Design of carbon supports for metal-catalyzed acetylene hydrochlorination

For decades, carbons have been the support of choice in acetylene hydrochlorination, a key industrial process for polyvinyl chloride manufacture. However, no unequivocal design criteria could be established to date, due to the complex interplay between the carbon host and the metal nanostructure. Herein, we disentangle the roles of carbon in determining activity and stability of platinum-, ruthenium-, and gold-based hydrochlorination catalysts and derive descriptors for optimal host design, by systematically varying the porous properties and surface functionalization of carbon, while preserving the active metal sites. The acetylene adsorption capacity is identified as central activity descriptor, while the density of acidic oxygen sites determines the coking tendency and thus catalyst stability. With this understanding, a platinum single-atom catalyst is developed with stable catalytic performance under two-fold accelerated deactivation conditions compared to the state-of-the-art system, marking a step ahead towards sustainable PVC production.

3 size of 1.5-3.0 nm were obtained, regardless of the type of carbon support Supplementary Fig. 21).
Catalysts after use in acetylene hydrochlorination for 12-h time-on-stream (TOS) are labeled metal/C-12h.
The composite Pt/CeO2+AC catalysts were prepared from a 1:1 weight ratio of Pt/CeO2 and AC (2 g each) via milling for 30 min at 20 Hz (ball-milled sample) and 5 min at 5 Hz (physical mixture sample) in a Retsch MM 500 nano mixer mill using a 50 cm 3 stainless steel jar and 12 stainless steel balls (0.9 cm diameter, total weight of 48.5 g).

Catalyst Characterization
An overview of the multiple characterization techniques employed in this study and their purposes are summarized in Supplementary Table 2. Powder X-ray diffraction (XRD) was measured using a PANalytical X'Pert PRO-MPD diffractometer with Cu-K radiation ( = 1.54060 Å). The data was recorded in the 10-70° 2 range with an angular step size of 0.017° and a counting time of 0.26 s per step. Raman spectroscopy was carried out in a confocal Raman microscope (WITec CRM 200) using a 532 nm diode laser. The microscope was operated in the backscattering mode with a 100× objective lens and 4 mW power. N2 and CO2 sorption isotherms were measured at 77 K and 273 K, respectively, in a Quantachrome Autosorb-6B equipment after degassing of the solids at 523 K for 4 h. Scanning transmission electron micrographs (STEM) with a high-angle annular dark-field (HAADF) detector were acquired on an aberration-corrected HD2700CS (Hitachi) microscope operated at 200 kV. Samples were prepared by dipping the copper grid supporting a holey carbon foil in a suspension of the solid in ethanol and drying in air. X-ray absorption fine structure (XAFS) measurements at the Pt L3-edge were carried out at the SuperXAS beamline at the Swiss Light Source (Villigen, Switzerland). The incident photon beam provided by a 2.9 T superbend magnet was selected by a Si(111) channel-cut Quick-EXAFS monochromator. 6 The rejection of higher harmonics and focusing were achieved with rhodiumcoated collimating and torroidal mirrors, respectively, at 2.5 mrad. The beamline was calibrated using Pt foil. The area of sample illuminated by the X-ray beam was 0.5 mm×0.2 mm. The Pt/C catalysts (250 mg) were finely ground and mounted in a 3 cm long, 0.5 cm wide cylindrical tube and measured in transmission geometry along the length. All spectra were recorded in transmission mode at room temperature. The extended X-ray absorption fine structure (EXAFS) spectra were acquired with a 1 Hz frequency (0.5 s per spectrum) and then averaged over 5 min. The X-ray 4 absorption near-edge structure (XANES) spectra were calibrated by measuring Pt foil simultaneously with each sample using the ProQEXAFS software. 7 The XAFS spectra were analyzed using the Demeter software package. 8 The background signal before the Pt L3-edge was subtracted using a victoreen function (fitting range between -200 and -70 eV). The post edge signal was normalized to the step of one after fitting it in the region between 150 and 1300 eV after the edge. We fitted the k 3 weighted Fourier transformed signal and determined an amplitude reduction factor (S0 2 ) of 0.84 from EXAFS fit of the Pt metal foil. All EXAFS spectra were fitted for the first coordination shell in the k-range of 3-16.8 Å −1 and R-range of 1-3 Å. To fit the Pt-Pt, Pt-Cl and Pt-(O/C) scattering paths, Pt foil, H2PtCl6 and PtO2 were used as the references. X-ray photoelectron spectra (XPS) were acquired on a Physical Electronics Quantera SXI instrument using monochromatic Al-K radiation, generated from an electron beam operated at 15 kV, and equipped with a hemispherical capacitor electron-energy analyzer. The samples were analyzed at an electron take-off angle of 45° and a constant analyzer pass energy of 50.0 eV with a spectra resolution step width of 0.1 eV. To suppress sample charging during analysis, an electron and an ion neutralizer were operated simultaneously. The spectrometer was calibrated for the Au 4f7/2 signal at 84.0 ± 0.1 eV. The spectra were fitted by mixed Gaussian-Lorentzian component profiles after Shirley background subtraction. The Pt 4f signal was fitted with two components assigned to Pt IV (73.5 ± 0.2 eV) and Pt II (72.4 ± 0.1 eV). The Pt 7/2 and Pt 5/2 doublet was constrained to be separated by 3.3 eV and at an areal ratio of 4:3. 2 Two contributions assigned to C-Cl (200.0 ± 0.1 eV) and Pt-Cl (198.1 ± 0.2 eV) were fitted to the Cl 2p signal constraining the Cl 3/2 and Cl 1/2 doublet to an areal ratio of 2:1 separated by 1.6 eV. 9 The O 1s signals were fitted using three components assigned to adsorbed water (536.5-536.0 eV), C-O (532.8 ± 0.1 eV), and C=O (531.1 ± 0.1 eV). 10 The Ru 3p3/2 signal was fitted with three components assigned to metallic Ru(0) (461.25 ± 0.15 eV), RuO2 (462.75 ± 0.15 eV), and RuCl3 (464.25 ± 0.15 eV). 11 The surface chemistry of the carbon supports was characterized with a SDT 2960 thermobalance from TA Instruments coupled to a Balzers GSD 300 T3 Thermostar mass spectrometer (TGA-MS).
After purging (T = 303 K, t = 20 min, FT = 20 cm 3 min −1 , flowing He), the desorption was initiated by increasing the temperature (T = 533 K, Ṫ = 5 K min −1 , flowing He) while monitoring the desorbed products by MS. To quantify the fraction of acetylene desorbing at temperatures >473 K (reaction temperature), we integrated (i) the whole data range between 303 K and 533 K (indicated as total acetylene adsorption capacity) and (ii) the fraction of the data range between 473 K and 533 K (referred to as acetylene adsorption capacity above 473 K). To assess the impact of HCl treatment on the acetylene adsorption capacity, the samples (0.1 g) were dried (T = 533 K,

Catalytic Evaluation
The hydrochlorination of acetylene was evaluated at atmospheric pressure in a continuous-flow fixed-bed micro-reactor ( Supplementary Fig. 24). The gases C2H2 (PanGas, purity 2.6), HCl (Air Liquide, purity 2.8, anhydrous), Ar (PanGas, purity 5.0, internal standard), and He (PanGas, purity 5.0, carrier gas), were fed using digital mass-flow controllers (Bronkhorst) to the mixing unit, equipped with a pressure indicator. A quartz micro-reactor of 10 mm inner diameter containing a porous frit was loaded with the catalyst (Wcat = 0.1 g for initial catalytic activity tests and kinetic tests, and 0.1-0.25 g for stability tests) and placed in a homemade electrical oven. A K-type thermocouple fixed in a coaxial quartz thermowell with the tip positioned in the center of the catalyst bed was used to control the temperature during the reaction. Prior to catalytic tests, the catalyst was heated (bed temperature, Tbed = 473 K, t = 30 min, flowing He). Thereafter, a total gas flow, FT = 15 cm 3 min −1 , containing 40 vol.% C2H2, 44 vol.% HCl, and 16 vol.% Ar, was fed into the reactor at bed temperatures, Tbed = 453-483 K, employing a high gas hourly space velocity 6 based on acetylene, GHSV(C2H2) = 650-1500 h −1 to assess the catalysts under accelerated deactivation conditions. Kinetic studies of acetylene hydrochlorination over Pt/C catalysts were conducted (Tbed = 453-483 K, FT = 20 cm 3 min −1 , 10-30 vol.% C2H2 and HCl in He) to determine the apparent activation energy (Ea) and the partial reaction order of the reactants (nC 2 H 2 , nHCl).
Carbon-containing compounds (C2H2 and C2H3Cl) and Ar were quantified on-line via a gas chromatograph equipped with a GS-Carbon PLOT column coupled to a mass spectrometer (GC-MS, Agilent, GC 7890B, Agilent MSD 5977A). For the 13 C isotope labeling study, the reactor outlet was directly connected to a Pfeiffer Vacuum Thermo-Star GDS 320 T1 MS. Since vinyl chloride (VCM) was the only product detected in all our tests, the catalytic activity is presented as After the tests, the reactor was quenched to room temperature in He flow and the catalyst was retrieved for further characterization. All catalytic data points were determined as an average of at least three measurements. The evaluation of the dimensionless moduli based on the criteria of Carberry, 12 Mears, 13 and Weisz-Prater, 14 confirmed that all the catalytic tests were performed in the absence of mass and heat transfer limitations (see for details Supplementary Discussion and   Supplementary Table 15). The deactivation constants, kD, were derived via a simple linear regression of the data range within the first 3 and 12 h time-on-stream in acetylene hydrochlorination (kD,3h and kD,12h, Supplementary Table 13).

C Isotope Labeling Study
To assess whether carbon could be directly involved in the catalytic cycle of acetylene hydrochlorination (e.g., via a Mars-van Krevelen mechanism), a metal-free nitrogen-doped carbon catalyst 1 was synthesized using 13

Assessment of Mass and Heat Transfer Limitations
The Carberry criterion (Ca) 4  , n the reaction order, and v,obs r and a' denote the reaction rate and the specific particle area, which are derived via Eq 7 and Eq. 8, respectively, To assess external temperature differences ( e T  ), 4 Eq. 9 was applied, where e  denotes the external Prater number, b T the temperature in the bulk material (473 K), h the heat transfer coefficient (estimated at a minimum value of 10 J m −2 s −1 K −1 ), and r H  the reaction enthalpy (99.3 kJ mol −1 ). Internal mass transfer limitations were evaluated using the Weisz-Prater criterion (  ), 5  where  is the tortuosity factor (estimated at 3),  is the particle porosity (estimated at 0.2), 22 C H ,HCl D is the molecular diffusion coefficient (1.96 10 -5 m 2 s −1 ), and K D is the Knudsen diffusion coefficient which is calculated for components i (i = HCl, C2H2) in a cylindrical pore according to Supplementary Fig. 9. Cl 2p XPS spectra of a, fresh and b, used Pt/C catalysts.
Supplementary Fig. 11. O 1s XPS spectra of a, fresh and b, used Pt/C catalysts.
Supplementary Fig. 13. a, Acetylene adsorption capacity determined by volumetric chemisorption at 303 K and 473 K. b, Correlation between acetylene adsorption capacity and surface area of Pt/C catalysts.
Supplementary Fig. 15. Acetylene adsorption capacity of Pt/CeO2 and Pt/AC4 before and after exposure to HCl (T = 473 K, t = 30 min, flowing HCl). While the pre-treatment only marginally affects the acetylene interaction with the carbon-based catalyst, the ability of the ceria-based catalyst to adsorb acetylene vanishes, as a consequence of the extensive chlorination of the support. 23 Supplementary Fig. 17. Kinetics of acetylene hydrochlorination on selected Pt/C catalysts. a, Arrhenius plots used to derive the apparent activation energy, Ea. b,c, Reaction rate, r, as a function of the inlet partial pressure of C2H2 or HCl. The partial reaction order of both reactants, n, is indicated by the slope of the fitting lines. Reaction conditions: Tbed = 473 K, FT = 20 cm 3 min −1 , Wcat = 0.1 g, and P = 1 bar. In order to circumvent the influence of catalyst deactivation, each point was obtained in a single experiment, after 15 min TOS.
Supplementary Fig. 18. Thermogravimetric analysis in 20 vol.% O2/Ar of fresh (colored profiles) and used Pt/C catalysts (black profiles). The difference in weight loss between the fresh and used samples suggests a comparable amount of coke deposits in the two catalysts, estimated at ca. 4 wt.%, which is well in line with their similar content of acidic oxygen functionalities.
Supplementary Fig. 20. Evolution of the m:z 62, m:z 63, m:z 64 ions over 13 C-labeled Ndoped carbon as a function of time-on-stream in acetylene hydrochlorination as determined by MS analysis. The solid black line indicates the moment when the C2H2 and HCl feed was stopped.
Supplementary Fig. 21. STEM of fresh Au/C and Ru/C catalysts. The gold nanostructure strongly depends on the choice of the carbon host and varies from large nanoparticles (C1) to single atoms (AC). In the case of ruthenium, small nanoparticles were obtained, regardless of the type of carbon support. Particle size distributions were derived from analysis of >100 particles.
Supplementary Fig. 23. Correlation between the deactivation constant of Ru/C catalysts in acetylene hydrochlorination and the density of surface oxygen functionalities in the fresh support, estimated as the total content of oxygen functional groups evolving as CO2 and CO during TPD-MS normalized by the surface area.