(M.J. Rost, J.W.M. Frenken, R.M. Tromp)
One of the prime strengths of Low-Energy Electron Microscopy is its ability to follow dynamic phenomena on surfaces with a superb time resolution (e.g. video rate) on a wide range of mesoscopic length scales, from several nm to 100 mm. In this sense, the LEEM/PEEM setup will be complementary to the high-speed Scanning Tunneling Microscopy instrumentation that has been developed in the Interface Physics Group, headed by Joost Frenken, Tjerk Oosterkamp and Marcel Rost. Although the high-speed STM can image these phenomena with atomic resolution at speeds up to video rate, the maximum field of view at these high STM frame rates is limited to a few tens of nm. This necessarily narrows the STM view to local, atomic-scale aspects. In much of the interesting physics of surface dynamics, including surface phase transitions, collective diffusion phenomena and nucleation and growth phenomena, larger length scales play a crucial role and a full understanding of these processes can be obtained only when one combines results from STM and LEEM observations. In the following we briefly describe two challenging subjects that call for such a combined Ångstrom-tomicrometer approach.
Thin polycrystalline metal films are becoming increasingly important, as is reflected in the multitude of applications in different fields,1 e.g. nanotechnology, nano-optics, micro-electronics, and catalysis. The intrinsic link between film properties and the precise film structure enables the production of films with tuned properties. Ideally one would like to control the texture (grain boundary orientation), grain size distribution, adhesion, intrinsic stress, and the overall morphology down to the nanometer scale. The great interest in the dependence between the developing film structure and the precise deposition parameters resulted in the establishment of the well-known structure zone models,2-7 that summarize the current background knowledge (see Figure 5). However, these phenomenological models certainly lack information on the atomistic details and processes that take place during thin film growth. The continuous miniaturization of thin film applications clearly demands fundamental research that links well-known atomic processes, such as diffusion and nucleation, with the mesoscopic evolution of film properties, both during film growth as well as during a post deposition treatment (such as heating, applying stresses, and additional coating). A typical example is given by the evolution of the intrinsic stress of thin films during and after the deposition.8-15 On growth interruption one observes sudden, significant changes of the film stress, which must be linked to morphological changes. However the underlying atomic processes are still under debate. In order to unravel such processes we need an in-situ, real-time technique that is nondestructive, surface sensitive and capable of obtaining atomic (step) resolution. Figure 5: Empirical structure-zone-model 2 : different film structures and surface roughness evolve during the deposition at different substrate temperatures T. (Tm: melting point of the depositing material)
Excited by this topic, Rost focused on polycrystalline film phenomena using scanning tunneling microscopy (STM). For the first time, STM technology was to observe grain growth and the migration of grain boundaries, during post-deposition anneal-treatment of a polycrystalline film.16 Also, the growth of a thin polycrystalline gold film could be studied with STM in real time, identifying important atomic processes that govern the growth.17 The major drawback of STM is its limited field of view. Due to this restriction it is hardly possible to combine the obtained atomic information with morphological changes in the grain boundary network on the mesoscopic scale. LEEM easily bridges this gap: within seconds the user can zoom out to 100 m, shift several mm over the sample, and quickly retrace to the very same starting location. In addition, atomic information is contained in the Low Energy Electron Diffraction pattern and intensities, which presently can be obtained on areas as small as 10 nm. Therefore LEEM forms the ideal complementary technique to STM. The proposed microscope with its improved spatial resolution (down to 2 nm) will make LEEM experiments on rather rough polycrystalline samples possible. To our knowledge nothing has been published on the application of LEEM to polycrystalline samples, probably due to the resolution difficulties one faces: the step distance is often less than the 10 nm spatial resolution of existing microscopes. Although we are not expecting to resolve all individual steps 1, we are confident that we can locate the position of the grain boundaries and thereby also the individual grains. In order to ensure the feasibility we performed a proof-of-principle test on the existing IBM microscope by using a polycrystalline bulk Ni foil, heated in situ to above 800 Celsius. The movement of grain boundaries, the growth of individual grains, and the overall change is crystallographic alignment and crystal quality could be easily observed at high temperatures. The superior spatial resolution as well as the time resolution of the new microscope will uniquely enable the application of LEEM to study aspects of polycrystalline phenomena in-situ and in real-time. Focusing on the understanding of the underlying atomic processes we are planning to unravel key questions concerning the morphological evolution of polycrystalline films. Special emphasis will be put on · grain boundary structure and migration · morphological surface changes on top of the individual grains · nucleation and growth of a polycrystalline film on an amorphous substrate · growth of a thin polycrystalline film during its deposition The present proposal perfectly fits in the large-scale research program on polycrystalline metal films of Dr. Rost. Recently; he received significant funding from STW for a project that is aimed at understanding the growth of thin vacuum-deposited, polycrystalline metal films and multilayers. In collaboration with FOM Rijnhuizen and ZEISS GmbH this projects aims on the development of the multilayer EUV optics that will enable the next generation lithography for the microelectronics industry. 1 On rough surfaces as well as in the vicinity of the grain boundaries the step distance can still be less than the improved lateral resolution of 2 nm.
1. C.K. Hu et al.; Mater.Chem.Phys. 52 5 (1998). 2. B.A. Movchan et al.; Fizika Metall. 28 83 (1969). 3. J.A. Thornton; Annu. Rev.Mater.Sci. 7 239(1977). 4. R. Messier et al.; J.Vac.Sci.Technol.A 2 500 (1984). 5. P.B. Barna et al.; Thin Solid Films 317 27 (1998). 6. C.V. Thompson; Annu.Rev.Mater.Sci. 30 159 (2000). 7. I. Petrov et al., J.Vac.Sci.Technol.A 21,5 S117 (2003). 8. A.L. Shull et al.; J.Appl.Phys. 80 6243(1996). 9. E. Chason et al.; Phys.Rev.Lett. 88 156103 (2002). 10. F. Spaepen; Acta Mater. 48 31 (2000). 11. C. Friesen et al.; Phys.Rev.Lett. 93 056104 (2004). 12. R. Koch et al.; Phys.Rev.Lett. 94 146101 (2005). 13. C. Friesen et al.; Phys.Rev.Lett. 95 229601 (2005). 14. C.-W. Pao et al.; Phys.Rev.Lett. 99 036102 (2007). 15. J.S. Tello et al.; Phys.Rev.Lett. 98 216104 (2007). 16. M.J. Rost et al.; Phys.Rev.Lett. 91 026101 (2003). 17. M.J. Rost; Phys.Rev.Lett. submitted