Polyvinyl chloride (PVC), manufactured with vinyl chloride monomer (VCM) as raw material, is widely used in many engineering areas [1, 2]. At present, the VCM is generally synthesized by acetylene hydrochlorination catalyzed by carbon-supported HgCl₂ in China. However, HgCl₂ is volatile and highly toxic, which is harmful to human health and the environment [3]. Therefore, the research on mercury-free catalysts for acetylene hydrochlorination is extremely urgent [4-6]. Gold catalysts have high activity among all mercury-free catalysts; however, they have not been widely applied in chloride alkali plants due to the high cost of gold and the ease of sintering during the reaction [7, 8]. Ruthenium catalysts are much less expensive with excellent sintering resistance, which makes them promising candidates as mercury-free catalysts for acetylene hydrochlorination [9, 10].

The stability of the mercury-free catalyst is very important for its industrial application in acetylene hydrochlorination. The identification of the deactivation mechanism is very important for the exploration of practical catalysts for industries [11-19]. The experiments and density functional theory (DFT) calculations confirmed that the metal ions are more inclined to adsorb acetylene rather than HCl and, through the triple bonds in acetylene, interact with metal ions to form p–s bonds that activate the acetylene. If the activated acetylene cannot react with HCl timely, the activated acetylene species are inclined to be polymerized or dehydrogenated into deposited carbon [20, 21]. To improve the stability of the ruthenium catalyst, researchers have taken many approaches to suppress the deactivation of the catalyst. Recently, some researchers have proposed that the coke deposition can be suppressed by activating HCl [22-24], which can suppress the polymerization of acetylene by accelerating the reaction between HCl and acetylene. Our research group found that the acidic site of the ruthenium catalyst can induce the polymerization of acetylene, which would result in the deactivation of the catalyst [25]. The stability of ruthenium catalysts can be greatly improved using the ammonia-assisted impregnation method.

The electronic structure of metal ions in supported metal catalysts is the main factor affecting the performance of catalysts [26-29]. Until now, several strategies such as ligand coordination or strong metal-support interaction have been reported to regulate the electronic structure of metal ions. [30-32] Tian et al. investigated the coordination behavior between the theophylline (THP) molecule and Au(III) ion. The stability constant, K_f, of the THP-Au(III) ion is 2.7 × 10³⁵ [33]. This indicates that regulating the electronic structure of gold ions can effectively suppress their reduction. Zuo et al. designed a series of Cu(I) complexes and found that the charge transfer of Cu(I) complexes occurs from the metal to the ligand through DFT calculations [34]. Suljagić et al. [35] found that the O, N, and S elements in imine, which coordinate with copper (II) ions, are donor atoms. In general, the electronic structure of metal ions can be effectively regulated by electron transfer between ligand and metal ions [36-41].

Up to now, no study on the regulation of the electronic structure of active sites in the ruthenium catalyst for acetylene hydrochlorination has been reported. In the present work, the effects of ligands with different coordination atoms (Cl, O, N, and S) are studied in coordination with ruthenium catalysts for acetylene hydrochlorination reaction. Special attention is paid to the effect of the electronic structure of the ruthenium ion on the catalytic performance of a carbon-supported ruthenium catalyst in acetylene hydrochlorination.

Activated carbon (AC) (coconut carbon, 12–24 mesh) was purchased from Hainan Yeqiu Co., Ltd. Ruthenium chloride hydrate (RuCl₃·3H₂O) was purchased from Sino-Platinum Metals Co. Ltd. Other reagents were obtained from Shanghai Chemical Reagent Inc. of Chinese Medicine Group.

AC was washed with a 1 mol/L HCl solution at 70 ℃ for 5 h to remove impurities, followed by washing with water to neutral pH and drying at 120 ℃ for 10 h. The dried AC was stored in a desiccator for further use.

Ruthenium catalysts (1.0 wt% Ru) were prepared by the following steps. Firstly, RuCl₃·3H₂O was dissolved in deionized water. Subsequently, a thiourea solution (the molar ratio of thiourea to ruthenium is 4) was gradually added to the ruthenium chloride solution under stirring, and the solution was kept for 3 h. Finally, the solution changed from brown to mazarine, and the above mixture solution was impregnated with AC at room temperature for 24 h, after which the impregnated sample was dried at 120 ℃ for 10 h. The catalyst was denoted as Ru-Thi/AC. The Ru-Phe/AC (phenanthroline as the ligand), Ru-Lac/AC (L-lactic acid as the ligand), and RuCl₃/AC (no ligand) with 1.0 wt% Ru loading were prepared using the same protocols as that for Ru-Thi/AC.

The catalytic performance was evaluated in a fix-bed glass reactor (i.d. of 10 mm). Prior to the reaction, purged nitrogen was imported into the microreactor to separate water and air from the catalyst. Clean HCl gas (99.99% purity) was passed through the reactor to pretreat the catalyst at a flow rate of 7.3 mL/min at a temperature of 170 ℃ for 4 h. Subsequently, acetylene (99.9% purity) was passed through a concentrated sulfuric acid solution to remove trace impurities. Clean acetylene (6.7 mL/min) and HCl (7.3 mL/min) were introduced into the heated reactor containing 1.0 mL of the catalyst with a C₂H₂ gas hourly space velocity (GHSV) of 400 h^‒1 at 170 ℃, and the feed volume ratio of V_HCl/V_C2H2 was 1.10. Finally, the reactor effluent was passed through a sodium hydroxide solution to remove the unreacted HCl. The gas effluents were analyzed using a GC-1690F gas chromatograph (GC) equipped with an FID detector and a GDX-301 column.

Ultraviolet-visible (UV-vis) absorption spectroscopy was performed using UV-7600S (METASH, China). During the experiments, the prepared solutions were injected into a standard quartz cuvette and scanned over the wavelength range of 200–800 nm. The concentration of the ligands aqueous solution is 1 × 10^‒5 mol·L^‒1.

Nitrogen adsorption isotherms were determined on a Quantachrome Autosorb-IQ apparatus. The specific surface area was obtained using the Brunauer-Emmett-Teller (BET) model for adsorption data in a relative pressure range of 0.05–0.30. The pore-size distribution was acquired from the desorption branches of the isotherms using the DFT model.

X-ray photoelectron measurements (XPS) were conducted on a Kratos AXIS Ultra DLD instruments using 300W Al K_α. The binding energy was calibrated by the contaminant carbon (C 1s, 284.6 eV).

X-ray powder diffraction (XRD) measurements were performed with a Rigaku D/Max-2500/pc powder diffraction system using Cu K_α radiation (40 kV and 100 mA) over the range of 10° ≤ 2θ ≤ 80°.

Transmission electron microscope (TEM) images of the samples were obtained using a FEI Tecnai G20 instrument. The samples were mounted and ultrasonically dispersed in ethanol, after which few droplets of the suspension were deposited on a copper grid coated by a carbon film, followed by drying under ambient conditions.

The temperature-programmed desorption (TPD), including ammonia temperature-programmed desorption (NH₃-TPD), acetylene temperature-programmed desorption (C₂H₂-TPD), acetylene and HCl temperature-programmed surface reaction (C₂H₂-HCl-TPSR), and HCl temperature-programmed desorption (HCl-TPD), for the ruthenium catalysts were carried out with a self-made TPD instrument. The mass spectra were collected by an on-line Hiden gas analyzer (QIC20). For the NH₃-TPD of various ruthenium catalysts, ammonia was introduced by a He stream containing 10 vol% ammonia at 100 ℃, and the physically adsorbed ammonia was removed by purging with an Ar flow at 100 ℃ until the baseline was flat. Subsequently, the temperature program was raised from 100 to 500 ℃ at a rate of 10 ℃ min^‒1 under Ar at a flow rate of 30 mL/min. The following mass signals were monitored simultaneously using a quadrupole mass spectrometer: m/e = 16, 17, 18, 28, and 40 amu.

For the C₂H₂-TPD of various Ru catalysts, prior to acetylene adsorption, the samples (200 mg) were purged at 120 ℃ for 1 h in a flow of Ar; subsequently, acetylene was introduced at 120 ℃ for 12 h. The physically adsorbed C₂H₂ was removed by purging with Ar flow at 120 ℃ until the baseline was flat. After cooling to room temperature, the sample was heated to 500 ℃ at a ramp rate of 10 ℃/min under Ar at a flow rate of 30 mL/min. The following mass signals were monitored simultaneously by a quadrupole mass spectrometer: m/e= 24, 25, 26, and 40 amu.

For the C₂H₂-HCl-TPSR of various ruthenium catalysts, the samples (200 mg) were activated at 120 ℃ for 1 h in a flow of Ar; subsequently, acetylene was introduced at 120 ℃ for 8 h. The physically adsorbed C₂H₂ was removed by purging with an Ar flow at 120 ℃. Thereafter, HCl was introduced at 180 ℃ for 8 h. After cooling to room temperature, the sample was heated to 500 ℃ at a ramp rate of 10 ℃/min under Ar flow at a flow of 30 mL/min. The following mass signals were monitored simultaneously by a quadrupole mass spectrometer: m/e=24, 25, 26, 35, 36, and 40 amu.

For HCl-TPD of various ruthenium catalysts, prior to HCl adsorption, samples (200 mg) were purged at 120 ℃ for 1 h in a flow of Ar; subsequently, HCl was introduced at 180 ℃ for 18 h. The physically adsorbed HCl was removed by purging under Ar flow at 180 ℃ until the baseline was flat. After cooling to room temperature, the sample was heated to 500 ℃ at a ramp rate of 10 ℃/min under Ar at a flow rate of 30 mL/min. The following mass signals were monitored simultaneously by a quadrupole mass spectrometer: m/e=35, 36, 37, and 40 amu.

Thermogravimetric analysis (TGA) was conducted using a TG-DTG simultaneous thermal analyzer (NETZSCH STA 449F3 Jupiter®, Selb, Germany) under air atmosphere at a flow rate of 30 mL/min. The temperature was increased from 30 to 850 ℃ at a rate of 5 ℃/min.

In the present work, four catalysts were studied: RuCl₃/AC (reference catalyst), Ru-Lac/AC (L-lactic acid as the ligand), Ru-Phe/AC (phenanthroline as the ligand), and Ru-Thi/AC (thiourea as the ligand). The above catalysts with various ligands were prepared via a simple incipient impregnation method, which is shown in Scheme 1. The aqueous solution of ruthenium chloride and ligands was mixed and impregnated with AC; afterward, the products were dried and used as fresh catalysts directly. The textural properties and dispersion of ruthenium in the ruthenium catalyst were characterized in detail and discussed below.

The AC with a surface area of 1235 m²/g and a pore volume of 0.53 cm³/g was used as the support. After loading of RuCl₃ with various ligands, the surface area and pore volume of the ruthenium catalysts were slightly decreased. However, from Fig. S1, it can be seen that all the samples have similar isothermal curves and pore-size distributions. The isothermal curves are all typical type IV curves with an H4 hysteresis loop in a high relative pressure range, which indicates that these samples have mesoporous and micropore structures. The above results on the loading of RuCl₃ with various ligands do not alert the main pore structures of carbon supports. The small decrease in the surface area and pore volume of the AC after the impregnation of the ruthenium complex occurs because parts of the pores of ACs are filled by the ruthenium complex.

XRD, TEM, and STEM techniques were applied to investigate the dispersion of ruthenium in the ruthenium catalyst. As shown in Fig. 1, there is no discernible reflection of ruthenium or other species diffraction peaks apart from the two diffraction peaks at 24^o and 43.5^o, which can be assigned to the (002) and (101) diffraction peaks of amorphous carbons. The results indicate a high dispersion of ruthenium species in the ruthenium catalysts. Similarly, the TEM images of the ruthenium catalyst are shown in Fig. 2. There are no aggregated ruthenium particles observed at both low and high-resolution modes. However, the distribution of ruthenium is clearly observed from elemental mapping. This indicates a highly dispersed status of ruthenium chlorides on the carbon support.

The catalytic performance of the above-mentioned ruthenium catalysts prepared with various ligands was tested in acetylene hydrochlorination reaction (GHSV_C2H2 = 400 h^‒1, T = 170 ℃, and V_HCl/V_C2H2 = 1.10). As shown in Fig. 3(A), the initial conversions of acetylene for RuCl₃/AC, Ru-Lac/AC, Ru-Phe/AC, and Ru-Thi/AC are 54.2%, 45.6%, 61.4%, and 85.1% respectively, and Ru-Thi/AC exhibits the best catalytic activity. The catalytic performances of AC and AC-supported thiourea were also tested (results are given in Fig. S2). From Fig. S2, it can be seen that there is no significant difference in the catalytic performance of AC and AC-supported thiourea, which indicates that the ligand itself has no activity in this reaction. The turnover frequencies (TOFs) of the ruthenium catalysts were calculated, and the results are shown in Table 1. The TOF of the ruthenium catalysts is in the order of Ru-Thi/AC > Ru-Phe/AC > RuCl₃/AC > Ru-Lac/AC. The kinetic experiment was applied, and the activation energy was calculated based on the reaction data of various temperatures. Fig. 3(B) shows the linear fitting of the Arrhenius equation (detailed information is shown in Fig. S3). The apparent activation energies of Ru-Lac/AC, RuCl₃/AC, Ru-Phe/AC, and Ru-Thi/AC are 35.19, 31.88, 26.17, and 21.54 kJ/mol, respectively.

The decomposition of ligands in ruthenium catalysts may affect the stability of the catalyst. Therefore, the thermal stability of the catalyst was studied by TPD techniques under Ar. The results of Ar–TPD (Fig. S4) show that the ligand begins to decompose at temperatures higher than 200 ℃; it is known that the reaction of acetylene hydrochlorination often occurs at temperatures lower than 180 ℃, indicating that the ligand in the ruthenium catalyst will not decompose during the reaction. As shown in Fig. 3(A), the deactivation rates of these catalysts are noticeably different. The deactivation rates of the ruthenium catalysts were calculated, and the results are shown in Table 1. The deactivation rate of Ru-Thi/AC was the lowest among all the catalysts tested. The acetylene conversion of the Ru-Thi/AC catalyst was still above 82.4% after the catalyst had been used for 30 h.

XPS was employed to identify the status of ruthenium species. The high-resolution Ru 3p spectra and deconvolution results for the ruthenium catalysts are given in Fig. 4(A) and Table 1, respectively. For RuCl₃/AC and Ru-Lac/AC, the Ru 3p peaks are fitted into three species, which are Ru, Ru³⁺, and Ruⁿ⁺, respectively. The Ru 3p peaks of Ru-Phe/AC and Ru-Thi/AC are fitted into two peaks, which can be assigned as Ru³⁺ and Ruⁿ⁺. As shown in Table 1, there are small amounts of metallic ruthenium in Ru-Lac/AC and RuCl₃/AC. Notably, the position of the binding energy peaks for these ruthenium catalysts is in the following sequences: Ru-Lac/AC > RuCl₃/AC > Ru-Phe/AC > Ru-Thi/AC (Table 1). The binding energy position of ruthenium ions was related to the TOF and activation energy, and an interesting linear relationship is shown in Fig. 4(B). The binding energy reflects the electron deficiency of the metal ions, and the shift of the binding energy can give the direction of the electron transfer between metal ion and ligands or supports [42, 43]. Therefore, the above results indicate that the catalytic performance is related to the electronic structure of the metal ions.

The high-resolution Cl 2p spectra and deconvolution results for the ruthenium catalysts are given in Fig. S5 and Table S1. As shown in Table S1, the contents of the Ru-Cl species in Ru-Thi/AC, Ru-Phe/AC, and Ru-Lac/AC are lower than that in RuCl₃/AC. This indicates that the chloride ions of RuCl₃ could be replaced by the ligands. To further identify the coordination behavior of the ligand and ruthenium ions, UV–vis absorption spectroscopy was applied to characterize the as-prepared catalysts. As shown in Fig. 5, there is an absorption peak at the 236 nm UV region corresponding to C=S (n→p* transition) of thiourea. For the ruthenium and thiourea mixed solution, the absorption peak of C=S in thiourea exhibits a blue-shift, which is ascribed to the increase in the energy gap between n and p*. This indicates that S in thiourea is coordinated to the ruthenium ion. For phenanthroline, the absorption peaks at 226 nm and 265 nm correspond to C=N (n→p* transition) and C=C (n→p* transition), respectively. For the ruthenium and phenanthroline mixed solution, the absorption peak of C=N in phenanthroline exhibits a blue-shift, which was caused by the increase in the energy gap between n and p*. The results show that N in phenanthroline is coordinated with the ruthenium ion. For lactic acid, there is no absorption peak observed in the UV region, because the absorption peaks of the OH and COOH (n→s* transition) groups are at the far UV region (0–200 nm). For the ruthenium and lactic acid mixed solution, there is an absorption peak in the UV region (298 nm), which is due to the decrease in the energy gap between n and s*. This shows that O in lactic acid is coordinated with the ruthenium ion.

Summarily, the electronic structure of the metal ions in the ruthenium catalyst can be effectively controlled using various ligands. Furthermore, the catalytic performance toward acetylene hydrochlorination is related to the binding energy position of Ru³⁺ in the ruthenium catalyst.

It has been reported that acetylene rather than HCl is more inclined to adsorption on metal sties [22, 23]. The adsorption state of acetylene is inevitably affected by the electronic structure of metal ions in the ruthenium catalyst. Therefore, the adsorption status of acetylene on the ruthenium catalysts was investigated by C₂H₂-TPD. As shown in Fig. 6(A), there are two desorption peaks of acetylene, which indicate the two kinds of active sites for acetylene on the ruthenium catalysts. Interestingly, when the binding energy of Ru³⁺ shifts to a lower position, the desorption temperature of acetylene gradually shifts to a lower temperature, and the amount of adsorbed acetylene gradually decreases. This result further explains the previous sequence of the activity with the binding energy position of Ru³⁺ of the catalysts since the activation of acetylene becomes easier on the ruthenium catalyst with a lower Ru³⁺ binding energy position, which will result in enhanced catalytic performance. C₂H₂-HCl-TPSR was applied to investigate the reaction performance of the two adsorbed acetylene species. As shown in Fig. S6, The C₂H₂-HCl-TPSR results show that the peak area of adsorbed acetylene at a low temperature (ca. 150 ℃) decreased notably after reaction with HCl, while the peak area of the adsorbed acetylene at a high temperature (ca. 260 ℃) had no noticeable change. The results show that only the acetylene species, which desorbed at low temperature are active for acetylene hydrochlorination.

HCl-TPD was used to study the adsorption status of HCl on the ruthenium catalysts. As shown in Fig. 6(B), there is only one desorption peak of HCl at 145 ℃ for AC, RuCl₃/AC, and Ru-Lac/AC. The results reflect that the adsorption of HCl can occur on the carbon support, and no new adsorption peak appears for RuCl₃/AC and Ru-Lac/AC. Interestingly, there are two desorption peaks of HCl observed at 145 and 190 ℃ for Phe/AC, Thi/AC, Ru-Phe/AC, and Ru-Thi/AC. The additional new peak observed for these samples at around 190 ℃ indicates that both phenanthroline and thiourea can adsorb HCl as well. This is understandable since phenanthroline contains pyridine nitrogen sites and thiourea contains –NH₂, and both can adsorb HCl.

NH₃-TPD was applied to characterize the acidity of the Ru-Lac/AC, RuCl₃/AC, Ru-Phe/AC, and Ru-Thi/AC catalysts. As shown in Fig. 7, there is only one NH₃ desorption peak at around 200 ℃. Since acetylene is also a basic probe molecule for acidic sites, the C₂H₂-TPD profile gives two distinguished sites of the catalysts, while the NH₃-TPD profile only gives one adsorption site. This is understandable since the alkalinity is stronger than that of acetylene. NH₃ as a probe molecule is unable to distinguish the acidic adsorption sites for acetylene with different acidic strengths. However, the acidity can be quantitatively determined via the NH₃–TPD technique. The concentrations of the acidic sites on the ruthenium catalysts (Table S2) are in the order of Ru-Lac/AC > RuCl₃/AC > Ru-Phe/AC > Ru-Thi/AC (exclusion of NH₃ from ligand decomposition by Ar-TPD shown in Fig. S4).

The TGA conducted under an air atmosphere was used to characterize the coke deposition for the ruthenium catalysts. As shown in Fig. 8, for both the fresh and used catalysts, the slight weight loss before 150 ℃ corresponds to the desorption of the existing water on the catalysts. Subsequently, there is noticeable weight loss for the used catalyst in the temperature range of 150–400 ℃, which is attributed to the combustion of the coke deposition. The weight loss at temperatures higher than 400 ℃ is due to the combustion of the carbon support. Therefore, the amount of coke deposition is calculated based on weight loss in the temperature range of 150–400 ℃. The amount of coke deposition for the ruthenium catalysts was calculated, and the results are shown in Table 1. The amounts of coke deposition for the Ru-Lac/AC, RuCl₃/AC, Ru-Phe/AC, and Ru-Thi/AC catalysts are 17.6%, 13.4%, 7.8%, and 3.6% (Table 1), respectively. The low content of coke deposition well explains the excellent stability of the ruthenium catalysts.

In brief, the excellent performance of the ruthenium catalyst with a low Ru³⁺ binding energy is analyzed in detail as follows. Compared to the case in RuCl₃/AC, the sulfur in thiourea coordinates with the ruthenium ion in Ru-Thi/AC, which causes the binding energy of the ruthenium ion to shift to a lower position. The activation of acetylene is easier for the ruthenium catalysts with low binding energy of Ru³⁺. Simultaneously, the thiourea in Ru-Thi/AC can adsorb HCl. Thus, the synergetic effect of the ruthenium ion and ligands in Ru-Thi/AC can accelerate the acetylene hydrochlorination reaction and suppress the polymerization of acetylene, which, consequently, enhance the catalytic performance of the Ru-Thi/AC catalyst.

In summary, we prepared a series of ruthenium catalysts with different electronic structures of the ruthenium ion using several ligands. The Ru-Thi/AC catalyst with thiourea as the ligand exhibited the best catalytic performance with high TOF and stability. The experimental results indicate that the adsorption behavior of acetylene is greatly affected by the electronic structure of the ruthenium ion in the ruthenium catalyst. The TOF and apparent activation energies of the acetylene hydrochlorination have a linear relationship with the binding energy of Ru³⁺ in the ruthenium catalysts. The synergetic effect of ruthenium ion and ligands play an important role in the acetylene hydrochlorination. The present study provides a simple method for regulating the electronic structure of supported metal catalysts with high catalytic performance achieved for the carbon-supported heterogeneous catalyst.