N-Doped TiO₂/CdS Nanocomposite Films: SILAR Synthesis, Characterization and Application in Quantum Dots Solar Cell

Introduction

The solar light-to-electricity conversion by photovoltaic technology offers an ideal problem solving for currently energy crisis because of the highly dependent on fossil fuel energy. Beside its unrenewable source, the use of fossil energy results in many serious environmental problems worldwide.¹ The dye-sensitized solar cells (DSSC) based on nanocrystalline TiO₂ has established an alternative concept to conventional solar cells owing to its low-cost production and high conversion efficiency.² DSSC system has successfully obtained an overall power conversion efficiency of 12%, but it is still having problems for commercial application due to the dye sensitizer and liquid electrolyte degradations causing the lower of solar cell efficiency and stability. The use of tandem semiconductors in the solar cells system became one of promising solutions to prevent the problems appeared from the dye degradation.³

Inorganic quantum dots (QDs) semiconductors can serve as sensitizer instead of dye followed by some advantages: high light absorption in the visible region, great stability, an effective band gap that can be controlled by the QDs size, the possibility for multiple exciton generation, and the utilize of hot electron whose having higher energy than TiO₂ conduction band.⁴ The small band gap semiconductors such as CdS, CdSe or CdTe have been used to sensitize TiO₂ in the QD-sensitized solar cells (QDSSC) and has successfully increased the visible light absorption of TiO₂. Among those QDs, CdS has been receiving much attention because it has a high established relationship between the optical absorption and the size of the particle.⁵ The QDSSC conversion efficiency has reach around 8% at present,⁶ being lower than the DSSC efficiency but it is highly potential to increase the efficiency of the QDSSC system by tuning several properties of quantum dots. In order to be applied in the solar cell system, it is desirable for preparing semiconductor nanoparticles which have appropriately functionalized with different size and optical properties.

The QDSSC based on wide band gap semiconductors, such as TiO₂, has demonstrated a promising commercial technology in solar cell application. However, TiO₂ is not ideal semiconductor for solar cell photoanode due to its nature intrinsic band gap (around 3-3.4 eV) which means TiO₂ has a weak absorption of visible light⁷. Doping TiO₂ is one of the most promising approaches to increase its visible light response. Asahi et al.⁸ firstly reported that N-doped TiO₂ films prepared by sputtering method showed significance visible light absorption at wavelengths less than 500 nm related to the band gap narrowing by mixing of N 2p states with O 2p states. The application of N-doped TiO­₂ in the DSSC has significantly improved the efficiency and stability of DSSC.^9,10

So far, both doping of TiO₂ and quantum dot sensitization have been explored separately for solar cells applications, while combining the two approaches has not been reported yet. In this work, we develop CdS quantum dot sensitization with nitrogen-doping of TiO₂ and its application in QDSSC system. In N-doped TiO₂/CdS nanocomposite, CdS acts as a visible sensitizer and N-doped TiO₂ being a wide band gap semiconductor that is responsible for the charge separation process. Therefore, the prepared N-doped TiO₂/CdS nanocomposites thin films can effectively capture the visible light and transfer the photogenerated electrons into the N-doped TiO₂ conduction band, and started the electrical flows. In this work, the N-doped TiO₂/CdS nanocomposite films have been fabricated by the successive ionic layer adsorption and reaction (SILAR) method as a relatively simple technique for large scale uniform coating to produce clean, dense and strong adhesion to substrate thin films.

Materials and Methods

Materials

Cd(NO₃)₂.10H₂O, (NH₄)₂S, ethanol absolute, acetyl acid glacial from Merck, Titanium Tetraisopropoxide from Aldrich, Dodecylamine from Fluka, ITO and electrolyte redox from Dyesol have been used as received without any further purification.

Methods

The N-TiO₂ has been synthesized following the method reported previously.¹¹ The prepared N-TiO₂ nanocrystalline was then applied on ITO glass substrate through doctor blading method and was calcined at 450^oC to obtain N-TiO₂ thin film. The CdS quantum dots were synthesized directly on the N-TiO₂ thin film surface by SILAR method as follows; the N-TiO₂ thin film was immersed in 0.2 M Cd(NO₃)₂ solution for 1 minute then washed with aquadest before further immersing in 0.2 M (NH₄)₂S solution for 1 minute and again washed with aquadest to remove the impurities materials. The process was called as one cycle and the N-TiO₂ was sensitized by CdS quantum dots with the cycle variation. The prepared CdS-sensitized N-TiO₂ was then applied in QDSSC system and the performance of the QDSSC was measured as inscident photon to current efficiency (IPCE) and light to electricity conversion efficiency.

To fabricate the QDSSCs, the N-doped TiO₂/CdS thin film TiO₂ was assembled with the Pt-counter electrode (CE) between the active areas. The electrolyte solution was introduced through drilled hole on CE by capillary action, and the hole was then sealed. The current-voltage (I-V) measurements were performed using Keithley-2000 instrument with 1000 W/m² power. Thus, during the measurement the Solar Cell was irradiated by the light with the power density of 1000 W/m², which was equivalent to Air Mass 1 (AM1).

The overall photo-conversion efficiency, h is calculated from the integral of the short-circuit photocurrent density (I_sc), the open-circuit photovoltage (V_oc), the fill factor of the cell (FF), and the intensity of incident light (I_s) using the formula of

where V_oc= open-circuit Voltage (V), I_sc = short- circuit-current density (mA/cm²), and P_inc = light intensity (W/cm²). Fill factor (FF) is given by

where V_pp is a maximum voltage (V), and I_pp is a maximum current density (mA/cm²).^12,13

To study the structure of resulted materials, the corresponding difractograms were recorded using a Rigaku Miniflex 600 40 kW 15 mA Benchtop Diffractometer, CuK_α, λ=1.5406 Å at scan rate of 2^o/min.

To study the electronic structure of CdS/N-TiO₂ nanocomposites, the light absorbances were recorded using UV/Vis Spectroscopy with diffuse reflectance of Shimadzu, UV-2550 model.

Results and Discussion

The N-doped TiO₂/CdS nanocomposite films have been successfully synthesized on the N-TiO₂ thin film surface through SILAR method with 1, 5, 10, 25, and 50 cycle’s variation resulted in CdS/N-TiO₂ nanocomposites. Figure 1 shows the selected powder XRD patterns of CdS/N-TiO₂ nanocomposites with cycle’s variation of 0, 5, and 50. The patterns indicated the existence of CdS quantum dots material on N-TiO₂ semiconductor. The powder XRD pattern of CdS for the higher cycles performed in the synthesis showed the typical (111), (220), and (311) peaks of the cubic zinc blend structure which match the data of JCPDS-10-0454. The detailed CdS peaks appeared on 2q of 26.8^o, 44.12^o, 52.14^o, and 72^o  corresponding to planes of (111), (220), (311), and (400), respectively, indicating the similar cubic structure of CdS as reported by Dhage et al¹⁴ and that of ZnS as reported by Soltani et al.¹⁵ The patterns also showed the peak of N-TiO₂ nanocrystalline that appeared at around 2q of 25.2^o, 38.5^o, 47.8^o, 54.1^o, 55.4^o, 63^o, 69-71^o, and 75^o, corresponding to the anatase crystal planes of (101), (112), (200), (105), (211), (204), (116), (220), and  (215), respectively.¹⁶ The diffractograms showed that there is not much peak change on the 5 cycles comparing to the pure N-TiO₂ (0 cycle), while on 50 cycles, the higher peak intensity might be associated with the higher amount of CdS on N-TiO₂ surface.

Figure 1: The Selected X-Ray Diffractograms of N-TiO2/CdS with 0 (a), 5 (b), and 50 (c) cycles, and CdS (d) Click here to View figure

 

The crystal size of both CdS and N-TiO₂ as estimated from Scherrer equation, are listed in Table 1. The result showed that all CdS synthesized on N-TiO₂ are quantum dots size. It was also found that the cycles applied in the synthesis process highly affect the crystal structure of CdS/N-TiO₂ nanocomposites. The more cycles lead to the lower CdS crystal size, while crystallite size of N-TiO₂ was not significantly changed. It might be because the more cycles formed higher CdS quantum dots on the N-TiO₂ surface, so it leads to the rearrangement of CdS quantum dots providing the lower particle size. In contrast to the smaller particle size with the more cycle synthesis process applied, it was also found that the more synthesis cycles resulted in smaller lattice parameter of the cubic crystal structure as listed in Table 1. The XRD patterns confirmed that both CdS and N-TiO₂ retained their identity as shown in both crystal structures and peaks appeared in all XRD patterns except on pure N-TiO₂. It means the resulted N-doped TiO₂/CdS system is a nanocomposite material.¹⁷

Table 1: Crystal parameters of N-TiO₂/CdS nanocomposites based on XRD analysis

Materials	N-TiO2 size (nm)	CdS size (nm)	N-TiO2 lattice parameters		CdS lattice parameters
			a	c	A
N-TiO2	16.8	–	3.7924	9.5183	–
N-TiO2/CdS (5)	16.6	10.2	3.8142	9.5097	6.5342
N-TiO2/CdS (50)	16.9	6.4	3.8068	9.5088	5.8324
CdS	–	8.0	–	–	6.6922

 

The absorption spectra of cycle’s variation on CdS/N-TiO₂ synthesis are shown in Figure 2. The analysis of the typical bands together with colors may be seen in Table 2. It was clearly shown that the CdS addition on the N-TiO₂ resulted in a visible active response, which the higher cycles on CdS synthesis provided higher increase on absorption and wider red-shift absorption bands (smaller band gap energy/E_g). It happened because the higher cycles resulted in the more CdS amount leading to the higher visible light response. The higher CdS amount could be also identified phisically from the deeper dark yellow intense color from the higher cycles applied (Table 2). The formation of CdS on N-TiO₂ surface greatly enhances the visible light response, so the resulted CdS/N-TiO₂ materials are highly potent to use as photoanode in solar cells system.

Table 2: The Color and Assignment of UV-Vis spectral bands of N-TiO2/CdS Click here to View table

 

Figure 2: The selected UV/Visible Spectra of N-TiO2/CdS from every cycle (1, 5, 25, and 50) Click here to View figure

 

The surface morphology of CdS/N-TiO₂ nanocomposites were evaluated using Scanning Electron Microscopy. The micrographs of N-TiO₂ and CdS/N-TiO₂ with 50 cycles are shown in Figure 3. The N-TiO₂ micrographs (black region) seemed to show the porous nature of the N-TiO₂ as a requirement for the electrode better adhesion in solar cell system, which was in fact to be true in preparing the electrode. The cross sectional micrographs indicated that the application of 50 cycles SILAR synthesis on N-doped TiO₂ thin film led to the adsorption of CdS on the N-doped TiO₂ as rough surface changed to smoother packed surface and the width of the film changed from around 55 µm to around 66 µm. The existence of cubic CdS can also be identified on the N-TiO₂ surface.

Figure 3: SEM Micrographs of N-TiO2 (a) and N-TiO2/CdS 50 cycles (b) Click here to View figure

 

To study the effect of SILAR cycles on the photoelectrochemical properties of the resulted materials, we applied N-TiO₂/CdS nanocomposites in the solar cells system. The solar cells performa was analyzed as I-V measurements (Figure 4) and the overall efficiency (Table 3). From the I-V characteristics, it can be shown that these nanocomposites exhibit much higher efficiency compared to N-TiO₂ only, being 4.5-8.3% with fill factors of 0.34-0.54. Within the cycles, the 25-SILAR cycle provided the best efficiency while the 1 and 50 cycles gave a lower efficiency even from the 5-SILAR cycles. It might be caused the higher CdS amount resulted in a low band gap energy which was easily to provide CdS as a center recombination which was then decreasing the overall solar cells performa. In general, these nanocomposites also exhibit much higher efficiency compared to the fluorine-doped-tin oxide (FTO) of FTO/TiO₂/CdS bilayers systems which was only 0.78% with fill factor of 0.40,¹⁸ and to the CdS/TiO₂-nanorod and CdS/TiO₂-nanorod/g-C₃N₄, which were to be 1.67-2.31% with fill factors of 0.43-0.51.¹⁹  However, this is slightly less than that recently reported by Wang et al²⁰ for different material, being 9.02% for Zn-Cu-In-Se QDSCs.

Figure 4: The I-V Curves of QDSSC based on N-TiO2/CdS nanocomposites Click here to View figure

 

Table 3: The performa of QDSSC based on N-TiO₂/CdS nanocomposites

QDSSC	Isc (mA)	Voc (V)	Imax(mA)	Vmax(V)	FF	Ƞ (%)
NTiO2	8.4	0.48	0.3	3.9	0.29	1.7
N-TiO2/CdS (1)	14.8	0.62	0.35	8.9	0.34	4.5
N-TiO2/CdS(5)	15.7	0.74	0.4	11.8	0.41	6.7
N-TiO2/CdS(10)	15.4	0.7	0.5	10.8	0.50	7.7
N-TiO2/ CdS(25)	15.6	0.7	0.5	11.8	0.54	8.3
N-TiO2/CdS (50)	15.1	0.7	0.4	9.6	0.36	5.5

 

Conclusion

Nanocomposites of N-TiO₂/CdS quantum dots have been successfully synthesized by SILAR method with the particle size of CdS is around 8 nm. The SILAR cycle was significantly influenced the visible light response, which the higher cycle applied in the synthesis provided higher red-shift absorption edge. The quantum dot solar cells based on N-doped TiO₂/CdS nanocomposite prepared with 25 cycles provided the highest solar cell performa with overall efficiency of 8.3% with fill factor of 0.54.

Acknowledgment

The authors should thank to DIKTI (2015-2017) due to the special research funding support.

References

1. Green, M. A. Solar Energy. 2003, 74(3), 181-192. https://doi.org/10.1016/S0038-092X(03)00187-7
   CrossRef
2. O’Regan, B.; Gratzel, M. Nature. 1991, 353, 737-740. doi:10.1038/353737a0
   CrossRef
3. Chen, X.; Lin, J.; Chen, J. Nanoscale Res Lett. 2011, 6(1), 475-480. doi:  10.1186/1556-276X-6-475
   CrossRef
4. Nozik, A. J. Physica E. 2002, 14, 115-120.
   CrossRef
5. Martinez-Castanon, G. A.; Loyola-Rodiguez, J. P.; Reyes-Macias, J. F.; Ruiz, N. N.-M. Superficies y Vacio. 2010, 23(4), 1-4. https://doi.org/10.1016/S1386-9477(02)00374-0
   CrossRef
6. Jun, H. K.; Careem, M. A.; Arof, A. K. (2013). Renewable and Sustainable Energy Reviews. 2013, 22, 148-167. https://doi.org/10.1016/j.rser.2013.01.030
   CrossRef
7. Qiu, X.; Burda, C. Chemical Physic. 2007, 339, 1-10. https://doi.org/10.1016/j.chemphys.2007.06.039
   CrossRef
8. Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Science. 2001, 293, 269-273. DOI: 10.1126/science.1061051
   CrossRef
9. Ma, T.; Akiyama, M.; Abe, E.; Imai, I. Nano Letter. 2005, 5(12), 2543-2547. DOI: 10.1021/nl051885l
   CrossRef
10. Kusumawardani, C.; Kartini, I.; Narsito. Thammasat Int. J.Sc.Tech. 2010, 14(5), 1-10.
11. Kusumawardani C.; Suwardi; Indriana, K.; Narsito. Asian Journal of Chemistry. 2012, 24(1), 255-260.
12. ANONIM, Chapter 9: Solar Cell Parameters and Equivalent Circuit. 113-121
13. Winkler, M. T.; Cox, C. R.; Nocera, D. G.; Bounasisi, T. Modeling integrated photovoltaic–electrochemical devices using steady-state equivalent circuits. 2012, E1076-E1082. www.pnas.org/cgi/doi/10.1073/pnas.1301532110
14. Dhage, S. R.; Colorado, H. A.; Hahn, T. Nanoscale Research Letters. 2011, 1-5. https://doi.org/10.1186/1556-276X-6-420
    CrossRef
15. Nayereh Soltani, N.; Saion, E.; Hussein, M. Z.; Erfani, M.; Abedini, A.; Bahmanrokh, G.; Navasery , M.; Vaziri, P.  Int. J. Mol. Sci. 2012, 13, 12242-12258; doi:10.3390/ijms131012242
    CrossRef
16. Jiang, H. B.; Cuan, Q.; Wen, C. Z.; Xing, J.; Wu, D.; Gong, X-Q.; Li, C.; Yang, H. G. Angewandte Chemie. 2011, 50(16), 3764-3768. https://doi.org/10.1002/anie.201007771
    CrossRef
17. Chi, Y.; Fua, H.; Qi, L.; Shi, K.; Zhang, H.; Yu, H. Journal of Photochemistry and Photobiology A: Chemistry. 2008, 195, 357-363
    CrossRef
18. Deshmukh, P. R.; Patil, U. M.; Gurav, K. V.; Kulkarni, S. B.; Lokhande, C. D. Bulletin of Material Science. 2012, 35(7), 1181-1186.
    CrossRef
19. Gao, Q.;  Sun, S.; Li, X.; Zhang, X.; Duan, L.; Lü, W. Nanoscale Research Letters. 2012, 1-9. DOI: 10.1186/s11671-016-1677-1
    CrossRef
20. Wang, W.; Jiang, G.; Yu, J.; Wang, W.; Pan, Z.; Nakazawa, N.; Shen, Q.; Zhong, X. ACS Appl Mater Interfaces. 2017, 9(27), 22549-22559. doi: 10.1021/acsami.7b05598. Epub 2017 Jun 28.
    CrossRef

---

This work is licensed under a Creative Commons Attribution 4.0 International License.