Comments on the paper by S. Westerberg, C. Wang and G.A. Somorjai, entitled “Heat of adsorption of CO on Pt(1 1 1) obtained by sum frequency generation vibrational spectroscopy—a new technique to measure adsorption isotherms” [Surf. Sci. 582 (20

It has been claimed [S. Westerberg, C. Wang and G.A. Somorjai, Surf. Sci. 582 (2005) 137] that the CO coverage on Pt(1 1 1) is proportional to the intensity of the band of linear CO observed by sum frequency generation vibrational spectroscopy, this assertion being based on an assumed agreement between a literature value of the adsorption heat of CO on Pt(1 1 1) and that calculated by applying the Langmuir model to the adsorption isotherms obtained by assuming the proportionality mentioned above. We show here, in the first place, that Langmuir’s equation was wrongly applied by Westerberg et al., the value of the adsorption heat obtained with a correct application of Langmuir’s model to the assumed adsorption isotherms being 73.5 kJ mol−1 instead of that of 113 kJ mol−1. In the second place, and far more importantly, we emphasize that Langmuir’s model, based on a constancy of the adsorption heat, cannot be applied to the adsorption of CO on Pt, a system for which the adsorption heat decreases markedly with increasing coverage. Therefore, the above value of 73.5 kJ mol−1 for the adsorption heat of CO on Pt(1 1 1), although obtained through a correct application of Langmuir’s equation, is invalid. Finally, application of the thermodynamic, and therefore model-independent, Clausius–Clapeyron equation to the adsorption isotherms given by Westerberg et al. yields adsorption heats that increase markedly with increasing coverage, which conclusively proves that the assumed adsorption isotherms are essentially wrong, and therefore that the claimed proportionality between the CO coverage on Pt(1 1 1) and the intensity of the sum-frequency-generation band of linear CO does not obtain.