Incorporating MoFe alloys into reduced graphene oxide as counter electrode catalysts for dye-sensitized solar cells

3.2.1 High-resolution scanning electron microscopy measurements

Fig. 4a shows the graphene-coated FTO electrode which has a 3D network structure with a graphene flake size in the range of 100 nm to 500 nm. Fig. 4(b–h) present Mo_xFe_1-x (0 ≤ x ≤ 1) alloys with different volume ratios of precursors under DPR. It was easily found that the Mo_xFe_1-x (0 ≤ x ≤ 1) alloys were immobilized on graphene flake surfaces without any agglomerations. We could not find the difference in the particle sizes for the samples prepared by different volume ratios of precursors under DPR. However, the formation of NPs on RGO surface is beneficial for reducing the diffusion impedance (Z_w) (Li et al., 2010). This is due to the triiodide ions can be rapidly reduced to iodide ions under catalysis of Mo_xFe_1-x (0 ≤ x ≤ 1) alloys to accelerate diffusion of triiodide ions (Li et al., 2010).

Close

Fig. 4 SEM images of MoFe alloy/RGO CEs: (a) GO, (b) Mo₀Fe₁/RGO, (c) Mo_0.1Fe_0.9/RGO, (d) Mo_0.3Fe_0.7/RGO, (e) Mo_0.5Fe_0.5/RGO, (f) Mo_0.7Fe_0.3/RGO, (g) Mo_0.9Fe_0.1/RGO and (h) Mo₁Fe₀/RGO.

Export to PPT

The intrinsic electrocatalytic activities of the CEs were evaluated by cyclic voltammograms (CVs) using a three-electrode system, as shown in Fig. 5(a). The values of the reduction currents (J_red) and the peak-to-peak separations (ΔE) were determined to estimate the catalytic activities of the electrodes. The calculated results are presented in Table 1. As can be seen, the |J_red| value increased with an increase in the volume content of the Mo precursors in the binary solution from 0% to 70%, whereas it decreased with further increases in the volume content of Mo precursors in a binary solution. The highest |J_red| value was 1.375 mA for the Mo_0.7Fe_0.3/RGO electrode. The obtained results indicate that the highest diffusion current of ions was found for Mo_0.7Fe_0.3/RGO electrode (Xiao and Han, 2015). Furthermore, it also demonstrated the lowest Z_w and the largest surface active area were obtained for Mo_0.7Fe_0.3/RGO electrode among of electrodes (Jeong et al., 2012). It is known that the ratio of |oxidation current (J_oxd)/J_red| presents the reversibility of the redox reaction toward I^-/I₃^- (Tributsch, 2004). The values obtained for different electrodes are presented in Table S1 (Supporting Information). All |J_oxd/J_red| values with different electrodes are close to 1.0. The obtained results indicate a more reversible redox reaction for I₃⁻ ↔ I⁻, resulting in improved long-term stability (He et al., 2014). In order to verify the stability of the electrodes, we carried out measurements of the CV curves over time of the redox system for one electrode. The results are presented in Fig. 5(b). The values of J_red and the J_oxd were extracted from the experimental current–voltage curves recorded over 100 cycles (Figs. S2–S8, Supporting Information). It is clearly seen that the Mo_xFe_1−x (0 ≤ x ≤ 1) alloys/RGO nanohybrid CEs did not exhibit noticeable changes in the values of J_red and J_oxd over 100 cycles. The results indicate that there was no degradation of the nanohybrids and no detachment of the films from the FTO glass, suggesting a better corrosion resistance in the iodine-based electrolyte (Huang et al., 2015). These results demonstrate the high stability of the Mo_xFe_1-x (0 ≤ x ≤ 1) alloys/RGO films under electrochemical reaction conditions.

Close

Fig. 5 (a) CV curves with different working electrodes at a scan rate of 50 mV s⁻¹, and (b) plots of J_red and J_ox as a function of the cycle number.

Export to PPT

The ΔE value is another important to reflect the electrocatalytic activities of the CEs that is inversely proportional to the standard electrochemical rate constant of a redox reaction (Zheng et al., 2014). Accordingly, a small ΔE means a rapid reduction of triiodide ions, which is important for accelerating dye regeneration. This enhances electron injection and thus the accumulation of electron density in the conduction band of TiO₂, resulting in a large photocurrent density (J_sc) (Zheng et al., 2014; 2017; Li et al., 2018). The ΔE values are extracted from CV and listed in Table 1. It was found that the smallest ΔE value was obtained for Mo_0.7Fe_0.3/RGO electrode. There is no significant difference in ΔE values with other electrodes. The observations of the CV curves were in good agreement with the R_ct levels, the open-circuit voltages (V_oc) and J_sc recorded for the DSCs (Dao, 2017).

To further investigate the charge transfer process at the electrolyte-electrode interface, we conducted EIS studies, as shown in Fig. 6(a and b). Note that the EIS measurements were conducted using different symmetrical dummy cells fabricated with two identical Mo_xFe_1-x/RGO electrodes. The equivalent circuit, which was used to fit the spectra with the Z-view software, is depicted in the inset of Fig. 6(b). It was recorded that there are three semicircles in the extended equivalent circuit. The first one was described at a high-frequency semicircle for electron transport resistance in the graphene layer (R_trns) and constant phase element in the graphene layer (CPE_trap). The second one, a middle semicircle, is attributed to the charge transfer resistance at the electrode/electrolyte interface (R_ct) and constant phase element at the electrode/electrolyte interface (CPE_dl). The last one, a low-frequency semicircle, is attributed to the Warburg impedance (W) (Kwon et al., 2012). The extended equivalent circuit also indicates the lower value of CPE_trap compared with CPE_dl because the trap sites only partially occupy the surface, as shown in Table S2 (Supporting Information). Simulated data of EIS spectra calculated by equivalent circuits are listed in Table 1.

Close

Fig. 6 (a and b) Nyquist plots of symmetrical dummy cells with two identical CEs. Note that the actual data and fitted data are presented by symbols and solid lines, respectively, (c) Bode phase plots of symmetrical dummy cells, and (d) Tafel curves of different symmetrical dummy cells that are similar to those used for the EIS measurements.

Export to PPT

As can be seen in Table 1, R_trns values of the reference electrodes are 1.57 and 1.18 Ω for Mo₀Fe₁/RGO and Mo₁Fe₀/RGO electrodes, respectively. The R_trns becomes small when the volume ratio of Mo to Fe is increased from 0:1 to 0.7:0.3 and it becomes larger with further increases in the volume ratio of Mo and Fe from 0.7:0.3 to 1:0. Rhee et al. reported that the R_trns is inversely proportional to electrode conductivity (Kwon et al., 2012). It was expected that the electrode conductivity becomes high when the volume ratio of Mo to Fe is increased from 0:1 to 0.7:0.3 and it becomes low with further increases in the volume ratio of Mo and Fe from 0.7:0.3 to 1:0. As a result, the highest electron conductivity was obtained for Mo_0.7Fe_0.3/RGO electrode. The reason for the increased conductivity of nanohybrid films is the synergistic effect of graphene with the deposited NPs (Dao et al., 2015).

Table 1 also presents the change in R_ct for all CEs under study. The change in conductivity displays the synergistic effect of graphene with the deposited Mo_xFe_1-x (0 ≤ x ≤ 1) NPs. Given that the R_ct at the interface between the Mo_0.7Fe_0.3/RGO CE and the electrolyte is the minimum value among the CEs because the synergistic effect of graphene with the Mo_0.7Fe_0.3 NPs was stronger than the effects of graphene with the other alloys. As can be seen in Table 1 and Table S2 (Supporting Information), the R_ct values followed the sequence of Mo_0.7Fe_0.3/RGO (0.11 Ω) < Mo_0.5Fe_0.5/RGO (1.32 Ω) < Mo_0.9Fe_0.1/RGO (5.48 Ω) < Mo_0.3Fe_0.7/RGO (7.65vΩ) < Mo₀Fe₁/RGO (8.84 Ω) < Mo₁Fe₀/RGO (11.42 Ω) < Mo_0.1Fe_0.9/RGO (11.81 Ω). The obtained results indicate the reaction rate for the reduction of triiodide ions followed the sequence of Mo_0.7Fe_0.3/RGO > Mo_0.5Fe_0.5/RGO > Mo_0.9Fe_0.1/RGO > Mo_0.3Fe_0.7/RGO > Mo₀Fe₁/RGO > Mo₁Fe₀/RGO > Mo_0.1Fe_0.9/RGO. This trend was further confirmed by the change in the electron lifetime (τ) data for the triiodide ion reduction (τ = 1/2πf_max, where f_max is the peak frequency in the Bode EIS plots) and the exchange current density (J_o = RT/nFR_ct, where R, T, n, and F are the gas constant, absolute temperature, number of electrons involved in the reduction of iodide electrolyte, and Faraday’s constant, respectively, in Tafel measurements), as shown in Fig. 6(c) for the Bode EIS plots and Fig. 6(d) for the Tafel curves. Analyses of the Bode EIS plots and Tafel measurements revealed that the trends of the f_max and J_o values followed the sequence of Mo_0.7Fe_0.3/RGO < Mo_0.5Fe_0.5/RGO < Mo_0.9Fe_0.1/RGO < Mo_0.3Fe_0.7/RGO < Mo₀Fe₁/RGO < Mo₁Fe₀/RGO < Mo_0.1Fe_0.9/RGO. The obtained results indicate the trends associated with the values of τ and J_o were similar to that found for the R_ct values.

To correlate the properties of the all developed catalysts with the photovoltaic (PV) characteristics of devices exploiting the developed nanohybrid materials as CEs, PV experiments were conducted. The J-V characteristics are described in Fig. 7(a), and the PV parameters are summarized in Table 2. It was found that the efficiencies become high when the volume ratio of Mo to Fe is increased from 0:1 to 0.7:0.3 and it becomes low with further increases in the volume ratio of Mo and Fe from 0.7:0.3 to 1:0. The highest efficiency was 5.44% for device assembled with Mo_0.7Fe_0.3/RGO electrode. This efficiency is also higher than that of the cell using Pt electrode (5.36%), as shown in Figs. S9, S10 (Supporting Information). Note that the efficiency of the cell using RGO is 1.26%. The efficiency of the cells fabricated with the Mo₀Fe₁/RGO and Mo₁Fe₀/RGO CEs showed the similar values of 4.53%. The efficiencies of DSCs fabricated by various graphene CEs in other studies are presented in Table S3 (Supporting Information). As shown in Table S3, the obtained efficiency was higher than those of cells using RGO (Zhang et al., 2016), Ni/graphene (2.39%), Ni_0.75Cu_0.25-decorated graphene (5.1%), Ni_0.6Cu_0.4-decorated graphene (2.87%), Cu_0.75Ni_0.25-decorated graphene (1.24%) (Motlak et al., 2015), FeNi/RGO (Sim et al., 2018), FeSn/RGO (Oh et al., 2018). It was comparable to cells using CoPd/RGO (Oh et al., 2017), CoNi/RGO (Park et al., 2017b), but still lower than those of devices using Pt-based alloy/RGO CEs (Dao et al., 2017, 2016; Yoon et al., 2016; Jin et al., 2016). The change in the V_oc values was explained by the change in ΔE (Dao et al., 2011) and the recombination of the conduction band electrons with ions in the redox electrolyte (Dao, 2017) because the rate of recombination is second order in I₃^- concentration, which is attributed to the disproportionation reaction 2I₂ → I₃⁻ + I⁻ with I₂ as the electron acceptor in the back-reaction (Huang et al., 1997). Note that, the positive shift in the I₃^-/I^- redox energy level results in the decrease of V_oc in DSCs. Accordingly, a decrease in ΔE means an increase in V_oc.

Close

Fig. 7 Characteristic current density-voltage curves of DSCs with different CEs measured (a) under standard conditions and (b) in the dark; (c) IPCE curves of the different DSCs.

Export to PPT

Fig. 7(b) shows the dark current of DSCs based on different CEs at a high forward bias. Yun et al. reported that a slightly lower dark current could contribute to a relatively low V_oc in the device (Yun et al., 2013). The dark current of all DSCs did not show any abnormal differences. The obtained results indicated that the developed materials would not react with the I^-/I₃^- electrolyte (Yun et al., 2013). The obtained results are in good agreement with the J-V characteristics. The enhancement of the J_sc values could be ascribed to the enhanced catalytic activity of the developed CEs (Dao and Choi, 2016) and to the decreases in the Z_w levels of the developed CEs (Li et al., 2011). To confirm the improvement in the J_sc value, incident photon-to-current efficiency (IPCE) performance tests were conducted, as depicted in Fig. 7(c). It was found that the current density of the devices for the Mo₀Fe₁/RGO, Mo_0.7Fe_0.3/RGO, and Mo_0.9Fe_0.1/RGO CEs were 11.15, 13.01 and 12.87 mA cm⁻², respectively. The calculated integral currents are in good agreement with the J_sc values. Therefore, this result is well aligned with the SEM, CV, EIS, and Tafel measurements.