*3.3. N2-Adsorption-Desorption*

BET surface area, average pore volume and average pore diameter of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors are enumerated in Table 2.


**Table 2.** Surface characterization of titanium modified FeMgO*x* catalysts with different precursors.

The BET surface area, BJH pore volume and average pore diameter of the catalysts SN, SS, CN and CS were similar to each other. The BET surface area of catalyst SN, which exhibited the best catalytic activity, was 55.1245 m<sup>2</sup> /g, the pore volume was 0.2291 cm<sup>3</sup> /g and the average pore diameter was 16.2670 nm. Nevertheless, the BET surface area of the catalysts NN and NS using Fe(NO3)<sup>3</sup> as a precursor were 181.3934 m<sup>2</sup> /g and 180.4130 m<sup>2</sup> /g, respectively, which were three times larger than the other four catalysts. The average pore diameter of the catalysts NN and NS were only 4.3710 nm and 4.4512 nm, which were almost four times smaller than the other four catalysts, but the pore volumes were close to each other. Generally speaking, more active sites could be provided by large surface area, pore volume and relatively small average pore diameter, which was beneficial for SCR reaction. For the catalysts NN and NS, large surface area could provide more active sites, but the small pore diameter would lead to the increment of diffusion resistance during the gas-solid reaction and would be bad for the adsorption-desorption process. Appropriate surface area, pore volume and average pore diameter in the catalyst SN could provide enough active sites and guarantee the diffusion and mass transfer processes, which were in favor of SCR reaction.

t-Plot microporous area and volume are enumerated in Table 3. It was obvious that the t-Plot microporous area was much smaller than t-Plot external surface area, and the microporous volume of all the catalysts was close to zero. It could be concluded that the t-Plot microporous area had less contribution to the surface area, and the t-Plot external surface area was much more important to produce large surface area. Meanwhile, mesopore (pore diameter ranging from 2 to 50 nm) was the main pore type in titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors, and there were substantially fewer micropores (pore diameter less than 2 nm).


**Table 3.** t-Plot properties of titanium modified FeMgO*x* catalysts with different precursors.

Distribution characterization of pore structures over titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors is shown in Figure 3. In Figure 3a,b, pore diameter of all the catalysts is distributed mainly from 2 nm to 10 nm, which also means that mesopores were the major pore type in the catalysts. In addition, the pore diameter distribution of the catalysts NN and NS is obviously distinguished from the other catalysts, and the intensities of the distribution peaks are a great deal stronger than other catalysts, indicating that the pores ranging from 2 to 10 nm made a greater contribution to surface area and pore volume. NS 3.1251 177.2879 2.54 × 10−<sup>3</sup> t-Plot microporous area and volume are enumerated in Table 3. It was obvious that the t-Plot microporous area was much smaller than t-Plot external surface area, and the microporous volume of all the catalysts was close to zero. It could be concluded that the t-Plot microporous area had less contribution to the surface area, and the t-Plot external surface area was much more important to produce large surface area. Meanwhile, mesopore (pore diameter ranging from 2 to 50 nm) was the main pore type in titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors, and there were substantially fewer micropores (pore diameter less than 2 nm).

**Figure 3.** Distribution characterization of pore structures over titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors: (**a**) Pore volume, (**b**) Pore area, (**c**) Cumulate pore volume and (**d**) Cumulate pore area. **Figure 3.** Distribution characterization of pore structures over titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors: (**a**) Pore volume, (**b**) Pore area, (**c**) Cumulate pore volume and (**d**) Cumulate pore area.

obviously distinguished from the other catalysts, and the intensities of the distribution peaks are a great deal stronger than other catalysts, indicating that the pores ranging from 2 to 10 nm made a

In Figure 3c and Figure 3d, it can be seen that the cumulative pore area and the cumulative pore volume of the catalysts SN, SS, CN and CS gently declined with pore diameter. However, the cumulative pore area and the cumulative pore volume of the catalysts NN and NS from 2 to 10 nm declined rapidly with pore diameter, and the cumulative pore area and the cumulative pore volume above 10 nm were pretty small compared with that from 2 to 10 nm. The intensive distribution of pore diameter from 2 to 10 nm could provide a large surface area, but the narrow distribution of pore diameter would lead to the hysteresis of the diffusion and mass transfer processes. For the catalyst SN, the reasonable distribution of pore diameter could guarantee enough surface area and

greater contribution to surface area and pore volume.

also the diffusion and mass transfer processes.

Distribution characterization of pore structures over titanium modified FeMgO*<sup>x</sup>* catalysts with

In Figure 3c,d, it can be seen that the cumulative pore area and the cumulative pore volume of the catalysts SN, SS, CN and CS gently declined with pore diameter. However, the cumulative pore area and the cumulative pore volume of the catalysts NN and NS from 2 to 10 nm declined rapidly with pore diameter, and the cumulative pore area and the cumulative pore volume above 10 nm were pretty small compared with that from 2 to 10 nm. The intensive distribution of pore diameter from 2 to 10 nm could provide a large surface area, but the narrow distribution of pore diameter would lead to the hysteresis of the diffusion and mass transfer processes. For the catalyst SN, the reasonable distribution of pore diameter could guarantee enough surface area and also the diffusion and mass transfer processes.

The N2-adsorption-desorption isotherms of titanium modified FeMgO*<sup>x</sup>* catalysts are shown in Figure 4. According to the International Union of Pure and Applied Chemistry classification, the N2-adsorption-desorption isotherms of the catalysts SN, SS, CN and CS were classified as V-shaped isotherms with an H3 hysteresis loop. The absorbed volume was really small when the pressure was low. Only when the pressure was approaching the saturated vapor pressure did the absorbed volume increase rapidly due to capillary condensation. The absorption characteristic was often observed by the weak solid-gas interaction in mesopores on the surface of catalysts. It could be concluded from the type of isotherm and hysteresis loop that disorderly wedge-shaped mesopores were formed by particles accumulated loosely on the surface of the catalysts [44]. *Catalysts* **2019**, 9, x FOR PEER REVIEW 7 of 15 The N2-adsorption-desorption isotherms of titanium modified FeMgO*<sup>x</sup>* catalysts are shown in Figure 4. According to the International Union of Pure and Applied Chemistry classification, the N2-adsorption-desorption isotherms of the catalysts SN, SS, CN and CS were classified as V-shaped isotherms with an H3 hysteresis loop. The absorbed volume was really small when the pressure was low. Only when the pressure was approaching the saturated vapor pressure did the absorbed volume increase rapidly due to capillary condensation. The absorption characteristic was often observed by the weak solid-gas interaction in mesopores on the surface of catalysts. It could be concluded from the type of isotherm and hysteresis loop that disorderly wedge-shaped mesopores

were formed by particles accumulated loosely on the surface of the catalysts [44].

**Figure 4.** N2-adsorption-desorption isotherms of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors: (**a**) SN, (**b**) SS, (**c**) CN, (**d**) CS, (**e**) NN and (**f**) NS. **Figure 4.** N<sup>2</sup> -adsorption-desorption isotherms of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors: (**a**) SN, (**b**) SS, (**c**) CN, (**d**) CS, (**e**) NN and (**f**) NS.

However, the N2-adsorption-desorption isotherms of the catalysts NN and NS were classified as IV-shaped isotherms with an H4 hysteresis loop. The absorbed volume gradually increased when

the type of isotherm and hysteresis loop that wedge-shaped mesopores were formed by lamellar

structures accumulated tightly on the surface of the catalysts.

precursors.

However, the N2-adsorption-desorption isotherms of the catalysts NN and NS were classified as IV-shaped isotherms with an H4 hysteresis loop. The absorbed volume gradually increased when the pressure was low. Only when the relative pressure was about 0.5 to 0.8 did the absorbed volume increase rapidly, and then the absorbing capacity became almost invariable, which meant there were mainly mesopores in the catalysts and less macropores were obtained. It could be concluded from the type of isotherm and hysteresis loop that wedge-shaped mesopores were formed by lamellar structures accumulated tightly on the surface of the catalysts.

*Catalysts* **2019**, 9, x FOR PEER REVIEW 8 of 15

### *3.4. SEM and Energy Dispersive Spectrometer (EDS)*

The SEM images of titanium modified FeMgO*<sup>x</sup>* catalysts are shown in Figure 5. The surface of the catalysts SN, SS, CN and CS presented spherical particle distribution and the particles were significantly more independent and regular. There was hardly any accumulation of particles occurring on the surface of these catalysts, and the regular distribution of particles benefitted the formation of intergranular pores, which were good for the mass transfer process. *3.4. SEM and Energy Dispersive Spectrometer (EDS)* The SEM images of titanium modified FeMgO*<sup>x</sup>* catalysts are shown in Figure 5. The surface of the catalysts SN, SS, CN and CS presented spherical particle distribution and the particles were significantly more independent and regular. There was hardly any accumulation of particles occurring on the surface of these catalysts, and the regular distribution of particles benefitted the

formation of intergranular pores, which were good for the mass transfer process.

**Figure 5.** SEM images of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors: (**a**) SN, (**b**) SS, (**c**) CN, (**d**) CS, (**e**) NN and (**f**) NS. **Figure 5.** SEM images of titanium modified FeMgO*x* catalysts with different precursors: (**a**) SN, (**b**) SS, (**c**) CN, (**d**) CS, (**e**) NN and (**f**) NS.

However, the surface of the catalysts NN and NS exhibited stratiform and nubby distribution. There were a huge number of pyknotic fine intergranular pores on the stratiform and nubby structure. Fine intergranular pores were favorable for large surface area and provided more active sites, but they would resist the diffusion and mass transfer process. Meanwhile the results implied that the choice of an iron precursor had significant influence on the surface morphology of the catalyst. The catalyst SN exhibited regular and distributed spherical particles with good pores connectivity which were favor of SCR reaction. However, the surface of the catalysts NN and NS exhibited stratiform and nubby distribution. There were a huge number of pyknotic fine intergranular pores on the stratiform and nubby structure. Fine intergranular pores were favorable for large surface area and provided more active sites, but they would resist the diffusion and mass transfer process. Meanwhile the results implied that the choice of an iron precursor had significant influence on the surface morphology of the catalyst. The catalyst SN exhibited regular and distributed spherical particles with good pores connectivity which were favor of SCR reaction.

The EDS composition analysis of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors is shown in Table 4. The EDS composition analysis of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors is shown in Table 4.

**Table 4.** EDS composition analysis data of titanium modified FeMgO*<sup>x</sup>* catalysts with different

**Fe Mg Ti O Fe Mg Ti O**

**Catalyst Percentage by Weight/ wt % Percentage by Atomicity/ at %**

SN 59.06 1.68 4.79 34.47 30.82 2.85 3.21 63.12 SS 59.25 3.10 4.73 32.92 31.72 3.81 2.95 61.52 CN 60.66 3.64 4.60 31.10 33.16 4.57 2.93 59.34 CS 59.33 3.18 4.40 33.09 31.68 3.90 2.74 61.68 NN 68.39 2.90 5.30 23.41 41.97 4.09 3.79 50.16 NS 67.93 3.47 4.88 23.72 41.53 4.45 3.17 50.85

was obviously lower than that of the catalysts SN, SS, CN and CS, which means that it was difficult for oxygen atoms to enrich on the surface of the catalysts NN and NS. The percentage of oxygen by atomicity of the catalyst SN could reach 63.12%. In general, the surface oxidation ability could be improved by the abundant lattice oxygen on the surface of the catalyst. Strong oxidation ability could promote NO oxidation to NO<sup>2</sup> and then induce the rapid SCR reaction, which effectively


**Table 4.** EDS composition analysis data of titanium modified FeMgO*x* catalysts with different precursors.

The catalysts prepared all consisted of iron, magnesium, titanium and oxygen, as expected. From Table 4, it can be seen that the percentage of oxygen by atomicity of the catalysts NN and NS was obviously lower than that of the catalysts SN, SS, CN and CS, which means that it was difficult for oxygen atoms to enrich on the surface of the catalysts NN and NS. The percentage of oxygen by atomicity of the catalyst SN could reach 63.12%. In general, the surface oxidation ability could be improved by the abundant lattice oxygen on the surface of the catalyst. Strong oxidation ability could promote NO oxidation to NO<sup>2</sup> and then induce the rapid SCR reaction, which effectively guarantees the catalytic performance of the catalyst. The results illustrated that using Fe(NO3)<sup>3</sup> as a precursor was not conductive to the enrichment of oxygen on the surface of the catalyst.

#### *3.5. X-ray Photoelectron Spectroscopy (XPS)*

An XPS test was used to better elucidate the spices, concentration and valency of different elements on the surface of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors, and the element surface concentration calculated is shown as Table 5. It can be seen that the main elements of titanium modified FeMgO*<sup>x</sup>* catalysts are iron, magnesium, titanium and oxygen, which is in agreement with EDS results.


**Table 5.** X-ray photoelectron spectroscopy (XPS) elementary surface concentration of titanium modified FeMgO*x* catalysts with different precursors.

The concentration of O 1s over the catalyst SN could reach 46.85%, which illustrates that using FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> as precursors was in favor of the enrichment of oxygen on the surface of the catalyst and enhanced the surface oxidation ability.

Mg 1s and Ti 2p spectra of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors are shown in Figure 6. The Mg 1s peaks at a binding energy of about 1303 eV in, and coincides exactly with Mg2+. For the catalysts SN, SS, CN and CS, the Ti 2p3/<sup>2</sup> peaks at 458.23 eV and the Ti 2p1/<sup>2</sup> peaks at 464.03 eV were attributed to Ti4+. The binding energy of Ti 2p shifted higher compared with those of the catalysts NN and NS. The Ti 2p3/<sup>2</sup> peaks appeared at 458.23 eV and the Ti 2p1/<sup>2</sup> peaks appeared at 464.03 eV.

at 464.03 eV.

ascribed to O2<sup>−</sup>

of the catalyst, which was in favor of the SCR reaction.

with Mg2+

guarantees the catalytic performance of the catalyst. The results illustrated that using Fe(NO3)<sup>3</sup> as a

An XPS test was used to better elucidate the spices, concentration and valency of different elements on the surface of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors, and the element surface concentration calculated is shown as Table 5. It can be seen that the main elements of titanium modified FeMgO*<sup>x</sup>* catalysts are iron, magnesium, titanium and oxygen, which is in

**Table 5.** X-ray photoelectron spectroscopy (XPS) elementary surface concentration of titanium

**Catalyst Fe 2p/% Mg 1s/% O 1s/% Ti 2p/%** SN 38.88 6.92 46.85 7.35 SS 46.35 7.89 37.20 8.57 CN 45.75 9.54 37.75 6.95 CS 47.49 7.47 37.48 7.55 NN 56.74 6.43 32.92 3.91 NS 56.65 6.20 32.86 4.29 The concentration of O 1s over the catalyst SN could reach 46.85%, which illustrates that using FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> as precursors was in favor of the enrichment of oxygen on the surface of the

Mg 1s and Ti 2p spectra of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors are shown in Figure 6. The Mg 1s peaks at a binding energy of about 1303 eV in, and coincides exactly

of the catalysts NN and NS. The Ti 2p3/2 peaks appeared at 458.23 eV and the Ti 2p1/2 peaks appeared

. For the catalysts SN, SS, CN and CS, the Ti 2p3/2 peaks at 458.23 eV and the Ti 2p1/2 peaks

. The binding energy of Ti 2p shifted higher compared with those

precursor was not conductive to the enrichment of oxygen on the surface of the catalyst.

*3.5. X-ray Photoelectron Spectroscopy (XPS)*

modified FeMgO*<sup>x</sup>* catalysts with different precursors.

catalyst and enhanced the surface oxidation ability.

agreement with EDS results.

**Figure 6.** Mg 1s and Ti 2p spectra over titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors. **Figure 6.** Mg 1s and Ti 2p spectra over titanium modified FeMgO*x* catalysts with different precursors.

The Fe 2p spectra of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors are shown in Figure 7. The Fe 2p spectra consisted of three overlapping peaks, and the binding energy of the Fe species were further analyzed by peak-fitting. The Fe 2p3/2 peaks at binding energy around 710.0 eV and 712.0 eV, the Fe2p3/2,sat peaks at binding energy about 718.8 eV and the Fe 2p1/2 peaks at binding energy of 724.2 eV were all ascribed to Fe3+ [21]. The Fe 2p spectra of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors are shown in Figure 7. The Fe 2p spectra consisted of three overlapping peaks, and the binding energy of the Fe species were further analyzed by peak-fitting. The Fe 2p3/<sup>2</sup> peaks at binding energy around 710.0 eV and 712.0 eV, the Fe2p3/2,sat peaks at binding energy about 718.8 eV and the Fe 2p1/<sup>2</sup> peaks at binding energy of 724.2 eV were all ascribed to Fe3<sup>+</sup> *Catalysts* **2019** [21]. , 9, x FOR PEER REVIEW 10 of 15

**Figure 7.** Fe 2p spectra of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors. **Figure 7.** Fe 2p spectra of titanium modified FeMgO*x* catalysts with different precursors.

The O 1s spectra of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors are shown on Figure 8. The O 1s peaks at the binding energy of about 529.8 eV and 531.5 eV were the lattice The O 1s spectra of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors are shown on Figure 8. The O 1s peaks at the binding energy of about 529.8 eV and 531.5 eV were the lattice oxygen

oxygen (denoted as Oβ) and the chemisorbed oxygen (denotes as Oα), respectively, which were both

. The intensity of the O 1s peaks of the catalysts NN and NS using Fe(NO3)<sup>3</sup> as

catalysts NN and NS, was significantly poorer, and the concentration of chemisorbed oxygen of the catalyst SN could be up to 21.2% on the surface, which was in accordance with the results above. Meanwhile, the shifting of the O 1s binding energy means that the oxygen deficit was on the surface (denoted as Oβ) and the chemisorbed oxygen (denotes as Oα), respectively, which were both ascribed to O2−. The intensity of the O 1s peaks of the catalysts NN and NS using Fe(NO3)<sup>3</sup> as precursor was apparently lower than those of the other catalysts, demonstrating that the concentration of oxygen on the surface of the catalysts NN and NS was low. Analyzed by peak-fitting, it could be seen that the chemisorbed oxygen, which was reported most active for the catalysts NN and NS, was significantly poorer, and the concentration of chemisorbed oxygen of the catalyst SN could be up to 21.2% on the surface, which was in accordance with the results above. Meanwhile, the shifting of the O 1s binding energy means that the oxygen deficit was on the surface of the catalyst, which was in favor of the SCR reaction. *Catalysts* **2019**, 9, x FOR PEER REVIEW 11 of 15

**Figure 8.** O 1s spectra of titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors. **Figure 8.** O 1s spectra of titanium modified FeMgO*x*catalysts with different precursors.

#### *3.6. NH3-TPD 3.6. NH3-TPD*

In the SCR reaction, NH<sup>3</sup> should have firstly adsorbed and activated on the active sites and then reacted with NO, such as the Eley–Rideal mechanism, or reacted with both NO and adsorbed NO, such as the Langmuir–Hinshelwood mechanism. The presence of acid sites was of great importance for catalytic performance [45]. The surface acidity and acid species of titanium modified FeMgO*<sup>x</sup>* catalysts were investigated by NH3-TPD and the results are shown in Figure 9. Usually the NH<sup>3</sup> desorption peaks below 300 °C were attributed to weak acid sites, and the NH<sup>3</sup> desorption peaks above 300 °C were ascribed to strong acid sites. The NH3-TPD profiles of all the catalysts exhibited a desorption peaks close to 170 °C, and the intensity of desorption peaks over the catalysts NN and NS In the SCR reaction, NH<sup>3</sup> should have firstly adsorbed and activated on the active sites and then reacted with NO, such as the Eley–Rideal mechanism, or reacted with both NO and adsorbed NO, such as the Langmuir–Hinshelwood mechanism. The presence of acid sites was of great importance for catalytic performance [45]. The surface acidity and acid species of titanium modified FeMgO*<sup>x</sup>* catalysts were investigated by NH3-TPD and the results are shown in Figure 9. Usually the NH<sup>3</sup> desorption peaks below 300 ◦C were attributed to weak acid sites, and the NH<sup>3</sup> desorption peaks above 300 ◦C were ascribed to strong acid sites. The NH3-TPD profiles of all the catalysts exhibited a desorption peaks close to 170 ◦C, and the intensity of desorption peaks over the catalysts NN and

deduced that the weak acid sites that appeared at 170 °C had impact on the catalytic performance below 200 °C, and the strong acid sites that appeared at 380 °C made a great contribution to the catalytic activity and inhibited the secondary reaction at a higher temperature range. However, it was difficult for NH<sup>3</sup> to absorb and activate on the strong acid site of the catalysts NN and NS, and appeared to exceed 450 °C. The total acidity of the catalyst SN reached 1362.03 μmol/g, and the improvement of the total acidity of the catalyst SN could provide more adsorption sites. The enhancement of strong acid sites around 380 °C could inhibit the secondary reaction above 300 °C,

was apparently stronger than that of the catalysts SN, SS, CN and CS. The desorption peaks ascribe to strong acid sites of the catalysts SN, SS, CN and CS appeared at about 380 °C, while those of the temperature range.

NS was apparently stronger than that of the catalysts SN, SS, CN and CS. The desorption peaks ascribe to strong acid sites of the catalysts SN, SS, CN and CS appeared at about 380 ◦C, while those of the catalysts NN and NS appeared above 450 ◦C. The amount of total acidity of all the catalysts was calculated as 1362.03 µmol/g, 883.81µmol/g, 933.96 µmol/g, 525.86 µmol/g, 1915.58 µmol/g and 1977.91 µmol/g, with the order from catalyst SN to NS. Considering the activity results, it could be deduced that the weak acid sites that appeared at 170 ◦C had impact on the catalytic performance below 200 ◦C, and the strong acid sites that appeared at 380 ◦C made a great contribution to the catalytic activity and inhibited the secondary reaction at a higher temperature range. However, it was difficult for NH<sup>3</sup> to absorb and activate on the strong acid site of the catalysts NN and NS, and appeared to exceed 450 ◦C. The total acidity of the catalyst SN reached 1362.03 µmol/g, and the improvement of the total acidity of the catalyst SN could provide more adsorption sites. The enhancement of strong acid sites around 380 ◦C could inhibit the secondary reaction above 300 ◦C, which was the reason that the catalyst SN exhibited excellent catalytic performance and had a wide temperature range. *Catalysts* **2019**, 9, x FOR PEER REVIEW 12 of 15 which was the reason that the catalyst SN exhibited excellent catalytic performance and had a wide

**Figure 9.** NH3-TPD profiles over titanium modified FeMgO*<sup>x</sup>* catalysts with different precursors. **Figure 9.** NH<sup>3</sup> -TPD profiles over titanium modified FeMgO*x* catalysts with different precursors.

#### **4. Conclusion 4. Conclusions**

correspondingly.

XRD patterns of Fe2O<sup>3</sup>

acquisition, D.W., Y.P. and C.L.

**Conflicts of Interest:** The authors declare no conflict of interest.

21777081 and 51576117.

**Reference**

The influence of different precursors on titanium modified FeMgO*<sup>x</sup>* catalysts was investigated and characterized. The results show that the crystalline phase of the active component was directly affected by the iron precursors. γ-Fe2O<sup>3</sup> formed as the main crystalline phase when FeSO<sup>4</sup> and FeCl<sup>2</sup> were used as precursors. The main crystalline phase would be amorphous iron-magnesium-titanium mixed oxides when Fe(NO3)<sup>3</sup> was used as precursor. The catalyst using FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> as precursors exhibited NO*<sup>x</sup>* conversion above 90% from 225 to 400 °C, while approaching 100% from 250 to 375 °C. The temperature window of the catalysts using Fe(NO3)<sup>3</sup> as a precursor shifted to lower temperature range, on which the secondary reaction occurred, leading to the decline of NO*<sup>x</sup>* conversion at a high temperature. The regular spherical particle distribution and the good pores connectivity were advantageous to the mass transfer process. The acid sites that appeared at 170 °C played an important role for catalytic performance below 200 °C. The acid sites that appeared at 380 °C inhibited the secondary reaction at high temperature range. However, it was difficult for NH<sup>3</sup> to absorb and activate on the strong acid sites using Fe(NO3)<sup>3</sup> as a precursor when exceeding 450 °C. The total acidity of the catalyst using FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> as precursors could reach 1362.03 μmol/g, and the surface oxygen concentration was also enhanced, thereby SCR reaction was improved The influence of different precursors on titanium modified FeMgO*<sup>x</sup>* catalysts was investigated and characterized. The results show that the crystalline phase of the active component was directly affected by the iron precursors. γ-Fe2O<sup>3</sup> formed as the main crystalline phase when FeSO<sup>4</sup> and FeCl<sup>2</sup> were used as precursors. The main crystalline phase would be amorphous iron-magnesium-titanium mixed oxides when Fe(NO3)<sup>3</sup> was used as precursor. The catalyst using FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> as precursors exhibited NO*<sup>x</sup>* conversion above 90% from 225 to 400 ◦C, while approaching 100% from 250 to 375 ◦C. The temperature window of the catalysts using Fe(NO3)<sup>3</sup> as a precursor shifted to lower temperature range, on which the secondary reaction occurred, leading to the decline of NO*<sup>x</sup>* conversion at a high temperature. The regular spherical particle distribution and the good pores connectivity were advantageous to the mass transfer process. The acid sites that appeared at 170 ◦C played an important role for catalytic performance below 200 ◦C. The acid sites that appeared at 380 ◦C inhibited the secondary reaction at high temperature range. However, it was difficult for NH<sup>3</sup> to absorb and activate on the strong acid sites using Fe(NO3)<sup>3</sup> as a precursor when exceeding 450 ◦C. The total acidity of the catalyst using FeSO<sup>4</sup> and Mg(NO3)<sup>2</sup> as precursors could reach 1362.03 µmol/g, and the surface oxygen concentration was also enhanced, thereby SCR reaction was improved correspondingly.

visualization, L.X., L.H. and Z.S.; supervision, J.L. and C.L.; project administration, D.W., Y.P. and C.L.; funding

**Funding:** This research was funded by the National Key Research and Development Program, grant number 2017YFC0212800 and 2017YFC0210200, the National Natural Science Foundation of China, grant number

**Supplementary Materials:** The following are available online at www.mdpi.com/xxx/s1, Figure S1: Powder

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/9/6/560/s1, Figure S1: Powder XRD patterns of Fe2O<sup>3</sup> .

**Author Contributions:** Conceptualization, D.W. and L.X.; methodology, D.W., Y.P. and L.X.; validation, J.L., Y.P. and D.W.; formal analysis, L.H. and Z.S.; investigation, L.X., Q.Y. and L.H.; resources, D.W. and C.L.; data curation, L.X. and Z.S.; writing—original draft preparation, L.X.; writing—review and editing, D.W.; visualization, L.X., L.H. and Z.S.; supervision, J.L. and C.L.; project administration, D.W., Y.P. and C.L.; funding acquisition, D.W., Y.P. and C.L.

**Funding:** This research was funded by the National Key Research and Development Program, grant number 2017YFC0212800 and 2017YFC0210200, the National Natural Science Foundation of China, grant number 21777081 and 51576117.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
