High Speed Atomic Force

Microscope Design Using DVD

Optics
A thesis submitted in partial fulfillment of the requirements for the degree of Master
of Science at Virginia Commonwealth University.

by
Thomas Beatty Carlson
Bachelors of Science, Physics, Virginia Commonwealth University 2012
Director: Dr. Jason Reed, Assistant Professor, Physics Department

Virginia Commonwealth University

Richmond, Virginia
May 2014

Acknowledgment

I wish to express my gratitude to several people. I would like to thank my wife, Angel
for her undying support and faith in me. I wish to thank my children and my parents
for their support and help without question. I wish to my advisor Dr. Jason Reed and
my collaborator Dr. Loren Picco. I also wish to thank my committee members Dr.
Marilyn Bishop and Dr. Christopher. Finally, I want to thank the entire VCU Physics
Department for sharing their knowledge.

II

Table of Contents

I.

Introduction
1.1: About this study ............................................................................ 2

II.

AFM
2.1: AFM background........................................................................... 4
2.2: Specific applications .................................................................... 6
2.3: Sources of noise in AFM ............................................................... 7

III.

Commercial OPU
3.1: Advantages of the OPU .............................................................. 10
3.2: How the OPU works.................................................................... 11
3.3: Powering the OPU ...................................................................... 14

IV.

VCU test setup

4.1: Characterization of Noises in System Design .............................. 17
4.2: Investigation of Noise................................................................... 19
4.3: Results showing thermally excited resonance peak. ................... 20

V.

Imaging Results
5.1: Imaging of DNA samples ............................................................. 22
5.2: Improvements to the imaging process ......................................... 24

VI.
VII.

Conclusion .......................................................................................... 26
Bibliography ........................................................................................ 29

III

List of Figures

2.1-1: Standard AFM Setup ............................................................................ 5

2.3-1: Example of thermal spetra of resonance peak for AFM cantilever. ....... 8
3.2-1: Laser path through the OPU ................................................................ 11
3.2-2: Laser path through the OPU ................................................................ 13
3.2-3: SFE versus defocus distance................................................................ 14
3.3-1: Voltage divider circuit ......................................................................... 15
3.3-2: Mounting setup for HSAFM ................................................................. 16
4.1-1: Bruker MSNL-10 scanning probe ....................................................... 17
4.1-2: FFT spectrum of the displacement or motion of the cantilever in the
initial setup .................................................................................................... 18
4.2-1: FFT spectrum of the displacement or motion of the cantilever. The
newer noise floor
4.3-1: FFT spectrum of the displacement or motion of the cantilever in the
current set up. Shows thermally excited resonance peak Design ................. 21
4.3-2: Current System Setup ........................................................................ 21
5.1-1 Topographic images of unidentified features formed on the surface of a
mica substrate ............................................................................................... 23
5.1-2 Topographic images of the same sample as above showing additional
unidentified features formed on the surface of a mica substrate. ................. 24
5.2-1: Topographic images of features formed on the surface of a mica
substrate. ...................................................................................................... 25

IV

List of Abbreviations

AFM:

Atomic Force Microscope/Microscopy

CD:

Compact Disk

DAQ:

Data Acquisition

DC:

Direct Current

DVD:

Digital Versatile Disk

FES:

Focus Error Signal

FFT:

Fast Fourier Transform

HSAFM:

High Speed Atomic Force Microscope/Microscopy

LD:

Laser Diode

LDC:

Laser Diode Controller

NI:

National Instruments

OPU:

Optical Pickup

PDIC:

Photo Diode Integrated Circuit

ROM:

Read Only Memory

VCM:

Voice Coil Motor

VDC:

Voltage Divider Circuit

Abstract

HIGH SPEED ATOMIC FORCE MICROSCOPE DESIGN USING DVD OPTICS

By Thomas Beatty Carlson, MS Applied Physics

A thesis submitted in partial fulfillment of the requirements for the degree of Master of
Science at Virginia Commonwealth University.

Virginia Commonwealth University, 2014

Director: Dr. Jason Reed, Assistant Professor, Physics Department

We examine the design of a high speed atomic force microscope using an optical
pickup from a commercially available compact disc/digital versatile disc drive. An
investigation of the commercial optical pickup is done with the goal of determining how it
can be used for dimensional measurements on nanometer scale. An evaluation of noise
sources, imaging capabilities, and functionality is performed.

I.

1.1

Introduction

About this study

This study is part of larger collaborative effort to develop a high speed atomic
force microscope (HSAFM) using a commercial optical pickup (OPU) unit from a
CD/DVD disk drive. This approach is being utilized in only two places throughout
the world (Virginia Commonwealth University and University of Bristol). This
project was performed under the direction of Dr. Jason Reed, Virginia
Commonwealth University.
The AFM is a widely used tool in nano-science. Its purpose is to image and
characterize surfaces on the nanometer scale. The biggest problem facing AFMs
is slow scan speeds. In recent years scientists have begun to approach these
problems, leading to the emerging field of high speed atomic force microscopy
(HSAFM). A major advantage of HSAFM is the extremely fast operating speed.
Using optics system contained within the Optical PickUp (OPU) from a
commercial digital versatile disc (DVD) drive, a system could provide real time
imaging, the opportunity for low cost (under $10) and a compact size
(approximately 40 x 50 x 12 mm). These features are extremely desirable
because they are beneficial to the end user.

The goal of this project was to construct this system in our lab, investigate the
workings of the various components, and evaluate noises in the system. We
then determined the best options for optimization of functionality, noise reduction
and image quality.

II.

2.1

AFM

AFM Background
The Atomic Force Microscope (AFM) uses a cantilever with a sharp probe
at its end to scan the surface of a given sample. In the AFM, the cantilever tip is
brought into close proximity of a sample surface. Once the tip is in range,
interatomic forces between the tip and surface result in a deflection of the
cantilever according to Hookes Law (Binning & Quate, 1986, Vol. 56). Typical
spring constants range between 0.001 to 100 N/m, and deflections of the
cantilever can range from microns to fractions of a nanometer. Typical forces
between the cantilever and sample can range from 10-11 N to 10-6 N. For
comparison, the interaction between two covalently bonded atoms is of the order
of 10-9 N at separations of an angstrom. Therefore, non-destructive imaging is
possible. (Atomic Force Microscopy, E. Meyer. Progress in Surface Science, Vol.
41, pp. 3-49)

AFM images are formed by scanning a sample relative to the probing tip
and then measuring the deflection of the cantilever (in the vertical or z direction),
as a function of the lateral positions (x and y). At small separations, of the order
of an angstrom, the tip of the cantilever is in contact with the sample. In this
mode, repulsion forces allow the surface topography to be traced with high
3

resolution. Atomic resolution can be obtained under ideal conditions (Meyer,

1992, Vol. 41).

FIGURE 2.1-1: Standard AFM Setup (Uni-Greifswald).

Currently, the most common method for measuring the deflection of the
cantilever is optical beam deflection. As the cantilever flexes, the light from the
laser is reflected onto the split photo-diode. By measuring the difference signal,
changes in the bending of the cantilever can be measured. The adoption of this
method has been successful, because the sensitivity (better than 0.1 nm) is
sufficient to produce quality imaging.
Recently it has been shown that using the optical pickup from a
commercial CD/DVD ROM drive is sufficiently sensitive for detecting deflections
of a cantilever tip down to an atomic step (HWU, Kuang-Yuh HUANG, &
HWANG, 2006, Vol. 45, No. 3B). This implies that under certain conditions the
4

optical system in the CD/DVD pickup has a sensitivity level comparable to that of
the beam deflection method in terms of the displacement measurement of the
cantilever.

2.2

Specific Applications
A primary concern in our lab is the investigation of biological materials
such as DNA and other molecular complexes. AFM is a useful technique for
observing DNA. As an example, the AFM can be used to image DNA molecules
and DNAenzyme interactions. AFM images of static DNAenzyme interactions
give information about enzyme locations on DNA (Hansma, 1996). Measured
heights of DNA in AFM have been reported to vary from 0.5 to 1.9 nm in air.
Thus we wanted to make our system sensitive enough to detect variations in
height on this scale.
High speed techniques allow for the possibility of exploring molecular and
cellular dynamics in real time. Time resolution is critical factor limiting the use of
AFM in live cell studies. Recently scientists have used AFM provide images of
the chemical bonds between atoms. In some situations, imaging can depict how
a molecules structure changes during a reaction (Sanders, 2013).
Another possible area of interest will be nanoparticles. Particles on the
nanometer scale have unique optical, electronic, and structural properties that
are not available in either isolated molecules or bulk solids. Nanoparticles are

also becoming increasingly important in many areas such as biomedical

applications (A Rao, 2007). Our setup should allow us to visualize individual
particles and groups of particles.

2.3

Noise sources in AFM

If the OPU is to be used effectively as a detection system, noise sources
must be considered. A primary source of AFM noise results from optical shot
noise. Optical shot noise is due to random fluctuations in optical phase due to the
quantization of light into wave packets (Daw, 2001). Shot noise sets the
fundamental limit of detection of the optical beam deection method and its
reduction plays a primary role in enabling atomic resolution imaging (Aleksander
Labuda, 2012). Radiation pressure of the laser on the cantilever was also
considered. AFM experiments using this setup are performed at the resonance
frequency of the thermally excited cantilever, thus the thermal noise of the
cantilever (~ 10-12

) is greater than noise originating from the detector or from

back-action forces, such as radiation pressure (~ 10-15

) (Smith, 1994). Noise

in AFM systems can also arise in part due to thermal Brownian vibration of the
cantilever. Brownian motion is random motion that results from collisions with
atoms or molecules in a surrounding gas or liquid. The noise in the our detection
system must be less than that of the thermal noise in order to obtain thermal
noise limited performance. Our detection system must ensure that the noise floor
is sufficiently low as to allow us to observe the thermally excited resonance peak
6

of the cantilever, an example can be seen in figure 2.3-1. A high noise floor will
conceal the presence of resonance peak and make quality measurements
impossible.

FIGURE 2.3-1: Example of thermal spectra showing the resonance peak for the AFM
cantilever.

Another source of noise in systems that use optical detection is known as

intensity noise. It is in part the result of quantum noise associated with laser gain
and resonator losses. Intensity noise also results from technical noise sources
such as excess noise of the power source, vibrations of the resonator, and
thermal fluctuations in the gain medium. The resulting intensity noise often
becomes weaker at higher power. In most cases, the lowest possible intensity
noise level for laser beams results in a higher shot noise (Paschotta, 2008).
Custom electronics are needed to power portions of the OPU. Electronic
and electrical noises were also considered in this setup. As is the case with
optical systems, noise is ever-present in electronics. Electronic or circuit noise is,
7

in large part, the result of two processes. The first produces thermal noise
effects. Thermal noise, also known is Johnson noise is due the result of electrons
moving about in a resistive material. The second is well known as shot noise.
Shot noise is due to random fluctuations in direct current circuits. As current
flows through a device, there are random fluctuations in the transit time of
electrons. Current flow also causes the random emission of electrons emitting
from material surfaces. Both of these factors contribute to the effect of shot
noise. It is noted that thermal noise is generated in active and passive
components, while shot noise comes only from active components, such as a DC
power supply (Vergers, 1987).
Though it is possible for us to measure the overall RMS noise from our
electronic circuitry, we cannot predict all of the sources of noise. Improper
grounding of circuits and other equipment plugged into the same circuit can pose
potential noise problems. Using quality components can be an effective means of
reducing circuit noise. Many components are made of materials best known for
their low noise qualities. Shielding our setup, electrically isolating our setup from
other laboratory equipment, and reducing wire length, a source of electronic
cross-coupling, can all be effective ways to reduce noise.
Micro and nano-scale measurement experiments can be severely
hampered by small vibrations. For example, many optical applications require a
laser with a spot size of just a few microns. If the position of this spot is critical to
system performance, such as its position on the tip of a cantilever, then
vibrations with amplitudes in the micron range can inevitably cause experimental
8

failure. In our laboratory environment vibrations may be the result of equipment

such as computers, power supplies, and fans. Thus, it was necessary to
vibrationally to isolate our system using commercially available equipment.

III.

3.1

The commercial OPU

Advantages of the OPU in AFM deflection sensing

Standard AFM in systems can be designed to run at high speeds;

however, they require electronics designed to operate at high bandwidths with
low noise. Such equipment can be a major operating cost at $50,000 and above.
The OPU from a CD/DVD disc drive is the commercially designed to operate at
speeds in the MHz range. Wide availability and low cost (under $10) of the OPU
provide an attractive alternative to much of the equipment need to develop a
HSAFM system. Commercial CD/DVD pickups are also compact, lightweight,
and reliable, and thus may greatly simplify the design of HSAFM systems.
Another advantage of the OPU in HSAFM setups is that the astigmatic
detection system (discussed in section 3.2) is used to make quantitative
measurement of the height displacement of the cantilever (E.-T. Hwu, 2007). In
the standard optical beam deflection detection system, angular changes are
measured and assumed to be due to only to the bending of the cantilever below
its resonance frequency. This assumption is not always correct. In high-speed
AFMs, the cantilever often has oscillations at the second, third and even higher
harmonics (O. D. Payton, 2012). This adds extra bending to the cantilever which
10

gets incorrectly interpreted as height changes from the surface. A common

problem induced by multiple probe effects is the shadowing or multiplication of
small structures produced (Dufrne, 2002).

3.2

How the OPU works

The commercial OPU is comprised of several optical and electronic
components that are manufactured in a single casing. In this set up we use the
Phillips OPU66.50 D Optical Pickup Unit (Royal Phillips Electronics, 2005). The
laser diode provides light at a wavelength of 658 nm. A schematic of the light
path in the OPU can be seen in figure 3.2-1.

FIGURE 3.2-1: Laser path through the OPU (Science Blogs, 2012)

11

As light leaves the laser diode it immediately enters a collimating lens,

converting the Gaussian profile of the beam into a uniform one. Processing
optical data requires a collimated uniform laser beam that maintains the phase of
the incident wave front. The collimating lens is made of neutral density glass to
produce the Gaussian absorption profile so that the output beam becomes
uniform (Chang-Yuan Han, 1983, Vol. 22, No. 22).
Once the beam has been made uniform, it passes through a diffraction
grating that is used to split the main beam into a central beam for reading
information and side beams for tracking disc position and keep the main beam in
place. We remove the diffraction grating, as it is not necessary in our application.
This gives main beam a more powerful signal at lower driving currents.
Laser light then passes through a polarizing beam splitter, also known as a
polarizing prism. This divides the incident light beam into a reflected beam and a
transmitted beam. The beam splitter is also used to combine two incoming
beams of light (Saleh & Teich, 2012). Additionally, the polarizing beam splitter
provides the means to extract polarized light from un-polarized light. To do this
two prisms of uniaxial anisotropic material are cemented together. When light
enters an anisotropic crystal, the ordinary and extraordinary waves refract at
different angles, causing separation. At the interface, a dielectric film is used to
modify the reflective properties of the surface by exploiting the interference of
reflections from multiple optical interfaces (Paschotta, 2008).

12

A quarter wave plate also known as a /4 wave retarder, then serves to

convert the light into circularly polarized light. The laser is then focused by a
lens to a single point, seen in figure 3.2-2. The focus distance of the OPU is
2.75mm.

FIGURE 3.2-2: Laser light focused on the disc surface (Science Blogs, 2012).

In the return path, light reflected from the surface passes back through the
lens and the quarter wave plate where it is now 90 degrees out of phase with the
incoming light. This keeps the incoming and outgoing light beams from interfering
with each other. The detection of the focusing condition of the laser beam on the
optical disk adopts the astigmatic method, which uses a cylindrical lens as a
beam-shaping device (F. Quercioli, 1997). When the laser beams are perfectly
focused on the disk surface, the laser spot is circular, and evenly centered on the
four photo-detectors (AD) as seen in Figure 3.2-3. When the disk surface is
slightly higher or lower than the focus of the laser beam, the laser spot on the
central quadrant sensors becomes more elongated. The shape change of the
laser spot on the photo-detector is detected by the Focus Error Signal (FES).

13

FIGURE 3.2-3: FES versus defocus distance. The corresponding light spot
projected on the photodiode is shown at three different defocus distances.
(HWU, Kuang-Yuh HUANG, & HWANG, 2006, Vol. 45, No. 3B)

The FES returns linear displacement information from the disk, which is given by
the following equation:
(

).

where SA, SB, SC, and SD are the output voltage of each quadrant respectively, as
shown in Figure 3.2-3. Plotting the FES as the laser passes through the focus
region results in an S-shaped curve as seen in the figure. The linear region of
this S-curve allows us to determine displacement of the cantilever as a linear
function of the FES. The detectable displacement range of the Phillips OPU66.50
D is 6.5 m. The focus point is 3.25 m about the center of the beam

14

3.3

Utilizing the OPU

To use the OPU for measurement we needed to provide power and be

able to read back the voltage signals from it. To power the laser diode, we use
the Thorlabs LDC 201 C laser diode controller. The FES is provided by a PhotoDiode Integrated Circuit (PDIC) that requires voltage levels of 2.1 V, 2.5 V, and 5
V. The 5 V signal powers the PDIC, 2.1 V is used as a reference voltage for the
photodiode, 2.5 V powers the various gain settings. To do this we have designed
a voltage dividing circuit capable of taking an input voltage and outputting the
necessary voltage levels. The Agilent U8001A regulated DC power supply is
used to supply to power the circuit. A diagram of the circuit used is shown in
Figure 3.3-1.

FIGURE 3.3-1: Voltage divider circuit (Payton, 2014).

15

The FES returned by the PDIC is read by a National Instruments (NI)

USB-6259, a Data AcQuisition (DAQ) interface with NI LabView Software. The
DAQ is capable of reading input signals up to 1.25 MHz which will allow us to
collect data at the desired speeds. Labview allows us to create the necessary
Virtual Instruments (VI) for gathering data and displaying image information.
The OPU is mounted onto a manual x-y-z micrometer stage such that the
laser is directed straight down, as seen in figure 3.3-2. This gives us the freedom
to position the laser spot directly over top of the cantilever. The cantilever is fixed
to a stationary, transparent polycarbonate film. The sample sits on an x-y
scanning stage that is have fixed to a z-stage, allowing the scanning stage to
engage the cantilever. The scanning stage is capable of moving at high
frequencies in the x and y directions, allowing a sample to be imaged in real time.
The entire setup is mounted on a breadboard with rubber feet, and sits atop a
vibration isolation table.

16

FIGURE 3.3-2: Mounting setup for HSAFM. 1) x-y-z positioning stage 2)

cantilever and holder 3) OPU 4) x-y scanning stage 5) x-z positioning stage 6) 6
x 12 ground isolating breadboard 7) vibration isolation table.

17

IV.

4.1

VCU test setup

Evaluation of System Noise

The first goal of the HSAFM setup is detection of the thermally excited
resonance peak of the cantilever. We used the Bruker MSNL-10 scanning probe.
The MSNL-10 contains six cantilevers with varying degrees of force constants,
ranging from 0.01 N/m to 0.6 N/m and resonant frequencies that range from 7
kHz up to 125 kHz. The back of the cantilever is coated with reflective gold,
which returns a high percentage of the incident light. Cantilever C was chosen
for measurement of the thermal peak. Its resonance frequency is 7 kHz and the
spring constant is 0.01 N/m. The thermally excited amplitude of the cantilever
motion was measured to be 2 nm using a Polytec OFV-5000 laser Doppler
vibrometer.

FIGURE 4.1-1: Bruker MSNL-10 scanning probe (A) cantilever C, 305 m long
and 17 m wide at its end (B) tip radius is nominally 2 nm (Bruker Corporation).

18

The initial set up and testing did not yield the desired results. A review of
the Fast Fourier Transform (FFT) motion spectrum revealed a higher than
desired noise floor and a high number of frequency spikes, seen in figure 4.1-2.
The strength of the FES was also observed to be too small at 20 mV peak-topeak vs. the expected 200 mV peak to peak signal.
Issues related to noise were hypothesized to be the result of several
factors, the first of which is electrical noise inherent in the building that houses
our laboratory. Considerable crosstalk was found between our system and other
electronic equipment in close proximity. A weak signal can arise from poor
reflectivity. Thus the transparency of the polycarbonate film on which the
cantilever was mounted was questioned. Laser noise was also considered as a
possible factor. A highly unstable resonance peak was observed with the LDC
current set at 72 mA, near the laser diode maximum current rating of 80mA. The
spectral noise floor increased rapidly when LDC currents moved above 64 mA.

FIGURE 4.1-2: FFT spectrum of the displacement or motion of the cantilever in

the initial setup showing undesired noises and a high noise floor.
19

4.2

Mitigation of Noise sources

In an effort to mitigate the effects of noise, several strategies were

employed. Equipment and wiring was isolated from other electronic equipment. A
line conditioner was used to reduce noise present in the building electrical
supply. Additionally, the use of battery power in the VDC reduced electronic shot
noises attributed to the regulated DC power supply. To increase reflectivity, the
polycarbonate film was replaced, and the new film cleaned with an ethanol
solution to improve transparency. The FES output of the OPU has been shown to
suffer from serious signal drift problems (Hwu, 2012). Due to temperature
fluctuation of the ambient environment, signal drift can be thousands of
nanometers right after the optical head is switched on; thus the OPU was allowed
warm up for at least 10 minutes after powering on. To reduce optical intensity
noise and stabilize the laser output, the LDC current was set to 60 mA, 70% of
the maximum operating current.
These methods reduced the overall noise floor from

to

, and yielded a slight increase of the peak to peak voltage of the S-

curve (refer to figure 3.2-3) from 20 mV to 40 mV. The results can be seen in
figure 4.2-1.

20

FIGURE 4.2-1: FFT spectrum of the displacement or motion of the cantilever.

The newer noise floor is below initial noise floor levels (superimposed to show
improvements).

Because we were not able to see the thermally excited resonance peak
after the initial improvements, further mitigation was required. The use of battery
power for the DAQ revealed a faulty DAQ power supply, which was replaced.
This led to an orders of magnitude reduction in the noise floor from
to

. The FES signal was dramatically improved. The peak to

peak voltage of the S-curve now spanned 280mV, vs. 40 mV previously. This is
in the expected range. In addition, we observed the elimination of most of the
unidentified noises with characteristic frequencies

21

4.3:

Results showing thermally-excited cantilever resonance peak

As a result of these improvements, we were ultimately able to resolve the
thermally excited resonance peak. The thermal noise peak of the cantilever was
detected at the expected normal resonance frequency of 7 kHz. The results can
be seen in Figure 4.3-1 along with the initial noise floor level, the noise floor
levels after initial improvements, and a the low noise floor achieved in the final
setup.

FIGURE 4.3-1: FFT spectrum of the displacement or motion of the cantilever in

the current set up. Shows thermally excited resonance peak at the expected
frequency of 7 kHz and a low noise floor. The upper half of the y-axis, above
3.00p, has been logarithmically scaled to show noise floor improvements from
the initial setup

22

A picture of the current system under operation can in figure 4.3-2. A

computer with NI Labview software interprets signals from the DAQ. Electronic
equipment is used to power the OPU and control the scanning stage. An
oscilloscope is used to monitor the scanning stage position.

FIGURE 4.3-2: Current System Setup (1) Computer with NI Labview

software. (2) Scanning stage controller. (3) DC power supply for VDC. (4)
Laser diode driver. (5) DAQ (6) Microscope (7) Oscilloscope

23

V.

5.1

Imaging Results

Imaging of DNA samples

To perform imaging using the HSAFM system, custom VIs are used to
acquire deection signals in a continuous time series. Images based on the
information from the deflection signal are then displayed on screen. Because
biological materials are an area of interest in our lab, samples of DNA are
prepared for imaging. Samples are deposited on MgCl2-derivatized mica
substrates by fluid shear flow and washed three times with purified water before
being blown dry for 30 seconds (Jason Reed, 2012). The sample is set atop a
Mad City Labs HS2 x-y nano-positioning scanning stage (Mad City Labs Inc.,
2014) operated in closed loop mode. The fast scan direction (parallel to
cantilever) is set to a frequency of 100 Hz. Initial data sampling rates were set at
200 kHz, more than sufficient for collecting data.
Initial imaging using the current system was able to detect aperiodic
features on the surface of the sample. Figure 5.1-1 shows the topographic
images of features formed on the surface of the sample during its fabrication.
These are most likely salt crystals formed during sample deposition of DNA from
24

a dilute salt solution. Some salt and impurities bind to the mica and are not
washed off. The same sample has been imaged with other AFMs in our lab and
the particles appear similar in shape. The scanning stage amplitudes of 3 V and
6 V correspond to a scan area of 3 and 6 microns in (a) and (b) respectively.
Diameters range from 100 to 500 nm with the OPU, currently. The height is not
calibrated, and the height can only be estimated based on the size of these
objects as seen on in other AFM images.

FIGURE 5.1-1 Topographic images of features formed on the surface of a mica

substrate. (a) Surface feature resolved at a scan size of 3 microns (b) Same
feature resolved in a scan area of 6 microns. Images contain 50 lines of
resolution at 1000 pixels per line.

The dark patches seen around the objects are artifacts of bad fitting in the
image flattening algorithms, which use polynomial line fitting (S. A. Tsaftaris,
2008). This can introduce a non-existent valley like deviation of the background.

The surface was further scanned for DNA. However, only additional salt
crystals were found, seen in figure 5.1-1. In both cases, smaller scan sizes
increased the size of the feature on the viewing screen and larger scan sizes
25

decreased the size of the same feature, confirming that imaging is actually
occurring.

FIGURE 5.1-2 Topographic images of the same sample as above showing

additional features formed on the surface of a mica substrate.

5.2

Improvements in the imaging process

Initial images confirmed that imaging was taking place. However, the
quality was insufficient and needed to be improved to image DNA on the surface
of the sample. To produce an image, analog signals provided to the computer by
the DAQ produce discrete scan lines. Each scan line is divided into discrete
pixels for image processing. The ordering of pixels by rows is known as raster
order, or raster scan order. There are more pixels per line in the fast scan
direction than there are lines, and thus the resolution is higher in the fast scan
direction, and the pixels are not square. The horizontal resolution can be
improved by increasing the sampling rate of the cantilever detection signal, thus
increasing the number of detected height changes over the course of the scan

26

line. The vertical resolution of the image can be increased by increasing the
number of scan lines in the image.
The number of pixels per line was increased from 1000 to 2000, and the
number of lines of resolution from 50 to 125. The sampling rate of the cantilever
deflection signal was increased from 200 kHz to 1 MHz. A new sample was also
prepared in same manner as the previous sample. A scan of several regions of
the new sample showed improved image quality, evidenced by resolution of finer
structured details in surface-bound salt crystals. The amplitude of the scanning
stage was set to 6 V, corresponding to a scan area of 6 microns. The two
images below show features of relatively high spatial frequency, which indicates
good quality scanning based on our observations with these samples on other
AFMs. However we were ultimately unable to identify DNA using these settings
with our current setup.

FIGURE 5.2-1: Topographic images of features formed on the surface of a mica

substrate. Similar features were resolved in a separate scan areas, both 6
microns. Images contain 125 lines of resolution at 2000 pixels per line.

27

VI.

Conclusion

We have demonstrated the capability of the CD/DVD OPU to detect the

vertical cantilever displacement in a HSAFM system. The imaging of sub-micron
sized nano-particles was achieved. Using the current setup the HSAFM system
using DVD optics provides an low cost option that is practical for applications
involving near molecular sized objects.
The HSAFM we have developed is capable of detecting nanoparticles, but
not yet capable of resolving DNA on the surface of our samples. Further
advancements could be made to the system in an effort resolve smaller feature
sizes. The salt crystals that appear in our images are known to be only slightly
taller than the DNA (0.06 nm). Comparison of the noise spectrum of our system
with that of other HSAFM systems will show how much further work is needed to
match the performance of other systems.
Image processing techniques could be used to improve image quality.
One technique that has been shown to remove common image artifacts is the
use of notch lters. (O. D. Payton, 2012). Once a noise peak present in the of the
deflection signal data has been determined, it is possible to remove specific

28

contributions made by that frequency by applying a notch filter at the resonance

frequency. This method decouples surface features from higher-mode
oscillations of the cantilever and enables cantilever-induced noise sources to be
filtered out. Image averaging algorithms are another image processing technique
that can be used to remove random noise from a series of images. Random
fluctuations above and below actual image data gradually even out if a sample
area is averaged over time series images.
System functionality could be improved by taking advantage of the Voice
Coil Motors (VCM) inside the OPU (Hwu, 2012). Compared to the manually
operated micro-positioning stages currently used for OPU alignment, control of
the laser spot can be performed more precisely by using the z and x axis
actuators of the VCM. Focusing can be performed in the z direction, which has a
range of motion 550 m. Positioning in the x direction offers a motion range of
400 m. The VCM has a sensitivity of 1.11 mm/V and can be precisely
controlled by using a DC signal.
System functionality could further be improved by adding the capability of
tapping mode. Our system currently operates in contact mode, meaning that the
tip is kept in constant contact with the surface. During the scanning process the
tip to sample force interactions of the order of 10-8 N can be enough to deform
some biological materials. Contact mode also can have the tendency to sweep
away weakly immobilized objects. An alternative to contact mode is found in
tapping mode, in which the cantilever is oscillated vertically while scanning
occurs. This reduces the lateral frictional forces that can apply to the sample
29

during contact mode. Tapping mode would be advantageous for imaging

biological materials because of the force interaction between the cantilever tip
and the sample is less than that of contact mode, of the order of 10-10 N,
decreasing the likelihood of damage to the sample (P. K. Hansma, 1994).

30

VII.

Bibliography

A Rao, M. S. (2007). Characterization of nanoparticles using Atomic Force Microscopy.

Journal of Physics, 971-976.
Aleksander Labuda, M. L. (2012). Stochastic noise in atomic force microscopy. Physical
Review.
Binning, G., & Quate, C. F. (1986, Vol. 56). Atomic Force Microscope. Physical Review
Letters, 930-933.
Bruker Corporation. (n.d.). Bruker AFM probes. Retrieved May 1, 2014, from Bruker:
https://www.brukerafmprobes.com/Product.aspx?ProductID=3710
Chang-Yuan Han, Y. I. (1983, Vol. 22, No. 22). Reshaping Collimated Laser Beams with
Gaussian Profiles to Uniform Profiles. Applied Optics.
Daw, E. (2001, August). An Introduction to Shot Noise in Laser Beams. Retrieved
February 2014, from http://ligo.phys.lsu.edu/edaw/homepage/talks/shot.pdf
Dufrne, Y. F. (2002). Atomic Force Microscopy, a Powerful Tool in Microbiology.
Journal of Bacteriology, 5205-5213.
E.-T. Hwu, S.-K. H.-W.-S. (2007). Simultaneous detection of translational and angular
displacements of micromachined elements. Applied Physics Letters, 91, 221908.
F. Quercioli, A. M. (1997). Correlation optical velocimetry. The Optical Society.
Hansma, R. K. (1996). Atomic force microscopy of long and short double-stranded,
single-stranded and triple-stranded nucleic acids. Nucleic Acids Research, pp.
713-720.

31

Hwu. (2012). Anti-drift and auto-alignment mechanism for an astigmatic atomic force
microscope system based on a digital versatile disk optical head. American
Institute of Physics, Review of Scientific Instrument.
HWU, E.-T., Kuang-Yuh HUANG, S.-K. H., & HWANG, a. I.-S. (2006, Vol. 45, No. 3B).
Measurement of Cantilever Displacement Using a Compact Disk/Digital Versatile
Disk Pickup Head. Japanese Journal of Applied Physics, 2368-2371.
Jason Reed, 1. C.-L. (2012). Identifying individual DNA species in a complex mixture by
precisely measuring the spacing between nicking restriction enzymes with atomic
force microscope. Journal of the Royal Society Interface, 9 74 2341-2350.
Mad City Labs Inc. (2014). Nano-HS series. Retrieved May 2014, from Mad City Labs:
http://www.madcitylabs.com/nanohsseries.html
Meyer, E. (1992, Vol. 41). Atomic Force Microscopy. Progress in Surface Science, 3-49.
O. D. Payton, L. P. (2012). Improving the signal-to-noise ratio of high-speed contact
mode atomic force microscopy. Review of Scientif Instruments, 83, 083710.
P. K. Hansma, J. P. (1994). Tapping mode atomic force microscopy in liquids. Applied
Physics Letters, 64, 1738 .
Paschotta, D. R. (2008). Encyclopedia of Laser Physics and Technology. Wiley & Sons,
Inc.
Payton, O. D. (2014). Voltage Divde Ciruit.
Royal Phillips Electronics. (2005). OPU66.50 D Optical Pickup Unit for DVD +R/RW, R/RW, +R DL Data sheet. Royal Phillips Electronics.
S. A. Tsaftaris, J. Z. (2008). AUTOMATED LINE FLATTENING OF ATOMIC FORCE
MICROSCOPY IMAGES. Retrieved May 2014, from
http://users.eecs.northwestern.edu/~jzu850/papers_bib/tsaftaris08icip.pdf
Saleh, B. W., & Teich, M. C. (2012). Fundamentals of Photonics; 2nd Edition. John
Wiley & Sons Inc. .
Sanders, R. (2013, May). Scientists capture first images of molecules before and after
reaction. Retrieved May 1, 2014, from UC Berkley News Center:
http://newscenter.berkeley.edu/2013/05/30/scientists-capture-first-images-ofmolecules-before-and-after-reaction/

32

Science Blogs. (2012, April). Science Blogs. Retrieved March 2014, from Science
Blogs: http://scienceblogs.com/principles/wp-content/blogs.dir/467/files/2012/04/i2c6394b95db042cdf484322b78cfba9d-cdplay.gif
Smith, D. P. (1994). Limits of Force Microscopy. IBM Research Division.
Uni-Greifswald, B. (n.d.). Biophysik Uni-Greifswald. Retrieved April 2014, from
Biophysik Uni-Greifswald: http://www3.physik.unigreifswald.de/method/afm/AFM_laser.gif
Vergers, C. A. (1987). Handbook of Electrical Noise Measurement and Technology, 2nd
Edition. Blue Ridge Summit, PA: TAP BOOKS, Inc.

33