Graphene-incorporated quasi-solid-state dye-sensitized solar cells

Abstract

The pursuit of technological implementation with no sacrifice of photovoltaic performances is a persistent objective for dye-sensitized solar cells (DSSCs). Herein, we report an experimental realization of a graphene-incorporated quasi-solid-state DSSC comprising a graphene/TiO₂ anode, a graphene integrated polyacrylate–poly(ethylene glycol) (PAA–PEG) gel electrolyte with I⁻/I₃⁻ redox couples, and a graphene counter electrode. An efficiency of 3.62% is measured under global air mass irradiation for the quasi-solid-state solar cell with a graphene/TiO₂ photoanode, a PAA–PEG/graphene gel electrolyte, and a graphene counter electrode. The new concept, along with promising results, demonstrates the potential application of the new solar cells for cost-effective electricity generation.

---

1 Introduction

Dye-sensitized solar cell (DSSC),^1–3 an electrochemical device that converts solar energy into electricity, is considered as one of the most promising solutions to energy depletion, environmental pollution, and ecological destruction. It has aroused intensive interests over the past few decades due to its merits of easy fabrication, cost-effectiveness, and relatively high power conversion efficiency. Since the first prototype DSSC was reported by O'Regan and Grätzel in 1991,¹ an efficiency of 13% has been measured for the cell with a Pt counter electrode (CE) and dye-sensitized TiO₂ photoanode, along with a liquid electrolyte containing I⁻/I₃⁻ redox couples.⁴ However, the bleak commercialization prospect of the DSSC mainly arises from its intrinsic limitations in terms of poor stability, resulting from the leakage of liquid electrolyte^5–7 and the high expense of Pt CE.⁸ Therefore, there is still a crucial need to enhance its stability, as well as realize a cost-effective electricity generation.

Versatile strategies have been employed to shorten electron transfer path length, reserve liquid electrolyte with I⁻/I₃⁻ redox couples into three-dimensional polymer frameworks, and replace Pt species by alternative cost-effective candidates. To the best of our knowledge, no systematic ageing studies have been carried out on efficient DSSC by combining these three components,⁹ particularly using the same species. We present here the experimental realization of a graphene-incorporated quasi-solid-state DSSC comprising a graphene/TiO₂ anode,^10–15 a graphene-integrated polyacrylate–poly(ethylene glycol) (PAA–PEG) gel electrolyte with I⁻/I₃⁻ redox couples, and a graphene counter electrode. The original intention of this design is to reduce the fabrication cost and to simplify the preparation technique. An efficiency of 3.62% was measured under global air mass 1.5 (AM1.5G) irradiation for the DSSC with a graphene/TiO₂ photoanode, a PAA–PEG/graphene gel electrolyte, and a graphene CE. The results are far from optimal but the preliminary photovoltaic performances make the strategy promising in efficient DSSC applications.

2 Experimental

2.1 Preparation of graphene CEs

An aqueous solution containing graphene oxide powder was prepared by agitating 2 g of graphene oxide powders in 20 mL of standard buffer solution (sodium tetraborate buffer solution; pH = 9.89). The electrochemical deposition was carried out on a conventional CHI660E setup comprising an Ag/AgCl reference electrode, a Pt CE, and a working electrode of FTO glass substrates (sheet resistance 12 Ω sq⁻², purchased from Hartford Glass Co., USA). The cyclic voltammetry mode was employed to conduct the electrochemical deposition at a scan rate of 10 mV s⁻¹. The pristine Pt CE was purchased from Dalian HepatChroma SolarTech Co., Ltd and used as a standard. All the chemicals were used as received. Deionized water was used throughout the study.

2.2 Synthesis of PAA–PEG/graphene gel electrolytes

A mixture consisting of acrylic acid monomer (AA, 8 mL, analytical reagent grade) and PEG (M_w = 20 000, 4.4 g, analytical reagent grade) was made by agitating them together in 15 mL of deionized water in a water-bath at 80 °C. A cross-linker, N,N′-(methylene)bisacrylamide (NMBA, 0.008 g), and initiator, ammonium persulfate (APS, 0.225 g), were subsequently added to the abovementioned mixture with vigorous agitation. As the polymerization proceeded, the viscosity gradually increased. Once the viscosity of the PAA–PEG reached approximately 180 mPa s⁻¹, the reaction mixture was poured into a Petri dish and cooled to room temperature, resulting in the formation of a transparent elastic gel. The PAA–PEG membranes were then molded into a φ 2.5 cm die. After rinsing with deionized water, the membranes were dried under vacuum at 50 °C for 24 h.

Microporous PAA–PEG membranes were prepared by immersing the dried PAA–PEG membranes in deionized water for 72 h to reach their swelling equilibrium. The membranes were subsequently freeze-dried under vacuum at −60 °C over 72 h. Subsequently, the membranes were immersed in a liquid electrolyte composed of a redox electrolyte and graphene for 10 days until they reached absorption equilibrium. The graphene concentration was controlled at 1.33 g L⁻¹. The redox electrolyte consisted of 100 mM of tetraethylammonium iodide, 100 mM of tetramethylammonium iodide, 100 mM of tetrabutylammonium iodide, 100 mM of NaI, 100 mM of KI, 100 mM of LiI, 50 mM of I₂, and 500 mM of 4-tert-butyl-pyridine in 50 mL acetonitrile.

2.3 Preparation of TiO₂ photoanodes

Graphene/TiO₂ colloid for anode fabrication was synthesized by a sol-hydrothermal method.¹⁶ Subsequently, a layer of TiO₂ nanocrystal anode film with a thickness of ∼10 μm was prepared by coating the colloid by a doctor-blade technique. The resultant anodes were further immersed in an ethanol solution containing dissolved graphene for 10 min. Then, after rinsing with ethanol and drying under N₂ gas flow, the TiO₂/graphene photoanodes were sensitized by immersing into a 0.25 mM ethanol solution of N719 dye for 24 h. As a reference, the pristine TiO₂ photoanode was coated by pristine TiO₂ coating and sensitized by N719 dye.

2.4 Cell assembly and photovoltaic measurements

Each DSSC was fabricated by sandwiching a quasi-solid-state electrolyte between a dye-sensitized TiO₂/graphene anode and a graphene CE. As shown in Table 1, six strategies were utilized to assemble the solar cell devices. The photovoltaic testing of the DSSC with an active area of 0.25 cm² was carried out by measuring the J–V characteristic curves using a CHI660E Electrochemical Workstation under an irradiation of a simulated solar light from a 100 W Xenon arc lamp (XQ-500 W) in ambient atmosphere. The incident light intensity was controlled at 100 mW cm⁻² (calibrated by a standard silicon solar cell). A black mask with an aperture area of around 0.25 cm² was applied on the surface of the DSSCs to avoid stray light. Each J–V curve was repeatedly measured at least five times to control the experimental error within ±5%.

Table 1 Various strategies for DSSCs assembly

Strategies	Photoanodes	Electrolyte	CEs
Route i	Graphene/TiO2	PAA–PEG/graphene	Graphene
Route ii	TiO2	PAA–PEG/graphene	Graphene
Route iii	TiO2	Liquid electrolyte	Graphene
Route iv	TiO2	PAA–PEG	Pt
Route v	TiO2	PAA–PEG/graphene	Pt
Route vi	TiO2	Liquid electrolyte	Pt

---

2.5 Characterizations

An atomic force microscopy (AFM) image of the graphene electrode was obtained with an NS3A-02 Nanoscope III. The top-view morphology of the microporous PAA–PEG matrix was observed on a Zeiss Ultra Plus field emission scanning electron microscopy (SEM).

3 Results and discussion

Considering that the function of a CE is to collect electrons from an external circuit and to reduce triiodides (I₃⁻) into iodide (I⁻),¹⁷ one of the criteria to evaluate CE alternatives is a high electron-conduction and electrocatalytic activity. Among various CE candidates, conducting polymers [such as polyaniline (PANi) and polypyrrole (PPy)] and carbonaceous materials [such as graphene and carbon nanotube (CNT)]^18–22 have attracted growing attention. In our experiments, we employed graphene as the CE catalyst, which was made by the potentiostatic electrodeposition method. As shown in Fig. 1a, homogeneously distributed graphene sheets were observed, indicating that the graphene was successfully deposited on the FTO glass by the electrochemical deposition method.

Fig. 1 (a) AFM image of the graphene CE and (b) SEM photograph of the microporous PAA–PEG matrix.

The CV curves reflecting the electrocatalytic activities of the graphene CE on the I⁻/I₃⁻ redox species are shown in Fig. 2a. The peak shapes of the CV curves from the graphene CE are very similar to that of planar Pt,²³ revealing that the graphene CE displays an electrocatalytic function toward the I⁻/I₃⁻ redox couples. However, the peak position of the graphene electrode is more negative in comparison with that of planar Pt CE, demonstrating that the graphene CE has a similar electrocatalytic function but lower activity than the Pt electrode. Considering that the task of the graphene CE is to reduce I₃⁻ to I⁻ ions, the peak from the I₃⁻ + 2e → 3I⁻ reduction reaction can be utilized to evaluate the electrocatalytic activity of the graphene CE. From stacking the CV curves of the graphene CE at the scan rates of 20, 50, 75, and 100 mV s⁻¹, one can find an outward extension of all the peaks (Fig. 2a). By plotting the peak current density corresponding to I₃⁻ ↔ I⁻ versus the square root of the scan rate, as shown in Fig. 2b, linear relationships are observed. This result indicates that the redox reaction on the surface of the graphene CE is controlled by ionic diffusion in the electrolyte. This result also suggests that the adsorption of I⁻/I₃⁻ species is hardly affected by the redox reaction on the graphene electrode surface and that no specific interaction occurs between I⁻/I₃⁻ and the graphene CE.²⁴

Fig. 2 (a) CV curves of the graphene CE at varied scan rates; (b) the relationship between peak current densities and the square root of the scan rate. The electrolyte contained 50 mM LiI, 10 mM I₂, and 500 mM LiClO₄ in acetonitrile.

Fig. 3a shows the J–V curve (Route iii) for a DSSC made from the graphene CE and the photovoltaic parameters are summarized in Table 2. As a reference, the J–V curve (Route vi) from the DSSC employing the pure Pt electrode is also provided. The cell assembled by Route (iii) yields an optimal efficiency of 5.08%, a J_sc of 14.38 mA cm⁻², a V_oc of 0.67 V, and an FF of 52.6%, which are at the same levels mentioned in a previous report.²⁵ Although the efficiency is lower than the solar cell with a pristine Pt electrode, it is still modest for a Pt-free-CE-based cell.

Fig. 3 Characteristic J–V curves of the graphene-incorporated quasi-solid-state DSSCs recorded (a) at AM1.5G irradiation and (b) in the dark. (c) Nyquist and (d) Bode EIS spectra for the DSSCs obtained with different assembly strategies. R_s: sheet resistance; R_ct1: charge-transfer resistance at the CE/electrolyte interface; R_ct2: charge-transfer resistance at the anode/electrolyte interface; W: Nernst diffusion resistance corresponding to the I⁻/I₃⁻ redox couples; CPE: constant phase elements.

Table 2 Photovoltaic parameters of graphene-incorporated quasi-solid-state DSSCs from various strategies

Strategies	η (%)	Voc (V)	Jsc (mA cm^−2)	FF (%)
Route i	3.62	0.65	17.75	31.3
Route ii	4.07	0.65	14.82	42.3
Route iii	5.08	0.67	14.38	52.6
Route iv	6.07	0.70	12.15	71.3
Route v	6.86	0.69	14.23	70.0
Route vi	7.13	0.73	15.70	62.2

---

The main drawbacks of a liquid electrolyte are the leakage and evaporation of the organic solvent.²⁶ Generally, organic solvents such as acetonitrile are always employed to dissolve the iodide salts; however, the evaporation of the organic solvent results in a loss of the medium used for charge transport. Therefore, the interconversion of I₃⁻ ↔ I⁻, and therefore the recovery of the dye are restricted by the persistently increased electrical resistance. To address this problem, a conducting gel electrolyte from PAA–PEG/graphene was synthesized by sealing the liquid electrolyte into a three-dimensional (3D) framework of the microporous PAA–PEG matrix, as shown in Fig. 1b. The 3D gel matrix provides interconnected frameworks for charge transport; in addition, the charge transportation mode is similar to that in a liquid system. Moreover, the incorporated graphene is expected to form channels for the electrons flowing from the electrolyte/CE interface to the whole conducting gel electrolyte; in this fashion, the reaction area for the I₃⁻ → I⁻ reduction is markedly enhanced. More importantly, the charge diffusion path length is also shortened. After synergistic evaluation, it was considered that the utilization of PAA–PEG/graphene for DSSC application can significantly enhance the catalytic and charge-transfer kinetics.

The J–V characteristics of the quasi-solid-state DSSC made from PAA–PEG/graphene and PAA–PEG gel electrolytes are displayed in Fig. 3a (Route iv, Route v). An efficiency of 6.07% was recorded for the PAA–PEG gel electrolyte under the solar irradiation of 100 mW cm⁻², which is at the level commonly reported in the literature.^27,28 In addition, the efficiency was significantly enhanced by incorporating graphene into the PAA–PEG matrix, and an efficiency of 6.86% was recorded for the DSSC made using the PAA–PEG/graphene gel electrolyte, which is comparable to the efficiency of 7.13% for the DSSC with a liquid electrolyte and pristine Pt electrode. The enhancement in photovoltaic performances is attributed to the shortened charge-transfer path length and expanded catalytic area. EIS has been widely employed to explore the potential photoelectrochemical processes within a DSSC device. Here, it was carried out on the CHI660E Electrochemical Workstation at a frequency range of 0.01 Hz–10⁶ Hz and an ac amplitude of 10 mV at room temperature. The resultant impedance spectra were analyzed using the Z-view software. As shown in Fig. 3c, an equivalent circuit (see the inset of Fig. 3c) was used to fit the Nyquist plots to estimate the electron transport parameters. It can be seen that the R_ct1 of the conducting gel electrolyte is considerably lower than that of pristine PAA–PEG electrolyte. A lower R_ct indicates elevated charge-transfer ability at the electrolyte/CE interface. The rapid transport of refluxed electrons (the electrons from external circuit to CE) from the Pt electrode to the conducting gel electrolyte is expected to accelerate the reduction reaction of the I₃⁻ species.²⁹

In the present study, graphene is also imbibed into the nanoporous structure of TiO₂ film. The original intention of this design is to directly conduct photogenerated electrons from the conduction band of the TiO₂ nanocrystallites to graphene, and subsequently to the FTO layer. As shown in Fig. 3a, the cell assembled by Route (i) yields an efficiency of 3.62% (J_sc of 17.75 mA cm⁻², V_oc of 0.65 V, and FF of 31.3%). The J_sc extracted from Route (i) is considerably higher than 14.82 mA cm⁻² from Route (ii); however, the FF decreased. J_sc is highly dependent on the accumulative electron density on the conduction band of TiO₂ injected from the excited dyes;³⁰ in this fashion, the enhanced J_sc in Route (i) suggests that the electron loss has been reduced by the direct transport of electrons along the graphene nanosheets. It is known that the decreased FF refers to an increased internal resistance of a solar cell. As shown in Table 3, the R_ct1 has been increased from 15.29 to 39.37 Ω cm², and W, corresponding to the diffusion resistance of I⁻/I₃⁻ couples, has also been elevated from 13.4 to 53.8 Ω cm², which may be a crucial factor in reducing the cell efficiency. In the synthesis of graphene/TiO₂ anode, we find that the addition of hydrophobic graphene into hydrophilic TiO₂ colloid leads to a weak combination of TiO₂ and FTO glass. Therefore, the interfacial resistance at the FTO/TiO₂ interface is believed to be markedly enhanced, which was confirmed by the R_ct1 enhancement.

Table 3 Electrochemical parameters of the graphene-incorporated quasi-solid-state DSSCs from various strategies. R_ct1, R_ct2, and W are extracted from the EIS spectra of the solar cells

Strategies	Rs (Ω cm^2)	Rct1 (Ω cm^2)	Rct2 (Ω cm^2)	W (Ω cm^2)
Route i	0.53	39.37	47.32	53.8
Route ii	0.40	15.29	34.55	13.4
Route ii	0.33	11.31	30.67	250.1
Route iv	20.38	3.68	26.81	7.2
Route v	13.99	1.17	14.86	5.0
Route vi	0.76	2.67	4.29	5.7

---

In a real DSSC device with a graphene/TiO₂ anode, a PAA–PEG/graphene gel electrolyte, and a graphene CE, N719 dyes absorb photons under AM1.5 irradiation and release electrons from their excited state. As shown in Fig. 4, the photogenerated electrons suffer from successive migration on the conduction band of TiO₂, graphene, and FTO layer; whereas, the electrons transfer along the percolating networks formed by TiO₂ nanoparticles in the absence of graphene. There are abundant interfaces between adjacent TiO₂ nanoparticles, which act as traps for electron loss. The electron flow from the TiO₂ conduction band to graphene can significantly enhance the accumulative electron number in the FTO layer, and therefore the short-circuit current density in the solar cell. Moreover, the integrated graphene species can form interconnected channels for conducting electrons from the CE/electrolyte to the conducting gel electrolyte for I₃⁻ reduction; in this fashion, the reaction area for I₃⁻ → I⁻ conversion has been expanded from the CE/electrolyte interface to both the interface and gel electrolyte. More importantly, the recovered I⁻ ions are expected to experience a shorter path length for dye regeneration. Furthermore, the utilization of graphene CE can realize the cost-effective fabrication of DSSCs.

Fig. 4 Schematic illustration of the energy levels and electron migration process in the graphene-incorporated quasi-solid-state DSSC with a graphene/TiO₂ anode, a PAA–PEG/graphene gel electrolyte, and a graphene CE.

4 Conclusions

In summary, a new class of graphene-incorporated quasi-solid-state DSSCs has been successfully fabricated by incorporating graphene into a photoanode, gel electrolyte, and CE. An efficiency of 3.62% was measured under AM1.5G irradiation for the DSSC with a graphene/TiO₂ anode, a PAA–PEG/graphene gel electrolyte, and a graphene CE. It was demonstrated that the graphene in the anode can directly transfer photogenerated electrons from the conduction band of TiO₂ to the FTO layer, where the PAA–PEG/graphene gel electrolyte has an ability of shortening the charge diffusion path length and expanding the electrocatalytic reaction from the electrolyte/CE interface to the conducting gel electrolyte. The resultant graphene CE shows a similar catalytic function but lower activity in comparison with the Pt CE. The investigation presented here is far from the optimized one, but these profound superiorities along with cost-effectiveness, mild synthesis, and scalable materials show the promise for graphene-incorporated DSSCs as a promising technological implementation.

Acknowledgements

The authors would like to acknowledge financial supports from Fundamental Research Funds for the Central Universities (201313001, 201312005), Shandong Province Outstanding Youth Scientist Foundation Plan (BS2013CL015), Shandong Provincial Natural Science Foundation (ZR2011BQ017), and Research Project for the Application Foundation in Qingdao (13-4-198-jch).

References

1. B. O'Regan and M. Grätzel, Nature, 1991, 353, 737.
2. U. Bach, D. Lupo, P. Comte, J. E. Moser, F. Weissortel, J. Salbeck and M. Grätzel, Nature, 1998, 395, 583.
3. M. Grätzel, Nature, 2001, 414, 338.
4. S. Mathew, A. Yella, P. Gao, R. Humphry-Baker, B. F. E. Curchod, N. Ashari-Astani, I. Tavemelli, U. Rothlisberger, M. K. Nazeeruddin and M. Grätzel, Nat. Commun., 2014, 6, 242.
5. P. Wang, S. M. Zakeeruddin, R. Humphry-Baker, J. E. Moser and M. Grätzel, Adv. Mater., 2003, 15, 2101.
6. S. S. Yuan, Q. W. Tang, B. B. Hu, C. Q. Ma, J. L. Duan and B. L. He, J. Mater. Chem. A, 2014, 2, 2814.
7. J. Tang, J. Hua, W. Wu, J. Li, Z. Jin, Y. Long and H. Tian, Energy Environ. Sci., 2010, 3, 1736.
8. X. X. Chen, Q. W. Tang, B. L. He, L. Lin and L. M. Yu, Angew. Chem., Int. Ed., 2014, 3, 10799.
9. Y. S. Shi, Y. H. Wang, M. Zhang and X. D. Dong, Phys. Chem. Chem. Phys., 2011, 13, 14590.
10. G. Cheng, M. S. Akhtara, O. B. Yang and F. J. Stadlera, ACS Appl. Mater. Interfaces, 2013, 5, 6635.
11. G. Cheng and F. J. Stadler, J. Colloid Interface Sci., 2015, 438, 160.
12. G. Cheng, M. S. Akhtara, O. B. Yang and F. J. Stadlera, Electrochim. Acta, 2013, 113, 527.
13. P. Dong, Y. Zhu, J. Zhang, F. Hao, J. J. Wu, S. D. Lei, H. Lin, R. H. Hauge, J. M. Tour and J. Lou, J. Mater. Chem. A., 2014, 2, 20902.
14. L. J. Chen, C. X. Guo, Q. M. Zhang, Y. L. Lei, J. L. Xie, S. J. Ee, G. H. Gui, Q. L. Song and C. M. Li, ACS Appl. Mater. Interfaces, 2013, 5, 2047.
15. S. Mao, G. H. Lu and J. H. Chen, Nanoscale, 2015, 7, 6924.
16. Z. Y. Tang, Q. Liu, Q. W. Tang, J. H. Wu, J. L. Wang, S. Chen, C. Cheng, H. Yu, Z. Lan, J. M. Lin and M. L. Huang, Electrochim. Acta, 2011, 58, 52.
17. T. W. Hamann, R. A. Jensen, A. B. F. Martinson, H. V. Ryswyk and J. T. Hupp, Energy Environ. Sci., 2008, 1, 66.
18. B. L. He, Q. W. Tang, M. Wang, C. Q. Ma and S. S. Yuan, J. Power Sources, 2014, 256, 8.
19. G. T. Yue, L. Wang, X. A. Zhang, J. H. Wu, Q. W. Jiang, W. F. Zhang, M. L. Huang and J. M. Lin, Energy, 2014, 67, 460.
20. H. H. Zhang, B. L. He, Q. W. Tang and L. M. Yu, J. Power Sources, 2014, 275, 489.
21. L. L. Shao, M. Chen, T. Z. Ren and Z. Y. Yuan, J. Power Sources, 2015, 274, 791.
22. B. L. He, Q. W. Tang, J. H. Luo, Q. H. Li, X. X. Chen and H. Y. Cai, J. Power Sources, 2014, 265, 170.
23. Z. Y. Tang, Q. W. Tang, J. H. Wu, Y. Li, Q. Hui, M. Zheng, Y. M. Xiao, G. T. Yue, M. L. Huang and J. M. Lin, RSC Adv., 2012, 2, 5034.
24. Y. Saito, W. Kubo, T. Kitamura, Y. Wada and S. Yanagida, J. Photochem. Photobiol., A, 2004, 164, 153.
25. G. Q. Wang, J. Zhang, S. Kuang, S. M. Liu and S. P. Zhuo, J. Power Sources, 2014, 269, 473.
26. J. Yoon, D. Kang, J. Won, J. Y. Park and Y. S. Kang, J. Power Sources, 2012, 201, 395.
27. P. Y. Chen, C. P. Lee, R. Vittal and K. C. Ho, J. Power Sources, 2010, 195, 3933.
28. X. Chen, Q. Li, J. Zhao, L. Qiu, Y. Zhang, B. Sun and F. Yan, J. Power Sources, 2012, 207, 216.
29. C. Bu, Q. Tai, Y. Liu, S. Guo and X. Zhao, J. Power Sources, 2013, 221, 78.
30. Y. Y. Duan, Q. W. Tang, B. L. He, R. Li and L. M. Yu, Nanoscale, 2014, 6, 12601.

---