Enhanced photovoltaic properties of perovskite solar cells by TiO₂ homogeneous hybrid structure

Abstract

In this paper, we fabricated a TiO₂ homogeneous hybrid structure for application in perovskite solar cells (PSCs) under ambient conditions. Under the standard air mass 1.5 global (AM 1.5G) illumination, PSCs based on homogeneous hybrid structure present a maximum power conversion efficiency of 5.39% which is higher than that of pure TiO₂ nanosheets. The enhanced properties can be explained by the better contact of TiO₂ nanosheets/nanoparticles with CH₃NH₃PbI₃ and fewer pinholes in electron transport materials. The advent of such unique structure opens up new avenues for the future development of high-efficiency photovoltaic cells.

1. Introduction

TiO₂ is an optimized candidate for the development of nanoscale architectures due to its appropriate electronic band structure, photostability, non-toxicity and high chemical inertness [1–9]. Among diverse TiO₂ nanostructure morphologies, TiO₂ nanosheets (TiO₂NSs) have been widely applied in the field of photovoltaic cells due to their sufficient surface area [10]. Many endeavours have been made to fabricate perovskite solar cells (PSCs) based on TiO₂NSs films [11,12]. Although such PSCs based on TiO₂NSs films show promising photovoltaic performance, poor properties are shown after the contact of CH₃NH₃PbI₃ (MAPbI₃) and fluorine-doped tin oxide (FTO) along the pinholes in the TiO₂NSs films. Given all that, microscopic and high specific surface area TiO₂ nanoparticles (TiO₂NPs) can be introduced to wipe out the pinholes and retain high specific surface area of TiO₂NSs [13–16]. Once TiO₂NPs were deposited onto TiO₂NSs, a peculiar TiO₂NSs/NPs homogeneous hybrid structure can be formed. This structure is anticipated to simultaneously possess higher specific surface area and fewer pinholes. Hence, it will definitely increase the power conversion efficiency of PSCs. However, up to now, scarce relevant works have been reported on those issues.

Here, we report a TiO₂NSs/NPs homogeneous hybrid structure as electron transport material (ETM) in PSCs, which was fabricated by the simple hydrothermal and chemical bath deposition (CBD) method. The photovoltaic properties of PSCs based on TiO₂NSs/NPs homogeneous hybrid structure are superior to those based on bare TiO₂NSs due to the introduction of TiO₂NPs. This homogeneous hybrid structure may provide a new strategy for improving the performance of PSCs.

2. Experimental procedure

2.1. Experimental

The compact TiO₂ (c-TiO₂) and TiO₂NSs films were synthesized by the simple CBD and hydrothermal method [17,18]. TiO₂NSs films were fabricated by a hydrothermal method at 170°C for 3 h and annealed at 550°C for 2 h in air. To deposit TiO₂NPs, the FTO/c-TiO₂/TiO₂NSs substrates were placed into an aqueous solution with 0.07 M TiCl₄ at 70°C for 30 min. In addition, the TiO₂NPs films were annealed at 450°C for 15 min in air atmosphere. The entire procedure of fabricating TiO₂NPs films was termed as one cycle (C). Here, the procedure was repeated for 1C, 3C, 5C, 7C and 9C. Finally, the samples were annealed at 450°C for 30 min under ambient conditions and taken out after cooling down to room temperature.

The TiO₂ substrates were treated by UV–ozone for 15 min before the MAPbI₃ was deposited. Then, 50 µl PbI₂ (462 mg ml⁻¹ in dimethylformamide) was spin-coated onto the substrates (1.5 cm ×1.5 cm) with a low speed of 500 r.p.m. for 5 s and with a high speed of 4000 r.p.m. for 30 s. After that, the substrates were annealed at 100°C for 20 min. CH₃NH₃I (MAI) was synthesized by the technique as described in [19]. In the next step, 200 µl MAI (10 mg ml⁻¹ in isopropanol) was spin-coated onto the substrates. After rinsing with isopropanol, the perovskite films were annealed at 100°C for 40 min. Thirty millilitre hole transport material (HTM) as described in [19] was spin-coated onto the perovskite films at 4000 r.p.m. for 10 s (0.073 g 2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenylamine)-9,9′-spirobifluorene, 29 µl of 4-tert-butylpyridine, 18 µl of a lithium-bis(trifluoromethanesulfonyl)imide (Li-TFSI) solution (520 mg Li-TFSI/1 ml acetonitrile) were dissolved in 1 ml chlorobenzene) [20]. Finally, the Ag back electrode was deposited by thermal evaporation.

2.2. Characterization

The microstructure and morphology of the films were observed by a MAGELLAN 400 scanning electron microscope. Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) measurements were conducted by a JEOL JEM-2100F microscope. The analysis of composition and crystal structure was conducted using X-ray diffraction (XRD; Rigaku D/max-2500) using Cu Kα radiation (λ = 1.5418 Å). Ultraviolet–visible (UV-vis) absorption spectra were measured in the range from 200 to 1000 nm by a UV-3150 double-beam spectrophotometer at room temperature. The current–voltage curves of PSCs were recorded by a Keithley 2400 source measure unit.

3. Results and discussion

3.1. Scanning electron microscope images

In figure 1, we present top-view scanning electron microscopy (SEM) images of the TiO₂NSs array films after being combined with TiO₂NPs for different cycles. TiO₂NSs are coated with TiO₂NPs uniformly, and the surface of the TiO₂NSs becomes rough gradually. There was a notable increase in the adsorption of TiO₂NPs deposited onto the TiO₂NSs films, and the TiO₂NPs architecture could result in a full TiO₂NSs/NPs homogeneous hybrid film. The inset of figure 1b represents an enlarged image of TiO₂NSs deposited with TiO₂NPs of 1C. Because of the high surface energy and reactivity of {001} facets, the TiO₂NSs can adsorb more particles on this facet. The inset of figure 1e represents an enlarged view of TiO₂NPs deposited on TiO₂NSs for 7C.

Figure 1.

Top-view SEM images of TiO₂NPs deposited on TiO₂NSs for (a) 0C, (b) 1C, (c) 3C, (d) 5C, (e) 7C and (f) 9C. The insets in (b) and (e) represent the enlarged images of TiO₂NSs/NPs of 1C and 7C, respectively.

As portrayed in figure 2a, the aligned TiO₂NSs films are relatively smooth and vertically grew on the FTO/c-TiO₂ substrates, providing the transfer path for electrons. The thickness of the film is approximately 370 nm. Figure 2b is the SEM image of TiO₂NSs film after the deposition of 7C TiO₂NPs; obviously, the surface of the TiO₂NSs became rougher, which benefits for the combination of MAPbI₃ [21]. The cross-sectional SEM images of PSC devices with and without TiO₂NPs are presented in figure 2c,d, respectively. MAPbI₃ crystals are compactly arranged on the surface of TiO₂ substrates. When the ETM films are not coated with TiO₂NPs, some pinholes are seen, as shown in figure 2c. However, no pinholes are seen in the TiO₂NSs/NPs/MAPbI₃ films which could be ascribed to the full TiO₂NSs/NPs film, and higher power conversion efficiency (PCE) was obtained.

Figure 2.

Cross-sectional SEM images of (a) TiO₂NSs films grown on FTO at 170°C for 3 h, (b) TiO₂NSs after coated with 7CNPs, (c) device based on TiO₂NSs and (d) device based on TiO₂NSs/7CNPs.

3.2. Transmission electron microscopy and high-resolution transmission electron microscopy observation

The detailed microscopic characterization of TiO₂NSs/NPs nanostructures was performed by TEM, as shown in figure 3. Figure 3a clearly displays that TiO₂NPs uniformly grew on TiO₂NSs. This homogeneous structure was not damaged through ultrasonic processing, signifying that the TiO₂NSs and TiO₂NPs bonded firmly together. To clarify the growth mechanism, HRTEM was carried out. The inter-planar distance between the neighbouring lattice fringes of HRTEM is 0.189 nm which is corresponding to the inter-planar distance from {200} planes of TiO₂NPs (JCPDS no. 21–1272), as shown in figure 3b. The TEM and HRTEM images indicate that TiO₂NPs exhibit homogeneous fine grain microstructure, with an average grain size of approximately 10–20 nm. The staggered lattice of TiO₂NSs and TiO₂NPs was also observed by HRTEM and the joint parts are pointed out by the arrows in figure 3c. The plentiful growth interfaces inevitably will result in more charge transfer channels, which can enhance the electronic injection efficiency.

Figure 3.

(a) TEM bright-field image of TiO₂NSs/NPs and (b,c) HRTEM images of TiO₂NSs/NPs.

3.3. Phase composition and phase structure

The prepared TiO₂ and TiO₂/MAPbI₃ films were examined by XRD to characterize the crystal structure (figure 4a). The diffraction pattern of TiO₂NSs array film matches fairly well with the existing reference data available for this crystal from the Joint Committee on Powder Diffraction Standards (JCPDS no. 21–1272). In addition, after depositing with TiO₂NPs for 7C, no diffraction lines of impurities were detected. The black curve shows the XRD pattern of FTO/c-TiO₂/TiO₂NSs/NPs/MAPbI₃ film. Except the diffraction peaks of FTO, TiO₂ and a weak peak at 12.7°of PbI₂, other diffraction peaks all correspond to MAPbI₃. This indicates that the MAPbI₃ has a high purity and has been successfully coated onto the surface of TiO₂ films. The peaks of MAPbI₃ were sharp, indicating that MAPbI₃ possesses good crystallinity.

Figure 4.

(a) X-ray diffraction patterns of (from bottom to top) FTO/c-TiO₂/TiO₂NSs, FTO/c-TiO₂/TiO₂NSs/7CNPs and FTO/c-TiO₂/TiO₂NSs/7CNPs/MAPbI₃. The plot shows the X-ray intensity as a function of 2θ. (b) UV-vis absorption spectra of films.

3.4. Ultraviolet–visible absorption spectroscopy

The UV-vis absorption spectra of bare TiO₂NSs array films, TiO₂NSs/MAPbI₃ films and TiO₂NSs/NPs/MAPbI₃ films are plotted in figure 4b as a function of wavelength. The peak maximum of TiO₂NSs films occurs at approximately 370 nm and the films have no significant absorption for visible light due to the large band gap of TiO₂ [22,23]. However, after coating with MAPbI₃, the TiO₂NSs/MAPbI₃ films present excellent absorption capacity from 350 to 800 nm, and the absorption intensity increased substantially. The variation indicates that the deposited MAPbI₃ has significantly extended the photoresponse of TiO₂NSs films to the visible light region. The most striking thing is that after TiO₂NSs films combined with TiO₂NPs for 7 C, the TiO₂NSs/NPs/MAPbI₃ films exhibit stronger absorption than TiO₂NSs/MAPbI₃ films. This result indicates that the MAPbI₃ films combine with TiO₂NSs/NPs films better than TiO₂NSs films.

3.5. Photovoltaic characterization of perovskite solar cells based on TiO₂NSs/NPs films

To probe the effect of different amount of TiO₂NPs on photovoltaic performance, successive deposition cycle (C) experiments of TiO₂NPs were conducted. The working area of devices was 0.15 mm². All current–voltage curves were recorded in air and derived from reverse scan (the bias scan rate was 100 mV s⁻¹). Figure 5 shows the J–V characteristics of the PSCs under one sun AM 1.5G irradiance, and the corresponding parameters are summarized in table 1. The J–V characteristics are enhanced with the increase of TiO₂NPs in the early cycles, suggesting that a higher incorporated amount of TiO₂NPs induced more superior photovoltaic characteristics. Surprisingly, as the amount of TiO₂NPs increased upon the cycles to seven, the PSCs exhibited a best PCE of 5.39%, with V_oc = 0.82 V, J_sc = 17.06 mA cm⁻² and FF = 0.40.

Figure 5.

J–V characteristic of the lead iodide perovskite solar cells based on TiO₂NSs/NPs films of different deposition cycles.

Table 1.

Photovoltaic device parameters of the TiO₂NSs/NPs/CH₃NH₃PbI₃ solar cells.

cycles	Jsc (mA cm−2)	Voc (V)	FF	PCE
0	4.27	0.68	0.30	0.86
1	6.59	0.8	0.38	2.07
3	12.91	0.78	0.37	3.72
5	14.65	0.79	0.42	4.8
7	17.06	0.82	0.40	5.39
9	14.87	0.79	0.35	3.89

In the early 7 cycles, the V_oc significantly increased along with the decreased pinholes which will generate charge loss. After TiO₂NPs are deposited onto TiO₂NSs, the MAPbI₃ grain size tends to be larger than that of pure TiO₂NSs. Larger grain size leads to fewer grain boundaries and, consequently, the corresponding J_sc improved [24]. With the increase of TiO₂NPs, the surface of TiO₂NSs array films appears rougher. This rougher surface is favourable for the combination of MAPbI₃ and it will improve the fill factor of PSCs. However, when the deposited cycles are more than seven, the device performance of PSCs dropped. This phenomenon is presumably attributed to the excessive TiO₂NPs grain boundaries. When TiO₂NSs/NPs films transported photo-induced electrons from MAPbI₃ to FTO, more grain boundaries caused more electron annihilation. This inevitably will decrease the J–V characteristics of PSCs. Etgar et al. [11] fabricated MAPbI₃/TiO₂NSs solar cell with J_sc = 16.1 mA cm⁻², V_oc = 0.631 V, corresponding to a PCE of 5.5% under glovebox conditions. Compared with their PSCs based on pure TiO₂NSs, the TiO₂ homogeneous hybrid structure PSCs synthesized under air conditions exhibit higher J_sc, V_oc and compatible PCE.

Holes are transported to the counter electrode by HTM in scheme 1a (the short black arrows). Electrons are transported to FTO by TiO₂NPs (the long blue arrows) and TiO₂NSs/NPs (the long brown arrows). Compared with pure TiO₂NSs films, fewer pinholes exist in TiO₂NSs/NPs homogeneous hybrid films, and these hybrid films have higher specific surface area. Therefore, the photovoltaic performance of PSCs based on TiO₂NPs/NSs homogeneous hybrid structure is significantly enhanced.

Scheme 1.

Schematic of PSCs based on (a) TiO₂NSs/NPs homogeneous hybrid structure films and (b) pure TiO₂NSs films.

4. Conclusion

We fabricated a TiO₂NSs/NPs homogeneous hybrid structure, the ETM in PSCs, by the hydrothermal and CBD method. When TiO₂NPs are deposited onto TiO₂NSs for 7 cycles, the PSCs based on TiO₂NPs/NSs homogeneous hybrid structure exhibit photovoltaic performance with a best efficiency of 5.39% under one sun AM 1.5G solar spectrum, which is two and a half times that of bare TiO₂NSs. Compared with pure TiO₂NSs films, the MAPbI₃ combined better with rough and free of pinholes TiO₂NSs/NPs homogeneous hybrid films. Hence, it demonstrates that this extraordinary structure will have potential application for enhancing photovoltaic performance of PSCs in the future.

Data accessibility

The datasets supporting this article have been uploaded as part of the electronic supplementary material.

Authors' contributions

P.S., W.F. and H.Y. conceived and designed the experiments, and analysed the data results. P.S., L.L., F.F. and S.F. measured the SEM images. P.S. and D.D. carried out the TEM experiments. P.S. carried out the XRD experiments. P.S. and Y.X. performed the device fabrication and photovoltaic performance measurements. W.F. and H.Y. helped draft the manuscript. All authors discussed the mechanism and gave their final approval for publication.

Competing interests

We declare that we have no competing interests.

Funding

This work was financially supported by National Natural Science Foundation of China (grant no. 51272086).