Semiempirical Correlation between Optical Band Gap Values of Oxides and the Difference of Electronegativity of the Elements. Its Importance for a Quantitative Use of Photocurrent Spectroscopy in Corrosion Studies F. Di Quarto,* C. Sunseri, S. Piazza, and M. C. Romano Dipartimento di Ingegneria Chimica dei Processi e dei Materiali, UniVersita ` di Palermo, Viale delle Scienze, 90128 Palermo, Italy ReceiVed: December 26, 1996 X A semiempirical correlation between the optical band gap of binary oxides and the difference of electronegativity between the oxygen and metallic elements (Pauling’s extraionic energy) is proposed. In the frame of the proposed correlation an estimate of the repulsive term in the total lattice energy of ionic oxides is obtained in very good agreement with the existing data. An extension of the correlation to the ternary oxides and hydroxides is proposed by using the concept of average cationic or anionic group electronegativity. The usefulness of the proposed correlation for the in situ characterization of passive films on metals and alloys by photocurrent spectroscopy is illustrated by reporting some preliminary experimental results on Sn and AlTi alloy. Introduction The use of photocurrent spectroscopy (PCS) for the physi- cochemical characterization of passive films is very attractive owing to some advantages with respect to other in situ optical techniques. In fact, (a) it is less demanding in terms of surface finishing, thus allowing monitoring of surface changes in long- lasting corrosion processes and/or in heavily attacked metal surfaces, and (b) it does not need high light intensity, thus minimizing the risk of modifications of very thin passive films. Apart from intrinsic limits, due to the fact that only photo- active (semiconducting or insulating) corrosion layers can be scrutinized, PCS suffers limitations in providing direct informa- tion on the composition and structure of passive films. It is our opinion that the use of PCS in a more quantitative fashion requires the extension of interpretative models for the photo- electrochemical behavior of crystalline bulk materials to the case of very thin and often strongly disordered or amorphous materials. In previous papers 1 we have shown that some new features in the photoelectrochemical behavior of passive films are attributable to the amorphous nature and/or to the extreme thinness of the passive films covering the metallic substrates. In this paper we propose a general correlation between the optical band gap of oxide films on passive metals and the average single-bond energy in oxides calculated on the basis of Pauling’s electronegativity scale. 2 On such a basis a possible quantitative correlation between optical band gap and composi- tion of passive films grown on metals and alloys will be delineated, which allows us to use in a more quantitative way the PCS technique. A Correlation between Optical Band Gap of Binary Oxides and the Difference of Electronegativity of the Elements To predict the physical properties of inorganic compounds, like the width of the forbidden gap in semiconductors, without applying theoretical models based on very complex quantum mechanical calculations, different authors have proposed semiem- pirical models and correlations for calculating or predicting energy effects and energy gaps in inorganic solids. Despite their evident simplicity and widely discussed limitations, 3 Pauling’s equations for the calculation of single-bond energies and heats of formation of inorganic compounds, based on the concept of electronegativity of the elements, are still a touchstone for more sophisticated and recent theories. 3-5 A short review of previous works on the argument is reported in ref 7c. Starting points of our considerations are the results reported in two papers. In the first one Manca 6 correlates the optical gap energies of different semiconducting compounds having diamond or zinc- blende structure to the single-bond energy, E s , through the relationship where a and b are constants characteristic of the series of investigated compounds. The single-bond energy was assumed by the author as given by Pauling’s equation: where the first term on the right represents the nonpolar contribution and the second one the polar contribution to the bond energy of an A-B compound. X A and X B are the electronegativity values of the elements A and B in Pauling’s scale, while D A-A and D B-B represent, according to Pauling, 2 the bond energy of molecules A-A and B-B in the gas phase. Really the use of the arithmetic mean between D A-A and D B-B in eq 1 should be more correct (see Chapter 3 and eqs 3-12 of ref 2). The second paper we like to mention is that of Vijh, 7a who in an attempt to rationalize Manca’s results and the empirical correlation between band gap and enthalpy of formation of a number of inorganic compounds reported by Ruppel et al., 8 derived for uni-univalent ionic compounds (e.g. alkali halides) the following theoretical relationship between the energy gap, E g , and the bond energy, E s : This last parameter was assumed coincident with the heat of atomization per mole, in the case of alkali halides, or with the heat of atomization per equivalent in the case of polyatomic * E-mail: diquarto@dicpm.unipa.it. X Abstract published in AdVance ACS Abstracts, March 15, 1997. E g ) a(E s - b) D A-B ) (D A-A D B-B ) 1/2 + (X A - X B ) 2 [eV] (1) E g ) 2(E s - R) (2a) 2519 J. Phys. Chem. B 1997, 101, 2519-2525 S1089-5647(97)00046-1 CCC: $14.00 © 1997 American Chemical Society