238
J. Eur. Opt. Society-Rapid Publ. 21, 24( 2025)
Figure 5. With the same source settings but different table rotation speed, a shift in absorption can be seen, while the refractive index and deposition rate stays constant.
combinations, we kept the process settings constant and only adjusted the table rotation speed, which is specified in seconds per rotation. This approach keeps the mixing ratio of the QNL within a set constant, but the thickness of the individual layers is altered. The lower the rotation time, the faster the table rotates, and the thinner the individual layers become. The transmission spectra of such a series are shown in Figure 5.
In the higher wavelength range, it can be observed that the measured spectra overlap, indicating that they must have the same refractive index as well as the same rate. In the lower wavelength range, the absorption edge shifts to shorter wavelengths for faster rotation speeds and correspondingly thinner individual layer thicknesses. Some of the layers were analyzed using HAADF STEM measurements. One of these layers is shown in Figure 6. This specific sample was coated with a table rotation speed of 6 s per pass for half an hour, resulting in 300 individual laminates and a total thickness of ~ 536 nm. Since we know the number of table rotations during the coating process and can determine via photospectrometry the thickness of the QNL stack, we are able to easily calculate the thickness of a single laminate pair. To calculate the thickness of each material within a laminate pair, we use the formula mentioned at the beginning. By knowing the refractive indices of the individual materials as well as from the overall block, it is possible to determine the volume fraction factor“ f”, which allows us to calculate the average laminate thickness of both SiO 2 and Ta 2 O 5. The ability to analyze these layers using STEM HAADF image allowed us to additionally verify this formula. The average calculated thickness for SiO 2 was found to be 1.35 nm, and for Ta 2 O 5 it was 2.18 nm, while the STEM measurements showed an average thickness of SiO 2 = 1.39nmandTa 2 O 5 = 2.12 nm, respectively, which shows a high level of agreement. A line scan is shown in Figure 7.
From the transmission and reflection spectra, the band gap of the layers was calculated using a Tauc plot. By using the software OptiChar, the refractive indices, extinction coefficients and physical thickness can be determined across a wide wavelength range from the transmission and reflection spectra. From the total thickness and the measured refractive index, the individual layer thickness of the Ta 2 O 5 laminates can be calculated using the presented formula for mixing refractive indices, since the number of table rotations during the 900 s coating time is known. The calculated band gaps are plotted against the Ta 2 O 5 laminate thicknesses in Figure 8.
The different colors represent different process settings and thus different mixing ratios. Points of the same color correspond to different rotation speeds and, accordingly, different Ta 2 O 5 thicknesses. All samples were annealed in ambient atmosphere at 450 ° C for 1 h to reduce potential defects. The refractive indices of the samples are also plotted against the Ta 2 O 5 laminate thickness in Figure 9.
3 Results
As described in the previous section, the refractive indices and extinction coefficients were determined for all coated samples. This made it possible to select QNL settings with desired values for multilayer systems and create designs using OptiLayer, an optical interference design software. For the first filter, we chose a setting that provides a significant difference in the absorption edge compared to pure Ta 2 O 5, while still maintaining a relatively high refractive index of n 1.8( QNL5), to design a short pass filter. We built a design with a transmission region in the UV range up to a transition range from 320 to 340 nm and a high reflection part from 340 to 380 nm. The transition edge was intentionally chosen to be close to the absorption edge