Introduction

Dye-sensitized solar cells (DSSCs) are one of the most promising candidates for future green energy alternative due to their facile, low-cost, and environmentally friendly fabrication process (O’Regan and Gratzel 1991; Zhang and Cao 2011; Duong et al. 2013; Xu et al. 2013). The DSSC assembly consists of dye-fabricated semiconductor oxide electrode sandwiched with platinum electrode in face-to-face manner; the space between the electrodes is filled with electrolyte-containing redox couple. Sandwich-type DSSC assembly has four major parts: oxide semiconductor, photosensitizers, redox couple, and counter electrode. By altering any one of these moieties, the cell parameters such as conversion efficiency, short-circuit current, open-circuit potential, and fill factor parameters can be tuned. Oxide semiconductors are preferentially used in solar cell because of their high stability against photocorrosion on optical excitation in the band gap (Hagfeldt and Grätzel 1995). The large band gap (>3 eV) of the oxide semiconductor is desired in DSSCs for the transparency of the semiconductor electrode for harvesting the solar energy into large spectrum. TiO2, ZnO, WO3, Fe2O3, SnO2, Nb2O5, and Ta2O5 (Hagfeldt and Grätzel 1995) are used as porous nanocrystalline structure in DSSCs. Out of these oxides, TiO2 is still in demand due to its better performance as high surface area for dye penetration and attracted major attention for their use as semiconductor in dye-sensitized solar cells (Hagfeldt and Grätzel 1995; Kalyansundaram and Grätzel 1998; Kong et al. 2007; Gratzel 2003; Polo et al. 2004; Hagfeldt et al. 2010). However, the natural drawbacks of the conventional TiO2 nanoparticle photoanode are numerous grain boundaries existing in the film and the slow electron mobility of TiO2 as compared with other materials such as ZnO and SnO2 (Hagfeldt and Grätzel 1995; Kalyansundaram and Grätzel 1998; Kong et al. 2007; Gratzel 2003). These drawbacks could result in a high recombination loss and slow electron transport, preventing further increment of the power conversion efficiency (PCE) for TiO2-based DSSCs. The ZnO oxide semiconductor can also be fabricated as photoanode due to its similar band gap level and higher electron mobility (~115–155 cm2 V−1 s−1) with respect to TiO2 (Zhang et al. 2009; Xu and Sun 2011; Hu et al. 2012; Sakai et al. 2013).

Recently, ZnO-based DSSCs have triggered considerable interest because ZnO has similar band gap level and higher electron mobility with respect to TiO2 (Polo et al. 2004; Hagfeldt et al. 2010; Zhang et al. 2009; Xu and Sun 2011). Among the various ZnO nanostructures, ZnO nanowire (NW) arrays have been suggested to be a promising photoelectrode configuration because they can provide a direct pathway for rapid electron transport (Hu et al. 2012; Sakai et al. 2013; Meng et al. 2005).

Many investigations have been carried out to synthesize ZnO nanoparticles via different techniques (Meng et al. 2005; Umar et al. 2005; Pan et al. 2001; Wahab et al. 2007a, b, 2008; Meulenkamp 1998; Spanhel and Anderson 1991; Zhang et al. 2002; Nyffenegger et al. 1998; Brinker and Scherer 1990) such as sol–gel technique (Meulenkamp 1998; Spanhel and Anderson 1991), hydrothermal (Zhang et al. 2002), and electrochemical techniques (Nyffenegger et al. 1998). Among these techniques, sol–gel technique is very attractive because it is relatively easy to achieve and allows us to tailor the morphology of the particles by controlling the rate of hydrolysis and condensation reactions (Brinker and Scherer 1990). Among all methods, precipitation and sol–gel technique provides suitable control of nucleation, aging, and growth of particles in solution. The direct precipitation is also one of the simple and cost-effective methods for bulk production of materials (Wahab et al. 2008). In this method, particle growth owes to interact between different aqueous solutions, and therefore very small particles are formed. Tiny size particles with lower solubility product dissolve and re-precipitate on the surface of larger particles in solution; consequently agglomeration takes place in solution as the particles clog together to minimize surface energy (Meulenkamp 1998).

In this paper, a simple and cost-effective precipitation method was used to synthesize ZnO nanoparticles using different surfactants (triethylamine, diethylamine, PEG 6000, and PEG 8000) as a precipitating agent. The synthesized ZnO powders were employed as photoanode sensitized with N719 dye in DSSC and measured the photovoltaic parameters (short-circuit current, open-circuit voltage, fill factor, and cell efficiency).

Experimental

Materials

Zinc acetate (Merck), triethylamine (Merck), diethylamine (Merck), polyethyleneglycol 6000 (Merck), polyethyleneglycol 8000 (Merck), and Ethanol (Merck) were used as received for preparing ZnO powder. For electrolytic solution, acetonitrile (Merck) was used as the medium of electrolyte; LiI (Aldrich) and I2 (BDH) were used as redox couple in electrolyte without any further purification. Platinum catalyst (T/SP) as counter electrode, the sealing agent (SX1170-60, 50 μm) for packing the assembly, and N719 dye used as photosensitizer in this experiment, all were purchased from Solaronix. Conductive glass plate (15 Ω/cm2) was obtained from Pilkington, USA.

Preparation of ZnO powder

Zinc acetate (5 mmol, 1.0 g) was dissolved in ethanol (50 mL) with constant stirring for 1 h at 80 °C. Subsequently, triethylamine (2 mmol, 0.5 mL), diethylamine (2 mmol, 0.36 mL), PEG 6000 (0.2 g), and PEG 8000 (0.2 g) were added into this solution with constant stirring for 8 h at 80–90 °C to obtain a white precipitate which was filtered and washed with water and ethanol. The as-prepared powder was dried in air and annealed at 450 °C and coded as Z-1 (triethylamine), Z-2 (diethylamine), Z-3 (PEG 6000), and Z-4 (PEG 8000) (Scheme 1).

Scheme 1
scheme 1

Flow chart of synthesis procedure of ZnO powders

Full size image

Preparation of ZnO electrode (photoanode) and counter electrode

The ZnO photoanode was prepared by making the paste of ZnO powders with ethanol. The paste was spread on conductive glass plate using doctor’s blade technique and annealed at 450 °C for half an hour in air. A thin film of ~5 μm thickness was obtained. The photosensitizer (N719 dye) was adsorbed onto the ZnO surface by dipping the ZnO thin film in an ethanolic solution of N719 dye for 8 h. The non-adsorbed dye was washed off with anhydrous ethanol. The platinum counter electrode was prepared by deposition of Pt catalyst (T/SP paste, Solaronix SA) using screen printing method on conductive glass plate and annealing at 400 °C for half an hour in air.

DSSC assembly

For sandwich-type DSSC assembly, dye-modified ZnO photoanode and platinum-coated counter electrode were positioned over each other in face-to-face sandwich manner leaving the space for making contact, to connect with external load. This assembly was sealed from three sides using the spacer leaving one side open through which the electrolyte solution (0.05 M iodine, 0.5 M LiI and 0.5 M 4-tert. butylpyridine in acetonitrile) was injected into the space between the electrodes. After electrolyte injection, the open side of the cell assembly was also sealed. Copper wires were fixed on both the electrodes using silver paste and araldite for getting electrical contact.

Instrumentation and characterization

The synthesized ZnO powder was characterized by pXRD using X-ray diffractometer (XRD, Bruker AXS model D-8 equipped with a monochromator and Ni-filtered Cu Ka radiation), field emission scanning electron microscope (FESEM, Hitachi S-4800), and UV–Vis (Shimadzu UV–Vis–NIR spectrophotometer-Model UV-3600).

The photoelectrochemical performance, including the short-circuit current (Jsc, mA cm−2), open-circuit voltage (Voc, V), fill factor (ff), and overall energy conversion efficiency (η) was determined from the J–V curve obtained using a digital Keithley 236 mm under an irradiation of white light from a 1,000 W/HS Xenon arc lamp with a 100 mW cm−2 light intensity.

The incident photon-to-electron conversion efficiency (IPCE) measured using a 300 W Xe lamp light source joined to a monochromator (Oriel). A reference Si photodiode calibrated for spectral response was used for the monochromatic power-density calibration.

The electrochemical impedance spectroscopy (EIS) was performed on a CH 660C electrochemical analyzer (CH Instruments, Shanghai, China) in a two electrode configurations. The photoanode was used as a working electrode and the Pt electrode as a counter electrode. The electron transport properties were investigated using electrochemical impedance spectroscopy (EIS) with 10 mV alternative signal in the frequency range of 10−2 to 105 Hz.

Result and discussion

UV–Vis spectrum and band gap energy

The absorption spectrum of the ZnO nanostructures is shown in Fig. 1. The band position at 373, 372, 365, and 366 nm corresponding to flakes-like (Z-1), hexagon-like (Z-2), particle-like (Z-3), and flower-like (Z-4) ZnO nanostructures was observed. The blue shift in the ZnO nanostructures, comparing to bulk ZnO (380 nm) is observed due to the size quantization effect (Koch et al. 1985). It is clear that the absorption edge systematically shifts to the lower wavelength or higher energy with decreasing size of the nanoparticle. This definite and orderly shift in the absorption edge is due to the quantum size effect. The flakes-like (Z-1) and hexagon-like (Z-2) ZnO have larger particle size than the others, which was reflected in the absorption spectrum.

Fig. 1
figure 1

UV–Vis spectra of different ZnO powders

Full size image

X-ray diffraction pattern

Figure 2 shows the XRD pattern of ZnO nanocrystals. The diffraction peaks indicate the nanocrystalline nature with scattering angles (2θ) of 31.72, 34.42, 36.24, 47.49, 56.53, 62.81, 66.28, 67.86, and 68.99 corresponding to the reflection from 100, 002, 101,102, 110, 103, 200, 112, and 201 crystal planes, respectively. The XRD pattern is identical to the hexagonal phase with Wurtzite structure with space group (C6V = P63mc) and unit cell parameters a = b = 3.253 Å and c = 5.2 Å (JCPDS File No. 80-0075).

Fig. 2
figure 2

X-ray diffraction pattern of the different ZnO powders (Z-1 to Z-4); JCPDS file no. 80-0075)

Full size image

FTIR spectra

Figure 3 shows the spectrum of the ZnO powders (Z-1 to Z-4) after annealing (450 °C). Only ZnO vibrations at ca. 490 cm−1 are observed indicating that the nanopowders consists of pure ZnO without carbon rests. The large absorption bands at 3,200–3,600 cm−1 are assigned to the OH-stretching frequencies of alcohol.

Fig. 3
figure 3

FTIR spectra for different ZnO powders

Full size image

Field emission scanning electron microscopy

The FESEM images of the ZnO nanostructures are shown in Fig. 4. The flakes-like, hexagon-like, particle-like, and flower-like morphologies have been obtained by different surfactant-assisted hydrothermal syntheses. The FESEM image of Z-2a and Z-2b shows the hexagons having small particles grow on the surface which helps to increase the surface area for sufficient dye penetration.

Fig. 4
figure 4

Typical FESEM images of ZnO nanostructures Z-1(flakes-like), Z-2 (hexagon-like), Z-3 (particle-like), and Z-4 (flower-like) at different magnifications

Full size image

The evaluations of surface area available for dye adsorption for the ZnO powders were investigated by measuring the loading of N719 via desorption experiments (Jennings et al. 2008). Table 1 shows the dye-loading amount for the ZnO powders prepared using different surfactants. The maximum dye loading was shown by Z-3 (particle-like) and Z-4 (flower-like) samples than the Z-1 (flakes-like) and Z-2 (hexagons), due to its greater surface area.

Table 1 The dye-loading amount and photovoltaic parameters derived from the J–V curves shown in Fig. 5
Full size table

Current–potential (J–V) curve

For obtaining J–V curve, the data were recorded under solar irradiation by changing the external load from zero (short-circuit conditions) to infinite load (open-circuit conditions). The DSSC assembly was packed using ZnO powders (Z-1 to Z-4) as photoanode sensitized by N719 dye to evaluate their photovoltaic performance (maximum output power, fill factor, and power conversion efficiency) at 100 mW cm−2 light intensity. The maximum power (Pmax) was obtained by choosing a point on J–V curve corresponding to which the product of current (Jmax) and voltage (Vmax) is maximum. The power conversion efficiency (η) and fill factor (FF) were evaluated using the following relations and

$${\text{FF}} = \frac{{J_{ \hbox{max} } \left( {A\;{\text{cm}}^{ - 2} } \right) \times V_{ \hbox{max} } \left( V \right)}}{{J_{\text{SC}} \left( {A\;{\text{cm}}^{ - 2} } \right) \times V_{OC} \left( V \right)}},$$
(1)
$$\eta \left( \% \right) = \frac{{J_{\text{SC }} \times V_{\text{OC}} \times FF}}{{I_{\text{inc}} \left( {W\;{\text{cm}}^{ - 2} } \right)}} \times 100,$$
(2)

here, Jsc, Voc, and Iinc are short-circuit photocurrent, open-circuit potential, and intensity of incident light, respectively. The photocurrent–potential curves for the cell assemblies are shown in Fig. 5.

Fig. 5
figure 5

Photocurrent-potential curves of the DSSCs fabricated with different ZnO powders as photoanode sensitized by N719 dye, illuminated with 100 mW cm−2 light intensity

Full size image

The photovoltaic performances have been summarized in Table 1 for the cell fabricated using prepared ZnO (Z-1 to Z-2) as photoanode modulated with N719 dye in DSSC assembly. The results show the marked efficiency difference between the cell assemblies fabricated by using different ZnO powders as photoanode. In comparison to Z-1 (1.06 %), Z-4 shows almost double efficiency (2.48 %) with the use of polyethyleneglycol 8000 as surfactant having flower-like morphology. The efficiencies difference between the ZnO powders is mainly due to its difference in surface area. The greater surface area enhances the greater dye loading which facilitates the improvement in cell efficiency.

Incident photon-to-current conversion efficiency (IPCE)

To determine the ratio of photons that generate electrons in the external circuit to incident photons of monochromatic light is incident photon-to-current efficiency (IPCE) or external quantum efficiency (EQE) for the cell. To quantify the IPCE values, the short-circuit photocurrent (Jsc) of the dye-sensitized cell was determined in monochromatic light from 400 to 700 nm. From the values of Jsc and the intensity of the corresponding monochromatic light (Iinc), the IPCE was calculated at each excitation wavelength (λ) using the following relation:

$${\text{IPCE}} = \frac{{1240{\kern 1pt} \; \times \;J_{\text{photo}} \;(A\;{\text{cm}}^{{{\mathbf{ - }}2}} )}}{{\lambda \;({\text{nm}})\; \times I_{\text{inc}} \;(W\;{\text{cm}}^{{{\mathbf{ - }}2}} )}}\; \times \;100.$$
(3)

Figure 6 represents the IPCE curve for the cell fabricated using synthesized ZnO powders sensitized by N719 dye, and the values are listed in Table 1. The maximum IPCE obtained was 46 % with cell fabricated by flower-like ZnO as photoanode in DSSCs.

Fig. 6
figure 6

Incident photon-to-current conversion efficiency (IPCE vs. λ) plots for different ZnO-N719/electrolyte/Pt DSSCs assembly

Full size image

EIS study

For investigating electron transport properties of ZnO/dye/electrolyte interface in DSSCs, the EIS has been used. In EIS, the study of semicircle in the middle frequency region is associated with the electron/charge transfer at ZnO/dye/electrode interface (Wang et al. 2005; Lee et al. 2008; Yanagida et al. 2009).

The EIS Nynquist plots of the DSSCs based on the ZnO/dye/electrolyte interface are shown in Fig. 7. According to the EIS spectra shown in Fig. 7, the charge-transfer resistance for ZnO/dye/electrolyte interface in light decreases with the decreasing semicircle diameter. The charge-transfer resistance decreases for the system using different ZnOs as photoanode in the following trend: Z-4 > Z-3 > Z-2 > Z-1 at ZnO/dye/electrolyte interface.

Fig. 7
figure 7

Electrochemical impedance spectra of DSSCs made of different ZnO as photoanode in light sensitized by N719 dye in the frequency range of 10−2 to 105 Hz

Full size image

Conclusion

In summary, different morphologies of ZnO powders have been synthesized successfully via surfactant-assisted precipitation routes. The different surfactants used in the synthesis provided different architectures for the ZnO as well as offered different surface areas. The pXRD for all ZnO nanostructure reveals the formation of single crystalline hexagonal phase. The optical study shows the blue shift than the bulk ZnO. The photovoltaic performances for the cell using these ZnO samples as photoanode was showing the cell efficiency trend Z-4 > Z-3 > Z-2 > Z-1. The Z-4 shows the highest efficiency than the others as it offers greater surface area for greater dye loading. The difference in efficiency is due to different surface areas, as particle size decreases surface area increases and dye loading also increases. The above efficiency trend was also supported by impedance study (EIS spectra) which shows the decrease in resistance in light for Z-4 sample which is more acute than the others. From this study, we can conclude that by modifying synthesis procedure one can get much better efficiency in future. From this study, we concluded that by varying surfactant we can easily tailor the shape, size, and morphology of ZnO and improve its performance as photoanode in DSSCs.