4.6 Problems and outlook
4.6.4 Electronics
RHK
At the moment the RHK electronics crashes sometimes. The PLLPro loses contact with the computer and the SPM1000 at random times. This causes the cantilever to oscillate with the last settings the PLLPro gave. During such a crash the PLL mode does not change the frequency any more so if it happens
during a measurement in PLL mode you might get a tip crash or at best a bad measurement. If it happens in lock-in mode or static mode the image can be completed but after that the system has to be rebooted. The crash occurs almost every hour but often already after 5minutes of a reboot. RHK is looking into the problem at the moment.
Attocube
The positioners are still controlled by the attocube hardware. To move the positioners at room temperature a voltage of 30V is used and at cryogenic temperatures a voltage of 40V. Three times we noticed that the voltage given out by the controller was 30V lower than the voltage it should give, which resulted in no voltage at room temperature and a 10V signal at lower temperatures. It looks like the connection between the positioners and the controller is broken but that was not the case. A reboot of the controller solves the problem.
An additional problem with the attocube controller is that at a certain tem-perature between 77K and 4K, when cooling down from liquid nitrogen to liquid helium, the controller gives a noisy voltage signal to the dither piezo. This noise is large and in addition it does not give the correct voltage to the dither. The problem was solved by switching the control for the dither and the y-scanner.
The breakdown of the dither controller at low temperatures is most likely caused by the low capacity of the dither piezo.
lift heigth
lift heigth
lift heigth (a)
(b)
c) (
Figure 4.21: Schematic scan lines in lift mode. The black lines are the forward scans and the red lines are the backward scans. The dashed lines indicate the first scan to measure the topography and the whole lines correspond with the lift mode to detect the magnetic contrast. (a) Ideal case where the forward scan and backward scan overlap and in lift mode the tip-sample distance is constant.
The phase is used to measure the magnetic contrast during the secons scan.
(b) The forward and backward scan do not overlap. The height at the start of the second forward scan is the height at the start of the first forward scan plus the lift heigth. (c) The forward and backward scan do not overlap. The height at the start of the second forward scan is the height at the end of the first backward scan plus the lift heigth. When the mismatch between forward and backward scan is larger then the lift height the tip crashes into the sample.
Chapter 5
Conclusion
In lock-in mode successful AFM measurement have been performed on the cal-ibration sample. This was done at room temperature and at liquid nitrogen temperature. The lock-in mode is hard to operate at liquid helium temperature due to the large Q. The large Q causes the resonance peak to become so narrow that when the tip approaches the sample the drive frequency is completely off resonance. The PLL mode makes it possible to image at liquid helium temper-ature but the resolution is not good due to the large tip-sample distance caused by the instability of the PLL mode.
The scanners show a large nonlinearity, in particular the y- and z-scanner at room temperature. To appoint the right sizes to the images an image correction needs to be made. All three scanners were calibrated for room temperature, liquid nitrogen temperature and liquid helium temperature. After the calibra-tion of the scanners a sample with clustered quantum dots was imaged and 4nm high features can clearly be distinguished and quantum dots 150nm apart can be imaged separately.
MFM measurements have also been performed at the three different tem-peratures. The hard disk was used to perform the MFM measurements. MFM measurements were done in lock-in mode at a constant height above the sample.
At cryogenic temperatures the signal to noise ratio of the phase signal increased resulting in a better MFM contrast. The lift mode was tested but because of its instability it was not used for the MFM measurements yet. Therefore the PLL mode was not used even at liquid helium temperature. Bits lying 500nm apart could be imaged on the hard disk. A new sample with magnetic bits closer together should be imaged to see what the limit of the resolving distance is for the MFM.
Appendix A
Image Correction
In chapter 4 the image correction was discussed. The distance difference per voltage difference was shown for the x-axis at room temperature. Here the other temperatures and axis will also be shown in figures A.1 to A.5as well as the Matlab script used to perform the image correction.
0 20 40 60 80 100 120 140 160
200 250 300 350 400 450 500
Dx/DV(nm/V)
77K x scanner Polynominal fit ofDx/DV
Position (V)
Figure A.1: The distance difference per voltage difference for the x scanner at 77K. The fit is a polynomial fit of the third order and gives ∆V∆x = 195 + 1.5V − 0.003V2+ 0.00003V3.
The Matlab script used to do the image correction can be see in figure A.6.
The script works as follows, first the raw data is loaded, then for each column in the x-direction the real distance instead of the equidistant voltage applied to the piezos is calculated. Then an interpolation is made to add 9 columns between all neighboring column. Next the original amount of columns is chosen that are equidistant in x-extension of the scanner. After that the process is repeated for the y-direction
0 20 40 60 80 100 120 140
Figure A.2: The distance difference per voltage difference for the x scanner at 4K. The fit is a polynomial fit of the third order and gives ∆V∆x = 145 + 0.4V +
Figure A.3: The distance difference per voltage difference for the y scanner at room temperature. The fit is a polynomial fit of the third order and gives
∆x
Figure A.4: The distance difference per voltage difference for the y scanner at 77K. The fit is a polynomial fit of the third order and gives ∆V∆x = 180 + 1.9V − 0.01V2+ 0.00007V3.
0 20 40 60 80 100 120 140 160
170 180 190 200 210
4K y scanner Polynominal fit ofDy/DV
Dy/DV(nm/V)
position (V)
Figure A.5: The distance difference per voltage difference for the y scanner at 4K. The fit is a polynomial fit of the third order and gives ∆V∆x = 130 + 0.85V − 0.005V2+ 0.000025V3.
close ;clear ;clc;
%the number of interpolation points between 2 pixels
%The maximum voltage to the scanner, for the maximum range at RT 60, at cryogenic temperatures 150V.
%Starting Voltage of the scanner.
%The voltage applied to the x scanner at each pixel
%RT correction
%x=195*V+0.5*1.5*V^2-(1/3)*0.003*V^3+0.25*0.00003*V^4;%77K correction
%x=145*V+0.5*0.4*V^2+(1/3)*0.003*V^3-0.25*0.000002*V^4;%4K correction
%The real distance at each line
%matrix of the z value at each line
%number of pixels after interpolation
%distance at each new pixel
%Matrix with two columns, first the line number then the distance.
%interpolation of the z value to the new pixels.
%Pixels are chosen equidistant from the newly generated pixels.
%the same for the y direction.
%RT correction
Figure A.6: The M-file used to do the image corrections
Bibliography
[1] C.F. Quate G. Binnig and Ch. Gerber. Atomic force microscope. Physical Review Letters, 56(9):930–933, 1986.
[2] P.F. Fewster R.W. Martin B.Ya Ber A.P. Kovarsky C.R. Staddon K.Y.
Wang K.W. Edmonds C.T. Foxon L.X. Zhao, R.P. Campion and B.L. Gal-lagher. Determination of the mn concentration in gamnas. Semiconductor Science and technology, 20(5):369–373, 2005.
[3] N.F. van Hulst C.A.J. Putman, B.G. De Groot and J. Greve. A detailed analysis of the optical beam deflection technique for use in atomic force microscopy. Journal of Applied Physics, 72(1):6–12, 1991.
[4] G. Verkerk. Binas. Wolters-Noordhoff Groningen, 1992.
[5] J. I. Springborn. Magnetometrie an Halbleiter-Nanostrukturen mit wenigen Elektronen. Cuvillier Verlag G¨ottingen, 2007.
[6] T. Glatzel S. Sadewasser, P. Carl and M. Lux-Steiner. Influence of uncom-pensated electrostatic force on height measurements in non-contact atomic force microscopy. Nanotechnology, 15, 2004.
[7] U. Hartmann. Fundamentals of non-contact force microscopy. Advances in Electronics and Electron Physics, 87, 1994.
[8] U. Hartmann. The point dipole approximation in magnetic force mi-croscopy. Physics letter A, 137(9):475–478, 1989.
[9] J. D. Jackson. Classical Electrodynamics. 1999.
[10] U. Hartmann. Magnetic force microscopy. Annual Review of Materials Science, 29, 1999.
[11] U. Hartmann. Magnetic force microscopy: Some remarks from the mi-cromagnetic point of view. Journal of Applied Physics, 64(3):1561–1564, 1988.
[12] Young and Freedman. University Physics. 2004.
[13] C. J. Chen. Introduction to scanning tunneling microscopy. Oxford Uni-versity Press US, 2007.
[14] M. Binggeli R. Imura J. Bruggerlj, J. Burger and N.F. de Rooijl. Lat-eral force measurements in a scanning force microscope with piezoresistive sensors. Transducers ’95 Eurosensors IX, 1995.
[15] C.C. Williams Y. Martin and H.K. Wickramasinghe. Atomic force microscope-force mapping and profiling on a sub 100-˚A scale. Journal of Applied Physics, 61(10):4723–4729, 1987.
[16] D. Horne D. Rugar. T. R. Albrecht, D. Gr¨utter. Frequency modulation detection using high-q cantilevers for enhanced force microscope sensitivity.
Journal of Applied Physics, 69(2):668–673, 1990.
[17] W. Allers U.D. Schwarz 1 H. Fuchs H. Holscher, B. Gotsmann and R. Wiesendanger. Measurement of conservative and dissipative tip-sample interaction forces with a dynamic force microscope using the frequency modulation technique. Physical Review B, 64(075402), 2001.