Min-A Kanga,
Seongjun Kima,
In-Su Jeona,
Yi Rang Lima,
Chong-Yun Parkb,
Wooseok Songa,
Sun Sook Leea,
Jongsun Lima,
Ki-Seok Ana and
Sung Myung*a
aThin Film Materials Research Center, Korea Research Institute of Chemical Technology, Daejeon 305-600, Republic of Korea. E-mail: msung@krict.re.kr
bDepartment of Physics, Sungkyunkwan University, Suwon, Gyeonggi-do 440-746, Republic of Korea
First published on 25th June 2019
Two-dimensional (2D) transition metal dichalcogenides (TMDs) such as molybdenum disulfide (MoS2) and tungsten diselenide (WSe2), have recently attracted attention for their applicability as building blocks for fabricating advanced functional materials. In this study, a high quality hybrid material based on 2D TMD nanosheets and ZnO nanopatches was demonstrated. An organic promoter layer was employed for the large-scale growth of the TMD sheet, and atomic layer deposition (ALD) was utilized for the growth of ZnO nanopatches. Photodetectors based on 2D TMD nanosheets and ZnO nanopatches were successfully fabricated and investigated, which showed a high photoresponsivity of 2.7 A/W. Our novel approach is a promising and effective method for the fabrication of photodetectors with a new structure for application in TMD-based transparent and flexible optoelectronic devices.
In order to observe the surface morphology of MoS2 nanosheets with ZnO nanopatches, scanning electron microscopy (SEM) was employed with regard to the process cycles of ZnO (Fig. 2(a)). MoS2 nanosheets produced by solution-based synthesis exhibited a uniform surface coverage and a continuous film on the SiO2 substrate. The surface morphologies of MoS2 nanosheets with ZnO were then investigated depending on the process cycles. At first, the partially aggregated grains and rough surface were observed, indicating that clearly displays 10-cycle ZnO nanopatches on MoS2 nanosheets over the entire area (Fig. 2(a)(ii)). Formation of granular surface is common for epitaxial ZnO films produced by the ALD process.17,18 When increasing the ZnO film thickness by the process cycles, ZnO nanopatches with island-shape disappear, and flat thin films were formed in the MoS2 nanosheets (Fig. S1†). Moreover, atomic force microscopy (AFM) analysis was carried out to confirm the surface roughness of MoS2 nanosheets with ZnO nanopatches (Fig. 2(b)). These results were similar to surface morphologies identified by SEM analysis. Interestingly, surface roughness of MoS2 nanosheets with ZnO nanopatches at 10 cycles was the highest (∼0.57 nm) due to the presence of numerous ZnO grains at the initial deposition stage (Fig. 2(c)). When the process cycles for ZnO growth were increased, the RMS roughness decreased. This phenomenon could be understood by the formation of ZnO film at the increased process cycles. In order to provide the exact height profiles of as-deposited ZnO, XRR (X-ray specular reflectivity) was employed to confirm the thickness of thin film more accurately (Fig. S2†). In the case of 10 cycles-ZnO, the thickness cannot be measured because the thin film was not formed. This result indirectly predicts that 10 cycles-ZnO is formed with nanopatches. Furthermore, we could identify that the thickness becomes increases gradually from several nm to ∼10 nm with increasing the number of cycles and then the ZnO thin film is formed eventually. Fig. 2(d) is a photograph of the large area-based MoS2 nanosheets with ZnO on an 8 inch SiO2 substrate. Notably, the hybrid film fabricated using this approach was uniform over the whole area. Furthermore, MoS2 nanosheets with ZnO were successfully transferred onto the quartz substrate (2 cm × 2 cm) in order to confirm the applicability to the transparent electronic devices (Fig. 2(e)). Optical transmittance characterization was carried out using UV-vis spectroscopy (Fig. 2(f)). Herein, there was a difference in the optical transmittance with and without MoS2. In general, the absorption spectra of ZnO nanoparticles depend on the method of fabrication, shape, and particle size. In this study, the optical transmittance spectra based on ZnO–MoS2 hybrid films showed absorption peaks at 335 nm and 345 nm, and the two absorption peaks gradually increased as the number of ZnO cycles increased. The decrease in transmittance was related to an increase in the thickness of the films and it reveals a relatively low thickness of the ZnO films. Since the relatively thin ZnO was formed using the ALD process and act as a nanopatch, it only contributes to the healing of numerous defect sites, such as imperfections and impurities, in as-synthesized MoS2 film.
In order to clarify the characteristics of chemical bonding in the MoS2 nanosheets with ZnO nanopatch, Raman spectroscopy is employed. There are two representative Raman modes in MoS2.19 The A1g mode is originated from the out-of-plane vibrations of molybdenum atoms, and the E12g mode is responsible for the in-plane vibrations of molybdenum and sulfur atoms. Remarkably, the Raman shift was almost constant regardless of the process cycle of ZnO film (Fig. 3(a)). Furthermore, the frequency difference of A1g mode and E12g dmode was approximately 21.6 cm−1 for the entire process condition (Fig. 3(b)). These results indicate that the structure of as-synthesized MoS2 nanosheets were well-maintained without any structural deformation during the deposition process of ZnO film. X-ray photoelectron spectroscopy (XPS) was also carried out to examine the chemical composition of the MoS2 nanosheets with ZnO nanopatch (Fig. 3(c)). Herein, the Mo 3d, S 2p core level spectra for the MoS2 nanosheets and the Zn 2p, O 1 s core level spectra related to the deposited ZnO were explored for the structural analysis of the MoS2–ZnO hybrid film. For the pristine MoS2 nanosheets, the Mo 3d3/2 and 3d5/2 peaks were located at binding energy (EB) of 232.5 eV and 229.3 eV, respectively. S 2p1/2 and 2p3/2 peaks at EB of 163.3 eV and 162 eV, respectively, were obtained in pristine MoS2 nanosheets, revealing that MoS2 was synthesized successfully using the solution process.14 Furthermore, the intensity of Mo 3d and S 2p core level spectra decreased as the process cycle of ZnO increased. In the Zn 2p core level spectra, Zn 2p1/2 at the higher binding energy and Zn 2p3/2 at the lower binding energy was observed. The difference in binding energy between the Zn 2p3/2 and Zn 2p1/2 was about 23 eV. This value, originated from the spin–orbit splitting of energy level, is consistent with previous reports that demonstrated the existence of ZnO.20 Moreover, the intensity of Zn 2p core level spectra increased as the deposition cycle of ZnO film increased. Typically, XPS is known as a powerful surface analysis tool with surface-sensitive technique. As increasing the number of ZnO cycles, the coverage region of ZnO on the MoS2 surface is increased. So, when the incident X-ray beam is injected on the MoS2–ZnO hybrid structures, the peak intensity and composition ratio could be altered. In other words, the Mo 3d and S 2p peaks become decreased as increasing the number of ZnO cycles. Whereas, the Zn 2p peaks related to ZnO become increased. Atomic concentration of MoS2–ZnO hybrid structure as a function of the number of ZnO cycles was represented in the Fig. S3.† In the O 1s core level spectra, O 1s core level spectrum was located at EB of 529.9 eV, indicating the presence of oxygen atoms in the crystal lattice without any oxygen vacancies. With increasing process cycles of ZnO film, the prominent extra peaks at EB of 531.6 eV increased. This phenomenon indicates the formation of oxygen atoms in the ZnO lattice with oxygen vacancies.
Fig. 4 (a–i) is an optical image of an electronic device based on a MoS2 nanosheet with a ZnO nanopatch. Here, the width and length of the channel were 40 μm, 200 μm, respectively. In order to clarify the electrical properties, output curves were examined depending on the number of ZnO cycles. Thereby, we could confirm the improved electrical properties with increasing the number of ZnO (Fig. S4†). This behavior could be understood by the fact that as-deposited ZnO contribute to heal the numerous defect sites in MoS2. Furthermore, transfer characteristics with back gate configuration were investigated for MoS2 nanosheets with a ZnO nanopatch (process cycle: 10, 20, 30, 40 cycles) under cyclic UV light irradiation (λ = 254 nm) (Fig. 4(a)(ii–vi)). Pristine MoS2 shows p-type characteristics due to the adsorbed oxygen atoms on the thin MoS2 surface in an arbitrary environment. With increasing the process cycles of ZnO films, a lot of photo-excited electrons were transferred to the ZnO film with a higher work-function than MoS2.21 Thus, the threshold voltage of MoS2 nanosheets was shifted to a more positive voltage, and their photocurrent significantly improved due to the high sensitivity of the ZnO film under UV exposure. The dependence of the photocurrent response in the UV region on the process cycle of ZnO at bias voltage of 10 V (Fig. 4(b)) was investigated. The photocurrent increased considerably with increasing process cycles. MoS2 nanosheets with 40 cycle-ZnO experienced the highest photocurrent owing to the unique properties of ZnO.22 The rise and decay time for photodetectors based on MoS2–ZnO heterostructure was calculated in Table S1.† With increasing the number of ZnO cycles, the deep level defect states (DLDS) originated form the sulfur vacancies in MoS2 crystal become disappeared and the recombination rate of photo-excited electron–hole pairs was decreased also. Our photodetectors based on MoS2–ZnO hybrid structures could improve the photoresponsivity as well as heal the several defect sites in MoS2 crystals by introducing ZnO. The relevant working principle of photodetector based on MoS2–ZnO hybrid film was depicted in Fig. S5.† The photocurrent response of MoS2 nanosheets with 40 cycle-ZnO was identified depending on the bias voltage (Fig. 4(c)). At increasing bias voltage, the photocurrent response of the MoS2–ZnO hybrid film increased dramatically and the highest photoresponsivity was estimated to 2.7 A/W at a bias voltage of 40 V. Table S2 in ESI† exhibits the parameters of MoS2 based ultraviolet photodetectors, indicating that performance for our photodetectors based on ZnO–MoS2 hybrid film is more competitive than previous reports. Furthermore, the variation of photocurrent was evaluated to confirm the applicability to flexible optoelectronic devices based on MoS2 nanosheets with ZnO nanopatch (Fig. 4(d)). PMMA-assisted wet transfer method was employed to transfer as-fabricated film to the flexible PET substrates. Fig. 4(d) shows the bending process of the MoS2–ZnO hybrid film. The variations of the photocurrent response of pristine MoS2 and MoS2–ZnO film (bending radius = 3 mm) as a function of the number of bending cycles (up to 10000 cycles) were examined. Remarkably, pristine MoS2 and MoS2–ZnO film do not exhibit dramatic changes upon increase of bending cycles. This reveals that the deposited ZnO film is relatively thin. When only ZnO exists, the variation of electrical properties were less than 3% after the bending process (Fig. S6†). Therefore, it does not influence the large structural deformation under the mechanical stress. We believe that the TMDs-based hybrid film in this study will be a promising candidate for applications requiring high efficient and flexible opto-electrical properties.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ra00578a |
This journal is © The Royal Society of Chemistry 2019 |