Download A selection of research topics in semiconductor physics at the Cavendish Laboratories

Document related concepts
no text concepts found
Transcript
A selection of research topics in
semiconductor physics at the
Cavendish Laboratories
G A C Jones
Cavendish Laboratory
University of Cambridge
Outline of talk
•
•
•
•
Split gate technology
Surface Acoustic Wave driven devices
Site control of quantum dots
Ultra high resolution lithography
Split gate technology for
2D, 1D and 0D electronic devices
(Wharam et al. J Phys C 21, L209 (1988))
Two-dimensional electron gas (2DEG)
Material Growth by MBE
2DEG mobility in modulation doped
structures
Pfeiffer et al APL 55,1888 (1989)
Quantum wire
-ve
drain
-ve
source
Quantized Conductance in 1D
•Making gate voltage more negative
decreases well width and increases
level spacing – depopulating energy
levels one by one.
-1V
Number of occupied subbands
EF
−∆V
0V
4
2D(ii)
2D(i)
1D
3
Vg more negative→
-1V
Wharam et al J
Phys C 21, L209
(1988)
Thomas et al APL
67, 109 (1995)
2
1
x
wire
Quantum dot
drain
source
Examples of devices based on
split gate technology
Quadruple quantum dot
Aharanov Bohm device with quantum dots
Making connections to isolated electrodes
Making connections to isolated electrodes
Surface Acoustic Wave Driven devices
Metallic surface gates
1 µm transducer, used to create a
mechanical surface acoustic wave
(SAW)
SAW devices can be tuned to
transport an integer number of
electrons in each minimum, which leads to
current quantisation.
Metallic surface gates are used to define 1D
channels. The spins of electrons in these
channels can be used as qubits.
Ohmic contacts
SAW speed is ~2800 m/s. GaAs
is piezoelectric, so a travelling
electric wave is created.
SAW driven single photon emitter
hν
SAW
n
i
p
hν
Ef
3GHz SAW transducer
3GHz SAW transducer
Quantum dots with non-invasive detectors
Dual channel SAW device for investigating
electron interactions and entanglement
Magnetic gates 3nm Au + 20nm NiFe + 3nm Au
Spin Manipulation Using Nanomagnets
Nanomagnetic fingers designed to redirect an external field to a perpendicular direction
have been fabricated.
The magnetic field in the gap has been imaged with electron holography. An external field
of 1 T leads to ~190 mT in the gap.
It should therefore be possible to rotate injected electrons to .
Similarly designed nanomagnets can also be used to create an alternating magnetic field for
use in ESR-type experiments.
InAs Quantum Dots for
optical applications
Motivation
• Discrete density of states makes quantum dots ideal
as single photon sources for quantum cryptography
or optical quantum computation schemes
Electrically Driven Single-Photon Source
Zhiliang Yuan, et al. Science (2002) 295 102-105
P. Atkinson, Cambridge University, 2005
Stranski-Krastanov growth of
InAs dots by MBE
results in near-random dot distribution
1.2 ×10 9 cm-2
2.55ML at 0.01ML/s
515°C
P(A)=2.1×10-7 mbar
V/III flux ~ 450
P. Atkinson, Cambridge University, 2005
Site-control of dot nucleation
• Electron-beam lithography to pattern
small holes (diameter ~60-140 nm)
• RIE etching (SiCl4)/ wet etching (weak
sulphuric acid based) to etch pits ~1040nm deep
• Resist removal (solvent rinses, oxygen
plasma ashing)
• In-situ low-temperature oxide removal by
hydrogen plasma (preventing surface
damage)
pre-growth after H plasma treatment
P. Atkinson, Cambridge University, 2005
Site-control of dot nucleation
[110]
pre-growth
P. Atkinson, Cambridge University, 2005
after 10nm GaAs buffer
after 2ML InAs dep
Site-control of dot nucleation
Good single
dot
occupancy
Holes
completely
infilled
Some
double dot
occupancy
Dots on the
verge of
coalescing
P. Atkinson, Cambridge University, 2005
No dots nucleating between
patterns
(less than the critical InAs
thickness deposited)
Photoluminescence
Observation of photoluminescence from nucleated dots
Data courtesy of Martin Ward, Toshiba Cambridge Research
P. Atkinson, Cambridge University, 2005
Nanoimprint masters for sub-10 nm patterning
M. S. M. Saifullah, Z. Zheng, W. T. S. Huck, D. Anderson, G. A. C. Jones & M. E. Welland, Unpublished data, (2005).
Nanoimprint masters for sub-10 nm patterning
M. S. M. Saifullah, Z. Zheng, W. T. S. Huck, D. Anderson, G. A. C. Jones & M. E. Welland, Unpublished data, (2005).
Acknowledgements
•
•
•
•
•
•
•
•
Dave Anderson
Paola Atkinson
Stephen Bremner
John Griffiths
Massaya Kataoka
Dave Ritchie
Saif Saifullah
Martin Ward
Collaborative projects with KAIST
Towards 1nm lithography
Prof H S Kim (Sun Moon University) is developing optics for a
low-voltage, high-resolution, e-beam micro lithography column.
Dr D. G Hasko (Cavendish), has been collaborating by providing
expertise for the resist processing which requires a deeper
understanding to achieve nanometer scale resolution. One
student has spent several weeks in Cambridge working on this
project learning resist processing etc.
• Dr S B Lee (Hanyang University) is responsible for system
integration of this tool and the Cavendish Lab has supplied
hardware for the mechanics of some of this system.
•
Dr G A C Jones who has a high-voltage electron beam
lithography tool at the Cavendish Laboratory is collaborating
with Dr S. B. Lee in an experimental program to create
patterned templates to act as a seeding base for clusters of
particles to be assembled out of solution. Arrays of etched pits
of various shapes and sizes, of order a few 10s of nanometers
have been fabricated on various substrates using the e-beam
facility in Cambridge for Dr Lee to investigate cluster formation
on these samples. (SBL)
• Mr Hongkee Yoon (KAIST student) spent ~8 weeks in the
Cavendish working on a new project - 3 Dimensional, grey
scale, e-beam lithography. (GACJ)
• Dr Jones has been working on other methods of achieving
nanostructure fabrication in the “towards 1nm scale”. (GACJ)
•
3 dimensional, grey scale,
electron-beam Lithography
Hongkee Yoon (Kaist)
Geb Jones (Cavendish)
David Anderson (Cavendish)
Motivation
• Applications in optical elements:
• blazed gratings
• sinusoidal, triangular gratings
• zone plates
• integrated surface lenses for optical interconnects
• …….?
• Applications in circuit elements
• air bridges
• MEMs
• …….?
• Suitable for mass replication using nanoimprint lithography
Dose vs resist thickness remaining
2000
0
1
49 97 145 193 241 289 337 385 433 481 529 577 625 673 721 769 817 865 913
- 2000
- 4000
- 6000
- 8000
- 10000
- 12000
• Blue line is measured height (nm)
• Red line is calibrated height with max height set as 0
Monte Carlo simulation of 60keV e-beam exposure
1000nm A9 PMMA on Si (Z = 500nm)
1.00E+04
1.00E+03
1.00E+02
1.00E+01
1.00E-02
1.00E-03
1.00E-04
1.00E-05
1.00E-06
1.00E-07
Radius Ri/um
90
80
100
1.00E-01
70
60
50
40
30
20
10
1.00E+00
0
Energy Density per Injected Electron
E*/(eV/um^3)
1.00E+05
1000nm A9 PMMA on Si (Z = 500nm)
1.00E+04
1.00E+03
1.00E+02
1.00E+01
1.00E-01
1.00E-02
Radius Ri/um
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
1.00E+00
0
Energy Density per Injected Electron
E*/(eV/um^3)
1.00E+05
Proximity effects in e-beam exposure
For double Gaussian approximation to PSF, the energy deposited per electron
(normalised to unity in the centre of a large area) is, E ∼ (gα + ηgβ)/(1+η)
1
E= ____
(1+η)
Normalised
Dose
1
1
__
2
(1+η)
E = ____ =1
(1+η)
( _14 + _14 η)
_1
E = _____ = 4
(1+η)
1 _
1
( _
+
η)
2 2
______
= _12
E=
(1+η)
Dose pattern for 3Ghz SAW transducer
3D test pattern with 10 levels of height
Calculated dose pattern for 10 x 10 array of
5um square pillars having 10 levels of height.
Corner of 10 x 10 pillar array showing
corrected doses calculated on 50nm grid
Exposed array without proximity correction
AFM image (uncorrected array)
AFM scan (uncorrected exposure)
Proximity corrected 10 by 10 array
AFM image (corrected exposure)
AFM scan (corrected exposure)
Sine wave grating slit pattern
Diffraction images from sine wave grating
slit diffraction images
“Towards 1nm lithography”
Sub-10nm, high aspect ratio patterning
of ZnO nanostructures using zinc napthenate
negative e-beam resist
G.A.C. Jones 1, M.S. Saiffullah 1,
K. R. V. Subramanian 1, D. Anderson 1,
D-J. Kang 2, W.T.S. Huck 1 and
M.E. Welland 1.
1
University of Cambridge, U.K.
2 Sungkyunkwan University, Korea.
Why direct writing of oxides?
¾Fundamental studies on nanoscale oxide systems
Basic transport properties, optoelectronics, nanophotonics,
nanoferromagnetism, nanoferroelectricity, etc.
¾To enable patterning of thick or complicated oxide layers
Conventional lift-off technique for patterning thick and complicated oxide
layers does not work.
Use a spin-coatable oxide resist for direct writing to circumvent the
problem associated with lift-off technique. It also improves tolerances and
reduces processing steps.
Resist Preparation
Alkoxide Route (TiO2, ZrO2, Al2O3)
Naphthenate Route (ZnO)
Preparation of spin-coatable oxide resists
of TiO2, ZrO2 and Al2O3
Alkoxide:
Titanium n-butoxide, Zirconium n-butoxide,
Aluminium tri-sec-butoxide
Stabilizers:
β-diketones and β-ketoesters
Solvents:
Methanol, Ethanol, Iso-propyl alcohol
Glove Box <5% relative humidity
alkoxide + stabilizer + solvent
Mixing for 2 hours
Ratio of alkoxide to stabilizer typically 1:1
Sub-10 nm lines of TiO2 using a
Leica VB6-UHR Nanowriter
M. S. M. Saifullah, K. R. V. Subramanian, E. Tapley, D-J. Kang, M. E. Welland & M. Butler.
Nano Letters, 3, 1587 (2003).
Sub-10 nm lines of ZrO2 using a
Leica VB6-UHR Nanowriter
K. R. V. Subramanian, M. S. M. Saifullah, E. Tapley, D-J. Kang, M. E. Welland & M. Butler.
Nanotechnology, 15, 158 (2004).
Resist Preparation
Alkoxide Route (TiO2, ZrO2, Al2O3)
Naphthenate Route (ZnO)
What are metal naphthenates?
Metal naphthenates consist of cyclopentanes or
cyclohexanes, methylene chains [-(CH2)-], carboxylates and
metals. They can be represented as:
[(cyclopentane) – (CH2)n – COO]–m – Mm+
where M is a metal atom.
They are sticky liquids at room temperature and are stable in
air. Hence they do not require any special treatment like
alkoxides.
E-beam damage of zinc naphthenate resist
Previous infrared studies suggest that the exposure of
naphthenate molecules to an e-beam results in building
bridges between them at the –(C=O)– and/or –(CH2)n–
groups thereby increasing the molecular weight of the resist.
This makes electron-beam exposed naphthenate resists
insoluble in toluene.
[Kakimi et al, Jpn. J. Appl. Phys., 33, 5301 (1994)]
Zinc naphthenate
_
H
H
H
H
C
H
C
C
H
O
C
C
O
C
C
H
_
H H
n
H H
X-linking occurs here
2+
Zn
ditto
Process details
• Zinc naphthenate (67 wt.-% in mineralised spirits)
diluted 1:20 in toluene.
• Filter and spin coat @ 4000rpm for 30s
• (~90nm thick).
• Expose at 30mC.cm-2 using 100kV 1.5nA beam.
• Develop in toluene for 10s
• Blow dry (no rinse)
• Characterise using DI NanoscopeTM AFM and
LEO1530VP SEM.
E-beam sensitivity of zinc naphthenate resist
1
Normalized Thickness
0.8
0.6
0.4
Sensitivity ~ 15 mC cm
Contrast, γ = 3.3
0.2
-2
0
1
10
Electron Dose mC cm
100
1000
-2
Sensitivity at half the normalized thickness is 15 mC cm-2
and the contrast (γ) is 3.3.
M. S. M. Saifullah, K. R. V. Subramanian, D-J. Kang, D. Anderson, W. T. S. Huck, G. A. C. Jones & M. E. Welland, Advanced
Materials, 17, p. 1757 (2005).
Comparative study of sensitivities
Resist
Sensitivity
Type
Voltage
TiO2
35 mC cm-2
Negative
100 kV
ZrO2
40 mC cm-2
Negative
100 kV
Al2O3
20 mC cm-2
Negative
100 kV
ZnO
15 mC cm-2
Negative
100 kV
Calixarene
20 mC cm-2
Negative
70 kV
SAL-601
0.07 mC cm-2
Negative
70 kV
HSQ
0.7 mC cm-2
Negative
100 kV
PMMA
0.6 mC cm-2
Positive
100 kV
Sub-10 nm lines of ZnO using a
Leica VB6-UHR Nanowriter
Sub-10 nm lines of ZnO in the 500 µm main field
High aspect ratio structures
Aspect Ratio (for narrowest features) > 11
Post Development Bake studies on
zinc naphenate resist
TGA & DTA studies of zinc naphthenate resist
0
Time, minutes
40
60
20
80
1100
TGA Trace
Step 3: -90.37%
-9.3 mg
8
Heating rate = 10 C min
6
Step 1: -32.4%
-3.34 mg
4
Heat Treatment Temperature = 500 C
Z
1000
Starting mass = 10.29 mg
-1
900
Step 2: -41.8%
-4.3 mg
Z
Intensity (A. U.)
Weight, mg
10
2
0
Z
S
800
50
Rate of Mass Loss, mg/min
X-ray diffraction data
100
12
150
250
350
450
550
650
Temperature, C
750
850
950
0.2
X
700
0
Z
First Derivative Of TGA Trace
-0.2
Z
Z
S
600
X
-0.4
-0.6
500
25
-0.8
30
35
40
45
50
55
60
Angle, 2 θ
-1
50
150
250
350
450
550
650
750
850
950
Temperature, C
Temperature Difference, C
6
4
Heat Treatment at 500°C gives
polycrystalline films of ZnO.
Z = ZnO, S = Substrate and
X = Contamination peaks
SDTA Trace
2
0
-2
-4
-6
-8
0
20
40
60
Time, minutes
80
100
65
Optimization of Post Development Bake for
photoluminescence in ZnO at 500°
500 C
540
Air
Ar
Ar + 1% O
Ar + 5% H
421
Peaks at 368nm (UV band-edge
emission), 421nm (blue
emission), 459nm (weak blue
emission) and 501nm (green)
were observed.
2
2
Normalized Intensity
501
368
459
300
350
400
450
500
550
Wavelength, nm
600
650
700
The presence of slightly broad
peaks (368nm and 421nm) and
visible range peak (501nm)
suggest that the material still has
some oxygen vacancies and
other defects that cannot be
completely removed by this
500°C heat-treatment process.
Sub-10nm lines of ZnO using a
Leica VB6-UHR Nanowriter
Conclusions
9Sub-10 nm, high aspect ratio structures (>10) have been directly
fabricated in ZnO by high resolution e- beam lithography using spincoatable, zinc napthenate as a negative resist.
9Exposure to an electron beam makes napthenate based resists insoluble
in organic solvents such as toluene.
9The electron beam sensitivity of these materials is comparable to that of
conventional electron beam resists such as calixarene and other metallic
oxide resists based on metalic alkoxides.
9Post development baking at 500C for 1 hour in an argon/hydrogen
ambient drives off the organic component of the resist, shrinking the
feature size and drives the formation of the Zn-O bond enhancing the
photoluminescence spectrum.
9The line edge roughness of the these patterns ~2.8nm (3σ) pre bake and
~2nm (3σ) post bake is the smallest measured value so far of any e-beam
resist.