3.1. Theoretical background of capacitive photodetector
The increase in the dielectric permittivity of the semiconductor with particle-polymer composites upon irradiation of light with a specific wavelength changes the capacitance of the composite sandwiched between the two facing electrodes. Two physical theories of “interfacial polarization” and the “Maxwell-Wagner-Sillars (MWS) effect” at particle–polymer interfaces can explain these changes in the dielectric response of the composites upon light irradiation32–34. The MWS effect is known to occur in the heterojunction structures owing to the accumulation of charges at the interfaces as well as the formation of dipoles on particles or clusters. In the heterostructures, additional charge accumulations occur at the particle boundaries under an external electric field owing to the differences in dielectric characteristics (ability to hold charges) and electrical conductivity35. It is well known that the huge difference in the conductivity of the particles and polymers will lead to a large charge accumulation at the interface because the current cannot flow freely across it. The accumulated charges are directly proportional to the difference between the two conductivities, and the interface behaves like a nanocapacitor33. Thus, the assembled movements of the interfacial dipoles at the interfaces amplify the response to the incoming electromagnetic field and thus enhance the microwave absorbing capability. Based on these theories, in this study, a composite film was fabricated by dispersing ZnS:Cu particles in polyvinyl butyral (PVB), a flexible free-standing polymer. The fabricated devices were mechanically flexible and able to withstand bending with a curvature radius of several millimeters.
3.2. Fabrication of AgNW/graphene capacitive photodetector
Figure 1 shows a schematic of the fabrication process for the AgNW/graphene-based capacitive photodetector. As can be seen in Fig. 1, the AgNW electrodes for the bottom surface of the composite film are embedded and are stable under external stimuli, whereas the AgNW electrodes on the top surface are vulnerable to the external environment because they are simply spin coated on the top. For this reason, stable bonding with the structure at the bottom is not achieved, and thus may fall off from surface. Therefore, the CVD-grown graphene monolayer was wet-transferred to the top of the capacitive photodetector. By transferring the graphene to the top, it is possible to increase the thermal and chemical durability of the entire device as well as the top AgNW electrodes. In addition, graphene has high flexibility; therefore, it can be advantage in the characteristics of the manufactured flexible capacitive photodetector.
Figure 2 shows SEM images of a cross-section of the photodetector as well as the upper and lower AgNW electrodes. ZnS/Cu particles with an average diameter of ~ 20 µm were dispersed in polyvinyl butyral (PVB), a flexible free-standing polymer, as shown in Fig. 2a. Because the particles are uniformly dispersed in flexible polymers, the fabricated photodetector is mechanically flexible and able to withstand bending with a curvature radius of several millimeters. Here, because the specific surface area of the spherical ZnS/Cu particles was much smaller than those of most nanosized particles (nanoparticles), aggregation between the particles was not observed. Figures 2c and 2d show the top and bottom AgNW electrodes, respectively. The morphologies of the AgNWs are quite different: The top AgNWs are simply spin-coated on top of the ZnS:Cu-PVB, whereas the bottom AgNWs are embedded into ZnS:Cu-PVB. Therefore, the bottom AgNWs featured a significantly flattened shape, and the SEM image of the bottom AgNWs seemed to be slightly blurred. By contrast, the top AgNWs coated on the ZnS:Cu-PVB surface did not appear to be firmly adhered; therefore, the graphene buffer layer was wet-transferred to the top of the AgNWs.
By transferring graphene to the top of the AgNWs, adhesion of the upper AgNWs can be increased, and an additional current path can be formed. The average sheet resistances of the AgNW electrodes obtained at the ZnS:Cu ratios of 0.3, 0.5, 0.7, and 0.9 wt% in PVB were 8.516, 9.09, 10.864, and 12.076 Ω/sq, respectively, as shown in Fig. 2e. After the introduction of graphene, the average sheet resistances were 7.356, 7.676, 8.042, and 10.34 Ω/sq, respectively, which are approximately 25% lower than those of the AgNW electrodes.
Here, for both AgNW electrodes and AgNW/graphene hybrid electrodes, the sheet resistance of the substrate increased as the ratio of ZnS:Cu particles in PVB increased. As the number of particles increases, the density of particles inside the thin polymer layer increases. This increases the roughness of the polymer layer, resulting in an increase in resistance.
Compared to those of the AgNW electrodes, the sheet resistance of the AgNW/graphene hybrid electrodes was lower, which suggests that the AgNW/graphene hybrid electrodes delivered a better electrical performance than the AgNW electrodes. This is because, as the AgNW electrodes feature a network form, there are void spaces where the electrodes are not present, which eventually leads to an open-circuit fault. In addition, AgNWs are known to have a large surface roughness, which can easily penetrate the thin film layers. For the AgNW/graphene hybrid electrodes, the CVD graphene layer covered the entire AgNW electrodes, and the current flowed throughout all non-empty spaces. In addition, covering the CVD graphene layer has the advantage of solving the problem of the large surface roughness of the AgNWs.
3.3. Optical characteristics of AgNW/graphene capacitive photodetector
To investigate the optical properties of the AgNW/graphene capacitive photodetector, the transmittance and reflectance were measured depending on the ZnS:Cu particle ratio in PVB (Fig. 3a and b). Overall, the transmittance decreases as the ZnS:Cu particle ratio increases because the embedded density of the particles also increases. Thus, devices with a ZnS:Cu loading ratio of 0.3 featured a high transmittance with a low reflectance over the entire visible range. For the AgNW electrode-based photodetector, the maximum transmittances obtained at ZnS:Cu ratios of 0.3, 0.5, 0.7, and 0.9 wt% were 54.8%, 34.0%, 25.3%, and 17.8%, respectively. For the AgNW/graphene hybrid electrode-based photodetector, the maximum transmittances were 45.9%, 31.8%, 22.5%, and 17.2%, respectively. On average, the photodetectors using the AgNW/graphene hybrid electrode exhibit a 4% lower transmittance than the photodetectors using the AgNW electrodes. It can be inferred that this is based on the light absorption rate of graphene, which is known to be 2.3% on average.
The measured transmittance of the photodetector can be evaluated as relatively low, and this is because the photodetectors were manufactured using 1.0 wt% of AgNWs. As the concentration of AgNWs increases, the overall device transmittance decreases, and the sensitivity decreases29. From this perspective, the AgNWs concentration should be lowered to increase the transmittance and sensitivity; however, in this study, the device was fabricated using a 1.0 wt% dispersion, which represents a relatively high density of AgNWs, to enhance the flexibility of the device. It is well known that the resistance of the AgNW electrodes is increased via repeated bending owing to the decreased number of conduction paths made by AgNWs.33 When a photodetector is manufactured with low-concentration-based AgNWs, the resistance increases rapidly upon mechanical stress, and an excessively large increase in resistance hinders the transfer of charges from the power source to the dielectric, which can cause the capacitance to decrease. In conclusion, photodetectors fabricated with a 1.0 wt% AgNWs dispersion with graphene exhibited a transmittance of approximately 45% of incident light with a 550-nm wavelength.
3.4. Photoswitching characteristics of AgNW/graphene capacitive photodetector
Based on the calculated ZnS:Cu-PVB composite bandgap energy from our previous study30, which is 2.95 eV regardless of the ZnS:Cu content, a lighting source with a wavelength of 420 nm and a power of 1.2 mW/cm2 was employed to test the photoresponsivity of the photodetectors. The photoresponsive capacitance was measured under ambient conditions using a two-probe method and a 50-kHz signal. The ZnS:Cu particle ratio-dependent photoresponsivities of the AgNW-based photodetector and that of the AgNW/graphene-based photodetector are shown in Figs. 4a and 4b, respectively. For the AgNW-based photodetector, photoresponsivities enhanced by 1.1-, 1.25-, 1.36-, and 1.47-fold are shown compared to those at the off state at ZnS:Cu particle ratios of 0.3, 0.5, 0.7, and 0.9, respectively. Similar to the AgNW/graphene-based photodetector, photoresponsivities enhanced by 1.1-, 1.24-, 1.35-, and 1.47-fold are shown at ratios of 0.3, 0.5, 0.7, and 0.9, respectively. The photoresponsivities revealed that the capacitance of the composite layer showed a higher photosensitivity with a higher particle density, which indicates the existence of a trade-off between the transparency and sensitivity of the fabricated photodetectors. So, to have the best trade-off condition, the only ratio of 0.7 is measured after this measurement of Fig. 4a and 4b.
To measure the photoresponsivities of the fabricated photodetectors, capacitances at different wavelengths were also observed, as shown in Figs. 4c and 4d. For both AgNWs and AgNW/graphene-based photodetectors, the capacitance did not effectively change upon irradiation by light with a wavelength of 550 or 650 nm, which shows that the implemented photodetectors were highly photosensitive. The photodetector reacted only at a wavelength of 420 nm because the band gap of the ZnS:Cu particle was 2.93 eV. The band gap of ZnS:Cu particles was calculated based on the Tauc plot equation, the details of which are described in our previous study.31
As mentioned earlier, as the thickness of the electrodes increases, the overall device transmittance decreases, and the sensitivity decreases. This is because as the thickness of the electrode increases, the light transmitted through the ZnS:Cu-PVB composite film decreases, and the amount of charge to be transferred decreases. In the case of the AgNW/graphene photodetector, compared to the AgNW electrode-based photodetector, because the thickness of the electrode becomes thicker owing to the additional graphene electrode, it can be predicted that the photoresponsivity will be relatively lower. In addition, the result indicates that the transmittance of the AgNW/graphene photodetector is approximately 4% lower than that of the AgNW electrode-based photodetector. Nevertheless, it can be seen that the measured photoresponsivity of both photodetectors shows little difference. It can be confirmed that the lowered transmittance, which is a disadvantage caused by the additional use of the graphene electrode, does not significantly affect the photoresponsivity of the photodetectors.
3.5. Durability of AgNW/graphene capacitive photodetector
The fabricated AgNW/graphene-based photodetectors exhibited a superior mechanical flexibility, which is essential for emerging flexible optoelectronic devices. A repetitive bending test was conducted to investigate their mechanical durability. During the test, photodetectors based on AgNW and AgNW/graphene electrodes were rolled at various bending radii (rb) or rolled and subsequently unrolled at a bending radius of 5 mm at a speed of 3 mm s− 1. The sheet resistance of each photodetector was compared to its initial value (ΔR/R0, where ΔR is the change in resistance after bending, and R0 is the initial resistance). The tensile strain applied to the bent substrate was estimated using36
where dsubstrate and delectrode are the thicknesses of the substrate and electrodes, respectively, and Rc is the radius of the curvature.
In Fig. 5a, the mechanical flexibilities of the photodetectors based on AgNW and AgNW/graphene electrodes are compared at various bending radii (rb). The AgNW-based photodetector exhibited changes in resistance of more than 100% at above rb = 6 mm and finally reached 700% of the initial resistance. This is because the polyvinyl pyrrolidone (PVP) on the surface of the nanowires generated during the manufacturing process weakens the adhesion to the substrate37, and as the bending radius decreases, the AgNWs at the top of the composite layer fall off. By contrast, there was little change in the resistance of the AgNW/graphene photodetector. Figure 5b compares the mechanical flexibilities of the photodetectors based on AgNWs and AgNW/graphene as functions of the number of wrapping cycles. The resistance of the AgNW/graphene-based photodetector was slightly changed after 1,000 bending cycles, whereas that of the AgNW-based photodetector was increased by more than 100% after 1,000 bending cycles and by 200% after 2,000 bending cycles. As shown in Fig. 5c, the photoswitching characteristics of the photodetectors after the mechanical test were measured. After the bending test, the sensitivity of the AgNW-based photodetector was significantly lower than that of the AgNW/graphene photodetector, despite the photoswitching curve remaining stable. It can be expected that, after the bending test, the AgNWs were detached from the composite film owing to the low adhesion of the AgNWs. When the AgNWs were detached, the void space of the electrode of the network type increased, and thus, the amount of charge accumulated in the composite film decreased, thereby reducing the overall sensitivity. However, in the case of the AgNW/graphene photodetector, the monolayer graphene film enhanced the adhesion of the AgNW electrodes to the composite film, and thus, the sensitivity of the photodetector was high.
To further evaluate whether the graphene layer acted as a protective layer against the AgNW electrodes, various solutions with pH values ranging from 2 to 12 were prepared. The AgNW and AgNW/graphene electrode-based photodetectors were then covered at each pH for 20 min. As shown in Fig. 6a, the resistances of the AgNW electrode-based photodetector immersed in the solutions with a pH of 6 to 12 were not significantly different, whereas, at a pH of 2 and 4, the changes in resistance were approximately 2.2% and 3.9%, respectively. The resistance of the AgNW/graphene photodetector was not considerably increased over the entire pH range of 2–12. The unique honeycomb lattice of the graphene acts as a protective barrier that prevents the penetration of small molecules; therefore, the AgNWs were able to maintain the properties of the electrode without oxidation. The AgNW/graphene hybrid electrode-based photodetector exhibited chemical stability. Figure 6b shows the photoswitching characteristics of the photodetectors. The significant gap in the change in photosensitivity between the AgNW and AgNW/graphene electrode-based photodetectors also verifies that graphene has high chemical stability, thereby improving the chemical durability of the overall photodetector.
Finally, to further evaluate the thermal stability of the AgNW/graphene-based photodetector through the thermal stability of the graphene protection layer, the photodetectors based on AgNWs and AgNW/graphene were annealed on a hot plate at 160°C for 120 h. The variations in resistance were measured every 24 h. As shown in Fig. 7a, the resistance of the AgNW electrode-based photodetector rapidly increased after 72 h and then reached a value that was almost 400% higher than the initial resistance. The AgNWs began to melt as high temperatures were applied, and the overlapping parts became welded. However, when a continuous high temperature was applied, dewetting occurred, and the shape of the nanowires was not maintained; they aggregated, forming round silver nanoparticles. Therefore, after the AgNWs melted at 160°C, the contacts between the AgNWs were initially welded, but after 72 h, the part in contact was cut off, and the overall resistance increased. By contrast, the AgNW/graphene hybrid electrode-based photodetector exhibited a slightly increased resistance, which demonstrates its superior thermal stability to that of the pristine AgNW electrode-based photodetector. Although both photodetectors exhibited a relatively stable photoswitching curve, as shown in Fig. 7b, the AgNW electrode-based photodetector showed a sensitivity decrease of approximately 20% compared to that of the AgNW/graphene hybrid electrode-based photodetector.
Overall, the fabricated AgNW/graphene hybrid electrode-based photodetector exhibited an excellent performance, compared to the AgNW electrode-based photodetector, and presented stable photoswitching responsivity without being affected by the surrounding environment, that is, it was unaffected by chemicals or temperature changes. The overall results suggest that the AgNW/graphene hybrid electrode-based photodetector is expected to be used in various next-generation optoelectronic devices.