Performance Improvement of Dye-Sensitized Solar Cells with Pressed TiO₂ Nanoparticles Layer

Abstract

A simple and low-cost fabrication method of dye-sensitized solar cells (DSSCs) was developed to improve the structure and performance of the photoanode with the pressed layer and compact TiO₂ thin film using spin coating, screen printing, and mechanical compression. In this study, four different TiO₂ layers were adopted to fabricate photoanodes: a mesoporous TiO₂ nanoparticles (NPs) layer, a pressed TiO₂ NPs layer, a mesoporous TiO₂ NPs layer on the TiO₂ compact thin film, and a pressed TiO₂ NPs layer on the TiO₂ compact thin film. The compact thin film was deposited on the fluorine-doped tin oxide (FTO) glass via spin coating, while the mesoporous TiO₂ NPs layer was deposited via the screen-printing method. The pressed TiO₂ NPs layer was produced by compressing the mesoporous TiO₂ NPs layer with a hydraulic press machine. When using the pressed TiO₂ NPs layer for the photoanode of DSSC, the power conversion efficiency of DSSC was enhanced the most. The electron lifetime for DSSC with photoanodes based on the pressed TiO₂ NPs and mesoporous TiO₂ NPs layers were 8.217 and 6.287 ms, respectively. The power conversion efficiency of DSSC with photoanodes based on the pressed TiO₂ NPs layer was 5.4%, while that based on the mesoporous TiO₂ NPs layer was 4.08%. DSSC with photoanodes based on the pressed TiO₂ NPs layer showed a significant increase in the power conversion efficiency by 36.16% compared to that based on the mesoporous TiO₂ NPs layer.

1. Introduction

The energy demand is increasing every year due to the development of the industry and the increasing global population. Due to the decreasing availability of fossil fuels, the significance of developing renewable energy sources has been underscored [1]. Renewable energy from recyclable sources includes solar energy, wind energy, hydroelectric energy, biomass energy, geothermal energy, tidal energy, etc. The use of solar energy is one of the main development projects, and it has been researched for several decades. Solar cells are classified into the silicon-based solar cell (the first generation), the thin-film solar cell (the second generation), the organic solar cell, the dye-sensitized solar cell (DSCC), and the perovskite solar cells (the third generation) [2]. The silicon-based solar cell, which is monocrystalline or polycrystalline are currently the most widely used as they have high efficiency, a long lifespan, and high energy yield. The thin-film solar cell uses compounds such as gallium arsenide (GaAs), cadmium telluride (CdTe), and copper indium gallium selenide (CIGS), the use of which enables thinner solar panels at a relatively low cost. However, rare metals contained in them may give toxicity. DSCC has lower energy needs and less carbon emission in the manufacturing process, so it is environmentally friendly. It also functions under lower light even without direct sunlight with better resistance to heat and damage. Such advantages allow less use of power lines and batteries in applications [2]. Thus, DSSC is replacing silicon-based solar cells. DSSC is fabricated on flexible substrates to improve portability through a simple process [3]. The photoanode of DSSC is made of different materials, such as zinc oxide (ZnO), niobium pentoxide (Nb₂O₅), tin dioxide (SnO₂), and tungsten trioxide (WO₃). Different shapes of such materials, including nanotubes, nanorods, nanowires, nanocrystals, and nanofibers, are prepared to enhance dye adsorption for the increase in light absorption [4]. Among them, TiO₂ nanoparticle is regarded as the most promising material for DSSC as it has high transmittance in the visible spectrum, is stable, non-toxic, and has a large specific surface area.

However, a low power conversion efficiency (PCE) of DSSC is observed due to the recombination effect between various interfaces inside its structure. When DSSC is irradiated by light, electrons are generated through four interfaces in sequence, including the dye/TiO₂, TiO₂/transparent conductive glass (TCO, usually fluorine-doped tin oxide (FTO)), counter electrode/electrolyte, and electrolyte/dye interface. The recombination in DSSC occurs by TCO electrons being contacted with the redox electrolyte. Thus, to avoid contact between TCO and redox electrolytes, researchers have proposed various blocking layers via different materials [3]. The blocking layer is made from different materials and preparation methods to have a dense and uniform structure that prevents direct contact between the electrolyte and the FTO surface. However, the previous processes and materials require complex chemical syntheses, which causes inconsistent results [5,6,7,8,9]. Therefore, it is necessary to find a more reliable and simpler way to produce the blocking layer.

Thus, in this study, we propose a new method to prepare the blocking layer using simple methods such as spin coating, screen printing, and mechanical pressing. Spin coating is a common and widely used coating method [10,11,12] and has the advantages of low power consumption, simple setup, and desired thickness in the preparation of the compact TiO₂ thin film. Screen printing is also a common coating method and is widely used in the preparation of the photoanode for DSSC and counter electrodes [13,14]. Mechanical compression is applied to reduce the cracks in the TiO₂ nanoparticles (NPs) layer, thereby lower the recombination effect [15,16]. In this study, a compact TiO₂ thin film as the blocking layer is prepared with spin coating, while a mesoporous TiO₂ NPs layer is fabricated using the screen-printing method. A pressed TiO₂ NPs layer is produced by applying pressure to the mesoporous TiO₂ NPs layer using a hydraulic press machine. The three distinct layers are combined and assembled into a DSSC as a photoanode. Their J–V curve and electrochemical characteristics are investigated to validate the improved performance of DSSC with photoanodes based on the pressed TiO₂ layer and compact TiO₂ thin film with the method proposed in this study.

2. Materials and Methods

2.1. Materials

The FTO substrate with a resistance of 7 Ω/sq and indium tin oxide (ITO) was procured from Ruilong in Taiwan. UNI ONWARD Corp. (New Taipei City, Taiwan), also in Taiwan, supplied acetone (95%), tert-butanol (99%), ethanol (95%), ethanol (99%), n-pentane (99%), and methyl alcohol (95%). Alfa Aesar provided lithium iodide (98%), while Sigma Aldrich supplied titanium diisopropoxide bis(acetylacetonate) (75 wt.%), iodide (99.8%), and 4-tert-butylpyridine (tBp, 99%). UniRegion BioTech manufactured the titanium dioxide P25 powder (TiO₂ P25), which is composed of 80% anatase and 20% rutile, with an average particle size of 25 nm. Avantor Performance Materials supplied acetonitrile (99.5%), and FLUKA provided 3-methoxypropionitrile. The spacer films were purchased from C.P SOLAR, and 1,2-dimethyl-3-propylimidazolium iodide (98%) was procured from TCI.

2.2. Electrolyte Fabrication

The electrolyte used in this experiment was obtained by mixing 0.1 M lithium iodide (LiI), 0.05 M iodine (I₂), 0.5 M 4-tert-Butylpyridine (tBP), and 1 M 1,2-dimethyl-3-propylimidazolium iodide (DMPII) in 3-methoxypropionitrile (MPN) and treating with ultrasonic cell disruptor.

2.3. Photoanode Fabrication

The FTO substrate was cleaned in acetone, methanol, and deionized (DI) water for 10 min in an ultrasonic bath. The preparation process of the photoanode was as follows.

2.3.1. Preparation of Mesoporous TiO₂ NPs Layer

A 15 wt% TiO₂ paste was prepared, which consisted of 1.2446 g TiO₂ P25 (the average particle size of TiO₂ NPs was 25 nm), 3 mL ethanol (99%), and 6 mL tert-butanol. The paste was stirred for 6 h to be homogeneous and then was coated on a clean FTO substrate by the screen printing method (the screen print size was 0.5 cm × 0.5 cm). The coated FTO substrate was annealed at 150 °C for 90 min and then at 450 °C for 30 min to form M TiO₂.

2.3.2. Preparation of Compact TiO₂ Thin Film

The 0.15 and 0.30 M precursor solutions for fabricating compact TiO₂ thin film were prepared with titanium diisopropoxide bis(acetylacetonate) (0.055 and 0.11 mL) with 1 butanol (1 mL). The 0.15 and 0.30 M precursor solutions of the compact TiO₂ thin film were spin-coated on a transparent FTO substrate at 1500 rpm for 30 s and annealed at 125 °C for 5 min. The process for making 0.30 M precursor solution of the compact TiO₂ thin film was repeated twice. Finally, the FTO substrate with the precursor solution was sintered at 500 °C for 30 min to form CL [17,18].

2.3.3. Preparation of Pressed TiO₂ NPs Layer

After the mesoporous TiO₂ NPs layer was prepared, it was pressurized at 138.4 kg/cm² for 60 s using a hydraulic pressurizing machine to obtain a pressed TiO₂ NPs layer. The exerted pressure was referred to in previous studies [15].

2.4. DSSC Fabrication

Four different photoanode structures were integrated, including a mesoporous TiO₂ NPs layer, a mesoporous TiO₂ NPs layer stacked on the compact TiO₂ thin film, a pressed TiO₂ NPs layer, and a pressed TiO₂ NPs layer stacked on the compact TiO₂ thin film (Figure 1) [17].

The prepared photoanode was immersed in 0.3 mM N3 ruthenium complex dye (cis bis (dithiocyanato) bis(4,4′ dicarboxylic acid 2,2′ bipyridine)ruthenium(II)) at 45 °C for 2 h in an oven and then moved to a dry cabinet at room temperature for 2 h. The counter electrode was prepared by sputtering platinum onto an indium tin oxide (ITO) glass for 60 s and at 10 mA. DSSC was assembled by placing the dye-sensitized photoanode and counter electrode together in a sandwich structure using a 60 μm thick shrinkable film as a spacer. Adherence of the shrinkable film to the substrates was induced by pressing the substrates on a heating plate. Next, the electrolyte was injected into the reserved hole on the counter electrode, which was sealed with a shrinkable film and a small glass to prevent electrolyte leakage. Finally, the conductive silver paste was applied to the photoanode and the counter electrode to enhance their conductivity.

2.5. Characterization

The morphology of the photoanode film was observed using a Field Emission Scanning Electron Microscope (FE SEM, JSM 7610FPlus, JEOL, Tokyo, Japan). The visible light absorption of the photoanode film of the dye-sensitized solar cell was measured using a UV visible spectrometer (U2900A, Hitachi, Tokyo, Japan). The power conversion efficiency of DSSC was measured with a solar simulator (XES 40S1, San Ei Brand, Osaka, Japan) under AM 1.5 (100 mW/cm²). Electrochemical Impedance Spectroscopy (EIS) (Zennium, Zahner, Kronach, Germany) was used to measure the impedance of DSSC.

3. Results

The deposited TiO₂ NPs layer is white and opaque if the layer is thick enough. The fabricated layer was semi-transparent under the light after pressurized with mechanical compression. The TiO₂ layer with mechanical compression had a reduced thickness and became denser, which reduced light scattering. Figure 2 shows the cross-sectional SEM images of the four different TiO₂ layers on the FTO glass. Figure 2a,c,e,g show the magnified images of a mesoporous TiO₂ NPs layer, a mesoporous TiO₂ NPs layer stacked on the compact TiO₂ thin film, a pressed TiO₂ NPs layer, and a pressed TiO₂ NPs stacked on the compact TiO₂ thin film, while Figure 2b,d,f,h show the interface of the FTO substrate and TiO₂ NPs layers at a higher magnification. In Figure 2a,c, valley-like voids are observed in the mesoporous TiO₂ NPs layer, and the voids between the particles are more prominent at higher magnification. The void might be caused by the rapid evaporation of the solution during the drying process in an excessive thickness of the mesoporous TiO₂ NPs layer. Figure 2c shows that the mesoporous TiO₂ NPs layer on the compact TiO₂ thin film is thinner than that of the mesoporous TiO₂ NPs layer (Figure 2a). This might be caused by the altered surface properties of the FTO by the compact TiO₂ thin film during the repeated screen printing processes. The TiO₂ layer treated with mechanical compression shows a denser structure and filled voids (Figure 2e,g). The voids observed between the FTO surface and the TiO₂ layer (Figure 2b,f) provide an opportunity for the electrolyte to contact the FTO directly. However, the compact TiO₂ thin film on the FTO surface (pointed by the white arrow in Figure 2d,h) shows a continuous and dense morphology and smaller particle size compared to the P25 TiO₂ NPs layer. This blocks direct contact between the electrolyte and FTO.

Figure 3 presents that the pressed TiO₂ NPs layer does not have cracks on its surface compared to the mesoporous TiO₂ NPs film. Thus, the compression helps prevent electrolyte infiltration into the FTO substrate surface, which allows the flat film surface to have higher power conversion efficiency [19,20]. The cracks on the surface of the mesoporous TiO₂ NPs layer (Figure 3a) are reported by other studies, too. The cracks were formed by the rapid evaporation of water or tert-butanol in the TiO₂ paste. These cracks serve as channels for the electrolyte to contact the FTO substrate, promoting charge recombination [15].

Figure 4 shows the UV-Vis absorption spectra of four different photoanodes after immersion in N3 dye. The photoanodes based on the mesoporous TiO₂ NPs layer and that on the compact TiO₂ thin film have a wider UV absorption bandwidth than the photoanode based on other TiO₂ layers. Its best absorption occurs at the wavelength of about 540 nm. The wider bandwidth can be explained by the scattering effects that enhance light absorption due to the thicker and looser structure of the mesoporous TiO₂ NPs layer. The photoanodes based on the pressed TiO₂ NPs layer and that on the compact TiO₂ thin film has a narrower light absorption bandwidth and show the best and second-best absorption at 420 nm and 530 nm, respectively. Two absorption peaks observed at 420 and 530 nm are attributed to the light absorption properties of TiO₂ and N3 dye, respectively [21]. Compared with the mesoporous TiO₂ NPs layer, the pressed TiO₂ NPs layer has a narrower light absorption bandwidth, as the thinner layer allows the light of a longer wavelength to pass through more easily, resulting in lower light absorption.

EIS analysis results are shown in

Table 1. Impedance and electron lifetime of DSSC with photoanodes based on four different TiO₂ layers.

TiO2 Layer	RS (Ω)	RPT (Ω)	RK (Ω)	RD (Ω)	${\mathbf{f}}_{\mathbf{m}\mathbf{a}\mathbf{x}}$ (s−1)	τeff (ms)
Mesoporous TiO2 NPs layer	14.78	3.78	12.94	5.26	25.32	6.287
Mesoporous TiO2 NPs layer on compact TiO2 thin film	13.04	2.54	13.57	4.28	25.32	6.287
Pressed TiO2 NPs layer	13.74	2.00	14.66	6.75	19.37	8.217
Pressed TiO2 NPs layer on compact TiO2 thin film	12.89	2.20	13.53	6.20	19.37	8.217

and Figure 5 and Figure 6. The Nyquist plot and Bode plot of DSSC with photoanodes based on different TiO₂ layers are shown in Figure 5 and Figure 6, respectively. In the plots, the horizontal axis is for the real part of the impedance, Z′(Ω), while the vertical axis is for the imaginary part of the impedance, Z″(Ω), with the frequency swept from 10 MHz to 100 kHz. The Nyquist plot in Figure 5 shows three curves. The maximum frequency, f_max is shown at the top of the middle curve. The electron lifetime (τ_eff) in the TiO₂ layer is expressed by the reciprocal of f_max multiplied by 2π [22]. The structure of the TiO₂ layer affects R_K and τ_eff. However, R_K involves more components that are difficult to analyze, so electron lifetime needs to be paid more attention to. The electron lifetime was 8.217 ms and 6.287 ms for DSSCs with photoanodes based on the pressed TiO₂ NPs layer and the mesoporous TiO₂ NPs layer. The longer electron lifetime in DSSC based on the pressed TiO₂ NPs layer implies a lower density of the layer in the working electrode. The lower density leads to the effective suppression of the recombination effect between the electrons injected into the conduction band of the pressed TiO₂ NPs layer on the compact layer and the ${\mathrm{I}}^{-}$/${\mathrm{I}}_3^{-}$ ions at the FTO surface [23]. Figure 6 shows that there is not much difference in the frequency of characteristic peaks appearing in DSSC based on four different photoanodes.

The J-V characteristics of DSSC were determined by Keithley 2400 and the solar simulator. The initial voltage was set to −0.1 V with a voltage increment of 0.01 V. The short circuit current (I_SC) was recorded when the measured voltage was 0, and the short circuit current density (J_SC) was calculated with I_SC divided by the photoanode area. The open circuit voltage (V_OC) was recorded when the measured current is 0 at a certain voltage. The product of the measured current and voltage at the maximum value is called the maximum power output (P_max) with the corresponding current density (J_max) and voltage (V_max). The fill factor (FF) is the ratio of P_max to the product of V_OC and J_SC and represents the ability of the device to provide the maximum output power. The PCE is obtained by multiplying the V_OC, J_SC, and FF and dividing the values by the power of the incident light (100 mW/cm²).

Table 2. J-V characteristics of DSSC with photoanodes based on four different TiO₂ layers.

TiO2 Layer	VOC (V)	JSC (mA/cm2)	F.F. (%)	PCE (%)
Mesoporous TiO2 NPs layer	0.74	7.66	71.66	4.08
Mesoporous TiO2 NPs layer on compact TiO2 thin film	0.76	8.36	71.31	4.51
Pressed TiO2 NPs layer	0.74	10.43	69.54	5.40
Pressed TiO2 NPs layer on compact TiO2 thin film	0.75	10.30	69.01	5.36

and Figure 7 present the V_OC, J_SC, FF, PCE, and current density, voltage curves for the photoanode with different TiO₂ layers. DSSC with photoanode based on the pressed TiO₂ NPs layer exhibited the highest PCE, demonstrating that the pressed layer and the compact layer effectively enhanced PCE by preventing electrolyte contact with FTO. DSSC with photoanode based on the pressed TiO₂ NPs layer has a higher J_SC than that based on the mesoporous TiO₂ NPs layer owing to the compact structure of the pressed TiO₂ NPs layer facilitating electron transfer [24].

4. Conclusions

The improvement of the DSSC with the photoanode based on the pressed and compact TiO₂ thin film was proposed in this study using simple and easy fabrication methods. Compressing a mesoporous TiO₂ NPs layer and using the spin coating method, DSSC with the photoanode with a pressed TiO₂ NPs layer and the compact TiO₂ thin film (as the blocking layer) of a nanometer-level thickness on the FTO glass substrate was obtained. The compression of the mesoporous TiO₂ NPs layer removed voids between TiO₂ NPs and improved electronic transmission, making the layer denser to block the contact between the electrolyte and FTO substrate. The compact TiO₂ thin film formed a uniform and dense structure on the surface of the FTO substrate to block the electrolyte. DSSCs with photoanodes based on four different TiO₂ layers, including a mesoporous TiO₂ NPs layer, a mesoporous TiO₂ NPs layer on the compact TiO₂ thin film, a pressed TiO₂ NPs layer, and a pressed TiO₂ on the compact TiO₂ thin film were compared in terms of impedance, electron lifetime, and J–V characteristics. The result shows that a pressed TiO₂ NPs layer allowed the best performance of DSSC. The SEM images showed that the loose and porous structure of the mesoporous TiO₂ NPs layer was changed to a dense structure of the pressed TiO₂ NPs layer. This structure effectively blocked the penetration of the electrolyte onto the FTO substrate and reduced charge recombination. The nanometer-level thickness of the compact TiO₂ thin film formed by spin coating also prevented contact between the electrolyte owing to its uniform structure on the FTO substrate. The analysis result of J–V characteristics and EIS showed that the compact structure increased short circuit current density and electron lifetime and thus enhanced the power conversion efficiency. Compared with the power conversion efficiency of DSSC with photoanode based on the mesoporous TiO₂ NPs layer, the power conversion efficiency of the DSSC based on the pressed TiO₂ NPs layer (5.40%) was increased by 36.16%.