N the past few decades, the wide use of non-renewable energy sources like fossil fuels become a universal concern both for harming the environment and, since the demand over availability is increasing drastically, for their rapid depletion. Dye sensitized solar cells (DSSCs), utilizing solar energy, the largest renewable energy source, have been extensively explored for their facile and inexpensive fabrication process, and environmentally friendly nature when compared with other PV technologies [1,2]. The typical structure of DSSC involves: a mesoporous TiO2 (m-TiO2) photoanode sensitized with N719 dye molecules, a 𝐼 ! /𝐼 " ! redox electrolyte, and a platinum (Pt) counter electrode (CE). Despite the promising features of DSSCs, the lesser absorption of incident sunlight by the dye sensitizers due to the lower reflectivity of transparent Pt CE limits their optimal PV performance. To date, (Corresponding author: D. Radu; email: dradu@fiu.edu) N. Kaur is a postdoctoral associate in the Department of Mechanical and Materials Engineering, Florida International University, Miami, FL 33174 USA (email: nakaur@fiu.edu) F.M. Syed is a M.S. in Engineering Management and graduate assistant in the Department of Mechanical and Materials Engineering Florida International University, Miami, FL 33174 USA. (e-mail: fsyed011@fiu.edu).
various research efforts have been made for the improvement of standard DSSCs via modifying or altering different components. However, factors affecting their PV performance such as the power conversion efficiency (PCE), materials stability, and charge transfer processes do not go hand in hand [3][4][5]. Considering the significance of standard materials, the external use of reflectors has attracted substantial attention [6][7][8]. Recently, Ibrahim et al. [9] demonstrated a significant 71.4% enhancement in the PCE of DSSCs using aluminum (Al) foil as reflector, as it contributed not only towards improving the light absorption of sensitized dye molecules via light scattering effect but also escalates the reduction of 𝐼 " ! via faster charge transportation. Further, Al reflectors have also been studied by Wu et al. [10] reporting an enhanced PV performance of DSSC with reflectors by 42% relative to the standard cell. Kim et al. [11]
In this work, the PV performance of the DSSCs using standard materials was improved by adding a reflective layer of metallic Ag film at the back of platinum CE. We demonstrated high PCE values, up to 6.65% for the Ag reflector based DSSC, a significant increase from the 5.61% measured for standard DSSC used as a baseline. EIS measurements were conducted to identify the mechanistic underpinnings of the PCE enhancement, supporting the hypothesis that it is caused by a reduction in charge transfer resistances at the interfacial layers.
All the chemicals used in the present work were used as purchased. For synthesizing Ag nanoparticles (NPs), poly(acrylic acid) 25 wt% in water (MW 50,000) and 50 wt% in water (MW 5000) were purchased from Polysciences, Inc., diethanolamine (99%) from Alfa Aesar, and AgNO3 from Chemsavers. For fabricating DSSCs, fluorine-doped tin oxide (FTO) substrates were procured from Hartford Glass Co., Inc., titanium (IV) isopropoxide (TTIP) and titanium dioxide (TiO2) paste from Sigma Aldrich, platinum (Pt) paste, Ditetrabutylammonium cis-bis(isothiocyanato)bis(2,2'-bipyridyl-4,4'-dicarboxylato) ruthenium(II) (N719) dye, and EL-HSE electrolyte from Greatcell Solar Co. Ltd.
Ag NPs were prepared through a procedure adapted from literature [14]. In a typical preparation, poly(acrylic acid) (25 wt% in water, MW = 50,000) (0.18 ml), poly(acrylic acid) (50 wt% in water, MW = 5,000) (0.36 ml), diethanolamine (8 ml) and DI water (5 ml) were added to a 250 ml Erlenmeyer flask and stirred for 15 min at room temperature, to form a homogeneous solution. Further, an amount of 4 ml of AgNO3 aqueous solution was added dropwise into the above-prepared solution and subjected to 24 h stirring under the same conditions. The resulting dark brown suspension was further ultrasonicated for 1.5 h at 65°C and allowed to naturally cool to room temperature. To collect the NPs, 30 ml ethanol was slowly added (at a 3 ml/min rate) into the above solution, for precipitation, followed by centrifugation. The Ag ink was synthesized by mixing the Ag nanoparticles with nanopure water and ethylene glycol in a 50 ml centrifuge tube, at a ratio of 1:0.55:0.35 (weight ratio) Ag nanoparticles: nanopure water: ethylene glycol. The ink was stirred continuously for 24 h.
Photoanode Preparation. FTO substrates were patterned by depositing zinc paste (zinc powder (100 mg) + nanopure H2O (0.5 ml)) onto the designated area. Once the zinc paste dried, aqueous HCl (18.25% v/v) was poured dropwise on the substrate, causing an exothermic reaction that led to the etching of FTO film on the paste-covered areas on the substrates. The FTOs were further washed with nanopure water. The patterned FTOs were thoroughly cleaned by dipping in individual beakers, containing soap solution (2% of Hellmanex III in nanopure H2O), nanopure water, acetone, ethanol, and isopropanol, and ultrasonicated for 15 min each. After every step, the FTO substrates were dried under N2 flow and finally placed in a UV ozone cleaner for 30 min. A compact layer of TiO2 (c-TiO2) was deposited on the pre-annealed FTO substrates, at 450ºC, using a spray-pyrolysis method, where 0.4 M solution of TTIP was sprayed on the effective area, further subjected to annealing for 30 min. Over the c-TiO2, m-TiO2 with an effective area of 0.15 cm 2 , was prepared by depositing as procured TiO2 paste using the doctor blading technique and annealed at 450ºC for 30 min, and a clear thin film was obtained. The thickness of c-TiO2 and m-TiO2 films are approximately 80 nm and 11.5 µm, respectively, as observed through the cross-sectional FESEM image (Figure 1). The m-TiO2 thin films were dipped overnight in 0.4 mM solution of N719 dye, and the resultant was the photoanodes of DSSCs. Counter Electrodes Preparation. For a comparative investigation, three types of CEs were prepared, first by doctor blading Pt-paste on pre-cleaned FTOs and annealing at 450ºC for 30 min, second with Ag ink doctor bladed on the glass side of Pt-CE FTOs and third with doctor blading Ag film on FTO substrate. Figure 2 illustrates the cross-sectional SEM image of Pt and Ag thin film which showed an approximate respective thickness of 450 nm and 7 µm, respectively. The thickness of c-TiO2, m-TiO2, Ag and Pt thin film is still under optimization. Further, the DSSCs were fabricated by assembling photoanodes and CEs in a sandwich-type geometry, along with inserting redox 𝐼 ! /𝐼 " ! electrolyte in between.
The PV performance of fabricated DSSCs was tested using an ORIEL LCS-100 TM solar simulator under 1 sun conditions (1.5 G AM) at an intensity of 100 mW•cm -2 , connected to a Keithley source meter (Model 2400). Before the device characterization, the solar simulator was calibrated with a standard silicon cell. EIS measurements were performed on the Ω Metrohm Autolab in a frequency range of 1 Hz to 1 MHz.
The DSSC modified with Ag reflector film showed a PCE of 6.65%, a significant efficiency increase (by 18.53%) from the reference DSSC (5.61%), which is mainly attributed to the increased JSC value from 11.67 mA•cm -2 to 14.52 mA•cm -2 . Ag CE-based DSSC showed negligible PV performance as Ag NPs are unstable in an iodine-based electrolyte. Figure 4 II. The underlying mechanism could be related to the reflective properties of the Ag film, as shown in Figure 3, which materializes in the amplification of the light propagating effect. This further leads to the creation of a large number of photogenerated charge carriers as the incident unabsorbed light is reabsorbed by the sensitized dye molecules, hence resulting in higher JSC and PCE. To investigate the PV performance reproducibility of both reference devices and Ag reflector based DSSCs, ten total devices (five within each configuration) were assembled. It was observed that the PCE within each category of fabricated devices is similar, with a deviation of around ± 0.16 (Figure 5), which indicates the reproducibility of DSSCs and validates the role of Ag reflector film in producing higher-performance solar cells.
To understand the improved PV performance of Ag reflector based DSSCs, the charge transfer mechanism has been estimated from the Nyquist plot (Figure 6) and Bode plots (Figure 7), obtained at respective VOC's under illumination. All the EIS and Bode plot parameters are summarized in Table III. Two distinct semicircles of fabricated DSSCs were observed in the Nyquist curves, where the first semicircle at the higher frequency region corresponds to the charge transfer resistance at the CE and electrolyte interface denoted as RCT, and the second semicircle at the lower frequency region represents the charge transfer resistance (R1) at photoanode and electrolyte interface. These semicircles were fitted with an equivalent > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < circuit model -RS+Q1/RCT+Q2/R1, where Q is the constant phase element, RS is the series resistance of the DSSCs involving FTO's, c-TiO2, m-TiO2, electrolyte, Pt CE, and interfaces, and its value is the intercept of the first semicircle at X-axis. RCT and R1 were obtained from the diameter of the first and second semicircle, respectively. A decrease in the R1 for Ag reflector DSSC highlights an improvement in the regeneration of dye molecules as well as the electrocatalytic action at CE/ electrolyte interface. Reduced R1 specifically is the result of the reduced recombination reactions occurring at the photoanode/electrolyte interface, and hence higher JSC in Ag reflector based DSSC. RS and RCT were observed to increase in the case of Ag reflector based DSSC in comparison to the reference DSSC, which is attributed to the increase in series resistance of device.
Further, the improved charge transport was confirmed through the electron lifetime at the interfacial layer in DSSCs estimated from Bode plots. The electron lifetime (𝜏) can be estimated from the characteristic back charge transfer frequency, which is the frequency maxima (𝑓 #$% ) of the peak observed in Figure 7, they are inversely proportional to each other as in equation ( 1):
It is observed that the 𝜏 for Ag reflector based DSSC (0.03 ms) is slightly higher than the reference DSSC (0.02 ms) which is in consonance with the higher series resistance of the fabricated devices.
A simple and elegant approach to improve DSSCs performance using a thin film of Ag as a reflector on Pt CE was successfully demonstrated. A significant improvement in the PCE of Ag reflector based DSSC (6.65%) is observed in comparison to reference DSSC (5.61%) and is attributed to the increased JSC, due to subsequent absorption of unused incident sunlight as it is being reflected towards the photoanode. This leads to further excitation of the dye molecules and aids in generating a large number of photogenerated charge carriers. The EIS measurements confirmed the reduction of the charge transfer resistances at the photoanode/ electrolyte interfacial layer in Ag reflector based DSSC, which indicates a lesser recombination reaction rate and faster catalytic reactions occurring at the electrolyte/CE interface. Hence, the Ag NPsbased thin films could be effectively utilized as reflectors for improving light harvesting in the current DSSC PV technology.