ZnS/SiO₂ Passivation Layer for High-Performance of TiO₂/CuInS₂ Quantum Dot Sensitized Solar Cells

Abstract

Suppressing the charge recombination at the interface of photoanode/electrolyte is the crucial way to improve the quantum dot sensitized solar cells (QDSSCs) performance. In this scenario, ZnS/SiO₂ blocking layer was deposited on TiO₂/CuInS₂ QDs to inhibit the charge recombination at photoanode/electrolyte interface. As a result, the TiO₂/CuInS₂/ZnS/SiO₂ based QDSSCs delivers a power conversion efficiency (η) value of 4.63%, which is much higher than the TiO₂/CuInS₂ (2.15%) and TiO₂/CuInS₂/ZnS (3.23%) based QDSSCs. Impedance spectroscopy and open circuit voltage decay analyses indicate that ZnS/SiO₂ passivation layer on TiO₂/CuInS₂ suppress the charge recombination at the interface of photoanode/electrolyte and enhance the electron lifetime.

1. Introduction

Semiconductor quantum dots (QDs) based on II-VI group such as CdSe [1], CdTe [2], CdS [3], PbS [4], PbSe [5], and etc. have been extensively studied for QD sensitized solar cell (QDSSC) and photocatalysis applications, due to tunable band gap, hot electron injection, higher absorption coefficients, and multiple excition generation (MEG) [6,7]. However, highly toxic Cd or Pb-containing QDs based solar cells show the excellent photostability and high power conversion efficiencies (PCEs). However, high toxicity of Cd or Pb still limit the commercial applications in consideration of environmental and health concerns. Therefore, the development of “green” QDs without carcinogenic heavy metal element is crucial for the practical utilizations of QDSSCs.

Less-toxic I−III−VI₂ group QDs, specifically CuInS₂ (CIS) QD has been attracted as “green” QDs due to high absorption coefficient (∼10⁵ cm⁻¹) and optimal band gap energy (1.0–1.5 eV), both of which make it a promising candidate as a sensitizer in QDSSCs [8,9,10]. There are two common approaches have been demonstrated for assembling CuInS₂ QDs onto TiO₂ electrodes: by direct adsorption or bifunctional-linker-assisted adsorption and by successive ionic layer adsorption [11,12]. Owing to its facile and reproducible preparation, the SILAR process gained much attention for depositing QDs onto TiO₂ surface with high QD loading and well controllable in size of QDs [13,14].

Chang et al. developed the TiO₂/Cu₂S/CuInS₂/ZnS photoanode using the SILAR process and achieved a PCE of 2.52% [15]. Zhou et al. prepared QDSSCs based on CuInS₂ and introduction of In₂S₃ buffer layer suing SILAR process, which presented as high as ~1.06% PCEs [16]. Meng et al. developed a CuInS₂ QDs on reduced graphene oxide sheets using facile one-pot solvothermal approach and delivered a PCE of 1.5% [17]. Han et al. fabricated PbS/CuInS₂/TiO₂ using SILAR process and obtained a PCE of 4.11% [18]. However, the lower performance of QDSSC is due to severe charge recombination process at the TiO₂/QD/electrolyte interface. To reduce the charge recombination in QDSSCs, thin wide band gap inorganic barrier layer was deposited over TiO₂ electrode, which acts as an energy barrier hindering electrons from recombining. Until now, ZnS is a promising passivation layer for suppressing the interfacial recombination in QDSSCs [19]. The ZnS over layer is introduced by the facile SILAR process, which covers the TiO₂ layer and the surface of QDs toward the electrolyte. Therefore, the deposition of ZnS layer is useful technique to improve the solar cells performance due to the passivation of the QD surface states, yielding in suppression of the recombination processes [20].

Herein, ZnS/SiO₂ double barrier coating was sequentially deposited on CuInS₂ QD to suppress the recombination in QDSSCs. TiO₂/CuInS₂/ZnS/SiO₂ structure favors the improvement of photovoltaic properties of the QDSSCs. The QDSSC based on the CuInS₂ QD sensitizer and ZnS/SiO₂ double layer exhibits a PCE of 4.63% (with short circuit current density (J_SC) = 12.83 mA cm⁻², an open circuit voltage (V_OC) = 0.603 V, fill factor (FF) = 0.598) under AM 1.5 G one full sun illumination, which is much higher than the CuInS₂/ZnS (PCE = 3.23%) and CuInS₂ (PCE = 2.15%).

2. Results and Discussion

2.1. Morphological Characterization

Figure 1(a,a1,b,b1,c,c1) show the scanning electron microscopy (SEM) images of the CuInS₂, CuInS₂/ZnS, and CuInS₂/ZnS/SiO₂ layers on the surface of TiO₂. The bare CuInS₂ film in Figure 1a,a1 exhibit uniformly distributed nanoparticles on the TiO₂ surface. All the samples exhibit the almost similar surface morphology and there is a slight increase in the particle size of with the deposition of ZnS and ZnS/SiO₂ layers; however, the specific change in size of particles is difficult to examine and was inconclusive. Therefore, elemental mapping from SEM analysis and X-ray photoelectron spectroscopy (XPS) measurements were conducted to identify the ZnS and ZnS/SiO₂ coatings on the surface of TiO₂. Moreover, the compositional distributions of a TiO₂/CuInS₂/ZnS/SiO₂ sample are further demonstrated by elemental mapping behavior, in which the homogeneous distribution and coexistence of Cu, In, Zn, S, Si and O elements are clearly observed in TiO₂/CuInS₂/ZnS/SiO₂ sample (Figure 2). Such elemental mapping results unambiguously indicate that the CuInS₂, ZnS and SiO₂ were deposited successfully on the TiO₂ surface.

The composition of the CuInS₂/ZnS/SiO₂ sample was investigated by XPS, as depicted in Figure 3. The XPS survey spectra in Figure 3a depict peaks for Ti_2p, O_1s, C_1s, Cu_2p, In_3d, Zn_2p, S_2p, and Si_2p, respectively. The binding energy of Cu 2p_3/2 and Cu 2p_1/2 were observed at 932.9 and 952.7 eV, respectively (Figure 3b), with no evident shake-up satellite signals in this Cu2p spectrum. The two major peaks of In are observed at 445.4 eV and 452.9 eV for In 3d_5/2 and In 3d_3/2, respectively (Figure 3c). Figure 3d depicts the binding energies for Zn 2p_3/2 and Zn 2p_1/2 of the prepared sample of CuInS₂/ZnS/SiO₂ at 1023.2 eV and 1046.2 eV respectively. The S 2p spectrum was yielded peaks of S 2p_3/2 and S 2p_1/2 at 162.0 eV and 163.0 eV binding energies, respectively (Figure 3e). In the Si 2p spectrum (Figure 3f), the main peak observed at 103.2 eV has been ascribed to Si in the oxidized form (SiO₂) and the other shoulder peak can be assigned to the presence of crystalline Si (elemental Si).

2.2. Electrochemical Characterization

The J-V curves of the QDSSCs based on various photoanodes were obtained under AM 1.5 illumination (100 mW cm⁻²) are displayed in Figure 4 and the corresponding photovoltaic parameters are tabulated in

Table 1. Photovoltaic properties and EIS results of QDSSCs fabricated various sensitized conditions.

Cell	Voc (V)	Jsc (mA cm−2)	FF	η%	RS (Ω)	RCE (Ω)	Rct (Ω)
CuInS2	0.509	7.87	0.537	2.15	9.34	0.84	30.65
CuInS2/ZnS	0.569	9.95	0.571	3.23	10.03	1.03	23.95
CuInS2/ZnS/SiO2	0.603	12.83	0.598	4.63	10.24	1.16	55.02

. When only CuInS₂ are deposited on TiO₂ film, the QDSSC exhibits a J_SC of 7.87 mA cm⁻², V_OC of 0.509 V, FF of 0.537, resulting a low PCE of 2.15%. However, when ZnS and ZnS/SiO₂ passivation layers were deposited, all the photovoltaic parameters were greatly improved; the QDSSCs with a ZnS/SiO₂ layer exhibit the good performance, with J_SC, V_OC, and FF reaching 12.83 mA cm⁻², 0.603 V, and 0.598, respectively and the highest PCE of 4.63%, which is much higher than the PCE of 3.23% with a ZnS passivation layer. It is observed that the ZnO/SiO₂ layer exhibiting higher performance than the bare and ZnS layers in QDSSCs, which is due to suppression of recombination losses in QDSSCs and increases the charge collection efficiency.

Electrochemical impedance spectroscopy (EIS) characterizations were conducted to identify the charge recombination processes in devices under forward bias (V_OC) and dark condition. Figure 5 depicts the EIS spectra of various photoelectrodes and the Nyquist plots were fitted using Z-view software with the equivalent circuit provided in the inset of Figure 5. The corresponding fitting results are shown in Table 1. The Nyquist plot consists of two semicircles and the first semicircle represents the resistance (R_CE) at the CE/electrolyte interface. The second semicircle denotes the charge transfer resistance (R_ct) at the interface of the TiO₂/QDs/electrolyte. At higher frequency, the intercept on the real axis corresponds to the series resistance (R_s) of FTO substrate and the resistance of FTO/TiO₂ [21,22]. It is noticed that there are no apparent differences observed in the R_S and R_CE due to the same CE and electrolyte used in these experiments. However, there is a noticeable difference in R_ct; the R_ct value for the CuInS₂/ZnS/SiO₂ based QDSSCs is 55.02 Ω, while R_ct value for the CuInS₂/ZnS and CuInS₂ based QDSSCs are only 30.65 Ω and 23.95 Ω, respectively. The charge recombination resistance at the TiO₂/QDs/electrolyte interface is mainly observed by R_ct. The higher R_ct value represents the suppressed recombination of the electrons and holes, and enhances the electron transfer process at the interface of TiO₂/QDs/electrolyte. Therefore, it is confirmed that the deposition of ZnS/SiO₂ layer on the CuInS₂ QDs favors the efficient electron transfer from CuinS₂ to TiO₂ photoanodes with suppression of the interfacial charge recombination processes, which is more effectively than that of the ZnS passivation layer.

Furthermore, Open circuit voltage decay (OCVD) measurements were carried out to study the charge recombination process in QDSSCs and the results. OCVD analysis of QDSSCs was performed during relaxation form an illuminated quasiequilibrium state to the darkness. Figure 6 depicts the OCVD plots of the QDSSCs based on CuInS₂, CuInS₂/ZnS and CuInS₂/ZnS/SiO₂ photoanodes. Apparently, the CuInS₂/ZnS/SiO₂ cell delivered considerably longer decay times than the CuInS₂ and CuInS₂/ZnS cells, indicating a suppression of charge recombination process. Moreover, the V_OC decay and electron life time are directly related according to following equation [23]:

$${\mathsf{\tau }}_{\mathrm{e}}=-\left(\frac{k_{B}T}{\mathrm{e}}\right){\left(\frac{d{\mathrm{V}}_{\mathrm{OC}}}{dt}\right)}^{-1}$$

(1)

where k_B is the Boltzmann constant, T is the absolute temperature, and e is the electronic charge. It can be noticed that the τ_e of all the devices increases with decreasing V_OC. Among the QDSSCs investigated, the CuInS₂/ZnS/SiO₂ delivers longer τ_e than the CuInS₂/ZnS and CuInS₂ devices, implying suppressed recombination and efficient electron transfer at the TiO₂/QDs/electrolyte, which is consistent with EIS analysis. Therefore, the slower V_OC decay and longer τ_e of the CuInS₂/ZnS/SiO₂ device efficiently suppressed the electron recombination from TiO₂ and QDs to electrolyte and higher charge collection efficiency contribute to the increased photovoltaic performance.

Several paths for charge recombination occur at the TiO₂/QDs/electrolyte interface, which suppress the performance of QDSSCs. The deposition of ZnS/SiO₂ layer on the surface of TiO₂/CuInS₂ can effectively suppress the charge recombination process at the photoanode/electrolyte interface (Figure 7) and enhance the QDSSCs performance.

3. Materials and Methods

3.1. Preparation of TiO₂ Electrodes

TiO₂ paste (20 nm, Ti-Nanoxide HT/TP, Solaronix) was doctor bladed on fluorine-doped tin oxide (FTO, 1.3 × 1.6 cm²) substrate and heated at 450 °C for 30 min. The film thickness was about 7.5 µm with an active area of 0.25 cm² [24].

3.2. Deposition of CuInS₂ QDs on TiO₂ Electrodes

CuInS₂ QDs was deposited on the TiO₂ substrate using the SILAR process. TiO₂ electrodes were immersed into the three different solutions: one of 0.05 M of Cu(NO₃)₂ for 2 min, another of 0.05 M of In(NO₃)₃ for 2 min, and a final one of 0.1 M of Na₂S for 5 min. Following each immersion, the samples were rinsed with deionized (DI) water for 1 min to remove the excess precursors. This procedure comprises one CuInS₂ SILAR cycle and was repeated six times.

3.3. Deposition of ZnS, ZnS/SiO₂ Passivation Layers on TiO₂/CuInS₂ Electrodes

The ZnS passivation layer was deposited on TiO₂/CuInS₂ electrodes by a SILAR process. Typically, TiO₂/CuInS₂ electrodes were successively immersed into aqueous solutions containing 0.1 M of Zn(NO₃)₂ and 0.1 M Na₂S for 1 min, respectively. This process was repeated for three times and the electrode is named as CuInS₂/ZnS. Furthermore, SiO₂ coating was deposited by dipping the CuInS₂/ZnS electrodes in ethanol solution containing 0.01 M tetraethylorthosilicate and 0.1 M NH₄OH for 1 h. The as-fabricated electrode is termed as CuInS₂/ZnS/SiO₂.

3.4. QDSSC Fabrication

CuS CE on FTO substrate was fabricated according to the literature [25]. The photoanode and CuS CEs were combined using sealant (SX 1170-60, Solaronix) and the space between the electrodes was filled with polysulfide electrolyte (1 M Na₂S, 2 M S, and 0.2 M KCl in methanol and water at a ratio of 7:3).

3.5. Characterizations

The morphology of the samples were evaluated by SEM (S-2400, Hitachi). XPS measurement was investigated using VG Scientific ESCALAB 250. The J-V measurements were examined under AM 1.5 sunglight (100 mW cm⁻²) using an ABET Technologies (USA) solar simulator. EIS was investigated using a BioLogic SP-150 work station under one sun illumination over the frequency of 100 mHz–500 kHz.

4. Conclusions

Introduction of ZnS/SiO₂ passivation layer on TiO₂/CuInS₂ QDs has been demonstrated to be an effective and promising approach to significantly suppress the charge recombination at the interface of photoanode/electrolyte and enhance the power conversion efficiency. Interestingly, an overall η of 4.63% was obtained for the TiO₂/CuInS₂/ZnS/SiO₂ device, which is 43% enhancement over the η = 3.23% for the TiO₂/CuInS₂/ZnS and more than 115% increment over the η = 2.15% for the TiO₂/CuInS₂ device. Overall, ZnS/SiO₂ passivation layer is an effective approach to enhance the overall power conversion efficiency of QDSSCs.