Nnano 2008 314 PDF
Nnano 2008 314 PDF
School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332-0245, USA, 2 Department of Chemical and Materials
Engineering, School of Engineering, University of Dayton, Ohio 45469, USA; * e-mail: [email protected]
34
NATURE NANOTECHNOLOGY
DOI: 10.1038/NNANO.2008.314
ZnO
Kapto
n film
Packa
ging fi
lm
Forward connection
40
SWG
20
Reversed connection
+
0
1.0
0.5
0.0
0.5
1.0
LETTERS
satisfying the switching-polarity and linear superposition criteria is a necessary and sufcient condition to conrm that the
output is indeed generated by the SWGs (see Supplementary
Information). Finally, increasing the straining rate signicantly
increases the electrical output (Fig. 3d; see Supplementary
Information, Figs S7,S8).
The ZnO PFWs used in our experiments were oriented along the
c-axis. It is known that the (0001) and (0001) facets of ZnO are terminated with zinc and oxygen, respectively, and a non-symmetric
Schottky contact usually appears at one end in transport measurement (see Supplementary Information). Figure 4 shows a proposed
mechanism for the generation of current in single-wire generators in
terms of the band structure of the system, with a Schottky
barrier introduced to represent a high local contact resistance
(50 1,000 MV) to block the ow of electrons through the wire
(see Supplementary Information, Tables S4,S5).
When the PFW is subjected to tensile strain, a piezoelectric eld
is created in the PFW due to polarization of atoms in the crystal to
create ionic charges (Fig. 4b; see Supplementary Information).
These piezoelectric charges, which are non-mobile, remain in the
PFW for an extended period of time without depletion by the
free carriers as long as the strain is preserved13,14. When the piezoelectric potential is positive (V ) at the Schottky barrier side (which
is on the left), and negative (V 2) at the ohmic side (which is on the
right in Fig. 4a), the conduction band and Fermi level of the electrode on the right is raised by DEp e(V 2 V2) with respect to
the electrode on the left, and electrons will ow from the right electrode to the left electrode through an external load resistor because
the resistance of the Schottky barrier is very high for voltages below
a threshold value and thus blocks the ow of electrons through
the wire (Fig. 4b). The electrons accumulate at the interfacial
region between the left electrode and the wire, and this process
continues until the potential created by the accumulated electrons
balances the piezoelectric potential and the Fermi levels of the electrodes reach a new equilibrium value (Fig. 4c). When the tensile
strain in the PFW is released, the immediate disappearance of the
piezoelectric potential lowers the Fermi surface of the right electrode
by DEp, and the electrons that had accumulated near the left electrode ow back through the external circuit to the right electrode
(Fig. 4d), returning the system to its original state. The peaks in
the electrical output associated with the stretching and release are
shown in Fig. 4e.
If the piezoelectric potential is negative (V 2) at the Schottky
barrier side and positive (V ) at the ohmic side (as a result of
switching the c-axis orientation of the PFW), the Fermi level of
the left electrode will be raised by DEp when the PFW is subjected
to tensile strain, and electrons ow from left to right through the
external load resistor (Fig. 4f ). The free electrons can enter the
PFW and screen the piezoelectric charges, reducing the local effec
tive potentials to V 2
1 and V1 , but the free electrons cannot completely neutralize/deplete the piezoelectric charges because the latter
cannot move (Fig. 4g)15,16. This process continues until an electric
potential due to the free moving electrons is created across the
PFW to balance the piezoelectric potential and the Fermi levels at
the two electrodes reach a new equilibrium value (Fig. 4g). When
the strain is released, the piezoelectric potential disappears, and
the free charges used to screen the piezoelectric charges are free to
move. At this moment, the Fermi level of the right electrode is
higher than that of the left electrode, resulting in the electrons
owing back from right to left through the external circuit. This
process ends when the Fermi levels of the two sides reach
equilibrium again.
The Schottky barrier acts as a gate that prevents the ow of the
electrons through the PFW; in addition to preserving the piezoelectric potential, this also leads to the accumulation of electrons, which
leads to a higher discharge rate. The Schottky barrier can either be at
35
LETTERS
NATURE NANOTECHNOLOGY
DOI: 10.1038/NNANO.2008.314
FS
0
10
FR FR
20
400
600
800
60
80
100
120
25.3 25.5
Time (s)
140
20
40
Time (s)
80
100
120
140
100
120
140
300
10
200
Current (pA)
Voltage (mV)
Reversed connection to
measurement system
60
Time (s)
15
FR FR
5
0
5
600
FR FR
50
40
400
200
40
20
FS
0
200
400
30
FS
200
Current (pA)
FS
Voltage (mV)
Voltage (mV)
Forward connection to
measurement system
10
0
10
20
30
40
50
25.2 25.4
Time (s)
Current (pA)
600
20
FS
FS
FR FR
100
0
100 FS
FS
10
200
15
0
20
40
60
80
100
120
140
20
40
Time (s)
60
20
1
0
1
2
3
10
0
Current (pA)
Current (pA)
Voltage (mV)
SS
0
10
20
20
Time (min)
40
2
4
6
8
4
60
Time (min)
30
90
120
FR
10
20 30 40
Time (s)
50
20 30
Time (s)
50
20
Current (pA)
2
1
0
1
15
10
10
Current (pA)
Voltage (mV)
80
Time (s)
0
5
10
2
3
0
6
8
Time (s)
10
12
14
4
2
0
15
20
4
0
6
8
Time (s)
10
12
FS
SR
0
10
40
Figure 2 | Electrical output of a single-wire generator (SWG). a,b, Open-circuit voltage (left) and short-circuit current (right) of an SWG when subject to
repeated cycles of fast stretching (FS) and fast release (FR) when forward-connected to the measurement system (a) and reverse-connected to the
measurement system (b). The insets in a show the detailed shapes of peaks in the voltage and current outputs. c, The electrical output of a SWG operated
at 22 cycles per minute demonstrates the stability of this approach. The bottom two panels show the detailed shapes of the voltage (left) and current (right)
outputs. d, Electric current generated when the PFW was subject to slow stretching (SS) and fast release (top), and fast stretching and slow release
(SR; lower curve). The data in a and b were acquired from the same SWG; the data in c and d are from different SWGs.
just one side of the PFW or at both sides. An insulator layer can also
perform the same role (see Supplementary Information, Fig. S9).
The PFW acts like a capacitor and charge pump, storing electrons
and then driving them back and forth through the external circuit as
the PFW is stretched and released. This means that the generated
voltage does not need to exceed a critical value (such as 0.4 V in
36
NATURE NANOTECHNOLOGY
Current (nA)
70
SWG1
200
0.5
Current (nA)
1.0
140
Current (nA)
LETTERS
DOI: 10.1038/NNANO.2008.314
SWG2
100
1.0
0.5
0.0
0.5
1.0
1.0
0.5
Voltage (V)
0.0
0.5
1.0
SWG1
FS
1.5
SWG2
1.0
FS
Voltage (mV)
0.0
0.5
FR
FR
FR
FR
0.5
0.0
0.5
FS
1.0
1.5
SWG2
SWG1
50
100
150
Time (s)
200
3.2 r.p.m.
250
Voltage (mV)
0.0
SWG1
0
50
FS SWG2
100
150
Time (s)
SWG1 SWG2
200
250
1.0
0.4
0.1
FR
1.5
SWG1 + SWG2
10 r.p.m.
60 r.p.m.
Voltage (mV)
0.2
1.0
FS
0.5
0.5
FR
FS
2.0
0.0
SWG1 SWG2
SWG2
1.5
1.0
0.5
Voltage (V)
SWG1 + SWG2
1.0
1.0
Voltage (V)
SWG1
Voltage (mV)
SWG1 + SWG2
0.0
Voltage (mV)
100
0.2
0.0
0.5
0.0
0.2
0.1
0.5
0
50
100
10
Time (s)
15
20
25
30
Time (s)
Time (s)
Figure 3 | Connecting two single-wire generators (SWGs) in series. a, IV characteristics of two individual SWGs (left, middle), and the same two SWGs in
series (right). b,c, Voltage output of two individual SWGs subject to fast stretching and release, and (for times after 180 s) the same two SWGs connected
constructively in series (b), showing that the output voltages are added, and then connected destructively in series (c), showing that the combined output is
the difference of the two individual outputs. d, As the number of stretchrelease cycles per minute is increased from 3.2 (left) to 10 (middle) to 60 (right),
the output voltage also increases, and is close to sinusoidal for 60cycles per minute. The data in a, b and c were acquired from the same two SWGs; the
data in d are from a different SWG.
Methods
Fabrication of the generator. The generator was fabricated by bonding a ZnO PFW
laterally on a Kapton polyimide lm (Fig. 1a). The ZnO PFWs were synthesized by a
physical vapour deposition method12, and typically have diameters of 3 5 mm and
lengths of 200 300 mm. We chose long PFWs because they were easy to manipulate,
but the same process applies to nanowires. The thickness of the Kapton lm received
from DuPont Company was 50 mm. The lm was rst washed with acetone,
isopropyl alcohol and ethanol under sonication. After that, the Kapton lm was
prepared for use as the substrate by dry cleaning using nitrogen gas and baking at
150 8C for 10 min. The ZnO PFW was placed at on the Kapton lm using a
probe station under an optical microscope. Silver paste was applied at both
ends of the ZnO PFW to x its two ends tightly on a exible substrate
(see Supplementary Information, Fig. S1). A current/voltage measurement
meter was connected to two ends of the PFW without introducing any external
power source in the circuit.
The role of the Schottky contact. To examine the role played by the Schottky
contact for electricity generation, we purposely created symmetric ohmic contacts at
two ends of a ZnO PFW using metallic indium. The output signals were very weak
and did not correspondence to the frequency of the fast stretch and release of the
substrate. More importantly, no reversal in output signal was observed in the
switching-polarity test. To examine the role played by the ZnO, we replaced the PFW
37
LETTERS
NATURE NANOTECHNOLOGY
DOI: 10.1038/NNANO.2008.314
CB
EF
VB
V+
V+
RL
10
b
Ep
FS
Current (pA)
5
R
L
a
e
FR
10
0
1
Time (s)
V+
V1
V1+
Figure 4 | Proposed mechanism for the generation of current in single-wire generators. a, Energy band diagram of a ZnO piezoelectric ne wire (PFW)
making a Schottky contact with a metal electrode on the left and an ohmic contact with an electrode on the right: the dotted line is the Fermi level, EF, of the
electrodes; CB and VB are the conduction and valence bands of the ZnO; and FSB is the height of the Schottky barrier. A realistic measurement circuit
containing a load resistor, RL , that is much smaller than the resistance of the Schottky barrier and the PFW, is also shown. bd, The piezoelectric potential of
the PFW is positive at the Schottky barrier side V, and negative at the Ohmic contact side V 2. When the PFW is tensile stretched, the Fermi levels of the
two electrodes shift relative to each other as described in the text, generating a positive voltage/current peak as a result of electrons owing from the right
electrode to the left electrode (b), and they re-reach equilibrium (c), and they shift reversely again when the PFW is released (d), generating a negative
voltage/current peak as the electrons ow back (from the left electrode to the right electrode). e, Experimentally measured output current from an SWG,
showing the four different stages described above in ad. fh, When the piezoelectric potential of the PFW is negative at the Schottky barrier side V 2, and
positive at the ohmic contact side V , the Fermi levels shift in a different way during the stretchrelease cycle, as described in the text.
by a carbon bre, which is not piezoelectric, and found that no electricity was
generated during the stretch release cycle. We then coated a Kevlar bre with a
500-nm-thick polycrystalline ZnO lm. The contacts still maintained their ohmic
behaviour, but no electricity was generated. This is because the ZnO nanocrystals in
the lm were randomly oriented without a measurable piezoelectric effect. After
these studies (see Supplementary Information, Figs S2,S3,S4) we concluded that the
presence of a Schottky contact at the interface and the piezoelectric property of the
material are mandatory for SWGs. Further experiments showed that an SWG also
produced output if both ends of the PFW were Schottky contacts.
4. Tian, B. et al. Coaxial silicon nanowires as solar cells and nanoelectronic power
sources. Nature 449, 885 890 (2007).
5. Patolsky, F. et al. Detection, stimulation and inhibition of neuronal signals with
high-density nanowire transistor arrays. Science 313, 1100 1104 (2006).
6. Javey, A., Guo, J., Wang, Q., Lundstrom, M. & Dai, H. J. Ballistic carbon
nanotube eld-effect transistors. Nature 424, 654 657 (2003).
7. Bachtold, A., Hadley, P., Nakanishi, T. & Dekker, C. Logic circuits with carbon
nanotube transistors. Science 294, 1317 1320 (2001).
8. Chen, J. et al. Bright infrared emission from electrically induced excitons in
carbon nanotubes. Science 310, 1171 1174 (2005).
9. Wang, X. D., Song, J. H., Liu, J. & Wang, Z. L. Direct-current nanogenerator
driven by ultrasonic waves. Science 316, 102 105 (2007).
10. Wang, X. D., Liu, J., Song, J. H. & Wang, Z. L. Integrated nanogenerators in
biouid. Nano Lett. 7, 2475 2479 (2007).
11. Qin, Y., Wang, X. D. & Wang, Z. L. Microber nanowire hybrid structure for
energy scavenging. Nature 451, 809 813 (2008).
12. Pan, Z. W., Dai, Z. R. & Wang, Z. L. Nanobelts of semiconducting oxides. Science
291, 1947 1949 (2001).
13. Liu, J. et al. Carrier density and Schottky barrier on the performance of DC
nanogenerator. Nano Lett. 8, 328 332 (2008).
14. Zhou, J. et al. Mechanical electrical triggers and sensors using piezoelectric
microwires/nanowires. Nano Lett. 8, 2725 2730 (2008).
NATURE NANOTECHNOLOGY
DOI: 10.1038/NNANO.2008.314
LETTERS
15. Zhou, J. et al. Flexible piezotronic strain sensor. Nano Lett. 8, 3035 3040 (2008).
16. Zhou, J. et al. Piezoelectric-potential-controlled polarity-reversible
Schottky diodes and switches of ZnO wires. Nano Lett. Doi: 10.1021/nl802497e
(2008).
Author contributions
Acknowledgements
Additional information
This research was supported by the US Department of Energy (Basic Energy Sciences), the
National Science Foundation, the Emory-Georgia Tech Cancer Centre for Nanotechnology
Excellence ( funded by the National Institutes of Health) and the US Air Force Ofce of
Scientic Research (FA9550-06-1-0384).
Z.L.W. and R.S.Y. designed the experiments. R.S.Y., Y.Q. and L.D. performed the
experiments. Z.L.W. and R.S.Y. analysed the data and wrote the paper. All authors
discussed the results and commented on the manuscript.
39