Abstract

The MoOx/MoS2 thin films were manufactured on conducting glass (FTO) from the ethanolic mixture of colloidal molybdenum disulfide (MoS2) and molybdenum oxides (MoOx) by electrophoretic deposition method and were used for counter electrode of quantum dot solar cells. Different ramp-rate conditions for electrophoretic deposition as well as bias potential were investigated in an attempt to get the highest possible electrocatalytic activity of polysulfide (S2-/Sn2-) redox couple. In this research, interestingly, by simply using CdS/CdSe/ZnS photoanode and polysulfide electrolyte under 1000 W.m−2 AM 1.5 G illumination, the power conversion efficiency of MoOx/MoS2-counter-electrode-based QDSC was achieved up to 2.01%, which was double compared to platinum-based counter electrode of QDSCs.

1. Introduction

Nowadays, quantum dot-sensitized solar cells (QDSCs) have emerged as a promising candidate. The construction of QDSCs is derived from dye-sensitized solar cells (DSCs), but using quantum dots (QDs) to replace organic dye molecules as the light harvesting material [1]. Compared to silicon solar cells or DSCs, QDSCs possess the largely improved absorption coefficient (for example, depending on the particle size and the excitation wavelength, those values of CdSe and ZnSe are 105–106 and , respectively [2, 3]) and the enhanced photoexcitation. They are also highly light-, thermal-, and moisture-stable, easy to fabricate, and available at low cost [4]; thus, they have been widely studied for harvesting a wide range of light and converting it to electrical energy. Moreover, the small size of QDs can guarantee an excellent film-forming capability [5, 6].

In order to enhance the performance of QDSCs, enormous available literatures commonly focused on improving the three parts of anode, cathode, and the electrolyte. To maximize the durability of QDSCs, polysulfide (S2-/Sn2-) [7, 8] and cobalt complex (Co2+/Co3+) [9, 10] redox couples have been investigated as active electrolytes to replace the iodide/triiodide (I-/I3-) electrolytes that usually cause photocorrosion [7, 8]. In terms of cathodic materials, so far, replacing noble platinum (Pt) has been the main objective of most studies, as Pt might be a superior cathode, but on its low abundance, corrosion by an electrolyte and high cost would prevent industrialization and commercialization [1113]. To overcome this problem, many attempts have been made in a search for suitable nanomaterial-based cathodes for QDSCs containing polysulfide or cobalt complex redox couples [7, 8, 10]. Compared to Pt, using some kinds of metallic sulfide to prepare the cathode has many advantages, such as low cost, low charge transfer resistance, and good catalytic activity towards polysulfide, etc. [14, 15]. Among them, CuS and CoS have been widely utilized as cathodes in QDSCs [16, 17], each of them as their own advantages and disadvantages. More specifically, CuS cathodes generally exhibit high values of open circuit potential () but show low electrocatalytic activity and stability due to sulfur corrosion and poor adhesion to the FTO substrate. On the contrary, CoS-based cathodes have displayed some of the highest fill factor (FF) and power conversion efficiency () in QDSCs to date, thanks to their outstanding electrocatalytic activity; yet, they possess relatively low open circuit voltages [18]. As a result, more intensive studies and researches should be conducted in order to develop better and more efficient cathodes. There has also been many attempts to exploit different types of metal sulfides, namely, PbS [14, 15, 19], NiS [20, 21], ZnS [21], FeS2 [21], and Cu2S [22], or the combination of different metal sulfides [2325], which all have the potential to replace expensive Pt with low-cost Pt-free catalysts for cathodes, leading to a promising route to enhancing the power conversion efficiency and reducing the fabrication cost of QDSCs.

For the synthesis of transition metal chalcogenides, numerous approaches have been investigated, such as solvothermal synthesis [26, 27], hydrothermal method [28, 29], in situ deposition methods [30], chemical vapor deposition [31], and many other methods [32, 33]. Among various preparation methods, electrophoretic deposition (EPD), the translation of colloidal charged particles moving to an appropriately biased electrode when applying an electric field across the two electrodes [34], appears to be a straightforward, facile, and low-cost process. Compared to other methods (e.g., coatings, ceramic formation, and chemical vapor deposition), it has many essential advantages such as significantly shorter processing time, simple equipment requirement and no need of binder, high versatility and easy modification, and cost-effectiveness as well as ideal suitability for the good distribution and high quality of products [3537]. Therefore, EPD is now becoming a captivating technique for the manufacture of biological materials (for example, proteins, enzymes, and cells), as well as solid inorganic nanoparticles, colloids, and polymers, both academically and industrially, which can be applied in many scientific fields, such as chemistry, biology, biochemistry, biotechnology, materials science, and pharmaceutical. Subsequently, various new applications for EPD in the development of both bulk materials and coatings are being reported [3739], with increasing interest on the utilization of its advantages on nanomaterials [35, 40, 41], biological materials [42, 43], and even solar cells (SCs) [44, 45].

Group 6 transition metal dichalcogenides, such as MoS2, are promising candidates for photocatalysis and photoelectrochemical applications, in the form of either bulk materials or monolayer thin films, as the photoexcited electrons in MoS2 can be transferred to the electrode while the holes are removed by the electrolyte, resulting in photocurrent generation [46]. However, up to now, there have been several studies on EPD of MoS2 thin films for electrocatalytic hydrogen evolution [47, 48], but no attempt on its utilization in manufacturing the cathode for quantum dot solar cells [49]. Therefore, in this work, we present the fabrication of MoOx/MoS2 thin films via electrophoresis, their electrochemical characterization, and their catalytic activity as cathodes in QDSC assembly with FTO/TiO2 photoanode and polysulfide electrolyte.

2. Experimental

2.1. Materials and Reagents

Conducting glass (FTO—2.2 mm thickness, 8 Ω) and TiO2 paste (DSL 18HR-T) were purchased from Dyesol (Australia). Cadmium nitrate tetrahydrate (Cd(NO3)2.4H2O), (NH4)6Mo7O24.4H2O and thiourea were obtained from Acros Organics (USA). Methanol (CH3OH), ethanol (C2H5OH), potassium chloride (KCl), selenium (Se), and sodium hydroxide (NaOH) were purchased from Merck (Germany), and sodium sulfite andydrous (Na2SO3), sulfur (S), and zinc acetate dihydrate (Zn(CH3COO)2.2H2O) were from China.

2.2. Synthesis of MoOx/MoS2 Thin Films via Electrophoresis
2.2.1. Synthesis of MoOx/MoS2 Colloidal Solution

In a typical reaction, ammonium heptamolybdate tetrahydrate, (NH4)6Mo7O24.4H2O, and thiourea, CSN2H4, were immersed in a solution of ethanol. The reaction mixture was rapidly heated and kept upon microwave-assisted condition for different time and power. When a great amount of black precipitate occurred, the reaction mixture was cooled down and diluted with ethanol. The solid product was separated by centrifugation, washed several times with absolute ethanol, and then dispersed in ethanol. The characterizations of MoOx/MoS2 colloidal solution were detailed in the Supporting Information.

2.2.2. Pretreatment of FTO Glass

FTO glasses ( cm) were washed with distilled water and ethanol twice and then were immersed in 0.1 M HCl/ethanol under ultrasonication. Finally, they were dried at 100°C, cooled down, and were coated with 5 mm perforated tape.

2.2.3. Electrophoresis

Electrophoresis process was conducted using Electrophoresis crosspower equipment (Atto, Japan), Autolab PGSTAT302N and Booster 20 A (Eco Chemie, Netherlands), and Solar Simulator OrielSol1A (Newport, USA). Typically, a pair of treated FTO glasses were parallelly put in a cell containing MoOx/MoS2 colloidal solution in ethanol, and the distance between two FTO glasses was 2.5 cm (Figure S3). The MoOx/MoS2 colloid was ultrasonicated within 2 hours before electrophoresis to have good dispersion. The fabrication of MoOx/MoS2 thin films was conducted under various conditions of (i) electrophoresis potential of 15, 20, 25, and 30 V and (ii) electrophoresis time of 5, 10, 15, 20, and 25 minutes. After the electrophoretic process, the obtained MoOx/MoS2 cathodes were calcinated at 350°C for 120 minutes.

2.3. QDSC Assembly

The pristine FTO/TiO2 photoanodes were prepared following our lab-made process using screen printing method [15]. The SILAR (successive ionic layer adsorption and reaction) method was employed for the preparation of FTO/TiO2/CdS/CdSe/ZnS electrode, which involved the immersing of the FTO/TiO2 substrate in a cationic precursor solution (M2+) and then an anionic precursor solution (S2- or Se2-), allowing them to react with each other. The complete process is detailed in the Supporting Information (Figure S4).

The FTO/TiO2/CdS/CdSe/ZnS QDs photoanode and the MoOx/MoS2 cathode were sandwiched under heating (150°C). The polysulfide electrolyte, consisted of 0.500 M Na2S, 0.200 M sulfur, and 0.200 M KCl in the mixture of MeOH/H2O (3/7, v/v), was injected into the cell, and the holes on the cathode side were sealed using small pieces of Surlyn and microscope cover glass (Figure S5).

2.4. MoOx/MoS2 Electrophoretic Thin Film Characterization

The electrochemical properties of MoOx/MoS2 thin films on FTO were evaluated via cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) in the three-electrode cell which included Pt as counter electrode and Ag/AgCl as reference electrode. The electrolyte was polysulfide solution containing 5.00 mM sulfur, 20.0 mM Na2S, and 0.100 M KCl in the solution of MeOH/H2O (3/7, v/v). A blank solution of 0.100 M KCl was also prepared for the comparison.

The crystalline structures of MoOx/MoS2 materials were characterized by X-ray diffraction (XRD) using an X-ray diffractometer D2 Phaser (Bruker, Germany) with a copper anode ( Å), and the XRD patterns were acquired in the 2θ range of 10°–80° (0.02° per second). Raman spectroscopy measurements were performed on a LabRAM HR 800 Raman Spectrometer (HORIBA Jobin Yvon) using the green line of an argon-ion laser ( nm) as the excitation source, and the Raman spectra were collected in the wavelength region ranging from 200 to 3000 cm-1 at room temperature. The morphology and chemical composition of the thin films were analyzed by scanning electron microscope (SEM) method with a Philips XL 30 FEG SEM (Japan).

3. Results and Discussion

3.1. Electrochemical Properties of MoOx/MoS2 Thin Films

Electrocatalytic activity of MoOx/MoS2 thin films and Pt electrode was evaluated by cyclic voltammetry (CV), and the characteristic parameters from cyclic voltammogram are detailed in Table 1. Figure 1 shows that the prepared thin films had no electrocatalytic activity in the blank solution, and FTO substrate was incapable of catalyzing the desulfurization reaction. Thus, it can be concluded that the change of current intensity in other cyclic voltammograms was caused by the catalysis of MoOx/MoS2 thin films for the desulfurization reaction of the polysulfide solution. Additionally, Pt electrode was less active towards polysulfide solution, so at the same measuring conditions, the current density of the Pt electrode was almost significantly lower than that of the MoOx/MoS2 thin films, as mentioned in the previous researches. Moreover, according to prior studies, the electrocatalytic activity of MoS2 in the thin film was positively influenced by the appearance of molybdenum oxides (MoOx), as a layer of substoichiometric molybdenum oxide (MoOx, ) and other -type transition metal oxides (TMOs) can be used as a Hole Extraction Layer (HEL) to improve the performance of the QDSCs [50]. MoO3, which was normally an insulator, was transformed into the oxygen-deficient species of MoOx and MoO2 due to surface passivation via the bonding between FTO silicon and TMO oxygen when the thickness was sufficient [5153]. This MoOx passivation layer could behave like a barrier for the interfacial recombination, thus could improve the photoelectrochemical performance of the thin films. As a result, molybdenum oxides such as α-MoO3 were proved to have a good and stable performance in polysulphide electrolyte due to their low charge transfer resistance at the counter electrode–electrolyte interface [54].

The cyclic voltammograms were used to determine the stability of the MoOx/MoS2 thin films (Figures S6 and S7 in the Supporting Information), where the spent film can be proved to be stable if the current shows no significant deterioration after 30 cycles of sweep. The current intensity of the 1st cycle and the 30th cycle was not much different. It tended to increase until the 12th cycle and then later gradually decreases. This could be explained that at first, the rate of diffusion of sulfur to the electrode surface was faster than the transfer of charge across the electrode boundary and the electrolyte solution. When a certain value was reached, the diffusion rate decreased and the sulfur concentration on the surface decreases strongly, so the current intensity decreased. At the 30th cycle, the current density was nearly 70% of value of the first cycle. Thus, it was possible to conclude that the MoOx/MoS2 thin film had good stability.

Moreover, the charge transfer resistance () of MoOx/MoS2 thin films was measured by electrochemical impedance, which is shown in Figure 2. Charge transfer resistance was the resistance against the process of electron transfer from one phase (MoS2 electrode) to another (polysulfide solution).

From Table 2, it can be seen that under the same preparation parameters, the resistance of the MoOx/MoS2 thin films increased when either increasing the voltage or the electrophoresis time. When the resistance was sufficiently high (greater than 110 Ω), the activity of the thin films decreased dramatically, as shown in Table 1. To achieve optimal electrocatalytic activity, the thin film must be of an appropriate thickness, as resistance is inversely proportional to thickness. If the film is too thick, it will have a high resistance, preventing electrons from moving efficiently from the FTO substrate to the polysulfide solution, lowering the ion conductivity. Electrophoretic deposition allows for easy modification of the film thickness; the higher the voltage applied and the longer the electrophoresis time, the thicker the film becomes [38]. Moreover, from Table 2, the charge transfer resistance of most MoOx/MoS2 thin films was much smaller compared to Pt electrode (approximately 20 times smaller), leading to the enhancement of electroactivity as measured in Table 1.

Due to the high electroactivity and stability of MoOx/MoS2 thin films, they can be used as a counter electrode for quantum dot solar cells, and the performance parameters of which, such as the I–V characteristics, are shown in Figure 3. The short circuit current density , open circuit voltage , fill factor, and power conversion efficiency of the thin films are summarized in Table 3. All of the MoS2 cathodes have the open circuit voltage of around 0.4–0.5 V, except for the ones deposited at 15 V/25 min and 30 V/15 min, whose open circuit voltages dropped to around 0.25 V. The decrease observed in both the short circuit current density and the open circuit voltage of these two thin films also leads to poor values of overall power conversion efficiency.

Several factors contribute to the efficiency of QDSC power conversion, including charge carrier recombination with the redox couple at the semiconductor interface, slower hole transfer, and counter electrode performance [55]. Our previous effort to design PbS-QDSC has already resulted in an improvement in the fill factor of the QD [15]. However, due to slower hole transfer to sulfide redox couple, the charge recombination may become the prime limiting factor in achieving higher efficiency of QDSC [56]. Hence, the midgap states created by the doping of semiconductor films such as MoS2 can cause electrons to get trapped and protect them from charge recombination with holes and/or oxidized polysulfide electrolyte [57]. As a result, numerous studies are being conducted to better understand the charge transfer dynamics and recombination processes in systems.

The maximum efficiency of QDSCs using MoS2 thin film (deposited at 25 V in 10 minutes) is 2.01%, which was nearly twice as high when compared to QDSCs with a Pt cathode (0.984%). Its higher fill factor also suggests that the sum of the charge-transfer resistance at the counter electrode–electrolyte interface and the series resistance for the MoS2 thin film is lower than for the Pt cathode, which is evidently associated with the higher catalytic activity towards the charge transfer reaction at the counter electrode–polysulfide electrolyte interface, as proved by the cyclic voltammetry (Figure 1) and EIS measurements (Table 2). This result is also consistent with the results of electrochemical activity that we measured in Table 1.

Table 4 compares the electrochemical parameters of QDSCs based on different types of metal chalcogenide cathode materials, as they not only exhibit excellent electrocatalytic activity towards the polysulfide electrolyte but also possess good physical and chemical stability, low cost, and ease of fabrication. The MoOx/MoS2 cathode exhibited satisfactory performance specifications compared to other types of sulfide counter electrodes constructed using different methods.

3.2. Characterization of MoOx/MoS2 Thin Film

In order to understand the good electrochemical properties and satisfactory QDSC application, the MoOx/MoS2 thin film was characterized. To begin with, nanostructure mixture of molybdenum disulfide and molybdenum oxide (MoOx/MoS2) has been synthesized by a facile and rapid method of microwave, and the diameter of the particles in the colloidal solution was measured by laser particle size analyzer LA-960A, which is shown in Figure S1 and Table S1. In this strategy, microwave irradiation accelerates the synthesis of MoS2 to several minutes owing to the special localized heating via dipolar polarization and ionic conduction, allowing higher heating rates and shorter processing time [59], with the potential to increase the amount of uniformly smaller nuclei which enables the formation of the final product in smaller diameter [60].

The structure and morphology of MoS2 were characterized by X-ray diffraction (XRD), Raman spectroscopy (Raman), and scanning electron microscopy (SEM). Firstly, MoS2 powder after electrophoresis from MoS2 colloidal solution was analyzed by XRD (Figure S2). The elemental analysis of the obtained powder indicates that the products seem to be a mixture of molybdenum sulfides and molybdenum oxides (Table S2). Since thiourea may act as a reducing agent, there might also be some undesired reduction by-products, as Afanasiev et al. showed that the reduction of aqueous ammonium heptamolybdate at different reaction conditions might yield different compounds such as binary sulfides of MoS5.6, MoS6, and MoS3 as well as binary oxides of MoO3 and MoO2 [61]; still there was no peak to be observed in the XRD diagram. There was no peak appearing to match with the peak positions of the crystalline phases of MoS2, especially the absence of the reflections at ° or 37.4°, which suggests a large extent of destacking in the synthesized MoS2 [62, 63]. The products exhibited their poor crystallinity and were almost amorphous. The experimental results suggest that the structure of the synthesized samples using microwave heating was randomly oriented, with crystal defects or strains owing to the curvature of the layers. The broad peaks at the low diffraction angle of less than 20° (which is commonly found in nanomaterials) have not been determined, but they also confirm the a highly disordered and poorly crystalline structure of the MoS2 materials [64, 65]. It is considered a satisfying result, as a study of Miki et al. concluded that amorphous MoS2 is better than other crystalline phases when used as a cathode in lithium batteries because of its much greater conductivity and significant life cycle of lithium batteries [66].

Subsequently, the thin films after electrophoresis from MoOx/MoS2 colloidal solution were also analyzed by XRD, and the result is shown in Figure 4. Similar to the XRD spectrum of MoS2, MoS2 films were believed to be amorphous. These results were also consistent with the results of XRD analysis for MoS2 colloidal solution [67]. The molybdenum oxide appeared to be a mixture of MoO2 and MoO3. As observed in the XRD spectrum, besides the peaks of FTO, there are some other peaks which may belong to MoO2 as the reduction–decomposition product of the heptamolybdate salt [68] and MoO3 as the calcination products at 350°C [69, 70]. More specifically, the peaks at , 17.5, 21.1, and 29.5° belong to the hexagonal h-MoO3 (ICSD #75417) [71], and other peaks at , 27.3, and 33.6° belong to orthorhombic α-MoO3 (ICSD #35076) [72]. The diffraction peaks of MoO2 at , 37.0, 53.6, 60.6, and 67.6° [73] may be overlapped or may be too small to be observed in the XRD diagram. Based on TGA results of Kovács et al., (NH4)6Mo7O24.4H2O will release H2O and NH3 to form (NH4)8Mo10O34 in the first decomposition step (from 25 to 180°C), then (NH4)2Mo4O13 in the second decomposition step (from 180 to 270°C), and finally MoO3 in the third decomposition step (from 270 to 370°C), with the increasing amount of α-MoO3 and decreasing amount of h-MoO3 when the temperature is elevated [74].

Next, the surface morphology and composition of MoOx/MoS2 thin film deposited at 25 V in 10 minutes were analyzed by field emission scanning electron microscope with Oxford Energy-dispersive X-ray Spectrometer (FE-SEM/EDS). Figure 5 shows that the surface of the thin film covered with MoOx and MoS2 was relatively uniform and was a little bit rough. It is possible that the surface defects were formed during the calcination process, as thin layer of molybdenum oxides appeared quickly during high temperature oxidation of the defective MoS2 layer and then vaporized leaving the flaws on the material surface [75, 76]. Nevertheless, these defects contribute as good catalytic sites for desulfurization reaction. Additionally, from the SEM images, MoS2 particle size was estimated about 200–300 nm. By EDS analysis (Figure 5(d) and Table S2), besides the Sn and O peaks of the FTO glass, Mo and S were present as major elements of the thin film. Together with the FE-SEM/EDS results, the Raman spectrum in Figure 6 also exhibited the vibration of Mo-O and Mo-S bonds of MoOx/MoS2 thin film, as well as the Sn-O bond of the FTO glass [77, 78]. There was also Sn-S bond due to the sulfidation of the FTO glass (SnO) at high temperature. These interactions help make the thin films become more adhesive to FTO substrate; thus, electrons can easily move from the substrate to the surface of the thin film to react with sulfur in the polysulfide solution.

4. Conclusion

MoOx/MoS2 thin films have successfully been fabricated via electrophoresis method with good catalytic activity towards polysulfide solution after thermal treatment process. The thin films contain molybdenum sulfide and different molybdenum oxides and possess high stability, uniformity, and adhesiveness on FTO substrate. The film thickness could be modified by electrophoretic deposition with different value of applied voltage and the electrophoresis time. Moreover, the charge transfer resistance on the boundary between the MoOx/MoS2 thin films and polysulfide solution is also much lower than the Pt electrode. The best condition for the electrophoresis of MoS2 colloid dispersed in ethanol solvent was also established to achieve the best electrocatalytic activity towards polysulfide solution, which is at 25 V in 10 minutes. QDSC assembly testing showed that the best cell fabricated with a FTO/TiO2(3)/CdS(3)/CdSe(5)/ZnS(2) anode and a MoOx/MoS2 cathode deposited at 25 V in 10 minutes yielded the highest efficiency of 2.01% with  mA/cm2,  V, , and an active area of 0.25 cm2 under AM 1.5 illumination. This result was nearly double compared to Pt cathode in the polysulfide electrolyte, suggesting that MoS2 thin films are capable of replacing Pt as a cathode in QDSCs. Furthermore, electrophoretic deposition method was shown to be an effective pathway to fabricate thin film, thanks to its simple equipment requirements, short processing time, and the ability to control the thickness and the distribution of the product. Future researches on this matter will focus on various preparation parameters, such as molybdenum precursors, reducing agents, or calcination temperature, in order to achieve different ratio of MoOx/MoS2 or MoO2/MoO3, therefore leading to better enhancement in the activity of the cathodes.

Data Availability

The research data used to support the findings of this study are included within the article.

Conflicts of Interest

The authors declare that there is no conflict of interest.

Supplementary Materials

Figure S1: scheme of particle size distribution. Figure S2: XRD pattern of MoS2. Figure S3: electrophoretic preparation of MoOx/MoS2 electrode on FTO. Figure S4: process of preparation of electrode FTO/TiO2/CdS/CdSe/ZnS by the SILAR method. Figure S5: schematic of the fabricated device. Figure S6: cyclic voltammograms of MoS2 ( V, minutes, and area 0.25 cm2) after 30 scanning cycles. Figure S7: cyclic voltammograms of MoS2 ( V, minutes, and area 0.25 cm2) after 30 scanning cycles. Table S1: diameter of the MoOx/MoS2 particles. Table S2: the elemental composition analysis of MoS2 powder. (Supplementary Materials)