High-efficiency dye-sensitized solar cells using ferrocene-based electrolytes and natural photosensitizers

According to a survey conducted in 2008, world energy consumption is projected to expand by 50% between 2005 and 2030 [1]. In addition to resource shortage, global warming, environmental contaminants, increase in oil prices and other similar factors have led to new research on renewable, environmentally sensitive and cheaper energy sources instead of non-renewable energy sources, such as coal, oil and natural gas. Today, intensive research is focused on the development of renewable alternative energies. Types of renewable energy resources include hydropower, wave power, thermal gradients, biomass, geothermal energy, wind energy and solar energy. These energy types are not only renewable, but also overall more environmentally friendly. The most remarkable energy source of these renewable types is photovoltaic energy, which converts solar energy directly into electrical energy through solar cells. Solar energy is the most powerful and unlimited source among renewable energy sources. Direct conversion of solar energy into electrical energy through photovoltaic devices is a promising approach for meeting the energy needs of the future. For a solar cell, efficiency, fill factor (FF) and production costs are the basic parameters that determine the quality of the cell. Since solar cells were first described, there have been a number of developments for increasing efficiency and decreasing production costs.

The history of solar cells began in 1839 when A E Becquerel, a French physicist, recognized the photovoltaic effect. After further investigation of the photoelectric effect by Heinrich Hertz in 1887, solar energy became an important area of growing research. Today, much progress has been made in solar cell production technologies, and high efficiencies have been achieved. The highest efficiency has been obtained with single crystal silicon-based solar cells. However, the excess consumption of material fabrication and high cost of material in single crystal silicon solar cells have led to the use of polycrystalline materials on a large scale. Research studies have focused on reducing the cost of fabrication, which has led to the development of thin-film technology. Therefore, silicon wafer-based solar cells, which are known as first-generation solar cells, have been replaced with a second-generation thin film-based solar cell technology that is cheaper to produce and involves different semiconductors, such as CdS, CdTe or CuInGeSe. However, despite the lower cost of production, the process of production is still expensive and difficult for thin-film technology.

In spite of the remarkable developments in solar cell technology over the past several decades, the high cost of producing solar cells has remained a limiting factor for the implementation of solar energy on a large scale. Therefore, there is a prevailing need to investigate new materials and concepts for photovoltaic conversion in order to lower the cost of producing solar cells. Dye-sensitized solar cells (DSSCs) have been intensively investigated in recent years as a promising source of energy in the future due to their low cost, efficiency in low light conditions and absorption across the visible light spectrum.

DSSCs were invented by Grätzel and Regan [2]. DSSC solar cells supply high conversion efficiencies and are known as third-generation solar cells. In general, a DSSC consists of a nanocrystalline semiconductor film electrode modified with a dye, a platinum (Pt) counter electrode and an electrolyte solution between the electrodes (anode and cathode). The electrodes of the standard DSSC are prepared on transparent conducting oxide (TCO)-coated glass substrates, between which the cell is assembled. The conductive coating of the substrate works as a current collector, and the substrate material itself functions as a support structure for the cell and as a sealing layer between the cell and ambient air. Fluorine-doped tin oxide (FTO; SnO₂ : F) and indium tin oxide (ITO; In₂O : Sn) are the most frequently used TCOs in thin-film photovoltaic cells. The standard preparation procedure of a nanostructured titanium oxide (TiO₂) electrode includes sintering of the deposited TiO₂ film at 400–500 °C. As the only TCO coating stable at these temperatures, FTO has been the material of choice for DSSCs [3]. The transparent substrate allows sunlight to enter the cell while its conductive surface collects charges. For electron transfer, the cathode electrode is coated with platinum (Pt) as a catalyst [4]. The space between the electrodes is filled with an electrolyte solution that ensures charge conduction through a redox couple. Although Br⁻/Br₂, SCN⁻/SCN² and SeCN⁻/SeCN² bipyridine cobalt are preferred as an electrolyte, an I⁻/I³⁻ redox couple is widely used because it has low stability and corrosion due to its high volatility [5, 6]. Therefore, alternative electrolytes have been sought to obtain solar cells with higher efficiencies. Fe²⁺/Fe³⁺ in a solvent can be used for this purpose, and the Fe²⁺/Fe³⁺ redox couple does not induce corrosion because it does not interact with oxygen during the fabrication process. Moreover, it enhances the light absorption of dyes or pigments and reduces recombination impact. Therefore, the Fe²⁺/Fe³⁺ redox couple provides advantages over I⁻/I³⁻.

A porous semiconductor film is preferred as the oxide of a semiconductor, particularly TiO₂, because it has good chemical stability [7] and its surface is highly resistant to continued electron transfer. Moreover, TiO₂ is transparent in the visible range due to a large band gap [8], and therefore it is sensitized by a light-absorbing dye in the visible range. Another advantage of the cheap and non-toxic TiO₂ material is the high dielectric constant [9, 10]. These characteristics provide good electrostatic shielding of the injected electron from the dye molecule stacked with TiO₂. Moreover, their recombination is prevented before reduction of the dye by the redox electrolyte [11]. However, electron transport in such a porous film occurs by trap-mediated diffusion, which is a slow mechanism [12]. Several studies have adopted a novel approach for improving the photovoltaic performance of DSSCs using TiO₂ nanomaterials, because these materials can improve the electron transport properties and reduce light scattering [13–15].

TiO₂ particles have to be sensitized with a layer of dye molecules that absorb light in the visible electromagnetic spectrum. To accomplish sensitization, TiO₂ is coated with a light-absorbing dye and immersed in the electrolyte solution. Dye molecules that attach to the semiconductor surface are used to absorb a great portion of the solar light. The desirable properties of a sensitizer can be summarized as follows:

• (i)  
  Absorption. The dye should absorb light at wavelengths up to approximately 920 nm (i.e. the energy of the excited state of the molecule should be approximately 1.35 eV above the electronic ground state, which corresponds to the ideal band gap of a single band gap solar cell) [16].
• (ii)  
  Energetics. To minimize energy losses and to maximize the photovoltage, the excited state of the adsorbed dye molecule should be only slightly above the conduction band edge of TiO₂, yet high enough to present an energetic driving force for the electron injection process. For the same reason, the ground state of the molecule should be only slightly below the redox potential of the electrolyte.
• (iii)  
  Kinetics. The process of electron injection from the excited state to the conduction band of the semiconductor should be fast enough to outrun competing, unwanted relaxation and reaction pathways.
• (iv)  
  Stability. The adsorbed dye molecule should be stable enough in the working environment (at the semiconductor–electrolyte interface) to sustain approximately 20 years of operation under exposure to natural daylight (i.e. at least 10⁸ redox turnovers) [17].
• (v)  
  Interfacial properties. Good adsorption to the semiconductor surface.
• (vi)  
  Practical properties (e.g. high solubility to the solvent used in the dye impregnation).

These properties can be considered to be prerequisites for a proper photovoltaic sensitizer [18]. The basic dye molecules consist of a ruthenium (Ru) metal atom and a large organic structure that provides a wide absorption range, fast electron injection, and stability. To date, several Ru complexes have achieved solar-to-electric power conversion efficiencies (η) above 11% in a standard global air mass 1.5 [19–21]. Although Ru metal complexes show high power conversion efficiencies, the cost, complex manufacturing process and environmental issues have limited the large-scale application of this type of solar cell. Therefore, scientists are interested in identifying alternatives to the expensive Ru-based dye sensitizers currently used in these solar cells, using natural dyes from plants and flowers as sensitizers in solar photo-conversion [22]. Hence, natural dyes have been recently used in solar cells. Compared with Ru dyes, natural dyes have many advantages, such as being environmentally friendly as well as having a simple manufacturing process and low fabrication costs [23–25].

In this paper, we report a new promising bilayer DSSC design using natural dyes extracted from Hypericum perforatum, Rubia tinctorum L. and Reseda luteola. The design also used an Fe²⁺/Fe³⁺ (ferrocene) liquid electrolyte coated on TiO₂ nanoparticles. To the best of our knowledge, no previous reports have described using a ferrocene liquid electrolyte with natural dyes. We first extracted dyes from the indicated species using a hot extraction method and then fabricated the DSSC solar cells using the ferrocene liquid electrolyte, extracted natural dyes and TiO₂ nanoparticles. The photovoltaic parameters controlling the device performance were then assessed.

2.1. Preparation of natural dye sensitizers

Quercetin, alizarin and luteolin dyes were extracted from Hypericum perforatum, Rubia tinctorium L. and Reseda luteola plants, respectively. The dyes were extracted using a hot water method as follows: fresh plants were washed with water and dried by mixing in an environment with air flow. The dried plants were then subjected to a grinding mill, and 1 g of the ground material was placed in 10 ml of distilled water. Finally, the plants were immersed in hot water as an extracting solvent at 50 °C for 24 h. The solids were subsequently filtered out to obtain clear dye solutions. The pH of alizarin, quercetin and luteolin dyes was adjusted to 5, 6 and 6, respectively, using oxalic acid and acetic acid. The chemical structures of the dyes are shown in figure 1.

Zoom In Zoom Out Reset image size

2.2. Preparation of DSSCs

In order to prepare a working electrode, the conductive glass plates (Asahi Glass, fluorine-doped SnO₂; sheet resistance: 15 Ω/sq) were first cleaned in a detergent solution using an ultrasonic bath for 15 min, rinsed with water and ethanol, and then dried. The photoanodes were prepared by depositing a TiO₂ film on the FTO conducting glass: two edges of the FTO glass plate were covered with a layer of adhesive tape (3M Magic) to control the thickness of the film and to mask electric contact strips. A TiO₂ (Solaronix, Co. Ltd) nanopowder (20 nm) paste was then spread uniformly on the substrate by sliding a glass rod along the tape spacer. TiO₂ nanoparticles were prepared using the sol–gel method. First, 1.2 ml titanium tetraisopropoxide [(Ti (OC₃H₇)₄, Merck] was added to 25 ml ethyl alcohol [C₂H₆O, Merck] and the solution was mixed using a magnetic stir plate for 1 h. Then, 5 ml glacial acetic acid [CH₃COOH, Merck] and 25 ml ethyl alcohol were added to the solution. After each additive component was added, the solution was mixed on a magnetic stir plate for 1 h. For the final step, 1.5 ml triethylamine [(C₂H₅)₃N, Merck] was added in the solution, and the final solution was again mixed for 1 h. The solution was aged at room temperature for 1 day before deposition. After sol formation, a spin coating process was used to cover TiO₂ nanoparticles on the front surface of the TiO₂ nanopowder/FTO conducting glass at 3000 rpm for 30 s. This process was repeated ten times. The TiO₂ nanoparticle films were then subjected to pre-heating at 500 °C for 5 min in a hot air oven after the spinning process. Finally, the films were subjected to an annealing process at 500 °C for 1 h in an oven after pre-heating. All photoanodes were immersed in the natural dye solutions at room temperature for 24 h, rinsed with distilled water, and subsequently dried. In order to prepare counter electrodes, a 10mM chloroplatinic acid hexahydrate solution in ethanol was spread uniformly onto another FTO conducting glass by sliding a glass rod along the tape spacer and pyrolysed for 15 min at 450 °C.

DSSCs were assembled as follows: the catalyst-coated counter electrode was placed on the top so that the conductive side of the counter electrode faced the TiO₂ film. The ferrocene electrolyte solution (0.05M ferrocenium hexafluorophosphate mixed with 0.1M ferrocene and 0.5M tert-butylpyridine in acetonitrile) was then placed at the edges of the plates. The liquid was drawn into the space between the electrodes by capillary action. Two binder clips were used to hold the electrodes together. The DSSC structure is shown in figure 2.

Zoom In Zoom Out Reset image size

2.3. Characterizations of dyes and DSSCs

The absorption spectra for all the dye molecules are shown in figure 3. All dyes exhibited broad absorption spectra ranging from 250 to 650 nm. Alizarin and luteolin dyes have a broad and strong absorption in the yellow region with maxima at 580 nm and in the near visible region with maxima at 354 nm, respectively. The quercetin dye exhibits two intense absorptions in the UV region with maxima at 268 nm and one in the visible region with maxima at 412 nm. Absorption bands around 300 nm can usually be attributed to the π–π^* transition, and bands around 450 nm can be attributed to the intramolecular charge transfer (ICT) between the donor and the acceptor [24, 26]. Importantly, the absorption results are in excellent agreement with those of similar studies [27–29].

Zoom In Zoom Out Reset image size

The standard characterization techniques of solar cells include the determination of the dc current–voltage characteristic under white light illumination of different intensities and the determination of the photocurrent under low intensity monochromatic light. Analysis of the photocurrent–voltage curves includes the determination of the following parameters:

• (i)  
  The FF is the ratio of the maximum power to the external short- and open-circuit values, and is typically calculated as

  $$\begin{equation} {\rm FF}=\frac{J_{\rm m} \times V_{\rm m}}{J_{{\rm SC}} \times V_{{\rm OC}}}, \end{equation} \tag{ 1 }$$

  where J_SC (short-circuit current density) is measured under the condition where the applied potential, V, equals zero volt. V_OC (open-circuit potential) is the cell potential measured when the current in the cell is zero, corresponding to almost flat valence and conduction bands. J_m and V_m are the maximum current density and voltage, respectively.
• (ii)  
  The efficiency (η) expresses the performance of a solar cell and is defined as the ratio of the maximum electrical power extracted to the radiation power incident on the solar cell surface.

  $$\begin{equation} \eta =\frac{P_{\max}}{P_{{\rm in}}}=\frac{J_{{\rm SC}} \times V_{{\rm OC}} \times {\rm FF}}{P_{{\rm in}}}, \end{equation} \tag{ 2 }$$

  where P_in is the power of incident light.

Photocurrent–voltage characteristics for DSSCs fabricated using natural dyes with a ferrocene electrolyte are shown in figure 4 and summarized in table 1. The efficiency of the cell sensitized using the quercetin dye was markedly higher than the other sensitizing dyes. This is due to a higher intensity and broader range of light absorption of the extract on TiO₂, and the greater interaction between TiO₂ and quercetin in the hypericum extract leads to a better charge transfer. Moreover, quercetin has a shorter distance between the dye skeleton and the point connected to the TiO₂ surface compared with other dyes. This could facilitate an electron transfer from quercetin to the TiO₂ surface and could account for better performance of hypericum extract sensitization. The efficiencies, J_SC and V_OC, of the cells sensitized using quercetin, alizarin and luteolin dyes were better than those reported for DSSCs fabricated using other natural dyes and liquid electrolytes [25, 30–36]. This improvement in cell parameters is due to the enlarged energy level difference between the redox potential of the Fe²⁺/Fe³⁺ and TiO₂ conduction band. In addition, this could be mainly attributed to the nature of the redox couple, because the increase is independent of the dye structure [37].

Zoom In Zoom Out Reset image size

The efficiency of the TiO₂-based cells is correlated with the maximum absorption coefficient of the dye, suggesting that the light-harvesting efficiency may limit the cell current. However, taking the high absorption coefficients of the UV–vis region into account, the cell current would be expected to be much higher, suggesting that the injection efficiency is not optimal. This could be due to several factors, such as a relatively short lifetime of the excited state related to the chemical structures of the dyes, limited electronic interaction, and overlap between the excited states and the conduction band, or unfavourable dye-regeneration kinetics. In addition, the observation of bleaching upon longer immersion in the dye solution indicates that dye-multilayer deposition and polymerization may occur, leading to a decrease in conjugation and, hence, a lowering of the absorption coefficient in the visible spectrum.

The incident photon to charge carrier efficiency (IPCE), sometimes referred to as the 'external quantum efficiency' (EQE), is an important characteristic of a device. In particular, using devices with the same architecture, it is possible to compare the light-harvesting performance of sensitizers. This performance parameter is defined as the number of electrons generated by light in the external circuit divided by the number of incident photons as a function of excitation wavelength, as in the following equation [38]:

$$\begin{equation*} {\rm IPCE}(\lambda )=\frac{{\rm Photocurrent~density}}{{\rm ~Wavelength}\times {\rm Photon~flux}} \end{equation*}$$

$$\begin{equation} \,\,\,\,\,={\rm LHE}(\lambda )\times \varphi _{{\rm inj}} \times \eta _{\rm c} , \end{equation} \tag{ 3 }$$

where LHE(λ) is the light-harvesting efficiency at wavelength λ, φ_inj is the quantum yield for electron injection from the excited sensitizer in the conduction band of TiO₂ and η_c is the efficiency for the collection of electrons. IPCE(λ) is determined experimentally from the measurement of the short-circuit photocurrent density I_SC under illumination at wavelength λ with incident power density P_in using the formula [39]:

$$\begin{equation} {\rm IPCE}(\lambda )=\frac{1240\times I_{{\rm SC}}}{P_{{\rm in}} \times \lambda}. \end{equation} \tag{ 4 }$$

Figure 5 shows the photocurrent action spectra, IPCE(λ), for three DSSCs containing quercetin, alizarin and luteolin as sensitizers, respectively, and ferrocene as the electrolyte solvent. The maximum IPCE values of the cells sensitized with quercetin, alizarin and luteolin dyes are approximately 69%, 63% and 59%, respectively, all obtained at the same wavelength of 580–600 nm. The quercetin-sensitized DSSC shows the highest IPCE.

Zoom In Zoom Out Reset image size

We have fabricated DSSC solar cells using ferrocene liquid as an electrolyte, extracted natural dyes and TiO₂ nanoparticles. Among the plant species investigated, hypericum gave significantly high photocurrent voltages with reasonable efficiency compared with the other plant species tested. This could be due to better interaction between the dye molecules and the surface of TiO₂. To date, raw anthocyanine and betalain extracts have been shown to be the best natural sensitizers, resulting in the generation of monochromatic photon to current conversion yields exceeding 60%. Maximum overall conversion efficiencies under simulated sunlight comparable to natural photosynthesis were increased by 15% using the components described in this study. Identification of appropriate additives for improving V_OC without causing dye degradation may result in further enhancement of cell performance, making the practical application of such systems more suitable for achieving economically viable solar energy devices.