Continuous synthesis of Cu/ZnO/Al₂O₃ nanoparticles in a co-precipitation reaction using a silicon based microfluidic reactor

Abstract

CuO/ZnO/Al₂O₃ catalysts were continuously synthesized in a microfluidic reactor, analyzed by X-ray diffraction (XRD), physisorption (BET), chemisorption, electron microscopy and X-ray absorption spectroscopy (XAS), and tested for methanol synthesis from CO-rich synthesis gas. The results were compared to those obtained from CuO/ZnO and CuO/ZnO/Al₂O₃ produced by conventional co-precipitation in a batch reactor. The predominant phase of the aged precursor from microfluidic co-precipitation was identified as zincian malachite. After calcination the microfluidically synthesized catalyst exhibited smaller CuO crystallites, a larger BET surface area, a rather uniform morphology and a homogeneous distribution of Cu and Zn compared to catalysts prepared by batch co-precipitation. H₂-Temperature programmed reduction (TPR) showed that Cu species in CuO/ZnO/Al₂O₃ from microfluidic co-precipitation were more easily reducible. In situ Cu and Zn K-edge XAS during the TPR indicated reduction of Cu²⁺ to Cu⁰ between 150 °C and 240 °C, without detectable reduction of Zn. N₂O pulse chemisorption evidenced an enlarged Cu surface area of the nanoparticles from the microfluidic synthesis. Based on activity tests in methanol synthesis, at 250 °C the microfluidically synthesized Cu/ZnO/Al₂O₃ catalysts showed better performance than the catalyst from batch preparation when 1 mol% CO₂ was present in the synthesis gas. Dimethyl ether formed as a side product. As the microreactor is specially designed for high X-ray transmission with a thin Si/glass observation window, this study opens interesting perspectives for investigating the formation of catalyst precursors at the early stage of precipitation in future.

---

Introduction

Methanol (MeOH) is not only an essential industrial bulk chemical but can also be used as a fuel additive or precursor for clean fuels.^1–4 Due to its high energy density methanol plays an important role in chemical energy and hydrogen storage including fuel-cell applications.^5–8 The most prominent catalytic system for industrial methanol synthesis since decades ago is based on Cu/ZnO/(Al₂O₃) and enormous efforts have been dedicated to the enhancement of its catalytic activity based on rational design via different preparation methods.^3,5,9–19 One widely applied method for the synthesis of Cu/ZnO/(Al₂O₃) is co-precipitation of soluble copper, zinc and aluminum precursors using e.g. sodium carbonate.^9,20–22

Specific synthesis parameters such as pH, temperature, ageing time and mixing conditions play key roles in the resulting catalytic performance.^5,9,21,23 Fundamental studies in this field e.g. by Behrens et al.²⁰ show that co-precipitation temperature between 60 °C and 70 °C and pH values between 6–7 as well as the ageing time and temperature are crucial to achieve optimum material properties. Sufficient ageing time is required for meso-structuring of the precipitate.^9,24 These parameters directly influence the formation of the – in this context important – zincian malachite (Cu_xZn_y)₂(OH)₂CO₃ phase. Calcination, typically in the range of 320 °C to 430 °C, promotes further nano-structuring during the phase transition to the mixed metal oxide.^20,25 Most preparation methods focus on batch procedures. However, for larger quantities and industrial applications a semi-batch or continuous process would be advantageous.^26–29

Catalytic performance of Cu/ZnO/(Al₂O₃) is particularly influenced by the surface area of copper (SA_Cu) and the area of reactive interface with ZnO in a nanoparticulate and porous structure.³⁰ Metallic Cu clusters/particles are considered to be the most active sites for methanol synthesis, strongly improved by ZnO,^2,25 and Al₂O₃ as a structural promoter^25,29 reducing NP sintering. Fast and homogeneous mixing of the soluble metal precursors and the precipitating agent are essential for the most uniform precipitate.²⁷ This is difficult to obtain in conventional stirred batch reactors where initial precipitation may already start before the reactants have been efficiently mixed, resulting in spatial and temporal material gradients. Every single added drop of reactants induces concentration changes during the entire process time. Redissolution, reprecipitation and ion exchange during ageing, washing and drying of the catalyst precursor may additionally influence the catalytic properties.^20,25,29 In contrast, during co-precipitation in a continuous microflow²⁹ the precipitates are formed in a small volume.^27,31 This results in defined and stable preparation conditions, once they are established. In the last decade, micromixing techniques have been increasingly applied for precipitation reactions.^26,27,29,32 Advanced micromixing devices such as microfluidic reactors, T-mixers and confined impinging jet reactors are designed for fast and homogeneous mixing based on high mass transfer and short residence time.^27–29

In earlier studies, we used a novel microfluidic reactor especially designed for X-ray spectroscopic in situ studies to investigate colloidal noble metal NP formation during fast reduction reactions, specifically focusing on the critical early stages of the process.^33–36 The experiments described here aimed at extending the application range of this type of microfluidic device towards co-precipitation reactions, focusing on Cu/ZnO/Al₂O₃ catalysts for methanol synthesis as a case study. For this purpose, two streams of the reactants were injected separately into the micromixers integrated in the microfluidic chip. For comparison, Cu/ZnO and Cu/ZnO/Al₂O₃ catalysts from conventional batch synthesis under conditions adopted from the literature were also characterized and tested in order to illustrate potential advantages of microfluidic mixing during co-precipitation.

Materials and methods

Materials

Copper(II) nitrate trihydrate (Cu(NO₃)₂·3H₂O, Sigma-Aldrich, >99% purity), zinc nitrate hexahydrate (Zn(NO₃)·6H₂O, Alfa Aesar, 99.99% purity), zinc oxide (ZnO, Fluka, >99% purity), aluminum nitrate nonahydrate (Al(NO₃)₃·9H₂O, Sigma-Aldrich, >98% purity), sodium carbonate (Na₂CO₃, Sigma-Aldrich, >99.95% purity) and nitric acid (HNO₃, Merck, 65% solution) were used without further purification.

Batch synthesis of Cu/ZnO and Cu/ZnO/Al₂O₃

Throughout the manuscript we denote all catalysts “Cu/ZnO/(Al₂O₃)-BR/MF”, irrespective of the copper phases present, where BR stands for batch reactor and MF for microfluidic synthesis.

Nitrate-derived Cu/ZnO and Cu/ZnO/Al₂O₃ reference catalysts were synthesized via co-precipitation in a batch reactor. For this purpose, a procedure adopted from Behrens et al.²⁰ was applied. 21.6 mmol Cu(NO₃)₂·3H₂O, 9.6 mmol ZnO and 4.4 mmol Al(NO₃)₃ were dissolved in 38 mL distilled water with 2.5 mL of concentrated nitric acid. 23 mL of 1.6 M aqueous solution of Na₂CO₃ was used as a precipitating agent. The two solutions were injected dropwise into the batch reactor using individual syringe pumps (neMESYS, Cetoni GmbH, with 50 mL glass syringes). The flow rates for injecting metal precursor and precipitating agent solutions were 1 and 0.56 mL min⁻¹, respectively. The batch reactor used was a round-bottom flask containing 100 mL distilled water stirred at ∼65 °C. The pH of the precipitate was kept constant at 6–7. Subsequently, the precipitate was aged for 3 h, filtered and washed several times. This process was followed by drying overnight at 70 °C and calcination at 330 °C for 3 h. This procedure resulted in 2 g of catalyst. For synthesizing Cu/ZnO the same procedure was applied, cf.Table 1.

Table 1 Preparation of the reactants for co-precipitation of Cu/ZnO/(Al₂O₃)

Catalyst	Synthesis technique	Metal precursor (mmol)			Na2CO3 (mmol)
		Cu^2+	Zn^2+	Al^3+
The amount of distilled water for dissolving the metal precursors and precipitating agent for batch reactor synthesis was 38 and 23 mL, and for microfluidic synthesis 40 and 40 mL, respectively.
Cu/ZnO	Batch reactor	21.6	9.6	—	36.8
Cu/ZnO/Al2O3	Batch reactor	21.6	9.6	4.4	36.8
Cu/ZnO/Al2O3	Microfluidic reactor	7.2	3.6	1.2	14.4

---

Microfluidic synthesis of Cu/ZnO/Al₂O₃

A nitrate-derived ternary Cu/ZnO/Al₂O₃ catalyst was prepared by co-precipitation of aqueous metal nitrate solution with Na₂CO₃ in a microfluidic chip (Fig. 1) fabricated in collaboration with GeSiM GmbH and the Institute of Semiconductors and Microsystems at Technische Universität Dresden (IHM TUD). It is made of Si-bonded glass and was specifically designed for in situ characterization using X-ray spectroscopic and scattering techniques.³⁶ A metal nitrate solution with a molar Cu : Zn : Al ratio of 60 : 30 : 10 was prepared by dissolving the precursors in 40 ml distilled water (total metal ion concentration 0.3 M). 1 ml concentrated nitric acid was added to the solution resulting in pH 1. The Na₂CO₃ concentration was 0.36 M. To avoid heavy deposition of precipitates in the channel and clogging, the precursors were diluted. However, the metal ratio was kept at the same level as for the batch synthesis. Each of the two reactant solutions (40 mL) was loaded in a 50 mL glass syringe and injected into the microfluidic chip at a flow rate of 22 mL min⁻¹ (neMESYS syringe pump, Cetoni GmbH), and the precipitate was transferred into a flask containing 100 mL distilled water under stirring at 65 °C. After ∼15 min a color change from blue to turquoise was observed. The pH of the precipitate was constant at 6–7, and the same ageing, washing, drying and calcination conditions as in the previous section were applied. This procedure provided 0.8 g catalyst.

Fig. 1 (a) Microfluidic synthesis of Cu/ZnO/Al₂O₃via co-precipitation at constant pH and a total flow rate of 44 mL min⁻¹, (b–e) color change during precipitation from 10 to 81 min.

Catalytic testing: methanol synthesis

MeOH synthesis was performed using the CuO/ZnO and CuO/ZnO/Al₂O₃ pre-catalysts. The pre-catalysts were pressed and sieved into fractions of 250–500 μm particle size before use. To minimize hot spot formation and potential sintering, the pre-catalyst was mixed with SiC (Mineraliengrosshandel Hausen GmbH, Telfs, Austria) in a mass ratio of 1 : 5. The reaction was performed in a continuous-flow, stainless-steel fixed bed reactor (460 mm length, 12 mm inner diameter) which was filled with 5 g of the pre-catalyst/SiC mixture. The temperature was measured at the outside of the reactor wall, details about the equipment are reported elsewhere.³⁷ Prior to the experiments, pre-catalysts were heated to 100 °C in Ar for 1 h. Catalyst reduction was performed by heating to 200 °C in 5% H₂ in Ar with a heating rate of 0.33 °C min⁻¹, followed first by increasing the H₂ content to 50% and then the temperature to 240 °C with a heating rate of 0.2 °C min⁻¹. These conditions were kept constant for 5 h. Every catalyst was tested at 50 bar, 230 °C and 250 °C, with two feed gas compositions (H₂/N₂/CO/CO₂/Ar: 34 : 15 : 15 : 0 : 36 and 34 : 15 : 14 : 1 : 36, respectively) and gas hourly space velocity (GHSV) of 24 000 mL h⁻¹ g⁻¹ to ensure conversion in the kinetic region. Further details on feed compositions and data recording during time on stream (ToS) are provided in Table S1.† Reaction products were analyzed using gas chromatography on an Agilent G1530A equipped with RT®-Msieve 5A and RT®-U-BOND columns, a thermal conductivity detector (TCD) and a flame ionization detector (FID). The performance indicators productivity and selectivity were calculated using the equations provided in the ESI.†

Characterization methods

X-ray diffraction (XRD). X-ray diffraction patterns of the pre-catalysts were acquired using a D8 Advance diffractometer (Bruker) ex situ in a 2θ range of 10–80° (2 s per step, step size 0.016°) using Cu-K_α radiation (λ = 0.15406 nm), a nickel filter and a graphite monochromator.

Inductively coupled plasma-optical emission spectroscopy (ICP-OES). The weight loadings of Al, Cu and Zn were determined by ICP-OES spectroscopy (Optima 4300 DV, PerkinElmer). 10 mg of the sample (accuracy ± 0.01 mg) was dissolved in 4 ml hydrochloric acid and 4 ml sulfuric acid at 240 °C for 6 h in the pressure digestion vessel DAB-2 (Berghof). Elemental analysis was performed with four different calibration solutions and an internal standard (Sc) using three wavelengths of each chemical element for calculation.

Physisorption (BET surface area). The specific surface area of the catalysts was determined by nitrogen physisorption at −196 °C on a Belsorp mini II apparatus using the Brunauer–Emmett–Teller (BET) method. Prior to the measurements the samples were degassed at 300 °C in vacuum.

Temperature programmed reduction in H₂ (H₂-TPR) and chemisorption. For chemisorption analysis, a Micromeritics AutoChem 2950 HP equipped with a thermal conductivity detector (TCD) and an MKS Cirrus 2 mass spectrometer (MS) was used. 160 mg of a sample were placed in a U-shaped quartz glass reactor between two quartz wool plugs. The sample was dried under Ar flow of 30 mL min⁻¹in situ by heating it from room temperature to 250 °C at 5 °C min⁻¹ and keeping the temperature constant at 250 °C for 2 h.

Afterwards, the sample was cooled down to 50 °C and temperature-programmed reduction (TPR) at ambient pressure in a flow of 5% H₂ in Ar (30 mL min⁻¹) through the sample was started. The temperature was increased at a rate of 1 °C min⁻¹ up to 250 °C and kept at 250 °C for 30 min. The effluent gas passed through a cold trap at 0 °C to condense any water formed. The dry gas was then analyzed by the TCD and the MS in series. The H₂ signal of the MS (m/z = 2) was calibrated with 5% H₂ in Ar before and after each measurement and used for quantification of the overall H₂ consumption during TPR.

Subsequently, N₂O pulse chemisorption was performed. The sample was cooled to room temperature and flushed with 30 mL min⁻¹ He for 30 min. Afterwards, 40 pulses with 0.32 mL of 30% N₂O in He were dosed onto the sample. The effluent gas from the sample passed through a cooling trap at −196 °C to separate any N₂O not converted to N₂. N₂ formed by partial Cu oxidation was then analyzed with the TCD and the MS in series. The TCD data were used to determine the metallic Cu surface area and Cu dispersion (Cu atomic area 0.068 nm², stoichiometry factor 2.0).

X-ray absorption spectroscopy (XAS). XAS measurements were performed at the CAT-ACT beamline at the KIT synchrotron light source.³⁸ CAT-ACT's superconducting multipole wiggler and a double crystal Si (111) monochromator provide high intensity (flux ∼ 5×10¹⁰ photons per s/100 mA) at the Cu K-edge energy. Each catalyst was diluted with Al₂O₃ in order to optimize the absorption step. Finely pressed and sieved (100–200 μm) catalysts were loaded in 1.5 mm quartz capillaries (20 μm wall thickness, catalyst bed length 6 mm) which serve as in situ reactor cells. The capillaries were carefully leak-tested by flowing He gas. TPR was performed in a temperature range between 20 °C and 260 °C (heating rate 1 °C min⁻¹) in a 5% H₂ in He flow (50 mL min⁻¹). X-ray absorption near edge structure (XANES) scans at the Cu and Zn K-edges were recorded continuously during the mentioned TPR conditions and cooling down. Extended X-ray absorption fine structure (EXAFS) scans were recorded at 20 °C and 260 °C. For energy calibration at the respective edges, spectra of Cu metal foil and ZnO were recorded simultaneously with the samples.

EXAFS data analysis was performed using the software packages Athena and Artemis.³⁹Athena was first used to process the raw data (absorption coefficient μ(E) vs. E) by background subtraction, normalization, converting μ(E) data to χ(k) data and Fourier transforming the resulting spectrum from k-space to R-space. For the EXAFS background extraction, the input parameter R_bkg, was set to 1.0 Å. EXAFS fitting was performed using Artemis by generating theoretical models from available crystallographic data of reference compounds, i.e. CuO and Cu. The models were fitted to the experimental data in k-space and R-space to determine energy shifts (ΔE₀) and structural parameters, including changes in the scattering path length (ΔR), passive electron reduction factor (S₀²), coordination number (N) and relative mean-square displacement of the atoms (Debye–Waller factor, σ²).

In situ XANES spectra were analyzed by linear combination fitting (LCF) using the spectra measured at 20 °C and 260 °C as internal references.

Transmission electron microscopy (TEM). The catalyst powder samples were directly dispersed on Cu grids covered with holey carbon film. Morphology and microstructure of the catalysts were characterized by high angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) and high resolution transmission electron microscopy (HRTEM), and their composition was analyzed by energy dispersive X-ray analysis (EDX) using an EDAX S-UTW EDX detector in a FEI Titan 80-300 microscope operating at 300 kV at the Karlsruhe Nano Micro Facility (KNMF). Analysis of STEM-EDX maps was carried out by using the TEM Image & Analysis (TIA 4.7 SP3) software.

Results and discussion

Comparison of Cu/ZnO/(Al₂O₃) produced in the microfluidic and batch reactor

The ternary Cu, Zn, (Al) hydroxycarbonate precursor materials were co-precipitated at constant pH in the magnetically stirred-batch reactor and the microfluidic reactor and then, after ageing, washing and drying, analyzed by XRD. In both cases the acquired XRD patterns clearly show reflections of zincian malachite [(Cu_1−xZn_x)₂(OH)₂CO₃] with x < 0.3 (Fig. 2a). The XRD data of the precipitates from the batch reactor show additional reflections attributed to gerhardtite [Cu₂(OH)₃NO₃],²⁰ whereas the diffraction patterns of the precipitates from the microfluidic reactor indicate presence of rosasite phase with poor crystallinity.²⁹ The formation of different phases is attributed to the efficient micromixing effect (homogeneous mixing in short time) and the spatially controlled nucleation of the primary precipitates in small volumes flowing in the microchannel.

Fig. 2 XRD patterns of the (a) uncalcined ternary CuZnAl precipitates prepared by different methods indicating formation of malachite/rosasite and malachite/gerhardtite phases in precipitates synthesized in (I) microfluidic and (II) batch reactor, respectively; (b) calcined (III) Cu/ZnO/Al₂O₃-MF, (IV) Cu/ZnO/Al₂O₃-BR and (V) Cu/ZnO-BR.

According to literature,^20,25,29 the formation of different crystalline phases during co-precipitation is a dynamic process, during or at the end of which other processes such as partial dissolution and reprecipitation can occur. It was reported earlier that such exchange reactions between malachite [Cu₂(OH)₂CO₃] and hydrozincite [Zn₅(OH)₆(CO₃)₂] lead to rosasite [(Cu, Zn)₂(OH)₂CO₃] and aurichalcite [(Cu, Zn)₅(OH)₆(CO₃)₂].^21,25,40,41 During the precipitation process the following main reactions occur:

2Cu²⁺ + CO₃²⁻ + 2OH⁻ → Cu₂(CO₃)(OH)₂ (1)

5Zn²⁺ + 2CO₃²⁻ + 6OH⁻ → Zn₅(CO₃)₂(OH)₆ (2)

2Cu²⁺ + NO₃⁻ + 3OH⁻ → Cu₂(NO₃)(OH)₃ (3)

During ageing, the following exchange, dissolution and reprecipitation proceed:

Cu₂(NO₃)(OH)₃ + CO₃²⁻ → Cu₂(CO₃)(OH)₂ + NO₃⁻ + OH⁻ (4)

Cu₂(CO₃)(OH)₂ + xZn²⁺ → (Cu_2−x, Zn_x)(CO₃)(OH)₂ + xCu²⁺ (5)

Zn₅(CO₃)₂(OH)₆ + xCu²⁺ → (Zn_5−x, Cu_x)(CO₃)₂(OH)₆ + xZn²⁺ (6)

During the following thermal treatment, rosasite and aurichalcite can decompose into materials with uniform distribution of Cu and Zn hydroxycarbonates and higher catalytic activity.

The pre-catalysts, including the Cu/ZnO-BR, calcined at 330 °C were also investigated by XRD. Reflections of Al₂O₃ were not observed for the Al containing materials due to its amorphous phase structure.^24,42 The results in Fig. 2b show the dominant CuO and ZnO phases in all three samples. In the case of Cu/ZnO/Al₂O₃-MF catalyst, the reflections are broader and less intense. The CuO reflection at 2θ = 38.8° in the XRD patterns of Cu/ZnO/Al₂O₃-MF was weaker compared to the Cu/ZnO/Al₂O₃-BR indicating smaller CuO crystallites. In the literature, it was repeatedly reported that small particle sizes and well-distributed copper clusters are important for the catalytic performance (activity and selectivity) in methanol synthesis via hydrogenation of CO/CO₂.^25,29 Thus, better catalytic performance is expected from Cu/ZnO/Al₂O₃-MF.

Table 2 summarizes the main textural characteristics of the Cu/ZnO/(Al₂O₃) samples prepared by different methods. Based on the ICP-OES results, the molar Cu : Zn : Al ratios in Cu/ZnO/Al₂O₃ catalysts produced in the microfluidic and batch reactors are similar (77 : 6 : 17 and 76 : 8 : 16, respectively). This enables sufficient comparability of their catalytic performance. In the case of Cu/ZnO-BR catalyst a molar Cu : Zn ratio of 77 : 23 was obtained.

Table 2 Comparison of calcined Cu/ZnO and Cu/ZnO/Al₂O₃ produced in the batch and microreactor

Catalyst	Synthesis technique	ICP-OES wt%			BET m^2 g^−1	N2O chemisorption
		Cu	Zn	Al		Metal dispersion%	SACu^a m^2 g^−1	Active particle diameter nm
a Per gram metal.
Cu/ZnO	Batch reactor	58.9	18.6	—	29	1.6	10.3	65
Cu/ZnO/Al2O3	Batch reactor	59.0	6.5	5.2	36	1.5	9.5	71
Cu/ZnO/Al2O3	Microreactor	59.2	4.5	5.2	49	2.6	17.0	39.5

---

BET surface areas of Cu/ZnO/Al₂O₃ synthesized in the batch reactor and the microreactor were 36 m² g⁻¹ and 49 m² g⁻¹, respectively. Cu/ZnO-BR had the lowest BET surface area (29 m² g⁻¹), which could be due to the absence of alumina.

H₂-TPR was applied to study the reduction of copper entities in the three calcined catalysts (Fig. 3). The Cu/ZnO/Al₂O₃-BR reduction peak is slightly broader with its maximum at higher temperature compared to the Cu/ZnO/Al₂O₃-MF TPR profile. This indicates a narrower size distribution²⁹ of CuO particles synthesized in the microreactor. Moreover, in the TPR data of the batch synthesized Cu/ZnO-BR and Cu/ZnO/Al₂O₃-BR a pronounced shoulder at 136 °C and 164 °C, respectively, suggests various CuO entities. N₂O pulse chemisorption (Table 2) shows a smaller Cu particle diameter, higher metallic surface area and higher metal dispersion for the Cu/ZnO/Al₂O₃-MF, which is beneficial for improved catalytic activity.

Fig. 3 TPR profiles of the calcined catalysts prepared in the batch and microfluidic reactor.

Finally, the samples were studied by transmission electron microscopy combined with EDX mapping. Fig. 4 shows TEM images of calcined Cu/ZnO produced in the batch reactor. STEM measurements combined with EDX mapping (Fig. 5a, S1 and S4a and Table S2†) on selected nanoparticles indicate variations in the Cu : Zn ratio and a non-uniform distribution of CuO and ZnO in the sample. The morphology and the EDX mapping of different regions of Cu/ZnO/Al₂O₃ nanoparticles produced in the batch reactor are shown in Fig. 5b, S2 and S4b and Table S3.† They reveal uniform distributions of O, Al, Cu and Zn as well as relatively uniform Cu : Zn ratio in the nanoparticles of this catalyst. This more uniform distribution compared to the Cu/ZnO-BR sample is attributed to the presence of alumina which prevents random sintering of nanoparticles.²⁵ Interestingly, metal distribution in the calcined Cu/ZnO/Al₂O₃-MF is somewhat more uniform (lower variance in Cu : Zn ratios at different positions) and the morphology looks more finely structured compared to nanoparticles produced in the batch reactor (Fig. 5c, S3 and S4c and Table S4†).

Fig. 4 TEM images of calcined Cu/ZnO nanoparticles produced in the batch reactor.

Fig. 5 STEM images and the corresponding elemental maps obtained from STEM-EDX spectrum imaging in the areas marked by the orange box of calcined (a) Cu/ZnO, (b) Cu/ZnO/Al₂O₃ nanoparticles produced in the batch reactor and (c) Cu/ZnO/Al₂O₃ nanoparticles produced in the microfluidic reactor. Scale bars for the elemental maps are 10 nm (parts a and b) and 5 nm (part c).

X-ray absorption spectroscopy on Cu/ZnO/Al₂O₃ nanoparticles

Fig. 6a shows normalized XANES spectra at Cu K-edge at room temperature (RT) for the two Cu/ZnO/Al₂O₃ catalysts produced in microfluidic and batch reactors and Fig. 6b shows their corresponding Fourier transforms (FT). A very weak pre-edge feature at 8976 eV is present indicating Cu in +2 oxidation state. Another feature is the shoulder S attributed to square planar geometry⁴³ that has been found to be weaker in the spectrum of the microfluidically synthesized sample (MF) indicating higher geometrical distortion. The EXAFS analysis (Table 3) shows that Cu is coordinated to 2 O atoms with a Cu–O bond distance of 1.95 Å, similar to the typical Cu–O bond distance of 1.95 Å in bulk CuO. We note that XAS spectra measured in transmission geometry show considerable damping effect probably attributable to higher harmonics in the beam in combination with strongly absorbing samples. This results in decreased coordination numbers (the true CN of oxygen in the first coordination shell should be 4 for square planar geometry typical for Cu(II)) but does not influence positions of features in EXAFS and XANES. Hence, the EXAFS-derived CN values can serve only as a trend for each specific catalyst rather than to identify particle sizes. The second shell consisting of Cu/Zn atoms at 2.94 Å slightly differs from the typical Cu–Cu bond distances of 2.90 or 3.083 Å in tenorite. Thus, the sample prepared from microfluidic synthesis probably has almost the same Cu structure as the catalyst obtained from the batch reactor. Fig. 6c shows normalized XANES spectra at the Zn K-edge for the two catalysts at RT. The shape of the edge for both Cu/ZnO/Al₂O₃ samples is quite different from that of the ZnO reference. These XANES features correspond to a mixed aluminate phase in which Zn possesses octahedral geometry.⁴⁴

Fig. 6 (a) XANES spectra at Cu K-edge, (b) corresponding magnitude of the Fourier transformed k³-weighted EXAFS data at Cu K-edge, and (c) XANES spectra at Zn K-edge of Cu/ZnO/Al₂O₃ produced in microfluidic (MF) and batch reactor (BR).

Table 3 EXAFS fitting results (k³-weighted) obtained from Cu K-edge X-ray absorption spectra of Cu/ZnO/Al₂O₃ catalysts at room temperature. S₀² is fixed to 0.85 as determined from Cu metal foil. k range for FT is 2.4–10.0 Å⁻¹ and R range for fitting is 1–3.5 Å. ΔE₀ = 7.8 ± 1.9, χ_ν² = 113 for MF and ΔE₀ = 6.7 ± 2.1, χ_ν² = 206 for BR

As prepared	Synthesis technique	Cu–O shell			Cu–(Cu/Zn) shell
		R (Å)	CN^a	DW factor	R (Å)	CN^a	DW factor
a CN = coordination number (affected by spectral damping), R = distance (Å), σ^2 = Debye–Waller factor (DW) (Å^−2). b Fixed during fitting.
Cu/ZnO/Al2O3	Microreactor	1.96 ± 0.01	2.1 ± 0.4	0.0036 ± 0.0020	2.94 ± 0.043.11 ± 0.04	3.6 ± 0.82^b	0.0150 ± 0.0030^b
Cu/ZnO/Al2O3	Batch reactor	1.96 ± 0.01	1.9 ± 0.4	0.0042 ± 0.0022	2.93 ± 0.033.11 ± 0.03	3.4 ± 0.72^b	0.0150 ± 0.0030^b

---

Fig. 7a and b shows background subtracted and normalized XANES spectra at Cu K and Zn K-edges, respectively, measured in situ during the reduction of the catalyst produced in the microreactor. The reduction of Cu(II) to Cu(0) is reflected by the decrease in the intensity of white line and shift of the edge to lower energies starting at a temperature close to 150 °C and is almost completed at 240 °C. The XANES features of the reduced catalyst at 260 °C (Fig. S5†) closely resemble the profile of the copper foil (spectral damping taken into account), indicating that the reduction has proceeded to the apparent full reduction of Cu²⁺ into Cu⁰ metal. This is confirmed by a Cu–Cu distance corresponding to metallic Cu (Table 4). In this case, no contribution from Cu(I) was observed during the TPR which is similar to the results recently reported by Frei et al.,⁴⁵ however there are examples in literature where relatively low (<10%) contribution from Cu(I) were also reported.²³

Fig. 7 In situ XANES spectra at (a) Cu K and (b) Zn K-edges of Cu/ZnO/Al₂O₃ produced in the microreactor measured during the TPR (from RT up to 260 °C), along with (c) first derivative XANES spectra at Zn K-edge and (d) comparison of the LCF results at Cu K-edge during TPR for catalysts produced in microfluidic and batch reactor.

Table 4 EXAFS fitting results (k³-weighted) obtained from Cu K-edge X-ray absorption spectra of Cu/ZnO/Al₂O₃ catalysts during the TPR. S₀² is fixed to 0.85 as determined from Cu metal foil. k range for FT is 2.7–10.2 Å⁻¹ and R range for fitting is 1–5.2 Å (up to 4 shells). ΔE₀ = 3.7 ± 1.0, χ_ν² = 13 for MF at 260 °C and ΔE₀ = 5.2 ± 0.7, χ_ν² = 30 for MF at RT. ΔE₀ = 4.6 ± 0.7, χ_ν² = 6 for BR at 260 °C and ΔE₀ = 5.0 ± 0.7, χ_ν² = 40 for BR at RT

Synthesis technique	T (°C)	Cu–Cu shell			Cu–Cu shell (2nd shell)
		R (Å)	CN	DW factor	R (Å)	CN	DW factor
a Fixed during fitting.
Microreactor	At 260	2.53 ± 0.02	5.4 ± 0.6	0.0134 ± 0.0012	3.56 ± 0.05	2^a	0.0163 ± 0.0042
	After cooling down to RT	2.54 ± 0.01	5.7 ± 0.5	0.0082 ± 0.0007	3.58 ± 0.03	3^a	0.0118 ± 0.0015
Batch reactor	At 260	2.54 ± 0.01	5.0 ± 0.4	0.0135 ± 0.0009	3.57 ± 0.03	2^a	0.0152 ± 0.0016
	After cooling down to RT	2.54 ± 0.01	5.2 ± 0.4	0.0085 ± 0.0007	3.59 ± 0.02	4^a	0.0156 ± 0.0016

---

Fig. 7c shows the derivative XANES spectra at Zn K-edge during the same TPR to 260 °C. The changes observed in the Zn K-edge in Fig. 7b are weak, and to highlight these the spectra are shown as derivative. The reduction of the copper oxide phase that occurs between 150 and 240 °C does not strongly affect the Zn K-edge XANES profile, showing that the segregation and reduction of the Cu²⁺ to Cu⁰ metal particles is an independent phenomenon. The decrease in the white line intensity accompanied by a shift of the edge to lower energies have been interpreted as indicators of the transformation of Zn(II) from octahedral to tetrahedral coordination.^46,47 These effects are more evident in the corresponding first derivatives that display a regular shift in energy and decrease in the intensity of the Zn²⁺ 1s → 4p peak from ∼9662.5 eV to ∼9662.0 eV.⁴⁸ As the reduction temperature increases, the Zn containing phase is undergoing a regular transformation from the mixed aluminate phase in which Zn possesses an octahedral geometry to a highly dispersed and amorphous ZnO phase with Zn in tetrahedral coordination. Also, the appearance of the ∼9658 eV feature, typical of metallic Zn,⁴⁵ is a clear indication that Zn(0) formation is just initiated at the end of the TRP experiment. By comparing the results of LCF of Cu K XANES spectra of TPR obtained for the two catalysts (Fig. 7d) it can be observed that the reduction of Cu species from the batch reactor starts at lower temperature than of those from the microreactor.

Methanol synthesis

After activation, catalysts prepared by the batch and continuous microfluidic precipitation were compared in order to assess their catalytic performance during methanol synthesis from CO-rich syngas. Reaction conditions are summarized in Table S1.† The three main reactions are as follows:

CO hydrogenation to methanol

CO₂ hydrogenation to methanolWater–gas shift reaction (WGSR)Table 5 summarizes the results of the experiments performed at 230 °C and 250 °C, each with two different feed gas compositions at 50 bar and a GHSV of 24 000 mL g⁻¹ h⁻¹. In addition to the CO syngas feed, a second feed was used with a CO/CO₂ ratio of 14 : 1, as it is known that small quantities of CO₂ increase methanol productivity for Cu/ZnO-based catalysts.^49–52 As expected, in pure CO synthesis gas feed the methanol productivity of all catalysts increased by a factor of between 2.1 and 3.4 at 250 °C compared to 230 °C. The methanol productivity for the two feed compositions increased by a factor of 6.9–14.8. Note that this factor is the highest for the catalyst prepared by microfluidic synthesis. The results also show that Cu/ZnO/Al₂O₃-MF in pure CO synthesis gas feed yields dimethyl ether (DME) with selectivity of 14.1% at 230 °C and 16.9% at 250 °C. This may be attributed to higher surface area of the Cu/ZnO/Al₂O₃-MF catalyst which results in more available acidic sites catalyzing dehydration of methanol (reaction (10)). For technical methanol synthesis DME formation would require a purification step and should be avoided by e.g. further optimizing the microfluidic synthesis procedure to decrease the amount of acidic sites.

Table 5 Methanol selectivity (mol%) and productivity (mmol h⁻¹ g_cat⁻¹) at 50 bar and a GHSV of 24 000 mL_N h⁻¹ g⁻¹

Reaction conditions	Catalysts	S DME mol%	S MeOH mol%	P DME mmol h^−1 gcat^−1	P MeOH mmol h^−1 gcat^−1
230 °C	Cu/ZnO/Al2O3-MF	14.1	41.5	0.2	0.6
15 mol% CO	Cu/ZnO/Al2O3-BR	0.0	47.7	0.0	1.3
	Cu/ZnO-BR	0.0	34.7	0.0	0.7
230 °C	Cu/ZnO/Al2O3-MF	1.3	98.7	0.1	8.9
14 mol% CO	Cu/ZnO/Al2O3-BR	0.0	100	0.0	13.4
1 mol% CO2	Cu/ZnO-BR	0.0	100	0.0	8.2
250 °C	Cu/ZnO/Al2O3-MF	16.9	65.9	0.4	1.5
15 mol% CO	Cu/ZnO/Al2O3-BR	0.4	60.2	0.0	2.7
	Cu/ZnO-BR	0.0	58.3	0.0	2.4
250 °C	Cu/ZnO/Al2O3-MF	1.8	98.1	0.4	21.9
14 mol% CO	Cu/ZnO/Al2O3-BR	0.0	99.9	0.0	19.6
1 mol% CO2	Cu/ZnO-BR	0.1	99.8	0.0	16.6

---

MeOH dehydration

By adding a small amount of CO₂ to the feed (CO/CO₂ = 14 : 1) CO₂ formation via WGSR can be suppressed in the investigated operation range, leading to nearly 100% MeOH selectivity (see Table 5), also the water produced through reaction (8) inhibits the formation of DME. At 230 °C and in presence of 1 mol% CO₂, methanol productivity over the Cu/ZnO/Al₂O₃-BR is by a factor of 1.5 higher than over the Cu/ZnO/Al₂O₃-MF, whereas at 250 °C in the same gas mixture the catalyst from the microfluidic reactor produces >10% more methanol compared to its counterpart from the batch reactor.

Overall, the Cu/ZnO/Al₂O₃-MF catalyst also shows the highest observed methanol productivity (21.9 mmol h⁻¹ g_cat⁻¹ at 250 °C, CO/CO₂ = 14) in all experiments, suggesting that the microfluidic co-precipitation method has a promoting effect on catalytic activity under these conditions, while further optimization of the MF synthesis is needed to control the DME selectivity.

Conclusion

A continuously operated microfluidic reactor was used to synthesize a nitrate-derived Cu/ZnO/(Al₂O₃) catalyst for CO₂ dehydrogenation to methanol and DME. The obtained catalyst was thoroughly characterized by multiple physicochemical techniques and benchmarked against materials made via co-precipitation in a conventional stirred batch reactor using the same precursors and the same metal ratios. XRD measurements revealed that in both materials zincian malachite was the main phase of the precipitated precursor. Additionally, some gerhardtite and rosasite phases were observed in the Cu/ZnO/Al₂O₃ synthesized in the batch and microfluidic reactors, respectively. After calcination, the sample made via microfluidic synthesis had smaller CuO crystallites leading to higher dispersion and active surface area of Cu in the final catalyst.

The catalyst synthesized in the batch reactor demonstrates higher MeOH productivity at 230 °C, whereas at 250 °C the microfluidically synthesized catalyst provides higher MeOH and DME yields. Hence, the study shows that a continuous synthesis using a micromixer is an attractive process for preparation of highly active methanol synthesis catalyst, as it is well-controllable and scalable. The used microfluidic reactor with optically and X-ray transmissible windows, furthermore, offers a possibility to investigate the early stages of methanol catalyst formation during co-precipitation with a multitude of X-ray spectroscopic and scattering techniques.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The Virtual Institute VI-403 “In situ Nano Imaging of Biological and Chemical Processes”, the BMBF (projects 05K10VK1, 05K13VK2) and KIT (Helmholtz-programs “Storage and Crosslinked Infrastructures” (SCI) and “Materials and Technologies for the Energy Transition” (MTET)) are gratefully acknowledged for financial support. We would like to thank the Karlsruhe Nano Micro Facility (KNMF), a Helmholtz research infrastructure at KIT for providing STEM-EDX and ICP-OES measurements and Angela Deutsch (ITCP) for BET measurements. Special thanks to Simon Wodarz and Sabrina Polierer (IKFT) for the N₂O chemisorption experiments. Finally, we thank the KIT synchrotron (operated by KIT-IBPT) for providing beamtime at the CAT-ACT beamline, and Dr. Tim Prüßmann and Dr. Anna Zimina (IKFT) for their help and technical support during XAS experiments.

References

1. G. A. Olah, A. Goeppert and G. S. Prakash, Beyond Oil and Gas: The Methanol Economy, John Wiley & Sons, 2018.
2. M. Behrens, F. Studt, I. Kasatkin, S. Kühl, M. Hävecker, F. Abild-Pedersen, S. Zander, F. Girgsdies, P. Kurr and B.-L. Kniep, Science, 2012, 336, 893–897.
3. T. Lunkenbein, F. Girgsdies, T. Kandemir, N. Thomas, M. Behrens, R. Schlögl and E. Frei, Angew. Chem., Int. Ed., 2016, 55, 12708–12712.
4. G. Centi and S. Perathoner, Catal. Today, 2009, 148, 191–205.
5. J. B. Hansen and P. E. H. Nielsen, in Handbook of Heterogeneous Catalysis, ed. G. Ertl, H. Knözinger, F. Schüth and J. Weitkamp, 2nd edn, Wiley-VCH, Weinheim, 2008.
6. J.-K. Lee, I.-B. Lee and J. Han, J. Ind. Eng. Chem., 2019, 75, 77–85.
7. J. Artz, T. E. Müller, K. Thenert, J. Kleinekorte, R. Meys, A. Sternberg, A. Bardow and W. Leitner, Chem. Rev., 2018, 118, 434–504.
8. K. F. Kalz, I. R. Kraehnert, P. M. Dvoyashkin, I. R. Dittmeyer, R. Gläser, I. U. Krewer, K. Reuter and J.-D. Grunwaldt, ChemCatChem, 2017, 9, 17.
9. C. Kiener, M. Kurtz, H. Wilmer, C. Hoffmann, H.-W. Schmidt, J.-D. Grunwaldt, M. Muhler and F. Schüth, J. Catal., 2003, 216, 110–119.
10. G. Chinchen, K. Waugh and D. Whan, Appl. Catal., 1986, 25, 101–107.
11. G. Chinchen, P. Denny, J. Jennings, M. Spencer and K. Waugh, Appl. Catal., 1988, 36, 1–65.
12. S. Vukojević, O. Trapp, J.-D. Grunwaldt, C. Kiener and F. Schüth, Angew. Chem., Int. Ed., 2005, 44, 7978–7981.
13. Q.-C. Zhang, Z.-W. Liu, X.-H. Zhu, L.-X. Wen, Q.-F. Zhu, K. Guo and J.-F. Chen, Ind. Eng. Chem. Res., 2015, 54, 8874–8882.
14. J. Schumann, T. Lunkenbein, A. Tarasov, N. Thomas, R. Schlögl and M. Behrens, ChemCatChem, 2014, 6, 2889–2897.
15. T. Lunkenbein, J. Schumann, M. Behrens, R. Schlögl and M. G. Willinger, Angew. Chem., Int. Ed., 2015, 54, 4544–4548.
16. B. L. Kniep, T. Ressler, A. Rabis, F. Girgsdies, M. Baenitz, F. Steglich and R. Schlögl, Angew. Chem., Int. Ed., 2004, 43, 112–115.
17. Y. Zhang, Q. Sun, J. Deng, D. Wu and S. Chen, Appl. Catal., A, 1997, 158, 105–120.
18. Z. Chu, H. Chen, Y. Yu, Q. Wang and D. Fang, J. Mol. Catal. A: Chem., 2013, 366, 48–53.
19. R. Van Den Berg, G. Prieto, G. Korpershoek, L. I. Van Der Wal, A. J. Van Bunningen, S. Lægsgaard-Jørgensen, P. E. De Jongh and K. P. De Jong, Nat. Commun., 2016, 7, 1–7.
20. M. Behrens, D. Brennecke, F. Girgsdies, S. Kißner, A. Trunschke, N. Nasrudin, S. Zakaria, N. F. Idris, S. B. A. Hamid and B. Kniep, Appl. Catal., A, 2011, 392, 93–102.
21. B. Bems, M. Schur, A. Dassenoy, H. Junkes, D. Herein and R. Schlögl, Chem. – Eur. J., 2003, 9, 2039–2052.
22. M. Behrens, Catal. Today, 2015, 246, 46–54.
23. J.-D. Grunwaldt, C. Kiener, F. Schüth and A. Baiker, Phys. Scr., 2005, 2005, 819.
24. M. Behrens, J. Catal., 2009, 267, 24–29.
25. M. Behrens and R. Schlögl, Z. Anorg. Allg. Chem., 2013, 639, 2683–2695.
26. S. Polierer, D. Guse, S. Wild, K. Herrera Delgado, T. N. Otto, T. A. Zevaco, M. Kind, J. Sauer, F. Studt and S. Pitter, Catalysts, 2020, 10, 816.
27. M. Schur, B. Bems, A. Dassenoy, I. Kassatkine, J. Urban, H. Wilmes, O. Hinrichsen, M. Muhler and R. Schlögl, Angew. Chem., Int. Ed., 2003, 42, 3815–3817.
28. T. Kirner, J. Albert, M. Günther, G. Mayer, K. Reinhäckel and J. Köhler, Chem. Eng. J., 2004, 101, 65–74.
29. Q.-C. Zhang, K.-P. Cheng, L.-X. Wen, K. Guo and J.-F. Chen, RSC Adv., 2016, 6, 33611–33621.
30. M. M. Günter, B. Bems, R. Schlögl and T. Ressler, J. Synchrotron Radiat., 2001, 8, 619–621.
31. X. Jiang, X. Qin, C. Ling, Z. Wang and J. Lu, AIChE J., 2018, 64, 2647–2654.
32. S. Marre and K. F. Jensen, Chem. Soc. Rev., 2010, 39, 1183–1202.
33. G. Tofighi, A. Gaur, D. E. Doronkin, H. Lichtenberg, W. Wang, D. Wang, G. N. Rinke, A. Ewinger, R. Dittmeyer and J.-D. Grunwaldt, J. Phys. Chem. C, 2018, 122, 1721–1731.
34. G. Tofighi, X. Yu, H. Lichtenberg, D. E. Doronkin, W. Wang, C. Wöll, Y. Wang and J.-D. Grunwaldt, ACS Catal., 2019, 9(6), 5462–5473.
35. G. Tofighi, D. Degler, B. Junker, S. Müller, H. Lichtenberg, W. Wang, U. Weimar, N. Barsan and J.-D. Grunwaldt, Sens. Actuators, B, 2019, 292, 48–56.
36. G. Tofighi, H. Lichtenberg, J. Pesek, T. L. Sheppard, W. Wang, L. Schöttner, G. Rinke, R. Dittmeyer and J.-D. Grunwaldt, React. Chem. Eng., 2017, 2, 876–884.
37. S. Wild, S. Polierer, T. A. Zevaco, D. Guse, M. Kind, S. Pitter, K. Herrera Delgado and J. Sauer, RSC Adv., 2021, 11, 2556–2564.
38. A. Zimina, K. Dardenne, M. Denecke, D. Doronkin, E. Huttel, H. Lichtenberg, S. Mangold, T. Pruessmann, J. Rothe and T. Spangenberg, Rev. Sci. Instrum., 2017, 88, 113113.
39. B. Ravel and M. Newville, J. Synchrotron Radiat., 2005, 12, 537–541.
40. B. Zhang and N. Yan, Catalysts, 2013, 3, 543–562.
41. Z. Li, S. Yan and H. Fan, Fuel, 2013, 106, 178–186.
42. P. Gao, F. Li, F. Xiao, N. Zhao, W. Wei, L. Zhong and Y. Sun, Catal. Today, 2012, 194, 9–15.
43. A. Gaur, W. Klysubun, N. N. Nair, B. Shrivastava, J. Prasad and K. Srivastava, J. Mol. Struct., 2016, 1118, 212–218.
44. R. Revel, D. Bazin and A.-M. Flank, J. Synchrotron Radiat., 1999, 6, 717–718.
45. E. Frei, A. Gaur, H. Lichternberg, L. Zwiener, M. Scherzer, F. Girgsdies, T. Lunkenbein and R. Schlögl, ChemCatChem, 2020, 12, 4029–4033.
46. L. Galoisy, L. Cormier, G. Calas and V. Briois, J. Non-Cryst. Solids, 2001, 293, 105–111.
47. G. Waychunas, C. Fuller and J. Davis, Geochim. Cosmochim. Acta, 2002, 66, 1119–1137.
48. D. Grandjean, V. Pelipenko, E. D. Batyrev, J. C. van den Heuvel, A. A. Khassin, T. M. Yurieva and B. M. Weckhuysen, J. Phys. Chem. C, 2011, 115, 20175–20191.
49. F. Studt, M. Behrens, E. L. Kunkes, N. Thomas, S. Zander, A. Tarasov, J. Schumann, E. Frei, J. B. Varley and F. Abild-Pedersen, ChemCatChem, 2015, 7, 1105–1111.
50. K. Klier, V. Chatikavanij, R. Herman and G. Simmons, J. Catal., 1982, 74, 343–360.
51. X.-M. Liu, G. Lu, Z.-F. Yan and J. Beltramini, Ind. Eng. Chem. Res., 2003, 42, 6518–6530.
52. M. Bowker, ChemCatChem, 2019, 11, 4238.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1re00499a

---