Fabrication of Dye-Sensitized Solar Cells (DSSC) Using Mg-Doped ZnO as Photoanode and Extract of Rose Myrtle (Rhodomyrtus tomentosa) as Natural Dye

A dye-sensitized solar cell (DSSC) device using Mg-doped Zn thin films as photoanode and fruit extract of rose myrtle (Rhodomyrtus tomentosa) as the natural dye was investigated. The effect of annealing temperature (400-550°C) on the films of photoanode was systematically studied using an X-ray diffractometer (XRD), UV-Visible Near Infrared (UV-Vis NIR) Spectrophotometer, scanning electron microscopy (SEM), and energy dispersive spectroscopy (EDS). XRD confirm that all sample has the wurtzite hexagonal with crystallite size of 25 nm. The SEM images reveal particles on the surface of the Mgdoped ZnO thin film of irregular shapes. Increasing the annealing temperature leads to a larger particle size and slightly increases bandgap energy. The dye sensitizer of extracted rose myrtle (Rhodomyrtus tomentosa) has a strong absorption at the visible light region. The maximum efficiency of the DSSC device is 3.53% with Mg-ZnO photoanode annealed at 500°C.


Introduction
The demands for renewable energy continually increase every year due to its eco-friendliness. Solar cells have been well known as a device to convert solar energy to electricity for decades. However, conventional solar cells are still high priced due to complicated fabrication process and expensiveness of raw materials. Dye-sensitized solar cell (DSSC) is one of the most promising solar cell types to produce renewable energy with a low-cost material and simple fabrication process [1][2][3]. After irradiation, the dye sensitizer harvests light and causes an electron to promote the conduction band leaving a hole in the valence band. There are numerous pigments of plant leaves, fruits, and flowers that have the potential to be utilized in DSSC. The variety of pigments with different absorption wavelengths and degrees of absorptivity in the UV-visible spectrum can cause different performances of DSSC. The molecules of the dye can be anchored into the surface areas of the semiconductor to form Lewis acid-base types of interaction to enhance electron transfer from HOMO of the dye molecule (pigment) to the conduction band of the semiconductor (anode) [4][5][6][7].
Zinc oxide (ZnO) semiconductor plays an important role as a photoanode to improve the conducting interface layer and to enhance the power conversion efficiency (PCE). According to the literature, ZnO has a high electron mobility, wide bandgap (3.37 eV), and large exciton binding energy of 60 meV [8]. Magnesium (Mg) is one of the metals that is used in many applications such as refractory materials and optical and heating apparatus [9,10]. Mg-doped ZnO material also has special properties to block the electron due to its wide bandgap [11,12]. There are several methods to grow thin film on a substrate, such as molecular beam epitaxy, metal-organic chemical vapor deposition, plasma-enhanced chemical deposition, sputtering method, spray pyrolysis, atomic layer deposition, pulse laser deposition, electron beam evaporation, and sol-gel [13]. The sol-gel method has several advantages compared to the aforementioned methods such as simple, cheap, and efficient [14]. By using a sol-gel spin coating technique, several parameters like concentration of precursor solution, annealing temperature, and annealing time can be easily tuned in order to achieve the desired properties [12,15].
In this work, the photoanodes of Mg-doped ZnO thin films were prepared by a sol-gel spin coating method. To the best of our knowledge, the natural dye from the fruit extract of Rhodomyrtus tomentosa has not been reported yet as the dye sensitizer for DSSC. The effect of different annealing temperatures on structural and optical properties of Mg-doped ZnO photoanodes as well as the efficiency of DSSC device was systematically investigated using necessary characterization tools. We find that the maximum efficiency of the DSSC device is 3.53% with Mg-doped ZnO photoanode annealed at 500°C.

Experimental Section
2.1. Synthesis of Mg-Doped ZnO Thin Films. Mg-doped ZnO thin films were fabricated using a sol-gel spin coating technique. Typically, zinc acetate dihydrate and magnesium chloride (2 wt.%) were dissolved in isopropanol (35 mL) under continuous stirring. After 10 min, 1.7 mL diethanolamine was added slowly into the solution. After refluxing process at 90°C for about 2 hours, the gel was dropped on top of FTO glass and spun at 5000 rpm for 60 s. After the drying process, the samples were annealed at different temperatures of 400, 450, 500, and 550°C for 5 hours.
2.2. Extraction of Natural Dyes. About 50 grams of Rhodomyrtus tomentosa fruit was ground using a mortar. After being moved into a beaker glass, 25 mL DI water, 21 mL ethanol, and 4 mL acetic acid were added and then stirred to form a homogeneous solution. The solution was then covered with aluminum foil to avoid photooxidation and socked at room temperature for 24 h. Finally, the solid and liquid parts were separated using filter paper. The filtered solution of Rhodomyrtus tomentosa fruit extract was ready to be used as the sensitizer in DSSCs. Figure 1 illustrates the schematic fabrication of the DSSC device. First, the as-prepared Mgdoped ZnO thin film was used as the photoanode electrode. Natural dye sensitized from Rhodomyrtus tomentosa fruit extract was adsorbed on the top of Mg-doped ZnO photoanode by immersing it into the extracted dye solution for several hours. After that, it was taken out and washed with ethanol to remove the unadsorbed dye and dried in the oven at 80°C. The commercial platinum coated on the glass FTO was used as the counter electrode. The DSSCs were assembled by attaching the photoanode and the counter electrode using thermoplastic sealant surlyn as glue and separator and then heated at 80°C to let the surlyn perfectly attach to the electrodes. The electrolyte was injected through a tiny hole that was drilled on the counter electrode. Finally, that hole was covered with transparent tape.

Characterization Tools.
To observe the surface morphology of Mg-doped ZnO thin films annealed at different temperatures, a scanning electron microscope (JEOL-6500) analysis was performed at an accelerating voltage of 15 kV. The X-ray diffraction (XRD) pattern of Mg-doped ZnO thin films was analyzed using an X-ray diffractometer (LabX XRD-6100, Shimadzu) with Cu Kα (λ = 1:54 Å). The transmittance and absorbance spectra were recorded using a UV-Vis NIR spectrophotometer. The efficiency of the DSSC was measured using an I-V measurement (Keithley Source Measure Unit) system by irradiating a photoanode electrode with a LED and input power of 35 mW/cm 2 . Several data such as open-circuit voltage (V oc ), short circuit current density (J sc ), maximum voltage (V max ), and maximum current (J max ) were recorded. Then, the fill factor (FF) and efficiency (ɳ) were determined using equations (1) and (2), respectively.

Results and Discussion
3.1. Electron Microscope Analysis. The surface morphology of Mg-doped ZnO with variation of annealing temperatures was investigated using a field-emission scanning electron microscope. With a magnification of 30k times, the top view images of Mg-doped ZnO thin films can be clearly observed in Figure 2. The surface microstructure of Mg-doped ZnO at different annealing temperatures shows nanoparticles with irregular shapes. It was clearly observed that by increasing the annealing temperature, the particle size was monotonically increased. Figure 2(d) shows the representative energy dispersive spectroscopy (EDS) spectra. The spectra exhibit five peaks to indicate the presence of zinc, oxygen, magnesium, platinum, and tin in the film. The appearance of platinum is contributed from the platinum coating before SEM analysis to improve the conductivity while tin comes from the substrate. The presence of a relatively low-intensity peak  International Journal of Photoenergy for Mg compared to zinc and O peaks confirmed the successful Mg doping into ZnO host. Furthermore, the EDS quantitative result depicted in Figure 2(d) has shown that the weight and atomic percentage of Mg are about 0.83 and 1.32%, respectively. To calculate the particle size more precisely, further analysis was conducted using ImageJ analysis. As shown in Figures 3(a)-3(c), the average particle sizes for Mg-doped ZnO thin films annealed at 400, 500, and 550°C are 30 ± 5, 53 ± 9, and 82 ± 17 nm, respectively. A larger particle size at a higher annealing temperature was reasonable. It could be explained due to a higher driving force from thermal energy that leads to a faster particle growth through Ostwald ripening mechanism. Our findings also well agree with some previous reports [16,17].

XRD Analyses.
The crystal properties of Mg-doped ZnO were studied by the XRD technique. The results are shown in Figure 4. The XRD patterns are similar to a wurtzite crystal structure based on the standard card of JCPDF #36-1451 (ZnO) [18]. Three peaks that located at 32.5°, 35.3°, and 37.0°for 2θ could be assigned as (100), (001), and (101) planes of ZnO, respectively. The other weak peak at about 38.5°might be contributed by impurity as reported in the previous work [19]. The intensity of peaks in Figure 4 grad-ually elevates as the temperature of annealing increases, which indicates an improvement in the crystallinity of Mgdoped ZnO films. Table 1 lists the summary of structural properties of Mgdoped ZnO thin films at different annealing temperatures. The average crystallite size of Mg-doped ZnO thin films was calculated at (101) plane using the Scherrer formula as shown in equation (3) [20]. Their crystallite size values are 20.60, 21.23, 16.83, and 22.9 nm at the annealing temperature of 400, 450, 500, and 550°C, respectively. Next, the dislocation density (δ) of Mg-doped ZnO was further determined by equation (4) [21]. The dislocation density of Mg-doped ZnO annealed at 400, 450, 500, and 550°C is 2:36 × 10 −3 , 2:22 × 10 −3 , 3:53 × 10 −3 , and 1:89 × 10 −3 nm -2 , respectively. Mg-doped ZnO annealed at 500°C has the highest dislocation density compared to other samples. Macrostrain value that indicates the peak shift position was calculated according to equation (5) [22]. Based on the database, the (101) plane for ZnO is located at 36.25°with an interplanar spacing of 2.4759 Å. However, the (101) plane for our Mg-doped ZnO is found at about 37.00°for 2θ with a calculated interplanar spacing of 2.4272 Å. The peak shifting of about 0.75°for 2θ also indicates that Mg as a dopant has been successfully doped into ZnO host lattice [23]  3 International Journal of Photoenergy with a value of 1:97 × 10 −2 because of their similarity in peak position. Based on the previous reports [24], the lattice parameters a and c of Mg-doped ZnO were estimated to be about 3.172 Å and 5.080 Å using equations (6) and (7), respectively. The lattice parameter a at the (100) plane did not significantly differ for different annealing temperatures because their peak position was located almost in the same diffraction angle. Similarly, the lattice parameter c at the (002) plane is also very similar at different annealing temperatures, as listed in Table 1.
where D is the crystallite size (nm), λ is the wavelength (nm), β is the full half maximum, FWHM (rad), and θ is the Bragg angle (°).
where d o is the interplanar spacing of pure ZnO without deformation while d is the calculated interplanar spacing for Mg-doped ZnO at the (101) plane using the Bragg law.
3.3. Optical Properties. To study the effect of different annealing temperatures on optical properties, the absorption and transmission spectra of Mg-doped ZnO thin films are measured and presented in Figures 5 and 6, respectively. The absorption peaks of all Mg-doped ZnO thin films are located at a wavelength of 350 nm, which is at the UV region. As clearly shown in Figure 5, the absorption of Mg-doped ZnO annealed at 400°C is quite low. However, after increasing the annealing temperature to 450°C and 500°C, the absorption sharply elevates. Further increasing the temperature of annealing to 550°C leads to a lower absorbance but is still higher than that at 400°C. Figure 6 exhibits transmission spectra of Mg-doped ZnO which also shows a similar trend to the absorption spectra in Figure 5. The thin films have transparency about 50-80% at visible light region. The bandgap energy of Mg-doped ZnO thin films was further derived based on the optical absorption data and plotted in Figure 7. As listed in Table 2   International Journal of Photoenergy 400, 450, 500, and 550°C, respectively. The slight increment of bandgap energy with increasing temperature might be due to the Burstein-Moss effect as reported in previous studies [25].

Absorbance of Rhodomyrtus tomentosa Dye Extract.
The optical absorption spectrum of the extracted rose myrtle (Rhodomyrtus tomentosa) natural dye was measured using a UV-Vis spectrophotometer to investigate its sensitivity to light. As shown in Figure 8, the natural dye has a strong absorption at the visible-light region with an intense absorbance peak at a wavelength of 610 nm. This property is very useful for DSSC to improve the light absorption ability. It is also well known that about 43% of the solar spectrum falls in the visible light range and only 4% is in the UV region [26]. The more light can be absorbed, the more electron-hole can be generated, which leads to a higher efficiency of a DSSC device.
3.5. Efficiency of DSSC. Figure 9 exhibits the J -V characteristic curve of DSSC with Mg-doped ZnO as photoanode at    Table 3. The open-circuit voltage (V oc ) at different annealing temperatures was similar with a value of 0.5 V. In contrast, the short-circuit current (J sc ) was significantly different. The J sc values are 3.83, 4.55, 8.77, and 4.03 mA/cm 2 at annealing temperatures of 400, 450, 500, and 550°C, respectively. The efficiency of the DSSC device annealed at 400°C was about 1.66%. By increasing the annealing temperature to 450°C, the efficiency also was increased to 2.36%. Further increasing the annealing temperature to 500°C, the efficiency of 3.53% can be achieved. However, the efficiency was observed to decline when the annealing temperature is 550°C. Therefore, the optimum annealing temperature is 500°C with a maximum power conversion efficiency of 3.53%. Based on the characterization results, the highest efficiency of 3.53% at 500°C could be contributed due to its highest optical absorption as revealed by UV-Vis analysis in Figure 5.

Conclusions
We have successfully fabricated a dye-sensitized solar cell (DSSC) device using Mg-doped ZnO thin film as the photoanode and natural dye of rose myrtle (Rhodomyrtus tomentosa) as the dye sensitizer. The scanning electron microscope analysis revealed that the surface of Mg-doped ZnO thin film was particles with irregular shapes. It is found that increasing the annealing temperature led to a larger particle size and slightly increased bandgap energy. The natural rose myrtle dye sensitizer had a strong absorption at the visible light region. The maximum efficiency of the DSSC device was 3.53% at an annealing temperature of 500°C. This work demonstrates that the annealing temperature of photoanode significantly affects the efficiency of the DSSC device.

Data Availability
The research data used to support the findings of this study are included in the article.

Conflicts of Interest
The authors declare that they have no conflicts of interest.