Large area growth of MoTe2 films as high performance counter electrodes for dye-sensitized solar cells

A cost effective and efficient alternative counter electrode (CE) to replace commercially existing platinum (Pt)-based CEs for dye-sensitized solar cells (DSSCs) is necessary to make DSSCs competitive. Herein, we report the large-area growth of molybdenum telluride (MoTe2) thin films by sputtering-chemical vapor deposition (CVD) on conductive glass substrates for Pt-free CEs of DSSCs. Cyclic voltammetry (CV), Tafel curve analysis, and electrochemical impedance spectroscopy (EIS) results showed that the as-synthesized MoTe2 exhibited good electrocatalytic properties and a low charge transfer resistance at the electrolyte-electrode interface. The optimized MoTe2 CE revealed a high power conversion efficiency of 7.25% under a simulated solar illumination of 100 mW cm−2 (AM 1.5), which was comparable to the 8.15% observed for a DSSC with a Pt CE. The low cost and good electrocatalytic properties of MoTe2 thin films make them as an alternative CE for DSSCs.

in semiconducting (2 H) and metallic (1 T′) phase in nature 22,23 . The hydrogen evolution reaction catalytic activity of MoTe 2 was reported 24 . In this work, we report the catalytic activities of MoTe 2 as counter electrode in DSSCs.
Herein, we have grown MoTe 2 thin films via sputtering combined with a post-deposition annealing process on conductive glass substrates with different thickness. This work is a continuation of our research focusing on TMDC material search and growth for DSSC applications. MoTe 2 films used as CEs in DSSC showed good electrical conductivity and electrocatalytic activity, and a DSSC employing a MoTe 2 CE synthesized under optimized conditions had a 7.25% PCE, which is comparable to the value of 8.15% obtained for the Pt CE under the same conditions. To the best of our knowledge, this is the highest PCE for (I − /I 3 − ) redox couple-based DSSCs employing MoTe 2 CE under a simulated solar illumination of 100 mW•cm −2 (AM 1.5).

Results and Discussion
In this study, we fabricated large-area and high-quality MoTe 2 directly onto FTO substrate by sputtering-CVD growth, as depicted in Fig. 1. Our synthesis method consists of two steps. Initially, Mo films were deposited onto FTO substrates using magnetron sputtering, and the film was annealed at 500 °C in a tellurium environment in a CVD chamber. Three samples were sputtered at three different times (20,30, and 40 min) and subsequently tellurized, and referred to S1 (~185 nm), S2 (~335 nm), and S3 (~668 nm), respectively.
Field emission scanning electron microscopy (FE-SEM) analysis was performed to reveal the surface morphology of the MoTe 2 /FTO structure. Figure 2(a-c) provide FE-SEM images of samples S1, S2 and S3, respectively. Samples exposed to the longest tellurization (40 min) exhibits the biggest grains in Fig. 2(c). Cross-sectional SEM images show that the thicknesses of the S1, S2 and S3 are ~185, ~355 and ~688 nm, respectively ( Fig. 2d-f). The low magnification FE-SEM image with EDS spectrum for the sample S2 is provided in supporting information ( Figure S1a,b). The cross-sectional view with their EDS profile is provided to confirm the presence of Mo and Te in the MoTe 2 film (Figure S1c,d).
The structures of the MoTe 2 films were characterized by Raman spectroscopy using a 514 nm excitation laser. Figure 3(a) shows prominent peaks at ~161, and ~267 cm −1 , which correspond to the A g mode. A shoulder peak was observed at ~189 cm −1 , and this was ascribed to the B g mode, for MoTe 2 in the 1 T′ phase. The spectrum agrees well with the previously reported results 22,25 . XRD measurements were performed to further evaluate the identity and structure of the film, as shown in Fig. 3b. The XRD patterns show that the synthesized MoTe 2 films were polycrystalline in nature with a monoclinic structure. The diffraction peaks were at 38.0°, 42.7°, 51.7°, 54.7°, 61.7°, 64.7°, 65.9°, 71.2°, and 78.8°, which correspond to (210), (106), (311), (022), (221), (411), (125), (219), and (504) lattice planes of MoTe 2 , respectively (JCPDS No. 71-2157). No impurities or other reflections from deleterious crystalline phases were observed, which suggests that well oriented MoTe 2 films were deposited.
X-ray photoemission spectroscopy (XPS) was used to verify the surface chemical compositions and valence states of 1 T′-MoTe 2 . The survey spectrum indicates the coexistence of Mo and Te elements in the MoTe 2 films ( Figure S2). High-resolution spectra of each element are also given in Fig. 4a,b. As shown in Fig. 4a, the Mo 3d spectrum exhibits two main peaks at 229.2 and 232.2 eV, corresponding to the doublet of Mo 3d 5/2 and Mo 3d 3/2 . For Te 2d specturm, peaks were observed at 573.1 and 583.6 eV, as shown in Fig. 4b. These can be assigned to the spin-orbit couple of Te 2d 5/2 and Te 2d 3/2 , respectively 18 . The stoichiometry of Mo and Te elements in our synthesized MoTe 2 film is confirmed by EDS spectrum (Fig. S1b). Hall measurements were performed on MoTe 2 /glass at room temperature (RT) with an active area of (1 × 1) cm 2 ( Figure S3). MoTe 2 CE revealed p-type behavior similar to that reported in the literature 26 . The conductivity = 3.3 × 10 −1 Ω −1 cm −1 , and charge mobility = 95 cm 2 V −1 s −1 were extracted from the device.
To investigate the application of the MoTe 2 as a CE in DSSCs, cyclic voltammetry (CV) studies were performed to estimate the reaction kinetics and electrocatalytic performance. CV was conducted using a three-electrode system in an acetonitrile solution consisting of 10 mM LiI, 1 mM I 2 , and 0.1 mM LiClO 4 at a scan rate of 20 mVs −1 . Figure 5a shows the CVs of the system for Pt and MoTe 2 (S1, S2, S3) in the potential interval between −0.2 to 1 V vs. Ag/AgCl. The Ox 1 and Red 1 peaks at low potential were attributed to the redox reaction of I 3 − + 2e − ↔ 3I − . The Red 1 peak corresponding to I 3 − + 2e − ↔ 3I − was used to evaluate the integral electrocatalytic ability of CEs to reduce triiodide ions to iodide ions. This reduction occurs in DSSCs, and the current density of this reaction is mainly determined by the number of reduction-active sites on the surface area of the electrocatalyst and the intrinsic electrocatalytic ability of each site. Ox 1 and Red 1 represent the same electrochemical reaction I 3 − + 2e − ↔ 3I − , in which Ox 1 indicates the left direction and Red 1 indicates the right direction. CV curves show that, like Pt, S1 and S2 also are catalytically active for the reaction that regenerates the redox couple. The higher cathodic peak current density can be used to evaluate the catalytic activity of the CE, and comparable peak current densities imply good electrocatalytic activity. The ~335 nm-thick (S2) CE showed higher current density than the ~185 nm CE (S1), suggesting faster reduction of triiodide ions in the S2 CE compared to the S1 CE (Fig. 4a). The higher cathode current density could be attributed to its relatively higher surface roughness compared to the much smoother S1. Furthermore, S1 and S2 samples displayed similar anodic and cathodic peaks to Pt CE, suggesting that they are active in catalyzing the reduction of I 3 − to I − . The peak current and peak to peak separation is important parameters for determining the catalytic activity of CE. The rate constant of a redox reaction is inversely proportional to its peak separation (Epp) [27][28][29] . Epp is calculated using the formula In DSSCs, the CE has more influence on the negative peak. So, we used this peak for Epp calculations. The Epp for the Pt CE was 295 mV, while those for S1, S2 and S3 were ~354, ~459 and ~308 mV, respectively.
To investigate the electrochemical stability of MoTe 2 S2 sample and Pt CE, CVs were recorded for 50 consecutive cycles with a potential range from −0.2 to 1 V vs. Ag/AgCl, as presented in Figure S4a. After 50 consecutive and (d-f) cross-sectional images of S1, S2 and S3. The observed thickness were ~185 nm, ~335 nm and ~ 668 nm for S1, S2 and S3, respectively. scans, the CV shape of sample S2 almost overlapped, and the redox peak current (cathodic and anodic peak current density) for sample S2 was almost constant, which suggests that the MoTe 2 CE possesses reversible redox activity, good electrochemical and chemical stability, and strong adhesion on the FTO glass substrate. The CVs of sample S1, S2 and S3 were measured using different scan rates from 10 to 100 mVs −1 for the (I − /I 3 − ) redox reaction, as shown in Figure S4b-d, respectively. There are a linear increment in the current peak value with increasing scan rate, indicating that the inner sites of MoTe 2 also become reactive and possess catalyst activity at higher scan rate 29,30 .
To further evaluate the charge transfer kinetics and internal resistance of DSSCs, EIS measurements were performed using symmetric cells fabricated with two identical electrodes (CE/electrolyte/CE). The equivalent circuit model used for fitting the resultant impedance data is illustrated in Fig. 5b. In each curve, there are two well-defined semicircles. The first semicircle at high frequency is related to impedance of charge transfer process occuring at CE/electrolyte and lower frequency range can be assigned to the Nernst diffusion impedance (Z w ) within electrolyte. The extracted charge-transfer resistance (R ct ) values of the Pt, S1, S2 and S3 CEs are 0.93, 27.01, 25.97, and 37.44 Ω cm 2 , respectively. The sample S3 has the largest R ct and S2 has the lowest one among the MoTe 2 samples. R s values of S1, S2, S3 and Pt are 16.47, 13.79, 13.83, and 7.05 Ω cm 2 , respectively. The sample S2 has the lowest R s . The R s value of S3 is largest probably due to the largest film thickness.
Tafel polarization analyses were also performed using symmetric cells at a scan rate of 50 mVs −1 for Pt, S1, S2 and S3 samples (Fig. 5d). The Tafel curve is usually divided into three regions. The lower potential zone is called the polarization zone, and the middle region (with a sharp slope) is the Tafel zone, which determines the catalytic activity of the electrode. The last zone is the diffusion zone, which determines the diffusion of ions in the electrode. The tangent slope in the anodic or cathodic branch provides information about the exchange current density (J 0 ) on the electrode 31 . The comparison indicates that S2 (S2 > S1 > S3) is more effective than S1 and S3 at catalyzing the reduction of I 3 − . The exchange current density, J 0 is inversely proportional to R ct from the equation where R is a gas constant, T is an absolute temperature, n is the number of electrons involved in the reaction, and F is Faraday's constant 27,32 . A higher J 0 for Pt and S2 CE implies a lower value of R ct in the impedance measurement. The schematic of DSSCs with MoTe CE is illustrated in Fig. 6a. The photocurrent density versus photovoltage (J-V) curves of the DSSCs are shown in Fig. 6b. The photovoltaic paramters including the short circuit current density (J sc ), open circuit voltage (V oc ), fill factor (FF) and PCE (η) of DSSCs with Pt and MoTe 2 (S1, S2 and S3) CEs under a simulated solar illumination of 100 mWcm −2 (AM 1.5) are summarized in the Table 1. The sample S2 CE exhibits the best performnce. The DSSC with S1 CE has lower FF than that with S2 CE, which is related to red-ox behaviour as discussed in earlier. The J sc and FF values are increased for the S2 CE, which leads to enhancing PCE from 6.38% to 7.25%. And, the low efficency of S3 CE is mainly due to low V oc and FF. This could be attributed to higher R ct value as confirmed by EIS analysis. The J sc values are decreased in the order of S3 > S2 > S1 > Pt, and PCE values are decreased in the order of Pt > S2 > S3 > S1. It is believed that thick film (S3) could affect electrolyte penetration and result in weaker adhesion to the FTO substrate 33 .
The observed PCE (7.25%) value of the S2 CE was much higher than those of earlier reports based on TMDCs, which include WS 2 films prepared by a doctor-blading technique (4.56%) 34 , multi-walled carbon nanotubes-MWCNTs@MoS 2 (6.45%) 35 , multi-wall carbon nanotubes decorated with tungsten sulfide-MWCNTs@WS 2 (6.41%) 36 , composite films of molybdenum disulfide (MoS 2 )/graphene flakes (5.98%) 29 , and molybdenum disulfide and graphene-MoS 2 /RGO (6.04%) 37 . The variations of V oc and J sc values for MoTe 2 and Pt CEs can be attributed to the nanoporous nature of the MoTe 2 CE in contrast to the planar Pt CE, and the high conductivity of Pt. Figures S5 shows incident photon-to current-conversion efficiency (IPCE) curves of DSSCs with the MoTe 2 CE and Pt CE. These results indicate that catalytic activities depend on the MoTe 2 thickness since active sites and morphology vary with the growth time, supporting that catalytic activities of thin MoTe 2 could be modulated by their film thickness and morphology.

Conclusions
In summary, we presented the sputtering-CVD post annealing route for synthesizing MoTe 2 as counter electrodes for DSSCs. Detailed electrochemical investigations were carried out using cyclic voltammetry, electrochemical impedance spectroscopy, and Tafel curve analysis to determine the suitability for CE for DSSCs. CV performance revealed that MoTe 2 CEs possess good electrocatalytic activity and fast reaction kinetics for the reduction of triiodide to iodide. It was found that catalytic activities of thin MoTe 2 could be modulated by their film thickness and morphology.
The optimum MoTe 2 CE in a fabricated DSSC exhibited a 7.25% PCE, which is comparable to the 8.15% Pt CE under the same illumination conditions. The presented work suggests that MoTe 2 would be a promising counter electrode material as a low-cost and highly efficient alternative to Pt in DSSCs.
Experimental Section and Device preparation. FTO/glass substrates were cleaned with a standard piranha solution and deionized water and were then baked at 120 °C for 5 min. After loading the FTO substrates in a sputter chamber, the chamber was evacuated by a rotary pump and a turbomolecular pump combination to a pressure of ~1 × 10 −7 torr. Next, Mo thin films were deposited onto FTO/glass substrates using a Mo target (99.99%) by magnetron sputtering. During the film deposition, the Ar gas flow ratio was maintained at 10 sccm, and the power was fixed at 100 W. Mo films were deposited at different sputtering times (such as 20, 30, and 40 min) at room temperature, and these are denoted as S1 (185 nm), S2 (335 nm), and S3 (668 nm) samples, respectively. After removing the samples from the sputter chamber, the as-sputtered films were placed downstream of the chemical vapor deposition (CVD) chamber and heated. The as-sputtered Mo films were annealed in tellurium vapor at 500 °C for 30 min to form MoTe 2 films and to improve the crystalline quality of the films. A pure tellurium powder (99.99%) was placed upstream of the CVD chamber, and a heating filament for the tellurium boat was fixed at 350 °C. The tellurium powder was evaporated at 350 °C using a mixture of argon and hydrogen (60 sccm -Ar and 30 sccm -H 2 ) carrier gases, and the pressure of the CVD chamber was kept at 2 × 10 −2 Torr.

Fabrication of DSSCs.
DSSCs were fabricated to evaluate the CE performance of the MoTe 2 films using our method [38][39][40][41] . Briefly, thin blocking layer TiO 2 was deposited onto a cleaned FTO glass substrate (15 × 15 mm 2 ) by dipping it in 40 mM TiCl 4 solution for 30 min at 70 °C and annealing it at 450 °C for 30 min. A homemade titanium dioxide (TiO 2 ) powder paste of P25 was coated on the cleaned FTO glass as the main layer (~12 µm thickness) using a simple doctor blade coating technique. The TiO 2 -coated FTO was then sintered in five steps of 70, 325, 375, 450, and 500 °C for 30, 5, 5, 15, and 15 min, respectively, in a high temperature furnace (Lab House Co.). Additionally, a scattering layer (~6 µm) was coated over the main layer and sintered using the same sintering steps. The TiO 2 film was then sensitized with 0.5 mM N 719 prepared in an absolute ethanol: acetonitrile (1:1) solution for 24 h. The polymer electrolyte, which was composed of 0.5 M LiI, 0.6 M 1-propyl-2, 3 − dimethylimidazolium iodide, 0.05 M I 2 , 0.5 M 4-tert-buylpyridine, and 3% w/w polyethylene oxide (Mw 250,000) with acetonitrile as the solvent was then injected between the two electrodes. The Pt-coated CE was prepared by spreading a drop of 2 mM chloroplatinic acid hexahydrate (H 2 PtCl 6 ) in isopropanol onto the FTO substrates using a simple brush method and heating it to 400 °C for 15 min in ambient air 42,43 . The dye-sensitized TiO 2 photoanode with an active area of 0.25 cm 2 and the as-fabricated CE were assembled using a 50-μm-thick spacer made of polyimide adhesive tape.   Table 1. Photovoltaic and EIS parameters of Pt, S1, S2 and S3 based DSSC CEs.