Realization of a Liquid Atomic Force Microscope

Realization of a Liquid Atomic Force Microscope Ivo de Rijk DCT 2008.004 Traineeship report Supervisor: prof. dr. H. Kawakatsu prof. dr. ir. M. Steinbuch Technische Universiteit Eindhoven Department Mechanical Engineering Dynamics and Control Technology Group Eindhoven, February, 2008

2 Acknowledgements This report is the result of a research project of four months at Tokyo University, Japan. The project could not have been successful without the help of my team members Kentaro Minemura and Yuki Nishimori. Furthermore, I would like to thank the laboratory members in general for the extremely educational and pleasant time. More specific I would like to thank Yasuo Hoshi for his effort on the software, Dai Kobayashi for his tremendous help with the control part, Shuhei Nishida for sharing his knowledge of the liquid AFM, Kazuhisa Nakagawa for the, always available, helping hand and of course professor Kawakatsu for giving me the opportunity to do this project and his inspiring help during the project.

Contents 1 Introduction 4 2 Liquid AFM 5 3 Mechanical head 8 3.1 Cantilever positioning....................... 9 3.1.1 Cantilever fixation..................... 9 3.2 Sample positioning......................... 11 3.2.1 Tube piezo......................... 12 4 Optics 15 4.1 Heterodyne Doppler interferometer............... 15 4.2 Photothermal excitation...................... 18 5 Control 19 5.1 Approach.............................. 20 6 Results 24 7 Conclusions & Recommendations 27 7.1 Conclusions............................ 27 7.2 Recommendations......................... 28 A Mechanical drawings LAFM mechanical head 30 A.1 Mechanical drawings lower part mechanical head........ 30 A.2 Mechanical drawings upper part mechanical head........ 44 B Piezo actuated stages 55 C Test results sample step size 57 3

Chapter 1 Introduction Since the introduction of the Atomic Force Microscope (AFM) in 1986 by Binnig et al.[1] the AFM has shown to be a versatile tool to gain knowledge about surfaces, down to atomic scale. One of the most promising applications of the AFM is imaging of surfaces in a liquid. For example, cells at atomic resolution in water or a real time observation of the behavior of water molecules on a surface with changing temperature. A liquid AFM (LAFM) was designed and build at Kawakatsu laboratory and completed in 2005 [7]. This LAFM has produced atomic scale resolution images in a liquid e.g. mica in water, which is shown in figure 2.1. The objective of this research project is to realize a new version of the LAFM, improving the mechanics where possible and realizing a home-built optical setup to measure the cantilever speed and excite the cantilever by photothermal means. This project is performed at Tokyo University, Institute of Industrial Science, Kawakatsu laboratory in Japan as a part of the Mechanical Engineering Master s program at Eindhoven University of technology in The Netherlands. The report starts with a basic introduction to the LAFM, describing its working principles. The mechanical part of the LAFM, the head, is discussed in chapter 3. The mechanical head performs the actuation and positioning of the sample and cantilever. The optics are explained in chapter 4, which incorporates the measuring and excitation of the cantilever together with camera view on the cantilever. The control circuitry is roughly explained in chapter 5, followed by the overall result in chapter 6 and the conclusions & recommendations in chapter 7. 4

Chapter 2 Liquid AFM The Frequency Modulated Liquid Atomic Force Microscope (FM-LAFM) enables imaging of a surface at high resolution in a liquid, for example biological molecules in their native environment. In a liquid, the highest resolution has been obtained in constant-force mode. In constant-force mode the tip-sample interaction force is kept constant by adjusting the distance between the tip and sample. However, scanning a sample with a constant vertical force implies lateral friction forces which can damage or detach molecules from the sample. Furthermore, in dynamic mode the (high frequency) signal can easily be distinguished from (low frequency) drift. (Hoogenboom 2006[2]) Figure 2.1: Constant frequency image of mica in water, size= 5x5 nm, f 0 = 982.8 KHz, f= -8 KHz, scan speed= 48.8 nm/s, amplitude= 0.18 nm p p The LAFM in this report uses Frequency Modulation (FM). In the FM detection system a cantilever oscillates at resonance frequency in the vicinity of 5

6 CHAPTER 2. LIQUID AFM a sample surface, typically 900 KHz in the 2 nd bending mode. The gradient of the conservative tip-sample interaction force leads to a shift, f, of the eigenfrequency, f 0, of the cantilever. Dissipative forces damp the amplitude of the oscillation, A, but do not change f. The conservative tip-sample interaction varies over the oscillation cycle of the cantilever. The interaction force can be quantitatively determined, provided that f is known as a function of the tip-sample distance and that A is kept constant by adjusting the cantilever excitation. (Albrecht 1991[3], Giessibl 1995[4]) A typical result of a surface scan by a FM-LAFM is shown in figure 2.1. The scan is performed on mica in water at the 2 nd bending mode of the cantilever. The cantilever is photothermally excited by an excitation laser and the speed is measured by a heterodyne Doppler interferometer. The excitation of cantilevers in a liquid obviously has its limitations compared to operation in air. Resonant vibration of the cantilever is impeded by the large damping and added inertial mass of the liquid. These effects result in a low quality factor (Q factor) and a smaller shift of the resonance frequency, resulting in degradation of the sensitivity. (Ramos 2006[5]) However, if piezoelectric actuation is chosen to excite the cantilever, this will result in coupling of the support vibration to acoustic modes of the liquid. (Ratcliff 1998[6]) Furthermore, photothermal excitation enables the use of the torsion mode, in which lateral forces between the cantilever and the sample can be measured. A schematic drawing of the realized LAFM is shown in figure 2.2. In this figure the LAFM is divided into three parts: 1. Mechanical head, to fixate and position: 2. Optics 3. Control The objective lens Cantilever Sample Heterodyne Doppler interferometer to measure cantilever velocity Laser Diode for photothermal excitation of the cantilever Camera view for vision on the cantilever and the laser spots Approach control Phase Locked Forced Oscillation Method for imaging The three parts will be discussed in detail in Chapter 3, 4 and 5 respectively.

Figure 2.2: Schematic drawing of the LAFM, adjusted from [7] 7

Chapter 3 Mechanical head The mechanical head of the LAFM enables the positioning and fixation of the objective lens, cantilever and sample. Figure 3.1 shows the realized mechanical head. Figure 3.1: Mechanical head The design is made with the use of CAD software (Unigraphics NX3). The mechanical design is a changed version of the first version of the LAFM [7]. Technical drawings are included in Appendix A. The AFM is of the inverted type, which means that the probe is oriented upwards and the sample approaches from the upside. The mechanical head is mainly constructed from magnetic stainless steel (SUS304). For parts that do not need to be magnetic, duralumin or brass is used to save manufacturing-time. Figure 3.3 shows the lower part of the mechanical head in which the reference, measurement and excitation laser, together with the green LED light 8

3.1. CANTILEVER POSITIONING 9 for the CCD camera, enter the mechanical head in x-direction from the optics. The reference part of the Doppler interferometer passes the 90 -mirror and hits a normal mirror in the lower part of the mechanical head. The other beams need to be positioned by manipulation of mirrors in the optics-part to hit the 90 -mirror and go over the central axis of the objective lens, in z- direction. The objective lens (Olympus LUCPLFLN 20X) is designed for observation through a substrate and can be manually adjusted to get best focus depending on the thickness of the substrate. Furthermore, this lens has a relatively large working distance (6.6-7.8 mm), where a minimum distance of 3 mm is needed to have the possibility to focus on the sample in the approach procedure (explained in more detail in section 5.1). This lens can easily be replaced by the 40X version instead, if needed in the future, which has a working distance of 2.7-4 mm and can increase the measurement and excitation laser performance. The objective lens can be moved in z-direction to focus by manually moving the commercial x-stage in the lower part of the mechanical head. The x-stage moves the mounted wedge in x-direction, which moves the objectiveholder in z-direction. The movement of the objective lens is guided by a commercial guide fixed on the lower base of the mechanical head. 3.1 Cantilever positioning On top of the lower base two stainless steel (SUS304) stages are stacked to position the cantilever relatively to the objective central axis. One plate to guide the cantilever x-stage is fixed to the lower base and two stages for the x- and y- direction movement are stacked on top (figure 3.2). The two stages are moved by shear piezo s (Fuji-ceramics C-22). The walking principle of the shear piezo s is explained in Appendix B. The stages and lower base have a hole around the laser path for the objective lens to approach the probe. A glass substrate is placed over the hole in the cantilever y-stage, on which the cantilever is fixed. The fixation is further explained in section 3.1.1. 3.1.1 Cantilever fixation The silicon chip of the cantilever is fixed by anodic bonding. By applying a voltage across the glass substrate and the silicon chip, alkali ions in the glass move towards the negative electrode, leaving a thin negatively charged region at the interface to silicon. In other words, all voltage applied is applied to the

10 CHAPTER 3. MECHANICAL HEAD Figure 3.2: Schematic drawing upper part mechanical head, adjusted from [7] Figure 3.3: Schematic drawing lower part mechanical head, adjusted from [7]

3.2. SAMPLE POSITIONING 11 glass-silicon interface. While the voltage is applied, the glass is heated up to several hundred degrees Celsius. Due to the dielectric attraction force, added to the interface, the alkali atoms of glass (which are at the interface) make covalent bonds with silicon atoms in the silicon chip of the cantilever. The anodic bonding setup is shown in figure 3.4. Pyrex #7740 is used as glass substrate, because it contains the Alkali ions and its thermal expansion coefficient is close to that of silicon, which leaves less stress in the silicon chip and substrate. The glass substrate is fixed by placing magnets on top of the substrate, which attract the cantilever y-stage under the substrate. Furthermore, a rubber ring is positioned on the substrate, around the silicon chip, to surround the liquid applied. Figure 3.4: Anodic bonding setup 3.2 Sample positioning The rough x- and y-movement of the sample is also realized with shear piezo s. This can be seen in figure 3.2. The sample x-stage is guided by two stainless steel bars in the top base and the y-stage is stacked on top of the x-stage. The tube piezo holder is on top of the wedge stage and is pushed up, in z-direction,when the wedge stage moves horizontally, in y-direction. The movement of the sample relatively to the cantilever can be monitored by a CCD camera (Sony-DXC107A), which has view on the cantilever through a hole in de sample stages. The camera is equipped with a variable zoom lens (Suruga seiki - V20-501, 0.75 4.5 variable zoom), the end of the lens can be seen in the left side of figure 3.5(a).

12 CHAPTER 3. MECHANICAL HEAD (a) (b) (c) Figure 3.5: Mechanical head; a) front view; b) left side view; c) right side view 3.2.1 Tube piezo The tube piezo holder is shown in figure 3.6. There are three shear piezo s between the wedge stage and tube piezo holder and four between the Z-base and tube piezo holder. The magnets are placed to: - Generate enough normal force between the piezo s and the surface, to increase the friction. - Keep the tube piezo holder attached to the Z-base when the wedge stage moves.

3.2. SAMPLE POSITIONING 13 The seven piezo s attached to the tube piezo holder and the four piezo s under the wedge stage are fed by the same signal. This means that slip occurs between the wedge stage and tube piezo holder and between the Z-base and tube piezo holder, when the tube piezo holder is pushed up. The construction to make the rough z-movement of the sample is over constrained, with the two guides and seven piezo s with line and surface contacts. The reason that this construction is used is the choice to use an existing well working construction: a custom designed piezoelectric stage at Kawakatsu laboratory, which uses the same construction on a smaller scale. But, there is room for improvement, considering that this rough sample positioning in z-direction is used in the approach procedure. This procedure will be further explained in section 5.1. Figure 3.6: Tube piezo holder To test the rough z-movement of the sample, the step size of the tube piezo holder is measured. A triangular wave input signal is applied to the wedge stage and the movement of the tube piezo holder is measured by a capacitance sensor (MESS-TEK M-2213). The test results are shown in table 3.1. More details can be found in Appendix C. Table 3.1: Rough test results sample step size Amplitude [V ] Step size [nm] 100 2 150 4 200 6 The tube piezo connected to the holder is glued to a steel plug on both sides. This can be seen in figure 3.7. The top plug is bolted to the tube piezo holder and the bottom plug is connected to the sample holder by a magnet. The magnet is located in a hole between the lower plug and the sample holder. A sample is glued to the sample holder.

14 CHAPTER 3. MECHANICAL HEAD The working principle of the tube piezo is shown in figure 3.7. By applying a voltage on the x- and y piezo s the scanning movements are performed. The z-piezo is used for probe to sample height adjustment. Figure 3.7: Tube piezo working principle; GND= ground

Chapter 4 Optics The optics of the LAFM consists of: A heterodyne Doppler interferometer to measure the cantilever deflection velocity. λ= 632.8 nm; red laser Photo thermal excitation of the cantilever. λ= 405 nm; blue laser CCD camera with green LED light The three parts will be discussed in more detail in the next sections. A overview of the optical system is shown in figure 4.1. The setup is fixed on the same base plate as the mechanical head. To position and focus the measurement and excitation beam on the cantilever, a CCD camera is added. The camera is in upright position at the end of the dichroic mirror-line. A normal mirror under the camera enables the positioning of the camera view through the dichroic mirrors and over the central axis of the objective lens. In the (temporary) setup of figure 4.1 an extra dichroic mirror is added to introduce the green LED light. However, in the final setup the LED will be combined in the upright camera construction. Figure 4.2 shows the camera view of the cantilever, with the measurement- (red) and excitation (blue) laser. 4.1 Heterodyne Doppler interferometer A heterodyne Doppler interferometer is implemented. (Kawakatsu 2002[8]) The optical setup is depicted in figure 4.3 To analyze the difference in performance, two types of He-Ne laser were used as source: 15

16 CHAPTER 4. OPTICS Figure 4.1: Overview of the optics, with light paths; red laser= measurement; blue laser= excitation; PBS= Polarizing Beam Splitter; AOM= Acousto Optic Modulator; APD= Avalanche Photo Diode; LED= Light Emitting Diode; CCD= Charge-Coupled Device - Power stabilized (Neoark - 2MSS), λ= 633nm - Frequency stabilized (Neoark - Model 430), λ= 633nm The laser beam is first split into a reference and measurement beam at the first polarizing beam splitter (PBS). The beat frequency of the laser source reference beam is shifted with 80 MHz (f 1 ), on the reference path, by an acousto optic modulator (AOM, Panasonic EFLM) and continues towards the mechanical head to hit the reference mirror. The measurement beam directly goes to the objective lens (under the mechanical head) and hits the cantilever, getting a frequency shift, f 2, due to the Doppler effect. At the point where the reference and measurement beam come back from the mechanical head and are recombined (f 0 +f 1 ), interference takes place by which the signal contains f 1 + f 2 and f 1 f 2 peaks. The signal is detected by the two Avalanche Photo Diodes (APD s) after which one peak is selected by a band path filter and frequency demodulated by a delay line descriminator. A schematic example of the input and output of the APD s is shown in figure 4.4. Figure 4.1 also

4.1. HETERODYNE DOPPLER INTERFEROMETER 17 Figure 4.2: Camera view of the cantilever s backside Figure 4.3: Schematic of the heterodyne laser Doppler interferometer shows a lens in front of the APD s to focus the beam on the diodes after it is scattered mainly due to the cantilever. To get the best result from the Doppler interferometer it is preferable to locate the reference mirror next to the cantilever. However, this could not be realized in this setup, which means that the difference between the measurement en reference beam causes an error. Amongst other things this difference is caused by the difference in speed of the laser beam due to temperature/ pressure- gradients. To compensate that difference, the reference mirror is actuated by a PI-controlled piezo.(nishida 2008[9]) By feeding back the phase difference between the AOM and APD, the length of the light path is actively adjusted to compensate the difference in light speed. Tests on the Doppler interferometer were performed to determine the noise density of the setup using a He-Ne laser source and a laser diode source. These results will be shown in chapter 6.

18 CHAPTER 4. OPTICS Figure 4.4: top) carrier, AOM and signal, Cantilever ; bottom) resulting output 4.2 Photothermal excitation The most common method to generate cantilever oscillations uses a piezoelectric transducer. However, in a liquid the use of a piezoelectric transducer leads to coupling of the support vibration to acoustic modes of the liquid. This produces spectra that contain many features unrelated to the cantilever vibrational modes.(ratcliff 1998 [6]). Photothermal excitation does not have vibration coupling. To excite the cantilever with photothermal excitation, a laser is aimed at the cantilever s backside. A laser with λ= 405 nm is selected to focus on the silicon cantilever. A significant part of this light is absorbed by the cantilever and converted to thermal energy. The intensity-modulated light focused onto a region of the cantilever produces a time-dependent temperature distribution due to the absorption of optical energy.(ramos 2006 [5]). Furthermore, the spot location can be chosen to excite both bending and torsional modes of the cantilever. In figure 4.1 it can be seen that the optical fiber enters the optical setup and goes through a collimator to make the beam from the fiber parallel. After the collimator a variable lens is located to focus the laser beam on the cantilever.

Chapter 5 Control The project focussed on the design and realization of the mechanical head and optics. The control-part is implemented, but designed and realized by other members of Kawakatsu laboratory, for that reason the control part will only be explained shortly. The control of a conventional FM-AFM can be divided into two parts: 1. The cantilever oscillation is generated by self-oscillation. The detected vibration signal is fed back to the actuation through filtering, amplitude adjustment and phase adjustment. In self-oscillation the cantilever s frequency automatically corresponds with the cantilever s resonance frequency. 2. The oscillation frequency is monitored by a FM demodulator and z-axis distance is controlled in order to keep the frequency constant. These parts are shown schematically in figure 5.1. In a liquid the Q-factor degrades to 10 or less, which results in large fluctuations of the oscillation frequency around the center frequency, a poor carrier to noise ratio and it is difficult to keep small amplitude oscillations stable. This is why a different method, the Phase Locked Forced Oscillation Method (PLFO) is used. In this method the cantilever is driven by an oscillating signal from a stable signal generator and the phase change between the driving signal and the resulting vibration is used as feedback signal, which is shown in figure 5.2.(Kobayashi 2007 [10]) The lower Q-factor has less effect on the phase change with changing oscillation frequency compared with the amplitude change, thus using the phase change as feedback signal gives better results at low Q factors. 19

20 CHAPTER 5. CONTROL Figure 5.1: Conventional FM-AFM control scheme; self-oscillation and z- distance control loop[10] 5.1 Approach While discussing the control of the LAFM it is assumed that the LAFM is operating in imaging-mode. However, before imaging can take place an approach procedure is needed. In the first step of the approach procedure the wedge stage is actuated while watching the approach with use of the camera at the side of the mechanical head (section 3.2). The second camera, from the optics, is focused somewhat closer to the sample, above and next to the cantilever, by which the further approach can be monitored by watching the sample getting into focus. The final approach step uses the piezo walkers of the wedge stage together with the tube piezo: 1. The tube piezo is maximal retracted 2. One approaching step is performed with the piezo walker of the wedge stage 3. The tube piezo is extended and the oscillation frequency is monitored 4. If a change in frequency is detected the approach is finished, otherwise step 1-3 is repeated

5.1. APPROACH 21 Figure 5.2: Control scheme of the forced oscillation method at imaging[10]; LD= Laser Diode When applying the total approach procedure, the control scheme shown in figure 5.2 can not be used since the Z distance control loop doesn t work before the tip-sample distance comes into a range which the Z-tube can cover. Therefore the control scheme shown in figure 5.3 is used in the approach procedure. In the approach procedure, the frequency of the driving signal is controlled, similar to the self-oscillation method. The total control scheme is shown in figure 5.4. A soft switch is implemented to switch between the approach and imaging control scheme.

22 CHAPTER 5. CONTROL Figure 5.3: Control scheme of the forced oscillation method at approach[10]; LD= Laser Diode

5.1. APPROACH 23 Figure 5.4: Schematic overview control scheme[10]; PLFO= Phase Locked Forced Oscillation; PLFO-1= Imaging path; PLFO-2 Approach path

Chapter 6 Results At the end of the project the LAFM is completely realized. Some practical imperfections are left, therefore imaging is not yet possible. Furthermore, the mechanical head and optics need testing and optimization. The software, which is developed in parallel at Kawakatsu laboratory, by laboratory staffmember Y. Hoshi, has not been tested with the DSP (Digital Signal Processor) and the LAFM. Figure 6.1 shows the total LAFM setup. On the isolation table an aluminum base plate is located, on which the mechanical head and the optics are mounted. Around the table control devices and monitoring equipment can be seen. The noise density of the optics with the mechanical head is investigated to compare the home-build system with a commercial system. A schematic of the test setup is shown in figure 6.2, no liquid is applied on the cantilever in the test setup. The excitation frequency applied to the cantilever is variated from 1 Hz to 2 MHz by a network analyzer via the excitation laser. The cantilever velocity-measurement is performed using a output power stabilized He-Ne laser(neoark - 2MSS). The resulting signal on the APD s is connected to the FM-demodulator and fed back to the network analyzer, without any filtering. The sensitivity of the FM-demodulator can be chosen at 0.1, 1, 10, 100 or 1000 mm/s/v. The resulting noise density is shown in figure 6.3. At the frequency-range of interest, approximately 0.3-1 MHz, the noise density lies between 50-100 fm/ Hz. This is already better then commercial Doppler interferometers, which offer typical values in de range 100-1000 fm/ Hz in air. (Fukuma 2005[11]) Furthermore, it should be noted that the presented noise density data is obtained before optimization of the optics. 24

25 Figure 6.1: Picture of the realized LAFM The noise density is also measured using a Laser Diode as laser source. The resulting noise density, with a sensitivity of 0.1 mm/s/v, is also shown in figure 6.3. The Laser Diode data shows good coherence with the He-Ne data and more research will reveal if the He-Ne laser source can be replaced by the laser diode. Figure 6.2: Schematic overview of the noise density test setup

26 CHAPTER 6. RESULTS Figure 6.3: Noise density of the LAFM excitation and measurement system

Chapter 7 Conclusions & Recommendations 7.1 Conclusions In this report a new design of a LAFM has been presented and the results of it s realization have been shown. The mechanical head of the LAFM handles the actuation and positioning of the cantilever and sample, these functions are performed by shear piezo s. The performance of the rough sample approach has been shown, with step sizes down to 2 nm. The measurement of the cantilever s frequency has been performed by a home made Doppler interferometer, which has a lower noise density level compared to commercial available systems. The excitation of the cantilever is realized by photothermal actuation, which is combined with the measurement optics and camera view on the cantilever into an all-in-one optical setup. The control circuitry for phase locked oscillation is implemented. The mechanical head, optics and control, together with the software left a fully equipped LAFM. It should be noted that debugging and optimization is needed before atomic resolution images of surfaces in a liquid can be obtained. 27

28 CHAPTER 7. CONCLUSIONS & RECOMMENDATIONS 7.2 Recommendations In the quest for ever increasing image quality the following recommendations should be considered. Experience with the first version of the LAFM at Kawakatsu laboratory has shown that an important limiting factor in getting high resolution images is the influence of temperature gradients. Redesign the mechanical head in such a way that the probe tip of the cantilever is as close as possible to the thermal center. This way temperature changes have minimal effect on the image. Furthermore this also gives opportunities for surface scans in a liquid with changing temperature. Low expansion coefficient and high conductivity materials should be selected to minimize temperature influence on the construction. To be able to predict the movement of the mechanical head under the thermal changes (and loads) the over constrained stages, with four piezo s each, should be replaces by a statically determined construction. More specific, the current rough sample positioning, which is a critical component in the approach procedure, should be replaced. The current wedge-construction is significantly over constrained, which leads to uncertainty in the rough steps. In general the mechanical head should be scaled down. The main reason for the current size is the accessibility to the cantilever to apply the liquid. The accessibility should be solved, since scaling down the construction can significantly decrease geometric and thermal errors.

Bibliography [1] G. Binnig, C. F. Quate, C. Gerber, Physical Review Letters, 56, 930-933 (1986) [2] B. W. Hoogenboom, H. J. Hug, Y. Pellmount, S. Martin, P. L. T. M. Frederix, D. Fotiadis and A. Engel, Applied physical letters, 88, 193109 (2006) [3] T. R. Albrecht, P. Grütter, D. Horne and D. Rugar, Applied physical letters, 69, 668 (1991) [4] F. J. Giessibl, Science, 267 pp. 68-71 (1995) [5] D. Ramos, J. Tamayo, J. Mertens and M. Calleja, Journal of applied physics, 99, 124904 (2006) [6] G. C. Ratcliff, A. E. Dorothy and R. Superfine, Applied physical letters, 72, 15 (1998) [7] S. Nishida, D. Kobayashi, T. Sakurada, T. Nakazawa, High-resolution imaging in liquid by Frequency Modulation Atomic Force Microscope (2005) [8] H. Kawakatsu, S. Kawai, D. Saya, M. Nagashio, D. Kobayashi, H. Toshiyoshi and H. Fujitao, Review of scientific instruments, 73, 6 (2002) [9] S. Nishida, D. Kobayashi, T. Sakurada and H. Kawakatsu, The 14 th International Colloquium on scanning Probe Microscopy, S7-5, 2006-12-08 [10] D. Kobayashi, presentation Control of a LAFM (2007) [11] T. Fukuma and S. P. Jarvis, Review of scientific instruments, 77 (2006) 29

Appendix A Mechanical drawings LAFM mechanical head A.1 Mechanical drawings lower part mechanical head 30

A.1. MECHANICAL DRAWINGS LOWER PART MECHANICAL HEAD31 - - -

32APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 1 2 Dural

A.1. MECHANICAL DRAWINGS LOWER PART MECHANICAL HEAD33 2 1 Brass

34APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 3 1 Brass

A.1. MECHANICAL DRAWINGS LOWER PART MECHANICAL HEAD35 5 1 Dural

36APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 6 1 SUS304

A.1. MECHANICAL DRAWINGS LOWER PART MECHANICAL HEAD37 7 1 SUS304

38APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 8 1 SUS304

A.1. MECHANICAL DRAWINGS LOWER PART MECHANICAL HEAD39 9 1 SUS304

40APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 10 1 SUS304

A.1. MECHANICAL DRAWINGS LOWER PART MECHANICAL HEAD41 11 1 SUS304

42APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 12 1 Dural

A.1. MECHANICAL DRAWINGS LOWER PART MECHANICAL HEAD43 13 1 Dural

44APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD A.2 Mechanical drawings upper part mechanical head

A.2. MECHANICAL DRAWINGS UPPER PART MECHANICAL HEAD45 - - -

46APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 14 1 SUS304

A.2. MECHANICAL DRAWINGS UPPER PART MECHANICAL HEAD47 15 1 SUS304

48APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 16 1 SUS304

A.2. MECHANICAL DRAWINGS UPPER PART MECHANICAL HEAD49 17-1 1 SUS304

50APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 17-2 1 SUS304

A.2. MECHANICAL DRAWINGS UPPER PART MECHANICAL HEAD51 17-3 1 SUS304

52APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 18 1 SUS304

A.2. MECHANICAL DRAWINGS UPPER PART MECHANICAL HEAD53 19 1 SUS304

54APPENDIX A. MECHANICAL DRAWINGS LAFM MECHANICAL HEAD 20 3 SUS304

Appendix B Piezo actuated stages The working principal of the shear piezo s is shown in figure B.1. By applying a trapezoidal input voltage with variable frequency to the piezo s, they walk with variable speed over the surface. Figure B.1: Working principal of shear piezo s Each stage rests on four shear piezo s. Two piezo s walk over the flat stainless steel surface and the other two are equipped with a bearing and walk over two bars. The two bars function as guide for the stage. The bottom view of a stage is shown in figure B.2. 55

56 APPENDIX B. PIEZO ACTUATED STAGES Figure B.2: Bottom view of a stage

Appendix C Test results sample step size The tests to determine the step size are performed at 1 Hz. The test setup is shown in figure C.1. The capacitance sensor used in the test setup has a sensitivity of 10µm/V. The resulting oscilloscope output, for 100, 150 and 200 V input amplitude, is shown below for the approach and retract situation. Figure C.1: Schematic overview sample z-movement test setup 57

58 APPENDIX C. TEST RESULTS SAMPLE STEP SIZE Figure C.2: V = 100V, approach Figure C.3: V = 100V, retract Figure C.4: V = 150V, approach Figure C.5: V = 150V, retract Figure C.6: V = 200V, approach Figure C.7: V = 200V, retract