Spectrophotometric methods for the determination of the band gap energy of materials for Dye-Sensitized Solar Cells

The energy band gap (Eg) is one of the properties of a semiconductor material that determines its possible application in energy devices such as Dye-Sensitized Solar Cells. There are several methods for Eg determination, such as XPS, UPS, PL and optical spectroscopy. One simple way is given by UV-Vis absorption spectroscopy on powder in suspension. Although simple, this technique is strongly dependent on the suspension stability: the precipitation of the powder can in fact affect the measurement leading to incorrect results. In order to avoid these complications, diffuse reflectance spectroscopy can be considered as a valid alternative. The aim of this work was to develop an easy and accurate method for Eg determination for powder materials through the combination of diffuse reflectance spectroscopy and graphical methods. The Eg determination through diffuse reflectance is possible using the Kubelka-Munk equation (K-M) generally indicated as F(R) = (1-R)2/2R (where R is diffuse reflectance). In literature several different methods have been used to graphically determine Eg from this equation, without obtaining a unified and conclusive approach. For this reason in this work the different approaches were tested and verified on well know TiO2 semiconductor in order to assess the easiest and most reliable one. This was found to be the one that takes into account the electronic transitions that occurs inside the semiconductor material (Tauc equation). The method consists in combining the F(R) of the K-M equation with the one obtained applying the Tauc equation in the equation ahv= A (hv-Eg)n where hv is the photon energy, A is a constant and n is proportional to the electronic transition considered. In a material with a perfect diffuse scattering the F(R) value becomes equal to the absorption coefficient (a). Although very simple, this method required the knowledge of the correct value of the "n" exponent and it is therefore necessary to establish the most convenient method for its determination. The n value was here calculated through a suitable graphical method and the band gap value was extrapolated by plotting (F(R)hv)1/n versus hv at (F(R)hv)1/n equal to zero. This approach provided an accurate Eg value for the semiconductor model and was validated on other common systems (e.g ZnO) showing good agreement with literature.