Effects of Ti3C2Tx MXene Addition to a Co Complex/Ionic Liquid-Based Electrolyte on the Photovoltaic Performance of Solar Cells

1. Introduction

Two-dimensional (2D) materials are typically crystalline solids consisting of a single layer of atoms. These materials are considered promising for various applications, which explains why they remain the focus of research. MXenes with a general formula of M_n+1X_nT_x (where n = 1−3; M denotes a transition metal; X is either carbon or nitrogen; and T_x indicates surface terminal groups such as −OH, −F, −Cl, and/or −O−) were created by selectively etching the “A” layers from layered MAX phases (M_n+1AX_n, where A is usually any element from among Cd, Al, Si, P, S, Ga, Ge, As, In, Sn, Tl, Pb, and S [groups 12–16]) and can be easily solution-processed in aqueous or polar organic solvents due to their hydroxyl- or oxygen-terminated surfaces [1,2,3,4]. Following the production of multilayered Ti₃C₂Tx MXene by etching the Al layers from the Ti₃AlC₂ MAX phase in 2011 at Drexel University [1], numerous related research results have been reported in the fields of energy storage, sensors, light-emitting diodes, electromagnetic shielding, and environmental applications [2,3,4]. In addition, MXenes have been extensively studied in relation to applications concerning solar cells, given their metallic conductivity, excellent charge carrier mobility, high optical transmittance, and tunable work function [5,6,7,8]. Among the various MXenes, Ti₃C₂T_x is the most commonly studied in terms of third-generation solar cells, such as dye-sensitized solar cells (DSCs) [9,10,11], perovskite solar cells [12,13,14], and polymer solar cells [15,16]. A conventional DSC is composed of a dye-adsorbed TiO₂ layer on a transparent electrode (i.e., a working electrode), a liquid electrolyte, and a Pt catalytic layer on a conductive electrode (i.e., a Pt counter electrode). Light absorption in dye molecules leads to the formation of excitons (electron–hole pairs), and the excited electrons are injected into the TiO₂ layer. The photoinjected electrons and the holes in the dye molecules are transported to the electrodes via the TiO₂ layer and the electrolyte, respectively. Finally, the electrons and holes are collected in the electrodes, allowing electron flows through the external circuits to occur [17,18]. The hole-conducting electrolyte is comprised of redox couples and electrical additives [19,20]. The redox couples, such as I⁻/I₃⁻, Co⁺²/Co⁺³, Cu⁺¹/Cu⁺², and Ni⁺³/Ni⁺⁴, are reduced near the Pt counter electrode and oxidized near the excited dye molecules, thereby allowing for hole collection and dye regeneration, respectively. Electrical additives such as 4-tert-butylpyridine (TBP) and cations (lithium [Li⁺] or guanidinium [C(NH₂)³⁺]) represent another important ingredient in a liquid electrolyte for enhancing the photovoltaic parameters of cells. These additives can control the potential of the redox couple, the surface state of the TiO₂ semiconductor, the shift in the conduction band edge, and the interfacial charge recombination through being incorporating in small amounts [20]. Ti₃C₂T_x MXene was introduced as an additive for electrolytes to improve the photovoltaic performance of quasi-solid-state DSCs [21,22]. Sun et al. reported that, via the addition of Ti₃C₂T_x MXene to a quasi-solid-state electrolyte composed of an I⁻/I₃⁻ redox couple and a melamine-formaldehyde (MF) sponge, the average power conversion efficiency (PCE) of the DSCs under a room light condition (1000 lux) was improved by 26.92% from that of the reference cell without MXene (23.35%) [21]. It was also reported that a PCE of 29.94% under a condition of 1000 lux was achieved through the incorporation of both reduced graphene oxide and Ti₃C₂T_x to a quasi-solid-state electrolyte containing an I⁻/I₃⁻ redox couple, polyethylene oxide (PEO), and poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP) [22].

In this study, we report the effects of Ti₃C₂T_x MXene addition to a liquid electrolyte on the photovoltaic performance of cells. Ti₃C₂T_x-dispersed liquid electrolytes based on a metal complex (tris(2-(1H-pyrazol-1-yl)-4-tert-butylpyridine)cobalt(III) tri[bis(trifluoromethane)sulfonimide] [FK209]) as a source of Co³⁺ and an ionic liquid (1-methyl-3-propylimidazolium iodine [MPII]) as a source of I⁻ (iodide) were first prepared. Then, DSCs with Ti₃C₂T_x-dispersed Co³⁺/I⁻ liquid electrolytes (redox mediators) were fabricated and their photovoltaic properties were compared with those of the reference cell without Ti₃C₂T_x MXene. To the best of our knowledge, this is the first report on the effects of Ti₃C₂T_x addition to Co complex (Co³⁺)/ionic liquid (I⁻)-based redox mediators. The reported photovoltaic parameters (short-circuit current [J_sc], open-circuit voltage [V_oc] and fill factor [FF]) of the DSCs with Ti₃C₂T_x MXene are summarized in Table 1, including those of our devices with FK209 (Co³⁺)/MPII (I⁻)/Ti₃C₂T_x-based liquid electrolytes.

2. Results and Discussion

2.1. Photovoltaic Performance of DSCs Based on Ti₃C₂T_x-Incorporated Co³⁺/I⁻ Redox Mediators

In a previous report, we revealed that, through the simple mixing of MPII and FK209, triiodide (I₃⁻) and Co²⁺(FK209) were produced via a chemical reaction (1) between the iodide (I⁻) of MPII and Co³⁺(FK209), where Co²⁺(FK209) and Co³⁺(FK209) are the Co²⁺ and Co³⁺ ions of FK209, respectively. Since the Co³⁺(FK209) was almost fully converted into Co²⁺(FK209), the redox mediators contained both I₃⁻ and Co²⁺(FK209) as well as I⁻ (originating from non-reacted MPII) [23]. Therefore, hole conduction occurs through reactions (2)–(5) when as-prepared DSCs are exposed to the AM 1.5 condition, indicating that the I⁻/I₃⁻ redox couple is involved in the dye regeneration (reaction [4]) and hole collection (reaction [5]). Here, we code the as-prepared DSCs based on the I⁻/I₃⁻ redox couple as I-DSCs.

3I⁻ + 2Co³⁺(FK209) → I₃⁻ + 2Co²⁺(FK209)

(1)

2Dye + hν → 2Dye^*

(2)

2Dye^* → 2Dye⁺ + 2e⁻ (TiO₂)

(3)

3I⁻ + 2Dye⁺ → I₃⁻ + 2Dye

(4)

I₃⁻ + 2e⁻ (Pt) → 3I⁻

(5)

Using the FK209 (Co³⁺)/MPII (I⁻) redox mediators with or without Ti₃C₂T_x MXene, DSCs were fabricated and the variations in the photovoltaic parameters were investigated. The average photovoltaic performance measured using the four I-DSCs (i.e., the as-prepared DSCs) are compared in Figure 1 and Table 2, while the raw data are presented in Table S1 in the electronic supplementary information (ESI). Through the incorporation of Ti₃C₂T_x MXene into the Co³⁺/I⁻ liquid redox mediators, the average J_sc value of the I-DSCs was enhanced, whereas the average V_oc value was decreased when compared with the values of the device without MXene, as shown in Figure 1a,b, respectively. As a result, the average PCE of the I-DSCs with Ti₃C₂T_x was very similar to that of the cells without Ti₃C₂T_x.

Moreover, it was also found that chemical reactions (6) and (7) occurred through the exposure of the I-DSCs to light for a certain time [23]. Under illumination, the Co²⁺(FK209) that was produced via reaction (1) reduced the oxidized dye (Dye⁺), thereby regenerating Co³⁺(FK209) at the dye/redox mediator interface (reaction [6]). The resulting Co³⁺(FK209) diffused to the counter electrode and then reduced to Co²⁺(FK209) through receiving an electron from the platinized FTO electrode (reaction [7]), indicating that the Co²⁺/Co³⁺ redox couple is related to the dye regeneration (reaction [6]) and hole collection (reaction [7]). This may lead to an increase in the V_oc value of the Co²⁺/Co³⁺-based cell when compared with that of the I⁻/I₃⁻-based cell, as the potential gap between the conduction band edge (CBE) of the TiO₂ and the redox potential (1.06 V versus a normal hydrogen electrode [NHE]) of the Co²⁺/Co³⁺(FK209) is wider than that (0.35 V versus a NHE) of the I⁻/I₃⁻ electrolyte [24,25,26]. Here, we code the DSCs based on the Co²⁺/Co³⁺ redox couple as Co-DSCs, which were transformed from the I-DSCs via exposure to AM 1.5 light. To determine the time taken to convert the I-DSCs into Co-DSCs, we measured the V_oc variations with the light-exposure time of the I-DSCs. As shown in Figure S1, the V_oc values improved with an increasing exposure time and then saturated after over a period of 150 min. This indicates that hole conduction mainly occurred through the action of the Co²⁺/Co³⁺ redox couple after exposure of the I-DSCs to light for over 150 min. To investigate effects of the MXene incorporation in Co-DSCs, the I-DSCs with or without Ti₃C₂T_x MXene were exposed to AM 1.5 condition for 150 min and their photovoltaic performance was measured. As a reference, the V_oc values of the Co-DSCs was sharply increased when compared with those of the I-DSCs due to the wider potential gap between the TiO₂’s CBE and the Co²⁺/Co³⁺(FK209)’s redox potential, as mentioned above. This suggests that the Co²⁺/Co³⁺(FK209) rather than the I⁻/I₃⁻ redox couple participates in the hole conduction (reactions [6] and [7]) in the Co-DSCs.

2Co²⁺(FK209) + 2Dye⁺ → 2Co³⁺(FK209) + 2Dye

(6)

2Co³⁺(FK209) + 2e⁻ (Pt) → 2Co²⁺(FK209)

(7)

Through the addition of MXene, a substantial enhancement of the J_sc and a slight decrement in the V_oc values were observed in the Co-DSCs with Ti₃C₂T_x when compared with those of the reference cell without MXene, as shown in Figure 1a,b, respectively. There was no meaningful variation in the FF value, as demonstrated in Figure 1c. As a consequence, an improvement in the PCE was recorded in the MXene-incorporated Co-DSCs because the increase in the J_sc overcame the decrement in the V_oc value, as shown in Figure 1d. Among the four cells, we focused on the best-performing cells to reveal the origins of the improvement in the PCE via the incorporation of Ti₃C₂T_x MXene into Co³⁺/I⁻ liquid redox mediators. Here, we denote the best-performing Co-DSCs with or without Ti₃C₂T_x MXene as Ti₃C₂T_x-Co-_bDSC or bare Co-_bDSC, respectively. Figure 2 presents the current density–voltage (J–V) curves of the Ti₃C₂T_x- and bare Co-_bDSC, while the cell performance is compared in Table 3.

2.2. Effects of Ti₃C₂T_x Incorporation on the J_sc of Co-DSCs

Through the incorporation of Ti₃C₂T_x MXene into the Co³⁺/I⁻ redox mediator, the J_sc value (18.45 mA/cm²) of the Ti₃C₂T_x-Co-_bDSC was substantially improved from that of the bare Co-_bDSC (14.41 mA/cm²). It is believed that the conductive Ti₃C₂T_x MXene present in the Co²⁺/Co³⁺(FK209) electrolyte reduced the charge transfer resistance and improved the electrocatalytic performance between the Co³⁺(FK209) and the platinized FTO electrode, thereby leading to a reduction in the internal resistances and, therefore, an increase in the J_sc value of the Ti₃C₂T_x-Co-_bDSC [21,22,27]. To confirm this, impedance spectroscopic (EIS) analysis was performed for the bare and Ti₃C₂T_x-Co-_bDSC. Figure 3a shows the Nyquist plots of the EIS spectra of the Co-DSCs, as measured at the open-circuit condition under under AM 1.5 one-sun illumination, providing the series (R_s) and internal resistances. Three typical arcs, corresponding to the resistance (R₁) of the redox reaction at the platinized FTO/electrolyte interface in the high-frequency region, the electron transfer resistance (R₂) at the TiO₂/dye/electrolyte interface in the medium-frequency region, and the ionic diffusion resistance (R₃) within the electrolyte, were observed. A smaller R₁ value (around 3.18 Ω) was measured in the Ti₃C₂T_x-Co-_bDSC when compared with that of the bare Co-_bDSC (around 4.11 Ω). This indicates that the incorporation of Ti₃C₂T_x MXene can lead to an improvement in the electrocatalytic performance, causing an increase in the hole collection efficiency, when considering that the R₁ value is related to the reduction of the Co³⁺(FK209) by the Pt catalyst (chemical reaction [7]) [21,22,27]. In addition, it is considered that the reduced resistance (R₁) associated with chemical reaction (7) can effectively produce Co²⁺(FK209), resulting in the promotion of dye regeneration (chemical reaction [6]) [27].

Furthermore, a shift in the TiO₂’s CBE could be estimated from the dark current curves of the DSCs [28,29,30]. Figure 4 presents the dark current–voltage curves of the bare- and Ti₃C₂T_x-Co-_bDSCs. The onset potential of the dark current for the bare Co-_bDSC was estimated to be approximately 0.771 V, whereas the onset potential for the Ti₃C₂T_x-Co-_bDSC was shifted to approximately 0.661 V. Through the incorporation of Ti₃C₂T_x MXene, a lower onset potential was recorded, indicating that the potential difference between the work function of the FTO and the TiO₂’s CBE in the Ti₃C₂T_x-Co-_bDSC was smaller than that in the bare Co-_bDSC. This suggests that the TiO₂’s CBE in the Ti₃C₂T_x-Co-_bDSC was located at a more positive potential than that in the bare Co-_bDSC. It is thought that the Ti₃C₂T_x MXene particles are adsorbed on the TiO₂ surface, causing a surface dipole to be formed and resulting in a positive shift in the TiO₂’s CBE. This positive shift (away from the vacuum level) of the CBE in the Ti₃C₂T_x-Co-_bDSC may lead to the more favorable injection of photoexcited electrons from the dye into the TiO₂ due to the larger potential difference between the dye’s lowest unoccupied molecular orbital level and the TiO₂’s CBE, thereby resulting in an improvement in the electron injection efficiency [31]. Thus, it is considered that the more positive shift of the TiO₂’s CBE in the Ti₃C₂T_x-Co-_bDSC yielded a higher electron injection efficiency than the bare Co-_bDSC. A similar result was reported in relation to the incorporation of Ti₃C₂T_x MXene into the mesoporous TiO₂ layer, which induced a positive shift in the TiO₂’s CBE, leading to an enhancement of the electron injection efficiency [32]. Overall, we attributed the enhanced J_sc value in the Ti₃C₂T_x-Co-_bDSC to the increases in both the hole collection and the dye-regeneration efficiency, which resulted from the reduced internal resistances, as well as to the improvement in the electron injection efficiency due to the positive shift of the TiO₂’s CBE.

2.3. Effects of Ti₃C₂T_x Incorporation on the V_oc of Co-DSCs

The V_oc value (0.760 V) of the Ti₃C₂T_x-Co-_bDSC was decreased when compared with that of the bare Co-_bDSC (0.780 V). As expressed in Equation (8), the V_oc value of the DSCs under constant illumination can be expressed as a function of the quasi-Fermi level (E_Fn) and the dark value (E_F0), which is related to the thermal energy (k_BT; 4.11 × 10⁻²¹ J at 25°C), Boltzmann constant (k_B), absolute temperature (T), positive elementary charge (e; 1.602 × 10⁻¹⁹ C), concentration in the dark (n₀), and free electron density of the TiO₂ photoelectrode (n) [33,34,35]. Equation (8) indicates that the V_oc is affected by n, which is closely related to the back electron transfer (BET) reaction that occurs between the photoinjected electrons and the ions in the electrolyte. As the BET reaction decreases the n value, suppression of the BET reaction is necessary to increase the V_oc. The n value can be estimated by measuring the lifetime of the electrons photoinjected into the TiO₂, where a longer electron lifetime can increase the n value and, therefore, the V_oc. Figure 3b shows Bode phase plots of the EIS spectra of the bare and Ti₃C₂T_x-Co-_bDSCs. Using the peak frequencies (f_max) of 38.2 Hz and 45.5 Hz obtained from the EIS Bode phase plots of the bare and Ti₃C₂T_x-Co-_bDSCs, respectively, the electron lifetime (τ_n) was estimated using Equation (9) [29,36]. The calculated electron lifetimes were 4.16 ms and 3.50 ms for the bare and Ti₃C₂T_x-Co-_bDSCs, respectively. A shortened lifetime on the part of the electrons injected from the photoexcited dyes was observed for the Ti₃C₂T_x-Co-_bDSC relative to that of the control cell (bare Co-_bDSC), indicating that the BET reaction (chemical reaction [10]) in the Ti₃C₂T_x-Co-_bDSC occurred faster than that in the bare Co-_bDSC.

$$V_{\mathrm{O}\mathrm{C}}=\frac{E_{Fn}-E_{F0}}{e}=\frac{k_{B}T}{e}\ln (\frac{n}{n_0})$$

(8)

$${\mathsf{\tau }}_n=\frac{1}{2\pi f_{max}}$$

(9)

Co³⁺(FK209) + e⁻ (TiO₂) → Co²⁺(FK209)

(10)

To further confirm that the BET reaction was faster in the Ti₃C₂T_x-Co-_bDSC, Nyquist plots of the EIS spectra measured at −0.7 V in the dark were obtained, as shown in Figure 5a. When the EIS measurement is performed in the dark, electrons are injected from the FTO into the TiO₂ under external applied voltage and then transferred to the electrolyte. Thus, the R₂′ value of the Nyquist plot measured in the dark corresponds to the resistance of the BET reaction between the Co³⁺(FK209) in the electrolyte and the electrons injected into the TiO₂ conduction band (reaction [10]). It was observed that the arc of the impedance component R₂′ in the Ti₃C₂T_x-Co-_bDSC was smaller than that in the bare Co-_bDSC. The smaller semicircle in the R₂′ suggests that the BET reaction at the TiO₂/dye/electrolyte interface was faster [34,35]. Moreover, from the peak frequencies (f_max) given in Figure 5b and Equation (9), the lifetime of the electrons injected from the FTO was calculated to be 11.9 ms and 8.4 ms for the bare and Ti₃C₂T_x-Co-_bDSCs, respectively. The same tendency in terms of a shortened electron lifetime was observed in the measurements under illumination (Figure 3b) and dark conditions (Figure 5b). Overall, a faster BET reaction resulted in a lower n value, leading to a decrease in the V_oc of the Ti₃C₂T_x-Co-_bDSC based on Equation (8). Furthermore, the reduced V_oc value in the Ti₃C₂T_x-Co-_bDSC can be explained by the positive shift in the TiO₂’s CBE, as discussed above (Figure 4). The lower n value can cause the positioning of the TiO₂’s CBE to shift in a positive direction, lowering the V_oc value due to the narrower potential gap between the TiO₂’s CBE and the redox potential of the electrolyte [39,40]. It is considered that this decrease in the V_oc (or n) value originated from the adsorption of the Ti₃C₂T_x MXene on the TiO₂ surface. More specifically, the electronically conductive Ti₃C₂T_x MXene particles could provide pathways for chemical reaction (10)—that is, the BET reaction between the photoinjected electrons and the Co³⁺(FK209) in the redox mediator.

As a reference, the shorter electron lifetime observed in the Ti₃C₂T_x-Co-_bDSC may decrease the electron collection on the FTO and, therefore, the J_sc value. In this study, it is believed that the enhanced hole collection, dye regeneration, and electron injection efficiencies surpassed the decreased electron collection efficiency.

2.4. Long Long-Term Stability of the Bare- and Ti₃C₂T_x-Co-_bDSCs

We compared the long-term stability of the bare- and Ti₃C₂T_x-Co-_bDSCs by evaluating their photovoltaic properties over time. Here, the fabricated devices were additionally sealed using hot-melt glue sticks to minimize the electrolyte leakage. Prior to the measurement of the photovoltaic performance, the I-DSCs with or without Ti₃C₂T_x MXene were converted into Co-DSCs through exposing them to the AM 1.5 condition for 150 min. Figure 6 compares the time-dependent performance variations of the cells stored at room temperature in the dark. Similar decay curves were noted in both devices, indicating that the incorporation of Ti₃C₂T_x MXene into the redox mediator did not affect the devices’ stability.

3. Materials and Methods

3.1. Materials

To fabricate DSCs, the same materials as those used in our previous report were utilized [23]. Detailed their information is provided in the ESI. The acetonitrile solvent used to prepare the liquid electrolytes was procured from Daejung Chemicals and Metals Co., Ltd. (Gyeonggi-do, Korea). Colloidal suspension of single-layer Ti₃C₂ in acetonitrile (2 mg Ti₃C₂/mL) (BK2020082105-08) was purchased from Beijing Beike New Material Technology Co., Ltd. (Jiangsu, China). All the chemicals used for DSC fabrication were exploited without further purification. The single-layer Ti₃C₂Tx MXene structure is illustrated in Figure S2 in the ESI. The chemical structures of the main components (FK209 and MPII) and additives (LiTFSI and TBP) of the electrolyte are shown in Figure S3 in the ESI.

3.2. Preparation of Co³⁺/I⁻-Based Liquid Electrolytes with or without Ti₃C₂T_x

The Ti₃C₂T_x-dispersed liquid electrolyte based on a Co³⁺/I⁻ redox mediator was prepared by dissolving MPII (0.6 M, 302.5 mg), FK209 (0.015 M, 15.2 mg), TBP (0.11 M, 32.4 mg), and LiTFSI (0.04 M, 22.8 mg) in the single-layer Ti₃C₂Tx colloid (2 ml). For the purpose of comparison, Co³⁺/I⁻-based liquid electrolytes without Ti₃C₂Tx were also prepared by simply replacing the Ti₃C₂Tx colloid with acetonitrile solvent (2 ml).

3.3. Fabrication of DSCs

Similar procedures to those described in our previous work were utilized to prepared the working (glass/FTO/TiO₂:dye) and counter (glass/FTO/Pt) electrodes for the DSCs [23]. A 60-μm-thick hot-melt adhesive was placed between the working and counter electrodes and then annealed for 10 min at 120 °C to seal the two electrodes. The Co³⁺/I⁻-based liquid electrolytes with or without Ti₃C₂T_x MXene were injected into the cells through one of the two small holes predrilled into the counter electrodes. By sealing the two holes, we were able to fabricate DSCs with a 25 mm² active area. Detailed procedures are mentioned in the ESI.

3.4. Measurements

Photovoltaic performance measurements, EIS analyses and dark current studies were carried out. All the measurements were performed under ambient conditions at room temperature. Detailed information for the measuring instruments is presented in the ESI.

4. Conclusions

The photovoltaic properties of DSCs with or without Ti₃C₂Tx MXene in Co³⁺/I⁻-based redox mediators were investigated in this study. Through the incorporation of Ti₃C₂T_x MXene into the Co³⁺/I⁻ liquid redox mediators, the average J_sc value of the I-DSCs, in which hole conduction occurred via the redox reaction of the I⁻ and I₃⁻ ions, was enhanced, whereas the average V_oc value was decreased when compared with that of the device without the MXene. As a result, the average PCE of the I-DSCs with Ti₃C₂T_x was very similar to that of the cells without Ti₃C₂T_x. To obtain Co-DSCs based on the Co²⁺/Co³⁺ redox couple, the I-DSCs were exposed to light for 150 min. Through the addition of Ti₃C₂Tx MXene into the Co³⁺/I⁻-based redox mediators, the hole collection, dye regeneration, and electron injection efficiencies of the Ti₃C₂Tx-Co-DSCs were all increased, leading to an improvement in both the J_sc and PCE when compared with those of the bare Co-DSCs without MXene. These results indicate that Ti₃C₂Tx MXene, as a J_sc-improver, is a good additive for improving the PCE of Co-DSCs.