FCRP - FENA & NRI-WIN Needs of Theory and Simulation for Nanoarchitectonics (FENA and WIN) Kang L. Wang Kang L. Wang Raytheon Professor of Physical Sciences Raytheon Professor of Physical Sciences Center on Functional Engineered NanoArchitectonics -- FENA (www.fena.org ) Western Institute of Nanoelectronics -- WIN (www.win-nano.org ) University of California - Los Angeles Los Angeles, California 90095 –1594 (Ph: 310-825-1609 // Fax: 310-206-7154 // E-mail: [email protected])
23
Embed
Needs of Theory and Simulation for N …phys.lsu.edu/~jdowling/qmhp/talks/wang.pdfRaytheon Professor of Physical Sciences ... Logic Switch replacement ... •This project aims to develop
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
FCR
P - F
ENA
& N
RI-W
IN
Needs of Theory and Simulation forNanoarchitectonics
(FENA and WIN)
Kang L. WangKang L. WangRaytheon Professor of Physical SciencesRaytheon Professor of Physical Sciences
Center on Functional Engineered NanoArchitectonics -- FENA(www.fena.org)
Western Institute of Nanoelectronics -- WIN(www.win-nano.org)
University of California - Los AngelesLos Angeles, California 90095 –1594
Close Collaborations among the theory and simulation talents Seamless interface Working close with Experiments
The needs: Key problems
Charge and Alternatestate variables for lowdissipation andvariability Spin Molecule/ phase transition
Device/performance Molecule devices Spintronics
Heterogeneous integration
5
FCR
P –
FE
NA
& N
RI -
- WIN
5WIN
Modeling Across Time and LengthScales
Quantum Mechanics
Atomic Kinetics
Continuum Modeling
Need efficient, accurate and general first-principlesmethods for realistic simulations of synthesis,processing (assembly), manufacturing and operationof nanoelectronic devices and nanosystems.
Length
Time
Non-equilibriumDissipative
6
FCR
P –
FE
NA
& N
RI -
- WIN
6WIN
Predicting Material Properties & Patterns
Left: Calculated effective inter-atomic interactions in Fe-Ag/Ru.
Simulated diffuse x-ray scatteringpattern for Fe-Ag nanowires on Ru. Simulated surface structure of Fe-
Ag/Ru, showing stripe formation.
Predictable regular patterned templates for directed self-assembly of nanostructures Be able to predict self assembled structures from ab initio (Self consistent
NEGF)VidvudsOzolins(UCLA)
GlennFredrickson
(UCSB)
• Example of a numerical SCFT (self-consistentfield theory ) simulation of 8 unit cells of theIa3d “gyroid” phase of diblock copolymers.
• This project aims to develop a similar high-resolution SCFT for thin block copolymerfilms relevant to nanoscale lithography
Simulation of block copolymer assemblyfor nanoscale lithography
7
FCR
P –
FE
NA
& N
RI -
- WIN
7WIN
Rotaxane: Mechanically-Interlocked Moleculefor Next Generations of nanosystems
• Develop rotaxane-based memory device• Device characteristics: Ultra-high Density
and Stable Response
Prof. William A. Goddard, and Seung Soon JangGraduate student: Hyungjun KimCollaboration with Prof. J. Fraser StoddartRing Location (x)
Ener
gy (e
V)
OFFON
-e–
OFF
ON
0
ΔG‡
IonizationPotential
ΔG
+e–
Molecular conformation as a state variable
• 160 Kbit memory fabricated and tested @ 1011
bits/cm2 +- 2 operation.
• Crossbar: 400 Si nanowire bottom electrodesand 400 crossing Ti nanowire top electrodes(wires: 16 nm diameter / 16.5 nm half pitch)
1 2 2~ exp ( )
store b
tun
ma E
P!
" #= $ %$ %
& 'h
mLtsw~
8
FCR
P –
FE
NA
& N
RI -
- WIN
8WIN
First-Principles Interface Engineering
Nanoelectronics will includemetal & semiconductornanostructures, organic (andbio) molecules.
Importance of interfacesincreases with decreasingfeature size.
Need first-principles methods to:• Understand, predict and optimize the structure and
thermodynamic stability of interfaces in nanosystems• Predict charge and spin transport across nano-
interfaces
Pentanethiol SAM on Au(111)
9
FCR
P –
FE
NA
& N
RI -
- WIN
9WIN
Rotaxane
TTF (OFF) DNP (ON)Positively chargedmonopyridine
-6
-4
-2
0
2
4
6
8
10
0 10 20 30 40 50
Coordinate (Angstrom)
Me
an
Fo
rce
(kcal/m
ol A
)
0
10
20
30
40
50
60
70
0 10 20 30 40 50
Coordinate (Angstrom)
Po
ten
tia
l o
f M
ea
n F
orc
e
(kc
al/
mo
l)
Free Energy Barrier ~ 60 kcal/mol
Rotaxane: Mechanically-Interlocked MoleculeCurrent work
( ) ( )( )R
rxn
rxn rxn
dF RF R F dR
dR!
"# $"= ! % %& '"( )
*Mean Force
Potential of Mean Force Approach
How to control the free energy barrier between the ON state and the OFF state
10
FCR
P –
FE
NA
& N
RI -
- WIN
10WIN
Self assembled nanoarchitectonics and theirheterogeneous integration on Si
Approach: Directed self-assembly using PNA- I-junction dsPNA, T-junction dsPNA
Electrical characterization of I- and T- junction
SWNT
PNA
100 nm
100 nm
SWNT
SWNT
PNA
Mihri Ozkan(UCR)
CNT – Molecular RTDRequires accurate modelingof Structure Excited states Non-equilibrium
potential Electron / phonon
interaction Vibrational modes Phonon (thermal)
transport
Roger Lake(UCR)
11
FCR
P –
FE
NA
& N
RI -
- WIN
11WIN
Simulations of Self-Synthesized FunctionalDevices.
Bio-assembled CNTFETs – DNA and PNAassembly First simulations of the CNT-ssDNA-CNT system.
Experimentally measured I-Vof CNT-ssDNA-CNT
FIREBALL / NEGF calculations oftransmission and spectral functions at
transmission peaks a and b.
CNTFET drain current vs.gate voltage for different
lead doping
Non-equilibrium Green’s function Dissipation
12
FCR
P –
FE
NA
& N
RI -
- WIN
12WIN
Example of Interface Geometry Effect
CNTs connected by conjugated molecules
Left: CNT-(CH)20-CNT with the polyacetylene co-planar with the tangential plane of the CNT (topstructure) and perpendicular to the tangentialplane of the CNT (bottom structure) at the pointof contact.
Transmission for the co-planar geometry is, on average, 3-ordersof magnitude larger than transmission for the perpendiculargeometry.
Electron transfer is a strong function of the interface geometry.
Relaxed
E.G. X. Guo et al., “Covalently Bridging Gaps in Single-Walled Carbon Nanotubeswith Conducting Molecules,” Science 311, 356 (2006).
Right: Transmission versus energy plotfor both structures. The CNTs aremetallic (12,0).
13
FCR
P –
FE
NA
& N
RI -
- WIN
13WIN
Simulation and Computation of NovelEngineered Nanomaterials and Devices
Phonon engineering: enhance electrontransport in nanoscale transistor channels andimprovement of heat removal and thermalmanagement to guide device design Alex Balandin
(UCR)
• The results (Nano Letters,2006) overturn conventionalbelieve that the phononconfinement effects arealways detrimental to thecarrier mobility.
• Carrier mobility in Sinanowires can be greatlyenhanced by embedding thenanowires within hardmaterials such as diamond.
14
FCR
P –
FE
NA
& N
RI -
- WIN
14WIN
Ultra-scaled device modeling needs
3D quantum mechanical electrostatics /band structure and physical transportmodels for devices-including strain, high-k/metal gate, UTB, and surface orientationeffects
Physical models for transport in beyond-Sidevices (Ge, III-V) enabling performanceprediction/analysis
Efficient simulation of dissipative QMtransport, especially using a comprehensiveset of scattering mechanisms
H.H. Hosack, Frontiers in Comp. Nanoelect., 2/20/07, Indianapolis,https://www.nanohub.org/resources/2380/
15
FCR
P –
FE
NA
& N
RI -
- WIN
15WIN
Theory and Simulation – WIN(Spintronics) Spin materials
DMS Multiferroic materials and devices Room temperature materials
Electric field control devices – Rashba (spin orbit interaction) Active control of dynamics: e.g., Spin torque
Spin Hall effects Theoretical foundation
Spin and Magnetic Devices – Empirical approach in simulation Switching mechanisms: Spin transport Need to have fundamental approach: collective phenomena
Device models for circuits
Self consistent – NEGF Non-equilibrium quantum mechanics Theory of damping, dissipation Many body effects
16
FCR
P –
FE
NA
& N
RI -
- WIN
16WIN
Atomic-Level Materials Design
Theory can suggest: Which specific material or ordered structure has desired
properties How to grow such a structure experimentally
Example: Raising the Curie temperature of a magnetic semiconductor(Ga,Mn)As for spintronics applications
A. Franceschetti, S.V. Dudiy, S.V. Barabash, et al., Phys. Rev. Lett. 97, 047202 (2006).
Unoptimized Ga0.75Mn0.25As(random alloy)
Optimized Ga0.75Mn0.25As:(201) superlattice
Tc~240K — too low Tc~360K — sufficient to use
17
FCR
P –
FE
NA
& N
RI -
- WIN
17WIN
Atomic-Level Materials
For spintronics, atomic-level optimization couldtarget:
Raising ferromagnetic transition temperature Tc Ensuring magnetic semiconductors indeed half-metals
(avoiding structures that mix spin channels in magnetic state) Adjusting doping-dependent profiles Increasing barriers for unwanted defects Impurity
Example: half-metallicity in magnetic (Ga,Mn)As
• Calculation predicts that in perfectly randomGa1-xMnxAs,both spin-up and spin-down densities of states(DOS) become non-zero at εF by x=0.125.
• Can atomic-level optimization bring back half-metallic properties that are vital for spintronicsapplications?
Figure adopted from:E.Kulatov, H.Nakayama et al., Phys. Rev. B 66, 045203 (2002).
Ga0.875Mn0.125As
18
FCR
P –
FE
NA
& N
RI -
- WIN
18WIN
Theoretical Work and NumericalModeling:
Empircal Logic device functionality using spin
wave superposition Nano architectures with spin wave bus
Gated Spin Wave Devices & Bus –A. Khitun and K Wang
Spin wave propagationestablished
Spin wave resonancefrequency occurring atf ~ H1/2
0( 4 )H H M! " #= +
0 50 100 150 200 250 300
External magnetic field (Oe)
Amplitude changes (dB)
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Fre
quency
(G
Hz)
-4dB
-3dB
-2dB
-1dB
0dB
1dB
2dB
3dB
4dB
Alex Khitun(UCLA)
19
FCR
P –
FE
NA
& N
RI -
- WIN
19WIN
Spin Transistors -- Spin current amplifierJoachim Stöhr
Achievement: Direct observation of spin transfer switching by x-ray microscopy.
Joachim Stöhr – SLACwith Yves Acremann
d) 8.6 ns e) 9.0 ns f) 9.6 ns
g) 12.0 ns h) 12.2ns
i) 13.2 ns
a) 0 ns b) 0.15ns
c) 0.6ns
a
b
c
def
ih
gb
c
de fg hh
i
Y. Acremann et al., PRL 96, 217202/1-4 (2006)
20
FCR
P –
FE
NA
& N
RI -
- WIN
20WIN
Spin Device Modeling
Spin Hall Field EffectTransistor: does not requireelectron transport and hencecan potentially be an lowdissipating device
Quantum Spin Hall Helicaledge states
Support and GuideExperiments
Quantum Spin Hall Field Effect Transistor
Science, 314 1757 (2006)
B (T) -0.06 0 0.06
θ K (µ
rad)
0
4
-4T = 20 K 3 mV/µm
y = -48 µm
y = +48 µm
xy
kjs
B E
David Awschalom
Quantum phase transitionSpin Orbit interaction
SC Zhang(Stanford)
T = 295 K-0.05 0.0 0.05
B (T)
θ K (µ
rad)
-0.3
0.3
0.0
21
FCR
P –
FE
NA
& N
RI -
- WIN
21WIN
In A Nutshell
Nanomaterials Assembly and nanopatterning
Alternate state variables Spin variables: electron, nuclear spin, spin
waves Molecular state, Phase transition, Dipole,
Phase, Spin FET, Spin torque, Spin wave packets
propagation
Devices New principles Non-equilibrium
Hetergeneous Nanosystems Integrated Efforts
Theoretical approach to come to closeexperimental
22
FCR
P –
FE
NA
& N
RI -
- WIN
22WIN
Acknowledgments
All the FENA and WIN participants All students, postdoctoral fellows and
visitors as well as collaborators aroundthe world
Support: SRC, NSF, Marco, NERC, ARO, AFOSR,ONR, DARPA and many industrial companies