M. Albert Vannice

A DRIFTS study of the formation of surface groups on carbon by oxidation

Abstract Diffuse reflectance spectroscopy (DRIFTS) was used to monitor the in situ oxidation of carbons after a heat treatment at 1230K under flowing Ar or H2 to remove sulfur impurities and provide oxygen-free carbon surfaces. After heating in Ar, a carbon black progressively oxidized at 623K using an O 2 Ar mixture yielded spectra very similar to that obtained from a carbon black oxidized in boiling HNO3. Oxidation by O2 gave evidence for initial formation of cyclic ethers which, upon further oxidation, underwent rearrangement to form ether-like structures cross-linking polyaromatic domains. More complete oxidation by either O2 or HNO3 produced bands assigned to cyclic anhydrides, lactones, ethers, and phenols; but no bands for carboxylic acids were observed, presumably because of the absence of water vapor under our conditions. Gasification by O2 of the HNO3-treated carbon enhanced the initial bands, and also produced quinone-like structures. Surprisingly, after a heat treatment at 1230K in H2, this carbon black provided no discernible spectroscopic features after oxidation by O2, even though it was gasified. Oxidation by O2 of an activated Saran carbon gave results similar to those for the Ar-treated carbon black, except that essentially no absorbance was observed in the carbonyl region.

Catalytic reforming of methane with carbon dioxide over nickel catalysts I. Catalyst characterization and activity

The reforming of methane with carbon dioxide was studied over nickel supported on MgO, TiO2, SiO2, and activated carbon. The influence of the support on catalyst activity and carbon deposition resistivity was markedly different ineach case. Although considerable formation of filamentous carbon was observed over Ni/SiO2 (confirmed by TEM and TPO), there was negligible initial loss of catalytic activity. The catalytic activity of Ni/C was very similar to that of Ni/SiO2, but no filamentous carbon appeared to be formed. In contrast to Ni/SiO2, substantially less coking was observed over either the Ni/TiO2 or the Ni/MgO catalysts. Evidence of strong metal-support interaction (SMSI) in the Ni/TiO2 catalyst indicated that large ensembles of nickel atoms, active for carbon deposition, are deactivated or removed by the presence of mobile TiOx species. The identification of two TiOx phases and a Ni5TiO7 phase was made possible by direct measurement of crystalline d spacings with TEM. The Ni/MgO catalyst was both active and very stable for up to 44 h time on stream. Chemisorption, XRD and TEM results indicate the formation of a partially reducible NiO—MgO solid solution, which appears to stabilize the reduced nickel surfaces and provide resistance to carbon deposition.

Metal-support effects on the intramolecular selectivity of crotonaldehyde hydrogenation over platinum

The vapor-phase hydrogenation of crotonaldehyde (CROALD) was studied at low temperature and low conversions over Pt powder and Pt dispersed on SiO2, η)-Al2O3, and TiO2 after either a high or a low-temperature reduction, HTR and LTR, respectively. The hydrogenation of butyraldehyde (BUTALD) and crotyl alcohol (CROALC) to butanol (BUTNOL) and the isomerization of CROALC to BUTALD were also investigated to obtain additional information about this network of reactions. The typical Pt catalysts produced 100% BUTALD, as expected from previous studies of CROALD and the awareness of the high hydrogenation activity of Pt for CC double bonds. However, the TiO2-supported Pt produced both CROALC and BUTNOL, and the best catalyst, sol(HTR) PtTiO2, gave a selectivity of 37% CROALC with no BUTNOL formation. In addition to this marked enhancement in selectivity, the turnover frequencies on the sol(HTR) PtTiO2 samples, based on sites counted by hydrogen chemisorption near reaction temperatures, were more than an order of magnitude higher than the catalysts displaying normal adsorption behavior. Activation energies were somewhat higher on the solPtTiO2 catalysts, however. This behavior is very similar to that observed previously for CO and acetone hydrogenation over these same catalysts, and it is consistent with the proposal that sites created at the Pt-titania interface are responsible for the activation of carbonyl bonds. Previous studies of liquid-phase hydrogenation of CROALD and other molecules with conjugated CC and CO double bonds support this model and indicate that a 1,4 diadsorbed species, rather than 1,2 and 3,4 diadsorbed species (counting the oxygen atom as 1), is formed in polar solvents. By analogy, defect sites on the titania at the metal-support interface may interact with the O atom, polarize the CO bond, and favor this intermediate which can react with hydrogen to give CROALC.

A kinetic and DRIFTS study of low-temperature carbon monoxide oxidation over Au—TiO2 catalysts

Titania-supported gold catalysts are extremely active for room temperature CO oxidation; however, deactivation is observed over long periods of time under our reaction conditions Impregnated AuTiO2 is most active after a sequential pretreatment consisting of high temperature reduction at 773 K, calcination at 673 K and low temperature reduction at 473 K (HTR/C/LTR); the activity after either only low temperature reduction or calcination is much lower. A catalyst prepared by coprecipitation had much smaller Au particles than impregnated AuTiO2 and was active at 273 K after either an HTR/C/LTR or a calcination pretreatment. Deposition of TiOx overlayers onto an inactive Au powder produced high activity; this argues against an electronic effect in small Au particles as the major factor contributing to the activity of AuTiO2 catalysts and argues for the formation of active sites at the AuTiOx interface produced by the mobility of TiOx species. DRIFTS (diffuse reflectance FTIR) spectra of impregnated AuTiO2 reveal the presence of weak reversible CO adsorption on the Au surface but not on the TiO2; however, a band for adsorbed CO is observed on the pure TiO2. Kinetic studies with a 1.0 wt.-% impregnated AuTiO2 sample showed a near half-order rate dependence on CO and a near zero-order rate dependence on O2 between 273 and 313 K with an activation energy near 7 kcal/mol. A two-site model, with CO adsorbing on Au and O2 adsorbing on TiO2, is consistent with Langmuir-Hinselwood kinetics for noncompetitive adsorption, fits partial pressure data well and shows consistent enthalpies and entropies of adsorption. The formation of carbonate and car☐ylate species on the titania surface was detected but it appears that these are spectator species. DRIFTS experiments under reaction conditions also show the presence of weak, reversible adsorption of CO2 (near 2340 cm−1) which may be competing with CO for adsorption sites.

Kinetics of catalytic reactions

Foreword Preface List of Symbols 1. Regular Symbols 2. Greek Symbols 3. Subscripts 1: Introduction 2: Definitions and Concepts 2.1 Stoichiometric Coefficients 2.2 Extent of Reaction 2.3 Rate of Reaction 2.4 Turnover Frequency or Specific Activity 2.5 Selectivity 2.6 Structure-Sensitive and Structure4nsensitive Reactions 2.7 Elementary Step and Rate Determining Step (RDS) 2.8 Reaction Pathway or Catalytic Cycle 2.9 Most Abundant Reaction Intermediate (MARI) 2.10 Chain Reactions 2.11 Reaction Rates in Reactors 2.12 Metal Dispersion (Fraction Exposed) 2.13 Meta1Support Interactions (MSI) References 3: Catalyst Characterization 3.1 Total (BET) Surface Area 3.2 Pore Volume and Pore Size Distribution 3.2.1 Hg Porosimetry Method 3.2.2 N2 Desorption Method 3.2.3 Overall Pore Size Distribution 3.3 Metal Surface Area, Crystallite Size, and Dispersion 3.3.1 Transmission Electron Microscopy (TEM) 3.3.2 X-Ray Techniques 3.3.2.1 Line Broadening of X-Ray Diffraction (XRD) Peaks 3.3.2.2 Extended X-Ray Absorption Fine Structure (EXAFS) 3.3.3 Magnetic Measurements 3.3.4 Chemisorption Methods 3.3.4.1 H2 Chemisorption 3.3.4.2 CO Chemisorption 3.3.4.3 02 Chemisorption 3.3.4.4 H2-02 Titration Techniques 3 3.5 Relationships Between Metal Dispersion, Surface Area, and Crystallite Size References Problems 4: Acquisition and Evaluation of Reaction Rate Data 4.1 Types of Reactors 4.1.1 Batch Reactor 4.1.2 Semi-Batch Reactor 4.1.3 Plug-Flow Reactor (PFR) 4.1.4 Continuous Flow Stirred-Tank Reactor (CSTR) 4.2 Heat and Mass Transfer Effects 4.2.1 Interphase (External) Gradients (Damkohler Number) 4.2.1.1 Isothermal Conditions 4.2.1.2 Nonisothermal Conditions 4.2.2 Intraphase (Internal) Gradients (Thiele Modulus) 4.2.1.1 Isothermal Conditions 4.2.2.2Nonisothermal Conditions 4.2.2.3 Determining an Intraphase (Internal) Effectiveness Factor from a Thiele Modulus 4.2.3 Intraphase Gradients (Weisz-Prater Criterion) 4.2.3.1 Gas-Phase or Vapor-Phase Reactions 4.2.3.2 Liquid-Phase Reactions 4.2.4 Other Criteria to Verify the Absence of Mass and Heat Transfer Limitations (The Madon-Boudart Method) 4.2.5 Summary of Tests for Mass and Heat Transfer Effects References Problems 5: Adsorption and Desorption Processes 5.1 Adsorption Rate 5.2 Desorption Rate 5.3 Adsorption Equilibrium on Uniform (Ideal) Surfaces-Langmuir Isotherms 5.3.1 Single-Site (Nondissociative) Adsorption 5.3.2 Dual-Site (Dissociative) Adsorption 5.3.3 Derivation of the Langmuir Isotherm by Other Approaches 5.3.4 Competitive Adsorption 5.4 Adsorption Equilibrium on Nonuniform (Nonideal) Surfaces 5.4.1 The Freundlich Isotherm 5.4.2 The Temkin Isotherm 5.5 Activated Adsorption References Problems 6: Kinetic Data Analysis and Evaluation of Model Parameters for Uniform (Ideal) Surfaces 6.1 Transition-State Theory (TST) or Absolute Rate Theory 6.2 The Steady-State Approximation (SSA) 6.3 Heats of Adsorption and Activation Barriers on Metal Surfaces: BOC-MP/UBI-QEP Method 6.3.1 Basic BOC-MP/UBI-QEP Assumptions 6.3.2 Heats of Atomic Chemisorption 6.3.3 Heats of Molecular Chemisorption 6.3.4 Activation Barriers for Dissociation and Recombination on Metal Surfaces 6.4 Use of a Rate Determining Step (RDS) and/or a Most Abundant Reaction Intermediate (MARl) 6.5 Evaluation of Parameter Consistency in Rate Expressions for Ideal Surfaces References Problems 7: Modeling Reactions on Uniform (Ideal) Surfaces 7.1 Reaction Models with a RDS Unimolecular Surface Reactions 7.2 Reaction Models with a RDS Bimolecular Surface Reactions 7.3 Reaction Models with a RDS Reactions between an Adsorbed Species and a Gas

Kinetics of liquid-phase hydrogenation reactions over supported metal catalysts — a review

Abstract The kinetics of liquid-phase hydrogenation reactions have been reviewed with a special emphasis on α, β-unsaturated aldehydes as reactants. These reactions can be complex and can be influenced by factors such as metal specificity, side reactions, and metal-support interactions as well as reaction parameters. The importance of results in the absence of heat and mass transfer limitations has been emphasized. Hydrogenation reactions are typically assumed to be structure-sensitive, but dependencies on metal crystallite size have been reported; however, this behavior has been attributed to side reactions which can inhibit activity. Finally, solvent effects can exist, but the effect of H 2 concentration in the liquid phase has infrequently been isolated. Thermodynamic arguments indicate that a solvent effect can enhance the surface coverage of hydrogen on the catalyst surface at a constant H 2 partial pressure, but in the absence of any solvent effects, surface coverage is independent of the liquid-phase H 2 concentration.

The effect of SMSI (strong metal-support interaction) behavior on CO adsorption and hydrogenation on Pd catalysts: II. Kinetic behavior in the methanation reaction

The specific activity of Pd in the methanation reaction can show a 200-fold variation depending upon the support used. In contrast, the CH4 turnover frequency is independent of Pd crystallite size, at least over the range of 3–30 nm particles. The most active catalyst is TiO2-supported Pd in which an SMSI state has been induced, and silica-supported Pd is the least active. Turnover numbers increase in the following order: PdSiO2 ⪡ PdSiO2-Al2O3, PdAl2O3 − PdTiO2 (448 K) < PdTiO2 (SMSI). All CO partial pressure dependencies were near zero, while H2 partial pressure dependencies shifted from unity to near one-half for the TiO2-supported catalysts. A reaction model for Pd is proposed which is consistent with this behavior, with ir spectra of CO adsorbed on these Pd surfaces, and with recent results in the literature on CO hydrogenation. This model involves no assumptions regarding CO coverage, and assumes CH4 forms from surface carbon produced by a hydrogen-assisted CO dissociation step. The higher N values for PdTiO2 in the SMSI state are consistent with this model and appear to be a consequence of much higher surface concentrations of hydrogen compensating for markedly reduced CO coverages, thereby resulting in a rate enhancement.

Furfural hydrogenation over carbon‐supported copper

Furfural hydrogenation over copper dispersed on three forms of carbon – activated carbon, diamond and graphitized fibers – were studied. Only hydrogenation of the C=O bond to form either furfuryl alcohol or 2‐methyl furan occurred at temperatures from 473 to 573 K. Reduction at 573 K gave the most active catalysts, all three catalysts had activation energies of 16 kcal/mol, and turnover frequencies were 0.018–0.032 s-1 based on the number of Cu0 + Cu+ sites, which were counted by N2O adsorption at 363 K and CO adsorption at 300 K, respectively. The Cu/activated carbon catalyst showed no deactivation during 10 h on stream, in contrast to the other two catalysts. A simple Langmuir–Hinshelwood model invoking two types of sites was able to fit all kinetic data quite satisfactorily, thus it was consistent with the presence of both Cu0 and Cu+ sites.

Benzene hydrogenation over supported and unsupported palladium: I. Kinetic behavior

Abstract The kinetics of benzene hydrogenation over supported Pd catalysts and unsupported Pd powder were examined under a wide range of reaction conditions. At temperatures below 433 K, apparent activation energies were routinely near 12.0 kcal mole −1 . The reaction order in hydrogen for supported catalysts increased from 0.5 to nearly 4 as the temperature increased from 353 to 573 K, and this same parameter for Pd powder showed a similar trend but was more temperature sensitive. The reaction order in benzene increased from zero to 0.8 over the same temperature range but a smaller difference was observed between the results obtained for supported and unsupported Pd. The activity of all the samples showed a temperature-dependent maximum near 495 K, but the unsupported Pd had a greater tendency to exhibit deactivation. which was attributed to carbonaceous surface species formed from reversible dehydrogenation of adsorbed benzene. The lowest turnover frequencies occurred on the Pd powder whereas values 10–50 times higher were obtained over low-temperature reduced Pd catalysts prepared using Cl-containing precursors and acidic supports such as SiO 2 Al 2 O 3 and TiO 2 . The enhanced activity over these last two catalysts is attributed to additional benzene adsorption sites present on the support surface in the metal-support interfacial region.

Methanol and methane formation over palladium/rare earth oxide catalysts

Lanthanide rare earth oxide (REO)-supported Pd systems can provide very good CO hydrogenation catalysts, especially for CH3OH formation. At low temperature and 1 atm, the major product was methanol but the selectivity shifted to methane at higher temperatures because of equilibrium constraints. The turnover frequency (TOP) values for CH4 varied more than tenfold, but all, except for PdEu2O3, were more than an order of magnitude higher on these catalysts than on PdSiO2 or pure Pd metal. Activation energies for CH4 on these PdREO samples were consistently near 32 kcal/mole and very close to the value on Pd powder, indicating that the methanation reaction mechanism may be similar on all these catalysts. At 14.6 atm total pressure, the major product over the PdREO catalysts was methanol and the selectivity for oxygenates ranged from 78 to 96% at 225 °C. Activation energies for CH3OH varied from 17 to 21 kcal/mole and were close to the value obtained for Pd powder. PdLa2O3 and PdNd2O3 were the most active CH3OH catalysts whereas PdCeO2 was the least active. All the PdREO catalysts deactivated initially at high pressure but reached steady state after about 30 hr on stream, with the stabilized activity being approximately half the initial activity in all cases. In contrast, the activity of Pd powder increased with time before reaching a steady state. The selectivity of Pd powder for oxygenates was lower compared to REO-supported Pd, and the CH3OH TOP over Pd powder was an order of magnitude lower than the least active PdREO catalyst. The role of these supports in CH3OH synthesis is consistent with that of promoting the concentration of formyl/formate intermediates on the surface, particularly at the Pd-support interface.