Meas. Sci. Techno!. 5 (1994) 976-984. Printed in the UK Design of a scanning probe microscope H Olin Department of Physics, Chaimers University of Technology, S-412 96 Goteborg, Sweden Received 17 February 1994, in final form 14 April 1994, accepted for publication 18 AprH 1994 Abstract. A compact scanning probe microscope for operation in air and liquid is described. The probe techniques implemented are scanning tunnelling microscopy and scanning ion conductance microscopy. The software, electronics, mechanical construction and some representative measurements will be presented here. The compact and concentric microscope head is built around a commercial piezoelectric inchworm motor. The scanner is a standard piezo tube. An analogue feedback system is used lor taking images, while digitai controi of the probe-sample distance is used for other experiments, such as measurements of current-voitage characteristics. A Macintosh personal computer is used for controi and presentation of data. A simple method to make scanning tunnelling microscope tips suitable for electrochemical use is described. The microscope has a high resonance frequency (9.6 kHz), low noise (0.01 nm Hz-'i 2 at 10 Hz), low thermal drift (less than 0.1 nm mln-'), and high acoustical noise suppression. The current-distance-dependency of the scanning ion conductance microscope was found to be linear. 1. Introduction Scanning tunnelling microscopes (STM) [1-4J and other scanning probe microscopes (SPM) [4,5J give topographic images by scanning a tip over a sample, with resolution down to atomic dimensions. In the STM a conductive tip is brought close to (about 0.5 nm) the sample. When a small voltage is applied between the tip and the samplc, a tunnelling current starts to flow. This current is strongly dependent on the distance and, by measuring the current, the tip-sample distance can be determined. The tip is then raster-scanned over the sample, while a feedback system keeps the tip-sample distance constant. A picture is generated by plotting this feedback signal. The introduction of the STM in 1982 by Binnig, Rohrer and co-workers [6 J has stimulated the develop- ment of other scanning probe microscopes, in which the tunnelling tip has been replaced by other probes, while keeping the rest of the concept. For example. in the scan- ning force microscope. electrostatic, magnetic or van der Waals forces are measured locally. Other SPMs mea- sure local temperature, Faradaic currents or optical near fields. A great deal of work is being done to improve understanding of these probes and to implement new ones. Another SPM, invented by Hansma and co-workers in 1989 [7J, is the scanning ion conductance microscope (SICM). The probe in the SICM is an electrolyte-filled micropipette. The sample is immersed in an electrolyte and a voltage drives an ion current through the aperture 0957-0233/94/080976+09$19.50 \'1;1 1994 lOP Publishing ltd of the pipette. The current decreases near the surface and the distance is monitorcd by measuring the current. An SPM consists of four main parts: the probe, the mechanical parts, the electronics and the computer. The mechanical part consists of essentiaJly a piezoelectric scanner, with a typical scanning range of a few micrometres, and a coarse positioning device that reduces the tip-sample to make it rcmain within the controlling range of the piezo. While the single piezo tube scanner invented by Binnig and Smith [8] has become a standard choice, a number of coarse positioning mechanisms have been implemented, for example, the early piezo electric walker ('the louse') [61. home-built [9] or commercial inchworms [10, II]. inertial gliders [12], lever reduced screw mechanisms [13] Or differential screws [14[. For more examples and references, see [2, 15J. Because of the mechanical nature of these micro- scopes, one of the major construction problems is vi- bration isolation. The main noise sources are low- frequency building vibrations and sound. The general method [15,16] (0 solve this problem is to construct the SPM with a high resonance frequency and to place it on a low-resonance-frequency damping table. An SPM with a high resonance frequency generally means a stiff and small SPM; the smaller the better. While the piezo scanner is small, the coarse approach mechanism usu- ally adds to the size, and it is often the design of this mechanism that determines the overall vibration sensi- tivity. Thermal drift is another problem that must be considered in designing an SPM. This prohlem can be