The layer boundary effect on multi-layer mesoporous TiO₂ film based dye sensitized solar cells

Abstract

Multi-layer mesoporous TiO₂ prepared by screen printing is widely used for fabrication of high-efficiency dye-sensitized solar cells (DSSCs). We compare the three types of ∼10 μm thick mesoporous TiO₂ films, which were screen printed as 1-, 2- and 4-layers using the same TiO₂ nanocrystal paste. The layer boundary of the multi-layer mesoporous TiO₂ films was observed in the cross-section SEM. The existence of a layer boundary could reduce the photoelectron diffusion length with the increase of layer number. However, the photoelectron diffusion lengths of the Z907 dye sensitized solar cells based on these different layered mesoporous TiO₂ films are all longer than the film thickness. Consequently, the photovoltaic performance seems to have little dependence on the layer number of the multi-layer TiO₂ based DSSCs.

---

1. Introduction

Dye sensitized solar cells (DSSC) have been one of the most well studied low-cost nanostructured photovoltaic devices since Grätzel and his coworkers reported their milestone in DSSC using a mesoporous TiO₂ nanoparticle layer in 1991.¹ A typical DSSC comprises a wide band gap semiconductor substrate, normally a nanostructured TiO₂ film as both a sensitizer matrix and electron transport pathway, a stable sensitizer dye with strong visible light absorption, redox electrolytes and a counter electrode.^2–7 A high-efficiency DSSC should absorb visible light as much as possible to achieve high photocurrent. Due to Ru(II)-bpy dye's absorption spectrum and extinguish coefficient limitation, the mesoporous TiO₂ nanocrystal film should be thick enough to absorb the dye.^5,8 Since the photoelectron diffusion length is around tens of micrometre, most dye sensitized solar cells with high efficiency currently use ∼10 μm thick mesoporous TiO₂ film to achieve the balance between high light absorption and high charge collection efficiency.

In most reports, the mesoporous TiO₂ films were prepared by doctor-blade and/or screen printing methods and the film thickness is controlled by tap/screen thickness combined with TiO₂ paste composition. Screen printing technique has been used in fabrication of TiO₂ films for DSSC due to its high reproducibility for large scale production.² Since it is more convenient to obtain TiO₂ films with a desired thickness by printing or doctor-blade for several times than by a one-time printing, lots of reported DSSCs with high efficiency were fabricated on these several layered mesoporous TiO₂ films. Although several layered mesoporous TiO₂ films work well in high-efficiency DSSC, one curious concern is still unanswered. Does the layer boundary matters in these multi-layer mesoporous TiO₂ based DSSCs? Whether and how one-layer TiO₂ films are different from multi-layer TiO₂ films with the same thickness? There is no such fundamental investigation to explore it yet. Here we investigate three types of 10 μm thick TiO₂ films by 1, 2 and 4 layer screen printing using the same TiO₂ nanocrystal paste with different screen. The effect of layer boundary on the photovoltaic performance and electron transferring properties is investigated.

2. Results and discussion

First, the thickness of these 1-, 2- and 4-layer TiO₂ films measured by profiler are 10.1 ± 0.2 μm, 9.8 ± 0.2 μm and 9.7 ± 0.3 μm respectively, which are almost the same. The cross-section SEM images of the 4-layer TiO₂ film are shown in Fig. 1. The results clearly reveal the obvious 4-layer morphology and corresponding layer boundary caused by multi-time screen printing process. This is understandable because there is heat curing step between the each screen printing. The layer boundary is schematically illustrated in the Fig. 1, which suggested that the multi-time screen printing process would introduce the more boundary layer. But the existence of layer boundary seems to have little influence on the photovoltaic performance. Table 1 compares the J–V characteristics measured under AM 1.5 illumination of the fabricated Z907 sensitized solar cells using 10 μm thick TiO₂ films screen printed respectively by 1, 2, 4 times using the same TiO₂ nanocrystal paste. All these device exhibited high reproducibility and low variation, which is the advantages of screen printing and is consisted with previous reports.⁹

Fig. 1 Schematic illustration of (A) 2-layer (B) 4-layer structured mesoporous TiO₂ films with layer boundary; cross-section SEM images of the (C) layer-1/2 and (D) layer-3/4 in the 4-layer mesoporous TiO₂ film.

Table 1 Detailed photovoltaic parameters of DSSCs based on different layered TiO₂ films; six devices from each type of TiO₂ film_s were used for comparison

Layer	Jsc (mA cm^−2)	Voc (mV)	FF	η (%)
1	14.52 ± 0.21	0.695 ± 0.010	0.76 ± 0.02	7.66 ± 0.19
2	14.44 ± 0.26	0.685 ± 0.011	0.75 ± 0.03	7.47 ± 0.23
4	14.24 ± 0.30	0.680 ± 0.012	0.75 ± 0.03	7.30 ± 0.29

---

With the increase of the layer number from 1 to 4, the average J_sc of the devices based on ∼10 μm thick TiO₂ decreases from 14.52 to 14.44 and 14.24 mA cm⁻², the average V_oc decreases from 0.695 to 0.685 and 0.680 V and average fill factor decreases from 0.759 to 0.755 and 0.754. Consequently, the overall photovoltaic efficiency of DSSCs prepared with the ~10 μm thick TiO₂ films decreases from 7.66% to 7.47% and 7.30% with the increase of layer number. The DSSC based on one-layer TiO₂ film seems to exhibit the best performance, but the difference of J–V characteristics among these DSSCs based on different layered TiO₂ films is less than 5%.

The above mentioned result seems to suggest that whether the TiO₂ films were prepared by one layer or multi-layers has insignificant or a little effect on their photovoltaic performance. It might account for the fact that previous reports adopted either multi-layer or one layer mesoporous TiO₂ for their high efficiency DSSCs using different dyes or electrolytes. However, we find that the electron transport properties depend on the layer structure of the TiO₂ electrode as discussed next.

Transport and recombination properties in DSSCs including recombination lifetime (τ_r), diffusion length (L_n) and diffusion coefficient (D_n) were investigated by intensity-modulated photocurrent spectroscopy (IMPS) and intensity-modulated photovoltage spectroscopy (IMVS).^10–12 The electron recombination lifetime τ_r showed in Fig. 2 was obtained by using IMVS under open-circuit conditions as function of light intensity, which can be directly calculated from IMVS response using equation:

τ_r = 1/2πf_min

where f_min is the lowest frequency of the IMVS. The electron lifetime plays a major role in DSSCs and the short lifetime means that recombination occurs before electrons reaching the collecting electrode. Fig. 2A shows that the electron recombination lifetime decreases as light intensity increases. It is slightly larger in 4-layer TiO₂ films than in 1- and 2-layer TiO₂ films but there is no significant difference. Previous studies showed that electron loses in mesoporous TiO₂ based DSSC occur predominantly at the TiO₂/electrolyte interface.^4,13–15 The redox electrolyte I₃⁻/I⁻ used in our study is known for slow I₃⁻ reduction kinetics accounting for tardy interfacial recombination.^16–18 Both the property of the electrolyte and the recombination mechanism confirm that layer boundary has little influence on electron recombination lifetime in these multi-layer TiO₂ films using the same TiO₂ nanocrystal paste.

Fig. 2 (A) Recombination lifetime viruses V_oc as function of light intensity and (B) diffusion length under different bias voltages in the different-layer mesoporous TiO₂ configuration DSSCs.

The photoelectron diffusion length (L_n) was calculated by using equation of L_n = (D_nτ_r)^1/2 defined from the probability of collection for a small macroscopic perturbation as shown in Fig. 2B.^19,20 The values of D_n were calculated from the transport times by following the procedure detailed previously.¹⁴ First of all, the diffusion length increases with potential when bias potential is below open-circuit voltage, which results from the free carrier lifetime. Interestingly, the L_n obviously decreases with the layer numbers in these multi-layer mesoporous TiO₂ films. The photoelectron's diffusion length in the 1-layer TiO₂ film based device is almost 50% larger than that of the 4-layer one. However, even the 4-layer one's diffusion length is more than three times larger than the film thickness of 10 μm when the bias voltage is below open-circuit voltage. It has been reported that L_n should be several times (generally more than three times) greater than the film thickness to obtain the optimum performance.^15,21 Thus, even the 4-layer mesoporous TiO₂ film based DSSC has the sufficient electron collection from the films. This could account for the fact that there is little variation on the photovoltaic performance of the 1-, 2- and 4-layer mesoporous TiO₂ based DSSCs as shown in Table 1.

To further explore the existence of layer boundary, we measured the amount of charge presented in the TiO₂ films by photocurrent transients under short-circuit conditions. Fig. 3 shows the extracted charge increases with bias voltage dependence of light intensity. The charge amounts of these different layer TiO₂ increase with the layer numbers, which reveal that more electrons are trapped in the multi-layer mesoporous TiO₂ than the 1-layer one. This indicates the layer boundary exist as seen in the SEM images of Fig. 1 and would trap the electron to affect their diffusion length in Fig. 2B. These photoelectron transfer results suggest that the layer boundary in the multi-layer mesoporous TiO₂ has little effect on the electron recombination lifetime but could affect the diffusion length due to the layer boundary trapping effect. The mechanism behind it is that the TiO₂/electrolyte interface accounting for the recombination lifetime in these different multi-layer TiO₂ mesoporous films is almost the same, while the electron diffusion coefficients could be affected due to the electron trap by the layer boundary.

Fig. 3 Extracted charge under short-circuit conditions as function of light intensity.

3. Conclusions

In summary, we investigate the layer boundary effect of the multi-layer mesoporous TiO₂ thin film for DSSC via cross-section morphology, photovoltaic and electron transfer properties of 1-, 2- and 4-layer TiO₂ films based DSSCs. The layer boundary in the multi-layer TiO₂ films can be clearly observed and such layer boundary should be related to the TiO₂ paste curing process between the screen printing steps. The layer boundary could induce more trap sites to deteriorate the photoelectron diffusion length but shows little impact on their recombination lifetime because the recombination lifetime is dominated by the TiO₂/electrolyte interface. However, the superior photoelectron transfer properties of Z907 lead to the insignificant photovoltaic performance variation in 1-, 2- and 4-layer mesoporous TiO₂ films based DSSCs because all the diffusion length in these devices are almost more than 3 times as the mesoporous TiO₂ film's thickness. Although the layer boundary seems to have little effect on the Z907 sensitized multi-layer mesoporous TiO₂ DSSCs, the layer boundary in multi-layer mesoporous TiO₂ structure might would exert a dramatic influence when some novel sensitizers with short diffusion length are used for sensitized solar cells. Thus, the 1-layer mesoporous TiO₂ films would be a better candidate than the multi-layer ones for DSSCs. In all, our study shows that more attention should be invested into the layer boundary effect when adapting multi-layer structure mesoporous TiO₂ for sensitized solar cells.

4. Experimental details

Preparation of 20 nm particle sized TiO₂ paste

20 nm sized TiO₂ nanoparticle pastes were synthesized via a modified procedure in accordance with the work of Seigo Ito et al.² 12 g of acetic acid was mixed with 58.6 g of titanium tetraisopropoxide (TTIP, Aldrich) for 30 minute stirring at room temperature. 290 mL of deionized water was then added into the mixture solution under one-hour vigorous stirring (700 rmp) to achieve a complete hydrolysis reaction. A sol–gel was obtained through reflux at 80 °C for 1.5 h after adding an amount of 4 mL concentrated nitric acid. The sol–gel was then hydrothermally treated at 250 °C for 12 h. Following this step, the TiO₂ nanocrystals was washed and centrifuged three times with pure ethanol to remove the water. The well dispersed TiO₂ ethanol suspension containing 16 g TiO₂ was then mixed with 100 mL of ethanol solution containing 55 g terpineol and 8 g ethyl cellulose by repeated probe sonicating and stirring overnight to form a homogenous suspension without any precipitate. A final homogenous TiO₂ paste was made after the ethanol was removed by rotary evaporator.

Device fabrication

36 units of 5.5 × 5.5 mm sized TiO₂ films with were fabricated on 10 cm × 10 cm FTO with compact TiO₂ layer by using the same prepared paste with three different screens with 2.5, 5 and 10 μm thickness on the same screen printer. The three TiO₂ films were screen printed respectively by 1, 2, 4 times on the same clean TEC 7 FTO with compact TiO₂ layer to obtain ∼10 μm thickness. The compacted TiO₂ layer was deposited by spray pyrolysis.²² The screen printed TiO₂ films were then followed by being baked at 120 °C for 10 min and subsequently annealed at 500 °C for 30 min with a heating rate of 5 °C min⁻¹. The annealed TiO₂ films was then treated by 0.04 mM TiCl₄ for 30 min at 65 °C. The square area of the device was 5.5 mm × 5.5 mm. For dye adsorption, the fabricated TiO₂ films were first annealed at 500 °C for 30 min with a heating rate of 5 °C min⁻¹. Once the temperature was cooled down to 100 °C, the TiO₂ films were immersed into anhydrous acetonitrile solution containing Z907 dye (Solaronix SA.) and kept at room temperature for 24 h in the dark. Pt-Coated FTO, used as a counter electrode, was prepared by dropping H₂PtCl₆ isopropanol solution (0.7 mM) on a clean FTO substrate followed by heating at 400 °C for 20 min in air. The electrolyte consisted of 0.6 M 1-butyl-3-methylimidazolium iodide (BMII, Merck Co.), 0.03 M I₂, 0.1 M guanidinium thiocyanate (GSCN, Aldrich Chemical Co.), 0.02 M LiI, and 0.5 M 4-tert-butylpyridine (TBP, Aldrich) in acetonitrile and valeronitrile (85 : 15 v/v). DSSC was assembled following the previous reports.^2,23 Briefly, the dye sensitized TiO₂ and counter electrode were integrated by a continuous square loop of Surlyn into a sandwich shaped cell. The electrolyte was driven into the cell through the hole made in the counter electrode by vacuum backfilling. Consequently, the hole was sealed by a glass slide and Surlyn.

Photovoltaic measurements

Photocurrent–voltage (I–V) measurements were performed using a Keithley model 2400 source measurement unit. A 300 W xenon lamp (Spectra-Physics) with a KG-5 filter for approximating global air mass 1.5 (AM 1.5, 1000 W m⁻²) was used as the light source and the light intensity was adjusted by using an NREL-calibrated Si solar cell. The dye sensitized cell was covered with a square shaped 0.25 cm² mask during the I–V measurement. The cross section morphology of the films was characterized on a FEI Nova 230 scanning electron microscope (SEM).

Acknowledgements

YZ acknowledges the support of the NSFC (Grant 51372151 and 21303103) and Huoyingdong Grant (151046).

Notes and references

1. B. O'Regan and M. Grätzel, Nature, 1991, 353, 737–740.
2. S. Ito, T. N. Murakami, P. Comte, P. Liska, C. Grätzel, M. K. Nazeeruddin and M. Grätzel, Thin Solid Films, 2008, 516, 4613–4619.
3. M. Gratzel, Nature, 2001, 414, 338–344.
4. A. J. Frank, N. Kopidakis and J. v. d. Lagemaat, Coord. Chem. Rev., 2004, 248, 1165–1179.
5. A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pettersson, Chem. Rev., 2010, 110, 6595–6663.
6. J. R. Swierk and T. E. Mallouk, Chem. Soc. Rev., 2013, 42, 2357–2387.
7. M. Wu and T. Ma, J. Phys. Chem. C, 2014, 118, 16727–16742.
8. M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Müller, P. Liska, N. Vlachopoulos and M. Grätzel, J. Am. Chem. Soc., 1993, 115, 6382–6390.
9. S.-W. Lee, K.-S. Ahn, K. Zhu, N. R. Neale and A. J. Frank, J. Phys. Chem. C, 2012, 116, 21285–21290.
10. C. Magne, T. Moehl, M. Urien, M. Grätzel and T. Pauporté, J. Mater. Chem. A, 2013, 1, 2079–2088.
11. Q. Liu, Y. Zhou, Y. Duan, M. Wang and Y. Lin, Electrochim. Acta, 2013, 95, 48–53.
12. Y.-R. Jung, J.-Y. Park and H.-B. Gu, Mater. Lett., 2015, 138, 268–271.
13. F. Fabregat-Santiago, J. Bisquert, G. Garcia-Belmonte, G. Boschloo and A. Hagfeldt, Sol. Energy Mater. Sol. Cells, 2005, 87, 117–131.
14. N. Kopidakis, K. D. Benkstein, J. van de Lagemaat and A. J. Frank, J. Phys. Chem. B, 2003, 107, 11307–11315.
15. Y. Fukai, Y. Kondo, S. Mori and E. Suzuki, Electrochem. Commun., 2007, 9, 1439–1443.
16. B. A. Gregg, F. Pichot, S. Ferrere and C. L. Fields, J. Phys. Chem. B, 2001, 105, 1422–1429.
17. D. Cahen, G. Hodes, M. Grätzel, J. F. Guillemoles and I. Riess, J. Phys. Chem. B, 2000, 104, 2053–2059.
18. S. A. Haque, Y. Tachibana, R. L. Willis, J. E. Moser, M. Grätzel, D. R. Klug and J. R. Durrant, J. Phys. Chem. B, 2000, 104, 538–547.
19. A. C. Fisher, L. M. Peter, E. A. Ponomarev, A. B. Walker and K. G. U. Wijayantha, J. Phys. Chem. B, 2000, 104, 949–958.
20. J. Bisquert and I. Mora-Seró, J. Phys. Chem. Lett., 2010, 1, 450–456.
21. P. R. F. Barnes, A. Y. Anderson, S. E. Koops, J. R. Durrant and B. C. O'Regan, Chem. Rev., 2009, 113, 1126–1136.
22. Y. Zhao, A. M. Nardes and K. Zhu, J. Phys. Chem. Lett., 2014, 5, 490–494.
23. K. Chen-Hao and W. Jih-Jen, Nanotechnology, 2007, 18, 505706.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra18830c

---