Zinc oxide (ZnO) is a direct transition semiconductor with wide bandgap of 3.37 eV and superior optical and physical features due to its high exciton binding energy of 60 meV at room temperature [1–3]. Because of its high on/off ratio, fast optical response speed, high absorption rate, physical and chemical stability against ultraviolet (UV) rays, ZnO has been actively investigated in the field of optoelectronics and for sensor applications [4,5]. ZnO can also be found in a variety of morphologies, such as nanowires, nanobelts, nanorods, and tetrapods [6–13].

Especially, Gröttrup et al. [14] and Mishra and Adelung [15] have developed light and gas detection devices with high sensitivity using ZnO tetrapods synthesized by flame synthetic methodology. A ZnO tetrapod is a three-dimensional structure comprising four nanowires; this tetrapod exhibits significantly enlarged surface area compared to a one-dimensional configuration. Thus, the ZnO tetrapod enable enhanced applications in electronic and sensor devices due to the facile interconnection and increased surface area with respect to the nanowire structure, respectively. For instance, the interconnections within the tetrapod arms facilitate a continuous path for carrier migration enabling rapid response in electronic and sensor applications. Previously, using micro-sized Zn powder under an atmospheric pressure microwave plasma (APMP) jet system, we have synthesized ZnO nanoparticles in the morphology of wires, rods, and tetrapods. As for the synthesis mechanism of ZnO tetrapod, the high energy environment of the atmospheric pressure plasma jet was responsible for rapid melting of micro-sized Zn powders and regrowth to tetrapod configuration. Next, reactive oxygen species in the plasma environment are adsorbed on Zn core and ZnO nuclei are finally formed with the tetrahedral structure. When ZnO nucleation occurs, an octahedral pyramidal crystal is formed with the eight facets of the octahedron consisting of alternating (0001) surfaces (+c direction) and (000−1) surfaces (−c direction). The positively charged surface (+c) was likely to be terminated by Zn atoms, resulting in the growth of legs along the four [0001] directions to form a wurtzite structure [14–18]. The dependence of the size on the photosensing characteristics were also evaluated based on the collection site and power applied in the APMP system [19].

UV detection devices should have not only fast photoresponse speed but also flexibility for portable uses, such as rollable, foldable, and wearable forms. Among the various polymers, polyimide (PI), owing to its excellent mechanical strength and high glass transition temperature, has been extensively investigated for use in flexible templates [20–22]. Spray coating is one of the most promising techniques for flexible device fabrication owing to its high production speed, compatibility with various substrates, and large-area fabrication. Furthermore, because it uses a designed coating sequence on various substrates, the spray coating process allows the fabrication of multilayered functional films with high thickness uniformity [23–26]. For example, Girma et al. [26] reported the preparation of a flexible sensor by spray coating without any vacuum system and demonstrated large-area fabrication at low-cost.

In this study, we demonstrate the synthesis of ZnO tetrapods and their availability as flexible photosensors through simple spray coating on PI films. Based on the high stability and reliability of the devices, we could suggest a facile strategy for developing flexible photosensors using ZnO nanomaterials on flexible polymeric substrates.

2.1. Synthesis of ZnO tetrapods using APMP system

To synthesize ZnO tetrapods, an APMP jet system ignited using a mixture of O₂ and N₂ (5 N purity, 50:50 vol.%) with a flow rate of 10 L/min. The frequency and applied plasma power were 2.45 GHz and 1,200 W, respectively. A spherical Zn powder as a starting material (Sigma Aldrich, 95 % purity, 10 µm diameter) was continuously introduced in front of the plasma plume. After the plasma process, we collected the deposited materials at a region 10 cm behind region from the Zn inlet, as shown in Fig. 1(a).

Figure 1. (a) Schematic illustration of APMP jet system. SEM images of (b) Zn powder and (c) ZnO tetrapods synthesized at plasma power of 1,200 W. (d) TEM image of ZnO tetrapod. Inset shows electron diffraction pattern showing high crystallinity of the ZnO tetrapod. (e) High resolution TEM image showing interplanar distance of 0.52 nm along the c-axis of ZnO tetrapod. Inset is a corresponding FFT pattern.

2.2. Fabrication of flexible photosensors and characterization

Flexible photosensors were fabricated by spray coating of ZnO tetrapods on PI (10 × 10 mm²) films. The synthesized ZnO tetrapods were dispersed in deionized (DI) water (20 mg/mL) and briefly ultrasonicated for dispersion. Then, the solution was sprayed 40 times onto the PI films with a spray gun having a nozzle diameter of 0.3 mm. The PI films were installed on a hot plate (80 °C) to dry during the spaying and distance between the PI and the nozzle was set to 20 cm. Au electrodes (500 nm thickness and 3 mm spacing) were made to fabricate the flexible photosensing devices.

Size and structural characterization of the materials were performed using scanning electron microscopy (SEM) (Hitachi, S-4800) and transmission electron microscopy (TEM) (JEOL, JEM-2100F). To confirm the purity and crystallinity of the synthesized ZnO tetrapods, X-ray diffraction (XRD) (PANalytic, X’Pert Pro) with Cu Ka radiation was used. Raman spectroscopy (Horiba, ARAMIS) was employed to investigate vibration modes of ZnO using a laser of 532 nm wavelength. To evaluate the transparency of the PI films before and after spray coating, a UV-visible (UV-vis) spectrophotometer (Biochrom Ltd., Libra S80) was used. Samples were irradiated with UV light (300−350 nm, 100 mW/cm²) and resultant photocurrent was measured using a source measurement unit (Keithley 2400). The distance between the UV source and PI film was set to 20 mm.

Figure 1(b) shows spherical Zn powders having average diameters of 10 µm and ZnO tetrapods synthesized by APMP system are shown in Fig. 1(c). Sizes of ZnO tetrapods were around 500 nm in size. Here, the size is defined as the distance between the two arms of a tetrapod. Previously, we reported size control of ZnO tetrapod by variation of applied plasma power: higher plasma power enhances both the dissociation of feedstock gases and the plasma temperature. Thus, the phase transformation from the Zn microspheres to the nanosized ZnO tetrapods can be accelerated in a high plasma power environment. Microstructure and selected area electron diffraction pattern were analyzed using a high resolution TEM, with results shown in Fig. 1(d). The results show that the thickness of a single arm was 80−120 nm and the ZnO tetrapods had high crystallinity. Figure 1(e) clearly reveals the lattice fringes of the (001) planes with an interplanar distance of 0.52 nm along the c-axis. The inset in Fig. 1(e), the corresponding fast Fourier transform (FFT) pattern, shows good similarity to the TEM image.

The crystal structure and crystallinity of the synthesized ZnO were determined using XRD and Raman spectroscopy. As shown in Fig. 2(a), the main diffraction peaks at 31.8, 34.5, and 36.4° correspond to the (100), (002), and (101) planes, respectively, of the hexagonal wurtzite phase of ZnO. The weak peaks were also well matched with ZnO (JCPDS 79-0205). We can expect that the synthesis yield of ZnO tetrapods was very high because Zn peaks were hard to observe. Figure 2(b) shows the Raman profile of the ZnO tetrapods with E_2L, E_2H, and E_2H-E_2L vibration modes at 99, 436, and 337 cm⁻¹, respectively. These peaks are associated with vibration between Zn and O in the ZnO lattice; strong peaks imply high crystallinity of the ZnO. The peak at 520 cm⁻¹ is caused by the silicon oxide substrate.

Figure 2. (a) XRD and (b) Raman spectra of ZnO tetrapods.

To fabricate a flexible photosensing device and evaluate the UV sensing characteristics, the ZnO tetrapods dispersed in DI water were spray deposited on PI films and photocurrent during UV light irradiation was measured. Figure 3(a) shows digital photos and UV-vis spectra from before and after ZnO spray coating. The transmittance of PI at 550 nm wavelength decreased from 49 to 18 % due to ZnO coating. This means that a flexible device with moderate transparency and excellent photosensing properties was realized using a simple spray coating process. Figure 3(b) shows the representative photoresponse curve after with intermittent exposure of UV ray. When the device was exposed to repetitive UV irradiation with every 10 s, a stable photocurrent of 29 nA was measured. The sharp increase and drop in photocurrent show the quick and stable photoresponse of the device.

Figure 3. (a) Digital photos and UV-vis spectra obtained before and after spraying of ZnO tetrapods on PI-films. (b) UV photoresponse curve measured from the PI-based device. A digital photo showing UV sensing system is inserted as inset.

To evaluate device reliability, we measured photocurrent at different levels of UV light intensity. Figure 4(a) shows a stepwise variation of device photocurrent with increase of UV light intensity. When the UV light intensity was 5 mW/cm², a photocurrent of 7.13 nA was measured. Then, the photocurrent increased to 15.8, 20.5, and 27.2 nA as a result of gradual increase of UV light intensity to 10, 25, and 100 mW/cm², respectively. In addition, stable and constant photocurrents were obtained with repetitive UV irradiation using sequential levels of UV lamp power, as presented in the inset of Fig. 4(a). Stable photocurrents of 9.0, 15.6, 24.6, and 26.2 nA were obtained by the repetitive UV exposure with set specific power levels of 5, 10, 50, and 100 mW/cm².

Figure 4. (a) Variation of photocurrent with different levels of UV lamp power. Inset shows change of photocurrent as function of cyclic change of UV light intensity. (b) Stable photocurrent shows high reliability of PI-based flexible device. Inset shows stable values of sensitivity over a period of two weeks.

Figure 4(b) shows the long term stability of the device. We obtained stable photocurrent of around 25 nA with cyclic UV exposure at 10 s on/off intervals over a period of 60 min. Moreover, we confirmed the high device reliability by measuring the device sensitivity for 14 days after device fabrication, as shown in inset of Fig. 4(b). Here, the sensitivity value is defined as the ratio of the photocurrent during the on/off states (S = I_on/I_off) of the UV light. Based on these results, we can conclude that the device made by spray coating of ZnO tetrapods on PI films shows high reliability and durability.

Finally, we investigated the UV sensing ability with respect to the strain (ε) of the substrate by convex and concave bending. We defined the strain of the PI substrate as the percentage of length change (l_o - l_f) of the device before (l_o) and after (l_f) the bending; thus, ε = [(l_o - l_f) / l_o] × 100. First, we prepared a PI-based device with 10 × 10 mm² size and gradually increased the strain with convex shape as shown in Fig. 5(a). The final length of the device was set to 7 mm and thus strain was 30 %. We found that the photocurrent gradually increased from 15.9 to 20.7 nA under strains of 5 and 20 %, respectively. Similar values of photocurrent were measured for the range of 15−25 % strain. In the case of 30 % strain, a relatively lower photocurrent of 19.1 nA was obtained. The change of the sensitivity was the same as that of the photocurrent. On the other hand, the photocurrent was not significantly changed in the case of concave strain as shown in Fig. 5(b). Although the absolute values of photocurrent between convex and concave samples were quite different, differences can be attributed to differences between individual samples. It is more important to consider the variation of photocurrent on the applied strain. The increase of photocurrent in the convex-shape situation can be attributed to the enlarged area of ZnO exposed to UV light. The decreased photocurrent in the 30 % strained sample could plausibly be the result of slippage or partial detachment of ZnO tetrapods on PI films. In the concave-shape situation, a reduced area of UV exposure compared to the flat case may have resulted in the gradual decrease in photocurrent. At present, to understand the strain dependence clearly and apply results to real and flexible situations, more work is needed using different device lengths and repetitive bending and release tests. However, we believe that our results propose a facile methodology to fabricate flexible photosensors.

Figure 5. Digital photos showing flexibility, variation of photocurrents with respect
to periodic UV illumination, and change of sensitivity on different bending for (a)
convex and (b) concave situations.

In summary, we demonstrated not only the synthesis of ZnO tetrapods using an APMP jet system but also the facile fabrication of flexible photodetectors. High quality ZnO tetrapods were synthesized using applied plasma power of 1,200 W and mixing gases of oxygen and nitrogen. The ZnO tetrapods were dispersed in DI water and subsequently spray coated on PI films. It was found that highly stable, reliable, and long-lasting photosensing ability can be achieved through this facile methodology. Finally, although more work is required, it can be concluded that ZnO-based PI films can be applied as flexible photosensors even in convex and concave shapes.

The authors declare no conflicts of interest.