Polystyrene spheres on mica substrates:AFM calibration, tip parameters and scan artefacts M. VAN CLEEF, S. A. HOLT, G. S. WATSON & S. MYHRA Faculty of Science and Technology, Griffith University, Nathan, Queensland 4111, Australia Key words. Force microscopy, polystyrene spheres, tip parameters, scan artefacts, noncontact mode. Summary Atomic force microscopy (AFM), in various versions, has had major impact as a surface structural and spectroscopic tool since its invention in 1986. At its present state of development, however, the interpretation of AFM images is limited by the current state of methodologies for calibration over the wide dynamic range of magnification. Also, the parameters of individual tips, as well as the generic characteristics of different kinds of tips, affect both the quality of the images and their interpretation. Finally, the very nature of the tip-to-surface interaction will generate artefacts, in addition to those associated with tip shape, which need to be fully understood by the practitioners of force microscopy. This project seeks to address and shed light on some of these issues. Polystyrene beads deposited on mica substrates form hexagonal close-packed layers. The unit cell parameters are suitable for calibration of the AFM in the lateral plane, while the perpendicular spacing of the layers is appropriate for calibration along the vertical axis. Using different size fractions, it is straightforward to determine the extents of linearity, orthogonality, thermal and instrumental drifts over distances from 100 nm to tens of micrometres. The present results show that the methodologies for contact mode operation can be adapted to noncontact modes. It is known that an AFM image arises from a convolution of surface topography and tip shape, and is mediated by the interaction. In principle it is possible to carry out a deconvolution, if we have complete knowledge about two of the three elements (i.e. tip, surface and interaction). In practice we rarely have the requisite information. Prominent artefacts will occur when the characteristic parameters of the tip are comparable to those of the surface topography, and/or if there is a variable strength, or extent of localization, of the interaction. The present results demonstrate artefacts due to effects of geometry as well as interaction. Introduction A generic scanned probe system consists of three essential elements: a sharp tip which must be locatable and controllable in space; a surface with particular topographic and structural properties; and a localized interaction between tip and surface having a strength which depends on the tip-to-surface separation, and which is responsive to topographic and/or structural changes. The most significant consequence of this description is that given complete information about any two of the three elements, there is the implication that, in principle, complete knowledge about the third must be obtainable. The first member of the recent family of scanned probe microscopes (SPM) was the scanning tunnelling microscope (STM), invented by Binnig et al. (1982). This development was anticipated by the work of Young et al. (1972), who achieved three-dimensional imaging with the ‘Topographi- ner’. The next member, the atomic force microscope (AFM), was demonstrated 4 years later by Binnig et al. (1986). Subsequently, other interactions (e.g. magnetostatic, electro- static, thermal radiation, etc.) have generated additional members, ably described by Wickramasinghe (1990). As well, each of the original techniques has had offsprings by way of variations on the theme – such as contact, noncontact, lateral force and force constant spectroscopic modes for the AFM. In many respects AFM is a technique at the steepest part of the learning curve. For instance, identification of image artefacts cannot be carried out on a routine basis, and establishing optimum imaging conditions remains very much an artform. Although these aspects have been the subjects of several recent studies (e.g. Wurster & Ocker, 1993; Odin et al., 1994; Schwartz et al., 1994; Xu & Arnsdorf, 1994), much work remains to be done in order to establish ‘best practice’ for AFM over the broad range of current and potential applications. The present study seeks to exploit the ready availability of polystyrene spheres in the nanometre size range. Journal of Microscopy, Vol. 181, Pt 1, January 1996, pp. 2–9. Received 22 November 1994; accepted 1 August 1995 2 1996 The Royal Microscopical Society