Stability and Deactivation Behavior of Cuprous Acetylide Containing Catalysts in Reppe Ethynylation

Abstract

Reppe’s ethynylation of formaldehyde uses coal-based acetylene to produce commercially valuable 1,4-butynediol with a silica-supported copper oxide-bismuth oxide catalyst. Cuprous acetylide (Cu₂C₂) is generally accepted to be the catalytically active phase, which is formed in situ from the CuO-Bi₂O₃/SiO₂ pre-catalyst under ethynylation conditions. The catalytic performance and stability of this sensitive Cu₂C₂ phase are evaluated by long-term experiments (up to 240 h) and by catalyst recycling (10 cycles of 22 h). Powder X-ray diffraction and Raman spectroscopy are found to be the best and the only applicable analytical tools for qualitative evaluation of Cu₂C₂’s crystallinity, purity, and morphology during in situ formation and for phase transformations during the ethynylation. They were continuously correlated with the catalytic performance (1,4-butynediol yield determined by gas chromatography). No catalyst deactivation was observed, indicating outstanding catalyst stability. Observed structural changes within the active Cu₂C₂ phase have obviously limited influence on the catalytic cycle and performance.

1. Introduction

The Reppe-type catalytic ethynylation of the formaldehyde process, developed in the 1940s, employs coal-derived gaseous acetylene as the starting material and copper-based catalysts to produce 1,4-butynediol (BYD) and has gained increased industry attention in recent decades [1,2,3,4,5,6]. This process offers a cost-effective and environmentally efficient alternative to the petrochemical production pathway for synthesizing industrially valuable 1,4-butynediol as a C4 feedstock, which is hydrogenated to 1,4-butenediol and 1,4-butanediol. It produces high-value biodegradable polymers, solvents, and chemicals like tetrahydrofuran, γ-butyrolactone, and polybutylene succinate [5,7,8,9,10,11]. Scheme 1 shows the catalytic ethynylation of formaldehyde and the subsequent hydrogenation reaction, with respective reaction parameters.

Silica-supported copper(II) oxide-bismuth oxide (CuO-Bi₂O₃/SiO₂), typically synthesized via co-precipitation, remains the most effective and widely applied ethynylation catalyst precursor in the industry [1,2,3,4,12,13,14]. Bismuth acts as a promoter and inhibitor and plays a crucial role in the ethynylation, exhibiting an electron synergistic effect with copper that weakens the CuO-to-carrier interaction, facilitating Cu²⁺’s reduction to Cu⁺, while inhibiting the over-reduction into inactive Cu⁰ [1,15,16,17,18,19]. Mesoporous, chemically inert silica with a large specific surface area and abundant pores effectively disperses and stabilizes the active copper(I) species. The unique copper-phyllosilicate (Si–O–Cu) nanotube structure provides a confinement effect, enabling slow release of the stable Cu⁺ centers. Nanocrystalline CuO, integrated into the well-structured SiO₂ network upon calcination, offers both outstanding initial catalytic activity and long-term stability [20,21,22,23,24].

The CuO catalyst precursor is activated in situ during ethynylation of formaldehyde, forming structurally complex polynuclear cuprous acetylide (Cu₂C₂), the commonly recognized catalytically active species in the ethynylation process [1,4,13,15,25]. Copper acetylide-containing species are highly sensitive and can explode when dried or exposed to energy sources like heat, impacts, sparks, and intensive X-ray and laser power [26,27,28]. Rigorous risk assessments and explosion prevention measures are essential. This explosive nature complicates interpreting the exact structure, formation mechanisms, influencing parameters, and catalytic behavior of copper acetylides and their catalytic role in Reppe ethynylation. Several copper acetylide-containing structures related to the catalytic ethynylation have been reported in the literature [28,29,30,31,32,33,34,35].

Brameld et al. [36] synthesized cuprous (Cu⁺) and cupric (Cu²⁺) acetylides from different copper salt solutions and on metallic copper surfaces. The cuprene (–(HC=CH)_n–) structure, formed via acetylene oxidative coupling (i.e., Glaser) and catalyzed by metallic copper (Cu⁰), was introduced [28,33,37,38]. Cataldo et al. [39,40,41,42,43] investigated the dicopper diacetylide (Cu–C≡C–C≡C–Cu) and copper polyynide structures with High-Performance Liquid Chromatography (HPLC), Raman, and Fourier Transform Infrared (FT-IR) spectroscopies. Judai et al. [29,44] confirmed Cu₂C₂ nanoparticles on the metallic surface and the nanowires via X-ray diffraction (XRD) and density functional theory (DFT) calculations. Bruhm et al. [34,45] experimentally tracked the activation progress from the CuO-Bi₂O₃/SiO₂ pre-catalyst into the Cu₂C₂ active phase, correlating its formation and catalytic performance using powder XRD and Raman spectroscopy.

The present research focuses on the stability of the assumed catalytically active phase Cu₂C₂ and the deactivation behavior of in situ-activated Cu₂C₂ from a CuO-based pre-catalyst by long-term and multi-cycle catalytic ethynylation experiments. Its phase transformation during activation and in catalysis is analyzed using powder XRD and Raman spectroscopy. The structural complexity and explosive nature of the polynuclear cuprous acetylide phases in the ethynylation catalysts, whether formed in situ from the CuO precursor or as precipitated Cu₂C₂ catalysts, pose significant challenges for qualitative and, in particular, quantitative analyses. Careful optimization of analytical parameters is crucial to ensure accurate and reliable analytical results and to prevent Cu₂C₂ explosions under the sensitive conditions during its handling and analysis (i.e., high laser and X-ray energy, heat, wet/dryness, vacuum, oxidative/reductive environments, and reactive solvents).

2. Results and Discussion

The following experiments were performed during the in situ transformation of a well-characterized industrial-like CuO-Bi₂O₃/SiO₂ (CBS) pre-catalyst into supported Cu₂C₂. Its role as the catalytically active phase driving the ethynylation reaction and its stability and potential deactivation during the reaction were then investigated in detail. A directly precipitated pure-phase Cu₂C₂ was used as a reference for XRD and Raman analysis. The activated Cu₂C₂-containing catalyst was tested in ethynylation by a 240 h long-term experiment and a 10-cycle 22 h experiment to examine the stability of the active Cu₂C₂ phase in catalysis, its catalytic performance, and the potential deactivation.

2.1. Structural Characterization of the Pre-Catalyst and of the Activated Catalyst—Cuprous Acetylide Formation

The CuO-Bi₂O₃/SiO₂ (CBS) pre-catalyst was synthesized via co-precipitation and calcined at 450 °C. A pure-phase Cu₂C₂ reference catalyst was synthesized via direct precipitation in cuprous ammonia solution under an acetylene atmosphere (refer to Section 3.1).

The standard CBS pre-catalyst was characterized, and its consistency and reproducibility were proven. It contains 35.3 wt.% copper and 3.7 wt.% bismuth, as determined by Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES). This aligned with the target of 35% and 4% for optimal catalytic performance in ethynylation [34,45]. It has a relatively large Brunauer-Emmett-Teller (BET) surface area of 256.6 m²/g and a high porosity, with a Barrett-Joyner-Halenda (BJH) pore volume of 0.26 mL/g and radius of 1.68 nm, measured by N₂ physisorption (refer to Appendix A, Figure A1). The results indicate a mesoporous silica matrix that supports well-dispersed CuO-Bi₂O₃ particles with a stable structure and potentially offers a large active surface area, also after catalyst activation.

The powder XRD diffractogram of the CBS pre-catalyst showed an XRD-amorphous structure with no characteristic CuO reflexes at 35.6° and 38.6°. The CuO phase was confirmed by hydrogen temperature-programmed reduction (H2-TPR), where a single signal of the typical CuO reduction peak existed at around 270 °C (refer to Appendix A, Figure A2). Thermogravimetric Analysis-Mass Spectroscopy (TGA-MS) further verified the purity of the CuO phase in CBS. The co-precipitated pre-catalyst was initially present as copper(II) basic nitrate (Cu₂(NO₃)(OH)₃) or carbonate (Cu₂(CO₃)(OH)₂) phases after drying, determined by XRD and TGA-MS (refer to Appendix A, Figure A3). Calcination at 450 °C in air is sufficient to eliminate those unwanted copper phases by thermal decomposition, reducing the complexity in analyzing the phase transformation from copper(II) precursors into Cu₂C₂. No further signals of thermal decomposition species, such as the sudden change in mass (TGA) and the ion current (MS), were observed up to 800 °C.

The well-characterized standard CBS pre-catalysts were activated in the ethynylation process under analogous conditions at 1.2-bar overpressure of acetylene and 90 °C in aqueous formaldehyde solution. It is hypothesized that the copper(II) phase (CuO) is pre-reduced into copper(I) sites (Cu⁺) under reductive ethynylation conditions and subsequently converts to the catalytically active Cu₂C₂ phase with acetylene, as depicted in Scheme 2.

Raman spectroscopy tracked the slow but successful in situ activation of the CuO-based pre-catalyst into a Cu₂C₂ phase within 180 min of ethynylation using 90 °C aqueous formaldehyde and 1.2-bar acetylene gas pressure, as shown in Figure 1.

Raman spectroscopy has been identified as the most suitable tool for the identification and characterization of the cuprous acetylide phase due to several reasons [34,45]. The unique vibrational frequencies of the bonds of Cu₂C₂-containing molecules are identified using the directly synthesized pure-phase Cu₂C₂ reference. Despite the high background noise, the characteristic C≡C π-bond is determined by its vibrational signal at a wavenumber of 1700 cm⁻¹, and the Cu–C≡C α-bond is at 420 cm⁻¹, aligned with the reported references [28,34,43,45]. Another vibrational signal is found at 573 cm⁻¹, which is proposed as a ≡C–C bond, indicating a carbon coupled with an acetylenic carbon. This can be attributed to the ethynylation products, BYD, or its intermediate, propargyl alcohol, (H(O–CH₂–)C≡C–CH₂–OH), and the acetylene coupling to polyynides (Cu⁺/H⁺–(C≡C–C≡C)_n–Cu⁺/H⁺). Additionally, the 1560 cm⁻¹ signal indicates the carbon G-peak, likely formed from decomposed acetylides due to its high sensitivity towards the laser power (minimal possible laser energy is applied), and also possibly through a complicated sample preparation procedure before Raman analysis [34,39,43,45,46]. Thus, Raman investigations can be used to identify a substantial number of potential intermediates and products that were possibly formed during activation, reaction, and deactivation of this catalyst.

While treating the CBS pre-catalyst in the aqueous formaldehyde solution under a nitrogen flow at 90 °C for an hour, no Cu₂C₂ formation was observed. After switching to acetylene and maintaining a 1.2-bar absolute acetylene pressure, the initially present broad and less intense signals disappeared completely. Cu₂C₂ signals, as well as the C–C coupling and the decomposed carbon signals, were observed and became more pronounced over 180 min, marking the Cu₂C₂ formation after activation under ethynylation conditions. Compared with the directly synthesized Cu₂C₂, the significantly lower intensities indicate less crystallinity and different crystallographic properties.

Powder XRD of in situ-formed Cu₂C₂ from CBS catalysts exhibits weak, poorly resolved reflexes compared to the more crystalline pure-phase Cu₂C₂ reference, indicating lower crystallinity. This highlights XRD’s reduced sensitivity (amorphous character) relative to Raman spectroscopy for detecting Cu₂C₂ and intermediates, particularly as surface species on silica, consistent with the detectable but less pronounced Cu₂C₂ reflexes observed in the XRD diffractograms in the next section.

2.2. Catalysis, Structural Changes, and Deactivation of Cuprous Acetylide in the Ethynylation Reaction

The Cu₂C₂-containing pre-activated CuO-Bi₂O₃/SiO₂ catalyst was evaluated in a batch Reppe ethynylation process to investigate its catalytic stability and potential deactivation behavior. The 1,4-butynediol yield (BYD-Y) in % was converted from Gas Chromatography (GC) analysis by calculating the moles of BYD produced from formaldehyde (FA) at a 1:2 stoichiometric ratio to evaluate the catalytic performance. The corresponding structural transformation of the highly sensitive Cu₂C₂ active phase was analyzed using XRD and Raman. While analytical techniques like Transmission Electron Microscope (TEM), Scanning Electron Microscope (SEM), X-ray Photoelectron Spectroscopy (XPS), and Attenuated Total Reflection Fourier Transform Infrared Spectroscopy (ATR-FTIR) could offer in-depth characterization of the unique but unrevealed Cu₂C₂-containing species, their requirements (e.g., vacuum, high-energy beams, non-IR-active symmetric C≡C²⁻ anion) conflict with Cu₂C₂’s handling constraints, which will pose safety risks. This is probably also the reason for the lack of direct active phase characterization in the literature using these tools and why up to now they have only been applied to supported copper oxide pre-catalysts.

For the 240 h long catalytic ethynylation experiment, the BYD yield over time plot is shown in Figure 2. To prevent high formaldehyde consumption from limiting the reaction rate, 120 g of aqueous formaldehyde, three times the standard amount, was used. As no further increase in BYD yield was observed from 150 h to 190 h, the reaction was paused, a fresh batch of formaldehyde solution was replaced, and the reaction was restarted at 200 h, continuing until 240 h.

The BYD yield versus time plot exhibits an increasing concave downward trend, indicating a decreasing reaction rate up to 190 h. At the initial stage of ethynylation, a high and constant reaction rate was found until ~50 h, achieving ~38% yield, reflected by a linearly increasing slope. The reaction rate gradually declined until 150 h, reaching a ~77% yield, with FA conversion rising to ~100%. From 150 h to 190 h, the yield plateaued at ~78%, indicating complete FA consumption and no further BYD production. With the replacement of fresh formaldehyde solution at 200 h, the catalytic reaction continued, and a higher reaction rate was achieved, reaching a ~40% yield in 40 h (200 h to 240 h).

A high FA concentration and freshly activated catalysts ensure optimal conditions to achieve the highest and most consistent BYD formation rate at the initial stage, suggesting catalytic activity as the primary rate-determining factor. As FA was consumed along with the reaction, it became the limiting reagent that restricted the BYD formation rate during the long-term experiments. Once the FA was fully consumed, BYD production stopped. A linear increase in BYD yield at an even higher rate after the introduction of fresh FA at 200 h supports the hypothesis that the catalyst retains its high activity, but a low formaldehyde concentration limits the reaction. This indicates excellent catalytic activity and long-term stability of Cu₂C₂ active species in the ethynylation. This conclusion is also supported by XRD and Raman analysis of the spent catalysts, taken at 6, 50, 150, 200, and 240 h during ethynylation (Figure 3 and Figure 4). An activated CBS catalyst and a directly synthesized pure-phase Cu₂C₂ are provided as references.

From the powder XRD diffractograms (Figure 3), it can be concluded that the directly synthesized pure-phase Cu₂C₂ produces high-purity and highly crystalline Cu₂C₂, with the most characteristic XRD reflexes at 2θ of 32.0°, 42.3°, and 45.8°, which agrees with the literature [34,45]. The in situ-activated Cu₂C₂ from the CBS pre-catalyst showed less significant characteristic reflexes of Cu₂C₂ and was slightly shifted due to different crystallographic structures. They showed poor resolution, low signal-to-noise ratio, and several undefined reflexes, suggesting minor bismuth-containing phases (e.g., Bi₂O₂ variants) or acetylide intermediates with varying crystallinities. They were challenging to identify, as no definitive matches were found when fitting the powder pattern against the existing database due to low intensity and signal overlap. From the activated CBS reference to ethynylation samples over 240 h, the diffractograms show minor changes, except for the disappearance of a signal at 2θ of 12° and slightly sharpened and more intense signals at around 2θ of 30°. More significant changes were observed in the Raman spectra, as shown in Figure 4.

Raman spectroscopy serves as a more reliable identification of Cu₂C₂, targeting chemical bonds (as discussed before, Figure 1). The intensity of the characteristic Cu₂C₂ signals at 420 cm⁻¹ and 1700 cm⁻¹, as well as 573 cm⁻¹, decreased over time, except for the one at 200 h, which showed some recovery, where the catalyst had been resting for 10 h in an acetylene atmosphere without heating. The intensity reduced again after resuming ethynylation at 240 h. The minor changes in the Raman spectra suggest the gradual, recoverable deformation of the Cu₂C₂ crystalline structure during ethynylation.

Both the XRD and Raman results align with the catalytic performance in BYD yield (Figure 2). The catalyst retained its activity, while the catalytically active Cu₂C₂ phase maintained its structure throughout the 240 h ethynylation. Reduced available formaldehyde is proposed as the sole factor limiting the catalysis from ~50 h to 150 h, with no BYD yield until 190 h, indicating complete FA consumption. After refilling with formaldehyde, catalysis resumed at a comparable rate, suggesting that no catalyst deactivation had taken place.

The Raman spectra for most samples show the broad carbon G-peak signal at 1560 cm⁻¹ and a slight carbon D-peak at 1350 cm⁻¹, indicating minor carbon deposition of about 2% (as shown in Figure 5). This may have contributed to the decomposition of the highly sensitive acetylides by laser energy, which switches the intense C≡C signals to carbon signals. Another possibility is the acetylene coupling and surface carbon deposition (coking) on the Cu₂C₂-containing catalyst during ethynylation. However, this does not cause catalytic deactivation, as evidenced by the sustained BYD formation rate after a fresh batch of formaldehyde was added at 200 h (shown as the comparable initial slopes in Figure 2). Consistent Raman signals also confirm no significant increase in Cu₂C₂ sites, supporting sustained catalytic activity. Additionally, the carbon content during activation was not quantified stepwise, but complete Cu₂C₂ formation was confirmed by Raman, where the full signals of the Cu₂C₂ phase were observed at 180 min (refer to Figure 1), and by consistent catalytic performance, indicating no residual non-activated material.

A carbon analysis during ethynylation was carried out to verify the carbon content in the catalyst samples over 240 h, as shown in Figure 5. In addition, the copper content was analyzed by ICP-OES at 30.0 wt.% in the activated catalyst, 29.4% at 50 h, and 29.1% at 240 h after ethynylation. The copper leaching into the reaction mixture was also determined by ICP-OES, with a negligible amount below 8 ppm or 0.03%, supporting the constant copper content in the catalyst.

The total carbon content of ~14.0 wt.% was analyzed in the activated catalyst and remained constant during the first 24 h. It rose to 14.9% at 50 h and 16.7% at 150 h but dropped to 15.6% at 200 h, since the catalyst had been separated for 10 h. It increased again to 16.0% at 240 h after restarting ethynylation, following the same trend as found in the Raman data.

The net gain in carbon mass may come from the surface carbon deposition, supported by the Raman signal at 1560 cm⁻¹, and the acetylide coupling (–C≡C–C≡C–) within the copper acetylide phase, suggested as a polycopper polyacetylide structure, which is also indicated by Raman analysis with a signal at 573 cm⁻¹, corresponding to a ≡C–C bond.

The net loss in carbon content during the interruption of the ethynylation at 200 h implies a catalyst recovery. It is proposed that the surface-deposited or weakly adsorbed carbon/acetylene species are desorbed when the acetylene pressure is released, with lower solubility and concentration of acetylene in the reaction mixture.

In another series of experiments, a ten-cycle 22 h catalytic ethynylation and a pre-test eight-cycle 15 h experiment were conducted, as shown in Figure 6. A fresh batch of diluted formaldehyde solution was applied for each cycle of experiments. The FA dilution prevents it from becoming a rate-limiting factor due to the rapid initial consumption, which ensures that the catalytic activity is the only rate-limiting step.

A generally reduced trend in BYD yield was observed over ethynylation cycles in both tests. The highest yield of 35.5% was found in Cycle 2, and the lowest of 30.6% in Cycle 9. The loss of the catalyst traces after each reaction cycle during catalyst separation, washing, purification, and transfer is expected to be the cause of the slightly reduced yield, instead of the loss of catalytic activity over each ethynylation cycle.

XRD and Raman analyses were also performed after each cycle, but they were not shown, since they are analogous to the ones in Figure 3 and Figure 4, and no additional information can be drawn.

The minor changes in Raman spectra and stable catalytic performance indicate that the Cu₂C₂ phase undergoes limited structural deformation without deactivation. To further elucidate the catalytic process, a probable reaction sequence for the ethynylation is proposed. The nucleophilic addition of the activated cuprous acetylide anions (Cu–C≡C^δ−) to the electrophilic carbonyls in the formaldehyde (^δ+COH) results in the formation (and desorption) of propargylic alcohol (Cu⁺/H⁺ –C≡C–CH₂OH). The subsequent reaction of this adsorbed intermediate species with formaldehyde is, however, fast and forms the final main product 1,4-butynediol (HOH₂C–C≡C–CH₂OH). The relative reaction rates are influenced by the catalyst, the number of active cuprous acetylide sites, the formaldehyde concentrations, and other reaction parameters, such as acetylene pressure, temperature, and stirring rate (affecting gas solubility and limiting the reagent and mass transport efficiency). This proposed reaction mechanism is consistent with the sustained BYD yield (Figure 3) and Raman analysis of bonds corresponding to Cu₂C₂ species (Figure 4), although the exact Cu₂C₂ phase composition and structure remain under investigation.

3. Materials and Methods

Caution: Copper acetylides, particularly when dry or crystalline, are highly sensitive and potentially explosive, even without energy exposure. Avoid contact with energy sources such as vacuum, pressure, heat, sparks, intense X-rays, or laser radiation. Synthesis, handling, storage, transportation, and analysis were performed using diluted (e.g., disperse on silica support) and wetted samples under mild, inert conditions (e.g., sealed, lowered heat/pressure/energy), following rigorous risk assessments. All copper acetylide-containing samples must be deactivated completely in excess concentrated nitric acid prior to safe disposal [26,28].

3.1. Preparation of Ethynylation Catalysts

A standard CuO-Bi₂O₃/SiO₂ pre-catalyst containing 35 wt.% copper and 4 wt.% bismuth was synthesized by co-precipitation. Copper from Cu(NO₃)₂·3H₂O was dissolved in distilled water, while bismuth from Bi(NO₃)₃·5H₂O was dissolved in concentrated nitric acid, and both acidic salt solutions were mixed in one beaker. Silicon from Na₂SiO₃ was dissolved in distilled water as a basic solution in another beaker and adjusted to the same volume as the acidic solution by adding water. A precipitation vessel with 50 mL of bottom water was heated to 60 °C with 500 rpm stirring on a magnetic stirring hotplate. Both salt solutions were added dropwise simultaneously at a flow rate of 2 mL/min to the vessel using a peristaltic pump. Co-precipitation was monitored using a Metrohm 906 Titrando automatic titrator (Mettler-Toledo GmbH, Giessen, Germany) at pH 7.0 by adding 2 M Na₂CO₃ solutions. After completion, the reaction mixture was aged for 1 h under the same conditions. The resulting precipitates were then cooled, filtered, and washed with distilled water multiple times, until the conductivity of the filtrate was below 4 mS/m. The residual precipitates were dried at 80 °C overnight and calcined at 450 °C for 4 h in a muffle furnace. The obtained pre-catalyst was then sieved to a diameter between 100 μm and 300 μm.

A pure-phase Cu₂C₂ model catalyst was synthesized by direct precipitation. Copper(II) sulfate was dissolved in 50 mL of 25% ammonia solution in a Schlenk tube. A reducing agent was added to reduce copper(II) into copper(I), observed by the color change of the solution from deep blue to colorless. The mixture solution was stirred by a magnetic stirrer with constant argon flow. The reddish-brown Cu₂C₂ precipitates formed once the acetylene was introduced. The acetylene flow was maintained for 60 min. The resulting precipitates, pure-phase Cu₂C₂, were removed and washed with water and methanol by centrifugation. They were then purified and dried in a vacuum.

3.2. Catalytic Ethynylation Procedures

The Carousel 6PLUS Reaction Station^TM (Radley & Co Ltd., Essex, UK), as shown in Figure 7, consists of six flask ports and was used for the ethynylation experiments. It allows for a six-fold testing capacity, enhances homogeneity, and synchronizes reaction conditions.

The CuO-Bi₂O₃/SiO₂ pre-catalyst was activated in situ in the reaction flask by treating 2.0 g of pre-catalyst (35 wt.% copper) in 37% saturated aqueous formaldehyde solution (pH 7.0) at 90 °C under nitrogen flow for 30 min, followed by acetylene purging at 1.2 bar for 30 min, forming the Cu₂C₂ active phase. The activated catalyst was tested for the catalytic performance in ethynylation under analogous conditions. A pure-phase Cu₂C₂ catalyst with the same copper content was tested for comparison. In the reaction flask, 1.5 g of 1,3-propanediol was added as an internal standard for GC analysis. A total of 120 g of aqueous formaldehyde solution was used for the 240 h long-term experiment; another 120 g of fresh formaldehyde was added as replacement after 190 h when 100% formaldehyde conversion was detected, and the reaction continued from 200 h to 240 h. A total of 40 g of aqueous formaldehyde solution diluted with 40 g of water was used for the ten-cycle 22 h experiment. A fresh batch of diluted formaldehyde solution was applied for every new cycle.

The reaction mixture was stirred and purged, first with nitrogen for 30 min, followed by acetylene for another 30 min while heating to 90 °C. The ethynylation was initiated once the outlet safety pressure relief valve was switched off. The static condition of the reaction system was maintained at 1.2-bar absolute acetylene pressure and 90 °C. After the reaction, the acetylene pressure was released, and the reaction system was cooled under the acetylene flow and further purged with nitrogen for 15 min at room temperature. The spent catalysts were separated via centrifugation and purified by washing with distilled water and methanol. They were stored accordingly under various conditions for analysis. The waste solution, Cu₂C₂-containing catalysts, and contaminated consumables were disposed of after being treated with concentrated nitric acids to completely disarm the explosive cuprous acetylide species for safety.

3.3. Sample Preparation, Characterization, and Analysis

Elemental Analysis: The copper composition in the catalysts was analyzed by Agilent 700 Series Inductively Coupled Plasma-Optical Emission Spectrometry (ICP-OES) (Agilent Technologies Deutschland GmbH, Waldbronn, Germany). Merck Certipur^® ICP multi-element standard solution IV (Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany) in diluted nitric acid was used for the calibration. The defined amount of catalyst was dissolved in concentrated nitric acid. It was heated to boil, or ultrasonically treated, if necessary, to ensure complete dissolution. The dissolved sample was diluted to a concentration of up to 50 ppm and an acidity of up to 10% nitric acid. Then, 10 mL of the sample solution was filtered through a 0.45-micron syringe filter and collected for analysis. Carbon content analysis was performed using a HEKATech Euro-EA elemental analyzer with HT pyrolysis (HEKAtech GmbH, Wegberg, Germany). Following this, 1.0–2.0 mg of an air-sensitive activated catalyst sample was weighed in the glove box and sealed tightly into two layers of tin capsules. It was then transferred into the analyzer for combustion with oxygen to 1200 °C and determined by gas chromatography coupled with a thermal conductivity detector.

Gas Chromatography: Ethynylation products (BYD) and formaldehyde were quantitatively analyzed in molar amounts by Agilent G1530A gas chromatography with a thermal conductivity detector (GC-TCD) using an Agilent J&W CP-Sil 5 CB column (Agilent Technologies Deutschland GmbH, Waldbronn, Germany). Then, 0.5 mL of the reaction mixture was collected from the reaction using a syringe in a test tube, and 0.5 mL of acetonitrile was added as a diluent to stop the further reaction. The sample solution was filtered over activated aluminum oxide (SASOL GmbH, Witten, Germany) and collected in GC vials. The conversion of formaldehyde and the yield of BYD were calculated in % using the molar value with the stoichiometric ratio.

Hydrogen Temperature-Programmed Reduction: The pre-catalyst samples were analyzed by a Micromeritics Autochem Analyzer, coupled with a TCD (Micromeritics GmbH, Unterschleißheim, Germany), and connected with a Pfeiffer ThermoStar Gas Analyzer (Pfeiffer Vacuum GmbH, Asslar, Germany). The sample was loaded into a U-shaped glass reactor, fixed with glass wool, and inserted into the device. H₂-TPR analysis was performed from room temperature to 500 °C with a 5 K/min heating rate and a 50 mL/min 10% H₂/Ar gas flow. An about −90 °C cold trap (liquid nitrogen mixture) was prepared to freeze the decomposed gases.

Nitrogen Physisorption (BET/BJH): The pre-catalyst samples were analyzed using a Quantachrome NovaTouch Analyzer (Anton Paar Germany GmbH, Ostfildern-Scharnhausen, Germany). The sample was weighed in a quartz tube. It was first degassed under vacuum for 3 h at 120 °C, followed by N₂ physisorption at liquid nitrogen temperature (−196 °C) over the full relative pressure range, 0 < p/p₀ < 1.

Powder X-ray Diffraction: The samples in dried powders or wet slurries were analyzed by a Rigaku MiniFlex X-Ray Diffractometer fitted with a high-sensitivity D/tex Ultra Si-strip detector (Rigaku Europe SE, Neu-Isenburg, Germany) (Kα(Cu), λ = 1.5419 Å) in a reflection (Bragg-Brentano) mode. The sample was flattened onto a silicon wafer plate and inserted into the device’s sample holder. The X-ray diffractograms were recorded with an angular range of 10° to 60° in a step size of 0.01°, 5°/min velocity, 40 rpm spin, and 50 mA power. Data evaluation was carried out with PANalytical HighScore Plus 3.0e software (Malvern Panalytical B.V., Almolo, The Netherlands).

Raman Spectroscopy: The samples in dried powders or wet slurries were analyzed by a Renishaw InVia Raman Microscope, equipped with an Andor Newton EMCCD Camera with a frequency-doubled Nd: YAG crystal laser (λ = 532 nm) and a Leica 50×/0.75 magnification was applied (Renishaw GmbH, Pliezhausen, Germany). The samples were analyzed using a laser intensity of 0.015 mW, 3–10 s exposure, and 3–10 repetitions.

Thermogravimetric Analysis-Mass Spectroscopy: The pre-catalyst samples were analyzed by a Mettler Toledo Thermal Analysis System (Mettler-Toledo GmbH, Giessen, Germany), coupled with a Pfeiffer ThermoStar Gas Analyzer (Pfeiffer Vacuum GmbH, Asslar, Germany). The sample was weighed in an alumina crucible and placed on the balance in the TGA device chamber. The analysis was performed from 25 °C up to 800 °C with a ramp rate of 10 K/min under inert (argon), oxidative (synthetic air), or reductive (10% H₂/Ar) atmospheres, with a gas flow of 20 mL/min.

4. Conclusions

This work verified the cuprous acetylide (Cu₂C₂) as the catalytically active phase being formed during in situ activation of CuO-Bi₂O₃/SiO₂ pre-catalysts under ethynylation conditions. The phase transformations during its activation and in catalysis can be recorded by XRD and Raman spectroscopy, despite the highly explosive nature of the acetylide species. Although the detailed mechanism on the Cu₂C₂ surface or of species in solution remains a black box, long-term experiments confirm that the presence of this Cu₂C₂ phase represents the prerequisite for catalysis. Its presence was correlated qualitatively as well as quantitatively with the catalytic performance. The standard catalyst system applied showed high catalytic activity, paired with outstanding stability. Although structural changes in the active phase were observed, they showed limited influence on the catalytic cycle and performance.