Joint Research Activity 3
Microfocusing Devices

JRA3 focuses on the development of advanced diffractive x-ray optics for x-ray imaging at synchrotron sources and the upcoming FEL sources. An overview of the individual work packages and the involved contractors are given below. At present, the request for diffractive x-ray optics by far exceeds their availability, and the recent progress in the field have even increased this misbalance. The reason for the shortness lie in the fact that a wide interdisciplinary set of skills is required covering the fields of diffractive optics, thin film and nanofabrication technology, and x-ray physics. Only very few institutions in Europe work in this field at present. Therefore, the objective of this JRA3 project is to develop advanced diffractive x-ray optics for use in the spectral range from the soft to the multi-keV photon range. Four European facilities are involved in the JRA3 activity: BESSY in Berlin, ELETTRA in Trieste, MAXLAB / KTH Stockholm, SLS in Villigen. Each of these partners exploit their specific expertise in different work packages as listed below. The various task of the JRA3 are:

Task A: Theory and design of diffractive x-ray optics aims at calculating the different optical elements developed within the project. In particular, Task A1 aims at understanding the diffraction properties of high-aspect ratio volume zone plates by electrodymanic theory.

State-of-the-art Fresnel zone plates with an outermost zone width of 20 nm can be described by scalar diffraction theory neglecting the three-dimensional shape of the zone structures. According to this theory their diffraction efficiency scales as 1/m2 where m is the diffraction order. While keeping the zone height constant, the aspect ratio of the zones increases inversely with decreasing outermost zone width. For photon energies below one keV, it is shown by applying electrodynamic theory that scalar theory is no longer suited to describe zone plates with outermost zone widths below 20 nm and aspect ratios of about 10:1. Additionally, in contradiction to scalar theory electrodynamic theory predicts diffraction efficiencies in higher diffraction orders of up to 50%. This result is obtained for high aspect ratio zone structures tilted to the optical axis according to the local Bragg condition. The resolving power of zone plates can be increased in two different ways. The conventional way is to use smaller zone periods in the 1st order of diffraction. Another way is to use high orders of diffraction, because the obtainable Rayleigh resolution =1.22drn/m scales inversely with the diffraction order m used for x-ray imaging. In this project, we define a zone plate as a volume zone plate if their diffraction efficiency is significantly higher than predicted by scalar diffraction theory. We will show coupled wave calculations for volume zone plates with high efficiency and resolving power. A stack-process for fabricating volume zone plates is proposed (see Fig. 1). With this process, tilted zones with high aspect ratios can be approximated by fabricating zone plates with low aspect ratios on top of each other with slightly decreasing zone radii. As shown in Fig. 2, the efficiency converges with increasing number of layer to the optimal diffraction efficiency for ideally tilted zones.

Fig.1: Proposed fabrication process for zone plates with tilted zones and high aspect ratios.

Fig.2: Plot of the obtainable diffraction efficiency as a function of the number of layers and the local line-to-space ratio of the zones. Note that the diffraction efficiency can be increased by decreasing the line-to-space ratio and that the number of layers forming the volume zone plate has to be in the range of 5 10 for high efficiency.

Task B: E-beam lithography aims at developing techniques required for high precision e-beam lithography for small and large area diffractive x-ray optics.

After testing different resists a ZEP 7000 e-beam resist was found to be suited to expose nanostructures with a line-width around 20 nm. Linear gratings have been exposed by e-beam lithography at 30 keV. Afterwards the resist was developed. Structures with good quality have been obtained down to a period of 36 nm in these experiments (see Fig. 3).

Fig. 3: Developed ZEP 7000 resist structures. The period of the linear grating is 36 nm (about 22 nm lines and 14 nm spaces).

Circular gratings with a pitch of 50 nm were exposed in two steps (see illustration in Fig. 4). In a first step only every second ring of the circular grating pattern was exposed. Afterwards the sample was unloaded and reloaded into the e-beam system. In a second overlay exposure the missing rings were exposed. The smallest deviation of the pitch was 0.66 nm. In general, the obtained overlay accuracy was better than 10 nm for 50 nm pitches between the rings (see Fig. 5).

Fig. 4: Overlay-exposure of two circular grating pattern with constant period. The superposition of both patterns is illustrated on the right side.
Fig. 5: SEM micrograph of the superposition of two circular gratings with 25 nm line width (50 nm pitch).
Task C: Pattern transfer and materials aims at developing optimized manufacturing parameters for zone plates providing high spatial resolution. In particular, the goal is to develop high aspect ratio nanostructures in nickel and gold by electroplating.

In the JRA3 project we tested different polymers which serve as a plating mold for electroplating in nickel or gold. In particular, it is important that the structured mold withstand the surface tension during the plating process. We found that an irradiated epoxy-based resist, SU-8, is well suited for this purpose. Fig. 6 shows the so far best result for electroplated structures. The line with is 35 nm and the height 400 nm which results in an aspect ratio of 11:1. This new process will be used for the zone plate fabrication to achieve higher efficiency. In addition, we introduced pulse-reverse (PR) plating techniques to improve nickel uniformity over the zone plates. With constant direct current (DC) plating, the thickness of the plated metal normally decreases with increasing radius of the zone plate. By tailoring pulse sequences for an individual set of zone plate parameters such as diameter, outermost zone width and aspect ratio, the zone plate height profile can be controlled. Fig. 7 shows an example between CC and PR plating for a micro zone plate. This improved uniformity will increase the diffraction efficiency from smallest outermost zones.

Fig. 6: A scanning electron micrograph of 35-nm half-pitch 400 nm high nickel gratings (60 degree view angle).

Fig. 7: Thickness profiles of two zone plates (a) plated with optimized pulse-reverse plating and (b) plated at constant current. The two zone plates are otherwise identical with a 75 m diameter and an outermost zone width of 50 nm.
Task D: Characterization of the imaging performance of diffractive optical elements aims at measuring the imaging performance of the diffractive optical elements developed within the JRA3.

The imaging performance of the manufactured x-ray optics can be measured with the full-field x-ray microscope installed BESSY. For this purpose a so-called Siemens star pattern is used as an object. Its smallest structures are 26 nm wide. Fig. 8 shows an x-ray micrograph of this test object demonstrating the expected resolving power.

Fig. 8: Right: TXM image of Siemens star with minimum line width of 26 nm, imaged with 25 nm nickel zone plate. The white, drawn line in the image is indicating the region for the intensity profile plot, shown on the left. Left: Intensity plot profile. The plot shows that structures of 32 nm are visable at the location of the white line in the Siemens star image. The data is fitted with a sine of 64 nm period.

Contact: Gerd Schneider, BESSY, email: Schneider@bessy.de, phone: ++49 30 6392 3131