Spano et al. [9] studied the variety of nonlinear absorption coefficient β in nc-Si films with changing the excitation intensities in a range of 1 to 5 × 1012 W/cm2; they found that TPA process dominated the nonlinear EGFR inhibitor optical process under the various laser excitation intensities and the β decreased as increasing the excitation power. It was explained in term of the banding filling effect at high pumping power if the TPA process dominated the nonlinear optical

absorption process. However, the different intensity-dependent optical nonlinearities are observed in sample E in our case. As shown in Figure 6a,b, the NLA of sample E changes from RSA to SA with increasing the excitation intensity. However, sample D keeps the SA characteristic with changing the excitation intensity while the transmittance increased,

as shown in Figure 6a. As mentioned before, the SA process is sensitive to the density of interface states. For sample with small-sized nc-Si, the more interface states are introduced due to the larger surface-to-volume ratio. We also measured the PL properties of samples D and E as displayed in Figure 7 to illustrate it. It is clear to find that the sample E displays stronger PL intensity than sample D, and a broad Selumetinib molecular weight luminescence band in the range of 700 to 1,000 nm was observed, which was attributed to the interface state-related recombination and radiative recombination in the previous work [13]. The more interface states introduced in the gap, the larger the saturation irradiance I s will be. When the excitation intensity (I 1 = 3.54 × 1011 W/cm2) is lower than the I s, the TPA dominates the NLA. Whereas, when the excitation intensity (I 2 = 3.54 × 1012 W/cm2)

is higher than the I s, the SA process appears and the TPA is suppressed. However, there are still two small valleys at the wings of the open aperture transmission trace, suggesting the TPA and SA processes co-exist, which is consistent with our model proposed in Figure 5. Figure 6 Open aperture Z-scan traces of samples D and E. (a) Sample D and (b) sample E under two laser intensity, I 1 = 3.54 × 1011 W/cm2 (open square) and I 2 = 3.54 × 1012 W/cm2 (full square). The solid lines are the fitting curves of the experimental data. Figure 7 The PL spectra of sample D (black line) and sample E (red line). Then, Metalloexopeptidase we will discuss the NLR behaviors in our samples. Accompanying with the change of NLA, the NLR characteristics are also tunable as shown in Figure 3e,f,g,h. Samples A and B show the negative nonlinear refraction index (n 2) while samples C and D have the positive nonlinear refractive index. We calculated the n 2 from the measured closed aperture transmittance data by using Equation 3 [18]: (3) where ΔΦ0 = k 0 n 2 I 0 L eff represents the nonlinear phase change. The nonlinear refraction index n 2 of sample A is -3.34 × 10-12 cm2/W. Spano et al.