PEDOT:PSS assisted preparation of a graphene/nickel cobalt oxide hybrid counter electrode to serve in efficient dye-sensitized solar cells

Abstract

Graphene/nickel cobalt oxide (Gr/NiCo₂O₄) hybrids with a nanostructure were prepared by the use of an in situ hydrothermal route, and applied as a counter electrode (CE) with poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) assisted preparation (here abbreviated as P-A) in dye-sensitized solar cells (DSSCs) for the first time. The surface morphology of the bilayer TiO₂ photoanode was confirmed using scanning electron microscopy. The superior structural characteristics of the photoanode were advantageous to fast mass transport for the electrolyte, increasing the contact area between the electrolyte and active materials, and enabling the (P-A) Gr/NiCo₂O₄ CE to speed up the reduction of triiodide to iodide. Also, the CE with a superior nanostructure was evaluated by electrochemical characterization, which indicated that the (P-A) Gr/NiCo₂O₄ CE possessed excellent electrocatalytic activity in the iodide/triiodide electrolyte and a lower charge transfer resistance of 3.04 ± 0.02 Ω cm² compared to that of the Pt electrode (3.63 ± 0.02 Ω cm²). Under optimum conditions, the DSSC based on the (P-A) Gr/NiCo₂O₄ CE achieved a remarkable power conversion efficiency of 8.10%, which is about 8.7% higher than that of the Pt-based DSSC (7.45%). The (P-A) Gr/NiCo₂O₄ CE can be considered as a promising alternative CE for Pt-free DSSCs.

---

1. Introduction

Dye-sensitized solar cells (DSSCs), electrochemical devices converting solar energy to electricity characterized by high conversion efficiency, are honoured as a promising solution to energy depletion, environmental pollution and ecological destruction due to them being low cost and environmentally friendly, and having good durability.^1–3 Typically, the device consists of dye-adsorbed TiO₂ nanoparticles, filled with an electrolyte containing iodide/triiodide (I⁻/I₃⁻) redox couples, and a counter electrode (CE) with excellent catalytic ability. Conventionally, the CE, as the most important component in a DSSC, collects the electrons from the external circuit and catalyzes the reduction of I₃⁻ to I⁻ between the CE and electrolyte interface.^4,5

Graphene and its hybrids have been believed to be an efficient approach for enhancing the photoelectric and electrochemical performance of DSSCs.^6–11 The large surface area, unique features of strong mechanical strength, and high electrical and thermal conductivity associated with graphene usually translate into high efficiency for photoelectric chemical devices,^12,13 so it is highly desirable for applications in DSSCs.¹⁴ Moreover, as one of the transition metal oxides, NiCo₂O₄ with a very remarkable electrical conductivity has attracted considerable attention as a low overpotential catalyst for oxygen reduction reactions and a high-performance electrode material in Li-ion batteries, supercapacitors and fuel cells due to its low cost and environmental friendliness.^15–21 It is therefore expected to offer richer redox reactions, including contributions from both nickel and cobalt ions, than those of the monometallic nickel oxide or cobalt oxide. Simultaneously, the electrochemical performance is also affected by the morphology of the NiCo₂O₄ catalyst, e.g., 1D nanomaterials have demonstrated better performance for Li–O₂ batteries than microparticles.^22,23

Therefore, taking into account all of these factors, in the search for more robust Pt-free CEs, here we reported graphene/nickel cobalt oxide (Gr/NiCo₂O₄) hybrids with a nanostructure through an in situ hydrothermal route and prepared an efficient Gr/NiCo₂O₄ CE with poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) assisted (here abbreviated as P-A) preparation by means of a slurry-coating technique. This type of electrode was generally anticipated to have many advantages, such as good electrical conductivity, easy electrolyte penetration, and high electrochemical activity. The electrocatalytic ability and photoelectric properties of the (P-A) Gr/NiCo₂O₄-based DSSC was systematically investigated. The DSSC assembled with the (P-A) Gr/NiCo₂O₄ CE exhibited a considerably improved performance with a photoelectric conversion efficiency of 8.10% under irradiation of 100 mW cm⁻² (AM 1.5 G).

2. Experimental

2.1 Materials

Nickel(II) nitrate hexahydrate (Ni(NO₃)₂·6H₂O, 98%), cobalt(II) nitrate hexahydrate (Co(NO₃)₂·6H₂O, 98.5%), poly(styrenesulfonate) (PSS), 3,4-ethylenedioxythiophene (EDOT), N-methyl-2-pyrrolidinone (NMP), polyvinyl-difluoride (PVDF), potassium persulfate (K₂S₂O₈) and titanium tetrachloride (TiCl₄) were purchased from Shanghai Chemical Agent Ltd, China. All reagents are of analytical reagent grade. The dye N719 cis-di(isothiocyanato)-bis-(2,2′-bipyridyl-4,4′-dicarboxylate)ruthenium(II)bis-tetrabutylammonium was obtained from Solaronix SA (Switzerland).

2.2 Preparation of the (P-A) Gr/NiCo₂O₄ hybrid counter electrode

A typical synthesis procedure of PEDOT:PSS is outlined below. 6.95 g PSS (18 wt%) aqueous solution was mixed with 75 mL deionized water at room temperature and treated with nitrogen bubbling for 0.5 h. Then 0.5 g EDOT and 0.0077 g Fe₂(SO₄)₃·9H₂O were added to initiate the polymerization. The mixture was stirred at room temperature for 2 h, after which a further 0.1673 g K₂S₂O₈ was added. After additional reaction for 12 h, the mixture was put into a dialytic bag for 48 h to remove the excess reactants. Then a blue solution of PEDOT:PSS was obtained.

Briefly, the preparation of the (P-A) Gr/NiCo₂O₄ counter electrode is outlined as follows. Firstly, 1.0 mmol Ni(NO₃)₂·6H₂O, 2.0 mmol Co(NO₃)₂·6H₂O, 6.0 mmol NH₄F and 15 mmol CO(NH₂)₂ were dissolved in 38 mL deionized water. After ultrasonication and stirring for 30 min a homogeneous solution was achieved. After that, 1.5 wt% of graphene was added into the reaction solution. This was further sonicated for another 30 min before being transferred into a Teflon-lined autoclave, and then heated in an oven at 100 °C for 10 h without intentional control of the ramping or cooling rate. The Gr/NiCo₂O₄ hybrids with a black color were collected by filtering and washed with ethanol and distilled water 5 times, followed by annealing at 200 °C for 4 h. Then, Gr/NiCo₂O₄ as the active material was mixed with acetylene black and PVDF in a weight ratio of 8 : 1 : 1. The paste of the above mentioned mixtures was prepared by the use of deionized water, ethanol, NMP and PEDOT:PSS as the solvent, respectively, and kept stirring for 12 h. Subsequently, the as-prepared paste was coated on an FTO substrate (8 Ω cm⁻², 350 nm thickness, Hartford Glass Co., USA) using a doctor blade method with an area of 0.64 cm² to form a film with a thickness of ∼6 μm. The obtained electrodes were dried at 100 °C for 12 h in a vacuum oven and marked as (W-A) Gr/NiCo₂O₄, (E-A) Gr/NiCo₂O₄, (N-A) Gr/NiCo₂O₄ and (P-A) Gr/NiCo₂O₄ CEs, respectively. For comparison, the Gr and NiCo₂O₄ electrodes mentioned in the manuscript were also prepared by the use of PEDOT:PSS as the solvent and marked as Gr* and NiCo₂O₄* CEs with a similar method. The Pt electrode was prepared using a three-electrode system.

2.3 Fabrication of the DSSC

The bilayer TiO₂ photoanode is prepared as described previously.^24,25 Firstly, a thin TiO₂ dense layer with a particle size of 36 nm and a thickness of 0.5 μm was prepared by spin-coating the solution (4 mL tetrabutyltitanate, 5 mL isopropanol and 1 mL acetylacetone in 25 mL n-butanol) on an FTO substrate at 1500 rpm for 30 s, followed by sintering at 450 °C for 30 min in air. Secondly, the TiO₂ nanorods with a diameter of about 300 nm grew on the dense layer as follows. Two pieces of FTO with a TiO₂ dense layer were placed into 60 mL of tetrabutyltitanate solution containing 1 mL of tetrabutyltitanate, 30 mL of deionized water and 30 mL of hydrochloric acid (36.5–38% by weight) at an angle against the wall of the Teflon-liner with the conducting side facing down. The hydrothermal synthesis was conducted at 160 °C for 6 h in an electric oven. The dye was loaded by immersing the bilayer TiO₂ photoanode in 0.3 mM of N719 ethanol solution for 12 h. Thus the dye-sensitized TiO₂ photoanode with a thickness of 4–5 μm was obtained. The DSSC was fabricated by injecting the liquid electrolyte (0.05 M of I₂, 0.1 M of LiI, 0.6 M of tetrabutylammonium iodide and 0.5 M of 4-tert-butylpyridine in acetonitrile) into the aperture between the dye-sensitized TiO₂ electrode and the CE. The two electrodes were clipped together and wrapped with thermoplastic hot-melt Surlyn (the schematic of the DSSC is shown in Fig. 1).

Fig. 1 Schematic of the DSSC with the bilayer TiO₂ photoanode and the (P-A) Gr/NiCo₂O₄ counter electrode.

2.4 Characterization

The surface morphology of the samples was observed using JSM-7001F field emission scanning electron microscopy (SEM). Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) were conducted with the use of a computer-controlled electrochemical analyzer (CHI 660D, Shanghai Chenhua Device Company, China). The electrolyte used in the DSSC test was also injected into the dummy cells for the EIS measurements. EIS was carried out under simulated open-circuit conditions at ambient atmosphere, sealing with thermoplastic hot-melt Surlyn and leaving an exposed area of 0.64 cm². The frequency of the applied sinusoidal AC voltage signal was varied from 0.1 Hz to 10⁵ Hz and the corresponding amplitude was kept at 5 mV in all cases.

The photovoltaic test of the DSSC with an exposed area of 0.4 × 0.7 cm² was carried out by measuring the photocurrent–photovoltage (J–V) character curve under white light irradiation of 100 mW cm⁻² (AM 1.5 G) from the solar simulator (XQ-500W, Shanghai Photoelectricity Device Company, China) in ambient atmosphere. The fill factor (FF) and the photoelectric conversion efficiency (PCE) of the DSSC were calculated according to the following equations:

where J_sc is the short-circuit current density (mA cm⁻²), V_oc is the open-circuit voltage (V), P_in is the incident light power (mW cm⁻²), and J_max (mA cm⁻²) and V_max (V) are the current density and voltage at the point of maximum power output in the J–V curve, respectively.

The diffusion coefficient (D_n) of the CE in the electrolyte is estimated in light of the Randles–Sevcik equation as illustrated in eqn (3):

I_pc = Kn^1.5AC(D_n)^0.5v^0.5 (3)

where I_pc is the cathodic current density, K is the constant of 2.69 × 10⁵, n means the number of electrodes contributing to the charge transfer, A is the area of the CE, and C and v represent the concentration of I₃⁻ species and the scan rate, respectively.

The exchange current density (J₀) for the Tafel curves can be estimated according to eqn (4):

where R is the gas constant, T is the temperature, F is Faraday’s constant, and R_ct is the charge transfer resistance.

3. Results and discussion

3.1 Surface morphology and composition of TiO₂

Fig. 2a presents the image of the TiO₂ dense layer, which shows that the TiO₂ dense layer with a size of about 36 nm was prepared using a spin-coating method to reduce the charge recombination and improve the fill factor of the DSSC. Top and side views in Fig. 2b and c show that the top surface of the TiO₂ nanorods is nearly perpendicular to the FTO substrate and hexagonal in shape with square top facets. After 6 h of growth, the average diameter and length for the nanorods, as determined from the SEM images, are about 300–340 nm and 3.5 μm, respectively. This distance favors the photoproduction electronic transmission. Moreover, the TiO₂ nanorods that grew on the TiO₂ dense layer by the use of an in situ hydrothermal route are of the anatase phase as referred to in the XRD patterns in Fig. 2d.

Fig. 2 SEM images of the TiO₂ dense layer (a), and TiO₂ nanorods (b and c), and XRD patterns of the TiO₂ nanorods (d).

3.2 UV-vis absorption spectra

The UV-vis absorption spectra of TiO₂ with nanorod and nanoparticle structures were measured and are displayed in Fig. 3a. As expected, TiO₂ nanorods show a more obvious absorption enhancement for the characteristic spectrum with its fundamental absorption of the Ti–O bond in the ultraviolet light range from 300 to 400 nm, mainly because of the more convenient channel for charge transport provided by the TiO₂ nanorods. Fig. 3b shows the specific surface areas (Brunauer–Emmett–Teller, BET) for the different morphologies of TiO₂ by nitrogen adsorption–desorption. It is clear that the BET surface area of the TiO₂ nanorods (105.4 m² g⁻¹) is a much larger specific surface area than that of the TiO₂ nanoparticles (71.3 m² g⁻¹), indicating that the TiO₂ nanorods have very uniform mesopores. Such a high BET surface area could facilitate the dye-loading capacity and probably endow the TiO₂ nanorod-based DSSCs with a higher current density.²⁶ Simultaneously, the absorption spectra of dye desorbed from the photoanodes with different structures are exhibited in Fig. 3c. It is clear that a stronger absorption of the dye for the TiO₂ photoanode with a nanorod structure can be obtained than that of the nanoparticle structure, which well proves the enhanced specific surface area for the TiO₂ nanorods as shown in Fig. 3b. Furthermore, the estimated adsorbed amount of dye N719 in the photoanode with the nanoparticle structure is about 0.52 × 10⁻⁷ mol cm⁻², while it can reach 0.86 × 10⁻⁷ mol cm⁻² for the TiO₂ nanorod photoanode.

Fig. 3 UV–vis absorption spectra (a and c) and the specific surface areas (b) of the TiO₂ nanorods and TiO₂ nanoparticles.

3.3 Surface morphology and composition of the samples

Fig. 4a and b show the SEM images of the NiCo₂O₄ nanorods and Gr/NiCo₂O₄ nanoparticles. As seen in Fig. 4a, the NiCo₂O₄ particles can exhibit a beautiful rod-like nanostructure. It is notable that when graphene was added into the hydrothermal reaction solution, the morphology of the Gr/NiCo₂O₄ nanoparticles interestingly changed to shorter nanorods and blended uniformly with the graphene flakes as shown in Fig. 4b. The unique nanostructure guarantees the full contact area and fast mass transport for the electrolyte, and enables highly enhanced catalytic activity toward the reduction of I₃⁻, which logically results in a considerably improved photovoltaic performance.²⁷ The EDS analysis of the sample indicates the presence of C, O, Co, and Ni elements with strong signals as shown as in Fig. 4c, which are responsible for the formation of the Gr/NiCo₂O₄ nanoparticles.

Fig. 4 SEM images of the NiCo₂O₄ nanorods (a), and the Gr/NiCo₂O₄ nanoparticles (b), and EDS of the Gr/NiCo₂O₄ nanoparticles (c).

3.4 Electrochemical properties of the counter electrode

As depicted in Fig. 5, two pairs of redox peaks are observed in the CV curves for the different CEs at a scan rate of 50 mV s⁻¹. The left and right redox pair peaks are associated with I₃⁻ + 2e⁻ ↔ 3I⁻ and 3I₂ + 2e⁻ ↔ 2I₃⁻, respectively. The main function of the CE is to be responsible for speeding up the reduction of I₃⁻ to I⁻, which is highly related to the left redox pair peaks. The cathodic current density (I_pc) and the peak to peak separation (E_pp) are two crucial parameters for comparing the electrocatalytic ability of various CEs.^28,29 Fig. 5a exhibits the CV curves of the Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ CEs at a scan rate of 50 mV s⁻¹. To our knowledge, the |I_pc| is positively correlated with the catalytic ability of the electrodes, and |E_pp| is inversely correlated with the electrocatalytic activity of the CEs. It was noticed that the (P-A) Gr/NiCo₂O₄ CE shows a higher I_pc than that of the Pt, Gr* and NiCo₂O₄* electrodes in Fig. 5a, indicating that the (P-A) Gr/NiCo₂O₄ CE effectively acted as a catalyst in the reaction of the I⁻/I₃⁻ electrolyte. The |E_pp| of the above-mentioned CEs increases in the order of (P-A) Gr/NiCo₂O₄ (0.23 ± 0.01 V) < Pt (0.28 ± 0.01 V) < NiCo₂O₄* (0.32 ± 0.01 V) < Gr* (0.51 ± 0.01 V), which is immediately responsible for the lower overpotential loss in the (P-A) Gr/NiCo₂O₄ CE than the others since the highly conductive PEDOT:PSS was incorporated into the Gr/NiCo₂O₄ nanoparticles. In the case of the (P-A) Gr/NiCo₂O₄ CE, NiCo₂O₄ evenly coated on the graphene surface and still basically kept the morphology of the graphene network, which would effectively facilitate the electron transport and the diffusion of the redox electrolyte within the CE. In addition, the synergistic effect of PEDOT:PSS also provides an efficient electron transport network and enhances the conductivity of the counter electrode. Furthermore, the reduction reaction of the I⁻/I₃⁻ redox couples is due to a diffusion-controlled transport process, and the diffusion coefficient (D_n) is positively correlated with I_pc from the Randles–Sevcik equation as illustrated in eqn (3).³⁰ Though the NiCo₂O₄* CE has a larger |E_pp| than the Gr* CE, it possesses the lowest |I_pc| among all the CEs. Thus, from this, the D_n of I₃⁻ for the samples follows the order of (P-A) Gr/NiCo₂O₄ > Pt > Gr* > NiCo₂O₄*. The larger D_n for the (P-A) Gr/NiCo₂O₄ CE than that of the other CEs presumably originates from its well-controlled surface structure and the improvement of the conductivity of the CE.

Fig. 5 CV for the Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ CEs (a), and the Gr/NiCo₂O₄ CE with different solvents (b) at a scan rate of 50 mV s⁻¹.

Fig. 5b presents CV of the Gr/NiCo₂O₄ CEs depending on deionized water, ethanol, NMP and PEDOT:PSS as the solvent, respectively. It is clear that the influence of the solvent on the electrochemical ability is great and the relationships between the solvent and the |I_pc| and |E_pp| values for the CEs are summarized in Table 1. Among them, |I_pc| of the above-mentioned CEs increases in the order of (W-A) Gr/NiCo₂O₄ < (E-A) Gr/NiCo₂O₄ < (N-A) Gr/NiCo₂O₄ < (P-A) Gr/NiCo₂O₄, while |E_pp| increases in the opposite order. The difference of electrochemical ability for the CEs prepared with various solvents mainly comes from the following several aspects. PVDF, as the binder of the conductive paste, dissolves easily in NMP and ethanol, but can not dissolve in deionized water very well. However, ethanol volatilizes quickly and causes its conductive paste to be unstable in air. Compared to NMP and ethanol, PEDOT:PSS with an excellent conductivity and catalytic ability could better dissolve PVDF and provide a good network for the charge transport.³¹

Table 1 The electrochemical performance of the samples based on the different CEs with various solvents

Solvents	|Ipc| (mA cm^−2)	|Epp| (V)	J0 (mA cm^−2)	Rs (Ω cm^−2)	Rct (Ω cm^−2)
Deionized water	2.69	0.42	1.08	8.46 ± 0.02	4.37 ± 0.02
Ethanol	2.97	0.39	1.16	8.13 ± 0.02	4.02 ± 0.02
NMP	3.24	0.34	1.20	7.69 ± 0.02	3.84 ± 0.02
PEDOT:PSS	3.41	0.23	1.32	6.82 ± 0.02	3.04 ± 0.02

---

The electrochemical stability of a CE is very important to the long-term performance of the DSSC. A 40-cycle consecutive cycle scan test has been performed for the (P-A) Gr/NiCo₂O₄ CE at a scan rate of 50 mV s⁻¹ using the same electrolyte mentioned in the experimental section. After 40 cycles, the normalized cathodic and anodic peak current densities remain scarcely changed and maximum redox peak current densities show a good linear relationship as presented in Fig. 6, suggesting that the (P-A) Gr/NiCo₂O₄ CE is stable and not corroded in the I⁻/I₃⁻ redox electrolyte.^32,33

Fig. 6 40-cycle CV (a), and the relationship between the cycles and the maximum redox peak currents for the (P-A) Gr/NiCo₂O₄ electrode (b).

CV of the (P-A) Gr/NiCo₂O₄ electrode at various scan rates from 30 to 90 mV s⁻¹ and the relationship between the redox peak current versus the square root of the scan rates are shown in Fig. 7. It can be obviously observed that both the oxidation and reduction peaks for the I_pc and potential gradually and regularly changed with the increasing scan rates, and the I_pc versus (scan rate)^1/2 plots have a good linear relationship. This phenomenon indicates that the adsorption of iodide species is almost not affected by the redox reaction on the (P-A) Gr/NiCo₂O₄ electrode surface.^34,35

Fig. 7 CV for the (P-A) Gr/NiCo₂O₄ electrode with different scan rates (from inner to outer: 30, 50, 70, and 90 mV s⁻¹) (a). The redox peak current versus the square root of the scan rates (b).

The electrochemical impedance properties of different CEs are measured using their symmetric cells. Fig. 8 exhibits the Nyquist plots of the symmetrical Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ CEs for the I⁻/I₃⁻ electrolyte and the corresponding EIS parameters are also listed in Table 2. The inset in Fig. 8 shows the equivalent circuit, in which R_s is the resistance value at the onset point of the first semicircle, R_ct is the radius of the first semicircle, and the semicircle at a low frequency represents the Nernst diffusion impedance (Z_w) corresponding to the diffusion resistance of the I⁻/I₃⁻ redox species.³⁶ As far as we know, R_s and R_ct, two crucial parameters for comparing the electrocatalytic ability of different CEs, are inversely correlated with the catalytic ability of the electrodes. It can be seen from Table 2 that the R_s values associated with the Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ CEs are 7.55 ± 0.02, 8.76 ± 0.02, 9.98 ± 0.02, and 6.82 ± 0.02 Ω cm², respectively. It is noted that the R_s of the Pt CE is comparable to that of the Gr* and NiCo₂O₄* CEs, but higher than that of the (P-A) Gr/NiCo₂O₄ CE. The lower R_s for the (P-A) Gr/NiCo₂O₄ CE means a better electrical conductivity and more contact surface area than the other CEs. On the other hand, the CE with the smaller R_ct means a lower overpotential for an electron transferring from the CE to the electrolyte, a facile electron transfer, and great electrochemical ability. From Table 2, the R_ct values for the Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ CEs are 3.63 ± 0.02, 4.61 ± 0.02, 6.78 ± 0.02, and 3.04 ± 0.02 Ω cm², respectively. The Pt CE has a better electrochemical ability than the Gr* and NiCo₂O₄* CEs, however, with the graphene and PEDOT:PSS introduction, a smaller R_ct is seen than that of the Pt CE produced from the (P-A) Gr/NiCo₂O₄ CE. Therefore, the graphene and PEDOT:PSS combinations are able to compete with the Pt CE for the DSSC. Additionally, it should be noted that Z_w for the (P-A) Gr/NiCo₂O₄ CE (1.41 ± 0.02 Ω cm²) is a little larger than that of the Pt electrode (1.39 ± 0.02 Ω cm²). This can be attributed to the conductive polymer’s relatively low electrical conductivity compared with that of the Pt catalyst.

Fig. 8 Nyquist plots of the symmetrical Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ CEs for the I⁻/I₃⁻ redox couple.

Table 2 EIS parameters of various CEs made from the impedance spectra

Electrodes	Rs (Ω cm^−2)	Rct (Ω cm^−2)	Zw (Ω cm^−2)	J0 (mA cm^−2)	|Ipc| (mA cm^−2)
Pt	7.55 ± 0.02	3.63 ± 0.02	0.63 ± 0.02	1.24	3.16
Gr*	8.76 ± 0.02	4.61 ± 0.02	0.51 ± 0.02	1.07	2.58
NiCo2O4*	9.98 ± 0.02	6.78 ± 0.02	1.39 ± 0.02	0.91	2.52
(P-A) Gr/NiCo2O4	6.82 ± 0.02	3.04 ± 0.02	1.41 ± 0.02	1.32	3.41

---

In addition to the EIS of the different materials, the influence of the solvent on the electrochemical performance was also studied, and the key parameters of the EIS are given in Table 1. As seen in Table 1, the values of R_s and R_ct for the CEs both follow the order of (W-A) Gr/NiCo₂O₄ > (E-A) Gr/NiCo₂O₄ > (N-A) Gr/NiCo₂O₄ > (P-A) Gr/NiCo₂O₄. These results are in good agreement with the CV, and ascribed to the same reasons.

To further investigate the catalytic activities of the samples, Tafel polarization curves were obtained for the electrodes with Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ in a dummy cell similar to the ones used in the EIS measurement as displayed in Fig. 9. Theoretically, the curve at low potentials (|U| < 0.12 V) represents the polarization zone, the one at the middle potentials (with a sharp slope) represents the Tafel zone, and the other at the high potentials (horizontal part) represents the diffusion zone. In a Tafel polarization curve, usually the logarithmic current density as a function of potential is presented, and thus the corresponding exchange current density (J₀) for each CE can be evaluated from the extrapolated intercepts of the anodic and cathodic branches of the corresponding Tafel curves at the exchange current density axes, as summarized in Table 2. J₀ for the (P-A) Gr/NiCo₂O₄ electrode is higher compared to that of the Pt, Gr* and NiCo₂O₄* CEs; this means that the (P-A) Gr/NiCo₂O₄ electrode can trigger the reduction of I₃⁻ to I⁻ more effectively than the electrodes mentioned above. R_ct can also be calculated by eqn (4) according to the relationship between J₀ and R_ct for the reduction of I₃⁻ ions to I⁻ ions.³⁷ According to the J₀ values and eqn (4), it can be deduced that the order of the R_ct is NiCo₂O₄* > Gr* > Pt > (P-A) Gr/NiCo₂O₄, which is greatly consistent with the CV and EIS results. This indicates that the (P-A) Gr/NiCo₂O₄ CE can be further explored to replace the expensive Pt as the catalyst for the Pt-free CE in DSSCs, in view of the fact that it is much cheaper than Pt-based DSSCs and exhibits just as high a power conversion efficiency. Furthermore, the influence of solvents on the performance of the Tafel polarization curve was also investigated and J₀ for the above-mentioned CEs follows the same order as |I_pc| (see Table 1).

Fig. 9 Tafel curves of the symmetrical Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ CEs.

3.5 Photovoltaic performance of DSSCs with different counter electrodes

The J–V characteristics of the DSSCs with different CEs were measured under the optimum conditions at 100 mW cm⁻² (AM 1.5 G) irradiation and are presented in Fig. 10a. The corresponding parameters are listed in Table 3. The DSSC with a Pt CE shows a J_sc of 15.23 mA cm⁻², a V_oc of 0.73 V, a FF of 0.67 and a PCE of 7.43%. The DSSCs based on the Gr* and NiCo₂O₄* CEs exhibit the poor PCEs of 5.75% and 5.73% due to their poor electrocatalytic activity, and the detailed photovoltaic performance parameters are described in Table 3. When PEDOT:PSS was added to graphene to assist with the conductivity of the catalyst, the DSSC with the (P-A) Gr/NiCo₂O₄ CE shows a better J_sc value of 16.12 mA cm⁻², a V_oc of 0.75 V, a FF of 0.67 and a PCE of 8.10%. Comparing the SEM image of the NiCo₂O₄ particles, it may be said that the unique nanostructure of the Gr/NiCo₂O₄ nanoparticles has a larger effective surface area, which could lead to the higher J_sc in favor of the (P-A) Gr/NiCo₂O₄-based DSSC. With the assisting PEDOT:PSS, the electrocatalytic activity is further enhanced, thus the device logically shows a better photovoltaic performance for the synergistic effect of graphene and PEDOT:PSS. Besides, the optimized (P-A) Gr/NiCo₂O₄-based DSSC shows a higher V_oc than that of the other DSSCs, which can be evidenced by the dark current characterization in Fig. 10a. The (P-A) Gr/NiCo₂O₄-based DSSC possesses a smaller dark current at the same potential, suggesting a lower electron recombination rate for the optimized (P-A) Gr/NiCo₂O₄-based DSSC than for the other DSSCs. Fig. 10b shows the EIS of the DSSCs based on the Pt, Gr*, NiCo₂O₄* and (P-A) Gr/NiCo₂O₄ CEs and the corresponding parameters are listed in Table 3, in which R_ct1 means the interfacial charge transfer resistance at the CE and electrolyte, and R_ct2 represents the interfacial charge transfer resistance at the dye-sensitized photoanode and the electrolyte. From Fig. 10b and Table 3, the (P-A) Gr/NiCo₂O₄-based DSSC exhibits a smaller R_ct1 and R_ct2 than the DSSCs with the Pt, Gr*, and NiCo₂O₄* CEs. This indicates that the (P-A) Gr/NiCo₂O₄ CE facilitates fast electron transport at the interface between the I⁻/I₃⁻ electrolyte and the electrodes, and it also can be deduced that the (P-A) Gr/NiCo₂O₄-based DSSC can indeed improve the charge recombination and has a more outstanding effect than the other DSSCs.

Fig. 10 J–V characteristics (a), and EIS spectra (b), of the DSSCs fabricated with the Pt, Gr*, NiCo₂O₄*and (P-A) Gr/NiCo₂O₄ CEs under the standard illumination. Inset in (a) is the DSSC assembled with (P-A) Gr/NiCo₂O₄ CE.

Table 3 The influence of different CEs on the electrochemical performance and photoelectric properties of the DSSCs

Electrodes	Voc (V)	Jsc (mA cm^−2)	FF	PCE (%)	Rct1 (Ω cm^−2)	Rct2 (Ω cm^−2)
Pt	0.73	15.23	0.67	7.45	8.37 ± 0.02	13.78 ± 0.02
Gr*	0.71	13.49	0.60	5.75	9.91 ± 0.02	22.77 ± 0.02
NiCo2O4*	0.72	12.44	0.64	5.73	16.08 ± 0.02	20.36 ± 0.02
(P-A) Gr/NiCo2O4	0.75	16.12	0.67	8.10	6.83 ± 0.02	10.94 ± 0.02

---

4. Conclusions

An efficient Gr/NiCo₂O₄ counter electrode with PEDOT:PSS assisted preparation as a Pt-free CE was used in a DSSC. An extensive electrochemical property study indicates that the (P-A) Gr/NiCo₂O₄ CE exhibits an amazing electrocatalytic ability and low charge transfer resistance for the reduction of I₃⁻, which can compare with that of a Pt CE. Comparative studies found that the dye-sensitized TiO₂ nanorods show a stronger absorbing capacity for UV-visible light and a greater loading capability to the dye. The (P-A) Gr/NiCo₂O₄-based DSSC produces a remarkable improvement in the short-circuit photocurrent, open-circuit voltage and power conversion efficiency. Under the optimum conditions, the DSSC based on the (P-A) Gr/NiCo₂O₄ CE achieved an enhanced power conversion efficiency of 8.10% under irradiation of 100 mW cm⁻², which is comparable to that of the Pt-based DSSC (7.76%). The research presented here is far from being optimized but these profound advantages along with the low-cost synthesis and scalable materials promise the new CE to have great potential and to be a strong candidate in robust DSSCs.

Acknowledgements

The authors are very grateful to the joint support by the National Natural Science Foundation of China (No. U1504624). This work is also supported by China Postdoctoral Science Foundation Funded Project (No. 2015M572102) and the Scientific Research Found of Henan Provincial Department of Science and Technology (No. 122300410107).

References

1. B. O’Regan and M. Grätzel, Nature, 1991, 353, 737–740.
2. A. Yella, H. W. Lee, H. N. Tsao, C. Yi, A. K. Chandiran, M. K. Nazeeruddin, E. W. Diau, C. Y. Yeh, S. M. Zakeeruddin and M. Grätzel, Science, 2011, 334, 629–634.
3. S. Mathew, A. Yella, P. Gao, R. Humphry-Baker, B. F. E. Curchod, N. Ashari-Astani, I. Tavernelli, U. Rothlisberger, M. K. Nazeeruddin and M. Grätzel, Nat. Chem., 2014, 6, 242–247.
4. M. Grätzel, Nature, 2001, 414, 338–344.
5. F. Bella, C. Gerbaldi, C. Barolo and M. Grätzel, Chem. Soc. Rev., 2015, 44, 3431–3473.
6. H. Choi, H. Kim, S. Hwang, Y. Han and M. Jeon, J. Mater. Chem., 2011, 21, 7548–7551.
7. L. Kavan, J.-H. Yum and M. Grätzel, Nano Lett., 2011, 11, 5501–5506.
8. H. Sun, Y. Luo, Y. Zhang, D. Li, Z. Yu, K. Li and Q. B. Meng, J. Phys. Chem. C, 2010, 114, 11673–11679.
9. M. K. Wang, A. M. Anghel, B. Marsan, N. C. Ha, N. Pootrakulchote, S. M. Zakeeruddin and M. Grätzel, J. Am. Chem. Soc., 2009, 131, 15976–15977.
10. M. X. Wu, Y. D. Wang, X. Lin, N. Yu, L. Wang, L. Wang, A. Hagfeldt and T. L. Ma, Phys. Chem. Chem. Phys., 2011, 13, 19298–19301.
11. L. Kavan, J. H. Yum and M. Grätzel, ACS Nano, 2010, 5, 165–172.
12. G. T. Yue, J. H. Wu, Y. M. Xiao, M. L. Huang, J. M. Lin, L. Q. Fan and Z. Lan, Electrochim. Acta, 2013, 92, 64–70.
13. J.-Y. Lin, C.-Y. Chan and S.-W. Chou, Chem. Commun., 2013, 49, 1440–1442.
14. G. Veerappan, K. Bojan and S.-W. Rhee, ACS Appl. Mater. Interfaces, 2011, 3, 857–862.
15. C. Yuan, J. Li, L. Hou, X. Zhang, L. Shen and X. W. Lou, Adv. Funct. Mater., 2012, 22, 4592–4597.
16. X. Y. Liu, Y. Q. Zhang, X. H. Xia, S. J. Shi, Y. Lu, X. L. Wang, C. D. Gu and J. P. Tu, J. Power Sources, 2013, 239, 157–163.
17. B. Hua, W. Zhang, J. Wu, J. Pu, B. Chi and L. Jian, J. Power Sources, 2010, 195, 7375–7379.
18. J. Li, S. Xiong, Y. Liu, Z. Ju and Y. Qian, ACS Appl. Mater. Interfaces, 2013, 5, 981–988.
19. Y. Q. Wu, X. Y. Chen, P. T. Ji and Q. Q. Zhou, Electrochim. Acta, 2011, 56, 7517–7522.
20. B. Sun, J. Zhang, P. Munroe, H.-J. Ahn and G. Wang, Electrochem. Commun., 2013, 31, 88–91.
21. J. R. Mayhew, D. J. Bozym, C. Punckt and I. A. Aksay, ACS Nano, 2010, 4, 6203–6211.
22. A. Debart, A. J. Paterson, J. Bao and P. G. Bruce, Angew. Chem., Int. Ed., 2008, 47, 4521–4524.
23. L. L. Zhang, X. B. Zhang, Z. L. Wang, J. J. Xu, D. Xu and L. M. Wang, Chem. Commun., 2012, 48, 7598–7600.
24. W. Q. Wu, B. X. Lei, H. S. Rao, Y. F. Xu, Y.-F. Wang, C. Y. Su and D. B. Kuang, Sci. Rep., 2013, 3, 1–7.
25. G. T. Yue, J. H. Wu, Y. M. Xiao, J. M. Lin, M. L. Huang, Z. Lan and L. Q. Fan, Electrochim. Acta, 2012, 85, 182–186.
26. Z. Lan, J. H. Wu, J. M. Lin and M. L. Huang, J. Mater. Chem., 2011, 21, 15552–15557.
27. G. T. Yue, F. R. Tan, F. M. Li, C. Chen, W. F. Zhang, J. H. Wu and Q. H. Li, Electrochim. Acta, 2014, 149, 117–125.
28. G. T. Yue, J. H. Wu, Y. M. Xiao, M. L. Huang, J. M. Lin and J.-Y. Lin, J. Mater. Chem. A, 2013, 1, 1495–1501.
29. J. Y. Lin, W. Y. Wang and S. W. Chou, J. Power Sources, 2015, 282, 348–357.
30. A. Hauch and A. Georg, Electrochim. Acta, 2001, 46, 3457–3466.
31. Z.-Q. Li, Y.-P. Que, L.-E. Mo, W.-C. Chen, Y. Ding, Y.-M. Ma, L. Jiang, L.-H. Hu and S.-Y. Dai, ACS Appl. Mater. Interfaces, 2015, 7, 10928–10934.
32. G. T. Yue, W. F. Zhang, J. H. Wu and Q. W. Jiang, Electrochim. Acta, 2013, 112, 655–662.
33. W. J. Ke, G. J. Fang, H. Tao, P. L. Qin, J. Wang, H. W. Lei, Q. Liu and X. Z. Zhao, ACS Appl. Mater. Interfaces, 2014, 6, 5525–5530.
34. S. Biallozor and A. Kupniewska, Electrochem. Commun., 2000, 2, 480–486.
35. Y. Saito, W. Kubo, T. Kitamura, Y. Wada and S. Yanagida, J. Photochem. Photobiol., A, 2004, 164, 153–157.
36. H.-M. Chuang, C.-T. Li, M.-H. Yeh, C.-P. Lee, R. Vittal and K.-C. Ho, J. Mater. Chem. A, 2014, 2, 5816–5824.
37. G. T. Yue, J. H. Wu, J.-Y. Lin, Y. M. Xiao, J. M. Lin, M. L. Huang and Z. Lan, Carbon, 2013, 55, 1–9.

---