| A simple theoretical two-parabolic band model was developed to predict the imaginary part [Wherrett 1984] and the real part [Sheik-Bahae 1990] of the third-order optical susceptibility (chi(3)) in semiconductors. This model yields a universal curve for this nonlinearity as a function of the ratio of photon energy to bandgap energy. Experimental results on different materials at a few different wavelengths confirmed the universality of the model. However, no systematic study was done on a single material with a sufficient number of measurements at different wavelengths to verify the local shape of the curve. Also, Hutchings et al. calculated third-order susceptibilities of many semiconductors based on the Kane model [Hutchings 1992] that takes the band structure into account and expected deviations from the universal curve.; Using the Z-scan method [Sheik-Bahae 1989], we took 46 measurements of the complex third-order optical susceptibility for wavelengths ranging from 738 nm (0,62 Eg, where Eg is the semiconductor bandgap energy) up to 965 nm (0,48 Eg) on the same sample of polycrystalline ZnSe. Fitting the two-photon absorption coefficient (alpha2) measurements using the two-parabolic band model yields Eg = 2.73 eV and a value for the fitting parameter introduced by Sheik-Bahae of K = 5429 cm (eV)5/2/GW while Eg = 2.67 eV and K = 3100 cm (eV)5/2/GW was expected. Fitting the nonlinear refraction index (n2) measurements based on the same model yields Eg = 3.02 eV and K = 1577 cm (eV)5/2/GW. The two-parabolic band model describes well both series of measurement in general, but some deviations may be explained better with a more complete model.; Finally, we proved that it was impossible to measure the electronic chi (5) of ZnSe with the Z-scan technique using the equipment we had. |
