preprint from: S. Morita, R. Wiesendanger, E. Meyer (eds.) Noncontact Atomic Force Microscopy Springer Verlag, Berlin to be published June 2002 6 Atomic resolution imaging on fluorides Michael Reichling and Clemens Barth Department Chemie, Universit��t M��nchen, Butenandtstra��e 5-13 81377 M��nchen, Germany reichling@cup.uni-muenchen.de Fluorides are an important class of inorganic materials that have been intensively studied in their bulk properties for several decades [1]. Among crystals with the fluorite structure, CaF2 is the most prominent prototype crystal that recently gained tremendous interest as a vacuum ultraviolet optical material. Components made of CaF2 are necessary for the development of next generations of laser lithography that is a key technology for the semiconductor industry [2]. Laser lithography optics requires materials with structural perfection and utmost purity of bulk and surface. In this context, highest resolution and sensitivity imaging on fluorides for structural characterization and surface defect detection became an important issue and is a major driving force behind the development of dynamic force microscopy for highest resolution imaging on fluorides. Beside this specific application, fluorides are important test materials for the development of atomic resolution force microscopy on insulators. Surfaces of fluorite type crystals prepared by cleavage along the (111) plane are atomically flat over large areas and very stable under ultra-high vacuum conditions. Their structure is simple but bears more details than that of cubic halides (see Chapter 5) and these details provide an excellent test for the resolution power of the scanning force probe but also for theoretical predictions on scanning results. As will be discussed in Chapter 17, scanning force imaging on the CaF2(111) surface has been modelled by theoretical simulations and this is presently the best understood insulator surface in terms of a quantitative interpretation of atomic scale contrast formation. Both, the ease of preparation and the detailed understanding of atomic contrast predestines the CaF2(111) surface as a standard for the calibration of atomic resolution scanning force measurements and the calibration of tips. In this chapter we review the state of the art in atomic resolution imaging of fluoride surfaces and describe main features and peculiarities while all aspects of quantitative imaging are discussed in Chapter 17. In the following sections we will briefly introduce details of our experimental techniques, then address the impor- tant issue of tip structure and tip instability and finally discuss atomic contrast on flat surfaces and at step edges and kinks. An important feature of experiments introduced here is that measurements are often not operated in the standard topography mode where the cantilever resonance frequency detuning is kept con- stant but in the constant height mode where the detuning is the primary output signal. Results obtained in the two modes are contrasted and it is discussed what kind of information and details can best be deduced from either method.

6.1 Experimental techniques All experiments reported here were performed with a scanning force microscope based on the design by Howald et al. [3] operating in an ultra-high vacuum (UHV) system with a base pressure in the low 10-8 Pa range. The principles of operation of this instrument are those outlined in Section 2.3 and a schematic representation of the electronics for cantilever self-excitation and measurement is shown in Fig. 6.1. The cantilever is excited to vibration at its resonance frequency of typically 75 kHz with an amplitude A stabilized to a pre-set value between 20 nm and 100 nm (peak to peak) by a positive feedback loop where the error signal of the amplitude loop yields a measure for the per-cycle dissipated energy during cantilever oscillation and is therefore called damping G. The cantilever resonance frequency detuning Df is detected by a PLL demodulator and used as an input for the distance control loop where the error signal of this ���f loop yields the topography z. The extent to what the tip follows the contour of the pre-set ���f value is determined by the gain of the topography (���f) loop. In standard topogra- phy mode imaging, the loop gain is high, ���f virtually vanishes and z represents the full topographic contrast. In the constant height mode, the ���f loop is switched off so that the tip scans effectively at constant height above the surface and ideally the topography signal z vanishes. In practice, the loop is often not switched off but operated at very low gain so that a minor inclination between the surface and the scanning plane is compensated to maintain a constant tip-surface distance while the cantilever response to small surface features and especially atomic corrugation is not influenced by the distance control. In this case, the detuning signal ���f is the primary output signal carrying information on the tip-surface interaction. During scanning, all four output signals are recorded for forward and backward scanlines and thus each image point can be characterized by a set of eight measured values. If relevant, the scanning direction is indicated in images and graphs by arrows for demod- ulator detuning ���f damping �� topography z z x - y Udc ���fpre-set amplitude loop Apre-set amplitude A loop gain ���f loop Fig. 6.1. Experimental setup for dynamic mode SFM with self-driven cantilever oscillation. In topography mode imaging, the topography signal z is recorded at constant detuning ���f while the primary output signal is the detuning ���f when operating in the constant height mode at fixed tip- surface distance