Transcript ONR-DNA-nanoelectronics

Goal: To use DNA self-assembly to overcome the challenges
of optical and e-beam lithography in creating nanoscale circuits.
PI Paul Rothemund, computer scientist, Senior Research Associate (research faculty)
Expert in creating complex shapes and patterns using DNA self-assembly.
Interested in scaling up DNA self-assembly, bridging nano, micro and macro scales.
PI Erik Winfree, computer scientist, Associate Professor
Expert in creating complex shapes and patterns using DNA self-assembly.
Interested in creating large, complex patterns using algorithmic self-assembly.
PI Marc Bockrath, applied physicist, Assistant Professor
Expert in nanoscale device fabrication, physics and properties of single molecules.
Interested in carbon nanotube (CNT) circuit fabrication and characterization.
Hareem Maune, graduate student synthesizing and testing CNT devices
PI William Goddard, theoretical chemist, Full Professor
Expert in atomistic simulation of chemical systems.
Interested in simulation of DNA-CNT device and circuit systems.
Andres Jaramillo-Botero, Director Caltech Center for Multi-scale Modelling
Siping Han, graduate student, synthesizing and modelling CNT devices
metallization of DNA nanostructures.
Challenge 1: "Functionalization" electronically or optically active materials
must be coupled to DNA nanostructures in high yieldat specificed locations.
We focus on carbon nanotubes (CNTs) but also work on metallization.
DNA noncovalently wraps CNTs
and allows them to disperse in
buffer solution.
An origami is made with 'red' and
'blue' DNA hooks, having different
DNA sequences on top and bottom.
Two batches of CNTs are made
with complementaryred or blue
strands.
Red and blue CNT assemble
into crossbar FETon the
origami.
+
+
Pd
A device measurement is made
Si/ SiO2
To make nanostructures more rigid and to avoid aggregation
origami-ribbon hybrids are used.
red and blue hooks
red tube
blue tubes
MOSFET geometry
50 nm
crossbar
Gate
Channel
Over 30% of tubes are within 10 degrees of the desired orientation
Orientation of SWNT
40
22%
35
2%
Red side (-1)
Unknown (0)
Blue side (1)
Frequency
30
25
76%
20
15
10
5
0
0-20
20-40
40-60
60-80
80-100 100-120 120-140 140-160 160-180
Angle
Characterization of DNA self-assembled CNT FET
b
a
ISD
ISD [nA]
VSD
300
ISD [nA]
200
400
100
-- - -- Vg=
-- -- Vg=
0.5V
-0.5V
------ Vg=
0.5V
0
-0.5
200
0.0
0.5
VSD [V]
100
Vg
ISD
VSD = 0.85V
0
-0.5
0.0
0.5
Vg [V]
1.0
Rothemund's Aims:
To continue work with IBM, to replicate positioning and
orienting work at Caltech, so that CNT devices can be more
easily characterized and integrated.
origami shapes
patterned DLC on Si
placed shapes
CNT organization
To combine multiple origami to create large origami breadboards.
B
A
jigsaw puzzle
stacking bonds
enable 800 pixels
D
C
A
B
C
D
To create DNA structures with features that bridge the nano and
microscales so that a complete device can be fabricated...
Bridging the nano and micro scales
cross-shaped origami
further
tile
additions
tiles
50 x 2
microns
100 nm
metallize strands with nanogold
mineralize strands bearing peptides
P
N
Patterning of nanotubes (wires) so that they
diverge and can be wired up at the microscale
+
=
CNT FET
+
=
(or colored tracks may be metallized)
Winfree's Aims:
To use combined existing DNA self-assembly techniques
(DNA origami, ribbons, algorithmic self-assembly, and periodic
DNA crystals) to create squares of programmable size.
These squares will have a pattern that is appropriate for a
memory with demultiplexers, an architecture perfectly suited for
useful circuits.
To explore the addition of actual nanoelectronic components to
the memory pattern, for example a crossbar lattice of
carbon nanotubes.
an origami seed n
encoding the number n is added to a soup of tiles
The 'computed'
output is a
square of size
n tiles x n tiles
with the origami
embedded
in one corner.
Self-assembly can compute:
a simple example is counting.
The pattern left behind is
a template for a demultiplexer
0
0
0
1
0
0
0
0
0
Counting tile set.
1
n
1
0
n
n
1
1
0
n
AND gate
0
1
n
0
0
c
c
0
1
0
L
c
c
R
AND gate,
lower input negated
S
NOT gate
0 1 1 0
Counting to a fixed length from an origami
enables programmed growth of NxN squares
A counter grown from origami
Full N x N squares remain an
important challenge.
Error rates must be reduced.
A termination scheme for
counters must be demonstrated.
0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
1
0
1 0 0 1
Input lines encode
binary values for
6 (vertical) and
9 (horizontal) which
are demultiplexed
to access the red
memory element.
The light gray
pattern underneath
which determines
the circuit would be
created by the
self-assembly of this
21 x 21 square.
Bockrath's Aims:
To use short length-sorted carbon nanotubes to increase the yield
of existing devices. (Many problems arise from very long tubes
acting as bridges between multiple origami).
To self-assemble and characterize circuits of more than one
carbon-nanotube based device to create elementary logic gates
and memory elements.
To self-assemble novel devices to explore transport physics in
nanostructures.
Rationally Engineered Logic gates and Memory
Elements Utilizing Multiple Nanotubes
Nanotube assembly
Schematic circuit diagram
Vs
Inverter
Vout
Vin
Vs
SRAM
Vout
Vs
Vs
NOR
Vout
Vin1
Vin2
Novel Devices Probing Transport Physics in Nanostructures: Phase
Coherence in Strongly-Interacting Electron Systems
Many possibilities exist for making novel devices.
DNA origami template for
parallel nanotubes
Tunable separation with desired values ~5-20
nm
B
Interferometer device
source
drain
I
V
Nanotubes act as a “which path?” interferometer enabling the study of phase
coherent transport in Nanotube-based Luttinger liquids via a transport experiment.
The setup is analogous to a double slit experiment in optics. The magnetic field B
tunes the phase by the Ahoronov-Bohm effect. Tubes must be closer together than
the phase coherence length in the electrodes, which is readily obtainable using
DNA based self-assembly.
Goddard's Aims:
Electron transport in DNA-carbon nanotube hybrids:
The effect of an insulating DNA layer between carbon nanotubes,
silicon nanowires, and quantum dots is unknown. In some cases
DNA may be removed from devices post-assembly, in other cases
may remain. Thus it is important to simulate electron transport in
carbon nanotube devices, with and without intervening DNA, starting
with atomistic simulation (Next slide).
Simulations of the placement process:
The interactions which bind DNA structures to technological
surfaces like silicon or diamond-like carbon are poorly understood.
The best choice of experimental conditions, as well as the best
choice of DNA shapes to bind can be explored by atomisticand
mesoscale simulation.Free energies of correct and mismatched
binding, and possible kinetic traps can be explored.
patterned DLC on Si
possible kinetic traps
and mismatches
DNA-origami CNT-based Transistor
V
Junctions
V
LUMO
DNA
1
2
EF
HOMO
I
CNT
CNT
Organic molecule
• Size of molecules << scattering lengths (e.g. mean free path, de Broglie
wavelength, etc.) -> quantum descriptions necessary.
Theory and Modeling to Describe…
•
•
•
•
•
Quantum chemistry of molecule(s) + nanotube -> charge flow & bonding -> geometry &
energy spectrum of the entire system.
Organo-metallic interface mechanics and transport. Need to treat molecule as finite and
nanotubes as semi-infinite electrodes.
Escape currents (through organic insulator layer).
Conformation effects on electronic transport.
Effect of finite bias.
• IV characteristic of self-assembled CNT-based transistor junctions.
Molecular Mechanics
Dynamics
d2R
geometry
Fm 2
dt
Multiscale Methodology:
1st-principles I-V validated by
rotaxane modeling
Density-functional theory (Hohenberg-Kohn-Sham)
1
Gm  (Em Sm  H m  1   2 )


T (E,V )  Tr 1G 2G
nA

@ gate voltage 2.55V
1
- 1
- 3
-4 0

0
4 0
-3
x 1 0
contact widening

2
0
- 2
self-energy
Lower current, asymmetric I-V
3
transmission
Ballistic transport theory (Landauer, Buttiker)
2e 
I
T (E,V ) f1 (E  2 )  f2 (E  1 )dE current


h
T(E,V)
2
dI 2e
G

T
conductance
dV
h
2
gate voltage (V)


1,2  i 1,2  1,2
2
1
-9
Green’s ftn. Formalism (Fisher-Lee)
m
1
x1 0
1  1  eV , n  H
electro-chemical potential
1
0
-1
e.g. Rotaxane switch
-2
-3
-6 0
dI/dV
-4 0
-2 0
0
20
40
60
source to drain voltage (mV)
dI/dV vs. junction bias and gate bias
Further validation: bi-phenyl-dithiol modeling
T(E)
contact
I (V ) 
2e  2
T ( E ,V )( f1  f 2 )dE
h 1
I(V)
molecule
contact
Au (111)
Relevance to the Office of Naval Research
Fundamental advances in microelectronics underlie all of
our country's defense systems, from networked warfare to
avionic systems. Eventually, self-assembly based methods may
be the only path forward to more powerful nanoelectronic systems.
DNA self-assembly uses non-hazardous "green" chemistries,
decreasing the Navy's environmental footprint.
DNA self-assembly techniques may yield lower cost
electronics "grown" from cheap components without
capital investment in conventional chip fabs.
Budget for 4 years, $2.6 million including:
PI: Paul Rothemund:
$200K/yr for Senior Research Associate salary
and materials
Co-PI: Mark Bockrath $100K/yr for 1 graduate student and
materials
Co-PI: Bill Goddard $100K/yr for 1 graduate student and
materials
Co-PI: Erik Winfree $100K/yr for 1 graduate student and
materials
Equipment $150K/yr including plasma etcher/cleaner ($20K),
wafer-scale Atomic Force Microscope ($200K)
temperature-controlled dynamic light scattering ($50K).