Accepted Manuscript
Atomic Force Microscope Scanning Head with 3-Dimensional Orthogonal Scanning to eliminate the curved coupling Yushu Shi , Wei Li , Sitian Gao , Mingzhen Lu , Xiaodong Hu PII: DOI: Reference:
S0304-3991(17)30354-6 10.1016/j.ultramic.2018.03.020 ULTRAM 12557
To appear in:
Ultramicroscopy
Received date: Revised date: Accepted date:
30 July 2017 17 February 2018 26 March 2018
Please cite this article as: Yushu Shi , Wei Li , Sitian Gao , Mingzhen Lu , Xiaodong Hu , Atomic Force Microscope Scanning Head with 3-Dimensional Orthogonal Scanning to eliminate the curved coupling, Ultramicroscopy (2018), doi: 10.1016/j.ultramic.2018.03.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Highlights
A novel probe scanning atomic force microscopy with 3 dimensional planar scanning has been developed.
New optical lever amplification optical path is designed to
CR IP T
compensate the shift of beam spot during scanning.
Compared with tube scanning AFM, the scanning head reduces the
AC
CE
PT
ED
M
AN US
curvature distortion significantly.
ACCEPTED MANUSCRIPT
Atomic Force Microscope Scanning Head with 3-Dimensional Orthogonal Scanning to eliminate the curved coupling
a
CR IP T
Yushu Shia,b, Wei Lib, Sitian Gaob* Mingzhen Lub, Xiaodong Hua
State Key Laboratory of Precision Measuring Technology and Instruments, Tianjin University, Tianjin 300072, China b
National Institute of Metrology, Beijing, 100029, China
AN US
* Corresponding author. E-mail address:
[email protected]
PT
ED
M
Abstract: An atomic force microscopy (AFM) scanning head is designed with the probe orthogonal scanning mode for metrological AFM to eliminate the curvature distortion. The AFM probe is driven by piezo stage and the scanning trajectory of the probe in 3 directions are orthogonal to reduce the cross coupling. A new optical lever amplification optical path is developed to eliminate the coupling error. The tracing lens and probe tip are moved as an integrated part. The AFM is operated at contacting mode. The step approach process of the probe tip is tested to the sample surface and the noise of the AFM head is analyzed. The response of the probe demonstrates a 0.5 nm resolution of the probe head in the z direction. Finally, the planar scanning performance of the scanning head is demonstrated compared with tube scanning AFM.
AC
CE
Keywords: Atomic force microscope, Scanning head, Nanometrology, Surface topography
ACCEPTED MANUSCRIPT 1. Introduction The AFM has been a powerful tool and widely used in nanoscience research and industry to measure the surface topography with nanometer resolution [1]. With the development of nanotechnology and semiconductor industry, metrological AFMs with high accuracy are demanded for quantitative characterization of nanostructure and quality control in nano manufacturing. Over the past 20 years, metrological AFMs are
CR IP T
developed in national metrological institutes (NMIs) around the world to trace the measurand to SI Unit by using interferometers to measure the scanning displacement [2,3]. Nano structure standards such as step height, grating pitch and linewidth artifacts are calibrated by metrological AFMs. These transfer standards are in turn used to calibrate the nano measurement instruments such as AFM in industrial and
AN US
scientific research laboratories [4].
The scanning range of AFMs is currently limited to hundreds of microns due to the piezo actuator used for scanning. However metrological AFM with scanning range above millimeter is desired for characterization of nanostructure and process and product quality control in the semiconductor industry. Recently, several NMIs have
M
developed large range metrological AFMs from 1 millimeter to tens of millimeters [3,5]. The large scanning range is either accomplished by direct stage scanning or
ED
image stitching [6]. Physikalisch-Technische Bundesanstalt (Germany) developed a metrological large range scanning probe microscope with a measuring range of
PT
25mm×25mm×5mm based on Nano Measuring Machine [7]. National Institute of Metrology (NIM) in China also developed a nano measurement machine with
CE
50mm×50mm×2mm scanning range [8]. Most of the metrological instruments are sample scanning mode, for the
AC
displacement of stage is orthogonal and can be measured with interferometers. However, with increasing of calibration standard adopted in microelectronics manufacturing, 300 mm sized wafer sample will be tested in metrology laboratory and the scanning speed will be limited by the sample size in sample scanning mode. A tip scanning AFM can provide low noise and high speed despite the sample size. However, common piezo tube driven probe is not suitable for metrological AFM, for the coupling of the driver in x-y plane and z-direction caused by the bent of the tube in scanning. The coupling between z and x-y direction can be eliminated by
ACCEPTED MANUSCRIPT separate the sample x-y scanning from the tip displacement in z direction [9]. 3-dimensional tip scanning mode metrological AFM head with low movement coupling is demanded for large sample measurement. One difficulty of probe scanning AFM is to keep the laser beam focused on the cantilever during lateral scanning. And the second challenge is that the laser spot position bounced on to the position sensitive detector (PSD) should not shift with the
CR IP T
lateral scanning motion of tip but only sensitive to the deflection of cantilever. The AFM scanning head with orthogonal planar scanning is presented here. NIM has established a large range metrological AFM with versatile AFM head. The instrument can operate in both sample scanning mode and tip scanning mode. The AFM probe is driven by a 3-dimensional piezo stage. The scanning movements of the probe in 3
AN US
directions are orthogonal to reduce the cross coupling. The special optical path is designed to focus the laser on to the cantilever and the reflected beam from cantilever is deflected to keep the constant position during lateral scanning. The performance of the head is tested in vertical and lateral scanning. The scanning head design introduced here can be applied not only in the large range metrological AFM but also
ED
2. The principle of instrument
M
the common AFM.
The structure of the AFM head is shown in Figure 1. The laser from laser diode is delivered by optical fiber with a collimating lens to minimize the thermal effects to
PT
the instrument. To reduce the laser fluctuation caused by the laser reflected back to the diode, a Faraday isolator is coupled to the output of laser diode. The collimated 3 mm
CE
diameter laser beam reflected by a mirror and focused on the cantilever by a tracing lens. The laser reflected from the cantilever is focused by a deflection lens to an 8 mm
AC
diameter spot onto the position sensitive detector (PSD) with a detection area of 10 mm×10 mm. The fiber, deflection lens, and PSD are fixed on the AFM frame. The tracing lens and probe are mounted on the 3-dimensional flexure piezostage (P-363.3CD, PI) as an integrated part to scanning. The stage guarantees the planar scanning in 3 orthogonal directions.
CR IP T
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN US
Figure 1. Scheme diagram of the AFM scanning head.
Figure 2. The principal of the laser spot tracing on the cantilever.
The mirror on the AFM frame is manually adjusted to focus the beam on the cantilever. The focus spot on the cantilever is 12 μm in diameter, which is smaller than the cantilever width, guaranteeing that the power of the laser is totally reflected by the cantilever. However, if the lateral scanning shifts the tip 5 microns relative to the focus, the laser will leak to sample surface and reduce the reflected light, even
ACCEPTED MANUSCRIPT more, cause noise by the interference of the laser from sample and cantilever reflection. So in this design, the probe cantilever is positioned at the focus of the tracing lens, then the lateral scanning of the tip and lens together will maintain the spot position on the backside of the cantilever (Fig 2). The customer designed lens is made of PMMA to reduce the mass of scanning part. With the tracing lens, the laser beam spot will vary with the scanning. A deflection
CR IP T
lens is introduced in the collection optical path before the PSD. To clearly demonstrate the design of the optics path and the tracing effects of the lens in scanning, the unfolded optical path of the optical lever amplification with lateral scanning is shown in Figure 3. The incident angle on the cantilever will shift with the lateral scanning and the beam will deviate from the center of PSD. The collection lens
AN US
L3 deflects the beam beak to the optical axis to compensate the shift. By placing the PSD at the intersection of the incident laser optical axis and the deflected beam, the
CE
PT
ED
M
spot center position on the PSD is insensitive to the lateral position of the tip.
AC
Figure 3. Principle of planner scanning AFM. L1 and L2 are tracing lenses following the lateral scanning of tip, L3 is deflection lens.
The deflection of the cantilever induced by the cantilever displacement z is given by [10] 3 z , (1) 2 l where l is the cantilever length. The beam deflection angle is 2. The displacement of the cantilever is amplified to the displacement of the spot on the PSD. The spot displacement is given by
ACCEPTED MANUSCRIPT
s
3zd ' , l
(2)
where d’ is the reduced distance between the PSD and cantilever taking account for the magnification of the collection lens. The amplification s/z of the measuring head is about 2000 by calculation from the instrument optics path. 3. Results and discussion
CR IP T
The scanning performance of this probe scanning AFM is tested by measure the PSD signal during scanning. The probe is driven by 3-dimensional piezostage with 5μm scanning range. However, the lateral scanning range can be extended to tens of microns with this design principally. The driving signal sent to the controller is generated by the function generator. The sample stage is raised to approach the probe
AN US
and sinusoidal and step wave are used to drive the probe movement in vertical direction. The position signal from capacitive sensor of the piezostage and reflected laser spot from PSD are measured simultaneously.
Snap off Retract
M
10
Snap in
Approach
ED
PSD signal (V)
5
0
Retract
-10 -150
-100
-50
0
50
100
150
Z (nm)
AC
CE
PT
-5
Figure 4. Response of the probe tip deflection when approaching to sample.
Figure 4 shows the optical deflection signal of the cantilever as a function of the
piezostage displacement. The probe oscillates with amplitude 200 nm at 10 Hz to repeat the progress. The approaching and retract progress is shown in Figure 4. As the tip approaches the sample surface, the snap in of tip due to the attractive force is observed. With the tip move towards the sample surface, the repulsive contact force will increase in the contact region, and the deflection of laser spot cantilever is
ACCEPTED MANUSCRIPT proportional to the z displacement of the cantilever. When the cantilever is retracted from the sample surface, the tip will bend to the sample due to the adhesive force until the cantilever overcomes the adhesive. The tip snaps off the sample to the free state. The sensitivity of the measuring head is defined as the change of the PSD output V with the probe position z,
V 3d ' P RIV z la
(3)
CR IP T
S
where P ,RIV, and a are the quantum efficiency of the PSD, the attenuation of the optical path, the laser power, gain of the PSD, current to voltage feedback resistor, and laser spot size assumed to be square respectively [10]. The sensitivity
AN US
calculated from the slope of the linear part in Figure 4 is 0.186V/nm. Then the resolution of the measuring head can be measured from the voltage signal. The resolution of the probe is measured by step motion of the probe in z direction. The PSD detector signal is recorded as the step movement. By driving the probe with piezostage, the distance of the probe to the sample is controlled within the range of
M
contact mode. The driving signal is step wave and the displacement of the stage is detected by capacitive sensor. The result is shown in Figure 5. In the range of the
ED
probe displacement 50 nm, the signal is linear with the probe displacement. The step of the motion of the probe in the z direction is 0.5 nm as shown in Figure 5(b). The
PT
PSD signal shows the consistent step as the displacement of the probe. The distinct step of the PSD output demonstrated that the resolution of the measuring head in z
CE
direction is better than 0.5 nm. The noise spectral density is shown in Figure 6. The density of the deflection noise
AC
was estimated to be <10-12 m / Hz . The curve is fitted with Equation 4, where kB, T, f , f0, Q, and k are Boltzmann constant, temperature, frequency, resonant frequency, quality factor and spring constant of the cantilever. The peak at 25 kHz corresponding to the resonance frequency of cantilever is caused by the thermal vibration.
nzB
2kBT 1 2 2 kf0Q [1 ( f / f 0 ) ] [ f /( f 0Q)]2
(4)
ACCEPTED MANUSCRIPT
10
100 PSD Signal Displacement/ nm
80
PSD signal /V
5
60
CR IP T
0 40
-5
20
0
5
10
15
0 20
AN US
-10
Displacement/nm
(a)
M
Time/s
ED PT
6
4
AC
CE
PSD Signal/V
(b)
5
2
Displacement/nm
6
PSD signal Displacement/nm
0.0
0.2
0.4
0.6
0.8
0 1.0
Time/s
Figure 5. PSD signal as the step motion of the probe in z direction. (a) PSD signal in 50 nm scanning range of the probe in z direction, (b) 0.5 nm step displacement.
Measured Fitting
-2
10
-3
10
-4
10
4
0
4
1x10
2x10
4
4
3x10
4x10
AN US
Frequency/Hz
CR IP T
Noise density/ nm/(Hz)
1/2
ACCEPTED MANUSCRIPT
Figure 6. Noise spectrum of the scanning head. The fitted parameter f0=25kHz, Q=100 and k=0.2N/m.
The lateral scanning performance of the AFM head is tested with a Si step height
M
standard sample. The sample is measured with both tube scanning AFM and the planar scanning AFM. The scanning range is 5 μm on the flat top of step height
ED
sample. As shown in Figure 7, the tube scanning causes about 1 nm coupling error, resulting in an obvious bow shaped artefacts. The planar scanning head reduces the
PT
cross coupling to about less than 0.5 nm on the same order of the noise.
Height (nm)
AC
CE
2
Tube scanning Planar scanning
1
0
0
1000
2000
3000
4000
5000
X (nm)
Figure 7. Influence of the lateral scanning of the probe to the deflected laser spot.
ACCEPTED MANUSCRIPT 4. Conclusion A probe scanning AFM measuring head with orthogonal scanning for the large range metrological atomic force microscopy is presented. The AFM probe is driven by piezostage and the scanning movements of the probe in 3 directions are orthogonal, eliminating the cross coupling of tube scanning. The tracing lens guarantees the focal spot on the tip and the deflection lens deflect the beam spot back to the PSD center.
CR IP T
The response of the probe demonstrates a 0.5 nm resolution of the probe head in z direction. The lateral movement of the probe has no influence on the spot position on PSD compared to tube scanning AFM head. The AFM head can not only be used in a metrological AFM, but also applied in commercial large sample AFM with high
AN US
scanning speed and range.
Acknowledgment
This work is supported by Special Funding Project for Public Industry (10-1); and
M
the National Key R&D Program of China (2016YFF0200602).
ED
References
[1] G. Binnig, C. F. Quate, Ch.Gerber, Phys. Rev. Letts.56 (1986) 930, doi: 10.1103/PhysRevLett.56.930.
PT
[2] J. A. Kramar, R. Dixson, N. G. Orji, Meas. Sci. Technol. 22 (2) (2011) 024001,
CE
doi: 10.1088/0957-0233/22/2/024001. [3] A. Yacoot, L. Koenders, Meas. Sci. Technol. 22 (12) (2011) 122001,
AC
doi:10.1088/0957-0233/22/12/122001. [4] V. Korpelainen, A. Lassila, Meas. Sci. Technol.18 (2) (2007) 395–403, doi: 10.1088/0957-0233/18/2/S11. [5] C. Werner, P. C. J. N. Rosielle,
M. Steinbuch , Int.
J. Mach. Tools Manuf, 50
(2) (2010) 183–190, doi: 10.1016/j.ijmachtools.2009.10.012. [6] P. Klapetek, M. Valtr, L. Picco, O. D. Payton, J. Martinek, A. Yacoot, M. Miles, Nanotechnology 26 (6) (2015) 065501, doi: 10.1088/0957-4484/26/6/065501.
ACCEPTED MANUSCRIPT
[7] G. Dai, H. Wolff, F. Pohlenz, H. Danzebrink, Rev. Sci. Instrum. 80 (4) (2009) 043702, doi: 10.1063/1.3109901. [8] M. Lu, S. Gao, Q. Jin, J. J. Cui, H. Du, H. T. Gao, Meas. Sci. Technol. 18 (6) (2007) 1735–1739, doi: 10.1088/0957-0233/18/6/S11. [9] J. Kwon, J. Hong, Y. Kim, D. Lee, K. Lee, S. Lee, S Park, Rev. Sci. Instrum. 74 (10) (2003) 4378–4383, doi: 10.1063/1.1610782.
AC
CE
PT
ED
M
AN US
CR IP T
[10] T. Fukuma, M. Kimura, K. Kobayashi, K. Matsushige, H. Yamada, Rev. Sci. Instrum. 76 (5) (2005) 053704, doi: 10.1063/1.1896938.