented the sum of NO and NO2.

#### *2.3. Catalyst Characterization*

A Rigaku D/max 2500 PC diffractometer (50 kV × 150 mA) with Cu Kα radiation was used to complete the X-ray Diffraction. The data of 2θ were collected from 10◦ to 90◦ by 4 ◦ /min with the step size 0.1◦ .

N2-adsorption-desorption was obtained by using an ASAP2020 Surface Area and Porosity Analyzer (Micromeritics Instrument Corp., Norcross, Georgia, USA) at −196 ◦C. The specific surface area and the average pore diameter were calculated by the Brunauer–Emmett–Teller (BET) method, and the specific pore volume and pore diameter distribution were calculated by Barrett–Joyner–Halenda (BJH) method.

Microstructure of the catalysts was conducted on a Japan JSM-6700F cold field emission scanning electron microscope. The elements on the surface of the catalysts were analyzed on an Oxford INCA X sight energy dispersive spectrometer (Be4-U92) with 5.9 KeV, UK.

To analyze the surface atomic concentration and distinguish the chemical states of the elements, a Thermo ESCALAB 250XI surface analyze system with Al Kα radiation (1486.6 eV, 150 W) was used to complete X-ray Photoelectron Spectroscopy. Prior to the measurement, each sample was degassed in vacuum to eliminate surface contamination.

Temperature-programmed Desorption of NH<sup>3</sup> (NH3-TPD) was performed on a TP-5080 instrument using a 100 mg sample. The sample was pretreated in flowing He at 300 ◦C for 1 h before the measurement. Then, the sample was He-cooled to 100 ◦C, then treated with 5% NH3/Ar at a flow rate of 30 mL/min for 0.5 h and flushed with He at 100 ◦C for 1 h. The desorption process was carried out by heating the sample from 100 ◦C to 700 ◦C at a rate of 10 ◦C/min.

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

#### *3.1. E*ff*ect of Di*ff*erent Precursors on Catalytic Performance Over Titanium Modified FeMgO<sup>x</sup> Catalysts*

The catalytic performance of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors is shown in Figure 1. It was obvious that temperature had a strong effect on titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors at the temperature range from 100 to 400 ◦C. Generally speaking, it was necessary to reach a certain temperature for the catalyst to exhibit good catalytic activity, while an excessively high reaction temperature would lead to a decrease in catalytic activity due to the secondary reaction. *Catalysts* **2019**, 9, x FOR PEER REVIEW 4 of 15 speaking, it was necessary to reach a certain temperature for the catalyst to exhibit good catalytic activity, while an excessively high reaction temperature would lead to a decrease in catalytic activity due to the secondary reaction.

Titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors revealed excellent catalytic activity from 100 to 400 ◦C and had wide temperature windows. Among all the catalysts, catalysts SN, SS, CN and CS showed similar catalytic performances, which was distinctly evident with catalysts NN and NS. When the temperature was blow 200 ◦C, catalyst NN and NS showed better catalytic activity than other catalysts, especially catalyst NN, whose NO*<sup>x</sup>* conversion could exceed 50% and 90% when the reaction temperature was close to 150 ◦C and 200 ◦C, respectively. However, when the temperature exceeded 350 ◦C, the NO*<sup>x</sup>* conversion of catalysts NN and NS apparently decreased to about 50–60% due to the oxidation and decomposition of NH3. The NO*<sup>x</sup>* conversion of catalysts SN, SS, CS and CN could be stable at 90% at high temperature range. Among all the catalysts, catalyst SN with precursors of FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> exhibited excellent catalytic activity and N<sup>2</sup> selectivity in a wide temperature range, with NO*<sup>x</sup>* conversion above 90% from 225 to 400 ◦C and N<sup>2</sup> selectivity above 90% in the whole temperature range. The NO*<sup>x</sup>* conversion of which approached 100% from 250 to 375 ◦C. Titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors revealed excellent catalytic activity from 100 to 400 °C and had wide temperature windows. Among all the catalysts, catalysts SN, SS, CN and CS showed similar catalytic performances, which was distinctly evident with catalysts NN and NS. When the temperature was blow 200 °C, catalyst NN and NS showed better catalytic activity than other catalysts, especially catalyst NN, whose NO*<sup>x</sup>* conversion could exceed 50% and 90% when the reaction temperature was close to 150 °C and 200 °C, respectively. However, when the temperature exceeded 350 °C, the NO*<sup>x</sup>* conversion of catalysts NN and NS apparently decreased to about 50–60% due to the oxidation and decomposition of NH3. The NO*<sup>x</sup>* conversion of catalysts SN, SS, CS and CN could be stable at 90% at high temperature range. Among all the catalysts, catalyst SN with precursors of FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> exhibited excellent catalytic activity and N<sup>2</sup> selectivity in a wide temperature range, with NO*<sup>x</sup>* conversion above 90% from 225 to 400 °C and N<sup>2</sup> selectivity above 90% in the whole temperature range. The NO*<sup>x</sup>* conversion of which approached 100% from 250 to 375 °C.

**Figure 1.** Catalytic performance of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors. **Figure 1.** Catalytic performance of titanium modified FeMgO*x* catalysts with different precursors.

#### *3.2. X-ray Di*ff*raction (XRD) Patterns*

*3.2. X-ray Diffraction (XRD) Patterns* Figure 2 shows the XRD patterns of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors. There were no diffraction peaks of magnesium or titanium in XRD patterns of all the catalysts according to JCPDF standard; it could be inferred that magnesium and titanium existed in a highly dispersed state or an amorphous state in the catalyst, or maybe that the crystallites formed were less than 5 nm. There were obvious sharp diffraction peaks of the catalysts SN, SS, CN and CS at 2θ = 30.2°, 35.5°, 43.2°, 53.7°, 53.7° and 62.8°, corresponding to maghemite (γ-Fe2O3) crystallite according to JCPDS PDF#39-1346 [39,40]. It could be inferred that maghemite crystallite was the Figure 2 shows the XRD patterns of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors. There were no diffraction peaks of magnesium or titanium in XRD patterns of all the catalysts according to JCPDF standard; it could be inferred that magnesium and titanium existed in a highly dispersed state or an amorphous state in the catalyst, or maybe that the crystallites formed were less than 5 nm. There were obvious sharp diffraction peaks of the catalysts SN, SS, CN and CS at 2θ = 30.2◦ , 35.5◦ , 43.2◦ , 53.7◦ , 53.7◦ and 62.8◦ , corresponding to maghemite (γ-Fe2O3) crystallite according to JCPDS PDF#39-1346 [39,40]. It could be inferred that maghemite crystallite was the main active component in these catalysts.

main active component in these catalysts. However, in catalyst NN and NS, the diffraction peaks were ascribe to amorphous oxides, comparing with that in Figure S1. It could be concluded that the active component was directly affected by the precursors. When Fe(NO3)<sup>3</sup> was used as precursor, the active components of the titanium modified FeMgO*<sup>x</sup>* catalysts prepared were iron-magnesium-titanium mixed oxides; however, when FeSO<sup>4</sup> and FeCl<sup>2</sup> were used as precursors, the active component of the titanium modified FeMgO*<sup>x</sup>* catalysts prepared was γ-Fe2O3. It was reported that γ-Fe2O<sup>3</sup> was an octahedral structure with vacancies and was in a metastable state with a lower activation energy, which resulted in better denitration activity. In general, different active components led to diversity between different catalysts. activity due to the secondary reaction.

approached 100% from 250 to 375 °C.

*3.2. X-ray Diffraction (XRD) Patterns*

Considering the result of the activity test, iron-magnesium-titanium mixed oxides and γ-Fe2O<sup>3</sup> as active components were the substantial cause of different catalytic performance. The mixed oxides made the temperature window of the catalysts NN and NS obviously shift to a low temperature range and was the main cause of the secondary reaction at high temperature range. highly dispersed state or an amorphous state in the catalyst, or maybe that the crystallites formed were less than 5 nm. There were obvious sharp diffraction peaks of the catalysts SN, SS, CN and CS at 2θ = 30.2°, 35.5°, 43.2°, 53.7°, 53.7° and 62.8°, corresponding to maghemite (γ-Fe2O3) crystallite according to JCPDS PDF#39-1346 [39,40]. It could be inferred that maghemite crystallite was the main active component in these catalysts.

catalysts according to JCPDF standard; it could be inferred that magnesium and titanium existed in a

**Figure 1.** Catalytic performance of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors.

Figure 2 shows the XRD patterns of titanium modified FeMgO*<sup>x</sup>* catalysts with different

*Catalysts* **2019**, 9, x FOR PEER REVIEW 4 of 15

speaking, it was necessary to reach a certain temperature for the catalyst to exhibit good catalytic activity, while an excessively high reaction temperature would lead to a decrease in catalytic

Titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors revealed excellent catalytic activity from 100 to 400 °C and had wide temperature windows. Among all the catalysts, catalysts SN, SS, CN and CS showed similar catalytic performances, which was distinctly evident with catalysts NN and NS. When the temperature was blow 200 °C, catalyst NN and NS showed better catalytic activity than other catalysts, especially catalyst NN, whose NO*<sup>x</sup>* conversion could exceed 50% and 90% when the reaction temperature was close to 150 °C and 200 °C, respectively. However, when the temperature exceeded 350 °C, the NO*<sup>x</sup>* conversion of catalysts NN and NS apparently decreased to about 50–60% due to the oxidation and decomposition of NH3. The NO*<sup>x</sup>* conversion of catalysts SN, SS, CS and CN could be stable at 90% at high temperature range. Among all the catalysts, catalyst SN with precursors of FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> exhibited excellent catalytic activity and N<sup>2</sup> selectivity in a wide temperature range, with NO*<sup>x</sup>* conversion above 90% from 225 to 400 °C and N<sup>2</sup> selectivity above 90% in the whole temperature range. The NO*<sup>x</sup>* conversion of which

**Figure 2.** Powder X-ray diffraction (XRD) patterns of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors.
