CHAPTER 4. RESULTS AND DISCUSSION
4.1 Silk fibroin based biodegradable piezoelectric composite sensor
4.1.2 Output enhancement for high sensitivity in self-powered sensor
To measure the piezoelectric power-generating performance of the device, a direct impact approach was used. A foot of human was used to apply vertical compressive force on the devices. Figures 6a and 6b show a photography of the actual experimental setup for scavenging energy from the foot motion,
be clearly seen that the change of the piezopotentials increased linearly as the external force increased from 100 to 500 µN for the composite film type nanogenerator.
Figure 6. (a) The original images indicating the experimental setup for power generation under the foot step in which silk fibroin-based composite-type generator was laid down a woodblock. (b) The output voltage and current density generated from the composite type nanogenerator with 30 wt% KNN:Mn nanoparticles. (c) The output voltage obtained with the dispersing agents (Ag, PVP). It shows that the output power increases with the degree of the dispersion, as shown in the inset. (d) The generated output voltages with the combination of dispersing agents. The inset shows the current-voltage (I–V) characteristics with the dispersing agents.
As mentioned above, we believe that the nanoparticles with same size distribution (400 ~ 500 nm) are well-dispersed in the silk fibroin matrix. However, it may not be easy to investigate how the nanoparticles are dispersed in the matrix. Furthermore, the silk fibroin solution is so low viscous that the nanoparticles naturally and easily sink into the surface of the substrate, as shown in the left inset of
Figure 6c. As expected, the composite nanogenerator with poor dispersion generates the low output voltage of less than 1 V in Figure 6c. To enhance the dispersion, Ag nanowires and PVP were used as dispersing agents and the optimization process was used to promote the dispersed distribution. It is well- known that the Ag nanowires are widely used in dispersing the nanoparticles in the composites, and developing the transparent and conducting electrodes [78-80]. Thus, the PVP was added to the solution to reduce the conductivity of the films (see the inset of Figure 6d), thereby, increase the output performance of the nanogenerator. Actually, when the nanoparticles are mixed into the silk fibroin solution with Ag nanowires (1 wt%) and PVP (12 wt%) together, a maximum output voltage and current density of 2.2 V and 0.12 µA/cm2 were obtained, 2 times higher than that for only KNN:Mn nanoparticles in Figures 6c and 6d. It is thought that the PVP plays role in dispersing nanoparticles in the silk fibroin matrix and wrapping a surface of Ag nanowires in order to prevent Ag nanowires from connecting with each other [81]. The cross-sectional image in the right inset of the Figure 6c confirms that the nanoparticles are well-dispersed in the silk fibroin matrix. In the Figure 8, the calculated piezopotential difference between of aggregated and well-dispersed nanoparticles in the silk fibroin matrix is in good agreement with experimental results.
significantly enhance the mechanical quality factor without deteriorating other piezoelectric properties, expecting the large Q-factor in composites embedded with KNN:Mn nanoparticles (the Q-factor of the composites embedded with various nanoparticles will be separate paper) [82-84]. The electromechanical coupling factor (k) can be expressed by;
where Y, dij, and εT are Young’s modulus, piezoelectric coefficient, and dielectric constant, respectively [85-86]. Nanoindentation measurements were conducted to evaluate the resistance to deformation by external force, and show that there is no significant change in the elastic modulus (0.2 ~ 0.3 GPa) in composites. We also believe that the Ag nanowires do not affect the dielectric properties of the silk fibroin materials because of the low volume fraction (as low as 7 %) [81]. Thus, it is essential in increasing the piezoelectric constant (dij) for the larger output power of the nanogenerator. In Figure 9, among the nanoparticles (BaTiO3, ZnSnO3, BNKT, and KNN:Mn), the KNN:Mn has the largest piezoelectric coefficient (d33) of ~ 270 pC/N and relatively small dielectric constant of about 730 [90].
Thus, the coupling factor of the material is expected to be the highest, leading to the largest output power, consistent with experimental results.
Figure 8. The calculated piezopotential difference between of (a) aggregated and (b) well-distributed ferroelectric particles in the silk fibroin matrix.
ij T
d Y
k
2= ´
2/ e
Figure 9. (a) The output voltages and (b) current densities of the various ferroelectric particles when the concentrations of ferroelectric particles are 30 wt%. The insets illustrate the enlarged graphs of the output signals from the boxed areas
4.1.3 Programmable control of biodegradable property via structural transition
The composite films formed from the silk fibroin and the nanoparticles are soluble in water because of the random coil structures, which make the composites generator useless within 10 min. This may be quite meaningful when the powering devices need to disappear with the implanted electronic devices when no longer needed. But, it is essential to develop the nanogenerators that can disappear in a controlled and programmable way. For example, the medical implants, that are needed for a few weeks and then disappear, may need the nanogenerators powering to the implanted devices during the same periods, without requiring an extra surgery to remove them from the body.
The optical images show that 30 wt% glycerol reduce the solubility of the composite film in water up to 2 days. The change in the output voltage with measuring time for the 0 wt% (b) and 30 wt% (c) glycerol composite films.
Glycerol has been used to control the biodegradable properties of the silk fibroin film. That is, it can accelerate silk gelation, making the silk fibroin film insoluble in water due to high content of β-sheet.
When the glycerol content in the blend films increases to 5 wt%, the composite films are completely dissolved after dipping in the water for 30 min. Further increases in glycerol to 20 wt% and more concentrations significantly reduced the solubility up to 48 hrs, as shown in the inset of the Figure 10a, and there is no significant change in the output voltage of the composite nanogenerators, as shown in Figure 10c. The result indicates that 20 wt% glycerol was a critical concentration in inducing the structural transition from random coil to β-sheet which is water-insoluble [88]. These results show that it is possible to disappear the composite films in a controlled way for transient electronics.
4.1.4 Self-powered wire type sensor
A sensor was also made as a form of a one-dimensional (1D) wire, and the output power was measured in a same fashion. It was done by dip-coating silk fibroin-based composite-type solution onto an aluminum wire and composite layer is cured at 50oC. Then semi-cured composite wire was taped with Al foils as electrode. In Figures 11a and 11b, the wire consists of Al wire (diameter 1 mm) at the inner, silk fibroin composite (~20 mm), and Al sheet (~ 20 mm) at the outer. The nanogenerator shows the output voltage of ~ 1.8 V and the output current density of ~ 0.1 mA/cm2 in Figure 11c, showing a similar output performance of thin film type nanogenerator, as shown in Figure 6. This will provide the feasibility of this 1D wire as a promising path toward the development of a robust textile nanogenerator for the future smart clothing industry.
Figure 11. (a) A schematic image and (b) a photograph of silk fibroin-based composite-type wire generator which consists of Al wire, silk fibroin composite layer and aluminum electrode. (c) The output voltage and current density of the wire generator.
4.2 Highly stretchable self-powered 2D fabric motion sensor 4.2.1 1D fiber motion sensor
The schematic diagrams of the fabrication process of fiber- and fabric-structured triboelectric nanogenerator are shown in Figure 12 and detail information described in Experimental section.
Basically, the fiber-structured nanogenerator is a coaxial and fully sealed device containing an Al wire
Figure 12. Schematic diagrams of the fabrication process for the FTENG. The fiber-structured nanogenerator is a coaxial and fully sealed device containing an Al wire in the core and a PDMS tube in the shell. ZnO nanowires were first grown on the Al wire by a hydrothermal process, followed by the deposition of Au thin film, producing three-dimensional Au/Al branched wires. The PDMS was etched away by using reactive ions etching process, producing the nanowire arrays with length of 1 - 2 mm inside the tube. The Au/Al wires were inserted into the PDMS tube and the resultant fiber was wrapped around the Al foil. FTENG were then fabricated by weaving the fibers, followed by bonding to a waterproof fabric. Scale bars in SEM images are 1 μm (top) and 2 μm (bottom).
Figure 13.(a) The output voltage and current of fiber-structured TENG. (b) The calculated electrostatic potential with radius of Al wire and PDMS tube simulated by the COMSOL multi-physics software.
The output voltages measured experimentally are also drawn as dotted lines. The change in (a) output voltage and (b) current of the unsealed and sealed fiber-structured TENG at relative humidity of 95 % RH.
The output voltage and the output current of the fiber-structured TENG of approximately 5 cm in length were measured under cycled compressive force around 50 N at an applied frequency of 10 Hz as shown in Figure 13a. The fiber-structured TENG shows a typical AC-type electrical output performance of approximately 40 V and 10 µA, respectively. And we also measured output performance of fiber without nanowires as shown in Figure 14. The output performance in the triboelectric nanogenerator is strongly dependent on the gap distance and the contact area between two materials, thus, on the radius of the PDMS tube and the Al wire. The COMSOL simulation in Figure 13b shows that the triboelectric potential increases with the radius of both materials and there is a maximum potential at 6 mm and 2.0 mm in the radius of tube and wire, respectively. This is evident by the change in output voltages measured experimentally, drawn as dotted lines in Figures 13b and 15. However, here, we used the PDMS tube with a radius of 2.5 mm and the Al wire with a radius of 0.1 mm to increase flexibility of the fiber, an important parameter for flexible practical wearable applications. The fiber was then sealed at both edges, as shown in the inset of Figure 13b. Through the homemade apparatus operating a stable harsh environment (Figure 16), it is clearly seen that the sealed device shows an enhanced performance under extreme conditions of high humidity (> 95 % RH), while the output signals in unsealed device are degraded within approximately 1 min in Figure 13c.
The charge generation of the fiber-structured nanogenerator can be understood by the coupling effect between triboelectrification and electrostatic induction in Figure 17. At the original state, the PDMS tube and the Al wire are fully separated by sealing both edges, in Figure 17a, and there is no charge transfer between them. When the PDMS tube is in fully contact with the wire, a charge transfer is brought about, resulting in positive charges on the wire and negative ones on the PDMS (Figure 17b).
When the external force is receded, the negative charges and the positive charges cannot be compensated. As a result, the positive charges are induced on the outer Al electrode attached onto the tube, by the flow of electrons from the outer Al electrode to the Al wire through the external circuit (Figure 17c). As the PDMS tube reverts back to the original position, the negative charges on the PDMS tube can be fully screened, inducing an equal amount charges on the Al electrodes (Figure 17d).
Subsequently, the external impact is applied once again, the PDMS tube is reverted to the Au/Al wires.
The induction of positive triboelectric charges on the Al electrode decreases. Consequently, the free electrons flow in a reversed direction from the Au/Al wire to Al electrode in order to eliminate the difference of electric potential (Figure 17e).
Figure 15.The output voltage (a) and the output current (b) as a function of radius of metal wire and PDMS tube. The output voltage (c) and the output current (d) as a function of radius of PDMS tube was also plotted at a 0.1 mm of Al wire radius.
Figure 16. The schematic image of homemade apparatus in order to measure the output performance under harsh environment.
4.2.2 Output performance in 2D Fabric motion sensor
Woven fabrics are then produced by the interlacing of the fibers and bonded to a waterproof fabric for all-weather use for the FTENG under vertical contact-separation mode, as shown in Figure 18a. When the electrical output performance was measured in parallel connection under the same mechanical force, the nanogenerator can generate an output voltage and current up to 40 V and 210 µA. However, the current are shown as asymmetric signal indicating leakage current. The reason of asymmetric output signal is the difference of pushing and releasing velocity. However, integration over time, namely the amount of induced charges is almost same, indicating no leakage current as shown in Figure 19. The current increases as the number of the fibers increases, as shown in Figure 18b. To investigate the output power of the nanogenerator, resistors were used as external loads from 1 Ω to 1 G Ω, the instantaneous power of the external resistance reaches a peak value of 4 mW at a resistance of 10 MΩ as shown in Figures 18c and 18d.
Figure 18.(a) Photograph and schematic image of FTENG. (b) The output current of the FTENG with the active area. (c) The output current and voltage, and (d) the instantaneous power of the FTENG as a function of resistance from 1 Ω to 1 GΩ.
Figure 19.The output current signal of fiber of one cycle. Generated charges are nearly same when vertical stress is applied and released.
In the FTENG, which is a two-dimensional network with the fibers, the fibers may be pressed together to produce a smooth-surfaced, stiff, dense material during the weaving process. This implies that the output signals from the FTENG depend on the position (I, II, and III) in the fabric, as shown in Figure 20a. Each position has flat fibers, curved fibers, and crossed fibers, respectively. Figure 20b shows the output voltage and the output current of the FTENG from the position (of area 0.5 cm × 0.5 cm). From the position I and II, an output current of about 3 μA was generated although the output current in position II is a little larger. The output current is the highest (~6 μA) in position III, almost 2 times enhancement of the output current from flat fibers. This indicates that the output power generation from each fiber in the FTENG is quite uniform and there is no significant influence on the output power of the fibers although the fibers may be pressed a little bit by the weaving process, showing the possibility of large-area nanogenerator based on the woven fibers without any degradation. Actually, 27 green
Figure 20.(a) Schematic image of FTENG composed of 6 ´6 fibers. The fibers may be pressed together to produce a smooth-surfaced, stiff, dense material during the weaving process, driving morphological change of the fibers. (b) The output voltage and output current from the fibers in each position. The photographs of (c) a foot step on the FTENG and (d) 27 commercial green LED lit up during a walking.
4.2.3 Strain sensing properties
Knowing the stretch and recovery characteristics of the fabric is also a very important first step in determining the possible applications. Figure 21a shows the photographs of stretching test with a length increase of over 25 %. It shows that the fabric can almost perfectly recover its original shape after release of the tensile force. This is clearly seen in the movie of Figure 22a. It means that plastic deformation does not occur in fabric, and elastic deformation and recovery occur during the process.
When it was stretched up to the strain of about 22 % along the both sides, the output current decreases by about 47.6 %, from 210 μA to 110 μA in Figure 21b. We ascribe this to the decrease in the radius of the PDMS tube induced by the stretching of the fiber. Actually, we showed the output power decreased as the radius of tube decreases in Figures 13b and 15. However, when it was released, the output power reversibly recovered its original value, showing a stable electrical output performance. The FTENG is then integrated to coats for power generating arm patch, becoming a power cloth. The generated current from nanogenerator can reach 1 μA when the FTENG bends at 60 o. When the bending angle increases to 130 o, the current increases to 7 μA, lighting up the 1 green LED during motion of the elbow in Figures 21d and 23
Figure 22.(a) The video shows a uniaxial tensile test of the FTENG over 25 %. (b) The change in output current as a function of the strain when it is stretched and released.
Figure 23. (a) The photographs of FTENG composed of 30 ´30 fibers for power generating human arm patch and a green LED lit up by the motion. (b) The generated current from the FTENG when the FTENG is bent at an angle of 135 o.
4.3 Highly stretchable self-powered sensor based on Au NPs capped with DMAP