Abstract

The surface modification of a TiO2 electrode with diethyldithiocarbamate (DEDTC) in dye-sensitized solar cells (DSSCs) was studied. Results from X-ray photoelectron spectroscopy (XPS) indicate that over half of the sulfur atoms become positively charged after the DEDTC treatment of the TiO2 surface. DSSCs were fabricated with TiO2 electrodes modified by adsorbing DEDTC using a simple dip-coating process. The conversion efficiency of the DSSCs has been optimized to 6.6% through the enhancement of the short-circuit current density ( mA/cm2). This is substantially higher compared to the efficiency of 5.9% ( mA/cm2) for the DSSCs made with untreated TiO2 electrodes.

1. Introduction

With the impending energy crisis and the growing concern about global warming, there has been considerable interest in the industrial sector and in the academia in dye-sensitized solar cells (DSSCs) since their discovery in 1991 [1]. DSSCs are presently the most cost-effective third-generation solar cell technology available with efficiencies reaching 12% [2]. Solar cells based on other thin-film technologies have efficiencies that are usually between 5% and 13%. The commercial silicon solar cells have efficiencies of 14%–18%. This makes DSSCs attractive as a substitute for present solar energy conversion technologies. While silicon solar cells are highly efficient, they also utilize advanced technology and, hence, have a higher cost. DSSC technology, however, is relatively simple and may become significant if certain technological problems can be solved.

The development of DSSCs is an important research area in alternative energy because of lower production cost and reasonable power conversion efficiency [3]. After O’Regan and Graetzel reported on low-cost DSSC solar cells with an efficiency of  7% [1], numerous investigators have attempted to enhance the performance of DSSCs by optimizing the properties of their constituents to improve their efficiency [3]. Although other semiconductors such as ZnO, SnO2, and CdS have been used in DSSCs, none of them have been as successful as TiO2 [46]. Hence, TiO2 is the most commonly used anodic material in DSSCs because of its high efficiency in DSSCs and its chemical stability [1].

In a DSSC, a dye molecule absorbs visible light and an electron is excited to a higher energy level. The excited electron is injected into the conduction band (CB) of TiO2. These electrons in the CB of TiO2 are transported towards the FTO surface along the interconnected particle matrix, through trap-mediated random diffusion process. During their journey, they could recombine with the oxidized dye molecules or the oxidized species of the redox couple, thus reducing the cell efficiency. Attempts have been made to change the surface properties of the TiO2 layer to reduce such recombinations. It has been demonstrated that the charge recombination can be considerably reduced by coating the TiO2 layer either with an ultrathin layer of an insulator [7, 8] or another wide band-gap semiconductor [9, 10].

A self-assembled monolayer (SAM) of a species such as dithiocarbonate adsorbed on a semiconductor surface plays a major role in determining the energy barriers between a semiconductor and a sensitizer adsorbed on its surface [11]. One of the most productive interactions involves the formation of an SAM on the TiO2 surface. Despite the decrease in the electron conductivity due to the SAM, the overall conversion efficiency of DSSCs has been observed to increase due to the enhancement of the short-circuit current density [7, 10, 11].

According to the literature, nitrogen or sulfur-containing molecules have been used as additives in the iodide/triiodide redox electrolyte in dye-sensitized solar cells [12]. These additives have contributed to a positive shift of conduction band edge and a decrease in the charge recombination rate [12, 13]. Further, a large increase in the photocurrent density along with a small decrease in photovoltage was also demonstrated in a study using thiourea [12, 13]. Additionally, thiourea has been used as an adsorbent in the synthesis of visible-light-responsive titanium dioxide thin films [14].

In this study, diethyldithiocarbamate (DEDTC) was coated on via a dip-coating process using ammonium diethyldithiocarbamate (ADEDTC) in order to modify the surface electronic properties of. We hypothesized that DEDTC could possibly generate a surface layer on the surface resulting in an increase in by reducing recombination. The adsorption configurations of DEDTC on the particles were studied by X-ray photoelectron spectroscopy (XPS). The effect of the surface treatment on the performance of DSSCs was also investigated.

2. Experimental Section

Fluorine-doped tin oxide (FTO) glass substrates were cleaned in a detergent solution in an ultrasonic bath for 15 min and thoroughly rinsed with ethanol. A layer with a 12 μm thickness was then deposited on the FTO glass by the doctor blade method using a paste (Ti-Nanoxide T, Solaronix). The -coated substrates were subsequently sintered at 450°C in air for 30 min. Electrodes were soaked in the aqueous solution (0.01 M) of ADEDTC for periods varying from 20 to 60 min to deposit the solution onto the particles. Afterwards, the resulting electrodes were rinsed with acetonitrile and dried at 50°C for 2 min. The /FTO and DEDTC-/FTO samples were separately immersed in a 0.5 mM N719 dye solution (acetonitrile/tertbutyl alcohol, ) for 12 h. Anhydrous electrolyte containing was sandwiched between the dye-adsorbed electrode and a platinum-coated FTO counter electrode to construct the solar cell with an active area of 0.25 cm2. The I-V characteristics and photocurrent action spectra were recorded using a calibrated solar cell evaluation system (JASCO, CEP-25BX) at AM 1.5, with a 100 mW/cm2 illumination. XPS analysis of the S 2p peak was performed for the prepared DEDTC-/FTO samples and ADEDTC powder. XPS experiments were performed by using a hemispherical electron energy analyzer (ESCALAB-MkII, VG) and an Al K X-ray tube (1486.6 eV). Each component of the S 2 core line consists of 2 and 2 peaks split by the spin-orbit coupling. The peaks show a relative intensity ratio of 1 : 2 and are separated by 1.18 eV [15]. A chemical shift was evaluated by the 2 peak position.

3. Results and Discussion

Chemical binding analysis surrounding sulfur atoms was performed by XPS. Figure 1(a) shows the XPS spectrum of the S 2 core levels obtained from an ADEDTC powder sample. The spectrum was fitted by two components corresponding to the C–S−1 bond (164.0 eV) and C=S bond (165.8 eV). Figures 1(b) to 1(d) show the S 2 spectra of the electrodes treated with ADEDTC solution for 20, 40, and 60 min, respectively. After the DEDTC treatment, two components were observed in these spectra at binding energies of 163.0 and 169.0 eV (Table 1). Dithiocarbamate (DTC) is a well-known sulfur-chelating agent that coordinates with a wide variety of metal ions. In these cases, large chemical shifts of the S 2 core level toward lower binding energies have been reported [16]. This chelating configuration given in Figure 2(a) does not agree with any of the two components observed in our XPS results. However, the S 2 binding energies of the bidentate configurations of DTCs on Au surface have been reported at 162 eV [17]. The component observed at 163.0 eV may be related to such bidentate sulfur (Figure 2(b)) on the Ti atoms of a nanocrystal. Here, it should be noted that the latter component has drastically shifted towards a higher binding energy. This core level shift indicates that an electronic charge transfer has taken place from the sulfur atom to surrounding atoms after the adsorption scheme proposed in Figure 2(c). The binding energy of 169.0 eV is close to that of sulfates [18].

Therefore, oxidized sulfur such as a sulfate ion should be related to the adsorption configuration in Figure 2(c). Taking into account the high binding energy, 2–4 oxygen atoms must be bonded to the sulfur atom. Therefore, we suggest the possibility of partial or full decomposition of DEDTC on the TiO2 surface with accompanying oxidation under the atmospheric conditions. nanocrystals are also well known for their photocatalytic activity. In reality, such decomposition of monoalkyl DTC has been reported to generate SO2, CS2, and alkyl-N=C=S on a TiO2 surface under the irradiation of UV light [19]. Thus, sulfur atoms bonded to a number of oxygen atoms may be expected after photocatalytic decomposition of DEDTC, as shown in Figure 2(d). X-ray radiation employed in the XPS experiment may be responsible for inducing photocatalytic activity in the . The absence of such decomposition with free DEDTC confirms the ability of the for the observed decomposition.

The relative amount of sulfur corresponding to a binding energy of 169 eV increased with dipping time. This difference may have an effect on the DSSC performance. The relative amount of sulfur related to the lower binding energy peak decreased. With increased dipping time, sulfur atoms in the dithiocarbamate which are in the −2 oxidation state may be getting oxidized to positive oxidation states through bonding to oxygen atoms.

The photocurrent-voltage curves of DSSCs using the bare and DEDTC-treated layers are compared in Figure 3. Attempts were made to optimize the deposition process of DEDTC by varying the coating time because it is one of the most important parameters determining the coating amount in the dip-coating process [12]. We prepared three different DEDTC-TiO2/FTO electrodes by varying the coating time from 20 to 60 min and used them as the photoelectrodes of the DSSCs. Photoconversion efficiencies (EFFs) of the DSSCs are presented in Table 2 together with fill factor (FF), short-circuit photocurrent , and open-circuit voltage . The conversion efficiency with the bare /FTO photoelectrode is 5.91%, and with the DEDTC-/FTO photoelectrode, the conversion efficiency increased to 6.56%. In particular, it is noteworthy that for the DEDTC-/FTO photoelectrode increased to 12.74 mA/cm2, whereas that of the bare photoelectrode is 11.26 mA/cm2.

Increase in can be explained as follows. It is possible that DEDTC is preferentially adsorbed at defect sites of the nanoporous structure, resulting in a decrease of the surface states in the band-gap region. Therefore, back donation of photoelectrons from to the electrolyte and N719 dye would be decreased. Consequent increase in by the DEDTC-treatment resulted in the improvement in the energy conversion efficiency. This effect is similar to the influence of 4-tertiary butyl pyridine in electrolyte, which results in a lowering in and an increase in . In this study, the efficiencies varied with the dipping time. A maximum efficiency of 6.56% is observed with a 30-minute dipping time, and the efficiency and the fill factor were decreased with further increase in the dipping time. Similar trend was reported recently where the fill factor values decreased with adding thiourea into the electrolyte [12]. This decrease was attributed to the decrease in dye adsorption on the surface. It was observed that the longer dipping time can present multilayer structure of DEDTC that can enhance decrease in dye— interaction.

Figure 4 shows the absorbance spectra of four nanoporous TiO2 electrodes: a bare electrode, DEDTC-treated (for 30 min.) electrode, a dye-adsorbed electrode, and a dye-adsorbed electrode after DEDTC treatment for 30 min. Compared to the bare electrode, the apparent increase in light absorption by the DEDTC-treated electrode (curves (a) and (b)) is possibly due to scattering effects as DEDTC does not absorb light in the visible region. It is also evident from the curves (c) and (d) in Figure 4 that DEDTC treatment has not adversely affected the dye adsorption by the TiO2 layer. In fact, the data in Table 2 clearly show that the DEDTC treatment has positively contributed to enhance the cell efficiency at an optimum dipping time of 30 min.

Electrochemical impedance spectroscopy (EIS) was employed to investigate the effect of DEDTC treatment on the internal resistance of the DSSCs. The results of EIS are shown in Figure 5. The charge transfer resistance at the /dye/electrolyte interface is 195  for the cell with the DEDTC-treated electrode and 307.5 Ω for the cell with the bare electrode. The lower internal resistance of the cell with the DEDTC treated electrode has clearly contributed to the improvement in fill factor.

We suggest here that DEDTC is a useful additive because it exhibits a dual functionality, namely, improving the visible light absorption and decreasing the photoelectrode resistance. The probability of back donation of photoelectrons from the to the electrolyte or dye through the surface layer would be reduced. Therefore, after the DEDTC treatment, the energy conversion efficiency was improved by increasing the short circuit current. Our results show that there is a slight decrease in the which is more than compensated by an increase in with DEDTC treatment for the cells with optimum performance. A recent study [20] has shown that a super-thin AlN layer has markedly reduced the dark reaction and greatly improved the forward electrical transport in the intrinsic InGaN/-InGaN solar cell where it has been suggested that the leakage current mechanism changes from a defect related one to an interface tunneling. DEDTC may be playing such a role by blocking the defect sites on electrode in the DSSCs described in this work. Yu et al. [21] have discussed the role of oxygen vacancy- defect sites as recombination centers in reducing both the open-circuit voltage and the fill factor.

Kim et al. [12] have studied the effect of incorporating thiourea into the electrolyte of -based DSSCs and observed a small decrease in , a substantial increase in , a reduction in fill factor, and an overall increase in efficiency. These results are therefore qualitatively identical to what we have observed by the incorporation of DEDTC. They have ascribed the improvement in cell performance to (i) minimizing recombination by adsorption of thiourea and (ii) reaction of thiourea with (present as triiodide ions, ) in the electrolyte forming ions and ions. This reaction reduces the concentration of triiodide ions which absorb part of the light. Therefore, a decrease in triiodide concentration increases the photocurrent. The release of ions lowers due to a positive shift of the conduction band of .

A similar reaction is possible between ADEDTC and triiodide ions as follows xy(1) This reaction converts part of the triiodide ions into iodide ions and contributes to a higher photocurrent as described earlier. Another consequence of the previous reaction is a negative shift of the redox potential of couple, due to the decrease in ion concentration and a corresponding increase in ion concentration. The observed decrease in can be explained as due to this negative shift of the redox potential.

4. Conclusions

The effects of DEDTC adsorption on the surface of /FTO electrodes via a dip-coating process were studied. XPS results indicate that DEDTC deposited on the surface results in the creation of positively charged sulfur. Use of these electrodes (DEDTC-/FTO) as photoanodes in DSSCs improved the cell performance due to enhanced visible light absorption and decreased internal resistance by reducing surface states. Furthermore, the presence of DEDTC can reduce back electron transfer and improve overall conversion efficiency because of short-circuit current enhancement. Finally, we obtained improved conversion efficiency by employing DEDTC-/FTO as the photoanode compared to a photoanode without DEDTC treatment.