1 Introduction

The transformation of synthesis gas into higher-value-added products is being increasingly investigated because of the substantial reservoirs of natural gas in remote areas and because synthesis gas is a clean source of low-sulfur-content fuels [1–8]. Currently, synthesis gas from sources such as biomass and coal are also being considered [9, 10]. Synthesis gas can be transformed directly into dimethyl ether (DME) over a hybrid catalyst based on a methanol synthesis catalyst and acidic porous materials, such as zeolites, silica-aluminas, and aluminas [11]. The synthesis of the hydrocarbons from synthesis gas via methanol is also possible over hybrid catalysts involving the follow reactions: Methanol synthesis:

$${\text{CO}} + 2{\text{H}}_{2} \leftrightarrow {\text{CH}}_{3} {\text{OH}} .$$
(1)

Water gas shift reaction:

$${\text{CO}} + {\text{H}}_{2} {\text{O}} \leftrightarrow {\text{CO}}_{2} + {\text{H}}_{2} .$$
(2)

Dehydration of methanol to DME:

$$2{\text{CH}}_{3} {\text{OH}} \leftrightarrow {\text{CH}}_{3} {\text{OCH}}_{3} + {\text{H}}_{2} {\text{O}} . { }$$
(3)

Formation of hydrocarbons:

$${\text{CH}}_{3} {\text{OCH}}_{3} \leftrightarrow {\text{Hydrocarbons}} + {\text{H}}_{2} {\text{O}} .$$
(4)

Net Reaction:

$$2n{\text{CO}} + \left( {n + 1} \right){\text{H}}_{2} \leftrightarrow {\text{C}}_{n} {\text{H}}_{2n + 2} + n{\text{CO}}_{2} ,$$
(5)

(n = 1, 2,…0.6; for LPG, n = 3 or 4)

Reactions (1) and (2) are catalyzed by a methanol synthesis catalyst, and Reactions (3) and (4) are catalyzed by an acid material. The formation of DME and the hydrocarbons (HCs) assist in overcoming the thermodynamic constraints inherent to methanol synthesis. Therefore, the direct synthesis of DME and these hydrocarbons could be performed over a hybrid catalyst that is composed of two components [7, 12, 13].

Hybrid catalysts based on the use of Cr–Zn and Pd/SiO2 as a methanol synthesis catalyst have been employed because of their stable activity at high temperatures [7, 12, 13]. Studies using Cu–ZnO as the methanol synthesis catalyst have also shown promising results [5, 6, 14–20]. A study by Zhang et al. [17] demonstrated that a catalyst using Cu–Zn was the most effective in synthesizing HCs from syngas.

Initial studies of the direct synthesis of HCs from syngas have focused on investigating the influence of the acid components (such as SAPO [2], HZSM-5, SAPO-34, HZSM-11, H-MCM-22, H-Beta [12], USY [1]) on the distribution of products. In short, the pore size and acidity of the second component may influence the distribution of products; however, no studies were found addressing the influence of the preparation method of the hybrid catalyst on its activity.

Selected works related to the DME synthesis from syngas using hybrid catalysts based on Cu–ZnO–Al2O3 and acid materials have studied the influence of the preparation method on the catalytic activity, and several important aspects have been emphasized. According to Li et al. [11], the proximity between the active phases of both components is a key to obtaining a higher activity. Ge et al. [21] suggested a close contact between both components and noted that their overlap may actually decrease the total catalytic activity. However, Moradi et al. [22] affirmed that the active sites for methanol synthesis and methanol dehydration must be formed in separate stages to obtain a hybrid catalyst with a higher overall activity.

In this work, several preparation methods were used to prepare the Cu–ZnO–Al2O3/H-ferrierite catalysts and to study their influence on the catalytic activity in syngas transformation. These methods are based on different types of mixtures between both components from a physical mixture until the formation of the methanol synthesis catalyst over the zeolite. The hybrid catalysts were tested in a fixed bed reactor and were characterized by nitrogen adsorption (BET), X-ray diffraction (XRD), temperature-programmed reduction (TPR), temperature-programmed desorption of hydrogen (H2-TPD), temperature-programmed desorption of ammonia (NH3-TPD) and transmission electron microscopy (TEM).

2 Experimental

2.1 Preparation of Catalysts

Hybrid catalysts were prepared by four different methods: physical mixture (PM), coprecipitation sedimentation (CS), coprecipitation impregnation (CI) and deposition precipitation (DP).

A commercial NaK-ferrierite zeolite provided by Toyo Soda Manufacturing (serial number 720 KOA) with a Si/Al ratio of 10 was used as the dehydration component. The acid form of this zeolite was obtained by ion exchange using an ammonium nitrate solution (1.7 mol dm−3) at 90 °C for 2 h. The zeolite was previously dried for 12 h and then calcined at 400 °C under N2 for 4 h.

The preparation of the hybrid catalyst by physical mixing consisted of first preparing the methanol synthesis catalyst by the coprecipitation method. Initially, two solutions (Cu, Zn and Al nitrates (1 mol dm−3) and Na2CO3 (1.5 mol dm−3)) were mixed slowly in a vessel containing 100 dm3 of water at 70 °C. This mixture maintained a pH of approximately 8 with continuous stirring. After the consumption of the reactants, the precipitate formed was aged for 1 h at the same temperature. The solid was then washed with warm water to eliminate the impurities and dried overnight at 90 °C. The atomic ratio Cu/Zn/Al used for the preparation of the methanol synthesis catalyst to form the hydrotalcite precursor was 55/30/15. This sample was then calcined at 400 °C under a flow of nitrogen (50 dm3 min−1) for 4 h before being mechanically mixed with the zeolite support. This product was named “PM”.

The coprecipitation sedimentation method consisted of mixing the two aqueous suspensions (the methanol synthesis catalyst after washing and the H-ferrierite zeolite after calcination) at room temperature. The methanol synthesis catalyst (MSC) was prepared under the same conditions mentioned above. The suspensions were mixed using vigorous stirring. The final solid was separated, dried overnight at 90 °C and finally calcined under the same conditions mentioned for the PM sample. The calcined sample was named “CS”.

In the coprecipitation impregnation method, the solutions of Cu, Zn and Al nitrates and Na2CO3 were slowly added to an H-ferrierite zeolite aqueous suspension under similar conditions of pH, temperature and stirring as mentioned above. The solid was also aged for 1 h, washed, dried, calcined and named “CI”.

The preparation by deposition precipitation consisted of slowly adding a solution of Cu, Zn and Al nitrates to an H-ferrierite aqueous suspension in Na2CO3 at 70 °C under intense stirring. After depletion of the nitrate solution, the solid was aged for 1 h at the same temperature with stirring. The sample was then calcined under the same conditions as the above samples and named “DP”.

2.2 Catalyst Characterization

The catalyst composition was determined by atomic absorption spectroscopy (AAS) using a Varian SpectraAA-200 instrument. The hybrid catalysts were dissolved in nitric and hydrochloric acid and diluted with distilled water to concentrations within the detection range of the instrument.

The BET surface area, pore volume and pore diameter were measured by N2 physisorption at −196 °C using a Micromeritics ASAP 2000. Prior to the adsorption–desorption measurements, all samples were degassed at 250 °C under vacuum.

A Rigaku diffractometer using CuKα radiation and operating at 40 kV was used for X-ray measurements. The range of the angle 2θ was between 5° and 80°.

The temperature-programmed reduction measurements (TPR) were performed in conventional TPR/TPD equipment outfitted with a thermal conductivity detector (TCD). Samples were treated with a flow of Ar up to 150 °C for 1 h. The reduction was produced with the H2/Ar mixture (1.53 % H2) at a flow of 30 cm3 min−1 and a heating rate of 10 °C min−1 up to 600 °C.

The temperature-programmed desorption of NH3 measurements (NH3-TPD) were also performed in conventional TPR/TPD equipment outfitted with a quadrupole mass spectrometer (Baltzer). The hybrid catalyst (600 mg) was initially heated at 400 °C for 2 h under a He flow of 50 cm3 min−1. The samples were then cooled to 175 °C. At this temperature, the NH3/He (3.84 % NH3) mixture was passed through the catalyst for 1 h at 300 mL min−1. The samples were then purged with He at 30 cm3 min−1 for 30 min and heated to 1000 °C at a heating rate of 10 °C min−1.

The Fourier-transform infrared (FTIR) spectra were recorded using sample waffles of thickness ca. 10 mg cm−2. Before analysis, the samples were calcined at 400 °C under an N2 atmosphere for 2 h and then evacuated to a pressure of 10−6 torr. The d3-acetonitrile was obtained from Aldrich Chem. Co. and was degassed by freeze–pump–thaw cycles. The adsorption of d3-acetonitrile on the evacuated samples was performed at 25 °C for 30 min. The adsorption was followed by evacuations at 25, 150 and 250 °C for 30 min. The spectrometer used was a Nicolet Nexus-470. The spectra were obtained at room temperature with a resolution of 4 cm−1 by collecting 120 scans for each spectrum. The bands were deconvoluted using the least squares minimization routine, approximating the bands with the Gaussian profile. Spectra using pyridine as a probe molecule were also recorded to investigate the ion exchange of the zeolite protons. Pyridine was obtained from Aldrich Chem. Co. and was degassed in the same manner as acetonitrile. In the case of pyridine, the adsorption was performed at 150 °C for 1 h, and the evacuation was made at 150 °C for 30 min.

The copper areas were measured by the temperature-programmed desorption of H2 (H2-TPD). The conditions of measurement were adapted from Mühler et al. [23], and the analysis was performed using a conventional TPR/TPD apparatus equipped with a Baltzer Mass Quadrupole. The samples (300 mg) were dried in He at 150 °C for 1 h and then reduced using pure H2 (60 cm3 min−1) at 250 °C for 1 h (heating rate 10 °C min−1). After the reduction, the samples were cooled to 0 °C and maintained at this temperature under a flow of H2 for 1 h. The samples were then cooled to −196 °C for 1 h using a liquid nitrogen bath. The samples were purged with He at a flow rate of 50 cm3 min−1 for 30 min. After the purge, the samples were heated to 350 °C to record the hydrogen desorption through the m/z = 2 fragment. Previous tests using this technique were made on commercial methanol catalyst to verify the reproducibility of the measurement.

Transmission electron microscopy was used to analyze the morphology of the hybrid catalyst. A JEOL 2010 operating at 200 kV was used. The samples were prepared using isopropyl alcohol and ultrasound waves to disperse the particles. The thinner particles were deposited over a specific sample holder grid.

2.3 Catalytic Activity Test

Catalytic activity measurements were conducted in a fixed-bed stainless-steel reactor with an internal diameter of 12.7 mm. The reaction products were analyzed on-line in a gas chromatograph with a thermal conductivity detector (TCD) for CO and CO2 and a flame ionization detector (FID) for products such as methanol, DME and hydrocarbons. The analysis was performed with Carboxen 1010 PLOT and SUPEL-Q PLOT fused silica capillary columns both with a length of 30 m.

The hybrid catalysts (500 mg) were put inside the reactor without inert glass bed, size of particle to 0.85–1.18 mm and 1.0 cm of bed height. The catalysts were activated in H2 at 250 °C (10 °C min−1) and 50 mL min−1 for 1 h under atmospheric pressure. Then, the H2/CO was introduced into the reactor, the pressure was adjusted to 2 MPa and the flow to 20 mL min−1. The tests were performed at 350 °C. Not axial temperature profile was considered because the temperature measured on the external reactor wall was the same observed inside the reactor. Furthermore flow tests carried out in empty reactor showed no mass transfer limitation through the catalyst bed.

3 Results

3.1 Composition and Textural Properties

All the samples showed Cu/Zn/Al atomic ratios similar to those specified in the experimental procedure (Table 1). The MSC/zeolite ratio shown in Table 1 was calculated considering the complete metal oxidation of MSC. The lower MSC/zeolite ratio observed in the CI and DP samples may be explained by the lack of precipitation of metals under the experimental conditions.

Table 1 Composition and textural properties of the hybrid catalysts
Full size table

The specific areas for the hybrid catalyst shown in Table 1 showed values smaller than the pure support (333 m2 g−1) and did not display any notable differences. These areas correspond to the combination of both components in the hybrid catalyst. To discriminate the support area, the specific areas of the hybrid catalysts were normalized using the weight of the zeolitic component present in the hybrid catalyst and compared with the pure support (Table 1). All samples displayed a decrease in the support area compared with the pure H-ferrierite, mainly in the CI and DP samples. Considering that the specific area of the pure support is mainly formed by micropores, the reduction in area in the hybrid catalysts can be attributed to the partial pore blockage by MSC by mixing of both components.

The pore-size distribution profiles of the hybrid catalysts in Fig. 1 presented slight differences when compared to the H-ferrierite zeolite. The differences observed in the mesoporous range were related to the inter-grain spaces of the oxide mixture. These spaces result from the agglomerates of oxide particles formed after calcination.

Fig. 1
figure 1

Pore size distribution of the hybrid catalysts

Full size image

The pore-distribution profile for the DP sample showed an intense and broadened peak indicating a high agglomeration of particles. The macropores of approximately 250 Å and of approximately 900 Å in the CI and PM samples can be attributed to the spaces formed by small particles of the oxide mixture and to the spaces formed between the MSC and the zeolite components, respectively.

The sample obtained by physical mixture presented pores with a size less than 20 Å as the pure zeolite, and the other samples (CS, DP and CI) had a decrease in these pores compared to pure zeolite. The partial blockage of the zeolite micropores can be attributed to the small particles of the methanol synthesis catalyst deposited inside the micropores while mixing the components during the formation of the hybrid catalyst. This blockage can also be attributed to the formation of a precipitate of the methanol synthesis precursor inside the micropores during the preparation process. The deposition of small particles inside the micropores seems to have been favored by the preparation method. In the physical mixture, the components are not in such intimate contact as in the coprecipitation-sedimentation method. In the coprecipitation-impregnation method, the support played an important role by acting as the seed to the methanol synthesis precursor for precipitation over its surface and the modification of the zeolite textural properties [24].

3.2 XRD

Well-defined peaks belonging to the hydrotalcite-like layered hydroxide phase (Cu x Zn6−x Al2(OH)16CO3)·4H2O) were observed in all samples including the MSC precursor synthesized for subsequent physical mixture with the zeolite (Fig. 2). These peaks were lower in the DP sample, but the samples showed high crystallinity indicating that a suitable aging time was employed. The hydrotalcite precursor provides highly dispersed CuO phases after calcination. The observed diffraction lines of the zeolite support indicated that its crystalline integrity was preserved in all the samples. The low intensity of these peaks is related to the amount of zeolite in the hybrid catalyst. The zeolite peaks did not appear in the PM sample because the MSC component in this catalyst was mixed after its calcination. Small malachite-phase peaks (noted as M in Fig. 1) were also observed.

Fig. 2
figure 2

XRD patterns of the uncalcined hybrid catalysts. H hydrotalcite, M malachite and F H-ferrierite

Full size image

After calcination, broad peaks for the CuO phase were found in all samples indicating the collapse of the precursor layered structure [24] (Fig. 3). An increase in baseline in 30°–45° range was ascribed to the presence of either highly nano-crystalline or X-ray amorphous ZnO [25–27]. This behavior can be considered as a decrease in the crystallinity and may be because of the Al present in the methanol synthesis catalyst; Al restrains the Cu and Zn component crystallites from enlarging during calcination. Al2O3 acts as a structure promoter; consequently, a high Al content favors a uniform dispersion of the different phases [28].

Fig. 3
figure 3

a XRD patterns of the calcined hybrid catalysts and b range of CuO peaks

Full size image

After calcination, the peaks of the CuO phase were more intense in the CI and PM samples, revealing a higher crystallinity and larger crystallite size compared with the CS and DP samples, as shown in Table 1.

3.3 TPR

The results of the TPR analysis are shown in Fig. 4, and their profiles were deconvoluted into Gaussian peaks. The peak positions and their contribution derived from the deconvolution are summarized in Table 2. The hybrid catalysts presented reduction profiles with two high reduction stages and a small shoulder at low temperature.

Fig. 4
figure 4

TPR profiles of the hybrid catalysts

Full size image
Table 2 Contributions of the reduction peaks, temperature and H2 to the TPR pattern hybrid catalysts
Full size table

The small shoulder is assigned as α and was ascribed to the reduction of highly dispersed CuO [29–31] or Cu2+ from the amorphous CuO dispersed on the catalyst surface as suggested by Palgunadi et al. [32]. Peaks β and γ were attributed to the reduction of CuO dispersed in a solid solution formed by CuO–ZnO [32] and to the reduction of bulk-like CuO phases, respectively [29–31].

As shown in Table 2, the contribution of each peak depends on the preparation method. A higher amount of easily-reducible well-dispersed CuO (peak β) was detected in all hybrid catalysts. The higher content of highly dispersed CuO found for the CI sample was probably caused by the exchange of Cu with the zeolite proton (H+) during the coprecipitation method used for this sample. The position of peak β was shifted to lower temperatures for the CI and CS samples. This indicated a higher reducibility of the CuO in these samples. The H2 consumption (Table 2) was similar for all hybrid catalysts samples and corresponded to the reduction of Cu(II) to Cu(0).

3.4 NH3-TPD

The NH3-TPD profiles of the hybrid catalysts are presented in Fig. 5. They showed two desorption peaks, which are in accordance with data already reported in the literature for the acidity of MFI materials [33, 34].

Fig. 5
figure 5

NH3-TPD patterns of the hybrid catalysts

Full size image

The desorption peak observed at approximately 260–270 °C were attributed to the ammonium desorbed from weak acid sites. A second well-defined peak was observed at 460–475 °C in the MP and CS samples corresponding to the ammonium desorbed from strong acid sites [35–37]. The DP and CI samples presented peaks at 440 and 370 °C, respectively. The profile from the H-ferrierite zeolite is also shown in Fig. 5, and as hybrid catalysts, the H-ferrierite zeolite presented the same two peaks with higher intensity, but they are situated at the higher temperatures of 295 and 550 °C. As the catalyst weight used for TPD analysis was normalized to have the same zeolite weight as that used for pure zeolite, the hybrid catalysts presented lower acidity than the pure zeolite. This decrease was more pronounced for the strong acid sites. The DP and CI samples showed a greater decrease in this type of site.

The quantification of acid sites is made by determining the peak area. Their strengths can be related to the desorption temperature. In Table 3, these values are presented, and the occurrence of weak acid sites in the catalysts can be ordered as PM > CI > CS > DP. For the relationship of strong acid sites, the order is PM = CS > DP > CI. This result can be explained by the lower degree of contact between the two components, MSC and zeolite, in the PM and CS samples when compared with DP and CI, demonstrating a greater preservation of the zeolite acidity.

Table 3 Acidity profile resulting from TPD-NH3 for the hybrid catalysts
Full size table

According to Mao et al. [38], metal oxides can react with the acidic bridging hydroxyl groups of the zeolite when impregnated. This process reduces the number of Brønsted (i.e., strong) acid sites and increases the number of Lewis (i.e., weak and/or medium strength) acid sites by forming the chemical species M(OH)+, where M corresponds to the metal oxide. Similar results were found by Khandhan et al. [39]. A certain amount of the metal oxides over the zeolite can also block the acid sites of the zeolite, as found in previous works [40]. This effect can be more intense in processes in which a precipitate is formed over the zeolite surface.

3.5 FTIR with d3-Acetonitrile

To identify the nature of acid sites of zeolite d3-acetonitrile was used as probe molecule. The typical probe molecule used to determine acid sites by infrared spectroscopy is pyridine, however previous work demonstrated d3-acetonitrile to be more suitable to determine Bronsted acid sites in H-ferrierite zeolite than pyridine molecule, [40, 41]. The adsorption of d3-acetonitrile on the hybrid catalysts presented different spectra depending on the preparation method. The PM and CS samples showed an intense band at 2325 cm−1 ascribed to strong Lewis acid sites, Fig. 6. A very intense band was also observed at approximately 2275–2276 cm−1 for the CI and DP samples corresponding to the interaction of the C≡N group from the physisorbed acetonitrile with either Si–OH or defect sites [42, 43]. Brønsted acid sites were identified by the deconvolution of the spectra as a small band corresponding to the d3-acetonitrile interaction with these sites at 2290–2296 cm−1 only for the MP and CS samples, with the MP sample demonstrating greater intensity.

Fig. 6
figure 6

Deconvolution of the IR spectra of adsorbed d3-acetonitrile on the hybrid catalysts

Full size image

The Brønsted acid sites were almost eliminated on the hybrid catalysts, with mainly the Lewis acid sites remaining on them. In general, the PM and CS samples showed a higher acidity than the CI and DP samples because the hybrid catalyst components (the MSC and zeolite) are not in intimate contact and only have a minor interaction between them. These results confirmed the reduction in acidity on the hybrid catalyst found by NH3-TPD and the presence of a higher remaining strong acidity in the PM and CS samples.

3.6 Catalytic Activity

The results of catalytic activity are presented in Table 3. A higher CO conversion was found for the CS sample indicating a better catalytic performance by the sample prepared by coprecipitation sedimentation. Samples prepared by impregnation and deposition coprecipitation showed the lowest activity. The observed values decreased in the order CS > PM > DP > CI.

Among the products, the HCs and CO2 were the most relevant in all samples (Table 4). A higher yield of these products was found for the CS sample. Methanol was not found among the products, indicating that the entire amount of methanol produced was converted to other products.

Table 4 Reaction performance of hybrid catalysts prepared by different methods
Full size table

The preparation method of the hybrid catalyst slightly influenced the distribution of hydrocarbons (Table 4). Among the hydrocarbons formed, ethane was the main product followed by propane and methane. Propane and butane, the principal components of liquefied natural gas (LNG), were produced in higher amounts for the CS sample. The distribution of HCs was different than that for the Fischer–Tropsch synthesis and consequently, did not follow the Anderson–Schulz–Flory distribution.

According to Zhang et al. [5], high temperatures can promote methane and ethane formation because the methanol equilibrium yield is not favored, and the rate of methanol formation becomes the controlling step for hydrocarbon synthesis from syngas over hybrid catalysts. Low methanol and DME concentrations, as intermediates from syngas transformations, would restrict the rate of hydrocarbon formation. Asami et al. [18] stated that the zeolite acidity and pore size would be responsible for the hydrocarbon distribution. Conversely, Mysov et al. [44] demonstrated that the hydrocarbon distribution depends mainly on the molar ratio H2/CO and GHSV.

4 Discussion

The syngas transformation over a hybrid catalyst based on an MSC and an acid material proceeds via methanol formation. This means that the conversion of CO must be related to the MSC catalyst and consequently, to the Cu surface area. A plot of CO conversion versus the Cu surface area showed that in exception of the DP catalyst, an almost linear correlation exists between these two properties in catalysts, as shown in Fig. 7, which is in agreement with other authors [10, 45, 46]. An increase in the Cu surface area produces an increase in CO conversion. However, this increase was not proportional to the CO conversion as found in a previous work [47]. Therefore, other active sites must be considered as described by Fichtl et al. [48]. The prepared samples showed different Cu surface areas. This behavior was ascribed to the preparation of the hybrid catalyst because the amounts of Cu and the MSC precursor in all the samples were similar. The highest Cu surface area, which was found in the CS sample, can be attributed to better CuO dispersion produced by the wet mixture of MSC and the zeolite support that did not favor an intimate contact between them, as shown in Fig. 8a, but rather proportioned a higher MSC dispersion over the zeolite after calcination (dark particles over a lighter particle). In the PM sample, the MSC was not well dispersed over the zeolite support, and some agglomerates were observed over the zeolite surface; thus, the Cu surface area was lower, as shown in Fig. 8b. This behavior was also observed in the DP sample, as shown in Fig. 8c. Conversely, amorphous material over the zeolitic component was observed in the CI sample; this finding was ascribed to the MSC formed that was poorly dispersed over the zeolite, as shown in Fig. 8d. These results confirm the influence of the preparation of hybrid catalysts on the dispersion of the MSC on the support.

Fig. 7
figure 7

Copper metal surface areas versus conversion

Full size image
Fig. 8
figure 8

TEM images of the hybrid catalyst. a PM, b CS, c CI and d DP samples

Full size image

The activity results did not show methanol as a reaction product indicating that the methanol produced in the first stage was completely transformed into other products, such as DME or hydrocarbons [10, 35]. The methanol conversion into dimethyl ether and then into hydrocarbons happens over the zeolite. An effective methanol conversion to hydrocarbons over hybrid catalysts can be favored by a high concentration of strong acid sites. Figure 9 shows what seems to be a relationship between the production of hydrocarbons and the amount of strong acid sites. An increase in the amount of acid sites showed an increase in the production of hydrocarbons. In Fig. 9 was observed that the DME production was smaller in samples where the hydrocarbon production was higher. From these results it is possible to assert that exist some correlation between the low DME production and the higher hydrocarbon production observed for these catalysts.

Fig. 9
figure 9

Strong acid sites versus yield of products. Reaction conditions: 350 °C; 2 MPa; GSHV = 79 h−1; 20 mL min−1; feed gas: H2/CO = 2/1; Cu/ZnO/Al2O3 (500 mg); H2/CO

Full size image

The availability of acid sites can be favored by a high dispersion of the MSC over the zeolite, [44]. To have both active sites (acid and metallic), the MSC must be dispersed over the zeolite surface without fully blocking the acid sites. Depending on the preparation method, the blocking of acid sites can be less significant. The different amounts of acid sites in samples prepared by different methods (Table 3) can be initially attributed to the partial blockage and low dispersion of the MSC over the zeolite. The PM and CS samples showed a higher acidity and higher amounts of strong acid sites because the contact between both components in these catalysts was lower. Consequently, they provide a hybrid catalyst with more acid sites. In the CS sample, the MSC precursor previously formed is mixed with the zeolite using a wet process and is deposited over its surface. This allows contact between both components. After calcination, the MSC precursor structure collapses, and nanoparticles of CuO–ZnO–Al2O3 are formed and are well-dispersed over the zeolite surface, thus preserving its textural properties. In the MP sample, the calcined MSC and zeolite are mixed physically, and the contact between them is lower.

The low acidity observed in the CI sample (Table 3) would be caused by the smaller MSC dispersion and also by the greater blockage of the area and zeolite micropores by the MSC. In this sample, the MSC precursor is formed during the addition of the solutions of the metals and the precipitating agent over a zeolite suspension. The zeolite surface acts as a seed favoring the formation of a precipitate and a greater blockage of the acid sites by the precipitate. After calcination, the metal oxides remain in intimate contact with the zeolite as can be observed in Fig. 8d. The zeolite surface loses part of its superficial acidity while the MSC loses part of its metallic surface as observed by the low dispersion (Table 1). In Fig. 8d, the image of the calcined CI sample shows the MSC deposited over the H-ferrierite zeolite. In the DP sample, precipitates of the MSC are immediately formed and deposited over the zeolite surface, growing up over it. Also, part of this precipitate may be formed outside the zeolite surface explaining the presence of strong acid sites on this sample. Moreover, the CI and DP samples were prepared by a process in which the coprecipitation was not completely controlled, and the exchange of the zeolite proton by Na+ or Cu2+ ions is favored as shown in Table 1 by the amount of Na [49]. A slight shift of the Lewis band toward lower wavenumbers (from 1450 to 1446 cm−1, the minimum band shift considered valid due to the measurement resolution of 4 cm−1) in the hybrid catalysts indicated the formation of new Lewis acid sites produced by the exchange of zeolite protons with Cu2+, Zn2+ or Na+ during the catalyst preparation, Fig. 10.

Fig. 10
figure 10

FTIR spectra of hybrid catalysts after pyridine adsorption

Full size image

Finally, there exists an aspect that also influenced the activity performance of the hybrid catalyst. This is the contact between both components, and the effect of the contact can be observed in Fig. 10. The PM and CS samples showed similar acidity, however, the CS sample showed a higher production of hydrocarbons. This result can be related to the intimate contact between the components of the hybrid catalyst. In the CS sample, the MSC is closer to the acid sites, and this fact favored the conversion into other products of the methanol formed on the MSC. On the other hand, although the physical mixture method preserved the textural properties and acid sites in the PM sample more than in coprecipitation-sedimentation, the contact between the active sites of each reaction stage did not favor a higher performance of the hybrid catalyst.

From these results, it is possible to assert that the availability of active sites to convert syngas to methanol and then to hydrocarbons depends on the mixture of the MSC and zeolite components during the preparation of the hybrid catalyst. A high hybrid catalyst performance requires a high Cu surface area and a sufficient amount of strong acid sites that can be obtained by the combination of three effects: (i) a high dispersion of the MSC on the zeolite, (ii) a low partial blockage of the acid sites by the MSC, (iii) a low proton exchange in the ferrierite zeolite during the coprecipitation process, and (iv) a suitable proximity between both active sites responsible for each reaction stage.

5 Conclusions

The preparation method of hybrid catalysts composed of MSC and zeolite for the direct transformation of synthesis gas into hydrocarbons was verified, and DME affected both the Cu surface sites and the amount of strong acid sites. The preparation methods used in this work favored different degrees of interaction between both components in the hybrid catalyst.

The coprecipitation-sedimentation method favored a higher Cu dispersion over the zeolite, thus preserving the acidity of the zeolite. Samples prepared by this method also showed a contact level between both components that allowed a suitable proximity between the active sites to transform the syngas into hydrocarbons. Conversely, the higher deposition on the MSC over the zeolite observed in the CI sample reduced the active sites of each component, thus decreasing the catalytic activity.

A high Cu surface area and enough available strong acid sites were the main factors required to achieve higher activity in these hybrid catalysts.