Background

Solar cell devices based on organometallic halide perovskite materials have exhibited unprecedented performance over the brief span of 6 years, and organometallic halide perovskite solar cells (PSCs) show promise as affordable alternative solar cells with high power conversion efficiency (PCE) [1,2,3]. The huge interest in this new class of solar cells is due to their high absorption coefficient, ambipolar charge transport, small exciton binding energy, and long diffusion length [4,5,6]. Despite these excellent properties, PSCs possess several drawbacks. The most important of these are the sensitivity of perovskite materials to moisture, heat, and UV irradiation. To address these drawbacks, it has been found that adding formamidinium and/or an inorganic cation (Cs or Rb) to a methylammonium cation improves the stability against these environmental factors [3], and the durability of PSCs thus depends on both the device configuration (n-i-p, p-i-n) and the metal oxide semiconductors [7]. Generally, TiO2 materials are widely used in PSCs as electron transport layers (ETLs) in the n-i-p device configuration because of their large band gap and band alignment, and highly efficient PSCs are realized using TiO2 ETLs [8]. Although PSCs with TiO2 ETLs exhibit remarkable efficiency, the UV sensitivity and electronic properties of TiO2 have been suggested as targets for improvement to reduce the hysteresis and obtain durable PSCs [9]. Specifically, Heo et al. reported that Li doping can enhance the carrier mobility and conductivity of TiO2 and thus yield PSCs without significant hysteresis [10]. Ito et al. reported that when TiO2 in a PSC is exposed to UV irradiation, electrons are extracted at the TiO2/perovskite interface, degrading the perovskite material [11].

Stannic oxide (SnO2) has been widely studied for diverse applications such as batteries, gas sensors [12], solar cells [13], and catalysts. It is regarded as a promising candidate for use as a transparent conducting material and photoelectrode in photovoltaic devices. Considerable attention has been drawn recently to its application in PSCs as an alternative ETL with the goal of enhancing device performance and light stability, as it has a larger band gap (~3.6 eV at 300 K), higher electrical conductivity, and greater chemical stability than TiO2 semiconductors [2]. Various synthetic routes to SnO2, including sol–gel methods [14], molten-salt synthesis [15], microwave techniques [16], atomic layer deposition (ALD), and electrochemical deposition (ED) [17,18,19,20] have been developed. ALD and spin-coating solution processes are the dominant methods for fabricating SnO2 ETLs in PSCs [21,22,23]. The fabrication of ETLs in photovoltaic devices is paramount for limiting production costs because of the requirements for its production, such as thermal treatment, multiple processing steps, operation control, and scalable processing.

Here, we report on the synthesis and ETL application of SnO2 thin films on fluorine-doped tin oxide (FTO) by ED. Among the available methods, electrodeposition has the advantages of reduced production cost and large-scale manufacturing because it does not require a vacuum environment or complex operation control. Considering that perovskite materials are suitable for roll-to-roll manufacturing, the application of electrodeposition to obtain SnO2 ETLs will demonstrate not only a simple, cost-effective, and scalable strategy for alternative ETLs but also facilitate development of a continuous roll-to-roll process for industrial application of PSCs.

Methods

Preparation of SnO2 Film

A chronovoltammetry technique (VSP 200, Biologic) was used for ED of Sn nanospheres onto an FTO substrate using a standard three-electrode system in a deionized water solution (50 mL) containing 0.05 M SnCl2∙2H2O [tin chloride (Π), Sigma Aldrich] and 1 mL of nitric acid (HNO3, Samchun Chemical). The nanospheres were then thermally treated in air at 400 °C for 30 min to obtain SnO2. The aqueous solution was stirred for 1 h at 60 °C on a hot plate. After N2 purging for 10 min, the solution was used for electrodeposition. In the standard three-electrode system, FTO was used as the working electrode, and a platinum plate was used as the counter electrode. The reference electrode was a Ag/AgCl electrode (CHI111) in 1 M KCl solution.

Device Fabrication

The prepared SnO2 thin films on FTO (TEC 8) were used in the fabrication of PSCs. The perovskite layer was processed in two steps. A mixture of PbI2 (99.999%, Aldrich) and PbCl2 (99.999%, Aldrich) was dissolved in N, N-dimethylformamide and stirred at 60 °C. The molar ratio of the precursor solution (PbI2:PbCl2) was 1:1 (1 M). The PbI2/ PbCl2 solution was spin-coated on the SnO2-coated FTO at 5000 rpm for 30 s in a glove box and dried on a hotplate at 70 °C. To convert it to a perovskite material, 120 μL of methylammonium iodide solution (40 mg/mL) was loaded at 0 rpm for 35 s and then spin-coated at 3500 rpm for 20 s; the sample was then annealed isothermally at 105 °C for 75 min in the ambient environment. After annealing, the films were moved into the glove box in N2 atmosphere, and a hole-transporting material (HTM) was spin-coated on the MAPbI3-xClx/SnO2/FTO film at 3000 rpm for 30 s. Poly[bis(4-phenyl)(2,4,6-trimethylphenyl)amine] (EM Index) solution (20 mg/1 mL) was used as the HTM with 15 μL of Li-bis(trifluoromethanesulfonyl)imide)/acetonitrile (170 mg/1 mL) and 15 μL of tert-butylpyrridine. Finally, Au was deposited via thermal evaporation. TiCl4 hydrolysis treatment was applied by immersing the electrodeposited SnO2 films in a 40 mM TiCl4 solution at 70 °C for 30 min and drying them at 150 °C in air.

Characterization

Cyclic voltammetry (CV, scan rate 50 mV/s) measurements were made to confirm the electrochemical behavior of the SnCl2∙2H2O solution from −1.5 to 2 V. The crystalline structure of the samples was characterized by X-ray diffraction (XRD, Rigaku, Dmax 2200, Cu Kα) and X-ray photoelectron spectroscopy (XPS, ULVAC-PHI 5000, VersaProbe II). The morphologies of the samples were observed by field emission scanning electron microscopy (SEM, Hitachi S4800). The J–V curves of the PSCs were obtained using an electrochemical station (VSP200, Bio-Logic) under 100 mW/cm2 AM 1.5G light (Sun 3000 class AAA, ABET Technology) with a metal mask 0.098 cm2 in area. Devices were scanned at a 20 mV/s scan rate. CV measurements of the blocking layer effect were performed using a three-electrode setup after nitrogen purging for 10 min. The aqueous electrolyte contained 0.5 M KCl and the electron redox couple K4[Fe(II)(CN)6]/K3[Fe(III)(CN)6] at a concentration of 5 mM. A Ag/AgCl electrode was used for the reference electrode, and a Pt wire was used for the counter electrode; the scan rate was 50 mV/s. An Oriel-calibrated Si solar cell (SRC-1000-TC-KG5-N) was used to adjust the light intensity to one-sun illumination. The external quantum efficiency (EQE) was measured using an Ivium potentiostat and a monochromator (DongWoo Optron Co., Ltd.) under a light support (ABET 150 W xenon lamp, ABET Technology). EQE data were acquired in DC mode. Photoluminescence (PL) spectra were measured using a luminescence spectrometer (LS 55, PerkinElmer) with excitation at 530 nm. The intensity-modulated photocurrent and photovoltage were measured by an Ivium potentiostat with a Modulight LED (Ivium).

Results and Discussion

We performed CV measurements of the SnCl2∙2H2O solution to identify suitable potential values. Figure 1a shows the CV curve, which was scanned from 2.0 to −1.2 V. All potential values were recorded with respect to the reference electrode (Ag/AgCl). As shown in Fig. 1a, an increase in the cathodic current was observed from −0.5 to −1.2 V. Generally, when the voltage is swept in a CV experiment from positive to negative voltage, the current first increases because of an electrochemical reaction on the working electrode surface and then decreases owing to local depletion of the chemical species close to the working electrode.

Fig. 1
figure 1

(a) CV curve measured at a scan rate of 50 mV/s and (b) XRD patterns of electrodeposited SnO2

Full size image

On the basis of the CV result, we performed ED using a chronovoltammetry technique. Note that the phase of the deposits depends on the concentration ratio of [HNO3] to [Sn2+] because nitric acid acts as an oxygen source in the phase [24]. The presence of HNO3 (as identified in the XRD pattern, Fig. 1b) facilitated generation of a SnO2–Sn co-phase. This will be referred to as SnO2–Sn nanospheres to distinguish it from pure SnO2. Figure 2 shows SEM images of the SnO2–Sn nanospheres deposited on FTO substrates at different potential values (−0.5, −0.6, −0.7, −0.8, −0.9, and −1 V). We found that the applied voltage is a very important parameter in the electrodeposition process, as the morphologies of the deposits were dramatically different. For relatively low absolute potentials (−0.5 and −0.6 V), few SnO2–Sn nanospheres formed. On the other hand, the FTO was overlaid with Sn having irregular shapes at −0.9 and −1 V. Even though comparable SnO2–Sn nanosphere formation occurred at −0.7 and −0.8 V, the uniformity was better at −0.7 V. As a result of these observations, −0.7 V was chosen as a suitable potential for electrodeposition of SnO2–Sn nanospheres.

Fig. 2
figure 2

Top-view SEM images of SnO2 films electrodeposited at various applied voltages. (a) − 0.5 V, (b) −0.6 V, (c) −0.7 V, (d) −0.8 V, (e) −0.9 V, and (f) −1.0 V vs. Ag/AgCl. Scale bar is 1 μm

Full size image

A potential of −0.7 V was also used to optimize the deposition time in the range of 150 to 210 s. Figure 3 shows SEM images of samples obtained at various deposition times and the corresponding device performance. Fewer particles formed at 150 s than at 180 s. For a longer deposition time (210 s), aggregation of SnO2–Sn nanospheres was confirmed. To evaluate the photovoltaic performance of PSCs with the electrodeposited SnO2 films, the SnO2–Sn nanosphere films were thermally treated in air at 450 °C for 30 min to obtain fully converted SnO2 films. A CH3NH3PbI3-xClx perovskite layer was fabricated via a PbICl-seed-layer-assisted interdiffusion process. Details are provided in the experimental section. As shown in Fig. 3e, f, for a deposition time of 150 s, the short-circuit current density (J sc), open-circuit voltage (V oc), fill factor (FF), and PCE (%) were 17.84 mA/cm2, 1.03, 0.496, and 9.11%, respectively. As the deposition time increased from 150 to 180 s, J sc improved, and a higher PCE of 10.0 was obtained. The use of a deposition time of 210 s mainly affected the J sc and FF value, leading to a lower PCE of 8.22. To gain further insight into parasitic resistances, we calculated series resistance (R s) and shunt resistance (R sh) from the J–V curves. R s values are 10.4, 5.2, and 12.5 (ohm cm2); R sh values are 194.9, 558.5, and 167.1 (ohm cm2) for the time of 150, 180, and 210 s, respectively. The calculated parasitic resistances explain device performances in operation obtained from different electrochemical deposition condition. As shown in the SEM image in Fig. 3d, the poor morphology of the SnO2 film at a deposition time of 210 s is expected to impede charge transfer between CH3NH3PbI3-xClx and FTO, resulting in a reduced J sc.

Fig. 3
figure 3

Top-view SEM images of substrates for different deposition time. (a) Bare FTO and SnO2 films deposited for (b) 150 s, (c) 180 s, and (d) 210 s. Corresponding photovoltaic performance: (e) J–V curves and (f) photovoltaic parameters of PSCs with electrodeposited SnO2 ETL. Scale bar is 1 μm

Full size image

Considering that the electrodeposition process depends on the ion mobility in an electrolyte solution, we also explored the effect of temperature on the morphology of the films. Figure 4 shows top-view SEM images of films deposited at different bath temperatures with −0.7 V for 180 s. As expected, the surface morphology of the SnO2–Sn nanospheres prepared at different bath temperatures varies. The nanosphere size, roughness, and thickness seem to be affected, as the migration of Sn2+ ions was enhanced at higher temperature. The photovoltaic efficiency of PSCs fabricated using these films is compared in Fig. 4e, f. A finer SnO2 film yields better performance, and the optimum efficiency was obtained for the film deposited at 60 °C. The SnO2 film morphology is expected to significantly affect the PSC performance because planar PSCs have a direct interface between the ETL and the perovskite layer. The improved conformality could result in good contact that affords enhanced electron transport [25]. The SEM images of perovskite layer fabricated from varied ETLs were provided in supporting information (SI) Additional file 1: Figure S1.

Fig. 4
figure 4

Top-view SEM images of SnO2 films electrodeposited at various bath temperatures. (a) RT, (b) 40 °C, e(c) 60 °C, and (d) 70 °C. Corresponding photovoltaic performance: (e) J–V curves and (f) photovoltaic parameters of PSCs with electrodeposited SnO2 ETL. Scale bar is 1 μm

Full size image

To further examine the effect of temperature on the morphology with respect to the blocking effect of the electrodeposited SnO2 films, we conducted CV measurements in an aqueous electrolyte containing [Fe(CN)6]3−/[Fe(CN)6]4− because the redox reaction depends on charge transfer between the FTO and the electrolyte [26]. The electron transfer kinetics can be interpreted by extracting the separation of the peak potentials and peak current of a redox system from the CV curves. If the redox reaction between [Fe(CN)6]3−/[Fe(CN)6]4− ions is hampered by the SnO2 layer, the oxidized and reduced forms of the redox couple exhibit peak potentials that are shifted away from the control on bare FTO and become semireversible; consequently, the peak current density will be reduced [27]. Figure 5a shows the CV curves of bare FTO and the SnO2 films. The CV curve of bare FTO clearly shows a reversible redox reaction, indicating a lower barrier to electron transfer. In contrast, the FTO with electrodeposited SnO2 exhibits a larger peak-to-peak separation (ΔE p) of the cathodic and anodic peak potentials compared to that of bare FTO. The ΔE p values of films deposited at room temperature (RT), 40, 60, and 70 °C are 125, 175, 207, and 230 mV, respectively. This indicates that the kinetics of the redox reaction are changed by the blocking effect of the SnO2 films. In contrast, charge transfer at the FTO is highly suppressed by the film deposited at 70 °C, implying that the SnO2 is densely deposited onto the FTO. The thick SnO2 film could result in less effective and slower electron transport, negatively affecting the photovoltaic performance. The cathodic peak current (I p) of the films decreased with increasing bath temperature, indicating that the FTO coverage was improved.

Fig. 5
figure 5

Various analyses for the films. (a) CV curves in redox solution system and (b) transmission spectra of bare FTO and SnO2 films electrodeposited at different bath temperatures in redox solution system. (c) XPS Sn 3d spectrum of thermally treated SnO2 film

Full size image

On the basis of the CV results and SEM images, we could speculate that the FTO electrode at low temperature is covered with fewer nanoparticles; therefore, we conclude that the SnO2 film fabricated at 60 °C has a suitable thickness and morphology for use in PSCs and has a dominant effect on the device performance. The optical transmission of the SnO2 films is also compared (Fig. 5b). As the bath temperature increases from RT to 60 °C, the transmittance of the SnO2 films is enhanced compared to that of FTO. At a high bath temperature of 70 °C, the transmittance is inferior to that of FTO, which is attributed to the increased film thickness, as evidenced by the SEM image.

XPS was performed to measure the composition of the electrodeposited films. The XPS spectrum of the thermally treated SnO2 film is shown in Fig. 5c. Sn 3d5/2 and Sn 3d3/2 peaks at binding energies of 486.6 and 495 eV, respectively, were observed, whereas the film without heat treatment showed Sn 3d5/2 and Sn 3d3/2 peaks at 484.8 and 493.2 eV, respectively (SI, Additional file 1: Figure S2) [21]. The SnO2 film is clearly obtained through heat treatment.

On the other hand, although SnO2 electrodeposition provides a versatile and low-cost route toward scalable manufacturing systems [28], the demonstrated photovoltaic performance of the electrodeposited SnO2 films is not impressive. To improve the device performance, TiCl4 treatment was used to modify the SnO2 surface. As shown in Fig. 6a, the device based on SnO2 without TiCl4 treatment shows a J sc value of 18.12 mA/cm2, a V oc value of 1.04 V, a FF of 57.3%, and a PCE of 10.83%. In comparison, the device based on SnO2 with TiCl4 treatment (SnO2–TiCl4) exhibits a J sc value of 18.65 mA/cm2, a V oc value of 1.02 V, a FF of 79.1%, and a PCE of 14.97% (a 38% enhancement). The efficiency improvement is attributed mainly to the improved J sc and FF.

Fig. 6
figure 6

Cell performance with IPCE and PL data. (a) J–V curves and (b) EQE spectra of PSC devices based on SnO2 and SnO2–TiCl4. (c) Steady-state PL spectra of FTO/SnO2/perovskite and FTO/SnO2-TiCl4/perovskite samples. (d) Recombination time versus current density

Full size image

To understand the mechanism by which TiCl4 treatment improves the J sc value, we measured the EQE (Fig. 6b). The EQE of the SnO2–TiCl4 device shows an increase from 17.8 to 18.6 mA/cm2 in the entire wavelength spectral region. The enhancement in the EQE after TiCl4 treatment is in good agreement with the improved J sc in the J–V curves, which implies efficient charge collection. The EQE enhancement is expected to be originated from a better injection of electrons at the ETLs/perovskite interface [29, 30]. To further investigate the electron injection, the steady-state PL was measured for substrates with both ETLs. Figure 6c shows the PL spectra of the FTO/SnO2/perovskite and FTO/SnO2–TiCl4/perovskite samples. Compared to the SnO2-based film, the SnO2–TiCl4-based film exhibited reduced PL intensity, indicating that electron transfer from the perovskite to the ETL was enhanced by TiCl4 treatment since the PL emission of perovskite layer is quenched by contact. Possibly, the enhanced electron injection in ETLs with TiCl4 treatment improved the EQE. To further examine the improved performance of the SnO2–TiCl4-based device, intensity-modulated photovoltage spectroscopy (IMVS, Additional file 1: Figure S3) was performed to characterize the recombination time (τ r) (Fig. 6d). The recombination lifetime depends on the concentration of charge carriers in the solar cell. Thus, the recombination time is influenced by the current density, which is modulated by varying the light intensity. The carrier recombination time for the SnO2–TiCl4-based device was 1.17 times longer than that of the SnO2-based devices. The longer time constant for recombination is expected to afford an increase in J sc, FF, and better device performance [31, 32]. The device statistics (30 samples for each) were provided in Additional file 1: Figure S4.

Conclusions

In summary, we demonstrated a versatile and scalable electrodeposition technique to obtain a SnO2 ETL for planar heterojunction PSCs. The properties of the electrodeposited SnO2 depended strongly on the deposition time, electrolyte bath temperature, and applied voltage. Moreover, devices based on SnO2 treated with TiCl4 showed significantly enhanced V oc and J sc, leading to a PCE enhancement of 42%.