We report a simple and convenient method for the preparation of Ag/TiO2 thin films supported on indium tin oxide, which was achieved by sonochemical deposition of Ag+ on aerosol-assisted chemical vapour deposited TiO2 thin films. Posttreatment was performed on the film by immersion in HCl. The as-prepared composite film was characterised by X-ray diffraction, ultraviolet-visible absorption spectroscopy, Raman spectroscopy, and field emission scanning electron microscopy. The photoelectrochemical measurements and - characterisation showed approximately fivefold increase in photocurrent density generation and approximately sevenfold enhancement in dye sensitiser solar cell (DSSC) conversion efficiency, which was achieved after modification of the TiO2 film with HCl posttreatment and Ag particle deposition. The improved photocurrent density of 933.30 A/cm2, as well as DSSC power conversion efficiency of 3.63% with high stability, is an indication that the as-synthesised thin film is a potential candidate for solar energy conversion applications.

1. Introduction

For decades, titanium dioxide has been substantially researched and has a wide range of applications including solar cells [1], environmental purification [2, 3], gas sensing [4, 5], water photolysis [4, 6], and biomedical devices [7, 8]. The large scope of applications is due to the excellent functionality, low cost, and stability of TiO2 in most environments [6, 9]. In addition, since the discovery of dye sensitised solar cells (DSSCs) by O’Regan and Grätzel in 1991 [10], DSSCs have attracted vast academic and industrial interest due to their high power conversion efficiency, low cost, and ease of fabrication.

The basic concept of the DSSC involves using the highly porous TiO2 thin film as a photoanode, allowing a large amount of sensitising organic dye to be adsorbed. The adsorbed dye will absorb light and photoexcited electrons will transfer to the conduction band of TiO2. Following this, the dye is regenerated by I/I3− [10, 11] through diffusion of electrons to the counter electrode. Thus, high solar conversion efficiency in DSSCs requires not only a large surface area of porous TiO2, for a large amount of dye adsorption, but also optimum construction or good networking between particles for light harvesting and fast electron transportation [1215].

Many methods have been reported for the preparation of TiO2 photoanodes including sol-gel hydrothermal synthesis method [16, 17]; the electrospinning technique [18]; sol-gel [19, 20]; the hydrothermal method [21]; the anodisation/electrodeposition method [1, 22]; the screen printing method [23, 24]. However, to the best of our knowledge, there has not been a study on TiO2 photoanodes for DSSC applications fabricated by aerosol-assisted chemical vapour deposition (AACVD).

Despite many advantages of the AACVD method including simple set-up, low maintenance, and low set-up costs, the method has not previously been utilised for the preparation of TiO2 thin films for DSSC photoanodes. This could be due to the difficulties of controlling the characteristics, such as the porosity, of the produced TiO2 thin films. Poor porosity will limit the dye adsorption and affect the transportation of electrons in the DSSC system. Hence, it is important to devise a way to fabricate an efficient DSSC using the AACVD method. In this report, we propose posttreatment of the TiO2 thin film with HCl, followed by surface modification with silver.

We report, for the first time, performance studies of DSSCs containing a TiO2 thin film prepared by the AACVD method. Photoelectrochemical properties and DSSC performance of the hybrid film were greatly increased following HCl posttreatment and Ag inclusion.

2. Experimental

2.1. Materials

Titanium isopropoxide (TTIP, 98%) was purchased from Acros Organics. Methanol and HCl were purchased from Systerm. Silver nitrate (AgNO3) was purchased from Merck. Indium tin oxide (ITO) conducting glass slides (7 Ω/sq.) were commercially supplied by Xin Yan Technology Limited, China. Deionised water was used throughout sample preparation.

2.2. Deposition of TiO2 Thin Film by AACVD

TiO2 thin films were prepared on the ITO substrates (1 × 1.5 cm) using the in-house AACVD assembly previously described [2]. The ITO substrates were ultrasonically cleaned with acetone, 1 M NaOH solution, and deionised water. TTIP was added to methanol to obtain a 0.1 M solution, and the solution was used to generate an aerosol at room temperature using an ultrasonic air humidifier operating at 60 Hz. Argon gas was passed through the aerosol mist at a flow rate of 200 mL/min to carry the aerosol droplets into the reactor chamber. Depositions were conducted at 450°C for 90 min. The exhaust from the reactor was vented directly into the extraction system of the fume cupboard. The aerosol line was closed toward the end of the experiment and pure argon was allowed to flow. The film was then cooled and stored in air.

2.3. Posttreatment of TiO2 Thin Film

Posttreatment of the TiO2 thin films (P-TiO2) was carried out by immersing the films in 0.1 M HCl. After 30 min, the film was removed from the HCl solution. The film was rinsed with deionised water and dried in an oven at 60°C.

2.4. Preparation of Ag-Loaded TiO2 Thin Film

To obtain the Ag-loaded TiO2 thin films Ag/TiO2 and P-Ag/TiO2, samples TiO2 and P-TiO2 were immersed in 0.001–0.01 M AgNO3 aqueous solution and stimulated using an ultrasonic generator (Kudos) at a frequency of 50 kHz for 30 min. The film was then washed with deionised water and dried in air.

2.5. Fabrication of Dye Sensitised TiO2 and Ag/TiO2 Films with and without Posttreatment

The films were immersed in an ethanol solution of 0.3 mM N719 (Ruthenizer 535-bisTBA, Solaronix) dye for 24 h. After sensitisation, the film was washed with ethanol. The dye absorbed electrode was assembled into a sandwich-type cell with a counter electrode (platinum-sputtered ITO glass). The counter electrode and dye sensitised TiO2 electrode were clamped firmly together, and redox electrolyte (Iodolyte Z-100, Solaronix) solution was introduced into the dye sensitised electrode by capillary action. An active area of 0.5 cm2 was used to measure the cell performance.

2.6. Characterisation

The crystalline phase of the samples was determined via X-ray diffraction XRD (D5000, Siemens), using copper Kα radiation ( Å) at a scan rate of 0.02 degrees s−1. The morphology of the Ag/TiO2 thin film was examined using field emission scanning electron microscopy FESEM (FEI Nova NanoSEM 400, operated at 10.0 kV). Optical absorption properties in the spectral region 190–900 nm were assessed using a UV-Vis spectrophotometer (Thermo Scientific Evolution 300). Raman spectra were collected using a Renishaw 2000 system with an argon ion laser emitting at 514.5 nm.

2.7. Cyclic Voltammetry

Electrochemical characterisation, to obtain linear sweep voltammetry (LSV) plots and chronoamperometry plots, was performed in the three-electrode VersaSTAT 3 electrochemical analyser from Princeton Applied Research. The system was equipped with an electrochemical cell consisting of Ag/AgCl (in 3 M NaCl) as the reference electrode and a platinum wire as the counter electrode. 0.5 M KCl was employed as a supporting electrolyte. The time dependence of the photocurrent was obtained under illumination from a 150 W Xenon arc lamp (Newport, model 69907), using an AM1.5G filter with a manual shutter.

2.8. DSSC Conversion Efficiency Measurement

The photovoltaic performance of the DSSCs was recorded using the VersaSTAT 3 electrochemical analyser from Princeton Applied Research. A 150 W Xenon arc lamp simulated solar illumination with the use of an AM1.5G filter. The intensity of illumination, measured using a fiber optic spectrometer (Avaspec-2048), was 20 mW/cm2.

3. Results and Discussion

By analysis of XRD patterns (Figure 1), it is observed that most of the peaks caused by the ITO glass slide are no longer visible after the AACVD process, indicating successful deposition of TiO2 onto the ITO glass slide (Figure 1(a)). The TiO2 films, TiO2 and P-TiO2 (Figures 1(b) and 1(d)), are composed of mixed anatase and rutile phases, which agrees with reference patterns JCPDS 83-2243 and 21-1276, respectively. The diffraction peak at 25.3° corresponds to the anatase phase of TiO2. The peaks at 27.4°, 36°, 39.1°, 41.2°, 44.1°, 54.3°, 56.7°, 62.74°, 64.1°, 68.9°, and 70° agree well with the rutile phase of TiO2. The diffraction peaks corresponding to (200), (210), and (310) peaks of TiO2 are very close to the peaks of Ag (JCPDS 41-1104) at (111), (200), and (220). It is therefore difficult to differentiate the Ag signals from the TiO2 signals due to the low Ag content in Ag/TiO2 and P-Ag/TiO2 (Figures 1(c) and 1(e)).

To further evaluate the phases in the prepared films, Raman spectroscopy was performed in the range 100–1000 cm−1, and the results are shown in Figures 2(a) and 2(b). The anatase TiO2 phase was observed at 143, 194, 398, 512, and 632 cm−1. The rutile TiO2 phase was detected at 449 and 611 cm−1. However, signals related to Ag particles were not identified for samples with and without HCl posttreatment, due to the relatively low concentration of Ag loaded onto the TiO2 thin film and its weak Raman scattering power [25]. An interesting observation is that the peak intensities are reduced with deposition of Ag, but the position of the peaks remains the same. This indicates that there is an interaction between Ag and TiO2 that affects the Raman resonance of TiO2 [26].

The UV-Vis spectra of the thin films are shown in Figure 3(a). It can be clearly seen that the addition of Ag onto TiO2 caused a red shift of the wavelength. This has the effect of lowering the band gap energy of TiO2. An estimation of the band gap energy (eV) was determined using a Tauc plot [() versus ], where is a constant, is the power of the exponent ( for a direct allowed transition and /2 for an indirect allowed transition), and is the incident photon energy. The resulting plots of direct allowed transitions, with an extrapolation to the -axis, for TiO2 and Ag/TiO2 are shown in Figures 3(b) and 3(c). The band gap energy of TiO2 decreased from 3.29 eV to 2.90 eV following deposition of Ag. Some studies report that the presence of Ag on a TiO2 thin film results in more excitation due to surface plasmon resonance [2729]. Moreover, Ag assists in reducing the band gap energy and retarding the charge recombination [25, 28, 30]. These improvements directly enhance the photocurrent. Modification of the TiO2 thin film has possible applications in dye sensitised solar cells (DSSCs) [3133], photocatalysts [34], hydrogen production [35], and sensing [36, 37].

The mechanism for the growth of Ag on the TiO2 surface is illustrated in Figure 4(a). The sonochemical growth of Ag particles consists of two steps: nucleation and growth. In the nucleation step, the TiO2 acts as the seed for the Ag particle. The collapse of the bubbles induced by the sonication process will result in high energy, exciting the electrons in the conduction band of TiO2 crystals [38]. The generated electrons will accumulate on the surface of TiO2 and attract the positively charged Ag+ due to the electrostatic interaction between the oppositely charged ions. The Ag+ ion is then reduced to an Ag particle. As the sonication process proceeds, the excited electrons will be generated continuously. The Ag seed will act as electron storage and attract more Ag+, hence the growth of Ag particles. The TiO2 and Ag/TiO2 films are shown in Figures 4(b) and 4(c). The physical appearance of the films on the ITO glass is translucent. The thicknesses of the as-deposited films are fairly uniform between 600 and 700 nm, measured using a surface profiler. The TiO2 films are strongly attached to the ITO glass, as they are able to endure the Scotch tape test. They also resist common solvents and do not dissolve in 1 M HCl or 1 M NaOH solutions.

The microscopic morphology of the thin films was studied using FESEM. Figures 5(a) and 5(b) show images of pure TiO2 and TiO2 with HCl posttreatment, respectively. The post-treated film appears pyramidal. Figures 5(c) and 5(d) show images of the Ag/TiO2 film, with 0.005 M Ag concentration, with and without HCl posttreatment. Interestingly, after the addition of Ag, the microstructure of TiO2, which initially appears pyramid-like, changes significantly to consist of numerous nanoparticles resembling stone blocks, attached to the pyramidal structure. No significant changes in morphology were observed for the film with and without HCl posttreatment. The presence of Ag on TiO2 was confirmed by EDX analysis, as shown in Figures 5(e) and 5(f). The untreated Ag/TiO2 thin film contains not only Ti, O, Ag, and C, but also Na. This implies that some chemical reactions occur at the glass substrate/thin film interface, and that sodium ions migrate from the glass substrate into the thin film [39, 40]. After HCl posttreatment, the Na peak disappears. The C peak, found in the spectra of both treated and untreated films, may come from the residual carbon of the precursor solution such as unhydrolysed alkoxide or hydrocarbon groups from the surrounding environment, possibly created during the sample handling process. It is observed that the C content decreases after HCl posttreatment. This is most likely due to the removal of the Na. Na+ will normally absorb CO2 in air to form sodium carbonate and cause an increase in the amount of carbon. However, sodium carbonate will dissolve easily in HCl solution to form sodium chloride, CO2, and H2O. The residue sodium and chloride were then removed by rinsing with deionised water and ethanol. The areas of bright contrast on the element mapping correlate with the Ti, Ag, and O signal maps for P-Ag/TiO2 shown in Figure 5(g).

LSV was conducted to study the redox properties of the samples. Figures 6(a) and 6(b) show the LSV voltammograms of the samples in darkness and under solar illumination. The results show that the photocurrent increases linearly with the potential in the low potential region and then remains constant. Oxidative peaks at around 0.22 V for the Ag/TiO2 and P-Ag/TiO2 films in the dark and under solar light suggest the presence of Ag with an oxidation state of O [41], corresponding to the reaction Ag→Ag++e. This suggests that the Ag on the surface of Ag/TiO2 can be oxidised electrochemically [42]. Hence, we can conclude that Ag particles grow on TiO2 thin films, not Ag+ or Ag2+ ions. From the voltammogram, it could be observed that the current for TiO2 and Ag/TiO2 films, with and without HCl posttreatment, was enhanced under solar illumination compared to the current in darkness.

To examine the photocurrent activity of the films, the photocurrent profiles were measured against time. First, the influence of the loading concentration of Ag+ on the photocurrent density was studied. As shown in Figure 7(a), the optimum concentration of Ag+ is 0.005 M. This result indicates that the Ag loading concentration has an important effect on the photoelectrochemical activity as it generates free carriers or electron trapping sites. Increasing the concentration of Ag+ to 0.01 M would decrease the photocurrent density because of the negatively charged Ag nanoparticles on the TiO2 film that result from the Schottky barrier. The Schottky barrier is the potential barrier formed at the Ag/TiO2 interface. It becomes dominant and retards the electron transport in the conduction bands. Consequently, the empty photoholes on the TiO2 surface may become trapped by the high extent of negatively charged Ag particles before they have a chance react with others [41]. A decrease in the concentration would cause the photocurrent density to decrease due to the smaller amount of Ag acting as the electron acceptor [43].

The current responses of the TiO2 and Ag/TiO2 films at 0 V potential, with and without HCl posttreatment, are shown in Figure 7(b). The photocurrent densities for TiO2 and Ag/TiO2 are 178.63 A/cm2 and 430.33 A/cm2, respectively, while the HCl post-treated films P-TiO2 and P-Ag/TiO2 have photocurrent density of 197.18 A/cm2 and 933.30 A/cm2, respectively. By depositing Ag onto the TiO2 thin film without HCl posttreatment, the photocurrent increased approximately 2.5-fold. The photocurrent density enhancement of as high as fivefold is achieved when Ag is deposited onto the TiO2 thin film after posttreatment. These significant enhancements of photocurrent density are most probably due to the posttreatment of the TiO2 thin film with HCl, which is essential to remove impurities such as Na. On the other hand, in presence of Ag, the TiO2 thin film can effectively transfer photoelectrons and restrict the recombination of photoexcited electron-hole pairs.

The photocurrent responses of Ag/TiO2 and P-Ag/TiO2 under solar illumination are shown in Figures 8(a) and 8(b). As illumination began, there was a significant increase in the photocurrent and a photocurrent spike was observed. This phenomenon is likely due to the surface recombination process [4448]. After the sudden increase in photocurrent after illuminating the film, a fast decay is observed which is likely attributed to accumulation of photogenerated holes on the surface of Ag/TiO2 films which in turn allows their recombination with photogenerated electrons [45]. No current was produced when the light was turned off, even at the highest applied potential 1 V. There were excellent reproducibility and repeatability of results. This indicates that there is no photodegradation of the sample and the response is not affected by photooxidisable surface species [49].

The operation principle of the DSSC is based on photoexcitation of the dye, followed by electron injection into the conduction band of the TiO2 surface. Figure 9(c) shows a schematic diagram of the operation principles of a DSSC based on an Ag/TiO2 thin film. During solar irradiation, the dye absorbs incident light and promotes electrons to the excited state. The excited electron is injected into the conduction band of the TiO2 particle. The dye is then oxidised by receiving the electron from the electrolyte through the redox system and is ready to be used again. The electrolyte itself will regenerate via the platinum counter electrode by an electron passed through the external circuit. In our study, the deposition of Ag onto the TiO2 thin film (Figures 9(a) and 9(b)) resulted in a change in the Fermi energy level; thus the electron in the conduction band of TiO2 can be effectively captured by the Ag until the Fermi level equilibrium is obtained. This migration of generated electrons to Ag increases the lifetime of the holes and suppresses the electron-hole recombination, which is beneficial to photoelectrochemical activity. Without the threat of electron-hole recombination, the valence holes will then oxidise water and form hydroxyl radicals. These hydroxyl radicals will eventually react with water and degrade organic matters efficiently.

To examine the improvement in the efficiency as a result of the proposed posttreatment with HCl and surface modification by deposition of Ag, we measured the density-voltage (-) curves, which are shown in Figure 10. From analysis of the - curves, the critical parameters of the cell, such as open circuit voltage (), short circuit photocurrent density (), and the cell’s fill factor (FF), can be obtained from the following equation [10]: where and are the voltage and current densities, respectively, at maximum power output, respectively. The cell’s overall energy conversion efficiency () is estimated as follows: where is the intensity of the incident light. The results are summarised in the inset of Figure 10. The power conversion efficiency () of the Ag/TiO2 film increased to 0.93%, compared with that of the TiO2 thin film (0.50%). Following HCl posttreatment, the power conversion efficiency drastically increased to 2.81%, and the conversion efficiency of P-Ag/TiO2 improved to 3.60%. The improvements are approximately twofold after deposition of Ag onto the TiO2 thin film, approximately sixfold after HCl posttreatment, and approximately sevenfold after HCl posttreatment and Ag deposition. In previous studies, it was confirmed that sodium ions are detrimental to the photoactivity. Even at low concentrations, sodium could produce surface and bulk recombination centres of photogenerated electron-hole pairs and hence retard the photoefficiency of TiO2 [50]. Therefore, after treatment with HCl to remove the sodium, the conversion efficiency was significantly enhanced. All Ag doped DSSCs exhibited better conversion efficiency. Moreover, the film subjected to both HCl posttreatment and Ag deposition showed the highest power conversion efficiency, almost sevenfold higher than that of the untreated film. This may be due to the synergetic effects of the removal of the sodium ion and the incorporation of silver.

4. Conclusion

We first reported on the DSSC performance of mixed rutile-anatase phase TiO2 thin films produced by the AACVD method. This was followed by deposition of Ag nanoparticles onto the film through a simple sonochemical method. It has been shown that the photoelectrochemical properties and DSSC performance of the hybrid film were greatly enhanced following HCl posttreatment. The photocurrent measurements revealed good photocurrent response for the Ag/TiO2 thin film. The photocurrent density of 933.30 A/cm2, with high stability, is an indication that the as-synthesised thin film is a potential candidate for solar energy conversion applications. A conversion efficiency of 3.63% has been achieved, with , , and FF of 1.95 mA/cm2, 0.87 V, and 0.43, respectively, owing to the reduced band gap energy and retarded charge recombination, which were contributing factors to the improved photoelectrochemical and DSSC performance.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.


This work was supported by the University of Malaya Research Grant (RG197/11AFR), Science Fund from the Ministry of Science, Technology and Innovation (06-01-04-SF1513), and a High Impact Research Grant from the Ministry of Higher Education of Malaysia (UM.C/625/1/HIR/MOHE/05).