Index

Introduction
The Research at Rijnhuizen
Results in 2008
Education, Training, Outreach and Public Information
Output
Appendix


website Rijnhuizen
 2.8 | Physics of thin films and multilayers (TFM)

Division: nanolayer Surface and Interface physics
Group leader: A.E. Yakshin
Scientific advisors: P.C. Zalm, J. Verhoeven, A.W. Kleyn
Senio scientist: R.W.E. van de Kruijs
Postgraduate students: T. Tsarfati, V. de Rooij-Lohmann, S. Bruijn, J. Bosgra
Undergraduate students: R. Stijl
Collaborators: I. Kozhevnikov (Crystallography Inst. Moscow)
Funding: FP-I10, Carl Zeiss SMT, SenterNovem

Research programme
The general goal of the thin films and multilayers group is to develop the basic physics of thin single and multilayered films and interfaces of nanometer thicknesses. This includes some more specific physics processes occurring inside the multilayers as resulting from the different applications of these films. The research in 2008 followed two running research programmes: the FOM Industrial Partnership Programme ‘Extreme UV multilayer optics’ and the SenterNovem funded ‘Advanced multilayer coatings for high volume EUV lithography’ (‘ACHieVE’).

The aim is to develop and apply the physics required for a particularly advanced application of multilayer structures: optical elements operating at Extreme UV wavelengths, notably 13.5 nm. The structures investigated represent fundamental challenges in thin layer and surface physics as well as in multilayer optics. The multilayer systems are targeted to meet the requirements of Extreme UV Lithography, following the development road map of the associated industrial partners. The TFM group is dealing with fundamental solid-state aspects of these research projects related to ultrathin films. Other aspects are carried out at the SIPC and AXO groups.

Following the programme descriptions of the two research projects mentioned above, the programme of TFM for 2008 consisted of two parts: thin film growth, and diffusion studies and chemical layer passivation. This included studies on the formation and properties of interface layers, the growth and structuring of artificial diffusion barrier layers at the interfaces, and the diffusion processes at elevated temperatures. Mastering these processes in the group has led to discoveries of chemical and diffusion mechanisms in the multilayers at the nanometer scale. 

The research is carried out using a unique suite of thin film deposition and surface analysis tools, including state-of-the-art UHV e-beam and magnetron deposition facilities, various surface ion treatment equipments, Angle Resolved-XPS, Auger Electron Spectroscopy and Scanning Tunneling Microscopy as surface analysis set-ups, as well as CuK X-ray reflectometry.

Highlight: Physics and chemistry side by side on a nanometer scale
Because of their very high absorption, conventional lenses cannot be used for X-ray optics. Instead, curved molybdenum/silicium multilayer mirrors with a period of only a few nanometer are used in a Bragg geometry. Well-defined and smooth interfaces are essential for the reflectance of these layer-structured mirrors. However, the incident X-ray radiation heats up the mirrors and causes interdiffusion of the Mo and Si. The resulting molybdenum silicide formation reduces both the reflectance and the period of the mirrors, thus limiting their lifetime. To slow down this interdiffusion, so-called diffusion barrier layers are introduced in between the Mo and Si layers. Boron carbide (B4C) is one such material, and it has the benefit that it does not only improve the thermal stability, but increases the reflectance as well. Although these benefits are well-known, the chemical and physical interaction of the B4C with the Mo and Si are not. In order to understand these interactions, physical and chemical changes in c‑Mo/B4C/α‑Si samples were studied during annealing at 500°C.

Earlier use of Low-Energy Ion Scattering led to the discovery of a two-stage diffusion process: at a certain moment, the diffusion rate increased suddenly. Our latest studies of the chemical and morphological changes upon diffusion have now led to understanding of the mechanism behind the transition from stage 1 to stage 2.

The chemical aspect was addressed with High Kinetic Energy X-ray Photoelectron Spectroscopy (HIKE-XPS) measurements at the Berlin synchrotron facility BESSY. The data revealed interesting chemistry, e.g. that B4C is not an inert barrier material, but decomposes to form MoBX and 



Figure 2.9: The width of the B1s peak decreases significantly, indicating chemical changes. However, these changes are mostly well before the transition from the first to the second stage of the diffusion, which is marked by the dashed line. This hints at a non-chemical cause for the transition from stage 1 to 2.

MoCX and/or SiCX. One of the indications is shown in Figure 2.9: the B1s peak becomes narrower upon annealing. However, the decomposition of B4C cannot explain the increased diffusion rate in the second stage, as it occurs well before the transition to the second stage.

Possible physics aspects of the transition were addressed with Cross-sectional Transmission Electron Microscopy (X-TEM) analysis. These measurements disclosed that the cause of the transition is crystallisation of the molybdenum silicide, the compound that is formed at the interface. As shown in Figure 2.10, this layer is fully amorphous just before the transition, whereas the structure has changed to fully crystalline just after the transition. This observation leads to the conclusion that the transition to the second stage is caused by crystallisation. The faster diffusion in this stage is believed to be facilitated by the formation of grain boundaries that provide an easy pathway for diffusion.



Figure 2.10: The interface (in between the Mo and the deposited Si) is amorphous just before the transition to the second stage (left), but has abruptly become crystalline just after the transition (right).


Highlight: Controlling interlayer chemistry in La/B4C multilayers
Thin film multilayers, with layer thicknesses of only several nanometers, can be used as reflective coatings in EUV lithography applications. Specifically, high reflectance La/B4C multilayers would enable EUV optics to operate at a reduced wavelength (6.7 nm), further reducing the printed feature size during the lithography process. Due to the chemical reactivity of La and B4C, formation of LaB6 and LaC2 occurs at the interfaces during the coating of a La/B4C multilayer. These interfacial layers strongly reduce the optical contrast, thereby reducing multilayer peak reflectivity.

By applying a N2+ plasma treatment to the La surface before depositing B4C, we expect surface passivation, resulting in a reduced La-B4C chemistry. From photo-electron spectroscopy, formation of La-N as well as B-N is indeed observed, leading to an enhanced optical contrast. 



Figure 2.11: A La/B4C multilayer structure exhibits sharp contrast between high (La) and low (B4C) density materials due to surface passivation using a N2+ plasma.

Transmission electron microscopy (Figure 2.11) indeed shows a high quality multilayer structure with clearly separated high (La) and low density (B4C) layers. 

This work provides the basis for further studies on thin film surface chemistry and ways to passivate thin film surfaces during multilayer development. A more detailed description of the work on La/B4C can be found in section 3.