Preparation of ZnO QDs

The synthesis of ZnO QDs was performed using the hot-injection method already reported elsewhere.^[Reference Kumar, Kumar, Rawat, Kumar, Sharma, Pal and Jit^15] The prepared solution was filtered using the polyvinylidene fluoride membrane (0.22 µm) to remove the large and unreacted materials from the solution to enhance the surface uniformity of the thin film. Furthermore, the filtered colloidal ZnO QDs was deposited on the cleaned patterned ITO-coated glass substrate using spin coating (SPM-150LC TSE- system GmbH, Germany). The prepared sample is placed in a pre-heated oven at 200 °C for 30 min under ambient environment. The monoethanolamine (MEA) used for the preparation of ZnO QDs defines the lower limit of temperature as the evaporation temperature of MEA is ~160–190 °C. The process of spin-coating and heating is iterated till the desired thickness of ~ 40 nm is achieved measured by Reflectometer (F-20 Thin Film UV Analyzer, Filmetrics). Afterward, the ZnO QDs thin film is annealed at 250 °C under the ambient environment for 30 min. The prepared samples are used with solution-processed MoO_x and thermally evaporated MoO_x to analyze the effect of heat treatment.

Preparation of MoO_x solution

MoO_x solution is prepared using the technique specified by Girotto et al.^[Reference Girotto, Voroshazi, Cheyns, Heremans and Rand^9] The 1 M solution of MoO₂ powder is stirred with H₂O₂ for ~2 h at 80 °C under the ambient environment and then allowed to cool down for 24 h to obtain a clear yellow liquid. Acetylacetone is further used to modify the viscosity and concentration of the solution. The concentration of H₂O₂ defines the quality of the solution; excess H₂O₂ makes the solution stable for a long duration, while it changes to dark-blue solution due to the insufficient conversion of MoO₃ into MoO₃(OH)(OOH) under the deficiency of H₂O₂.^[Reference Girotto, Voroshazi, Cheyns, Heremans and Rand^9] The prepared solution is spin-coated on an ITO/ZnO QDs substrate, and then annealed at 270 °C. The process is repeated to achieve the thickness of 30–35 nm measured using Reflectometer (F-20 Thin Film UV Analyzer, Filmetrics).

Device fabrication

Thermal evaporation (vacuum evaporation) is used for the deposition of high-purity Ag (99.99%) and MoO_x using HHV, FL400 SMART COAT 3.0 A. Thermal deposition of MoO_x was performed on the ITO/ZnO QDs to achieve a thickness of ~30 nm. Afterward, the thermal deposition of Ag was performed on the solution and thermally prepared ITO/ZnO QDs/MoO_x using a shadow mask technique, and the thickness of ~50 nm is achieved. Thermal deposition of both the materials (MoO_x and Ag) was performed at a vacuum of ~10⁻⁶ bar at the variable rate (for first 5 nm 0.1 Å/s, for 5–20 nm 0.5 Å/s, and for 20 nm onwards 1 Å/s). The thickness of the deposited thin film and the rate of evaporation are monitored using digital thickness monitor (INFICON model no. SQM-160) attached to the coating unit. Furthermore, the complete device is heated at 280 °C for 30 min under the ambient environment after the one set of results are obtained as discussed in further sections. The effect of post-heat treatment on the Ag electrodes is shown in the Supplementary Material.

Structural characterization

The high-resolution scanning electron microscopy (HRSEM: SUPRA 40, Zeiss) image of the complete device structure with solution-processed and thermally grown MoO_x is performed before heat treatment of the complete device and after heat treatment of the complete device at 280 °C under the ambient environment for 30 min as shown in Fig. 1. The HRSEM of the solution-processed MoO_x-based diode shows the stability of the device as the layers are distinguished after the heat treatment and indicates separable interface between MoO_x and ZnO QDs as shown in Figs. 1(a) and 1(b). Similarly, HRSEM of the diode with thermally evaporated MoO_x is performed and shown in Figs. 1(c) and 1(d). The device after heat treatment shows the continuous interface between thermally grown MoO_x and ZnO QDs, which may be due to the diffusion of MoO_x into the ZnO QDs shown in Fig. 1(d). The morphology of as-grown ZnO QDs is also analyzed using transmission electron microscopy (TEM: TECNAI G2 20 TWIN) image shown in Supplementary Fig. S1 and diffraction pattern is shown in Fig. S2. The detailed analysis of the ZnO QDs size and orientation is reported in our previous work.^[Reference Kumar, Kumar, Mukherjee, Rawat, Kumar, Pal and Jit^24] The effect of heat treatment on the Ag electrode is shown for both devices in the Supplementary Figs. S3–S6. Furthermore, the annealing of the thin films greatly improves the morphology and interface characteristics of the devices and materials. Thermally deposited MoO_x does not show any grains in Fig. 1(c), but after the heat treatment of the complete device the grain boundaries are visible leading to the crystallization of the MoO_x thin film shown in Fig. 1(d). In the case of solution-processed MoO_x, annealing of the film is performed at 270 °C under ambient environment for 30 min. The annealing of solution-processed MoO_x crystallizes the MoO_x film shown in Fig. 1(a).

Figure 1.

HRSEM image of (a) solution-processed MoO_x-based device before annealing at 280 °C, (b) solution-processed MoO_x-based device after annealing at 280 °C, (c) thermally grown MoO_x-based device before annealing at 280 °C, and (d) thermally grown MoO_x-based device after annealing at 280 °C.

Optical characterization

The study of optical properties of MoO_x and ZnO QDs is analyzed in this section. The photoluminescence excitation (PLE) and emission spectrum of ZnO QDs and MoO_x are shown in Figs. 2(a) and 2(b), respectively. The PLE spectrum of ZnO QDs shows a peak at 365 nm (3.39 eV), the curve of absorption spectra is fitted to match the wavelength axis. The peak of PLE spectrum for MoO_x is obtained at ~400 nm, and the same is also shown in Figs. 2(a) and 2(b). The PL spectrum of the material provides great details about the monodispersity of the nanocrystals, quality of the film or synthesized material, and interface defects if any. The PL spectrum of ZnO QDs shown in Fig. 2(a) shows the narrow-band emission spectrum with a full-width at half-maximum (FWHM) of 72 nm. The selectivity of ZnO QDs PL indicates the high-quality and monodispersity of the ZnO QDs. The PL of QDs can be treated as a surface-dependent phenomenon due to the dependence of band gap upon the particle size of the QDs.^[Reference Klimov^25] The PL spectrum of MoO_x shown in Fig. 2(b) has the FWHM of 125 nm and shows no trailing edges indicates that the MoO_x are free from any surface defects. The involvement of surface defects increases the charge-trapping sites.

Figure 2.

(a) Photoluminescence (PL) excitation and emission spectrum of colloidal ZnO QDs solution and (b) PL excitation and emission spectrum of MoO_x-solution.

Electrical characterization

The current versus voltage (I–V) relationship for the devices with MoO_x grown by solution processing and thermal evaporation are shown in Fig. 3, measured using Keysight B1500A from −1 to +1 V. The device under study for MoO_x grown by solution processing and thermal evaporation is also shown in the inset of Figs. 3(a) and 3(c), respectively. The I–V curve is plotted under the dark and under the illumination of UV light (UV LED source with an optical power density of 800 µW/cm² at 365 nm). The effect of heat treatment on the complete device is evident in both devices as the dark current (noise current) increases with the annealing. The solution-processed MoO_x-based device shows superior photoresponse after heat treatment compared with the device without heat treatment. Furthermore, the stability of the photocurrent is also observed in Fig. 3(b) compared with Fig. 3(a) and the contrast ratio (I _ph/I _dark) of the device enhanced from 3.88 to 38.72 at an applied bias of −1 V. The thermally grown MoO_x-based device shows deterioration in rectifying nature of the device and the noise current is increased by ratio of greater than ~100 times shown in Figs. 3(c) and 3(d). This deterioration of the device is due to the diffusion of MoO_x into the ZnO QDs, which varies the barrier between the ZnO QDs and MoO_x from graded to the linear junction. The photoresponse of the devices can be compared using the responsivity of the device calculated using:

(1)$${\mathop{\rm Re}\nolimits} (v,\lambda ) = \displaystyle{{\mathop I\nolimits_{{\rm ph}} (v,\lambda )} \over {\mathop P\nolimits_{{\rm opt}} (\lambda )}},$$

where I _ph is the photocurrent obtained at a bias of V volts at wavelength λ. The P _opt is the optical power density of the light source measured at wavelength λ using the power meter (PM100D, Thorlabs). The measured responsivity of the devices with and without heat treatment under the illumination of UV LED at an applied bias of −1 V are indicated in Table I. The A ²/C ² versus voltage curve for solution-processed and thermally grown MoO_x-based device after heat treatment is shown in Fig. 4. The heterojunction between MoO_x and ZnO QDs is assumed to be a Schottky junction, and therefore most of the depletion region will reside in ZnO QDs thin film. The curves shown in Fig. 4 can be interpreted as

(2)$$\displaystyle{{A^2} \over {C^2}} = \displaystyle{2 \over {q\varepsilon}} \left( {\displaystyle{{\phi - V} \over {N_{\rm d}}}} \right),$$

where N _d is the donor concentration in n-type layer, ε is the effective dielectric constant of the ZnO QDs film, q is the electron charge, and ϕ is the built-in potential. From the projected intersection of capacitance curve in voltage axis we find the value of built-in potential and found to be 0.25 and 0.046 V for solution-processed and thermally grown MoO_x-based device, respectively. Thus, the low built-in potential shown by a diode (for thermally grown MoO_x) indicates the junction between ZnO QDs and MoO_x is not abrupt but the linear in nature. The ideality factor of the diodes are also calculated and found to be 3.85 and 6.14 for solution-processed and thermally grown MoO_x-based device, respectively. The ideality factor defines the quality of the diode (junction). Ideally, the value of ideality factor should be equal to 1, but due to interface defects this result deviates from 1 and tends to increase >1. The high value of ideality factor 6.14 also indicates the low quality of junction formed between thermally grown MoO_x and ZnO QDs.

Figure 3.

Logarithmic current–voltage curve under light and dark condition for (a) solution-processed MoO_x-based diode without annealing, (b) solution-processed MoO_x-based diode after annealing, (c) thermally grown MoO_x-based diode without annealing, and (d) thermally grown MoO_x-based diode with annealing. The complete device structure of the solution-processed and thermally grown MoO_x are shown in the inset of (a) and (c), respectively.

Figure 4.

A ²/C ² versus voltage curve for (a) solution-processed MoO_x-based diode and (b) thermally processed MoO_x-based diode.

Table I.

Responsivity of the fabricated devices under the optical power density of 800 µW/cm² at 365 nm.