8254
Langmuir 2004, 20, 8254-8260
Identification of Surface States on Finely Divided Supported Palladium Catalysts by Means of Inelastic Incoherent Neutron Scattering Peter W. Albers,*,† Ju¨rgen G. E. Krauter,‡ D. K. Ross,§ Roland G. Heidenreich,⊥ Klaus Ko¨hler,⊥ and Stewart F. Parker| Degussa AG, Wolfgang Industrial Site, Department of Physical Chemistry, 63457 Hanau, Germany, Degussa AG, Exclusive Synthesis and Catalysts, 63457 Hanau, Germany, Centre for Materials Research, University of Salford, Joule Laboratory, Salford M5 4WT, United Kingdom, TU Mu¨ nchen, Institute of Inorganic Chemistry, 85747 Garching, Germany, and ISIS Facility, Rutherford Appleton Laboratory, Chilton, Didcot, Oxfordshire, OX11 0QX, United Kingdom Received March 29, 2004. In Final Form: June 10, 2004 The purpose of the present investigation was to utilize the inelastic incoherent neutron scattering (INS) technique to reveal changes at the surface of technical catalysts under the influence of hydrogen in gas/ solid interactions and during chemical reactions in a liquid-phase process. The formation and the properties of supported palladium hydride and changes of the hydrogen-related surface chemistry of the corresponding activated carbon supports in 20% Pd/C catalysts after short-term and long-term hydrogen cycling at different hydrogen pressures and temperatures were studied. The spectra indicate that hydrogenation of the activated carbon support by hydrogen spillover occurs to, partly, give a material that strongly resembles a-C:H (amorphous hydrogenated carbon). Indications for different relaxation phenomena and long-range phase coherence inside of supported particles of palladium hydride compared to hydrogenated palladium black were obtained. A 5% Pd/C catalyst after use in C-C coupling reactions, the Heck reaction of bromobenzene and styrene to stilbenes, was also studied after subsequent solvent extraction. Evidence for a preferential adsorption and accumulation of cis-stilbene at the catalyst surface was obtained. INS allows identification of a certain isomer from a complex reaction mixture preferentially adsorbed at the surface of a finely divided industrial heterogeneous catalyst.
Introduction Supported precious metal catalysts are commonly used in chemical technology and industrial catalysis. Typical applications include catalytic hydrogenation reactions for the synthesis of fine chemicals, intermediates, vitamins, or pharmaceuticals. Owing to the technological importance of these materials, their properties have been extensively investigated including carbon-based precious metal catalysts and the properties of the support materials such as activated carbon or carbon blacks themselves.1-6 The detailed identification of the specific properties in the topmost atomic layers of carbonaceous supports used for the preparation of (e.g.) Pd/C catalysts are of paramount importance in catalyst development as well as in quality control in the industrial production of catalysts. However, * Corresponding author. E-mail:
[email protected]. Tel: 0049-6181-59-2934. † Degussa AG, Department of Physical Chemistry. ‡ Degussa AG, Exclusive Synthesis and Catalysts. § Centre for Materials Research, University of Salford, Joule Laboratory. ⊥ TU Mu ¨ nchen, Institute of Inorganic Chemistry. | ISIS Facility, Rutherford Appleton Laboratory, Chilton. (1) Boehm, H. P.; Kno¨zinger, H. In Catalysis: Science and Technology; Anderson, J. R., Boudart, M., Eds.; Springer: Berlin, 1983; Vol. 4. (2) Bansal, R. C.; Donnet, J. B.; Stoeckli, F. Active Carbon; Marcel Dekker: New York, 1988. (3) Radovic, L. R.; Rodriguez-Reinoso, F. In Chemistry and Physics of Carbon; Thrower, P. A., Ed.; Marcel Dekker: New York, 1987; Vol. 25. (4) Toebes, M. L.; van Dillen, J. A.; de Jong, K. P. J. Mol. Catal. A: Chem. 2001, 173, 75. (5) von Kienle, H.; Ba¨der, E. Aktivkohle; Enke: Stuttgart, 1980. (6) Rylander, P. Catalytic Hydrogenation in Organic Synthesis; Academic: New York, 1979.
in studying the surface properties of finely divided, black materials such as activated carbons which originate from various natural sources, carbon blacks and also the final catalysts which are derived from these carbons, wellproven methods may come up against physical limits: with increasing sp2 character of the carbons, the use of infrared or Raman spectroscopies is limited by the strong absorption of electromagnetic radiation. Studies on strongly adherent molecular species directly at the catalyst surface by means of (e.g.) NMR to monitor changes of the catalyst surface in process can also be hampered by the electrical conductivity of these support materials. These limitations in gaining analytical information are complicated by the fact that the grades of activated carbons which are suitable as support materials in catalytic processes have to be prepared and modified by dedicated methods: (1) activation by different industrial procedures to obtain the adequate surface area, porosity, and pore size distribution; (2) purification (e.g., by acid treatments and elution processes) to remove ash, extractable sp3 material, and contaminants; (3) post-treatments and conditioning to prepare suitable surface chemistry for obtaining the best precious metal/ support interactions during impregnation and improved dispersibility in the given reaction media. During these processes, various inorganic and organic contaminants are removed and the contribution of sp3type species is lowered relative to aromatic/graphitic species. This is accompanied by increasing blackness value and, hence, strong absorption of electromagnetic radiation as mentioned above. Similar limitations can occur in the detection of chemical species in the topmost atomic layers
10.1021/la040054c CCC: $27.50 © 2004 American Chemical Society Published on Web 08/12/2004
Surface States of Supported Palladium Catalysts
Langmuir, Vol. 20, No. 19, 2004 8255
Table 1. Sample Numbers and Pretreatments sample no.
composition
1
20% Pd/C
2
20% Pd/C
3
20% Pd/C
4
palladium black
5
5% Pd/C
treatment before INS (1) 4× in situ cycling with H2 at 22 °C in the INS can up to 1.5 bar (2) INS measurement under 800 mbar H2 under sorption equilibrium (3) removal of hydrogen for 4 h at 200 °C in vacuo (4) INS measurement (1) 20× in situ cycling with H2 at 22 °C in the INS can up to 1.5 bar (2) INS measurement under 800 mbar H2 under sorption equilibrium (1) pressurized for 5 days under 300 bar H2 in an autoclave, during days 2-4 annealing at 300 °C, cooling for 1 day (2) transfer into the INS can (3) evacuation for 2 days (4) INS measurement after 4× in situ cycling with H2 at 22 °C in the INS can up to 1.5 bar, INS measurement under 800 mbar H2 under sorption equilibrium (1) after use in the Heck reaction in liquid phase, extracted (2) transfer into the INS can (3) evacuation for 2 days (4) INS measurement
of used precious metal/carbon catalysts. In studying phenomena of selective adsorption from a complex reaction mixture or in investigating deactivation phenomena, uncertainties can remain about the presence of strongly adsorbed entities which may not be extracted. A direct and nondestructive identification of these species is desirable but is also often hampered by the properties of the material. Furthermore, the detection of monolayer or submonolayer adsorption requires high sensitivity and selectivity in probing small amounts of molecular species. These limitations are circumvented by using inelastic incoherent neutron scattering (IINS, or for short, INS).7,8 The hydrogen-related properties of activated carbons and of carbon blacks can be identified. INS is also useful to identify the vibrational states of protons on and inside of small supported precious metal particles such as Pd or Pt. In a recent paper, the fine features in the region of the out-of-plane aryl-C-H wagging modes of the structural units of activated carbon catalyst supports which were derived from different natural sources and had varying hydrogen content were studied and shown to reflect both the origin of the material and the nature of the sites.8 The purpose of the present INS investigation was to utilize the INS technique to reveal changes at the surface of technical catalysts under the influence of hydrogen in gas/ solid interactions at low and high hydrogen pressure and during chemical reactions in a liquid-phase process. Experimental Section Materials. Steam-activated and acid-washed wood-based activated carbon was loaded with 5% and 20 wt % of palladium, respectively, in a standard procedure.7 The N2 surface area of the support was >900 m2/g. Its hydrogen content was determined using a LECO RH-404 hydrogen analyzer. The experimental value of 5600 ppm ((500 ppm) indicates an enhanced sp2 character and a comparatively low content of sp3-type species compared to activated carbons of hydrogen contents of about 10.000 ppm and more. Additional data on the surface chemistry of this material were given in ref 7, Table 1. The primary particle size of the supported Pd entities was determined by means of transmission electron microscopy. Average values of 5.4 nm for the 20% Pd/C catalyst and 4.5 nm for the 5% Pd/C catalyst were detected. High surface area palladium black (Umicore) with a N2 surface area of 35 m2/g was hydrogenated and was also measured as a reference to obtain the neutron spectrum of unsupported palladium under the same conditions as in the case of the supported Pd/C catalysts. This material consists of aggregates (7) Albers, P. W.; Burmeister, R.; Seibold, K.; Prescher, G.; Parker, S. F.; Ross, D. K. J. Catal. 1999, 181, 145. (8) Albers, P. W.; Pietsch, J.; Krauter, J.; Parker, S. F. Phys. Chem. Chem. Phys. 2003, 5, 1941.
and agglomerates in the range of about 100-1000 nm, formed by primary crystallites of about 20-80 nm. Inelastic Incoherent Neutron Scattering and Catalyst Pretreatment. The INS spectra were measured at the TOSCA spectrometer at the spallation neutron source ISIS (Rutherford Appleton Laboratory, U.K.).9,10 The resolution is now 1-1.5% in ∆E/E as compared to 2-3% for its predecessor TFXA. The beam size was 4.0 cm × 4.0 cm so that a representative macroscopic amount of a sample could be monitored in each single experiment. Sample cans of 99.9% Al with integral filter elements and welded bellows valves (Nupro) were used throughout this study after leak testing with helium gas. After transfer into an Al can, each sample was evacuated for 24 h using a turbomolecular pump and a two-stage rotary pump with a zeolite adsorption trap to remove highly volatile matter, solvents, and residual moisture. The potential influence of adsorption of oil vapor which might affect the INS results due to the higher H/C ratios of such species was, therefore, excluded. The high selectivity of INS to hydrogen reveals the proton-related properties of bulk quantities of activated carbons and catalysts on the scale of several tens of grams. The neutron scattering cross sections of the samples were such that less than a few percent of the incident neutrons were scattered. Multiple scattering events are thus negligible. The sample pretreatments and sample numbers are summarized in Table 1. The INS spectra of samples 1-4 are focused on the effects of gas-phase treatments. Samples 1, 2, and 4 were treated with hydrogen under in situ conditions inside of the INS sample containers. Each sample was slowly exposed to hydrogen (99.999% Messer-Griesheim) at 22 °C in steps of a few millibars each over several days using a gas volumetric device including capacitive pressure transducers (MKS Baratron). By means of thermocouples, the adsorption was controlled such that a pronounced macroscopic heating of the catalyst powder due to the heat of adsorption at the surface and the heat of absorption in the palladium particles did not occur. This was done to avoid an influence of pronounced heating on the precious metal dispersion. In the first step, the pressure in the can was limited to 100 mbar. Afterward, the can was pumped out again, ending with the vacuum of the turbomolecular pump. The procedure was repeated in a second hydrogenation cycle ending with 300 mbar. Afterward, the final hydrogenation/dehydrogenation cycles were carried out 4 times and 20 times up to 1.5 bar, respectively. After evacuation, the catalysts were loaded to the final equilibrium pressure of 800 mbar. The final pressure of
