No abstract

Density functional theory (DFT) calculations on stoichiometric, high-symmetry clusters have been performed to model the (100) and (111) surface electronic structure and bonding properties of titanium carbide (TiC), vanadium carbide (VC), and titanium nitride (TiN). The interactions of ideal surface sites on these clusters with three adsorbates, carbon monoxide, ammonia, and the oxygen atom, have been pursued theoretically to compare with experimental studies. New experimental results using valence band photoemission of the interaction of O(2) with TiC and VC are presented, and comparisons to previously published experimental studies of CO and NH(3) chemistry are provided. In general, we find that the electronic structure of the bare clusters is entirely consistent with published valence band photoemission work and with straightforward molecular orbital theory. Specifically, V(9)C(9) and Ti(9)N(9) clusters used to model the nonpolar (100) surface possess nine electrons in virtually pure metal 3d orbitals, while Ti(9)C(9) has no occupation of similar orbitals. The covalent mixing of the valence bonding levels for both VC and TiC is very high, containing virtually 50% carbon and 50% metal character. As expected, the predicted mixing for the Ti(9)N(9) cluster is somewhat less. The Ti(8)C(8) and Ti(13)C(13) clusters used to model the TiC(111) surface accurately predict the presence of Ti 3d-based surface states in the region of the highest occupied levels. The bonding of the adsorbate species depends critically on the unique electronic structure features present in the three different materials. CO bonds more strongly with the V(9)C(9) and Ti(9)N(9) clusters than with Ti(9)C(9) as the added metal electron density enables an important pi-back-bonding interaction, as has been observed experimentally. NH(3) bonding with Ti(9)N(9) is predicted to be somewhat enhanced relative to VC and TiC due to greater Coulombic interactions on the nitride. Finally, the interaction with oxygen is predicted to be stronger with the carbon atom of Ti(9)C(9) and with the metal atom for both V(9)C(9) and Ti(9)N(9). In sum, these results are consistent with labeling TiC(100) as effectively having a d(0) electron configuration, while VC- and TiN(100) can be considered to be d(1) species to explain surface chemical properties.

Copper has been evaporated onto chemically different single-crystal surfaces of zinc oxide in ultrahigh vacuum to model Cu/ZnO methanol synthesis catalysts. The formation of the copper overlayers from less than 0.1 monolayer (ML) to several ML on the (0001), (0001), and (10 0) surface planes is followed with core-level X-ray photoelectron spectroscopy, valence band photoelectron spectroscopy with both resonance discharge sources and synchrotron radiation, and low-energy electron diffraction. At room temperature, the first monolayer grows in a two-dimensional fashion, with low coverage copper existing as isolated atoms/small islands on all surfaces. Surface perturbations show submonolayer copper supported on the Zn2+-terminated (0001) surface to be most prone to high-temperature clustering and reaction with molecular oxygen. Copper on the oxide-terminated (0001) surface is much less reactive, while copper on the (1010) dimer surface shows intermediate reactivity. Low-temperature carbon monoxide (CO) chemisorption experiments indicate that highly dispersed copper on the (0001) and (lOTO) surfaces chemisorb CO with approximately the same affinity as copper metal (A/fads = 15-16 kcal/mol), while chemisorption on Cu/(0001) was much weaker (A/7ads < 12 kcal/mol). High-affinity (A7/ads = 21 kcal/mol) CO chemisorption, often associated with the catalytic active site, is shown to occur at a coordinatively unsaturated tetrahedral Cu+ site created on the (0001) surface upon annealing in oxygen. Chemisorption to the Cu+ site perturbs the CO electronic structure to a much greater extent than chemisorption to either Cu°or Zn2+, with valence band PES indicating both strong and 7 interactions. The implications of these results with respect to CO activation and the catalytic activity of the Cu/ZnO system are discussed.with Cu/ZnO as compared to ~30 kcal/mol on ZnO.8 Metallic copper is reported to show no measurable activity.8,9 Additionally, a significant amount of CO chemisorbs to the Cu-promoted catalyst more strongly than to either Cu or ZnO, indicating that the Cu/ZnO combination contains a detectable number of surface

Titanium carbide (TiC) and titanium nitride (TiN) possess remarkable physical properties, such as extremely high hardness and melting point, that promote their use as antiwear materials under harsh tribological conditions. These physical properties must arise from chemical bonding phenomena that result from the inclusion of the non-metal atom within the metallic matrix, and these bonding phenomena should be apparent in measurements of the valenceband electronic structures of TiC and TiN. This paper explores the surface electronic structure and bonding in TiC(lOO) and TiN(l 10) with core and valence level photoelectron spectroscopies (PES's) using X-rays (1486.6 eV) and synchrotron radiation in the range 28-180 eV. Intensity changes in the valence-band features are followed as a function of incident photon energy; these changes are then compared to theoretical atomic photoionization cross sections to determine the atomic origins of these features. Resonant PES at the Ti 3p absorption edge is used to determine titanium 3d contributions to the valence band and to show differences in the electronic structures in TiC and TiN. A new resonance phenomenon near the Ti 3s edge in TiC was observed, and its possible assignment is discussed. The electronic structure and bonding in these materials is well described by molecular orbital theory, where the Ti and non-metal ions in their formal oxidation states (e.g., Ti4+ and C4-in TiC) undergo covalent bonding interactions. Overall, the PES results indicate greater covalent mixing for TiC as compared to TiN, consistent with the differences in the electronegativities of the atoms. Specifically, stronger covalent interactions between the C 2s, 2p and the Ti 3d, 4s, 4p levels must occur to explain the spectroscopic differences between TiC and TiN. In addition, there is no evidence for an occupied TiC valence level having predominantly Ti character (unlike TiN), precluding the existence of direct Ti-Ti bonding in TiC. Any such orbital overlap is significantly affected by the carbon atoms in the lattice.

The interaction of water with two transition metal carbides, titanium carbide (TiC) and vanadium carbide (VC), has been investigated. The adsorption, reaction, and desorption of water on the (100) face of singlecrystal samples of these materials have been studied as a function of substrate temperature over the range 100-600 K. The adsorption state of water on these surfaces has been probed with high resolution electron energy loss spectroscopy (HREELS). The reactivity of water has been directly measured with HREELS and X-ray photoelectron spectroscopy (XPS). The desorption of molecular water and the products of surface reactions has been followed with temperature programmed desorption. Collectively, these measurements indicate that water adsorbs both molecularly and dissociatively on TiC and VC; however, a greater degree of reactivity at cryogenic temperatures is observed on TiC. Dissociation of water produces surface bound hydrogen and hydroxyl groups on both surfaces and a fully dissociated surface oxide on TiC. Furthermore, a greater participation of the surface carbon atoms is observed at the TiC surface through the evolution of CO x species at elevated temperatures. The differences in surface bonding and desorption profiles are discussed in terms of differences in electronic structure of the two metal carbides. Some possible implications of these studies for the use of TiC and VC as tribological materials are also discussed.

of inhibitor. At a fivefold excess, 3 inhibited approximately 50% of both thrombin and factor Xa activity; this level of inhibition remained relatively unchanged for more than 20 h.'* However, under similar conditions, the inhibition curves for 4 with thrombin and factor Xa showed marked reactivation (Figures 2 and 3). These data suggest that the enzyme deacylation rate is slow but real. Experiments in progress with 4 will address the question of reversibility of enzyme acylation and will provide rates of deacylation.Several aliquots of both a-thrombin and factor Xa inhibited with compound 4 were subjected to photolysis. Irradiation of 4 itself resulted in lactonization of the inhibitor to produce 3-methylcoumarin and pamidinophenol. Since neither the coumarin nor the phenol are good inhibitors of serine proteinases,% photolysis of enzyme solutions containing compound 4 effectively removed inhibitor from the system. Enzyme solutions inhibited with low concentrations of 4 (0.6-10 1 M ) were fully reactivated in less than 1 h of irradiation, but photolysis of samples containing higher inhibitor concentrations (30-97 mM) resulted in only partial reactivation (50-65%) after the same period of irradiation. The longer photolysis periods which would be required to photoisomerize higher concentrations 4 (30-100 mM) were not feasible in presence of enzyme, since all proteinases also absorb energy within the mercury-xenon broad band emission spectrum. Control studies confirmed that prolonged irradiation of a-thrombin or factor Xa did result in decreased enzyme activity.Although chymotrypsin preferably cleaves peptide bonds at aromatic amino acids (phenylalanine, tryosine, and tryptophan), information on its affinity for 4 was needed to better target future chymotrypsin photoactive inhibitors. The altered specificity of this proteinase can be readily accounted for by an important difference in the amino acid sequences of chymotrypsin and trypsin-like enzymes such as trypsin, thrombin, and factor Xa. In general, trypsin and chymotrypsin accommodate substrates within their active sites in exactly the same manner. However, trypsin contains a key aspartic acid residue which can form a salt bridge with the charged amino acid side chains of lysine or arginine. The formation of these ionic bonds conveys added specificity to trypsin peptide bond hydrolysis. In chymotrypsin, this residue has been replaced by Ser189.27 Since the pK, of the serine hydroxyl is normally around 14, this amino acid does not form ionic bonds to charged amino acid side chains. Thus, pamidinophenyl esters such as 4 are not readily recognized by chymotrypsin. The speed and efficiency of all photoreactivation experiments with derivative 4 were limited by the (1) extensive overlap between enzyme and inhibitor absorbance spectra and (2) the intensity of the light source. As noted earlier, the extended periods of broadband irradiation required to lactonize high concentrations of 4 eventually degraded the enzyme. Use of a monochromator tuned to appropriate emissio...

The reaction of methanol on the (100) surfaces of single crystal vanadium carbide (VC) and titanium carbide (TiC) has been studied using high resolution electron energy loss spectroscopy (HREELS), temperature programmed desorption (TPD), and X-ray photoelectron spectroscopy (XPS). Methanol forms a mixed monolayer of molecular methanol and a methoxy intermediate upon adsorption at 153 K on both VC(100) and TiC(100). With increasing temperature, methanol is evolved from both surfaces through molecular and recombinative desorption. Approximately half of the methoxy intermediate reacts with the VC surface to produce formaldehyde and hydrogen, with a small amount of methane and persistent oxygen surface species. By contrast, very little of the methoxy intermediate reacts with the TiC surface, producing methane and hydrogen. A model of the surface reactions has been constructed based upon differences in the electronic structures of the carbide substrates.

No abstract