Gregory L. Griffin

Growth Kinetics of CVD TiO2: Influence of Carrier Gas

This paper reports on the growth rate of TiO{sub 2} thin films deposited by the decomposition of titanium tetraisopropoxide (TTIP) in the presence of N{sub 2} carrier gas. Experiments are performed at TTIP partial pressures between 0.005 and 0.7 torr and a substrate temperature of 573 K, conditions where second-order kinetics have previously been observed in the presence of TTIP alone. When 5 torr of N{sub 2} is present as a carrier gas, the kinetics become first order in TTIP concentration. By fitting the observed rates to the rate expression for the bimolecular reaction mechanism proposed in the earlier study, the authors obtain a value of {phi} = 0.43 for the relative efficiency of N{sub 2} for collisional energy transfer, compared to TTIP.

Photo-oxidation of methanol using MoO3TiO2: Catalyst structure and reaction selectivity

Abstract Laser Raman spectra, UV-visible spectra, and photo-catalytic activity measurements are described which show that supported MoO 3 TiO 2 catalysts prepared by impregnation possess a layer of surface molybdate species coordinated to the titania support. The saturation coverage of this surface molybdate monolayer corresponds to a surface density of ca. 4 × 1014 Mo atoms per cm2 of TiO2. The properties of this surface monolayer are markedly different from bulk MoO3, as reflected in different vibrational frequencies for the surface molybdate species and for bulk MoO3, and in the higher photo-catalytic activity of the surface molybdate relative to bulk MoO3. The photo-efficiency of the TiO2 catalyst containing the surface molybdate monolayer is about one-fifth of the photo-efficiency of pure TiO2, but the selectivity for suppressing secondary oxidation reactions is significantly greater. The major photo-assisted oxidation product on TiO2 is methyl formate, independent of methanol conversion. In contrast, the molybdate monolayer catalyst has nearly 100% selectivity for dimethoxymethane (the reversible condensation product of CH2O + 2 CH3OH) at low conversions. For both catalysts the selectivity of the primary oxidation step appears to be insensitive to the mode of excitation (i.e., thermal vs photo-assisted). Photo-induced changes in selectivity at a given conversion are due primarily to differences in adsorbate coverage at the different operating temperatures required to give equal conversion for dark vs light reactions.

Methanol synthesis activity of Cu/ZnO catalysts

Abstract We have measured the rate of CH 3 OH synthesis on a series of Cu ZnO and related catalysts prepared by different techniques. Under reaction conditions of 523 K, 50 atm, and H 2 /CO/CO 2 = 70/24/6, the CH 3 OH rate varies approximately linearly with Cu surface area. The specific activity is 0.03 g CH 3 OH/m 2 Cu/hr, in reasonable agreement with the results of other laboratories. Temperature-programmed desorption experiments using CO and H 2 , together with the results of previous IR studies, appear to deny the existence of any unique Cu surface sites beyond those associated with metallic Cu clusters that contain a mixture of high- and low-Miller-index crystal planes.

Coadsorption studies of CO and H2 on ZnO

We have studied the adsorption of pure CO and CO:H2 mixtures on powdered ZnO using the combined techniques of transmission infrared spectroscopy and temperature programmed desorption (TPD). When CO is adsorbed alone, the vibrational frequency ωCO decreases from 2192 to 2178 cm−1 with increasing CO coverage, and a repulsive CO–CO interaction is observed in the TPD spectrum. When CO is adsorbed on an H2‐covered surface, ωZnH decreases from 1709 to 1653 cm−1, ωOH increases from 3490 to 3523 cm−1, and the zero‐coverage limit of ωCO increases from 2191 to 2196 cm−1. There is also an increase in the CO adsorption energy due to an attractive CO–H2 interaction. Analysis of TPD spectra for CO yields an expression for the CO adsorption energy as a function of CO and H2 coverage: ΔHCOads (kcal/mol) = 12.2−0.16 nCO+0.08 nH2, where nCO and nH2 are the coverages in μmol/gm. We attribute the coverage dependence of the CO adsorption energy, as well as the observed IR frequency shifts, to both electrostatic and chemical i...

Combined temperature-programmed desorption and infrared study of H2 chemisorption on ZnO

Abstract The adsorption of H 2 on ZnO powders has been studied using the combined techniques of temperature-programmed desorption (TPD) and transmission infrared spectroscopy. The single previously observed ir-active state of dissociatively adsorbed H 2 , which has stretching frequencies at ν Zn-H = ~1710 cm −1 and ν O-H = ~3490 cm −1 , is found to desorb in two stages, characterized by maxima in the TPD spectra at ~240 and ~300 K. A new, low-temperature ir-active state has been observed with ν Zn-H = ~1710 cm −1 and ν O-H = ~3455 cm −1 , which has a TPD maximum at ~170 K. Additional measurements of adsorption isotherms and adsorption rates indicate that the two states with TPD maxima at 240 and 300 K have comparable binding energies of 12.6 ± 1.0 kcal/mole, and are distinguished by different activation barriers for adsorption. The low-temperature adsorption state, which also exhibits an activation barrier for adsorption, has a binding energy of ~7 kcal/mole, and cannot be populated at room temperature.

Adsorption studies of H2 isotopes on ZnO: Coverage‐induced IR frequency shifts and adsorbate geometry

The coverage dependence of the IR stretching frequencies for dissociative type I adsorption of H2 and D2 on ZnO powders has been measured using transmission IR spectroscopy. By comparing the frequency shifts observed when the isotopic composition of the adsorbate is varied at constant total coverage with the shifts observed when the total coverage of either pure component is varied, we can separate the dynamic and static contributions to the coverage‐induced frequency shifts. The ZnH and ZnD shifts are due primarily to electrodynamic interactions. The observed shifts are in good agreement with the Hammaker model for dynamic dipole–dipole interactions, if adsorption is assumed to occur on (2×2) reconstructed ZnO(0001) surface planes. In contrast, the OH and OD shifts are due to electrostatic and inductive interactions. The electrostatic contribution can be estimated using a modification of Buckingham’s treatment of local environment effects. A qualitative model based on electron localization effects is pre...

Gas-phase kinetics for TiO2 CVD: hot-wall reactor results

Abstract We have studied the growth kinetics of TiO2 chemical vapor deposition using the decomposition of Ti(i-OC3H7)4 (TTIP, titanium tetra-isopropoxide) in a hot-wall, axial flow low-pressure chemical vapor deposition reactor. Under conditions of high reactant conversion, we obtain polycrystalline, fully dense anatase TiO2 films at growth rates up to 0.2 μm h−1. The kinetic results (i.e., measured growth rates and axial film-thickness profiles) are analyzed using a two-dimensional reactor transport model and the gas-phase reaction mechanism that we proposed previously. This analysis yields a value for the rate constant of the gas-phase activation step (k1 [cm3 mole−1 s−1]= 4.0 × 1011 exp(-40 [kJ mole−1]/RT) that is consistent with the value obtained from our earlier kinetic experiments performed using a cold-wall, impinging flow reactor. In particular, the present results confirm the seemingly low value for the activation energy of the proposed gas-phase activation step obtained in our earlier work.

Direct alcohol synthesis using copper/cobalt catalysts

The authors have measured the yields of methanol, higher alcohols, and hydrocarbons during CO hydrogenation on a Cu/Co/ZnO/Al{sub 2}O{sub 3}/K catalyst. The overall yield of higher alcohols increases linearly with CO conversion, and the individual higher alcohol yields obey a Schulz-Flory distribution through n-C{sub 6}H{sub 13}OH. The selectivity for hydrocarbon products also increases with CO conversion. The product alcohols do not undergo a significant degree of secondary reaction at the concentrations formed under these conditions. The overall results are consistent with a mechanism for higher alcohol synthesis that involves chain growth of a common surface alkyl intermediate, followed by a chain termination step which determines whether the final product desorbs as an alcohol or hydrocarbon. The selectivity of the termination step shifts toward greater hydrocarbon formation with increasing CO conversion. 20 refs.

Methanol decomposition on CuZnO oriented thin films

We have studied the temperature programmed decomposition of CH3OH adsorbed on model CuZnO surfaces which have been prepared by evaporating small amounts of Cu onto RF plasma grown ZnO thin films. Desorption products assigned to the decomposition of CH3O(a) and HCOO(a) intermediates on exposed ZnO are observed respectively at 585 K (H2 and CH2O evolved) and 635 K (H2 CO and CO2 evolved). Products from CH3O(a) decomposition on metallic Cu clusters are observed at 410 K (H2 and CH2O). In addition to these results, which are consistent with the previously determined behavior of the separate components, we observe two new phenomena: The CH3OH adsorption step on the Cu clusters does not require pre-exposure of the sample to O2, and large CO2 desorption signals are observed at 510 K and 635 K. The former result is attributed to the spillover of H atoms onto O2− anions of the basic ZnO support, stabilizing the CH3O(a) species which remain on the Cu clusters. The enhanced CO2 desorption peaks are attributed to the decomposition of HCOO(a) species on Cu clusters (the 510 K peak) and at dispersed Cu cation sites on the ZnO surface (the 635 K. peak).

A Simple Phase Transition Model for Metal Passivation Kinetics

We present a minimal parameter kinetic model which describes the corrosion and passivation of metal surfaces. Two elementary steps are included, namely, the oxidation of surface metal atoms to produce adsorbed cations, and the subsequent dissolution of these cations into the electrolyte. Oxide layer formation is represented by two‐dimensional condensation of adsorbed cations to produce a continuous phase; statistics of the condensation process are treated at the mean field level. The structure and magnitude of the i‐V curve in the active‐passive transition region is calculated for two cases: the limit of maximum passivation hysteresis (spinodal decomposition of the oxide phase), and the limit of zero hysteresis (equilibrated phase transition). Methods for analyzing experimental i‐V curves to yield the rate constants associated with the model are presented.