Your Name
Journal Citation in the following form: First-author surname, First-author given name et al (if
>1 authors). Article Title. Journal Name (year of publication) Volume, Issue, page numbers. [e.g.Wujcik, Evan K. et al. An acetylcholinesterase-inspired biomimetic toxicity sensor. Chemosphere
(2013) 91, 1176–1182.]
Overview paragraph This paragraph will answer the following questions, in relation to the article:
Why are the researchers conducting this research/who will this work benefit? What was done in
the past, and how does this research improve upon it? What is innovative about this work...if
anything? What fabrication methods are used? What materials are used? What is the
analyte/species being sensed? What characterization methods are used? What is the detection
method? What is the transduction method? Can the sensor be classified by a lecture (or
combination) we saw in class? What is concluded? What is the planned direction of the work/what
are the researchers’ future plans for the work? This paragraph's information will be strictly taken
from the article...but in your own words, not copy-pasted. The entire journal review (including
headings and paragraphs) should be single spaced and typed with a 10-12 point common font
(Arial or Times New Roman). The entire review (not including heading and citation) should fall
in the range of 550-600 words. I do not need a hard copy.
Impression paragraph This paragraph will be your “review” of the article. Things to include/look
for within the article: Is the article in well-written, proper English? Did the researchers actually
accomplish what they set out to do? Is their work a large or insignificant improvement over what
was done before? Is everything labeled correctly or do some parts (e.g. - references to figures,
labels on figures, table headings, etc.) not match up? Is the concept of the article clear and does
it come across easily/is it well explained or confusing? Were the methods completed ethically?
Lastly: Did you enjoy the article? What did you like/dislike about it?
Full paper
Nanogenerators
www.advenergymat.de
Elastic-Beam Triboelectric Nanogenerator for HighPerformance Multifunctional Applications: Sensitive Scale,
Acceleration/Force/Vibration Sensor, and Intelligent Keyboard
Yuliang Chen, Yi-Cheng Wang, Ying Zhang, Haiyang Zou, Zhiming Lin, Guobin Zhang,
Chongwen Zou, and Zhong Lin Wang*
harvesting techniques has turned into
one of the best alternatives for the new
era—the era of Internet of Things.[1–3]
On the other hand, mechanical energy
is ubiquitous in the ambient environment, such as vibration, breeze, and water
wave. If the energy can be converted into
electrical output, it could become sustainable power sources to meet the lower
power requirement of various sensors.
In 2012, Dr. Zhong Lin Wang’s group
first proposed an emerging technique
called triboelectric nanogenerator (TENG)
that can efficiently convert mechanical
energy from the ambient environment
into electrical output, taking advantage
of contact-electrification and electrostatic
induction.[4] The converted electrical
signals can then be used as power sources
or sensor signals. For example, the
mechanical energy of ocean waves, human
motions, wind, and even raindrop can be
converted to electrical energy based on
various types of TENGs.[5–9] In addition, the self-powered sensors based on TENGs have been successfully applied to motion,
vibration, environmental, and biological monitoring.[10–15]
There are four basic operation modes for TENGs.[16] Among
those modes, contact-separation and sliding are two foremost
and fundamental principles, both depend on the motions
between two dissimilar materials.[17–22] The working principles
of the former depend mainly on the separated distance between
two dissimilar materials, since the area of the two materials are
Exploiting novel devices for either collecting energy or self-powered sensors
is vital for Internet of Things, sensor networks, and big data. Triboelectric
nanogenerators (TENGs) have been proved as an effective solution for both energy
harvesting and self-powered sensing. The traditional triboelectric nanogenerators are usually based on four modes: contact-separation mode, lateral sliding
mode, single-electrode mode, and freestanding triboelectric-layer mode. Since the
reciprocating displacement/force is necessary for all working modes, developing
efficient elastic TENG is going to be important and urgent. Here, a kind of
elastic-beam TENG with arc-stainless steel foil is developed, whose structure
is quite simple, and its working states depend on the contact area and separating
distance as proved by experiments and theoretical calculations. This structure
is different from traditional structures, e.g., direct sliding or contact-separation
structures, whose working states mainly depend on contact area or separating
distance. This triboelectric nanogenerator shows advanced mechanical
and electrical performance, such as high sensitivity, elasticity, and ultrahigh
frequency response, which encourage applications as a force sensor, sensitivity
scale, acceleration sensor, vibration sensor, and intelligent keyboard.
1. Introduction
With the rapid development of Internet of Things and sensor
network over the past decades, enormous electronic devices
such as sensors, transmitters, and actuators have been used all
over the world. Each device requires a small power (in microwatt
to milliwatt range) for operation; however, providing persistent power sources for such electronics remains challenging.
Therefore, the exploration for self-powered sensors and energy
Prof. Z. L. Wang
Beijing Institute of Nanoenergy and Nanosystems
Chinese Academy of Sciences
Beijing 100083, China
Prof. Z. L. Wang
College of Nanoscience and Technology
University of Chinese Academy of Sciences
Beijing 100049, China
Y. L. Chen, Dr. Y.-C. Wang, Prof. Y. Zhang, H. Y. Zou, Z. M. Lin,
Prof. Z. L. Wang
School of Materials Science and Engineering
Georgia Institute of Technology
Atlanta, GA 30332-0245, USA
E-mail: zhong.wang@mse.gatech.edu
Y. L. Chen, Prof. G. B. Zhang, Prof. C. W. Zou
National Synchrotron Radiation Laboratory
University of Science and Technology of China
Hefei 230029, China
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/aenm.201802159.
DOI: 10.1002/aenm.201802159
Adv. Energy Mater. 2018, 1802159
1802159 (1 of 9)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advenergymat.de
www.advancedsciencenews.com
generally fixed.[18,19] For the sliding mode, its working principle
is mainly associated with the contacted area of two dissimilar
materials, since there is no vertical separation between the two
materials.[20,23] However, sliding fraction may lead to low durability of the TENGs and the most of the induced charges are
transferring at the moment of contacting or separating between
the two materials.[18,19,24,25] In addition, in order to continuously
harvesting mechanical energy or using TENGs as self-powered
sensors, it is very critical for a TENG to recover to its original
state automatically after the mechanical motion was removed.
Researchers have applied spring or elastic polymers to deal with
this problem;[11,26–29] however, those methods could increase
the complexity of the device fabrication that may prevent them
from mass production.
To address this issue and further improve the performance of
TENGs, herein, we integrated an elastic-beam (EB) arc-stainless
steel foil (arc-SSF) into the TENG structure and developed
a novel EB-TENG. According to our experimental, theoretical,
and numerical results, the working principles of the EB-TENG
depend not only on the area of contact but also on the separation
distance between the arc-SSF and dielectric layer of the
EB-TENG.[14] In addition, after the contact event was removed, e.g.,
releasing the pressed SSF, the system could recover to its original state automatically owing to the elasticity of SSF. We also
showed that the electrical output of the designed EB-TENG
is highly sensitive to external force and can provide excellent
response at high-frequency. By virtue of the structural design
and material properties, we further demonstrated that this EBTENG can be used as force sensor, sensitive scale, acceleration
sensor, vibration sensor, or smart keyboard. We believe this
novel design of the EB-TENG with arc-SSF can be generalized
to other TENGs that can further improve the performance of
energy harvesting and self-powered sensors.
2. Results and Discussion
2.1. Structural Design and Working Principle of EB-TENG
event happens between SSF and PTFE triggered by mechanical force, I) the SSF becomes positively charged and the
PTFE becomes negatively charged, based on the triboelectric series.[30] II) These triboelectric charges (Q) stay on both
SSF and PTFE, due to the open-circuit condition (for SSF)
and the dielectric properties of the PTFE layer. Pressing
the SSF would not change the amount of triboelectric charge
but change the capacitance of EB-TENG (C), which is affected by
increasing the contacted area (A) and separated distance (h)
between the SSF and the PTFE. The open-circuit voltage (U)
is inversely proportional to C. Hence, U would change when
pressing the SSF. III) When the SSF is fully contacted with
the PTFE, the A reaches maximum and h to minimum. IV)
Releasing the SSF to recover A and h partially. Finally, the SSF
moves back to initial position to complete a cycle. The defined
variables are listed in Table S1 in the Supporting Information.
2.2. Performance Characterization, Theoretical Calculation, and
Finite Element Analysis (FEA)
To take advantage of the structural design and further
apply the EB-TENG to other applications, it is necessary
to test the mechanical behaviors and the electrical output
of the EB-TENG. Figure 2a shows the experimental setup:
an analytical balance was placed beneath the EB-TENG
and was used to record the external force (F); an oscilloscope was used to measure the open-circuit voltage U.
The height (h) is defined as the vertical distance from the
vertex of the SSF to the PTFE. The displacement (d) is defined
as the variation of h, d = h0 − h, where h0 is the initial height
of the vertex of SSF. When pressed the SSF, the h was then
reduced (d increased), and we can get the corresponding U, F,
and d (Figure 2b). Intuitively, the relationships of U versus F,
U versus d, and F versus d are illustrated in Figure 2c,e and
Figure S2 in the Supporting Information, respectively. Figure 2b and
Figure S2 in the Supporting Information display that the last
point is beyond the range of measurement. Therefore, we did
not plot the last point in Figure 2c,e, which indicates that the
U increases with increasing F and d. In addition, it should be
Figure 1a is the schematic of EB-TENG based on arc-SSF.
A copper (Cu) film and a polytetrafluoroethylene (PTFE) film were pasted on an acrylic
(polymethyl methacrylate, PMMA) board successively. Then, one end of an arc-SSF was
fixed on the PTFE/Cu/PMMA structure to
complete the fabrication of EB-TENG. The
width and length of the EB-TENG are 7.5
and 60 mm, respectively. Other details of
the fabrication processes are presented in
the Experimental Section. Figure 1b shows
the photo of the EB-TENG device. Two
wires were connected to the SSF and the
copper film with the conjunctions winded
by Kapton tape. The length of EB-TENG
can be adjusted by a black rubber ring
(Figure S1a, Supporting Information). The
Figure 1. a) The schematic of the elastic-beam triboelectric nanogenerator (EB-TENG). b) The
working principle of the EB-TENG at open- picture of the actual EB-TENG device. A black rubber ring is used to adjust the length of
circuit condition is summarized as follows EB-TENG. c) The working principle at open-circuit condition. The output of EB-TENG depends
(Figure 1c). After the first contact-separation on the contact area and separation distance between the SSF and the PTFE.
Adv. Energy Mater. 2018, 1802159
1802159 (2 of 9)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advenergymat.de
Figure 2. a) The experimental setup: an analytical balance is placed beneath the EB-TENG and is used to record the F applied; an oscilloscope used
to record the U. b) The experimental results of U versus d and F. c) The results of U versus d, the experimental data and fitting data indicated by olive
diamond and orange curve, respectively. And the corresponding differential curve in (d). e) The U versus F, the experimental data and fitting data
indicated by navy dot and pink curve, respectively. And the corresponding differential curve in (f). g) The potential distribution of the initial state of
the EB-TENG under open-circuit condition. h) The potential distribution of the partially pressed state of the EB-TENG under open-circuit condition.
pointed out that the relationships of U versus F and U versus d
are not linear in Figure 2e,c, which is different from the results
of the regular contact-separation TENGs in ideal plane-parallel
capacitor mode,[18] showing linear relationships. Figure S3a in
the Supporting Information shows the standard deviation of U
with increasing d from independent measurements and one
of the measurements shows in Figure S3b in the Supporting
Information. Figure S1b in the Supporting Information shows
the experimental results of U versus d and F after reducing the
length of SSF by moving the rubber ring.
In order to further understand the working mechanism of
the EB-TENG, we conducted theoretical calculation and FEA. In
previous reports, the open-circuit voltage U of TENGs is always
associated with contacted area and separated distance between
two dissimilar materials.[18,20,31,32] For our device, the width of
SSF is fixed at 7.5 mm, so the contacted area (A) between the SSF
Adv. Energy Mater. 2018, 1802159
and PTFE is proportional to the length (l) of the contacted part of
SSF. The separated distance between the vertex of SSF and PTFE
could be presented as h, since the SSF has a symmetrical shape.
Moreover, d = h0 − h, hypothesized the U can be written as follows
U = k1l + k2d + k3
(1)
where k1, k2, and k3 are fitting parameters; k3 is introduced as
a constant that includes other factors, such as the roughness of
friction layers, edge effect. The fitting results can be described
as follows
F
15.16 − 0.00678
U = −158 arcsin
− 351.03F + 63.34
10
1802159 (3 of 9)
(2)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advenergymat.de
www.advancedsciencenews.com
15.16 − d
U = −158 arcsin
− 2.38d + 63.34
100
(3)
It can be found that the nonlinearity for U(F) and U(d) attributing to the antitrigonometric functions in Equations (2) and
(3). The fitting results are plotted in Figure 2c,e with each corresponding measured data. The detail derivation is described in
Note S1 in the Supporting Information.
Figure 2g is the potential distribution of the initial state
under open-circuit condition by FEA. The colors correspond
to the magnitude of potential. In the top panel of Figure 2g, it
indicates a higher potential for SSF and a lower potential for
PTFE as expected due to the negative charges
transferred from SSF to PTFE after contacting (followed by separating event). The
arrows show the electric field lines. Their
directions from high potential to low potential match the potential distribution. Interestingly, we found the electric field lines are
perpendicular to the surface of SSF, whereas
they are inclined to the surface of PTFE.
As we known, SSF is a conductor and PTFE is
an insulator, which means SSF is an equipotential body and PTFE is not. So there is no
electric field component along the surface of
SSF. The equipotential lines in the bottom
part of Figure 2g verify the speculation,
i.e., the potential of the right part of PTFE
is lower than that of the left part. Similar
results can be found in the partially pressed
state (Figure 2h), whose potential difference
(U) is reduced compared to initial state,
which matches our experimental results in
Figure 2c,e. It should be noted that the SSF
was connected to the ground in our experimental setup. And the potential difference
U of the initial state was set to the baseline.
This is the reason that U increasing with
larger F (Figure 2e) and d (Figure 2c).
and 1.867 ± 0.03 g for two screws, which is well accordant
with the actual values. In practical scenarios, we usually want
to weigh the mass to an expected value. In the up panel of
Figure 3b, we increased the mass to the expected value (8 g)
step by step. And the below panel of Figure 3b shows the scenario of increasing and reducing mass randomly. For these two
scenarios, the tested data well match the actual results, which
demonstrate great performance in our EB-TENG. According to
the results in Figure 2e, the corresponding range of measured
mass is 0–9.85 g for the present EB-TENG. The wider measured range can be achieved by applying high Young’s modulus
SSF in similar structure of EB-TENG.
2.3. Sensitive Scale and Acceleration Sensor
Figure 2f shows the differential curve of
dU versus dF, which could be used to estimate the sensitivity of EB-TENG. It suggests
the sensitivity dU/dF is about 30–900 V N−1.
The calculated sensitivity dU/dd can be up
to maximum ≈6 N mm−1 in Figure 2d. The
high sensitivity suggests the excellent performance of the EB-TENG (the higher sensitivity
can be achieved by changing the fabrication
of EB-TENG, e.g., the wider SSF applied).
To further take advantage of this property,
we then applied the EB-TENG as a scale. In
Figure 3a, two same screws (each is 0.9350 g)
were weighed by using EB-TENG. The measured values derived from the relationship of
U versus F are 0.932 ± 0.04 g for single screw
Adv. Energy Mater. 2018, 1802159
Figure 3. a) Two identical screws are put onto the tip of the SSF and weighed by the EB-TENG.
b) The two different situations of weighing. The up panel shows weighing mass to the expected
value step-by-step. The bottom part shows increasing and decreasing mass randomly during
the weighing process. c) The schematic for the EB-TENG as an acceleration sensor. The EBTENG is fixed on a disk and could spin with the disk. The actual picture of the EB-TENG spinning d) at low revolving speed and e) at high revolving speed. The red dashed line shows the
PTFE and the blue dashed line shows the arc-SSF. A tiny 0.4 g weight is fixed at the vertex of
SSF. f) The experimental results of U versus revolving speed (R) and acceleration (a). g) The
calculated acceleration (gray triangle) and the tested acceleration by EB-TENG (orange circle).
The error bars represent standard deviation from independent measurements.
1802159 (4 of 9)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advenergymat.de
According to above discussion, the fabricated EB-TENG definitely can be used as a force sensor. Based on F = ma, m is mass
and a denotes acceleration, if set a at a constant value, the EBTENG can be used to measure m as discussed in last section
(a = g, where g is gravitational acceleration and treated as a constant value). On the other hand, if set m at a constant value, the
EB-TENG can be used as an acceleration sensor.
As Figure 3c shown, an EB-TENG is fixed at the edge of a
disk. Then the combined structure was installed on a rotary
motor. The EB-TENG would subject to different centrifugal
forces when the disk was spinning at different speeds, in
another word, the different acceleration can be applied on
the EB-TENG in terms of only a same 0.4 g weight fixed at
the vertex of SSF. When the revolving speed (R) was fast
(Figure 3e), the pressed degree was deeper than that of the
pressed state at slow revolving speed (Figure 3d). Figure 3f
shows the measured results of U versus R (pink curve) and
a (olive curve) when the EB-TENG was put on a radius (r)
30 mm disk, which indicates faster R or bigger a would bring
greater U. Figure S4 in the Supporting Information shows
one of the measurements, U versus time at different R.
These relationships are accordant to the positive correlation
of U versus F in Figure 2e. Furthermore, in order to test the
performance of the EB-TENG, it was fixed on another radius
50 mm disk. In Figure 3g, the measured data derived from
the relationship of a and U in Figure 3f can well match the
calculated acceleration (by formula v2/r = (2πrR)2/r, where v is
speed), and the inconformity between them could attribute to
the influence of wind when the disk was spinning. The influence of wind could be avoided, for example, packaging the EBTENG in a box.
2.4. Frequency Simulation and Vibration Sensor
Furthermore, we checked the frequency performance of EBTENG. Taken a fixed shape to consider, the eigenfrequency
(or resonant frequency) f ∝ E / ρ , where E notes Young’s
modulus and ρ is density.[33] For usual elastic materials, such
as rubber, the typical value of E is in the range of ≈106 Pa. However, the E of SSF is about five orders of magnitude higher (in
≈1011 Pa range). More importantly, the densities of SSF and
rubber are generally in the same order of magnitude
(≈103 kg m−3). This result implies that our SSF-based EB-TENG
could have a better performance at high-frequency. Figure 4a
shows the relationship of eigenfrequency versus the noncontacting arc length of SSF by FEA, which suggests the eigenfrequency can be from 53 to 85 279 Hz when the length varies from
40 to 1 mm. The eigenfrequency values of the SSF are almost
two orders of magnitude higher than that of rubber (Figure S5,
Supporting Information, details in Note S2, Supporting Information). Figure 4b is the simulated results for 40 mm arc length
(initial state) operated at the eigenfrequency, which shows
the vertex has a bigger displacement from the color mapping
(Figure S6, Supporting Information). Similar results for
20 mm arc length are displayed in Figure S7 in the Supporting
Information.
To apply our theoretical analysis and simulation results,
we conducted an additional set of experiments. The vertex of
Adv. Energy Mater. 2018, 1802159
Figure 4. a) The simulated results of eigenfrequency versus the uncontacting arc length of SSF. b) The simulated results for 40 mm arc length
(initial state) working at the eigenfrequency. The color displays the displacement distribution of each position qualitatively.
the SSF on the EB-TENG was set to contact with a vibrator as
shown in Figure 5e. The signals were input into the vibrator to
drive the EB-TENG. The measured signals by EB-TENG can be
obtained by an oscilloscope or played by a loudspeaker through
an amplifier. Figure 5a–d shows the measured voltage signals
by EB-TENG when the sine waves at different frequencies were
input into the vibrator (more results in Figure S8, Supporting
Information). Compared to the background noise (the initial
small signals in Figure 5a–d), the detected signals are quite
prominent with high signal-to-noise ratio even up to ultrahigh
frequency (e.g., 18 kHz). The amplitude of voltage decreasing
with time in Figure 5b, Figure S8b in the Supporting Information, and increasing in Figure 5d probably originate from the
measuring instrument rather than the EB-TENG. When a commercial vibration sensor was adopted to measure the voltage
signal at 1 kHz under the same conditions, the results show
the same decreasing trend as shown in Figure S9 in the Supporting Information. The frequencies that can be heard by
human being are between 20 Hz and 20 kHz. Then, we input
1802159 (5 of 9)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advenergymat.de
www.advancedsciencenews.com
Figure 5. a–d) The measured voltage signals by EB-TENG at different frequencies. e) The experimental setup of the frequency response for EB-TENG.
f) A piece of music input to a vibrator. The up panel shows the detected wave by EB-TENG. The bottom panel shows the original wave. The detailed
comparison between original results and detected results by EB-TENG is presented in (g).
a piece of music into the vibrator, and the detected results are
demonstrated in Figure 5f,g. Compared to the original signals,
the recorded signals show a good accordance (Figure S10, Supporting Information), which can be used to generate an audio
file (Audio S1, Supporting Information). In this audio, the different sounds of instruments can be distinguished, such as
drum, trombone.
2.5. Intelligent Keyboard Application
For traditional keyboards, the method of input is based on the on/
off states of each switch (unit), corresponding to 1 and 0, respectively, as shown in Figure 6a. Before the switch is connected, the
all states of switch only represent 0. However, in the big data
and artificial intelligence (AI) era, an intelligent keyboard that
can provide additional information is highly desired.[28] Herein,
Adv. Energy Mater. 2018, 1802159
we developed a multiple states input devices (MSIDs) based on
EB-TENG, taking advantage of the successive voltage variation
of the pressed states (Figure 6b). The inset of Figure 6b shows
the fabricated numeric keyboard integrating numbers from
1 to 9 (the fabricated details in the Experimental Section). First,
we tested the on/off performance for individual key. In Figure 6c,
two binary sequences based on Morse code (Table S2, Supporting Information) were input to express “SOS” and “911”,
respectively. The on/off states can be switched sharply, which
suggests the fabricated EB-TENG keyboard is compatible with
the function of traditional keyboards. Then a numeric sequence
“298 157 436” was input randomly with different pressing
force. In Figure 6d, it can be observed that the input magnitudes of numbers are different, which can achieve the function
of MSIDs. For instance, in Figure 6e, the “3” has the largest
magnitude and “1” has the smallest magnitude. The magnitudes can carry additional information in numerous practical
1802159 (6 of 9)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advenergymat.de
Figure 6. a) A traditional keyboard based on on/off switch that only has binary states. b) A smart keyboard based on EB-TENG that can have multiples
states. c) Two binary sequences based on Morse code were input to express “SOS” and “911” by EB-TENG keyboard. d) A numeric sequence input
by EB-TENG keyboard with different signal magnitudes due to different pressing forces. e) The extracted normalized magnitude of the voltage output
from (d). f) The numbers in different size, which is proportional to the extracted normalized magnitude of the voltage output. g) The measured and
fitting results of transferred short-circuit charge (Qsc) versus d. h) The short-circuit current (Isc) of periodic pressing SSF (d = 1.5 mm) at different typing
frequencies. i) Three typical typing results for indicating the basic principle of keystroke dynamics by EB-TENG keyboard.
Adv. Energy Mater. 2018, 1802159
1802159 (7 of 9)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advenergymat.de
www.advancedsciencenews.com
applications. For example, if we want characters with different
sizes, we can set the size of the characters to be proportional to
the electrical output magnitude. Figure 6f shows the numeric
sequence with different sizes, based on different output
magnitudes.
Based on people’s typing attributes, keystroke dynamics
as a kind of unconditioned reflex, for example, the typing
force and typing speed, can be utilized as the “fingerprint” to
develop personal identity verification, cyber security, and AI
technology.[34–36] Here, the short-circuit condition was applied
to our EB-TENG and the short-circuit working processes are
specified in Figure S11 in the Supporting Information. We
found that the transferred short-circuit charge (Qsc) and shortcircuit current (Isc) can reflect the vital information of keystroke
dynamics. Taken individual key of EB-TENG keyboard to
consider, the increase of Qsc accompanies the more forceful
pressing due to the short-circuit working condition (Figure 6g),
which implies the information of typing force. The fitting equation of Qsc versus d (olive curve in Figure 6g) is
Q sc = 0.275d 2 + 0.282d
(4)
When the pressing state keeps constant (d = 1.5 mm), the
higher typing frequency produces bigger Isc as Figure 6h shown
(more results based on other d and frequencies in Figure S12,
Supporting Information), which can be explained by formula
Isc = dQsc/dt, in a shorter time (t) to obtain a certain Qsc, the
bigger Isc would be generated. Moreover, three typical examples are presented to expound the basic principle of keystroke
dynamics by EB-TENG keyboard in Figure 6i. Compared the
results between user 1 and user 2, the Qsc are almost same,
which means that the two users have near the same typing
force. However, the amplitude of Isc of user 2 is almost four
times that of user 1. Obviously, user 2 is typing faster than
user 1. The Qsc of user 3 is more than that of other users, so
user 3 has the biggest typing force in the three users. Although
the amplitude of Isc of user 3 is highest, the typing speed of
user 3 is difficult to directly compare with other users through
Isc because the bigger Isc of user 3 could originate from the
increased amount of Qsc or faster typing speed. In order to
exclude the influence of increased amount of Qsc, a parameter
ϕ was introduced
ϕ=
I scm
dQ sc
dd
(5)
where Iscm is the amplitude of Isc. According to Equations (4)
and (5), the calculated ϕ are 7.83, 31.19, and 31.37 for these three
users, respectively. With these results, in fact, user 3 has near
typing speed to user 2 even if user 3 has the biggest typing force.
The more discussion is in Note S3 in the Supporting Information.
our experiments and theoretical calculation, which is different
from traditional structures, for example, contact-separation
mode depending on the separation distance and sliding mode
depending on the contact area. Moreover, we developed this
EB-TENG to be used as a force sensor, sensitive scale, acceleration sensor, vibration sensor, and intelligent keyboard, which all
showed good performance. This novel structure of EB-TENG
with arc-SSF can be applied to other TENGs as devices of collecting energy and self-powered sensor in future.
4. Experimental Section
Fabrication of the EB-TENG: A 7.5 mm × 60 mm × 5 mm acrylic board
was used as the substrate. First, a copper film as an electrode was
pasted on the top of the acrylic board. Then, the PTFE (with 0.127 mm
thickness) was pasted on top of the copper film. The SSF (Alfa Aesar,
type 304, width of 7.5 mm, arc length of 45 mm, thickness of 0.1 mm),
bent to an arc with the radius of 50 mm, was used as both contacting
layer and electrode. The one end of SSF, 5 mm length, was fixed on the
PTFE/Cu/PMMA structure. The actual maximum contacting length for
arc-SSF is then 40 mm. Two wires were connected to SSF and copper
film with the conjunctions winded by Kapton tape. A black rubber ring
was used to bundle the device tightly, which can be used to adjust the
length of SSF.
Fabrication of Multiple States Keyboard: An 85 mm × 85 mm × 3.5 mm
acrylic board was used as the substrate. The nine identical EB-TENGs
were fabricated on the acrylic board, and the whole fabricated processes
were similar to the above section. The arc-SSFs (width of 6.5 mm, arc
length of 20 mm, thickness of 0.1 mm, radius of 50 mm) were used.
One end of the SSF, 4 mm length, was fixed, so the actual maximum
contacting length is 16 mm.
Characterization: The electric measurement was conducted by a
Keithley 6514 system electrometer. A software platform programmed
using LabVIEW to achieve real-time data acquisition and analysis. A
commercial linear mechanical motor was used to precisely adjust the
separated distance between PTFE and SSF. A commercial analytical
balance was used as a force sensor (OHAUS, PA224C), with readability
1 × 10−4 g. The actual force applied was calculated by the following formula,
F = mg, where m is mass; g is gravitational acceleration, value adopted
9.8 m s−2. The EB-TENG was put on the analytical balance and the SSF
was pressed by linear mechanical motor in the experiments. In order to
collect signal of acceleration sensor, two copper electrodes were glued
beneath the disk, which respectively connected to the two terminals of
EB-TENG, as shown in Figure 3d,e. Then two electric brushes contacted
the electrodes, respectively. The electrometer collected the data by two
leads connecting to the electric brushes. The vibrator was based on a
commercial multimedia speaker system (MPC, MMP001176-01). The
signals with different frequencies were input into the vibrator through
a computer. The input music is part of “Georgia Tech (GT) fight song”.
Finite Element Analysis: The FEA is carried out by utilizing the COMSOL
software 5.2a. The potential distribution was calculated at different
states under the open-circuit condition. It should be emphasized that
the simulations in previous reports applied ideal average charge density
(≈10−4 cm−2) without considering the real micromorphology.[37,38] Here,
it is thought that the tested voltage results can also be the reference
criterion for the selection of average charge density. The 5 × 10−8 cm−2
was adopted in our simulation. The simulation of eigenfrequency was
conducted for each different length of SSF (40–1 mm). The whole SSF
was set at free state except one of the ends was constrained.
3. Conclusion
In this work, we fabricated a novel EB-TENG based on SSF. The
structure of EB-TENG is quite simple, and its working states
depend on the contact area and separation distance proved by
Adv. Energy Mater. 2018, 1802159
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
1802159 (8 of 9)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
www.advenergymat.de
Acknowledgements
The authors acknowledge support from King Abdullah University of
Science and Technology (KAUST), the Hightower Chair foundation,
and the “thousands talents” program for pioneer researcher and his
innovation team, China. Y.L.C. thanks China Scholarship Council for
supplying oversea scholarship (201706340019).
Conflict of Interest
The authors declare no conflict of interest.
Keywords
elastic triboelectric nanogenerators, intelligent keyboards, multifunctional
sensors, self-powered sensors
Received: July 13, 2018
Revised: July 27, 2018
Published online:
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
Z. L. Wang, Mater. Today 2017, 20, 74.
Z. L. Wang, J. Chen, L. Lin, Energy Environ. Sci. 2015, 8, 2250.
S. Wang, L. Lin, Z. L. Wang, Nano Energy 2015, 11, 436.
F.-R. Fan, Z.-Q. Tian, Z. L. Wang, Nano Energy 2012, 1, 328.
S. L. Zhang, M. Xu, C. Zhang, Y.-C. Wang, H. Zou, X. He, Z. Wang,
Z. L. Wang, Nano Energy 2018, 48, 421.
J. Wang, S. Li, F. Yi, Y. Zi, J. Lin, X. Wang, Y. Xu, Z. L. Wang, Nat.
Commun. 2016, 7, 12744.
M. Xu, Y.-C. Wang, S. L. Zhang, W. Ding, J. Cheng, X. He, P. Zhang,
Z. Wang, X. Pan, Z. L. Wang, Extreme Mech. Lett. 2017, 15, 122.
J. Bae, J. Lee, S. Kim, J. Ha, B. S. Lee, Y. Park, C. Choong, J. B. Kim,
Z. L. Wang, H. Y. Kim, J. J. Park, U. I. Chung, Nat. Commun. 2014,
5, 4929.
Z. H. Lin, G. Cheng, S. Lee, K. C. Pradel, Z. L. Wang, Adv. Mater.
2014, 26, 4690.
K. Dong, J. Deng, Y. Zi, Y. C. Wang, C. Xu, H. Zou, W. Ding, Y. Dai,
B. Gu, B. Sun, Z. L. Wang, Adv. Mater. 2017, 29, 1702648.
J. Chen, G. Zhu, W. Yang, Q. Jing, P. Bai, Y. Yang, T. C. Hou,
Z. L. Wang, Adv. Mater. 2013, 25, 6094.
M. Xu, P. Wang, Y.-C. Wang, S. L. Zhang, A. C. Wang, C. Zhang,
Z. Wang, X. Pan, Z. L. Wang, Adv. Energy Mater. 2018, 8, 1702432.
J. Wang, W. Ding, L. Pan, C. Wu, H. Yu, L. Yang, R. Liao, Z. L. Wang,
ACS Nano 2018, 12, 3954.
Adv. Energy Mater. 2018, 1802159
[14] P. K. Yang, L. Lin, F. Yi, X. Li, K. C. Pradel, Y. Zi, C. I. Wu, J. H. He,
Y. Zhang, Z. L. Wang, Adv. Mater. 2015, 27, 3817.
[15] J. Yang, J. Chen, Y. Su, Q. Jing, Z. Li, F. Yi, X. Wen, Z. Wang,
Z. L. Wang, Adv. Mater. 2015, 27, 1316.
[16] Z. L. Wang, Faraday Discuss. 2014, 176, 447.
[17] S. Niu, S. Wang, Y. Liu, Y. S. Zhou, L. Lin, Y. Hu, K. C. Pradel,
Z. L. Wang, Energy Environ. Sci. 2014, 7, 2339.
[18] S. Niu, S. Wang, L. Lin, Y. Liu, Y. S. Zhou, Y. Hu, Z. L. Wang, Energy
Environ. Sci. 2013, 6, 3576.
[19] R. D. I. G. Dharmasena, K. D. G. I. Jayawardena, C. A. Mills,
J. H. B. Deane, J. V. Anguita, R. A. Dorey, S. R. P. Silva, Energy
Environ. Sci. 2017, 10, 1801.
[20] S. Niu, Y. Liu, S. Wang, L. Lin, Y. S. Zhou, Y. Hu, Z. L. Wang, Adv.
Mater. 2013, 25, 6184.
[21] C. Xu, Y. Zi, A. C. Wang, H. Zou, Y. Dai, X. He, P. Wang,
Y. C. Wang, P. Feng, D. Li, Z. L. Wang, Adv. Mater. 2018, 30,
1706790.
[22] M. Sow, D. J. Lacks, R. M. Sankaran, J. Electrost. 2013, 71, 396.
[23] S. Wang, L. Lin, Y. Xie, Q. Jing, S. Niu, Z. L. Wang, Nano Lett. 2013,
13, 2226.
[24] L. Lin, S. Wang, S. Niu, C. Liu, Y. Xie, Z. L. Wang, ACS Appl. Mater.
Interfaces 2014, 6, 3031.
[25] L. Lin, S. Wang, Y. Xie, Q. Jing, S. Niu, Y. Hu, Z. L. Wang, Nano Lett.
2013, 13, 2916.
[26] L. Xu, T. Z. Bu, X. D. Yang, C. Zhang, Z. L. Wang, Nano Energy 2018,
49, 625.
[27] L. Lin, Q. Jing, Y. Zhang, Y. Hu, S. Wang, Y. Bando, R. P. S. Han,
Z. L. Wang, Energy Environ. Sci. 2013, 6, 1164.
[28] C. Wu, W. Ding, R. Liu, J. Wang, A. C. Wang, J. Wang, S. Li, Y. Zi,
Z. L. Wang, Mater. Today 2018, 21, 216.
[29] M. Ma, Q. Liao, G. Zhang, Z. Zhang, Q. Liang, Y. Zhang, Adv. Funct.
Mater. 2015, 25, 6489.
[30] A. F. Diaz, R. M. Felix-Navarro, J. Electrost. 2004, 62, 277.
[31] M. M. Apodaca, P. J. Wesson, K. J. Bishop, M. A. Ratner,
B. A. Grzybowski, Angew. Chem., Int. Ed. 2010, 49, 946.
[32] S. Niu, Y. Liu, S. Wang, L. Lin, Y. S. Zhou, Y. Hu, Z. L. Wang, Adv.
Funct. Mater. 2014, 24, 3332.
[33] S. M. Han, H. Benaroya, T. Wei, J. Sound Vib. 1999, 225, 935.
[34] M. L. Ali, J. V. Monaco, C. C. Tappert, M. Qiu, J. Signal Process. Syst.
2016, 86, 175.
[35] S. P. Banerjee, D. Woodard, J. Pattern Recognit. Res. 2012, 7, 116.
[36] F. Monrose, A. D. Rubin, Future Gener. Comput. Syst. 2000,
16, 351.
[37] J. Wang, C. Wu, Y. Dai, Z. Zhao, A. Wang, T. Zhang, Z. L. Wang, Nat.
Commun. 2017, 8, 88.
[38] J. Chun, B. U. Ye, J. W. Lee, D. Choi, C. Y. Kang, S. W. Kim,
Z. L. Wang, J. M. Baik, Nat. Commun. 2016, 7, 12985.
1802159 (9 of 9)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Purchase answer to see full
attachment