A Study of CO Oxidation over IR(110) with Surface Sensitive Probes

Author: Taylor, James Long

Year: 1979

Degree: Dissertation (Ph.D.)

Advisor: Weinberg, William Henry

Committee Member: Unknown, Unknown

Option: Chemical Engineering; Chemistry

DOI: 10.7907/9jhk-5j44

Abstract

The heterogenously catalyzed reaction of CO and O2 to form CO2 over (110), as well as the chemisorption of the individual reactants, has been studied at low pressures (<10-5 torr). The experiments were performed with four surface sensitive probes - thermal desorption mass spectrometry (TDS), contact potential difference (CPD) measurements, Auger electron spectroscopy (AES) and LEED. Oxidation of the surface complicated several aspects of this study; that is, a second, tightly bound, dissociatively chemisorbed state of oxygen forms irreversibly on Ir(110) near 700 K. Hence, several of the experiments were performed for both clean and oxidized surfaces of Ir(110).

Chapter 2 presents a method for analyzing thermal desorption spectra. The method stresses the coverage dependence of the activation energy and pre-exponential factor in an Arrhenius expression. Techniques for assessing, avoiding and offsetting errors in TDS are presented also. Although this method was used for determining rate parameters throughout this study, the method is applied specifically to the desorption of CO in this chapter. As the coverage increases from 0 to 0.75 ML (1 ML = 1015 cm-2), the activation energy decreases continuously from 35 to 20 kcal/mole whereas the pre-exponential factor decreases continuously from 1013 and 109 sec-1.

Chapter 3 considers the chemisorption of CO on Ir(110). The adsorption kinetics of CO are described by a second-order precursor model that accounts for ordering of the adlayer through intermolecular repulsion. The second-order model is apposite because two sites are needed for a CO molecule to adsorb, not because adsorption is dissociative. A model is developed to fit the coverage and temperature dependence of CPD measurements for CO. The CPD changes non-linearly with coverage because of the depolarization of the overlayer from dipole-dipole interactions and changes in temperature through the thermal population of translational and vibrational modes. Inferences about the binding sites for CO could be drawn from CPD measurements because the CPD and the translational motion of the adsorbate are related. Also, a method is presented for evaluating the error in measurements of isosteric heats that arise when the Clausius-Clapeyron equation is evaluated for points of constant CPD, rather than constant coverage.

Chapter 4 describes the chemisorption of oxygen on Ir(110). Oxygen may chemisorb molecularly, may chemisorb dissociatively or may bind into the oxide layer. Above 200 K, oxygen chemisorbs dissociatively since the activation energy for dissociating molecularly chemisorbed oxygen is about 8 kcal/mole. Oxygen desorbs molecularly with an activation energy between 45 and 70 kcal/mole, which decreases continuously with increasing coverage. A second-order precursor model consistently describes the chemisorption of oxygen; that is, the logarithm of measured lifetime for the precursor is inversely proportional to the surface temperature between 300 and 700 K. Moreover, oxygen chemisorption is not activated. Measurements of the CPD upon oxygen chemisorption, which were analyzed by extending the methods presented in the previous chapters, indicate that the overlayer undergoes a structural transformation at a fixed coverage, which depends on whether or not the surface is oxidized.

Chapter 5 discusses the reaction of gaseous CO and O2 to form CO2 over Ir(110). Four elementary reactions - the desorption of CO, the oxidation of CO via the langmuir-Hinshelwood mechanism, the adsorption of CO and the adsorption of oxygen - may limit the rate of CO oxidation under equilibrium conditions. The kinetics of each reaction are modeled from measurements involving TDS. The pre-exponential factors and activation energies for CO desorption and CO oxidation depend markedly upon the composition of the adlayer. Moreover, the diffusion of chemisorbed CO and oxygen may limit the kinetics of these processes. The co-adsorption of CO and oxygen is not, in the strictest sense, competitive; that is, CO blocks sites for oxygen adsorption, but oxygen does not block sites for CO adsorption. Despite these complications, a model developed from the kinetic expressions for the elementary reactions qualitatively and quantitatively predicts trends in the rate of CO2 production under equilibrium conditions. These trends include the dependence of the rate upon surface temperature and the partial pressures of the reactants. However, hysteresis in the "equilibrium" rate of CO2 production, which cannot be attributed to the oxidation of the Ir(110) surface, limits the applicability of this model.

Appendix B details an interactive programming language, POL, for monitoring experiments with a PDP-11/10 minicomputer. This language allows the user to control data acquisition, experimental parameters, the storage of data and the display of data, as well as to analyze data in an on-line environment. This language was used to expedite all experiments described above.

Files