Transcript Introduction to Electron Beam Lithography 5A5676

Overview of Nanofabrication
• Material depostion methods
– Thin films of materials
– Thickness measurement
• Lithography
– Pattern transformation on to planar suface
– Direct write, or mask reproduction
• Imaging and Metrology methods
– Electron Microscopy
– Scanning probe microscopy
Thin film deposition techniques
• Vacuum deposition Methods
– UHV (<10-8), HV
– Sputtering
– CVD
– Laser Oblation
– Thermal deposition
• Boat or crucible, E-gun
– Epitaxy, growth models
Sputtering
substrate
B
Target material
Ar
Vacuum + 10-3 Torr Ar
E
•RF plasma rectifies RF power,
gives DC acceleration voltage
Ar, N2
•Ions circle B field lines,
increase colisson probability
RF Power
E-beam evaporator
B
Permanent
Magnet
HV
E-gun
Filament
Thermal CVD system
Precurser Gas
For growing
Carbon Nanotubes
http://www.iljinnanotech.co.kr/en/material/r-4-4.htm
Carbon Nanotubes
http://www.iljinnanotech.co.kr/en/material/r-4-4.htm
MBE
"Molecular Beam Epitaxy is a versatile technique for
growing thin epitaxial structures made of
semiconductors, metals or insulators."
In a ultra-high vacuum, a beam of atoms or, more
general, a beam of molecules is directed towards a
crystalline substrate such that the atoms or
molecules stick at the substrate’s surface forming a
new layer of deposited material. But where is the
difference between MBE and other material
deposition methods as e.g. thermal vacuum
evaporation?
http://www.wsi.tu-muenchen.de/E24/resources/facilities.htm
MBE and surface analysis chamber
The Knudsen Cell (effusion cell)
http://www.grc.nasa.gov/WWW/RT2002/5000/5160copland.html
Lithography
•
•
•
•
Spin coat radiation sensitive polymer - Resist
Expose layer (through mask or direct write)
Develop
Etch away or deposit material
Positive and negative resist
Liftoff requires undercut
Resist Contrast Curve
 D2 
Contrast:   log 
 D1 
1
Logarithmic measure of slope of contrast curve
100%
100%
Film Retention
Positive Resist
Film Retention
Negative Resist
0%
0%
D1
D2
D1
D2
Positive Resist Chemistry
Molecular weight shift
Typical Positive Resist process
EXAMPLE PROCESS: AZ5206 POSITIVE MASK PLATE
• Soak mask plate in acetone > 10 min to remove the original photoresist.
Rinse in isopropanol, blow dry.
• Clean the plate with RIE in oxygen. Do not use a barrel etcher.
RIE conditions: 30 sccm O2, 30 mTorr total pressure, 90 W (0.25 W/cm2), 5 min
• Immediately spin AZ5206, 3 krpm.
• Bake at 80 C for 30 min.
• Expose with e-beam, 10 kV, 6 C/cm2, Make sure the plate is well grounded.
(Other accelerating voltages may be used, but the dose will be different.)
• Develop for 60 s in KLK PPD 401 developer. Rinse in water.
• Descum - important Same as step 2 above, for only 5 seconds
Or use a barrel etcher, 0.6 Torr oxygen, 150W, 1 min.
• If this is a Cr plate, etch with Transene Cr etchant, ~1.5 min.
If this is a MoSi plate, then RIE etch:
0.05 Torr total pressure, 0.05 W/cm2, 16 sccm SF6, 4.2 sccm CF4,1 min.
• Plasma clean to remove resist: same as step 2 above, for 3 min.
Negative Resist Cemistry
Typical Negative resist process
EXAMPLE PROCESS: SAL NEGATIVE MASK PLATE
•Soak mask plate in acetone > 10 min to remove photoresist.
•Clean the plate with RIE in oxygen. Do not use a barrel etcher.
RIE conditions: 30 sccm O2, 30 mTorr total pressure, 90 W (0.25 W/cm2), 5 min.
•Immediately spin SAL-601, 4 krpm, 1 min.
•Bake in 90 C oven for 10 min. This resist is not sensitive to room light.
•Expose at 50 kV, 11 C/cm2. Be sure the plate is grounded.
•Post-bake for 1 min on a large hotplate, 115 C.
•Cool for > 6 min.
•Develop for 6 min in Shipley MF312:water (1:1) Be sure to check for underdevelopm
•Descum 30 s with oxygen RIE: same as step 2, 10 s.
•Etch with Transene or Cyantek Cr etchant, ~1.5 min.
•Plasma clean to remove resist: Same as step 2, 5 min.
Photo Lithography
• Project UV light through Mask
– Non contact with optical reduction (typical 4X)
– Contact with one-to-one pattern transfer
– Mask – very flat SiO2 plate with Cr thin film
– Resolution limited by wave length (phase shift)
– Optics hard for short wave lengths
Electron Beam Lithograpy
• Literature, Resources
– Handbook of Microlithography Micromachining and
Microfabrication, ed. P. Rai-Choudhury, SPEI press (chapter two
is on the web, linked from home page
– J C Nabity web site: http://www.jcnabity.com
Course material is posted on web site in restricted area:
http://www.nanophys.kth.se  education  Intro. to e-beam Lithography
 Link to restricted area (password protected)
Username: ebeamlecture
Password: lithogr
Some things you can do with EBL
Circuit of SQUIDs and Josephson Tunnel Junctions
1.5 mm
Bonding
Pads
Contact “cage” to nano-circuit -- for rapid testing
Connecting
Strips
Ferromagnetic Normal metal
tunnel junctions
100 nm
Co
Al
Circuit to measure spin injection from ferromagnet (Co) to normal metal (Al)
Innerdigitated Capacitor in
coplanar waveguide
Cooper Pair Transistor
All these structure were made with
one layer of e-beam lithography and one
vacuum deposition cycle!
Block Diagram on an EBL system
Electron Optics
detector
sample
Scanning the electron beam
Aperture
Lens
a – convergence angle
Beam diameter
d  d g  d s  dc  d d
2
a
2
2
2
Electron scattering limits resolution
Higher energy electrons have larger back-scattering range
Double Gaussian profile
Overview of systems
• SEM conversion (NPGS)
• SEM modification (Raith)
• High end system
– SEM conversion limited in speed by slow beam
deflection system (induction in magnet coils).
– Laser stage is big step in price, but necessary for
accurate pattern writing and stitching.
– The more complex the system, the more service and
higher user costs
– Industry Fab. machines not always well suited to
research needs.
NPGS
•
•
•
•
Joe Nabity, one man company, good reputation, very helpful, good support
Works with many SEMS
Can do stage control, many SEMs come with micrometer, motor control
(accuracy)
Can do precision alignment in single field by scanning in reduced area to find
mark. Manual mark detection.
Good Web site: http://www.jcnabity.com
list of references, pictures, ideas
Fabricated with NPGS
This picture shows part of a circular grating
with a period of 0.15 microns. The lines
appear almost straight, because they are
near the outer edge of the grating where the
radius is ~100 microns. The pattern was
written in PMMA and has been coated with
gold for viewing. The lithography was done
at the Optical Sciences Center at the
University of Arizona.
This image shows a pattern of radially placed dots in
PMMA after development. The white bar at the bottom of
the image is 1 micron long. The pattern was designed as
radial lines, but the spacing of the exposure points was
set 0.3 microns to produce discrete dots. Notice how the
dot size and spacing is very consistent in all
directions. The exposure was done with an SEM with no
beam blanker and the image was taken with the NPGS
digital imaging feature. The pattern was written by Dr.
ChiiDong Chen at the Institute of Physics, Academia
Sinica, Taiwan.
Raith 150
(KTH and Lund in Sweden)
•
•
•
•
•
•
•
Expanding company, niche for mid range system
Based on conversion of Zeiss FE Sem, high quality SEM, good detector
Also sell conversion system (Elphay Quantum)
Control system has bugs, poor software support
Software has good features: simple cad, position list, direct exposure control
Laser stage not perfect, but accurate overlay and stitching has been achieved
Can take a 6 inch (150mm) wafer
Proven resolution with our Raith 150
courtesy of Anders Holmberg
L= 80 nm
50 nm
L=Line width (pitch = 2L)
30 nm
25 nm
18 nm
16 nm
15 nm
20 nm
High End system, designed for Industry Fab.
Nanophys positive process for onecycle tunnel junction fabrication
• Two layer resist, selective developers
• Very large undercut – suspended bridge
• Tunnel junction (top and base layer) in one layer
Top view of pattern
Exposed areas
Undercut region
Next slides:
Cut on this axis
Supporting resist
Lithography and shadow evaporation
ZEP 520
PMGI SF7
SiOx
Si
Lithography and shadow evaporation
Irradiate with electron beam
Lithography and shadow evaporation
Develop the two layers selectively
Top layer:
Bottom Layer:
Lithography and shadow evaporation
Evaporate Al at an angle
Lithography and shadow evaporation
Oxidize the first layer
Lithography and shadow evaporation
Evaporate Al at opposite angle
Lithography and shadow evaporation
Lift off the resist and excess metal
Tunnel junctions
Voilà
Circuit of SQUIDs and Josephson Tunnel Junctions
3D structuring using contrast curve
•Accurately measure thickness of film
•Do test pattern with dose profile to accurately measure
contrast curve
Patterning in third dimension
Dose
thickness
Desired structure:
Holography
Positive electron resist SAL 110
Developer SAL 101
(Shipley)
Chalmers Group, S. Hård et al.
Applied Optics vol. 33 p 1176, 1994
Optical Kinoforms
Chalmers Group, S. Hård et al.
Applied Optics vol. 33 p 1176, 1994
Optical Comm. Vol. 88, p 37, 1992
Two basic types of pattern methods
• Direct Writing
–
–
–
–
Change pattern with each run
Slow, serial method of fabrication
Good for research and development
Low through-put, too costly for large scale production
• Lithography
–
–
–
–
pattern copying one process step
Fast, parallel method
High through-put makes low cost in large scale prod.
Not flexible enough for research and development.
Comparison of Lithographic methods
• Photo Lithography
– UV, deep UV
– Projection or contact
• Micro contact printing
– Stamp formed from Soft material
– Molecules (ink) is wet on to stamp, transferred
to surface
• Printing Press
Micro Contact printing
http://www.research.philips.com/technologies/light_dev_microsys/softlitho/
Stamp fabrication
•Master made by direct writing methods (EBL on Si + etch)
•Stamp gets dirty, wears out
•Essentially old-style printing methods scaled to nm dimensions
SAM’s and molecular electronics
Optical Stepper
http://www.sematech.org
http://www.nanonet.go.jp/
For example: Nikon optical steppers
High through-put direct writing tool
Sigma 700 series from Micronic Laser systems, Täby Sweden
http://www.micronic.se
Spatial Light
Modulator
(SLM) chip
10 6 electronically
addressable
mirrors
Alignment and overlay
• Alignment and overlay are more serious
problems than actually making the small
structure!
• Large area with fine detail requires “stitching”
write fields together – laser interferometer stage,
nm position and metrological accuracy!
• Overlay requires accurate alignment marks,
mark detection, registration and extremely
accurate pattern placement over large area
(scaling accuracy 1 part 106).
3-layer process done in Albanova
Industry has MUCH more sophisticated circuits with 1520 layers, 108 components, with very accurate overlay
Metrology and Imaging
• Laser interferometers on Stage
– 5nm “resolution”
– Reproducibility
• Thickness measurement
– Profilometer, demonstration
• Scanning Probe microscipe
– Vertrical resolution 1 Å level
– Latteral resolution depends on tip sharpness
SPM system overview
Scanning Tunneling Microscopy (STM)
Binnig and Rohrer 1981 (Nobel Prize in Physics 1986)
E(z)
 (z )
E0
U
z
d
Electric current proportional to quantum
mechanical probablility amplitude of ”tunneling”
through the energy barrier
I  P  exp( 
2
2m(U  E0 )d )

Wave´function decays eponentially in barrier region
Single Atom imaging possible
•Sharp tip
•Pristine surface
•Ultra High Vacuum
Check out this web page
http://www.almaden.ibm.com/vis/stm/gallery.html
The making of a Quanum Corral
Fe atoms on a Cu (111) surface
Atomic Force Microscopy (AFM)
Two Basic AFM Modes:
Contact mode (no vibrating tip)
Tapping mode (vibrating tip)
Many variations on Scanning Force Microscopy :
Liquid AFM
Magnetic Force Microscopy (MFM)
Latteral Force Microscopy (LFM)
Intermitant and non-contact AFM
Force Modulation Microscopy (FMM)
Electrostatic Force Microscopy (EFM)
Atomic Forces
Force
Hard core repulsion
Contact region
z
Seperation between tip and surface
Attractive force: van der Walls
Non-contact retion
Image molecular monolayers in liquid
•Molecules must be immobilized on surface
•Local force measurements possible
S-layer protein monolayer on Si surface
in liquid environment, 500 nm x 500 nm
Zentrum für Ultrastrukturforschung - Universität für Bodenkultur. Austria
Two basic scanning modes
1.
Feedback off: Scan over
surface with constant z0
(piezo voltage), control
signal changes with tipsurface separation.
2.
Feedback on: circuit
regulates z piezo voltage to
constant value of control
signal (constantly changes
tip-surface separation).
AFM
Contact mode
AFM tapping mode
Free space oscillation of cantilever
resonance 10-100 kHz
Cantilever hits surface
smaller amplitude of oscillation
Feedback loop
tapping mode
Free oscillation
Large amplitude
Hitting surface
lower amplitude
Digital
Insturments
Multi-Mode
head, scanner
and base
•
•
Turn on the controller (the
computer should be left on)
Remove the scanner from
under the microscope.