INFRARED SPECTRA OF CARBON MONOXIDE
343
Infrared Spectra of Carbon Monoxide Adsorbed on Some Evaporated Metal Films
by J. F. Harrod,' R. W. Roberts, and E. F. Rissmann General Electric Research and Development Center, Schenectady, New York (Received July 86, 1966)
Infrared spectra of CO adsorbed on evaporated thin films of Rh, Pd, Ir, and Pt have been measured. A technique was devised for measuring such spectra by transmission through several very thin films (ca. 100 A each) onto which CO was adsorbed. Initial studies to prove the feasibility of the method and to determine effects of contaminants such as HzO, 02,and HZon the spectra were performed with films evaporated at ca. torr and r e duced with Hz prior to adsorption of CO. The spectra of H-COOH and H-CHO on Rh, Pd, Ir, and Pt were also measured in such a system. After the feasibility of the multiple transmission technique was established, it waa adapted for use with an ultrahigh-vacuum system and spectra of CO adsorbed on atomically clean Rh, Pt, and I r were successfully measured. The latter spectra leave little doubt that those observed on contaminated samples of these metals are primarily due to CO and not some reaction product of GO with surface contaminants. Some preliminary kinetic studies of the oxidation of CO adsorbed on I r are also discussed.
Introduction During the past decade, many studies have been reported concerning the infrared spectra of carbon monoxide adsorbed on highly dispersed metal particle^.^-^ Such studies provide information on the structure of chemisorbed carbon monoxide, the effects of coadsorbates on the metal-GO interaction, the effect of the support material on metal-GO interaction, and the kinetics of a variety of chemical reactions involving either gaseous or adsorbed GO. Our initial interest in this subject was directed toward the use of infrared spectroscopy for studying rates of, and intermediates in, surface chemical reactions. The high extinction coefficient and strong chemisorption of CO made it particularly attractive for such studies, but there was serious doubt in our minds that the various methods used for preparing surfaces and observing spectra would be suitable for meaningful kinetic studies. The pressed-pellet method, first described by Eischens, lo suffered two possible serious disadvantages : (1) the rate of gas diffusion into and out of the pellet may be rate controlling, and (2) it is virtually impossible to obtain an atomically clean surface in this type of system. Our ultimate objective was to make measurements on
CO that had been adsorbed onto an atomically clean metal. Eischens had reported2 the successful observation of the transmission spectrum of GO adsorbed on a sputtered platinum film. Since the preparation of similar films, followed by the adsorption of GO onto them, seemed feasible as an ultrahigh-vacuum torr) experiment, we decided to investigate the utility of such a method. (1) To whom correspondence should be addressed at Chemistry Department, McGill University, Montreal, Can. (2) R. P. Eischens and W. A. Pliskin, Advan. Catalysis, 10, 1 (1958). (3) (a) A. C. Yang and C. W. Garland, J . Phys. C h m . , 61, 1504 (1957); (b) C. E. O'Neill and D. J. C. Yates, ibid., 65, 901 (1961). (4)J. B. Sardisco, Perkin-EZmr Instrument News, 15, No. 1, 13 (1963). (5)G. Blyholder, J . C h m . Phys., 36, 2036 (1962). (6) H.L. Pickering and H. C. Eckstrom, J . Phys. Chem., 63, 512 (1959). (7) R. A. Gardner and R. H. Petrucci, J . Am. Chem. SOC.,82, 50 (1960). (8) C. W. Garland, R. C. Lord, and P. F. Troiano, J . Phys. Chen., 69, 1188 (1965). (9) N. N. Kavtaradze, E. G. Boreskova, and V. I. Lygin, Kinetica i Kataliz., 2 , 378 (1961). (10) R.P. Eischens, 8.A. Francis, and W. Pliskin,J . Chem. Phys., 2 2 , 1786 (1964).
Volume 71, Number d
January 1087
344
Since we began our study, other workers have reported the difficulty that may be encountered in observing the transmission spectrum of CO adsorbed on metal films which have been evaporated under good vacuum conditions.8 These workers resorted to evaporating the metal in the presence of CO in order to obtain detectable spectra. Because occlusion and disproportionrLtion of CO on hot metal filaments are pos-~ sible under these conditions, we chose the alternative method of obtaining detectable spectra by means of stacking a number of very thin clean metal films in series in the beam of the spectrometer. We shall refer to this method as "multiple transmission" as opposed to the multiple-reflection method used by Pickering and Eckstrom.6
Experimental Section All measurements were made on a Model 21 PerkinElmer spectrometer, with variable attenuation in the reference beam. High-purity wires for filaments were obtained from Sigmund Cohn Inc., Aft. Vernon, N. Y. Minimum purities were Pd, 99.5% and Pt, Rh, and Ir, 99.9%. Gases were obtained from Airco Reduction. Preliminary Observations on Contaminated Metal Films. Preliminary experiments to determine the feasibility of using the multiple-transmission technique were performed using the apparatus shown in Figure 1. Films of the desired metal were evaporated onto 10mm x 1-mm circular NaCl disks. The films were evaporated from filaments in a bell jar at ca. 10-6 torr in such a way that both sides of the disks were coated with a film ca. 100 A thick. A stack of ten such disks was found to have a transmission of ca. 10% compared to ten uncoated disks at 2000 cm-'. After deposition of the films, the disks were removed from the bell jar and ten of them were stacked in the aluminum holder (A) with a ring spacer 0.5-mm thick between eaeh disk. The holder was placed inside the Pyrex cell and the cell was attached to a conventional vacuum system capable of rapidly achieving ca. torr. The cell was made from a standard 29/42 Pyrex ground-glass joint (F). The ends were flanged and closed by sealing on 1.75-in. rock salt windows (E) with epoxy resin. A piece of heating tape wrapped around the cell in the vicinity of the aluminum disk holder, and about 15 mm removed from the window, allowed heating of the sample to ca. 150". All pressures were measured with electronic gauges (1.0-760 torr, Alphatron gauge; 0.001-1.0 torr, thermistor gauge; and
