*2.2. The Preparation of Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> Catalysts*

The Al2O<sup>3</sup> and SiO<sup>2</sup> supports with a near Brunauer-Emmett-Teller (BET) surface area (157–160 m2/g) were chosen in this study. The catalysts were prepared using a sequential impregnation method. Typically, for the Cu/Al2O<sup>3</sup> and K-Cu/Al2O<sup>3</sup> catalysts, 6.74 g of Cu(NO3)<sup>2</sup> × 3H2O (10 wt % CuO) was dissolved in 20 mL of deionized water. Twenty grams of Al2O<sup>3</sup> were added into the above copper nitrate solution and impregnated by the ultrasonic treatment for 1 h at room temperature. Afterward, the mixture was dried at 120 ◦C for 10 h and calcined at 900 ◦C for 5 h in air. The obtained solid was Cu/Al2O<sup>3</sup> catalyst. K-Cu/Al2O<sup>3</sup> catalyst was prepared through the second impregnation of Cu/Al2O<sup>3</sup> in K2CO<sup>3</sup> aqueous solution. Simply, 0.61 g of K2CO<sup>3</sup> (4 wt % K2O loading) was dissolved in another 10 mL of deionized water. The desired amount of the Cu/Al2O<sup>3</sup> catalyst obtained above was impregnated in K2CO<sup>3</sup> aqueous solution, along with the ultrasonic treatment, for 1 h at room temperature. Then, the resulting mixture was dried at 120 ◦C for 10 h and calcined at 500 ◦C for 5 h in air. The Cu/SiO<sup>2</sup> and K-Cu/SiO<sup>2</sup> catalysts were also prepared using a method similar to one above. For comparison, the Al2O<sup>3</sup> and SiO<sup>2</sup> supports were also calcined at 900 ◦C for 5 h in air and denoted as Al2O3-900 and SiO2-900, respectively.

#### *2.3. Catalyst Characterization*

The textural properties of the as-prepared catalysts were measured with N<sup>2</sup> absorption-desorption at −196 ◦C on a Tristar 3000 Micromeritics (Atlanta, GA, US) instrument. The specific surface area (SBET) was calculated by the BET method. The micropore volume was obtained from the t-plot method. The pore size distributions were evaluated by using the density functional theory (DFT) method applied to the nitrogen adsorption data. The measurement of textural properties is accuracy (±1%). The experiments were repeated three times.

Powder XRD patterns of the catalysts were collected on a Rigaku MiniFlex II X-ray diffractometer (Tokyo, Japan), using Ni-filtered Cu-Kα radiation (k = 0.15418 nm) with a scanning angle (2θ) of 10–90◦ .

H2-TPR was carried out on an automatic temperature-programmed chemisorption analyzer (TP-5080, Tianjin Xianquan Industrial Trade and Develpment Co. Ltd, Tianjin, China) equipped with a thermal conductivity detector. The catalyst with 100 mg was pretreated at 300 ◦C under a flow of N<sup>2</sup> (32 mL/min) for 1 h to remove traces of water and then cooled to 50 ◦C. Subsequently, the gas flow was switched to a 10% H2/N<sup>2</sup> (*v*/*v*, 35 mL/min). The sample was heated to 900 ◦C at a rate of 10 ◦C/min.

NH3-TPD was carried out on a TP-5080 chemisorption instrument in order to evaluate the acidity of the catalysts. The catalyst (100 mg) was pretreated at 400 ◦C under a flow of N<sup>2</sup> (32 mL/min) for 1 h

and then cooled down to 100 ◦C. After that, sample was exposed on NH<sup>3</sup> flow for 15 min. The TPD spectra were recorded from 100 to 600 ◦C, using a heating rate of 10 ◦C/min.

Characterizations of XPS and Auger electron spectroscopy (XAES) were conducted on an AXIS ULTRA DLD instrument (Kratos, Manchester, UK) equipped with Al Kα (hν = 1486.6 eV). The binding energy values were corrected for charging effects by referring to the adventitious C1s line at 284.5 eV.

In situ FTIR spectra of CO adsorption and desorption were obtained with a TENSOR-27 in the range from 4000 to 1000 cm−<sup>1</sup> with 4 cm−<sup>1</sup> resolution. Before CO adsorption, all catalysts were reduced at 400 ◦C for 0.5 h in a 10% H2/N<sup>2</sup> (*v*/*v*, 15 mL/min). CO adsorption was taken at 400 ◦C and after 30 min of pure Ar flow at the same temperature. IR spectra were collected after evacuation for 30 min.

#### *2.4. Catalytic Performance Evaluation*

The catalyst test of CO hydrogenation was performed in a stainless fixed-bed reactor. In a typical run, 5 mL of the prepared catalyst (30–40 meshes) was placed in the center of the reactor. The catalyst was reduced according to the designed temperature program, i.e., from room temperature to 400 ◦C in a 10% H2/N<sup>2</sup> (*v*/*v*, 35 mL/min) mixture at 400 ◦C for 4 h. The reaction was conducted at 400 ◦C, 10 MPa and 5000 h–1. The flow rate of fed syngas (with CO/H<sup>2</sup> ratio of 1 to 2.7) was controlled by a mass flow controller, and the exit gases were measured using a wet test meter. The products were analyzed using four chromatographs during the reaction. The organic gas products, consisting of hydrocarbons and methanol, were detected online on GC4000A (EastWest, Beijing, China) equipped with flame ionization detector and GDX-403 column (EastWest, Beijing, China) (3 mm, 1 m). The inorganic gas products were detected online by thermal conductivity measurements using an EastWest GC4000A (carbon molecular sieves column, 3 m, 3m). The H2O and methanol products in the liquid phase were detected by thermal conductivity measurements using a GC4000A (Shimazduo, Kyoto, Japan) (GDX-401 column, Shimazduo, Kyoto, Japan, 3 mm, 3 m). The alcohol products in the liquid phase were detected by flame ionization measurements using a Shimazduo GC-7AG (Shimazduo, Kyoto, Japan) (Chromosorb 101, Shimazduo, Kyoto, Japan, 3 mm, 4 m).

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

### *3.1. Catalyst Characterization*

### 3.1.1. XRD

The XRD patterns of Al2O3, Al2O3-900, Cu/Al2O3, K-Cu/Al2O<sup>3</sup> catalysts (Figure 1a) and SiO2, SiO2-900, Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts (Figure 1b) are shown in Figure 1. No obvious changes were observed in the XRD patterns of the supports (Al2O<sup>3</sup> and SiO2) before and after calcination, revealing that tuning calcination temperature did not influence the phases of the supports. In the case of the Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> catalysts, the XRD patterns were very different from that of the supports. Specifically, the diffraction peaks (2θ = 31.3, 39.4, 42.6, 52.5, and 55.7◦ ) of the CuAlO<sup>2</sup> phase (JCPDS no. 39-0246) [43] and the peaks (2θ = 31.3, 36.9, 44.9, 55.7, 59.5, 65.3, 77.2, and 80.8◦ ) of the CuAl2O<sup>4</sup> phase (JCPDS no. 33-0448) [16,43] appeared in the Cu/Al2O<sup>3</sup> catalysts (as shown in Figure 1a). Unlike the Cu/Al2O<sup>3</sup> catalyst, the Cu/SiO<sup>2</sup> catalyst showed the peaks (2θ = 35.7 and 38.9◦ ) of CuO phase (JCPDS no. 05-661) [22,44] and the peaks (2θ = 31.4, 57.5, and 62.4◦ ) of copper phyllosilicate [33,34] (as displayed in Figure 1b). In addition, potassium introduction (such as the K-Cu/Al2O<sup>3</sup> catalyst and the K-Cu/SiO<sup>2</sup> catalyst) did not seemingly induce obvious changes in the diffraction peaks. These findings clearly revealed that copper species reacted with the support to form new phases when calcined at 900 ◦C, and the supports strongly affected the forms of copper species, but the potassium had no obvious effect on the phases of the catalysts.

**Figure 1.** X-ray diffraction *(*XRD) patterns of (**a**) Al2O3, Al2O3-900, Cu/Al2O3, K-Cu/Al2O3 catalysts and (**b**) SiO2, SiO2-900, Cu/SiO2, K-Cu/SiO2 catalysts. **Figure 1.** X-ray diffraction (XRD) patterns of (**a**) Al2O<sup>3</sup> , Al2O<sup>3</sup> -900, Cu/Al2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> catalysts and (**b**) SiO<sup>2</sup> , SiO<sup>2</sup> -900, Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts.

#### 3.1.2. N2 Absorption-Desorption 3.1.2. N<sup>2</sup> Absorption-Desorption

The textural properties of the Al2O3, Al2O3-900, SiO2, SiO2-900, Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts were listed in Table 1. In comparison of the parent Al2O3 (157 m2/g) and SiO2 (160 m2/g), the BET surface areas of Al2O3-900 and SiO2-900 dramatically decreased to 87.0 and 25.8 m2/g, respectively, via calcination at 900 °C, which was strongly associated with the collapse of porous structure during the high-temperature calcination process. In addition, when copper species were introduced into the uncalcined supports, the surface areas of the Cu/Al2O3 and Cu/SiO2 catalysts sharply dropped to 41.6 and 6.69 m2/g, which were much smaller than that of Al2O3-900 and SiO2- 900, respectively. The decrease in surface areas was probably due to both the formation of interfacial composite phases and copper as a sintering agent. As also shown in Table 1, compared with the Cu/Al2O3 and K-Cu/Al2O3 catalysts, the Cu/SiO2 and K-Cu/SiO2 catalysts showed considerably lower values of the surface area (4.63–6.69 m2/g), smaller pore volume (0.006–0.008 cm3/g), and average pore diameter (5.26–5.66 nm). When potassium was added, the surface area, pore volume, and average pore diameter of Al2O3 supported catalysts further decreased (from 41.6 to 40.7 m2/g, 0.20 to 0.18 cm3/g, and 19.1 to 17.8 nm, respectively). The surface areas and pore volume of SiO2 supported catalysts also showed a decreasing trend (from 6.69 to 4.63 m2/g and 0.008 to 0.006 cm3/g, respectively), but the value of average pore diameter slightly increased from 5.26 to 5.66 nm, which was probably related to the corrosion of potassium to SiO2. The results indicated that the textural parameters of the samples were greatly affected by both supports (Al2O3, SiO2) and potassium. The N2 adsorption-desorption isotherms of the Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 The textural properties of the Al2O3, Al2O3-900, SiO2, SiO2-900, Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts were listed in Table 1. In comparison of the parent Al2O<sup>3</sup> (157 m2/g) and SiO<sup>2</sup> (160 m2/g), the BET surface areas of Al2O3-900 and SiO2-900 dramatically decreased to 87.0 and 25.8 m2/g, respectively, via calcination at 900 ◦C, which was strongly associated with the collapse of porous structure during the high-temperature calcination process. In addition, when copper species were introduced into the uncalcined supports, the surface areas of the Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> catalysts sharply dropped to 41.6 and 6.69 m2/g, which were much smaller than that of Al2O3-900 and SiO2-900, respectively. The decrease in surface areas was probably due to both the formation of interfacial composite phases and copper as a sintering agent. As also shown in Table 1, compared with the Cu/Al2O<sup>3</sup> and K-Cu/Al2O<sup>3</sup> catalysts, the Cu/SiO<sup>2</sup> and K-Cu/SiO<sup>2</sup> catalysts showed considerably lower values of the surface area (4.63–6.69 m2/g), smaller pore volume (0.006–0.008 cm3/g), and average pore diameter (5.26–5.66 nm). When potassium was added, the surface area, pore volume, and average pore diameter of Al2O<sup>3</sup> supported catalysts further decreased (from 41.6 to 40.7 m2/g, 0.20 to 0.18 cm3/g, and 19.1 to 17.8 nm, respectively). The surface areas and pore volume of SiO<sup>2</sup> supported catalysts also showed a decreasing trend (from 6.69 to 4.63 m2/g and 0.008 to 0.006 cm3/g, respectively), but the value of average pore diameter slightly increased from 5.26 to 5.66 nm, which was probably related to the corrosion of potassium to SiO2. The results indicated that the textural parameters of the samples were greatly affected by both supports (Al2O3, SiO2) and potassium.

catalysts were shown in Figure 2a. As observed, Cu/Al2O3 catalyst showed a type IV adsorption **Table 1.** Textural properties of the representative samples.


The N<sup>2</sup> adsorption-desorption isotherms of the Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts were shown in Figure 2a. As observed, Cu/Al2O<sup>3</sup> catalyst showed a type IV adsorption isotherm [9]. When potassium was added, the shape of the isotherms of Al2O<sup>3</sup> supported catalysts did not change significantly. Unlike Al2O<sup>3</sup> suppored catalysts, SiO<sup>2</sup> supported catalysts had no N<sup>2</sup> adsorption-desorption isotherms. Cu/Al2O3 41.6 0.20 19.1 K-Cu/Al2O3 40.7 0.18 17.8 SiO2 160 0.54 13.7 SiO2-900 25.8 0.07 10.6 Cu/SiO2 6.69 0.008 5.26 K-Cu/SiO2 4.63 0.006 5.66

Al2O3-900 87.0 0.39 18.1

*Catalysts* **2019**, *9*, x FOR PEER REVIEW 6 of 17

**Table 1.** Textural properties of the representative samples.

**Figure 2.** (**a**) Nitrogen adsorption-desorption isotherms and (**b**) pore size distribution of Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.* **Figure 2.** (**a**) Nitrogen adsorption-desorption isotherms and (**b**) pore size distribution of Cu/Al2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts.

3.1.3. H2-TPR The reduction behaviors of the Al2O3, SiO2, CuO, Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts were studied by H2-TPR, and the results were presented in Figure 3. No reduction peak was Figure 2b presented the pore size distribution curves of the Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts. It was clearly observed that the Cu/Al2O<sup>3</sup> catalyst had a wide range of 10–120 Å, while the addition of potassium, such as the K-Cu/Al2O<sup>3</sup> catalyst, led to no obvious change in pore size distribution. In Figure 2b, note that no pore size distribution existed in the SiO<sup>2</sup> supported catalysts.

observed in the Al2O3 and SiO2, and one reduction peak at 299 °C was clearly detected in the CuO

#### phase. As displayed in Figure 3, the H2-TPR profile of Cu/Al2O3 catalyst showed three reduction 3.1.3. H2-TPR

peaks at around 280, 540, and 800 °C, which corresponded to the reduction of CuO [45], CuAl2O4 [16], and CuAlO2 [43], respectively. When the potassium was introduced into the catalyst, only the reduction temperature of the CuO phase in the Cu/Al2O3 catalyst shifted towards a higher temperature. The observed shift could be attributed to that the chemical interaction between the copper species and alumina, which was somewhat affected by the addition of potassium, in agreement with the observations of Tien-Thao et al. [40], who reported an increase in the reduction temperature of copper in Co-Cu catalysts with increasing amounts of alkali additives. The above XRD results revealed that the diffraction peaks ascribed to CuO were not observed factually for Al2O3 supported catalysts. It was thought that CuO particles with small size were probably dispersed on Al2O3 support. From Figure 3, four reduction peaks at 435, 540, 700, and 770 °C were clearly found in the Cu/SiO2 catalyst, suggesting that four types of copper species formed on the catalyst [46]. Apparently, the addition of potassium to the Cu/SiO2 catalyst led to no obvious change in the position of all the reduction peaks, suggesting a weak influence of potassium on the interactions between Cu and Si. In comparison of two kinds of the catalyst with different supports (in Figure 3), SiO2 supported catalysts (Cu/SiO2, K-Cu/SiO2) showed a much narrower reduction temperature range than Al2O3 supported catalysts (Cu/Al2O3,K-Cu/Al2O3). It was easily understood that the copper oxide interacted with Al2O3 or SiO2, and different supports always led to different interactions, which implied different reaction behaviors on these catalysts. The reduction behaviors of the Al2O3, SiO2, CuO, Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts were studied by H2-TPR, and the results were presented in Figure 3. No reduction peak was observed in the Al2O<sup>3</sup> and SiO2, and one reduction peak at 299 ◦C was clearly detected in the CuO phase. As displayed in Figure 3, the H2-TPR profile of Cu/Al2O<sup>3</sup> catalyst showed three reduction peaks at around 280, 540, and 800 ◦C, which corresponded to the reduction of CuO [45], CuAl2O<sup>4</sup> [16], and CuAlO<sup>2</sup> [43], respectively. When the potassium was introduced into the catalyst, only the reduction temperature of the CuO phase in the Cu/Al2O<sup>3</sup> catalyst shifted towards a higher temperature. The observed shift could be attributed to that the chemical interaction between the copper species and alumina, which was somewhat affected by the addition of potassium, in agreement with the observations of Tien-Thao et al. [40], who reported an increase in the reduction temperature of copper in Co-Cu catalysts with increasing amounts of alkali additives. The above XRD results revealed that the diffraction peaks ascribed to CuO were not observed factually for Al2O<sup>3</sup> supported catalysts. It was thought that CuO particles with small size were probably dispersed on Al2O<sup>3</sup> support. From Figure 3, four reduction peaks at 435, 540, 700, and 770 ◦C were clearly found in the Cu/SiO<sup>2</sup> catalyst, suggesting that four types of copper species formed on the catalyst [46]. Apparently, the addition of potassium to the Cu/SiO<sup>2</sup> catalyst led to no obvious change in the position of all the reduction peaks, suggesting a weak influence of potassium on the interactions between Cu and Si. In comparison of two kinds of the catalyst with different supports (in Figure 3), SiO<sup>2</sup> supported catalysts (Cu/SiO2, K-Cu/SiO2) showed a much narrower reduction temperature range than Al2O<sup>3</sup> supported catalysts (Cu/Al2O3, K-Cu/Al2O3). It was easily understood that the copper oxide interacted with Al2O<sup>3</sup> or SiO2, and different supports always led to different interactions, which implied different reaction behaviors on these catalysts.

*Catalysts* **2019**, *9*, x FOR PEER REVIEW 7 of 17

*Catalysts* **2019**, *9*, x FOR PEER REVIEW 7 of 17

**Figure 3.** H2-temperature-programmed reduction (H2-TPR) profiles of Al2O3, SiO2, CuO, Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.* **Figure 3.** H<sup>2</sup> -temperature-programmed reduction (H<sup>2</sup> -TPR) profiles of Al2O<sup>3</sup> , SiO<sup>2</sup> , CuO, Cu/Al2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts. **Figure 3.** H2-temperature-programmed reduction (H2-TPR) profiles of Al2O3, SiO2, CuO, Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.*

#### 3.1.4. NH3-TPD 3.1.4. NH3-TPD

The acidity of Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts was studied by NH3-TPD measurements, and the results were shown in Figure 4. No NH3 desorption peak was found over the CuO phase, revealing that the acid of the CuO phase was very weak. It was clearly observed that the NH3-TPD profiles of the Al2O3 and Cu/Al2O3 catalysts were exactly the same. Specifically, two peaks at 270 and 500 °C, ascribed to the weak acidic sites and the strong acidic sites, respectively, were obviously observed in the Al2O3 and Cu/Al2O3 catalysts. These results indicated that the acid stemmed mainly from the Al2O3 support. When the potassium was introduced, the peak of weak acidic sites shifted to a lower temperature, and yet that of strong acidic sites slightly shifted towards a higher temperature, revealing that the strength of weak acidic sites decreased and the strength of strong acidic sites increased slightly. In comparison of the area for NH3 desorption, it was apparent that the ammonia amounts of both weak acidic sites and strong acidic sites obviously decreased when adding the potassium, due to the partial neutralization of the surface acidity by alkali compounds [36,39,40]. As known, SiO2 possesses remarkably weak acidity. Therefore, the present SiO2 supported Cu catalysts (eg., Cu/SiO2, K-Cu/SiO2) showed no NH3 desorption peak, with or without the potassium addition [47]. These findings indicated that the acid-base property of the prepared catalysts was closely related to the support employed, such as SiO2 and Al2O3, wherein, the difference in acid-base property easily resulted in obviously different reaction behaviors. The acidity of Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts was studied by NH3-TPD measurements, and the results were shown in Figure 4. No NH<sup>3</sup> desorption peak was found over the CuO phase, revealing that the acid of the CuO phase was very weak. It was clearly observed that the NH3-TPD profiles of the Al2O<sup>3</sup> and Cu/Al2O<sup>3</sup> catalysts were exactly the same. Specifically, two peaks at 270 and 500 ◦C, ascribed to the weak acidic sites and the strong acidic sites, respectively, were obviously observed in the Al2O<sup>3</sup> and Cu/Al2O<sup>3</sup> catalysts. These results indicated that the acid stemmed mainly from the Al2O<sup>3</sup> support. When the potassium was introduced, the peak of weak acidic sites shifted to a lower temperature, and yet that of strong acidic sites slightly shifted towards a higher temperature, revealing that the strength of weak acidic sites decreased and the strength of strong acidic sites increased slightly. In comparison of the area for NH<sup>3</sup> desorption, it was apparent that the ammonia amounts of both weak acidic sites and strong acidic sites obviously decreased when adding the potassium, due to the partial neutralization of the surface acidity by alkali compounds [36,39,40]. As known, SiO<sup>2</sup> possesses remarkably weak acidity. Therefore, the present SiO<sup>2</sup> supported Cu catalysts (eg., Cu/SiO2, K-Cu/SiO2) showed no NH<sup>3</sup> desorption peak, with or without the potassium addition [47]. These findings indicated that the acid-base property of the prepared catalysts was closely related to the support employed, such as SiO<sup>2</sup> and Al2O3, wherein, the difference in acid-base property easily resulted in obviously different reaction behaviors. 3.1.4. NH3-TPD The acidity of Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts was studied by NH3-TPD measurements, and the results were shown in Figure 4. No NH3 desorption peak was found over the CuO phase, revealing that the acid of the CuO phase was very weak. It was clearly observed that the NH3-TPD profiles of the Al2O3 and Cu/Al2O3 catalysts were exactly the same. Specifically, two peaks at 270 and 500 °C, ascribed to the weak acidic sites and the strong acidic sites, respectively, were obviously observed in the Al2O3 and Cu/Al2O3 catalysts. These results indicated that the acid stemmed mainly from the Al2O3 support. When the potassium was introduced, the peak of weak acidic sites shifted to a lower temperature, and yet that of strong acidic sites slightly shifted towards a higher temperature, revealing that the strength of weak acidic sites decreased and the strength of strong acidic sites increased slightly. In comparison of the area for NH3 desorption, it was apparent that the ammonia amounts of both weak acidic sites and strong acidic sites obviously decreased when adding the potassium, due to the partial neutralization of the surface acidity by alkali compounds [36,39,40]. As known, SiO2 possesses remarkably weak acidity. Therefore, the present SiO2 supported Cu catalysts (eg., Cu/SiO2, K-Cu/SiO2) showed no NH3 desorption peak, with or without the potassium addition [47]. These findings indicated that the acid-base property of the prepared catalysts was closely related to the support employed, such as SiO2 and Al2O3, wherein, the difference in acid-base property easily resulted in obviously different reaction behaviors.

**Figure 4.** The temperature-programmed desorption of ammonia (NH3-TPD) profiles of Al2O3, SiO2, CuO, Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.* **Figure 4.** The temperature-programmed desorption of ammonia (NH3-TPD) profiles of Al2O3, SiO2, CuO, Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.* **Figure 4.** The temperature-programmed desorption of ammonia (NH<sup>3</sup> -TPD) profiles of Al2O<sup>3</sup> , SiO<sup>2</sup> , CuO, Cu/Al2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts.

#### 3.1.5. XPS the peak (933.3 eV) of CuO and the peak (935.0 eV) of copper phyllosilicate [33,34,48] were clearly

3.1.5. XPS

3.1.5. XPS

XPS measurements of the representative catalysts were carried out in order to investigate the chemical state of the elements at the catalyst surface and the results were shown in Figures 5–8. observed from Figure 5. Further, the BEs values of Cu2+ species shifted to the lower position with the addition of potassium. The results revealed that the chemical states of the Cu element were strongly affected by the supports employed, as well as potassium addition.

Cu/Al2O3, the peaks of the Cu2+ species shifted to lower BEs values. In the case of the Cu/SiO2 catalyst,

*Catalysts* **2019**, *9*, x FOR PEER REVIEW 8 of 17

*Catalysts* **2019**, *9*, x FOR PEER REVIEW 8 of 17

XPS measurements of the representative catalysts were carried out in order to investigate the

The Cu2p3/2 binding energies (BEs) spectra of the catalysts were displayed in Figure 5. For Cu/Al2O3 sample, the peak appearing at ~934.0 eV, along with the shakeup satellites (940–945 eV), suggested the presence of Cu2+ species [30]. The asymmetry of the Cu2p3/2 envelope could be deconvoluted into two peaks centered at around 933.0 and 935.0 eV, which were ascribed to Cu2+ in CuO and Cu2+ in CuAl2O4 respectively [16,17]. It indicated that copper oxides reacted with Al2O3 to form interfacial composite phases, consistent with the XRD results. When potassium was added into Cu/Al2O3, the peaks of the Cu2+ species shifted to lower BEs values. In the case of the Cu/SiO2 catalyst, the peak (933.3 eV) of CuO and the peak (935.0 eV) of copper phyllosilicate [33,34,48] were clearly observed from Figure 5. Further, the BEs values of Cu2+ species shifted to the lower position with the addition of potassium. The results revealed that the chemical states of the Cu element were strongly

chemical state of the elements at the catalyst surface and the results were shown in Figure 5–8.

affected by the supports employed, as well as potassium addition.

chemical state of the elements at the catalyst surface and the results were shown in Figure 5–8.

XPS measurements of the representative catalysts were carried out in order to investigate the

The Cu2p3/2 binding energies (BEs) spectra of the catalysts were displayed in Figure 5. For Cu/Al2O3 sample, the peak appearing at ~934.0 eV, along with the shakeup satellites (940–945 eV), suggested the presence of Cu2+ species [30]. The asymmetry of the Cu2p3/2 envelope could be deconvoluted into two peaks centered at around 933.0 and 935.0 eV, which were ascribed to Cu2+ in

**Figure 5.** Cu2p3/2 X-ray photoelectron spectroscopy (XPS) spectra of Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.* **Figure 5.** Cu2p3/2 X-ray photoelectron spectroscopy (XPS) spectra of Cu/Al2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts. catalysts. However, it was very obvious that the potassium addition induced an increase in Cu+ species on two kinds of catalysts, which is in agreement with the observations of Lopez et al. [41].

**Figure 6. Figure 6.**  Cu LMM spectra of the reduced Cu/Al Cu LMM spectra of the reduced Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.* 2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts. states of the O element on the Cu/Al2O3 catalyst changed, apparently, by adding potassium.

**Figure 7.** O1s XPS spectra of Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.* **Figure 7.** O1s XPS spectra of Cu/Al2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts.

Figure 8 displayed Al2p XPS spectra of Cu/Al2O3, K-Cu/Al2O3 and Si2p XPS spectra of Cu/SiO2, K-Cu/SiO2 catalysts. Three peaks attributed to aluminum species were obviously observed in the Cu/Al2O3 catalyst, revealing that three forms of aluminum species were present in this catalyst, which

were found on the Cu/SiO2 catalyst, indicating that the Si element possessed two chemical states in the catalyst [34,48]. Also, adding potassium did not change the BEs values of Si2p. The results revealed that the addition of the potassium promoter altered the chemical states of the Al element,

**Figure 8.** Al2p XPS spectra of Cu/Al2O3, K-Cu/Al2O3 and Si 2p XPS spectra of Cu/SiO2, K-Cu/SiO2

while it had no influence on that of the Si element.

catalysts*.*

while it had no influence on that of the Si element.

**Figure 7.** O1s XPS spectra of Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts*.*

Figure 8 displayed Al2p XPS spectra of Cu/Al2O3, K-Cu/Al2O3 and Si2p XPS spectra of Cu/SiO2, K-Cu/SiO2 catalysts. Three peaks attributed to aluminum species were obviously observed in the Cu/Al2O3 catalyst, revealing that three forms of aluminum species were present in this catalyst, which were related to the Al2O3, which formed CuAl2O4 and CuAlO2, respectively [43]. The BEs values of Al2p obviously decreased when potassium was added, suggesting that the three chemical states of the Al element were affected by potassium. It was noted that two peaks centered at 103.1 and 103.9 eV were found on the Cu/SiO2 catalyst, indicating that the Si element possessed two chemical states in the catalyst [34,48]. Also, adding potassium did not change the BEs values of Si2p. The results

O1s XPS spectra of Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts were given in Figure 7. The O1s signal of Cu/Al2O3 catalyst showed three overlapping peaks at around 530.1, 531.2, and 532.4 eV, indicating that three oxygen compounds formed on the catalyst surface [43]. After adding the potassium, a remarkable decrease in the BEs values was observed. However, different than the O1s XPS patterns of Al2O3 supported catalysts, two forms of oxygen compounds with higher BEs values were monitored over SiO2 supported catalysts. Furthermore, the BEs values did not change with the addition of potassium. These results clearly revealed that the electronic environments of the O element on Al2O3 and SiO2 supported catalysts were significantly different, and only the chemical

states of the O element on the Cu/Al2O3 catalyst changed, apparently, by adding potassium.

**Figure 8.** Al2p XPS spectra of Cu/Al2O3, K-Cu/Al2O3 and Si 2p XPS spectra of Cu/SiO2, K-Cu/SiO2 catalysts*.* **Figure 8.** Al2p XPS spectra of Cu/Al2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> and Si 2p XPS spectra of Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts.

The Cu2p3/2 binding energies (BEs) spectra of the catalysts were displayed in Figure 5. For Cu/Al2O<sup>3</sup> sample, the peak appearing at ~934.0 eV, along with the shakeup satellites (940–945 eV), suggested the presence of Cu2+ species [30]. The asymmetry of the Cu2p3/2 envelope could be deconvoluted into two peaks centered at around 933.0 and 935.0 eV, which were ascribed to Cu2+ in CuO and Cu2+ in CuAl2O<sup>4</sup> respectively [16,17]. It indicated that copper oxides reacted with Al2O<sup>3</sup> to form interfacial composite phases, consistent with the XRD results. When potassium was added into Cu/Al2O3, the peaks of the Cu2+ species shifted to lower BEs values. In the case of the Cu/SiO<sup>2</sup> catalyst, the peak (933.3 eV) of CuO and the peak (935.0 eV) of copper phyllosilicate [33,34,48] were clearly observed from Figure 5. Further, the BEs values of Cu2+ species shifted to the lower position with the addition of potassium. The results revealed that the chemical states of the Cu element were strongly affected by the supports employed, as well as potassium addition.

The distinction between the Cu<sup>+</sup> and Cu<sup>0</sup> species is feasible through the examination of Cu LMM XAES spectra. From the Cu LMM XAES of the reduced Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts shown in Figure 6, each Cu LMM spectrum contained two peaks centered at about 914–915 eV and 917–918 eV, with respect to the Cu<sup>+</sup> and Cu<sup>0</sup> species [34,48]. From the deconvolution results (inset), SiO<sup>2</sup> supported catalysts showed a slightly higher ratio of Cu<sup>+</sup> than Al2O<sup>3</sup> supported catalysts. However, it was very obvious that the potassium addition induced an increase in Cu<sup>+</sup> species on two kinds of catalysts, which is in agreement with the observations of Lopez et al. [41].

O1s XPS spectra of Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts were given in Figure 7. The O1s signal of Cu/Al2O<sup>3</sup> catalyst showed three overlapping peaks at around 530.1, 531.2, and 532.4 eV, indicating that three oxygen compounds formed on the catalyst surface [43]. After adding the potassium, a remarkable decrease in the BEs values was observed. However, different than the O1s XPS patterns of Al2O<sup>3</sup> supported catalysts, two forms of oxygen compounds with higher BEs values were monitored over SiO<sup>2</sup> supported catalysts. Furthermore, the BEs values did not change with the addition of potassium. These results clearly revealed that the electronic environments of the O element on Al2O<sup>3</sup> and SiO<sup>2</sup> supported catalysts were significantly different, and only the chemical states of the O element on the Cu/Al2O<sup>3</sup> catalyst changed, apparently, by adding potassium.

Figure 8 displayed Al2p XPS spectra of Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Si2p XPS spectra of Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts. Three peaks attributed to aluminum species were obviously observed in the Cu/Al2O<sup>3</sup> catalyst, revealing that three forms of aluminum species were present in this catalyst, which were related to the Al2O3, which formed CuAl2O<sup>4</sup> and CuAlO2, respectively [43]. The BEs values of Al2p obviously decreased when potassium was added, suggesting that the three chemical states of the Al element were affected by potassium. It was noted that two peaks centered at 103.1 and 103.9 eV were found on the Cu/SiO<sup>2</sup> catalyst, indicating that the Si element possessed two chemical states in the catalyst [34,48]. Also, adding potassium did not change the BEs values of Si2p. The results revealed

that the addition of the potassium promoter altered the chemical states of the Al element, while it had no influence on that of the Si element.

#### *3.2. Catalyst Evaluation*

The performances of the Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts for higher alcohols synthesis from syngas were presented in Table 2. It was observed that the supports, such as Al2O<sup>3</sup> and SiO2, could profoundly influence catalytic behaviors. CO conversion of 84.6% and total alcohol selectivity of 7.7%, wherein the percentages of methanol and C2+ alcohols were 44.0 and 56.0 wt %, respectively, were achieved over the Cu/Al2O<sup>3</sup> catalyst corresponding to CO<sup>2</sup> selectivity of 23.0%. Conversely, the Cu/SiO<sup>2</sup> catalyst showed relatively low CO conversion (18.2%) and CO<sup>2</sup> selectivity (2.4%), but much higher total alcohol selectivity (26.7%) in spite of slight lower percentage of C2+ alcohols (40.3 wt %). These results indicated that the supports had obvious effects on CO conversion, CO<sup>2</sup> selectivity, total alcohol selectivity, and alcohol distribution.



Reaction conditions: 10 MPa, 400 ◦C, 5000 h−<sup>1</sup> .

Further, one could observe from Table 2 that the potassium addition induced obviously different reaction behaviors over the Al2O<sup>3</sup> and SiO<sup>2</sup> supported catalysts. Seemingly, for Al2O<sup>3</sup> supported catalysts, the potassium addition resulted in a dramatic decrease in CO conversion, CH<sup>4</sup> selectivity, and C2–5 hydrocarbons selectivity (from 84.6 to 48.5%, from 42.2 to 23.2%, and from 27.1 to 19.5%, respectively) but an obvious increase in CO<sup>2</sup> selectivity and total alcohol selectivity (from 23.0 to 32.5% and from 7.7 to 24.1%, respectively). Besides, the K-CuO/Al2O<sup>3</sup> showed a relatively lower selectivity to methanol but a higher one to C2+ alcohols, indicating that alcohol chain-growth was enhanced. Chain-growth probabilities (α) of alcohols were calculated, as shown in Figure 9. In the case of SiO<sup>2</sup> supported catalysts, the change trend in the CH<sup>4</sup> selectivity, CO<sup>2</sup> selectivity, and alcohol chain-growth, induced by potassium addition, was similar with that in Al2O<sup>3</sup> supported catalysts. Whereas, different than the Cu/Al2O<sup>3</sup> catalyst, adding the potassium into the Cu/SiO<sup>2</sup> catalyst did not significantly change the values of CO conversion (~17.0%), total alcohol selectivity (~27.0%), and C2–5 hydrocarbons selectivity (~28.0%). Conclusively, potassium introduction did not only promote the formation of CO<sup>2</sup> and inhibit the CH<sup>4</sup> formation, but also enhanced the carbon chain growth probability of products; moreover, the potassium had a greater impact on the CO conversion, total alcohol selectivity, and C2–5 hydrocarbons selectivity over the Al2O<sup>3</sup> supported catalysts than that over SiO<sup>2</sup> supported catalysts.

*Catalysts* **2019**, *9*, x FOR PEER REVIEW 11 of 17

**Figure 9.** Anderson-Schulz-Flory (A-S-F) plots for the distribution of alcohols for catalysts: (**a**) Cu/Al2O3, (**b**) K-Cu/Al2O3 and (**c**) Cu/SiO2, (**d**) K-Cu/SiO2. Wn = n × (1 − α)2 × α<sup>n</sup>−1, where Wn stands for the mass fraction of alcohols containing n carbon atoms. **Figure 9.** Anderson-Schulz-Flory (A-S-F) plots for the distribution of alcohols for catalysts: (**a**) Cu/Al2O<sup>3</sup> , (**b**) K-Cu/Al2O<sup>3</sup> and (**c**) Cu/SiO<sup>2</sup> , (**d**) K-Cu/SiO<sup>2</sup> . Wn = n × (1 − α) <sup>2</sup> <sup>×</sup> <sup>α</sup> n−1 , where Wn stands for the mass fraction of alcohols containing n carbon atoms.

#### *3.3. In situ FTIR 3.3. In Situ FTIR*

To further obtain detailed information on the molecular events that occur on the surface of the catalyst, CO adsorption over reduced Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts was monitored using in situ FTIR. As displayed in Figure 10, there were two absorption bands in the region of 3200–2850 cm−1 (ν C-H) and 1650–1300 cm−1 (ν COO) observed, assigned to the adsorbed hydrocarbons and formate species (C1 oxygenate species), respectively [49,50]. For the Cu/Al2O3 catalyst, very weak peaks at 1650–1300 cm−1 were detected, revealing the presence of only a trace of C1 oxygenate species on the catalyst surface, which were approved to contribute to the formation of higher alcohols [2,41,50]. It was also noted that the peaks at around 3200–2850 cm−1, ascribed to the hydrogenation, was strong on the Cu/Al2O3 catalyst. When potassium was added, the band peaks of the C1 oxygenate species obviously increased, whereas the intensity of hydrocarbons decreased. As Santos et al. [2] reported, potassium, in close vicinity to an adsorbed methyl group, stabilized oxygenate species that were found to play an important role in the syngas to alcohol route. Thus, the reaction shifted towards alcohols, rather than hydrocarbons, when potassium modified the Cu/Al2O3 catalyst. When SiO2 was used as support, it was noted that the peak at 1650–1300 cm−1 ascribed to the C1 oxygenate species was enhanced on the Cu/SiO2 catalyst, which was very different with that of Cu/Al2O3 catalyst, on which only trace amounts of the C1 oxygenate species were formed (Figure 10); moreover, it seemed that adding potassium had no effect on the amount of C1 oxygenate species formed. This explained why the Cu/SiO2 and K-Cu/SiO2 catalysts exhibited the similar alcohol selectivity of ~27% (Table 2). To further obtain detailed information on the molecular events that occur on the surface of the catalyst, CO adsorption over reduced Cu/Al2O3, K-Cu/Al2O<sup>3</sup> and Cu/SiO2, K-Cu/SiO<sup>2</sup> catalysts was monitored using in situ FTIR. As displayed in Figure 10, there were two absorption bands in the region of 3200–2850 cm−<sup>1</sup> (ν C-H) and 1650–1300 cm−<sup>1</sup> (ν COO) observed, assigned to the adsorbed hydrocarbons and formate species (C1 oxygenate species), respectively [49,50]. For the Cu/Al2O<sup>3</sup> catalyst, very weak peaks at 1650–1300 cm−<sup>1</sup> were detected, revealing the presence of only a trace of C1 oxygenate species on the catalyst surface, which were approved to contribute to the formation of higher alcohols [2,41,50]. It was also noted that the peaks at around 3200–2850 cm−<sup>1</sup> , ascribed to the hydrogenation, was strong on the Cu/Al2O<sup>3</sup> catalyst. When potassium was added, the band peaks of the C1 oxygenate species obviously increased, whereas the intensity of hydrocarbons decreased. As Santos et al. [2] reported, potassium, in close vicinity to an adsorbed methyl group, stabilized oxygenate species that were found to play an important role in the syngas to alcohol route. Thus, the reaction shifted towards alcohols, rather than hydrocarbons, when potassium modified the Cu/Al2O<sup>3</sup> catalyst. When SiO<sup>2</sup> was used as support, it was noted that the peak at 1650–1300 cm−<sup>1</sup> ascribed to the C1 oxygenate species was enhanced on the Cu/SiO<sup>2</sup> catalyst, which was very different with that of Cu/Al2O<sup>3</sup> catalyst, on which only trace amounts of the C1 oxygenate species were formed (Figure 10); moreover, it seemed that adding potassium had no effect on the amount of C1 oxygenate species formed. This explained why the Cu/SiO<sup>2</sup> and K-Cu/SiO<sup>2</sup> catalysts exhibited the similar alcohol selectivity of ~27% (Table 2).

**Figure 10.** In situ Fourier-transform infrared (FTIR) spectra of CO adsorption over reduced Cu/Al2O3, K-Cu/Al2O3 and Cu/SiO2, K-Cu/SiO2 catalysts during CO flow at 400 °C for 30 min and then under Ar at 400 °C for 30 min. **Figure 10.** In situ Fourier-transform infrared (FTIR) spectra of CO adsorption over reduced Cu/Al2O<sup>3</sup> , K-Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> , K-Cu/SiO<sup>2</sup> catalysts during CO flow at 400 ◦C for 30 min and then under Ar at 400 ◦C for 30 min.

#### *3.4. Discussion 3.4. Discussion*

The results obtained by the test and characterizations of Cu/Al2O3 and Cu/SiO2 samples clearly demonstrated that the physicochemical properties and catalytic performances of the Cu-based catalyst were strongly affected by the types of supports employed (Al2O3, SiO2). In comparison to the Cu/Al2O3 catalyst with only 7.7% of total alcohol selectivity, the Cu/SiO2 catalyst possessed a much higher total alcohol selectivity of 26.7%. As confirmed by the FTIR result, only a trace of the C1 oxygenate species was detected on the Cu/Al2O3 catalyst, but a relatively large amount of the C1 oxygenate species existed on the surface of the Cu/SiO2 catalyst. Some authors have pointed out that the oxygenate species played an essential role in directing the synthesis toward alcohols, rather than hydrocarbons [39–41,50,51]. Zhang et al. [52,53] conducted a series of DFT studies to assess the mechanism of CO hydrogenation to higher alcohols on Cu (110) [52] and Cu (211) [53], and pointed out that the CHxO species as key intermediates for higher alcohols synthesis could give the CHx species through the C-O cleavage, and CHx monomers subsequently combined with CO or CHO to form alcohol. Based on the above, it was apparent that the more adsorbed C1 oxygenate species on the Cu/SiO2 catalyst inevitably resulted in an increase in the concentration of CHx species via the C-O cleavage, and, finally, promoted the formation of alcohol. Hence, the Cu/SiO2 catalyst with a higher amount of the C1 oxygenate species exhibited a higher total alcohols selectivity than the Cu/Al2O3 catalyst. Further, by comparing to the Cu/SiO2 catalyst, the Cu/Al2O3 catalyst showed higher selectivities towards C2+ alcohols. As reported [54–56], aldol condensation as one of the key steps for carbon-chain growth of alcohol products easily proceeded on basic oxide catalysts or acid oxide catalysts. The NH3-TPD result revealed that the surface acidity of the Cu/Al2O3 catalyst was much stronger than that of the Cu/SiO2 catalyst. Therefore, the possibility of carbon-chain growth occurred on the acid Cu/Al2O3 catalyst more easily. Despite that the ethanol selectivities were both 32.6 wt % above on Al2O3 and SiO2-supported catalysts, which was probably due to high Cu+/(Cu++Cu0) values, Cu/SiO2 with a relatively high Cu+/(Cu++Cu0) value did not give higher ethanol selectivity than Cu/Al2O3. Our previous work [16,17] has demonstrated that the amounts of Cu+ species were somewhat responsible for the formation of ethanol. These results suggested that the ethanol formation was affected by many factors, such as Cu+/(Cu++Cu0) value, the physical properties, the reduction behaviors, acidity, and electronic properties on the catalyst, which are synergetic. The results obtained by the test and characterizations of Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> samples clearly demonstrated that the physicochemical properties and catalytic performances of the Cu-based catalyst were strongly affected by the types of supports employed (Al2O3, SiO2). In comparison to the Cu/Al2O<sup>3</sup> catalyst with only 7.7% of total alcohol selectivity, the Cu/SiO<sup>2</sup> catalyst possessed a much higher total alcohol selectivity of 26.7%. As confirmed by the FTIR result, only a trace of the C1 oxygenate species was detected on the Cu/Al2O<sup>3</sup> catalyst, but a relatively large amount of the C1 oxygenate species existed on the surface of the Cu/SiO<sup>2</sup> catalyst. Some authors have pointed out that the oxygenate species played an essential role in directing the synthesis toward alcohols, rather than hydrocarbons [39–41,50,51]. Zhang et al. [52,53] conducted a series of DFT studies to assess the mechanism of CO hydrogenation to higher alcohols on Cu (110) [52] and Cu (211) [53], and pointed out that the CHxO species as key intermediates for higher alcohols synthesis could give the CH<sup>x</sup> species through the C-O cleavage, and CH<sup>x</sup> monomers subsequently combined with CO or CHO to form alcohol. Based on the above, it was apparent that the more adsorbed C1 oxygenate species on the Cu/SiO<sup>2</sup> catalyst inevitably resulted in an increase in the concentration of CH<sup>x</sup> species via the C-O cleavage, and, finally, promoted the formation of alcohol. Hence, the Cu/SiO<sup>2</sup> catalyst with a higher amount of the C1 oxygenate species exhibited a higher total alcohols selectivity than the Cu/Al2O<sup>3</sup> catalyst. Further, by comparing to the Cu/SiO<sup>2</sup> catalyst, the Cu/Al2O<sup>3</sup> catalyst showed higher selectivities towards C2+ alcohols. As reported [54–56], aldol condensation as one of the key steps for carbon-chain growth of alcohol products easily proceeded on basic oxide catalysts or acid oxide catalysts. The NH3-TPD result revealed that the surface acidity of the Cu/Al2O<sup>3</sup> catalyst was much stronger than that of the Cu/SiO<sup>2</sup> catalyst. Therefore, the possibility of carbon-chain growth occurred on the acid Cu/Al2O<sup>3</sup> catalyst more easily. Despite that the ethanol selectivities were both 32.6 wt % above on Al2O<sup>3</sup> and SiO2-supported catalysts, which was probably due to high Cu+/(Cu++Cu<sup>0</sup> ) values, Cu/SiO<sup>2</sup> with a relatively high Cu+/(Cu++Cu<sup>0</sup> ) value did not give higher ethanol selectivity than Cu/Al2O3. Our previous work [16,17] has demonstrated that the amounts of Cu<sup>+</sup> species were somewhat responsible for the formation of ethanol. These results suggested that the ethanol formation was affected by many factors, such as Cu+/(Cu++Cu<sup>0</sup> ) value, the physical properties, the reduction behaviors, acidity, and electronic properties on the catalyst, which are synergetic.

In addition, CO conversion and CO2 selectivity of the Cu/Al2O3 catalyst also showed significant differences from that of the Cu/SiO2 catalyst. As revealed by the XRD and H2-TPR results, although In addition, CO conversion and CO<sup>2</sup> selectivity of the Cu/Al2O<sup>3</sup> catalyst also showed significant differences from that of the Cu/SiO<sup>2</sup> catalyst. As revealed by the XRD and H2-TPR results, although

the refractory phases formed over Al2O3 or SiO2 supported Cu catalysts under 900 °C, the two kinds of the catalysts with different supports presented differences in the types of copper species and the the refractory phases formed over Al2O<sup>3</sup> or SiO<sup>2</sup> supported Cu catalysts under 900 ◦C, the two kinds of the catalysts with different supports presented differences in the types of copper species and the reduction behaviors of copper species. N<sup>2</sup> absorption-desorption results indicated that the textural parameters of the Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> catalysts, despite uncalcined supports with similar surface areas (157–160 m2/g), were significantly different. Furthermore, the Cu/Al2O<sup>3</sup> catalyst with a type IV adsorption isotherm showed a wide pore size distribution, but the Cu/SiO<sup>2</sup> catalyst had no N<sup>2</sup> adsorption-desorption isotherms and pore size distribution. The NH3-TPD result suggested that the acidity was obviously detected in the Cu/Al2O<sup>3</sup> catalyst, but no acidity was observed in the Cu/SiO<sup>2</sup> catalyst. Moreover, the XPS results showed that the Cu, O elements of the Cu/Al2O<sup>3</sup> catalyst differed from that of the Cu/SiO<sup>2</sup> catalyst in electronic environments. As confirmed by the FTIR result, only a trace of C1 oxygenate species, contributing to alcohols formation, was detected on the Cu/Al2O<sup>3</sup> catalyst, while a relatively large amount of C1 oxygenate species existed on the surface of the Cu/SiO<sup>2</sup> catalyst. These characterization results indicated that the structural and chemical properties of Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> catalysts showed obvious differences and thus affected the catalytic behaviors of the catalysts synergistically.

The performances of two kinds of catalysts with the support employed, such as SiO<sup>2</sup> and Al2O3, were also significantly affected by the potassium introduction. Doping potassium into the Al2O<sup>3</sup> and SiO<sup>2</sup> supported catalysts improved the carbon chain growth probability of alcohol products. This was probably because potassium introduction provided new basic sites for the aldol condensation of lower alcohols to higher alcohols [40,57]. Moreover, alkali elements are known to be good promoters for the WGS reaction (CO + H2O → CO<sup>2</sup> + H2) when introduced at optimum content [9,15,39,50]. As a result, the addition of potassium to Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> catalysts resulted in the improvement of CO<sup>2</sup> formation.

Additionally, according to the test results, doping potassium into Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> catalysts induced distinct differences in the total alcohol selectivity. For Al2O<sup>3</sup> supported catalysts, total alcohol selectivity increased up to 24.1 from previous 7.7%, clearly, when the potassium was added. It indicated that the presence of potassium promoted the formation of alcohols. As confirmed by the FTIR result, by adding the K promoter, the relative amount of adsorbed C1 oxygenate species, as intermediates in higher alcohols synthesis [41,50], increased obviously by tuning the reduction behavior, neutralizing the surface acidity, and altering of the electronic properties, whereas the formation of hydrocarbons (mainly CH<sup>4</sup> and C2–5 hydrocarbons) was severely inhibited [39,50]. Anton et al. [39–41] reported that alkali (K/Cs/Rb) modified Cu-based catalysts can enhance the stability of CHx intermediate species. Therefore, total alcohols selectivity on the K-Cu/Al2O<sup>3</sup> catalyst increased dramatically with potassium introduction. However, for the SiO<sup>2</sup> supported catalyst, the potassium addition hardly had an obvious effect on the relative amount of C1 oxygenate species, thus the total alcohol selectivity on K-Cu/SiO<sup>2</sup> catalyst remained almost unchanged. Due to very weak acidity of the SiO<sup>2</sup> supported catalyst, only CH<sup>4</sup> selectivity decreased sharply after potassium introduction, implying that potassium addition more preferentially inhibited CH<sup>4</sup> formation, compared with that of C2–5 hydrocarbons. Obviously, potassium addition could somehow modify the structural and chemical properties of the Cu/Al2O<sup>3</sup> catalyst and enhance the amount of C1 oxygenate species, which narrowed the gap in performances of the two Al2O<sup>3</sup> and SiO<sup>2</sup> supported catalysts.

#### **4. Conclusions**

In this work, the Al2O<sup>3</sup> and SiO<sup>2</sup> supported Cu catalysts prepared by a facile impregnation method were used for higher alcohols synthesis from CO hydrogenation. Based on the remarkably different reaction behaviors over the Cu catalysts supported on Al2O<sup>3</sup> and SiO2, some systematical investigations were carried out to understand the main reasons. The Cu/SiO<sup>2</sup> catalyst possessed a higher amount of the C1 oxygenate species and showed higher total alcohols selectivity than the Cu/Al2O<sup>3</sup> catalyst. Compared to the very weak acidity of the SiO<sup>2</sup> supported catalyst, the carbon chain growth probability of alcohol products occurred on the acid Cu/Al2O<sup>3</sup> catalyst more easily. Further, the Cu/Al2O<sup>3</sup> and Cu/SiO<sup>2</sup> catalysts showed obvious differences in the structural and physicochemical properties, such as the types of copper species, the reduction behaviors, acidity, and electronic properties. As a result, the CO conversion, alcohol distribution, and CO<sup>2</sup> selectivity of the Cu/Al2O<sup>3</sup> catalyst were different from that of the Cu/SiO<sup>2</sup> catalyst. Additionally, the performances of the Cu catalysts supported on Al2O<sup>3</sup> and SiO<sup>2</sup> became more similar when the potassium was introduced. Wherein, the potassium was approved to modify the structural and chemical properties of the catalysts to some extent.

**Author Contributions:** The idea was conceived by Y.T. and X.L.; X.L. performed the experiments and drafted the paper under the supervision of Y.T. and J.Z.; M.Z. (Min Zhang), W.Z., M.Z. (Meng Zhang), H.X. and Y.W. helped to collect and analyse some characterization data. The manuscript was revised and checked through the comments of all authors. All authors have given approval for the final version of the manuscript.

**Funding:** This research was funded by the National Natural Science Foundation of China (No. 21573269), the Key Research and Development Program of Shanxi Province (No. MD2014-10).

**Acknowledgments:** The authors thank the financial support of National Natural Science Foundation of China (No. 21573269), the Key Research and Development Program of Shanxi Province (No. MD2014-10).

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