Enhancement of light-harvesting efficiency of dye-sensitized solar cells via forming TiO₂ composite double layers with down/up converting phosphor dispersion

Abstract

The light-conversion efficiency of dye-sensitized solar cells (DSSCs) was improved about 13.5% and 6.0% by dispersing a down-conversion phosphor (DCP) and an up-conversion phosphor (UCP) in a TiO₂ layer, respectively. Also, further improvement in the efficiency could be achieved by forming DCP/UCP/TiO₂ double composite layers (η = 8.88%), which is about 21% higher than that of the cell (η = 7.36%) without the phosphor dispersion.

---

Dye-sensitized solar cells (DSSCs) have attracted great attention as one of the potential technologies for renewable electrical power generation.^1–3 Solid-state solar cells based on crystalline or amorphous silicon are dominant players in the market. DSSCs suffer from the relatively low light-conversion efficiency compared with the crystalline Si or thin-film CIGS (copper indium gallium selenide). Thus, many researchers have been focused on improving the light-harvesting efficiency of DSSCs.^4–8 In general, DSSCs consist of a dye-loaded photoanode (semiconductor), a redox electrolyte, and a platinized counter electrode. Among those parts, the photoanode properties are critical to determine the efficiency of DSSCs because the performance of the cell depends strongly on the light harvesting capacities of dye molecules adsorbed on the surface of anode materials. The development of anode materials is a critical issue.

Recently, luminescent materials (phosphors) were explored as an alternative strategy for enhancing the light-conversion efficiency of DSSCS.^9–14 For example, Ce-doped Y₃Al₅O₁₂ (YAG:Ce) phosphors effectively convert the high energy photons to the visible light that is suitably absorbed by dye molecules.¹⁵ As a result, when Y₃Al₅O₁₂:Ce particles are used as a scattering layer, the DSSC efficiency was reported to be enhanced about 13.5% mainly due to improved light absorption and scattering. The most commonly used dyes (Ru-based dyes: N3, N749, and N719) have a large optical band gab of about 1.8 eV. As a result, Ru-based dyes have an absorption threshold below ∼700 nm. To achieve the improvement in the efficiency of DSSCs, it is necessary to extend the light utilization of dye molecules into the infrared (IR) region without any loss of the cell performance in the visible region. Up-conversion phosphors (UCP) can successfully convert IR photons to visible light. In 2010, Shan and Demopoulos first demonstrated the effectiveness of lanthanide-doped UCPs for the utilization of IR photons in DSSCs.¹⁶ They synthesized the Er³⁺/Yb³⁺ co-doped LaF₃–TiO₂ nanocomposite and proved the potential of the near-IR sunlight harvesting in dye-sensitized solar cell. In 2011, Li et al. introduced the Tm³⁺/Yb³⁺ co-doped Lu₂O₃ into the TiO₂ film in the DSSC and reported the cell efficiency could be improved about 11.1%.¹⁷ To our best knowledge, there was no report using both the down- and up-converting phosphors together as a photoanode.

In this communication, we report a simple way to improve the light-harvesting efficiency of DSSCs. The purpose of this work is to demonstrate the effectiveness of the photoanode composed of DCP/UCP-dispersed TiO₂ double layers in terms of improving the light harvesting efficiency of DSSCs. Various anode structures as shown in Fig. 1 were constructed and the cell efficiency was evaluated with intention of determining the most effective way in simultaneously using the DCP (Y₃Al₅O₁₂:Ce³⁺) and UCP (Gd₂O₃:Er³⁺/Yb³⁺).

Fig. 1 Schematic diagrams of electrode films with UCP and DCP.

We synthesized submicron-sized Y₃Al₅O₁₂:Ce³⁺ and Gd₂O₃:Er³⁺/Yb³⁺ particles by an ultrasonic spray pyrolysis process which is known as a promising tool for the multi-component ceramics like phosphor.^18,19 Submicron-sized titania granules were also synthesized by using spray pyrolysis of TiO₂ nanoparticles (P25). In the preparation of phosphor/TiO₂ composite anodes, the weight percentage of the phosphor was fixed at 10 wt%. Fig. 2 shows SEM images and particle size distribution of the synthesized particles. The prepared granules consisting of nano-sized TiO₂ have spherical shape and the average particle size of 403 nm. The synthesized Y₃Al₅O₁₂:Ce³⁺ and Gd₂O₃:Er³⁺/Yb³⁺ particles exist as an aggregate. The average particle sizes of Y₃Al₅O₁₂:Ce³⁺ and Gd₂O₃:Er³⁺/Yb³⁺ phosphors were 641 nm and 288 nm, respectively.

Fig. 2 SEM photos ((a) TiO₂, (b) YAG:Ce, (c) Gd₂O₃:Er,Yb) and particle size distribution (d).

Fig. 3 shows the emission spectra of Y₃Al₅O₁₂:Ce³⁺ and Gd₂O₃:Er³⁺/Yb³⁺. The band emission with a peak at 535 nm for Y₃Al₅O₁₂:Ce³⁺ under the excitation wavelength of 450 nm is assigned to a typical 5d¹ → 4f¹ transition of Ce³⁺ and in good agreement with the absorption range of N719 dye as shown in Fig. 3. Thus, Y₃Al₅O₁₂:Ce³⁺ as an efficient light down-converting phosphor can improve the light harvesting of dye molecules. The prepared Gd₂O₃:Er³⁺/Yb³⁺ nanoparticles showed strong green and red emission when excited by a 980 nm laser. The green emission bands at 534 and 564 nm are due to ²H_11/2 → ⁴I_15/2 (534 nm) and ⁴S_3/2 → ⁴I_15/2 (564 nm) of Er³⁺ ions, respectively.²⁰ The red emission band located at 650–700 nm is due to the transition of ⁴F_9/2 → ⁴I_15/2. The strong green emissions of Gd₂O₃:Er³⁺/Yb³⁺ nanoparticles fall in the absorption band of the N719 dye. Therefore, the IR photons of the incident light can be utilized in the enhancement of DSSCs by using the synthesized up-conversion Gd₂O₃:Er³⁺/Yb³⁺ phosphor.

Fig. 3 Emission spectra of YAG:Ce and Gd₂O₃:Er,Yb phosphor, and UV/visible absorption spectrum of dye (N719).

The current density–voltage (J–V) curves of DSSCs with different types of photoanodes were obtained under simulated 1.5 AM solar illumination (100 mW cm⁻²) and shown in Fig. 4(a). Detailed photovoltaic characteristics were summarized in Table 1. The dispersion of DCP or UCP into the TiO₂ anode led to the increase in the current density (J_sc), whereas the open-circuit voltage (V_oc) was reduced. No significant change in the fill factor (FF) was observed with changing the anode materials. The DCP (YAG:Ce) dispersed anode (type II) exhibited a current density of 14.5 mA cm⁻² and an cell efficiency (η) of 8.35%, indicating a 13.5% improvement in efficiency compared with the pure TiO₂ anode (type I, η = 7.36%). This improvement in the cell efficiency via the DCP dispersion is similar to the previous reported literature in which the YAG:Ce DCP was used as a back scattering layer.¹⁵ For the UCP (Gd₂O₃:Er³⁺/Yb³⁺) dispersed anode (type III), the cell efficiency was 7.80% corresponding to about 6.0% improved value compared with the cell without a phosphor dispersion. When the anode was prepared a DCP/TiO₂–UCP/TiO₂ bilayer (type IV), the current density was largely increased, and the light-harvesting efficiency (η = 8.88%) was improved about 20.7% compared with the photoanode type I.

Fig. 4 J–V curves (a), UV/visible diffuse reflectance spectra before the dye loading, absorption spectra of anode films after the dye loading (c), IPCE (d), EIS spectra (e), and Bode plots of the EIS (f).

Table 1 Photovoltaic properties and parameters of the EIS data for the four electrodes consisting of different

Anode type	I	II	III	IV
a Δη = (η − ηtype I)/ηtype I × 100.b Electron life time = 1/(2πfmax).
Materials	TiO2	DCP–TiO2	UCP–TiO2	DCP/UCP–TiO2
Thickness, μm	25	25	26	27
Voc, V	0.82	0.78	0.77	0.77
Jsc, mA cm^−2	12.0	14.5	13.9	15.9
FF	0.75	0.74	0.73	0.73
η, %	7.36	8.35	7.80	8.88
Δη,^a %	—	13.5	6.0	20.7
R2, Ω	8.18	4.56	7.38	4.77
τeff,^b ms	8.6	8.6	13.6	10.0

---

The UV/visible diffuse reflectance of the photoanodes before the dye loading was measured in order to check the light scattering effect and the results were shown in Fig. 4(b). The DCP or UCP-dispersed TiO₂ composite films (type II, III, and IV) showed no improved reflectance because the refractive indexes of YAG and Gd₂O₃ are smaller than that of TiO₂. For the YAG:Ce-dispersed anode (type II), the light absorption in the wavelength range from 400 nm to 500 nm is due to the 4f → 5d transition of Ce³⁺. Fig. 4(c) shows the absorption spectra for the photoanodes after the dye loading, which comprises the effect of porous photoanodes. In the wavelength shorter than 450 nm, the absorbance of the phosphor-dispersed photoanodes was smaller than that of the pure TiO₂ layer (type I) because the adsorption of dye molecules on the surface of phosphors (DCP or UCP) is negligible. But, in the longer wavelength region than 550 nm, the DCP-dispersed films (type II and type IV) showed a little higher absorbance compared with the film without the phosphor dispersion. This result do not reflect the increase of the dye-loading quantity, but be attributed to the enhanced penetration of long-wavelength light by the phosphor particles having the lower refractive index than titania. From the reflectance data, it was confirmed that the improvement in the light-harvesting efficiency of the photoanodes comprising the phosphor is not due to the enhancement of light scattering or the increment of dye loading.

As shown in Fig. 4(a), the dispersion of DCP or UCP in the nano-sized TiO₂ anode results in the increment in the current density, indicating that the light-harvesting capability of dye molecules is improved. Given this, it was clear that the phosphors effectively work as a light converter. That is, the YAG:Ce and Gd₂O₃:Er³⁺/Yb³⁺ phosphors convert blue light (400–500 nm) and IR photons in the incident light into the visible light more suitable for the light utilization of dye molecules, respectively. To elucidate the phosphor effect, the incident-photon-to-current conversion efficiency (IPCE) was given in Fig. 4(d). The photoanode with the DCP dispersion (type II) has the highest IPCE values in the wavelength region shorter than 450 nm. According to the excitation spectrum of YAG:Ce shown in the inset of Fig. 4(d), the phosphor has good emission under the excitation of UV (325–375 nm) and blue (400–500 nm) light. Accordingly, the enhancement in the IPCE values of the DCP-dispersed anode (type II) in the short wavelength region (<450 nm) indicates that the YAG:Ce phosphor was working good as a light converter. Interestingly, compared with the pure TiO₂ electrode (type I), the UCP-dispersed anode (type III) shows higher IPCE values in a longer (>600 nm) wavelength range. This can be attributed that the UCP dispersion makes it possible for the light of longer wavelength to more deeply penetrate into the porous anode layer.

Electrochemical impedance spectroscopy (EIS) measurements at the open-circuit voltage (V_oc) under the light illumination (100 mW cm⁻²) were conducted in order to obtain information about the interfacial characteristics of the photoanodes. The EIS spectra exhibited three typical semicircles in the Nyquist plot (Fig. 4(e)). An equivalent circuit was shown as an inset in Fig. 4(e). R_s is the resistance at the TiO₂/FTO interface. R₁ is corresponding to the first semicircle in the high frequency region, reflecting the charge transfer resistance at the electrolyte/Pt-FTO interface. The large semicircle in the intermediate frequency is related to the resistance (R₂) of the TiO₂/dye/electrolyte interface, and the Nernst diffusion resistance of electrolyte (R₃) is observed in the low-frequency region.^21,22 There was no significant changes in the R₁ and R₃ values, whereas the resistance R₂ was lowered by the dispersion of phosphors, especially by the YAG:Ce dispersion (type II and IV). The reduction in the R₂ value can accelerate the recombination rate. Nevertheless, the photocurrent was increased in the phosphor-dispersed photoanodes. This result indicates that the reduction of the resistance R₂ is due to the increase of the photo-excited electrons injected into the conduction band of TiO₂. N719 dye molecules are hard to be adsorbed on the surface of DCP particles. Thus, the increase in the dye-loading quantity was not achieved by the phosphor dispersion. Otherwise, the DCP or UCP particles can supply additionally the visible light to the dye molecules, which is responsible for the enlargement of electrons injected into the conduction band of TiO₂. Given this, the dispersion of YAG:Ce phosphor into the nanocrystalline TiO₂ layer is helpful to accelerate the electron transfer in the photoanode. The light-conversion efficiency did not have a linear relation to the R₂ values of photoanodes. This result says that a crucial factor is involved in the determination of the cell efficiency in the phosphor/TiO₂ composite photoanodes.

The electron lifetime (τ) can be estimated by using the relation τ = 1/(2πf_max), where f_max is the peak frequency of the intermediate-frequency semicircle in Fig. 4(e). To clearly show the changes in the lifetime of photo-excited electrons, the Bode plots of the EIS data obtained for the prepared four electrodes were shown in Fig. 4(f), in which the peak position (f_max) in the middle frequency region (1–100 Hz) is connected with the charge transfer TiO₂/dye/electrolyte. Then the peak position should be shifted to the low frequency as increasing the electron lifetime.²³ The resistance (R₂) and the electron lifetime calculated from the EIS data are summarized in Table 1. The lowest resistance was observed in the type II (DCP–TiO₂), whereas the type III electrode (UCP–TiO₂) had the longest electron lifetime. The DCP (YAG:Ce) dispersion was effective to reduce the interfacial resistance, but did not increase the electron lifetime. On the contrary, the UCP (Gd₂O₃:Er³⁺,Yb³⁺) dispersion was helpful for the elongation of the electron lifetime rather than the decrease of the interfacial resistance. As a result, the type IV electrode fabricated by dispersing both UCP and DCP particles showed an large decrease of the resistance as well as increasing the electron lifetime compared with the pure TiO₂ electrode. As a result, the highest cell efficiency was achieved at the photoanode consisting of two phosphor/TiO₂ composite layers (DCP–TiO₂//UCP–TiO₂).

In summary, DCP and UCP phosphor particles were dispersed in the TiO₂ layer with the intention of improving the light-harvesting efficiency of DSSCs. The improvement in the cell efficiency was due to the increase in the current density, indicating that the improvement in the light harvesting of dye molecules is achieved by the dispersion of phosphor particles in the TiO₂ layer. The phosphor dispersion did not improve the light scattering but reduced the resistance related to the electron charge transfer in the photoanode. Also, the UCP dispersion was effective to increase the electron lifetime. Therefore, the dye-sensitized solar cell with the DCP/UCP–TiO₂ composite double layers showed a remarkable improvement in the light-conversion efficiency (8.88%) compared with the cell without the phosphor dispersion (7.36%).

Notes and references

1. S. K. Balasingam, M. Lee, M. G. Kang and Y. Jun, Chem. Commun., 2013, 49, 1471.
2. S. Zhang, X. Yang, Y. Numata and L. Han, Energy Environ. Sci., 2013, 6, 1443.
3. A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pattersson, Chem. Rev., 2010, 110, 6595.
4. T. T. T. Pham, T. Bessho, N. Mathews, S. M. Zakeeruddin, Y. M. Lam, S. Mhaisalkar and M. Grätzel, J. Mater. Chem., 2012, 22, 16201.
5. C.-Y. Cho and J. H. Moon, Adv. Mater., 2011, 23, 2971.
6. N. Tétreault, É. Arsenault, L.-P. Heiniger, N. Soheilnia, J. Brillet, T. Moehl, S. Zakeeruddin, G. A. Ozin and M. Grätzel, Nano Lett., 2011, 11, 4579.
7. L. Tu, H. Pan, H. Xie, A. Yu, M. Xu, Q. Chai, Y. Cui and X. Zhou, Solid State Sci., 2012, 12, 616.
8. F. Sauvage, D. Chen, P. Comte, F. Huang, L.-P. Heiniger, Y.-B. Cheng, R. A. Caruso and M. Grätzel, ACS Nano, 2010, 8, 4420.
9. X. Haung, S. Han, W. Huang and X. Liu, Chem. Soc. Rev., 2013, 42, 173.
10. H. Sun, L. Pan, X. Piao and Z. Sun, J. Mater. Chem. A, 2013, 1, 6388.
11. E. Lee, J. Ryu and J. Jang, Chem. Commun., 2013, 49, 9995.
12. L. Liang, Y. Yulin, Z. Mi, F. Ruiqing, Q. LeLe, W. Xin, Z. Lingyun, Z. Xuesong and H. Jianglong, J. Solid State Chem., 2013, 198, 459.
13. J. Yu, Y. Yang, R. Fan, H. Zhang, L. Li, L. Wei, Y. Shi, K. Pan and H. Fu, J. Power Sources, 2013, 243, 436.
14. H.-Q. Wang, M. Batentschuk, A. Osvet, L. Pinna and C. J. Brabec, Adv. Mater., 2011, 23, 2675.
15. G. Zhu, X. Wang, H. Li, L. Pan, H. Sun, X. Liu, T. Lv and Z. Sun, Chem. Commun., 2012, 48, 958.
16. G.-B. Shan and G. P. Demopoulos, Adv. Mater., 2010, 22, 4373.
17. Q. Li, J. Lin, J. Wu, Z. Lan, Y. Wang, F. Peng and M. Huang, Electrochim. Acta, 2011, 56, 4980.
18. M. C. Maniquiz, K. Y. Jung and S. M. Jeong, J. Electrochem. Soc., 2010, 157, H1135.
19. S. H. Lee, D. S. Jung, J. M. Han, H. Y. Koo and Y. C. Kang, J. Alloys Compd., 2009, 477, 776.
20. C. Joshi, A. Rai, Y. Dwivedi and S. B. Rai, J. Lumin., 2012, 132, 806.
21. J. van de Lagemaat, N.-G. Park and A. J. Frank, J. Phys. Chem. B, 2000, 104, 2044.
22. R. Kern, R. Sastrawan, J. Ferber, R. Stangl and J. Luther, Electrochim. Acta, 2002, 47, 4213.
23. T. Peng, K. Fan and J. Chen, J. Phys. Chem. C, 2010, 114, 22346.

---