Co/SiO2 Catalyst for Methoxycarbonylation of Acetylene: On Catalytic Performance and Active Species

Reppe carbonylation of acetylene is an atom-economic and non-petroleum approach to synthesize acrylic acid and acrylate esters, which are key intermediates in the textile, leather finishing, and polymer industries. In the present work, a noble metal-free Co@SiO2 catalyst was prepared and evaluated in the methoxycarbonylation reaction of acetylene. It was discovered that pretreatment of the catalyst by different reductants (i.e., C2H2, CO, H2, and syngas) greatly improved the catalytic activity, of which Co/SiO2-H2 demonstrated the best performance under conditions of 160 °C, 0.05 MPa C2H2, 4 MPa CO, and 1 h, affording a production rate of 4.38 gMA+MP gcat−1 h−1 for methyl acrylate (MA) and methyl propionate (MP) and 0.91 gDMS gcat−1 h−1 for dimethyl succinate (DMS), respectively. Transmission electron microscopy (TEM), X-ray diffraction (XRD), and diffuse reflectance infrared Fourier transform spectra of CO adsorption (CO-DRIFTS) measurements revealed that an H2 reduction decreased the size of the Co nanoparticles and promoted the formation of hollow architectures, leading to an increase in the metal surface area and CO adsorption on the catalyst. The hot filtration experiment confirmed that Co2(CO)8 was generated in situ during the reaction or at the pre-activation stage, which served as the genuine active species. Our work provides a facile and convenient approach to the in situ synthetization of Co2(CO)8 for a Reppe carbonylation reaction.


Introduction
Reppe carbonylation reactions of acetylene (including hydrocarboxylation, alkoxycarbonylation, etc.) (Scheme 1) provide atom-economic and non-petroleum routes to produce acrylic acid (AA), methyl acrylate, and dimethyl succinate [1,2], which are raw materials and key intermediates that are widely used in the textile, leather finishing, and polymer industries, with a global demand of 10 million tons annually [3][4][5].

Introduction
Reppe carbonylation reactions of acetylene (including hydrocarboxylation, alkoxycarbonylation, etc.) (Scheme 1) provide atom-economic and non-petroleum routes to produce acrylic acid (AA), methyl acrylate, and dimethyl succinate [1,2], which are raw materials and key intermediates that are widely used in the textile, leather finishing, and polymer industries, with a global demand of 10 million tons annually [3][4][5].Various catalysts have been developed for the carbonylation reactions of acetylene, of which Ni-based catalysts are the original ones [6].Reppe and coworkers were the first to discover that Ni(CO)4 was able to catalyze the hydrocarboxylation of acetylene to pro-also reported.Among others, Badische Anilin-und-Soda-Fabrik (BASF) developed the commercial catalytic system NiBr 2 -CuBr 2 -CH 3 SOOOH for hydrocarbonylation of acetylene [11], which, however, suffers from a difficulty in the separation of catalyst products [12].To tackle this problem, metal oxides (such as SiO 2 , Al 2 O 3 , vermiculite, and MCM-41) [13][14][15] and zeolites (e.g., NaY and ZSM-5) [16] were applied as support materials of Ni.However, these catalysts have one or more drawbacks, which include a short catalyst lifetime, harsh reaction conditions, and metal leaching.For example, Shi and coworkers reported that Ni/Y served as an efficient catalyst for the hydrocarboxylation of acetylene [17], and the AA space-time yield reached 62 g acrylic acid g cat −1 h −1 under reaction conditions of 235 • C and 3.6 MPa.Although they claimed that the catalysis followed a heterogeneous pathway, the Ni/Y material suffered from severe coking.Yan and coworkers prepared a NiO/AlOOH catalyst, which gave an AA space-time yield as high as 412 g acrylic acid g cat −1 h −1 , yet it was discovered that the leached Ni-carbonyls were the true active species [12].Given the high toxicity and instability of Ni(CO) 4 , which might generate in situ during a reaction, the potential application of Ni-based catalysts is limited.
Other catalysts based on group VIII transition metals, such as Pd, Ir, and Ru, have also been explored [18].For instance, Drent and coworkers developed a Pd complex with the ligand of 2-pyridylphosphine (2-PyPPh 2 ) [19], which exhibited excellent catalytic performance in the methoxycarbonylation of propyne, and the TOF reached as high as 40,000 h −1 with 99.95% selectivity to methyl methacrylate.The 2-pyridyl ring was found to be crucial for the catalytic activity, which, if replaced by a phenyl or 4-pyridyl group, led to a drastic decrease in TOF.Nonetheless, strong acid promoters, e.g., p-toluenesulfonic acid, are indispensable for the reaction, which would cause serious environmental issues [20].Ding and coworkers developed porous organic polymer-supported single-site Pd catalysts, in which the 2-pyridylphosphine ligand and p-toluenesulfonic acid were incorporated into the framework of the support [21,22], and the methoxycarbonylation reaction of acetylene was allowed to run in the absence of an extra acid promoter.However, this approach suffered from tedious procedures for the preparation of the support [23,24].In addition, the high cost and low abundance of Pd compromised its efficiency.
Cobalt is also the metal of choice for Reppe carbonylation reactions due to its earth abundance and high catalytic activity [25][26][27].Pyridine-promoted cobalt carbonyls demonstrated high efficiency in converting butadiene to adipic acid in a pilot-scale process developed by BASF [28,29].Cobalt carbonyls were also competent catalysts in the alkoxycarbonylation reactions of alkenes [30].Nevertheless, to the best of our knowledge, few reports have been produced on the Co-catalyzed methoxycarbonylation of acetylene.In the present work, we prepared SiO 2 -supported Co catalysts and investigated their catalytic performance in the methoxycarbonylation of acetylene in the absence of acid promoters.It was discovered that the pre-activation atmosphere (i.e., H 2 , C 2 H 2 , syngas, CO, and NH 3 ) had a profound impact on the catalytic activity, which increased following the trend of C 2 H 2 < CO < syngas ≈ H 2 .Under reaction conditions of 160 • C and 0.05 MPa C 2 H 2 and 4 MPa CO, Co/SiO 2 exhibited high catalytic activity, leading to a production rate of 4.2 g MA+MP g cat −1 h −1 for MA and MP and 0. 9 g DMS g cat −1 h −1 for DMS, respectively.Mechanism studies by XRD and hot filtration test revealed that Co nanoparticles underwent dynamic evolution during the reaction, where cobalt carbonyls were formed in situ and served as the genuine active species.Our work provided an approach for the in situ generation of catalytically active species for the methoxycarbonylation reaction, thus circumventing the employment of costly and unstable conventional Co 2 (CO) 8 catalysts, which are generally synthesized under harsh conditions of high temperatures and high pressures [31][32][33].

Results and Discussion
The Co/SiO 2 catalysts were prepared by the one-pot synthetic method, followed by calcination and reduction [34].The XRD patterns (Figure 1a) exhibited characteristic diffraction peaks of metallic Co with an fcc structure (JCPDS 00-01-1259), and the size of the Co was calculated to be 28 nm according to the Scherrer equation [35].No diffraction peaks ascribed to SiO 2 (JCPDS 00-33-1161) were present, probably owing to the amorphous nature of SiO 2 in the sample.The TEM images (Figure 1b) and energy-dispersive X-ray (EDX) spectra (Figure 1c) showed that Co nanoparticles were homogeneously distributed on SiO 2 , most of which had an average size of 18.6 nm.However, the examination of different regions showed that there were also smaller Co nanoparticles with sizes of 1-2 nm (Figure S1).The lattice spacing was measured to be 2.07 nm and 1.93 nm (Figure 1d), corresponding to the (111) and (101) plane in fcc Co, respectively, which were in agreement with the XRD measurements.

Results and Discussion
The Co/SiO2 catalysts were prepared by the one-pot synthetic method, followed by calcination and reduction [34].The XRD patterns (Figure 1a) exhibited characteristic diffraction peaks of metallic Co with an fcc structure (JCPDS 00-01-1259), and the size of the Co was calculated to be 28 nm according to the Scherrer equation [35].No diffraction peaks ascribed to SiO2 (JCPDS 00-33-1161) were present, probably owing to the amorphous nature of SiO2 in the sample.The TEM images (Figure 1b) and energy-dispersive X-ray (EDX) spectra (Figure 1c) showed that Co nanoparticles were homogeneously distributed on SiO2, most of which had an average size of 18.6 nm.However, the examination of different regions showed that there were also smaller Co nanoparticles with sizes of 1-2 nm (Figure S1).The lattice spacing was measured to be 2.07 nm and 1.93 nm (Figure 1d), corresponding to the (111) and (101) plane in fcc Co, respectively, which were in agreement with the XRD measurements.The as-prepared Co/SiO2 catalysts were evaluated in the methoxycarbonylation reaction of acetylene under conditions of 0.05 MPa C2H2, 4 MPa CO, 5 mL CH3OH, and 160 °C.However, no products were detected in the three catalysts with different Co loadings after The as-prepared Co/SiO 2 catalysts were evaluated in the methoxycarbonylation reaction of acetylene under conditions of 0.05 MPa C 2 H 2 , 4 MPa CO, 5 mL CH 3 OH, and 160 • C.However, no products were detected in the three catalysts with different Co loadings after a 1 h reaction (Table 1, entries 1-3).It is assumed that the passive layer of CoO x outside of the metallic Co might suppresses the reactant to access the catalytically active sites [36]; therefore, 1 MPa H 2 was charged into the reactor to activate the catalyst in situ.As expected, under this condition, the Co/SiO 2 samples demonstrated high catalytic performance for the target reaction.Three main products were detected (i.e., MA, MP, and DMS) after a 1 h reaction (Table 1, entries 4-6), of which MA and DMS were produced by the methoxycarbonylation and dimethoxycarbonylation of acetylene, respectively, while MP was derived from the hydrogenation of MA.Specifically, in the 41.7% Co/SiO 2 catalyst (Table 1, entry 4), all the MA had been completely hydrogenated to MP, which was produced with a rate as high as 4.42g MP g cat −1 h −1 , and DMS was yielded with a relatively lower rate of 0. 58 g DMS g cat −1 h −1 .The other two Co/SiO 2 catalysts with different Co loadings were also tested in the reaction.The 31.6% Co/SiO 2 sample showed similar catalytic performance to the 41.7% Co/SiO 2 sample (Table 1, entry 5).However, the catalytic activity of the 25.6% Co/SiO 2 sample, with an even lower Co loading, decreased drastically (Table 1, entry 6), and only the methoxycarbonylation product was yielded, with a rate of 0.39 g MA+MP g cat −1 h −1 .The reason might be that with a decrease in Co loading, the size of the Co became smaller, and in turn, the interaction between Co and SiO 2 became stronger (e.g., cobalt silicate might be formed) [34,37], leading to a more difficult reduction of the CoO x species.The other support-material-loaded Co catalysts were also evaluated in the reaction.However, Co/WO x , Co/HAP, Co/TiO 2 , and Co/S-1 were totally inactive for the reaction (Table 1, entries 7-10).Metallic Co powder was inefficient for the reaction as well (Table 1, entry 11).As the pre-activation of Co/SiO 2 by H 2 improved the catalytic activity remarkably, the other reductants, i.e., C 2 H 2 , CO, syngas, and NH 3 , were also employed to pre-activate the sample to see if the catalytic activity could be further improved.Except for NH 3 , the pre-activation was conducted in an autoclave in CH 3 OH under different reductant gases at 160 • C for 1 h, and after that, the autoclave was flushed with N 2 and recharged with the reactant gas (0.05 MPa C 2 H 2 and 4 MPa CO), and the reaction was allowed to run at 160 • C for 1 h.The pretreatment by NH 3 was carried out according to the literature [38].As shown in Table 2, the pretreatment under Ar made no contribution to the activity (Table 2, entry 1), thus precluding the possibility that CH 3 OH served as the reductant.By contrast, the pre-activations under C 2 H 2 , CO, or syngas atmosphere all led to obvious yields of the three products, and the production rate increased following the trend of Co/SiO 2 -C 2 H 2 (Table 2, entry 2) < Co/SiO 2 -CO (entry 3) < Co/SiO 2 -syngas (entry 5) ≈ Co/SiO 2 -H 2 (entry 4).The improvement in catalytic activity by pretreatment with C 2 H 2 or CO was surprising, as no products were detected using the as-made Co/SiO 2 catalyst under 0.05 MPa C 2 H 2 and 4 MPa CO at 160 • C for 1 h, as mentioned before (Table 1, entry 1).We suppose there might be an induction period of >1 h; accordingly, the reaction was once again conducted using an as-made Co/SiO 2 catalyst under 0.05 MPa C 2 H 2 and 4 MPa CO for a long time (i.e., 2 h); however, no products were detected.On the contrary, the Co/SiO 2 -NH 3 sample showed no catalytic activity for the reaction (Table 2, entry 6).The reason for this might be that NH 3 adsorbed strongly on the surface of metallic Co, and thereby, the adsorption of the reactants on the catalyst was prevented.To verify this, we carried out a control experiment, where the methoxycarbonylation of acetylene was conducted under conditions of 0.1 MPa NH 3 , 0.05 MPa C 2 H 2 , 4 MPa CO, 5 mL CH 3 OH, 160 • C, and 1 h using Co 2 (CO) 8 as the catalyst (Co 2 (CO) 8 was used here because it has been proven to be the genuine active species in the following section).It was found that no product was yielded, suggesting that the Co 2 (CO) 8 catalyst was poisoned by NH 3 .On the basis of the above results, it is inferred that pretreatment with a distinct reducing gas should modify the structure of the Co/SiO 2 catalyst, where the active species are more prone to be formed.Accordingly, TEM, XRD, and CO-DRIFT measurements of the activated catalysts were conducted.Figure 2 shows the TEM images and particle size histograms for different pre-activated Co/SiO 2 samples.The average particle size of metallic Co was 24.9 nm, and it was 20.3 nm for Co/SiO 2 -C 2 H 2 (Figure 2a,d) and Co/SiO 2 -CO (Figure 2b,e), respectively.Compared with fresh Co/SiO 2 (18.6 nm), it is obvious that the C 2 H 2 and CO pre-activation induce an increase in particle size.In sharp contrast, in the Co/SiO 2 -H 2 sample (Figure 2c,f and Figure S2), not only had the size of the Co decreased to 17.5 nm, but also, a lot of Co nanoparticles had been reconstructed to hollow shells.This reconstruction during H 2 reduction might result from the strain induced by a variation in the metal-metal distances between the CoO x surface layer and the inner metallic Co [39].
As shown in Figure 3a, the intensity of the diffraction peaks at 44.4 • and 47.3 • , attributed to the (111) and (101) plane of metallic Co [40] in the samples of Co/SiO 2 -C 2 H 2 and Co/SiO 2 -CO, increased greatly compared with those of the fresh Co/SiO 2 , indicating an increase in the size of the Co nanoparticles.In sharp contrast, for the Co/SiO 2 -H 2 sample, the diffraction peaks attenuated instead, suggesting a decrease in the Co particle size, or a reduction in the crystallinity of the metallic Co.These results were consistent with the TEM examination results but quite different from conventional experimental results, where H 2 reduction generally results in the sintering of metal nanoparticles [41].CO-DRIFTs were also carried out to study the electronic properties of Co.As shown in Figure 3b, four adsorption bands appeared on different samples.The band at 2048 cm −1 was ascribed to the ν(C≡O) of CO molecules that were linearly adsorbed on metallic Co [42], while the one at 1793 cm −1 was assigned to the ν(C≡O) of CO molecules in multi-bonded carbonyls on Co [43].The other two bands at 1627 and 1510 cm −1 were attributed to the ν asymm (C=O) of bicarbonate species and ν asymm (C=O) of carbonate species, respectively [44,45].It was found that the integrated area of the CO band at 2048 cm −1 and 1793 cm −1 increased with the trend of Co/SiO 2 -C 2 H 2 < Co/SiO 2 -CO < Co/SiO 2 -H 2 , indicating an increased surface area of the metallic Co, which should come from the reduction of the passive layer of CoO x .In addition, this peak red-shifted to 2030 cm −1 on Co/SiO 2 -CO and further to 2019 cm −1 on Co/SiO 2 -H 2 , which, according to the TEM images, might result from the increased proportion of corner and edge sites in the Co nanoparticles with a decrease in sizes.
[44,45].It was found that the integrated area of the CO band at 2048 cm and 1793 cm increased with the trend of Co/SiO2-C2H2 < Co/SiO2-CO < Co/SiO2-H2, indicating an increased surface area of the metallic Co, which should come from the reduction of the passive layer of CoOx.In addition, this peak red-shifted to 2030 cm −1 on Co/SiO2-CO and further to 2019 cm −1 on Co/SiO2-H2, which, according to the TEM images, might result from the increased proportion of corner and edge sites in the Co nanoparticles with a decrease in sizes.increased with the trend of Co/SiO2-C2H2 < Co/SiO2-CO < Co/SiO2-H2, indicating an increased surface area of the metallic Co, which should come from the reduction of the passive layer of CoOx.In addition, this peak red-shifted to 2030 cm −1 on Co/SiO2-CO and further to 2019 cm −1 on Co/SiO2-H2, which, according to the TEM images, might result from the increased proportion of corner and edge sites in the Co nanoparticles with a decrease in sizes.Stability is a key criterion for judging the quality of a catalyst [46].Therefore, we also studied the recyclability of the Co/SiO 2 catalyst.When the used catalyst was separated from the reaction mixture by filtration, it was found that the filtrate was dark red in color, suggesting severe leaching of Co [47,48].The concentration of Co in the filtrate of different samples increased, following the trend of Co/SiO 2 -C 2 H 2 < Co/SiO 2 -CO < Co/SiO 2 -H 2 (Figure 3c), which was in agreement with that of the catalytic activity.In addition, the Co/SiO 2 catalyst was discovered to already undergo Co leaching after the pretreatment with the reductant.In particular, for the Co/SiO 2 -syngas sample, after the pre-activation using syngas, the solid sample was separated by filtration and was then subjected to a batch of reactions (Figure S4); however, no products were yielded at all.By contrast, when the dark-red filtrate (containing 200 ppm Co) was employed as a catalyst, production rates of 3.76 g MA+MP g cat −1 h −1 of MA and MP and 0.49 g DMS g cat −1 h −1 of DMS were achieved (Table 3), which was quite similar to that of the Co/SiO 2 -syngas (Table 2, entry 5).These results indicated that it was the leached Co species that contributed to the overall catalytic activity. 13C NMR measurement was then conducted to identify the leached Co species; however, the filtrate did not show any signal because of the paramagnetic nature of Co.The yellow color of the filtrate implied that the leached Co species might be Co 2 (CO) 8 .To confirm this point, we used commercial Co 2 (CO) 8 as a catalyst for the methoxycarbonylation of acetylene.It was discovered that Co 2 (CO) 8 demonstrated similar catalytic activity and product distribution to the filtrate (Figure 4), verifying that Co 2 (CO) 8 was the true active species for the reaction.samples increased, following the trend of Co/SiO2-C2H2 < Co/SiO2-CO < Co/SiO2-H2 (Figure 3c), which was in agreement with that of the catalytic activity.In addition, the Co/SiO2 catalyst was discovered to already undergo Co leaching after the pretreatment with the reductant.In particular, for the Co/SiO2-syngas sample, after the pre-activation using syngas, the solid sample was separated by filtration and was then subjected to a batch of reactions (Figure S4); however, no products were yielded at all.By contrast, when the darkred filtrate (containing 200 ppm Co) was employed as a catalyst, production rates of 3.76 gMA+MP gcat −1 h −1 of MA and MP and 0.49 gDMS gcat −1 h −1 of DMS were achieved (Table 3), which was quite similar to that of the Co/SiO2-syngas (Table 2, entry 5).These results indicated that it was the leached Co species that contributed to the overall catalytic activity. 13C NMR measurement was then conducted to identify the leached Co species; however, the filtrate did not show any signal because of the paramagnetic nature of Co.The yellow color of the filtrate implied that the leached Co species might be Co2(CO)8.To confirm this point, we used commercial Co2(CO)8 as a catalyst for the methoxycarbonylation of acetylene.It was discovered that Co2(CO)8 demonstrated similar catalytic activity and product distribution to the filtrate (Figure 4), verifying that Co2(CO)8 was the true active species for the reaction.Co2(CO)8 is known as an efficient catalyst for the alkoxycarbonylation and hydroformylation of alkenes [49-51], although no reports on its catalytic performance in the methoxycarbonylation of acetylene have been published.Our work thus demonstrated that Co2(CO)8 is also a competent catalyst for this transformation to produce MA and MP.Traditionally, Co2(CO)8 is quite unstable, and its synthesis requires either harsh reaction conditions (i.e., 140-230 °C, 10-70 MPa) or stoichiometric reducing reagents (such as metal Co 2 (CO) 8 is known as an efficient catalyst for the alkoxycarbonylation and hydroformylation of alkenes [49-51], although no reports on its catalytic performance in the methoxycarbonylation of acetylene have been published.Our work thus demonstrated that Co 2 (CO) 8 is also a competent catalyst for this transformation to produce MA and MP.Traditionally, Co 2 (CO) 8 is quite unstable, and its synthesis requires either harsh reaction conditions (i.e., 140-230 • C, 10-70 MPa) or stoichiometric reducing reagents (such as metal powders, NaBH 4 , or Na 2 S 2 SO 3 ) or phosphine oxide promoters.Li and coworkers recently reported that Co 2 (CO) 8 could be formed in situ from Co/MoS 2 under conditions of 6 MPa CO, 140 • C, and 15 h, and MoS 2 was claimed to be critical for this process [28].This work demonstrated that when using conventional SiO 2 as a support material, Co 2 (CO) 8 could also be generated in situ under conditions of 4 MPa CO, 1 MPa H 2 , 160 • C, and 1 h, thus providing a facile, straightforward approach to prepare Co 2 (CO) 8 for Reppe carbonylation reactions.

Materials and Methods
The Co/SiO 2 was prepared according to the literature [34].Briefly, 2.91 g Co(NO 3 ) 2 •6H 2 O (10 mmol) and 2.08 g TEOS (10 mmol) were dissolved in 50 mL mixed liquor of water and ethanol (3/1, v/v) and stirred for 10 min.After, 5 mL of NH 3 •H 2 O was added to the above solution, and the suspension was stirred at room temperature for another 8 h.The precipitate was separated and collected by filtration, washed with deionized water, and dried at 100 • C overnight.The obtained solid was calcined at 500 • C in a muffle furnace for 4 h and was then reduced in flowing pure hydrogen (100 mL/min) for 3 h at 600 • C. The obtained Co/SiO 2 catalyst was labeled as 41.7% Co/SiO 2 .The other two catalysts with different Co loadings (i.e., 31.6%Co/SiO 2 and 25.6% Co/SiO 2 ) were prepared using similar procedures, except for changing the amount of TEOS to 15 and 20 mmol, respectively.
The other support material (i.e., WO x , hydroxyapatite (HAP), and TiO 2 )-loaded Co catalysts were synthesized by the impregnation method.First, 5 mL Co(NO 3 ) 2 •6H 2 O solution (10 mg Co /mL) and 5 mL ultrapure water were mixed and stirred to form a transparent solution, and then, 1 g carrier was added to the above solution.After the mixture was stirred at room temperature for 24 h, the excess water was removed by rotary evaporation until dry.The resulting power was further dried at 373 K overnight and then reduced flowing pure hydrogen (100 mL/min) for 3 h at 600 • C.
The preparation of the Co 2 (CO) 8 solution in methanol from Co/SiO 2 started with adding 10 mg of fresh Co/SiO 2 to a 30 mL autoclave with 5 mL methanol, and then, the autoclave was purged with N 2 repeatedly (4 times) to remove gaseous and dissolved oxygen.After that, the autoclave was charged with 4 MPa CO and 1 MPa H 2 , and the mixture was stirred (600 rpm) at 160 • C for 4 h.Then, the autoclave was placed into an ice bath until the temperature was below 5 • C. The solid was separated by filtration, and the yellow filter liquor was collected and kept at 0 C.
Other comparison materials, including Co power and Co 2 (CO) 8 , are commercial samples at analytically pure levels.

Characterization
The concentration of Co in the filtrate after pretreatment or reaction was determined by inductively coupled plasma spectroscopy (ICP-AES) on an IRIS Intrepid II XSP instrument (Thermo Electron Corporation, Waltham, MA, USA).The patterns of XRD were recorded on a PANalytical X' pert diffractometer with a Cu-Kα radiation source, operated at 40 kV and 40 mA under a continuous mode in the 2θ range of 10 • ~90 • .The morphology and element distribution of samples were observed by STEM and EDS experiments, which were performed on a JEM-2100F Transmission Electron Microscope (JEOL, Singapore) with a spatial resolution of 0.19 nm at 20 kV, equipped with the Oxford Instruments ISIS/INCA EDS system with an Oxford Pentafet Ultrathin Window (UTW) detector.Before the microscopy examination, the sample was demagnetization-treated using a magnet.After that, the sample was ultrasonically dispersed in ethanol for 5-10 min, and then, a drop of the suspension was dropped on a copper TEM grid coated with a thin holey carbon film.CO-DRIFTS (diffuse reflectance infrared Fourier transform spectroscopy) experiments were performed by means of a Bruker Equinox70 spectrometer (Bruker, Singapore), equipped with a mercury-cadmium-telluride detector at a resolution of 4 cm −1 , using 16 scans in a range of 400-4000 cm −1 .Prior to the measurement, the catalysts were pretreated in situ with H 2 (30 mL/min) at 160 • C for 30 min, and then, the flow was switched to He (30 mL/min) at 170 • C for another 30 min to remove surface H.After that, the catalyst cooled down to room temperature, and background spectra were recorded at 20 • C in He flow.

Catalyst Evaluation
The acetylene methoxycarbonylation (AMC) reaction was performed in a 30 mL autoclave equipped with a quartz lining.The autoclave was filled with 10 mg of a catalyst and 5 mL methanol, and then, the autoclave was purged with N 2 repeatedly (4 times) to remove gaseous and dissolved oxygen.After that, it was charged with 1 MPa 5% C 2 H 2 -95% He, 1 MPa H 2 , and 4 MPa CO.The reaction was started by heating the mixture to 160 • C under vigorous stirring at a speed of 600 rpm.After the reaction, the reactor was cooled to

Figure 1 .
Figure 1.(a) The XRD patterns of the fresh Co/SiO2 catalyst and the references of fcc Co and SiO2; (b) the transmission electron microscope (TEM) images of Co/SiO2 and the size distribution of Co nanoparticles; (c) the EDX spectra of Co, Si, and O; (d) the TEM image of Co/SiO2 and exposed lattice plane.

Figure 1 .
Figure 1.(a) The XRD patterns of the fresh Co/SiO 2 catalyst and the references of fcc Co and SiO 2 ; (b) the transmission electron microscope (TEM) images of Co/SiO 2 and the size distribution of Co nanoparticles; (c) the EDX spectra of Co, Si, and O; (d) the TEM image of Co/SiO 2 and exposed lattice plane.

Figure 3 .
Figure 3. (a) XRD patterns of fresh and pre-activated Co/SiO2, as well as reference Co samples; (b) DRIFT spectra of CO adsorption on different samples; (c) amount of leaching Co in reaction filtrate from pre-activated Co/SiO2 (determined by ICP-OES).

Figure 3 .
Figure 3. (a) XRD patterns of fresh and pre-activated Co/SiO2, as well as reference Co samples; (b) DRIFT spectra of CO adsorption on different samples; (c) amount of leaching Co in reaction filtrate from pre-activated Co/SiO2 (determined by ICP-OES).

Figure 3 .
Figure 3. (a) XRD patterns of fresh and pre-activated Co/SiO 2 , as well as reference Co samples; (b) DRIFT spectra of CO adsorption on different samples; (c) amount of leaching Co in reaction filtrate from pre-activated Co/SiO 2 (determined by ICP-OES).

Table 1 .
Catalytic performances of various catalysts in methoxycarbonylation reaction of acetylene a .

Table 3 .
Catalytic performances of filtrate and residual solids in methoxycarbonylation reaction of acetylene a .

Table 3 .
Catalytic performances of filtrate and residual solids in methoxycarbonylation reaction of acetylene a .