236
J. Eur. Opt. Society-Rapid Publ. 21, 24( 2025)
Figure 1. In this figure, the bandgap of the two different materials, building a QNL system are shown [ 5 ].
valence electrons of the respective material. When a photon passes through these QNL, it experiences a periodic sequence of different potential energies, as shown in Figure 1.
If the layer thickness of the“ well” material is chosen to be sufficiently thin, an effect occurs where the bandgap of the entire system increases [ 1 ]. This can be understood as a restriction of the valence electron mobility within the material due to the thinness of the material layers, thereby making the absorption of light more difficult. The effective refractive index of the entire stack, on the other hand is only given by the volume ratio of the two materials and their own specific refractive indices [ 6 ].
n 2 QNL ¼ f n2 Barrier þ ð1 � f Þ n2 Well:
This means, that by structuring mixed materials into ultrathin layers instead of homogeneous mixtures, it is possible to obtain different bandgaps with the same refractive indices, since the band gap change depends only on the thickness of the well layers, but the refractive index depends on the mixing ratio, visualized in Figure 2.
2.1
Experimental setup
The properties of QNL have already been demonstrated several times in different experiments and working groups, using different coating technologies. As each individual layer of the two different materials is extremely thin, a large number of them are required to produce a layer with a thickness in the range required for optical thin film designs. This leads to a problem for many standard manufacturing methods when it comes to producing QNL stacks economically and efficiently. One reason for this is that different processes are required for the different materials used, which have to be carried out alternately. Each of these must be set, stabilized, carried out and ramped down again before the next one can begin, which can greatly reduce the coating rate. Furthermore, it is hardly possible to monitor individual layers, as the sensitivity of both optical monitoring and the oscillating quartz method does not allow such high resolution to reliably measure layers below one nanometer. Due to possible drifts in processes, there is always a high risk that the thickness of the material layers cannot be kept
Figure 2. The graph shows the expected effect of QNL. Both systems shown schematically have the same mixing ratio of materials and consequently the same refractive index. However, the thinner layer structure increases the band gap of the left system. The figure has been modified from [ 2 ].
constant in a stack of hundreds of them, which severely limits reproducibility. We benefit from a structural advantage provided by the Clusterline 200 BPM from Evatec, a magnetron sputtering system equipped with a rotating substrate table. To understand the benefits of this coating tool, it is necessary to first examine its working principles, as illustrated in Figure 3.
The white wall separating the clean room from the gray room is visible at the front. Next to it is the control panel, positioned beside two automatic load locks, each capable of holding sixteen 8-inch substrates. Behind these load locks lies the vacuum transfer chamber, equipped with a handler, an aligner for precise substrate positioning, and a flipper. This flipper allows substrates to be turned for backside coating without breaking the vacuum. In the background, the large coating chamber can be seen. It houses a rotating substrate table with a capacity for sixteen 8-inch substrates, as shown in the picture on the right. The sputter sources are mounted on the chamber, with the system supporting up to four sources, each equipped with a different target material. The additional plasma source( PSC) in the background is capacitively coupled and operates with various gases, mainly for etching, increasing density, or oxidizing the sputtered layers. In our experiments to coat QNL, we use the two magnetron sputter sources at the front, fitted with a Si-target and a Ta-target. For producing pure SiO 2 layers, we use the same Si-target in combination with the PSC source. To understand why our magnetron sputtering system is particularly well-suited for QNL production, it is helpful to first consider the standard process for creating single-layer materials. The system is designed for continuous table rotation during normal coating processes. Initially, the process gas is introduced, and the sputter source is activated with the shutter closed. Once the plasma stabilizes, the shutter opens, and the sputtering material is evenly deposited onto all substrates as the table rotates.