Dye-sensitized solar cell (DSSC) using multiwalled carbon nanotube/titanium dioxide (MWCNT/TiO2) was successfully synthesized using sol-gel method. In this method, it has been performed under various acid treatments MWCNT concentration level at (a) 0.00 g, (b) 0.01 g, (c) 0.02 g, and (d) 0.03 g. Atomic force microscopy (AFM) was used to study surface roughness of the MWCNT/TiO2 thin films. The average roughness results for 0.00 g, 0.01 g, 0.02 g, and 0.03 g were 10.995, 18.308, 24.322, and 25.723 nm, respectively. High resolution transmission electron microscopy (HR-TEM) analysis showned the inner structural design of the MWCNT/TiO2 particles. The TiO2 nanoparticles covered almost all the area of MWCNT particles. Field emission scanning electron microscopy (FESEM) gave the morphological surface structure of the thin films. The thin films formed in good distribution with homogenous design. The DSSC with MWCNT/TiO2 electrode containing 0.03 g MWCNT were resulted in the highest efficiency of 2.80% with short-circuit current density of 9.42 mA/cm2 and open-circuit voltage of 0.65 V.

1. Introduction

Dye-sensitized solar cell (DSSC) is considered as a relatively new type of solar cell being discovered, in 1991, by Grätzel et al. DSSC shows greater promise compared to Si solar cells due to low cost, environmental friendly and simple manufacturing process. DSSC has attracted a lot of attention worldwide. Nevertheless, Grätzel’s cell has a solar conversion efficiency of ~13% [1], which is significantly lower than that of Si solar cells. To improve the performance of DSSC devices, a number of aspects are considered. Electron transport across a TiO2 electrode is one of the most important factors affecting the conversion efficiency of DSSC; the greater electron mobility is, the higher the DSSC efficiency will be [2]. On the other hand, charge recombination processes generally inhibited injected electrons from TiO2 to the conducting glass substrate, thus decreasing the performance of DSSC. Therefore, the rapid photo induced electron transport in the working electrode, TiO2, and the suppression of charge recombination processes can ensure a higher conversion efficiency of DSSC [3, 4].

In recent years, multiwalled carbon nanotube MWCNT has attracted considerable attention worldwide due to its excellent mechanical properties and electrical and thermal conductivity making it a high potential candidate in various applications, for example, field emission display [5], photo catalysis, photovoltaic devices, and DSSC [6]. Several researches [7, 8] reported the incorporation of MWCNT within nanocrystalline TiO2 working electrodes to enhance the solar energy conversion efficiency of DSSC. Furthermore, some studies [9] have revealed that the better performance of DSSC fabricated using MWCNT/TiO2 electrodes is because of the higher electron mobility at the electrodes than that of conventional TiO2 electrodes. This resulted in a higher short-circuit photocurrent () of DSSC. However, the performance of DSSC dropped when high MWCNT content was applied to the MWCNT/TiO2 electrode, possibly due to a severe aggregation of MWCNT in the nanocomposite electrode [10, 11].

In this study, we aim to synthesize and fabricate MWCNT/TiO2 dye-sensitized solar cell using sol-gel method with different concentrations of MWCNT. The different MWCNT concentrations were the main key in order to get better photovoltaic efficiency. We investigate the MWCNT concentration effect on the solar cell in terms of their photo-conversion efficiency performance. The samples were analyzed using FESEM, TEM, AFM, and IV curve analysis to see the effect of MWCNT doped into TiO2 photoelectrode.

2. Methodology

2.1. Materials

Titanium (IV) tetraisopropoxide (TTIP) (98%) was purchased from Sigma-Aldrich, Belgium, and used as the main material. MWCNT, 98% carbon basis with length of 6 to 13 nm, was purchased from Sigma-Aldrich, USA. The MWCNT will go through the acid treatment process before being used. The MWCNT will be sonicated for 2 h in a beaker containing 50 mL of concentrated nitric acid and boiled at 90°C on a hotplate. The MWCNT powder achieved after filtering was washed several times using distilled water to remove residual acid and dried in oven for 24 h. In this study, other chemicals were used as ethanol anhydrous I2 (99.5%), ruthenium 620-1H3TBA dye obtained from Solaronix SA, adult MPN-100 purchased from Solaronix SA, nitric acid (95%), and electrode substrate fluorine tin oxide (FTO) glasses, 30 Ω, obtained from Solaronix SA and used as received.

2.2. Preparation of MWCNT/TiO2 Solution

TiO2 nanoparticles are produced from titanium (IV) tetraisopropoxide that acts as a precursor solution. The TTIP solution mixed with anhydrous ethanol solution in the ratio of (0.1/2) using magnetic bar stirrer for 30 min. The acid treatment MWCNT powder with different weights of (a) 0.00 g, (b) 0.01 g, (c) 0.02 g, and (d) 0.03 g was added into the TiO2 colloidal solution paste and dispersed using an ultrasonic horn machine for 60 min. MWCNT/TiO2 nanocomposite stirred vigorously for another 2 h to get homogeneous solution. The MWCNT/TiO2 paste coated onto the fluorine tin oxide (FTO) conductive glass using doctor-blade technique to generate 0.25 cm2 active area is followed by evaporation of ethanol in air at room temperature for a few minutes. All the MWCNT/TiO2 thin films were annealed in a dry furnace at 300°C for 30 min forming noncrack and uniform thin film electrode. Figure 1 shows the flowchart of this preparation.

2.3. Fabrication of Dye-Sensitized Solar Cell

The annealed MWCNT/TiO2 electrode thin films were then being immersed in 0.5 M N719 ruthenium dye for 1 day to make sure that all the MWCNT/TiO2 particles were covered with the N719 dye particle. The electrodes were then sandwiched with another electrode glass called counter electrode which was covered with platinum (Pt) thin film on top of the FTO glass. Both electrodes separated using parafilm barrier to place electrolyte between the electrodes. DSSC based MWCNT/TiO2 nanocomposite is completed and is ready to be analyzed. Figure 2 shows the flowchart of this preparation.

2.4. Characterization

Field emission scanning electron microscopy (FESEM, Zeisz Supra-15KV) was used to analyze the morphological structure of the MWCNT/TiO2 thin film. Transmission electron microscopy (TEM, CM12 Philips, 1990) is a device used to see the internal structure of a sample. This microscopy has expanded until 660,000x with resolution around 100 nm at 120 keV. Atomic force microscopy (AFM) analysis showed the roughness of thin film in 3D images. Photovoltaic analysis (GAMRY Instruments G300) was assessed as simulated AM 1.5 xenon illumination with a 100 mW/cm2 light output. All experimental results were characterised by good repeatability.

3. Results and Discussions

3.1. Field Emission Scanning Electron Microscopy (FESEM)

By referring to other researches and studies [12], we expect to improve the solar cell energy conversion efficiency with the presence of the MWCNT as favorable electrical conductivity on the metal oxide nanocomposite. MWCNT nanoparticles can extend the electron lifetime and enhance the electron transport rate in the photovoltaic metal oxide electrode. In addition, we notice that the MWCNT nanoparticles that go through acid treatment process in the concentrated nitric acid could produce MWCNT with terminal COOH group. This phenomenon can improve the solar cell electron collection due to better interconnection between MWCNT and TiO2 nanoparticles.

Figure 3: (a) 0.00 g, (b) 0.01 g, (c) 0.02 g, and (d) 0.03 g show the morphological images of MWCNT/TiO2 thin films, where else for (e) and (f) images show the film cross-section and EDX graph. From FESEM images, we found that the MWCNT and TiO2 nanoparticles there are in good contact. MWCNT nanoparticles are well dispersed and highly compact after annealing at 300°C for 30 min. The porosity can be observed from the morphological structure of all thin films. The thin films are connected randomly which in lack dissemble long range order for the pore arrangement. We also found that, with every additional MWCNT, the porosity of the thin film became bigger and larger, while the TiO2 nanoparticles become smaller and thinner [13, 14]. The thin films also show, with increasing the amount of MWCNT, the structural morphology of TiO2 nanoparticles changes from spherical to oval nanoparticles structure, which existed around the long range MWCNT nanoparticles. The average thickness of TiO2 and MWCNT/TiO2 thin film samples around 14.88 μm and 18.79 μm, respectively. The addition of MWCNT can improve the structural and morphological design of the films. However, the amount of MWCNT added must be controlled and optimized to produce high quality electrode thin film. The amount of MWCNT can affect the films in terms of large crack and inhomogeneous arrangement. Referring to others [15], with the increase in the amount of MWCNT, the number of cracks on the surface of the films is increased subsequently. It is thought that the cracks generated on the surface could be reducing the number of adsorption sites on TiO2 film as well as causing the discrimination in the conversion efficiency of DSSC.

Table 1 shows the data collected for energy dispersive X-ray (EDX) spectroscopy analysis. The EDX data prove that every additional MWCNT added into the samples, the carbon weight percentage in the samples increased from 0.00%, 4.15%, and 2.25% to 6.55% of samples (a) 0.00 g, (b) 0.01 g, (c) 0.02 g, and (d) 0.03 g, respectively. These results confirm that the sample (d) 0.03 g MWCNT added has a larger MWCNT weight percentage compared to the other samples. Figure 4 shows the MWCNT weight percentage in graphical diagram. Both oxygen and titanium compound weight percentage data were decreased in small amounts from 40.09% to 38.84% and from 59.91% to 54.61%, correspondingly.

3.2. High Resolution Transmission Electron Microscopy (HR-TEM)

In Figure 5, the HR-TEM images of TiO2 nanoparticles (a) 0.00 g, (b) 0.01 g (c) 0.02 g, and (d) 0.03 g MWCNT/TiO2 nanocomposite are shown. Images in Figures 5(a), 5(b), 5(c), and 5(d) are carried out to show the comparison between undoped TiO2 nanoparticles and TiO2 doped MWCNT nanocomposite in terms of their inner structure design. HR-TEM investigation on selected MWCNT partially covered by TiO2 aggregates indicates that simple or multiple connections of aggregates of TiO2 nanoparticles to MWCNT are possible. In Figure 3(a), TiO2 nanoparticles with low grain density and compact agglomerate configuration were observed. The TiO2 nanoparticles sizes vary compared to the TiO2 attached to MWCNT particles. From images (b, c, and d), by using chemical treatment process, we can see the MWCNT particles successfully formed in nanotube structure. From these images, we can find out that the type of CNT used in this research is MWCNT not SWCNT [16, 17]. This can be confirmed by the multiple thin walls formed along the CNT. The TiO2 nanoparticles and TiO2 sphere-like shape are formed and positioned mostly around the MWCNT particles. Another researcher [18] also implies that there is a good contact between MWCNTs and TiO2 particles. MWCNTs are directly coupled with the uniform anatase shell, which are together embedded in the TiO2 aggregates. Such a structural feature is beneficial for efficient electron transfer and hole-electron separation, as if there is a “conducting wire” acting as a readily accessible electron-transfer channel.

Table 2 illustrates the data size parameter for TiO2 doped MWCNT nanocomposite and TiO2 nanoparticles alone. The MWCNT inner and outer structure diameters were approximately 4.49 nm and 17.71 nm in length, respectively. The diameter size for TiO2 doped MWCNT was slightly smaller with only 3.91 nm compared to the TiO2 nanoparticles with 12.00 nm. This might be due to the combination with MWCNT nanoparticles that reduce the TiO2 size and hence produce highly porous thin film as presented and shown in the FESEM analysis. HR-TEM images suggest a good affinity between the TiO2 and the MWCNT, which is important in view of limiting the MWCNT loading required to improve DSSC performance. MWCNT improves the roughness factor of the electrode and limits the charge recombination of electron/hole (e/h+) pairs [19]. Another advantage of combining MWCNT with TiO2 was to get higher photo response due to decrease in resistivity of the thin film and resulting in a higher current collection at the electrode thin film [20, 21].

Nevertheless, a high loading of MWCNT causes light-harvesting competition that affects the light absorption of the dye sensitizer and consequently reduces the cell efficiency. Moreover, an excess of MWCNT can result in a less compact TiO2 layer, in which large pores are formed at the micron scale [22] or in the formation of quite disconnected aggregates of MWCNT covered in conformance by TiO2 nanoparticles.

3.3. Atomic Force Microscopy (AFM)

The MWCNT/TiO2 thin films upper texture prepared using doctor-blade technique are observed by AFM instrument. Images in Figure 6: (a) 0.00 g, (b) 0.01 g, (c) 0.02 g, and (d) 0.03 g demonstrate that, with increasing MWCNT concentration in the MWCNT/TiO2 sample, the texture of the thin films became rougher and the particle arrangement became compact. This condition can be appointed from the root mean square (rms) value. The AFM analysis results unveil that the nonadded MWCNT 0.00 g has roughness average value of only 10.995 nm. The roughness average slightly increases with additional MWCNT added with 18.308 nm for 0.01 g, 24.322 nm for 0.02 g, and 25.723 nm for 0.03 g. These image data also reveal that rms value drastically increases for each sample. The rms values of the MWCNT/TiO2 thin film are listed in Table 3 and the data are illustrated with graphical diagram in Figure 7.

AFM measurement investigates the surface morphological roughness of the film. By AFM roughness analysis, roughness factors obtained are increased as increasing the content of MWCNT in TiO2 film with the maximum value in 0.03 g carbon-content TiO2 as listed in Table 1. This characterization is an important characteristic to investigate the surface reflection phenomenon in DSSC [23]. Due to the increase in roughness average, the enlargement for surface texture angle in the thin film will bounce the light on surface films and causing the light to reflect indirectly back to the electrode surface. This phenomenon can increase the light absorption in the metal oxide photovoltaic and improve the light of electrical conversion energy because the light reflectance had been reduced [24].

3.4. I-V Curve Efficiency

Figure 8 shows the curve graph of MWCNT/TiO2 thin film photoelectrode dye-sensitized solar cell. This process is performed under 100 mW/cm2 illuminations using xenon lamp. The measurement of light to electrical process was executed after completing the sandwich-look DSSC between electrode and counter electrode part. The parameters included in Table 4 were open-circuit voltage (), short-circuit photocurrent density (), fill factor (), and energy conversion efficiency () [25]. As stated in Table 4, open-circuit voltage () data increased from 0.49 V to 0.67 V but decreased to 0.64 V for samples (a) 0.00 g, (b) 0.01 g, and (c) 0.02 g MWCNT. At the end, the () data slightly increased back to 0.65 V for sample (d) 0.03 g MWCNT. Short-circuit photocurrent density () data for sample (a) 0.00 g is 7.60 mAcm−2. The data for samples (b) 0.01 g and (c) 0.02 g decreased from 7.25 mAcm−2 to 8.11 mAcm−2. However, data for sample (d) 0.03 g increased rather large about 19% from sample (c) 0.02 g from 8.11 mAcm−2 to 9.42 mAcm−2 successively. The fill factor percentage data for (a) 0.00 g, (b) 0.01 g, (c) 0.02 g, and (d) 0.03 g MWCNT are 38%, 43%, 47%, and 45% in that order. An increase in gas values implied enhanced electron transfer in DSSC [26]. The and data gradually increased for sample (d) 0.03 g, which might be attributed from enlargement of thin film porosity and alignment of MWCNT in the photoelectrode thin film. As referred to other researches [27], MWCNT particles might cause a significant change in efficiency via the fluctuation of the short-circuit photocurrent and the open-circuit voltage for the DSSC. Incorporating MWCNT into TiO2 nanoparticles electrode might also affect the quantity of dye adsorption and the e/h+ recombination process in this dye-sensitized solar cell.

The efficiency data for samples (a) 0.00 g, (b) 0.01 g, (c) 0.02 g, and (d) 0.03 g MWCNT are 1.43%, 2.12%, 2.46%, and 2.80%, respectively. Sample (d) 0.03 g CNT gives the highest efficiency percentage 2.80% with 0.65 V (), 9.42 mAcm−2 (), and 45% fill factor. Table 4 demonstrates the solar cell performance trend. It showed that, with every additional MWCNT powder, the efficiency of the solar cell exponentially increased. The MWCNT/TiO2 nanocomposite combination tends to improve the electrical conductivity of the photoelectrode and also helps to enhance the interconnectivity between the TiO2 and MWCNT nanoparticles, thus increasing the short-circuit current density and providing an alternative route for efficient electron transfer between the TiO2 nanoparticles. It shows that the MWCNT can operate as an electrochemical catalyst to improve the energy conversion efficiency of DSSC [28]. The overall light to electrical effectiveness of the dye-sensitized solar cell can be efficiently enhanced with every additional MWCNT added. By referring to another researcher [29], the advantages of adding MWCNT in the photoelectrode are that the MWCNT can absorb over almost the entire visible light spectrum and act as photo sensitizers, providing the MWCNT/TiO2 nanocomposites with an electron transfer mechanism similar to that of DSSC based TiO2. As a result, the MWCNT/TiO2 samples can transfer excited electrons from the MWCNT to the conduction band of TiO2 when illuminated with visible light, thereby increasing the photocurrent. Second, the conductivity of MWCNTs is superior to that of TiO2; therefore, we can expect a high transport rate of electrons in the CNT/TiO2 composites.

However, the concentration of MWCNT should be maintained ~0.03 g optimum MWCNT to avoid MWCNT agglomeration within the films. This is because higher concentration of MWCNT can cause light-harvesting competition between the dye and MWCNT particles. Thus, increases the charge transport resistance and consequently reduces the solar cell efficiency [30]. Besides, excess MWCNT may cause aggregation of TiO2 grains (as observed in the FESEM result), leading to a decrease amount of dye being adsorbed on the working electrode. In addition, as reported by another researcher [31], excessive amount of MWCNT may cause the working electrode to be less transparent, which leads to reduced efficiency of DSSCs. Furthermore, this condition will decrease the crystallinity of the TiO2 samples, thereby inhibiting the transport of electrons and increasing the probability of electron trapping by the crystal defects. Moreover, the shielding and scattering effects of excess MWCNT might have prevented the photo absorption of other visible light-active species. Nevertheless, the optimum amount of MWCNT needs to be considered for fabrication of MWCNT/TiO2 working electrodes being used in DSSC applications.

4. Conclusion

The MWCNT/TiO2 nanocomposite DSSC was successfully fabricated using sol-gel method and doctor-blade technique. The films were uniform and highly adherent. Photochemical and structural properties of the thin film were improved by incorporate MWCNT powder into TiO2 nanoparticles. The 0.03 g is the best and optimum concentration of CNT added to this research. Excessive amount of CNT may cause the working electrode to be less transparent, which leads to reduced efficiency of DSSC. FESEM morphological analysis indicated that the TiO2 and MWCNT/TiO2 thin films were in compact alignment and highly porous with thickness around 14.88 μm and 18.79 μm. HR-TEM inner structural analysis confirms that the thin film is composed of TiO2 nanoparticles that existed around the multiwalled carbon nanotube particle. The average thickness for MWCNT and TiO2 nanocomposite was about 17.71 nm and 3.91 nm, respectively. AFM analysis proved that the roughness factor can significantly improve the photoelectrode performance in the solar cell. The AFM topography reveals a very compact and rough surface; the rms values of all films are in the range of 10–25 nm. From analysis, the highest efficiency () wassuccessfully obtained from sample (c) 0.03 g added MWCNT with 2.80% efficiency, 0.65 V open-circuit voltage (), 9.42 mAcm−2 short-circuit photocurrent density (), and 45% fill factor (). This result indicated that the solar cell efficiency can be enhanced by adding optimum concentration of MWCNT into TiO2 photoelectrode solar cell.

Conflict of Interests

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


The authors would like to thank the Laboratory of Photonic Institute of Microengineering and Nanoelectronic (IMEN), Universiti Kebangsaan Malaysia, for providing the facilities.