Abstract

Compared with the crystalline Si solar cells, the - characteristics of CH3NH3PbI3 perovskite solar cells are different under forward and reverse scan, and the CH3NH3PbI3 film exhibits some polarization properties. To explore those performances of the mesoporous TiO2 layer based perovskite solar cells, we focus on the effect of modulating the spin-coating rate of the TiO2 precursor for mesoporous layer on - hysteresis of solar cells with the polar film by - curves, atomic force microscopy topographic images, and piezoresponse force microscopy phase images. Firstly, the AFM images illustrate that the polarization behaviors exist and the deformation scale is large at the corresponding position when the DC bias voltage increases. Secondly, it is suggested that the polar films which applied the positive DC biases voltage show a tendency to 0° phase angle, while the polar films which applied the negative DC biases voltage show a tendency to −180° phase angle. Thirdly, a weak polar hysteresis loop relation for CH3NH3PbI3 film was observed. Finally, the hysteresis index for the 1500 rpm mesostructured solar cell shows relatively low - hysteresis compared with the 3000 rpm mesostructured and the planar-structured solar cell. Our experimental results bring novel routes for reducing the hysteresis and investigating the polar nature for CH3NH3PbI3 material.

1. Introduction

Aiming at next-generation solar cells, the CH3NH3PbI3 perovskite as a light harvester has been widely used in solar cells owing to its low cost and outstanding desirable properties [1]. Currently, the efficiency, stability, and large area of the CH3NH3PbI3 perovskite solar cells are the main focuses, while the polarization behaviors of CH3NH3PbI3 perovskite film and hysteresis of CH3NH3PbI3-based perovskite solar cells receive insufficient attention and evolve slowly. Dualeh et al. first reported the hysteresis of CH3NH3PbI3-based perovskite solar cells [2]. Snaith et al. reported effect of three factors including scan rate, architectures, and mesoporous components on hysteresis [3]. Tress et al. reported the role of a compensated electric field on the rate-dependent hysteresis [4]. O’Regan et al. reported an evolution in band offset at the TiO2/CH3NH3PbI3 interface during Hysteresis [5]. Meloni et al. investigated the origin of this phenomenon with a combined experimental and computational approach [6]. Kutes et al. showed the presence of ferroelectric domains in high-quality β-CH3NH3PbI3 perovskite thin films that have been synthesized using a new solution-processing method [7]. Kim and Park investigated effects of perovskite crystal size and mesoporous TiO2 layer on Hysteresis of CH3NH3PbI3 perovskite solar cells [8]. Then they reported on ferroelectric polarization behavior in CH3NH3PbI3 perovskite in dark and under illumination [9]. Coll et al. reported on polarization switching and light-enhanced piezoelectricity in lead halide perovskites [10]. The low conductivity of the contact materials (TiO2 or HTMs) is one of reasons for the high hysteresis [11, 12].

In this work, we have investigated the polarization behaviors and local hysteresis phenomenon of the CH3NH3PbI3 films, as well as the polarization orientation by AFM and PFM. In addition, a weak polar hysteresis loop between the polarization and the external electric field of FS and RS for CH3NH3PbI3 film was observed. We also study the effect of spin-coating rate of TiO2 precursor for mesoporous layer on HI of CH3NH3PbI3 perovskite solar cells.

2. Experimental and Methods

2.1. Materials

Unless stated otherwise, all materials were purchased from Sinopharm Chemical Reagent Beijing Co., Ltd., or Shanghai MaterWin New Materials Co., Ltd., and used as received. Co(III) PF6 Spiro (spiro-OMeTAD) solution was purchased from Beijing HuaMin New Materials Technology Co., Ltd.

2.2. Device Fabrication

The fluorine-doped tin oxide (FTO) conductive glasses were cleaned by ultrasonic in detergent, rinsed with deionized water, followed by washing with the mixture of deionized water, ethanol, and acetone at 1 : 1 : 1 by volume, and suffered from an O3/ultraviolet treatment for one hour. The TiO2 blocking (bl-TiO2) underlayer was deposited on the cleaned FTO conductive glasses by spin-coating method at 2000 rpm for 30 s and sintered at 500°C for 30 min.

In our trials, the following two parts were carried out. () In terms of polarization behaviors measurements, we took the FTO/bl-TiO2/mp-TiO2/CH3NH3PbI3 mesostructured device and the FTO/bl-TiO2/CH3NH3PbI3 planar structured device for measurements. A TiO2 paste was diluted by ethanol at 2 : 7 by weight to prepare the mp-TiO2 precursor. The mp-TiO2 precursor was deposited on top of the FTO/bl-TiO2 substrates with the spin-coating rates for 3000 rpm and 1500 rpm for 30 s and sintered at 500°C for 30 min to form mp-TiO2 layers. The perovskite films were prepared by two-step sequential deposition method [14]. The PbI2 (1 mol/L) solution which was stirred for one hour at 70°C was deposited on the mp-TiO2 film at 5500 rpm for 15 s, followed by thermal annealing at 100°C for 30 min on hot plate to form a stable film. The PbI2 solution was kept at 70°C while spin-coating. The methylammonium iodide solution (MAI 10 mg/ml dissolved in 2-isopropanol) was spin-coated on top of PbI2 layers at 0 rpm for 10 s, followed by 3000 rpm for 30 s to form CH3NH3PbI3 films, and then the as-fabricated films were dried at 100°C for 30 min. () In terms of the hysteresis in measurements, we took the CH3NH3PbI3 perovskite solar cells with different thickness mp-TiO2 layers and without mp-TiO2 layers for measurements. To prepare the CH3NH3PbI3 perovskite solar cells, the following steps were carried out. The spiro-OMeTAD solution was deposited on top of CH3NH3PbI3 films by spin-coating at 2000 rpm for 30 s. Finally, a candle burning method was used to prepare carbon electrode as counter electrode. The FTO-candle soot film was directly deposited by putting a cleaned FTO glass in the candle flame for about 3 s [15]. Then the FTO-candle soot film and the CH3NH3PbI3 perovskite film were directly clamped to fabricate the solar cell.

2.3. Characterization

The curves were measured by a VersaSTAT3 source meter under standard air-mass 1.5 global (AM 1.5G) illumination (100 mW·cm−2) provided by a solar simulator (Oriel Sol 3A). The monochromatic incident photon-to-electron (IPCE) was measured with an IPCE measurement tool (Institute of Physics, CAS, China) in our laboratory. 3D surface topography AFM images were obtained by Dimension AFM (Bruker). PFM measurements require NanoScope software 7.20 R1 SR1 with the NanoScope V Controller using a Dimension AFM (Bruker). The photoluminescence (PL) diagrams were obtained by the LabRAM HR800, which was made by HORIBA Jobin Yvon. The ultraviolet-visible (UV-Vis) absorption spectrums were obtained by the test system, which consisted of AvaSpec-ULS2048XL, test software, AvaLight-DH-S-BAL as a light source, AvaSpere-50-REFL, AvaSphere-IRRAD, and so on.

3. Results and Discussions

3.1. 3D Surface Topography AFM Images Illustrate the Polarization Behaviors and Local Hysteresis Phenomenon of the CH3NH3PbI3 Films

To investigate the polarization behaviors and local hysteresis phenomenon of the CH3NH3PbI3 films, the topography and the dielectric response of the CH3NH3PbI3 films were measured by AFM.

A dielectric can be polarized by an applied electric field. When a dielectric is placed in an electric field, electric charges do not flow through the material but only slightly shift from their average equilibrium positions causing dielectric polarization. Because of dielectric polarization, positive charges are displaced toward the field and negative charges shift in the opposite direction [16].

In AFM images, the DC bias voltage applied an electric field, which is capable of rotating the orientation of the polarization. Thus, it is dielectric polarization.

To understand the polarization behaviors of CH3NH3PbI3 films, we applied a DC bias voltage between the CH3NH3PbI3 film and the conductive tip during the AFM measurement; the DC biases voltages were set to 0 V, +3 V, +6 V, and −10 V, respectively. In addition, for poling the samples, an AC voltage was applied from the conductive tip to the film’s surface during the AFM and PFM measurements.

We take the CH3NH3PbI3 perovskite films with the spin-coating rate of the mp-TiO2 precursor for 1500 rpm as the typical sample for the AFM measurement. Figures 1 and 2 are the same substrate. Figures 1(a), 1(b), 1(c), and 1(d) are the same substrate and the same position, and Figures 2(a) and 2(b) are the same substrate and the same position. In order to see clearly, Figures 1(c) and 1(d) are the enlargement of the center of Figures 1(a) and 1(b). Because when applying negative DC voltage of −10 V, a distortion may occur. The same substrate and the same position were selected for Figures 2(a) and 2(b). Figure 1 shows the 3D surface topography AFM images of the CH3NH3PbI3 perovskite films with mp-TiO2 layer which applied different DC bias voltage and AC voltage groups. The contrast of the CH3NH3PbI3 perovskite films in 3D surface topography AFM images varies depending on the height of the CH3NH3PbI3 perovskite crystals; that is, the brighter the color of AFM image, the higher the peak on the surface of CH3NH3PbI3 perovskite layer; in contrast, the darker the color of AFM image, the deeper the valley on the surface of CH3NH3PbI3 perovskite layer. Compared with the sample which applied (0 V, 0 V) (Figure 1(a)), the surface topography of the samples is more uneven with the increasing of applied (~8 V, 0 V) (Figure 1(b)) and (~10 V, 0 V) (Figure 2(a)) AC voltage, because the AC voltage provides the energy of the vibration of the CH3NH3PbI3 perovskite films, resulting in deformations. Huang et al. stated that the piezoresponse originates only from the polarization induced by the crystal structure deformation because of the applied AC bias [17]. When a DC bias voltage (+3 V) (Figure 1(c)) is applied, the surface topography of the sample has a clearly significant polarization behavior in the CH3NH3PbI3 film. Furthermore, with the DC bias voltage increasing to +6 V (Figure 1(d)), the polarization behavior is stronger at the corresponding position in 3D surface topography AFM images. Moreover, when the voltage group is increased to (~10 V, −10 V) (Figure 2(b)), a serious distortion occurs, because the breakdown voltage of the CH3NH3PbI3 film we fabricated is lower than 10 V. These results presented here demonstrate that the CH3NH3PbI3 perovskite films have a scale of vertical deformation under the DC bias voltage and AC voltage groups, and the deformation scale is large at the corresponding position when the DC bias voltage increases. Huang et al. stated that when the bias field strength exceeded a certain threshold, a permanent breakdown occurred in CdS/CdTe film [17]. In this work, similarly, when applying negative DC voltage of −10 V on perovskite film, a serious distortion occurred in AFM images. According to the AFM image, we infer that the −10 V is probably up to breakdown voltage. The deformation is reversible if the voltage applied is below or far below the breakdown voltage. When the voltage applied is higher or much higher than the breakdown voltage, the distortion is not reversible. Huang et al. reported that CdS/CdTe film can withstand a large bias field which could reach 8 × 104 V/cm; this suggests that there is less possibility of electrical failures during future applications [17]. In our work, in terms of perovskite film, it can bear 6 V in our experiments, if we take ~500 nm for its thickness, perovskite film can withstand a larger bias field of 1.2 × 105 V/cm. So there is less possibility of electrical failures during future applications of the perovskite solar cells.

(a)
(a)
(b)
(b)
(c)
(c)
(d)
(d)
Figure 1 
3D surface topography AFM images of the CH3NH3PbI3 perovskite films with mp-TiO2/bl-TiO2/FTO structure which applied DC biases voltage and AC voltage groups are (a) (0 V, 0 V), (b) (~8 V, 0 V), (c) (~8 V, +3 V), and (d) (~8 V, +6 V). (a), (b), (c), and (d) are the same substrate and the same position. In order to see clearly, (c) and (d) are the enlargement of the center of (a) and (b).
(a)
(a)
(b)
(b)
Figure 2 
DC biases voltage and AC voltage groups are (a) (~10 V, 0 V) and (b) (~10 V, −10 V). Figures 2(a) and 2(b) are the same substrate and the same position.
3.2. Effect of Spin-Coating Rate of TiO2 Precursor for Mesoporous Layer and Additional DC Bias Voltage on the PFM Phase Orientation of CH3NH3PbI3 Films

Two vector diagrams, shown in Figure 3, illustrate a number of properties of the digital lock-ins, without going into the details of the actual lock-in implementation. The digital lock-in implementation phase angle varies between −180 and +180 degrees, not 0 to 360 degrees, and the raw outputs of the lock-ins are the inphase and quadrature values [13].

(a)
(a)
(b)
(b)
Figure 3 
Vector diagram of field and sample oscillation vectors for (a) 0° and (b) 90° drive phase [13].

The CH3NH3PbI3 films were deposited on the FTO/bl-TiO2 substrates without the mp-TiO2 layer and the FTO/bl-TiO2/mp-TiO2 substrate with the spin-coating rates of the mp-TiO2 precursor for 1500 rpm and 3000 rpm, respectively.

We can compare contrast in PFM image with the scale bar. The contrast in PFM image represents the tendency of the polarization. In order to further investigate the polarization-switching behaviors, the switching measurements were performed by applying positive biases voltage of +3 V and +6 V and negative biases voltage of −3 V and −6 V to the conductive tip. The orientation of the polarization of the CH3NH3PbI3 perovskite films can be characterized by the corresponding PFM phase images. To estimate the effect of mp-TiO2 layer on the polarization behaviors, the CH3NH3PbI3 perovskite films with planar structure (w/o mp-TiO2 layer) and mesoporous structure (with the spin-coating rates of the mp-TiO2 precursor for 3000 rpm and 1500 rpm, resp.) were selected as samples for the PFM measurements. Note that overall CH3NH3PbI3 perovskite thin film is fabricated by the same preparation process for all devices regardless of mp-TiO2 layer.

The PFM phase images of the CH3NH3PbI3 perovskite films which applied (0 V, 0 V) without mp-TiO2 layer and with the different spin-coating rates of the mp-TiO2 layers are depicted in Figures 4(a)–4(c), which implies the absence of polarization in CH3NH3PbI3 perovskite films in the absence of electric field. When the samples applied an AC voltage of ~8 V (Figures 4(d)–4(f)), it is found that the stimulated polarization shows a slight enhancement. Moreover, the sample with the spin-coating rate of the mp-TiO2 precursor for 1500 rpm shows the strongest polarization compared to other samples.

(a)
(a)
(b)
(b)
(c)
(c)
(d)
(d)
(e)
(e)
(f)
(f)
Figure 4 
PFM phase images of the CH3NH3PbI3 perovskite films which applied (0 V, 0 V) (a) w/o mp-TiO2 layer and (b) with the spin-coating rate of the mp-TiO2 precursor for 3000 rpm and (c) with the spin-coating rate of the mp-TiO2 precursor for 1500 rpm. PFM phase images of the CH3NH3PbI3 perovskite films which applied (~8 V, 0 V) (d–f) correspond to those which applied (0 V, 0 V) without mp-TiO2 layer and with the different spin-coating rates of the mp-TiO2 layers.

Figure 5 shows the PFM phase images of the CH3NH3PbI3 perovskite films which applied (~8 V, +3 V) and (~8 V, +6 V) without mp-TiO2 layer and with the different spin-coating rates of the mp-TiO2 precursor. Compared with the CH3NH3PbI3 perovskite films which applied (~8 V, +0 V) (Figures 4(d)–4(f)), PFM phase images of the CH3NH3PbI3 perovskite films show a clear distinction in the presence of an additional DC bias voltage of +3 V (Figures 5(a)–5(c)). Furthermore, the DC bias voltage of +6 V (Figures 5(d)–5(f)) is advantageous to the enhancement of the stimulated polarization. Interestingly, the positive DC bias voltage is capable of rotating the orientation of the polarization, which has a tendency to 0° phase angle. Besides, the higher positive DC bias voltage (+6 V) is beneficial to the enhancement of the tendency. Additionally, despite +3 V or +6 V of the positive DC biases voltage, the CH3NH3PbI3 perovskite film with the spin-coating rate of the mp-TiO2 precursor for 1500 rpm shows the most promising tendency to 0° phase angle among these samples.

(a)
(a)
(b)
(b)
(c)
(c)
(d)
(d)
(e)
(e)
(f)
(f)
Figure 5 
PFM phase images of the CH3NH3PbI3 perovskite films which applied (~8 V, +3 V) (a) w/o mp-TiO2 layer and with the spin-coating rate of the mp-TiO2 precursor for (b) 3000 rpm and (c) 1500 rpm. PFM phase images of the CH3NH3PbI3 perovskite films which applied (~8 V, +6 V) (d–f) correspond to those which applied (~8 V, +3 V) without mp-TiO2 layer and with the different spin-coating rates of the mp-TiO2 precursor.

Figure 6 shows the PFM phase images of the CH3NH3PbI3 perovskite films which applied (~8 V, −3 V) and (~8 V, −6 V) without mp-TiO2 layer and with the different spin-coating rates of the mp-TiO2 precursor. Similar stimulated polarization behavior is observed showing the capacity of rotation (Figures 6(a)–6(f)) compared to the CH3NH3PbI3 perovskite films which applied positive DC biases voltage (Figures 5(a)–5(f)) except for the orientation of the polarization, which has a tendency to −180°. Besides, the films which applied the negative DC bias voltage of −6V (Figures 6(d)–6(f)) show a stronger polarization than the films which applied −3 V (Figures 6(a)–6(c)). In addition, regardless of −3 V or −6 V of the negative DC biases voltage, the CH3NH3PbI3 perovskite film with the spin-coating rate of the mp-TiO2 precursor for 1500 rpm shows the most promising tendency to −180° phase angle among these samples.

(a)
(a)
(b)
(b)
(c)
(c)
(d)
(d)
(e)
(e)
(f)
(f)
Figure 6 
PFM phase images of the CH3NH3PbI3 perovskite films which applied (~8 V, −3 V) (a) w/o mp-TiO2 layer and with the spin-coating rate of the mp-TiO2 precursor for (b) 3000 rpm and (c) 1500 rpm. PFM phase images of the CH3NH3PbI3 perovskite films which applied (~8 V, −6 V) (d–f) correspond to those which applied (~8 V, −3 V) without mp-TiO2 layer and with the different spin-coating rates of the mp-TiO2 precursor.

As a comparison of Figures 5 and 6, it is suggested that the CH3NH3PbI3 perovskite films which applied the positive DC biases voltage (Figure 5) show a tendency to 0° phase angle, while the films applied the negative DC biases voltage (Figure 6) showing a tendency to −180° phase angle. Moreover, the CH3NH3PbI3 perovskite film which applied +6 V and −6 V bias with the spin-coating rate of the mp-TiO2 precursor for 1500 rpm shows the most promising tendency to 0° and −180° phase angle, respectively.

3.3. A Weak Polar Hysteresis Loop Relation between the Polarization and the External Electric Field of FS and RS for CH3NH3PbI3 Film

Figure 7 shows a weak polar hysteresis loop relation between the polarization and the external electric field of FS and RS. In this measurement, a DC bias varying between −6 V and +4 V was applied between the CH3NH3PbI3 film and the conductive tip, in addition to the 8 V AC bias. As shown in Figure 7, a DC bias varying between −6 V and +4 V changes the phase orientation of the polarization from about 110° to −65° phase angle, which indicates that the DC bias voltage is capable of rotating the orientation of the polarization. Moreover, the separated curves of FS and RS are reasonable to infer that the CH3NH3PbI3 perovskite could potentially be a polar material.


Figure 7 
Phase-voltage curve of the typical CH3NH3PbI3 perovskite film, collected by PFM at the fixed conductive tip with DC bias from −6 V to +4 V and AC bias of 8 V.
3.4. Effect of Spin-Coating Rate of TiO2 Precursor for Mesoporous Layer on HI of CH3NH3PbI3 Perovskite Solar Cells

Figures 8(a)–8(c) show the curves of the CH3NH3PbI3 perovskite solar cells incorporating mp-TiO2 layers with different spin-coating rates for 3000 rpm and 1500 rpm together with a planar structure without mp-TiO2 layer, and their curves depend on voltage sweep direction. Owing to the hysteresis in measurements, it is crucial to report both RS and FS measurements. It is noted that the measurements are performed in both RS and FS directions at a scan rate of 0.1 V/s and overall back contact is carbon electrode, as well as under simulated AM1.5G (100 mW/cm2 irradiation). Moreover, we also present IPCE spectrums and the integrated of the CH3NH3PbI3 perovskite solar cells depending on their corresponding films (Figures 8(d)–8(f)). As shown in Figures 8(d)–8(f), the integrated proved by IPCE measurement is in good agreement with that derived from the measurements (Table 1).

Table 1 
Solar cell performance parameters w/o mp-TiO2 layer or with the spin-coating rates of the mp-TiO2 precursor for 3000 rpm and 1500 rpm.

Figure 8 
curves of the CH3NH3PbI3 perovskite solar cells with (a) planar structure and the spin-coating rate of the mp-TiO2 precursor for (b) 3000 rpm and (c) 1500 rpm. The corresponding IPCE spectrum and the integrated short circuit current density () of the CH3NH3PbI3 perovskite solar cells with (d) planar structure and the spin-coating rate of the mp-TiO2 precursor for (e) 3000 rpm and (f) 1500 rpm.

Figures 9(a)–9(c) show the HIs of six perovskite solar cells with planar structure (w/o mp-TiO2 layer) and mesoporous structure (the spin-coating rates of the mp-TiO2 precursor for 3000 rpm and 1500 rpm, resp.). And Figures 9(d)–9(f) show the PCEs. The figures showed good reproducibility of the hysteresis and characteristics of the devices.

(a)
(a)
(b)
(b)
(c)
(c)
(d)
(d)
(e)
(e)
(f)
(f)
Figure 9 
HIs of six perovskite solar cells with (a) planar structure and the spin-coating rate of the mp-TiO2 precursor for (b) 3000 rpm and (c) 1500 rpm. PCEs of six perovskite solar cells with (d) planar structure and the spin-coating rate of the mp-TiO2 precursor for (e) 3000 rpm and (f) 1500 rpm.

Here reported modified HIs are defined by where and represent photocurrent density at 80% of for the RS and FS, respectively [8].

The HIs for CH3NH3PbI3 perovskite solar cells were calculated by (1). As shown in Table 1, the HIs by (1) are calculated to be 0.611, 0.255, and 0.181, for planar structure without mp-TiO2 layer, the spin-coating rates of the mp-TiO2 precursor for 3000 rpm and 1500 rpm, respectively. Compared with the planar structure without mp-TiO2 layer, the HI for the mesostructured device with mp-TiO2 layer is smaller, which indicates that the mp-TiO2 layer reduces the hysteresis. Meanwhile, the HI for the 1500 rpm mesostructured device is smaller than that for 3000 rpm, which indicates that the mp-TiO2 layer for 1500 rpm has smaller hysteresis than that for 3000 rpm. It is reasonable to believe that the relatively low spin-coating rate in a proper range of mp-TiO2 precursor reduces the hysteresis. Thus, the interface difference between mp-TiO2 (1500 rpm and 3000 rpm) and perovskite is the main reason for the different hysteresis.

Figure 10(a) shows that three PL diagrams with the main PL peaking at 750~800 nm are attributed to the intrinsic band gap of approximate 1.6 eV from the CH3NH3PbI3 crystal lattice [18]. The structure of planar device is glass/FTO/bl-TiO2/CH3NH3PbI3. The mesostructured device is glass/FTO/bl-TiO2/mp-TiO2/CH3NH3PbI3. Planar structure device is different from mesoporous structure device in structure, so we do not compare their PL intensity, and we just take the planar structure as a reference. Compared with the PL intensity for 1500 rpm, the PL intensity for 3000 rpm is higher, which is in good agreement with the measurements. Jiang et al. provided the PL diagrams of glass/perovskite and glass/bl-TiO2/perovskite. A higher PL intensity was observed for glass/perovskite than glass/bl-TiO2/perovskite [19]. This is certainly a theoretical result because photo-generated electrons are more easily transferred to bl-TiO2 layer than glass; thus, the radiation recombination of photo-generated electrons and holes with glass/bl-TiO2/perovskite structure is lower than glass/perovskite structure, so a lower PL intensity was observed for glass/bl-TiO2/perovskite structure than glass/perovskite structure.

(a)
(a)
(b)
(b)
Figure 10 
(a) PL diagram and (b) UV-Vis absorption spectrum of the CH3NH3PbI3 perovskite thin films with planar structure and the spin-coating rates of the mp-TiO2 precursor for 1500 rpm and 3000 rpm.

In terms of absorption spectrum of the CH3NH3PbI3 perovskite solar cells shown in Figure 10(b), the mesostructured devices with mp-TiO2 layer have better absorption than the planar structured device. In addition, the mesostructured device with mp-TiO2 layer for 1500 rpm has the best absorption in our experiments. The mp-TiO2 layer is a skeleton structure. The perovskite layer can grow in and on mp-TiO2 layer. In the condition of the same method for perovskite formation, the perovskite layer is thick when the spin-coating rate of the mp-TiO2 precursor is slow. The thicker the perovskite film is, the stronger the UV-Vis absorption behavior is.

Tripathi et al. reported their curves of perovskite devices with different scan rates. Little difference of hysteresis exists with scan rates of 0.02 V/s, 0.05 V/s, 0.15 V/s, and 0.30 V/s [20]. Snaith et al. reported influence of scanning conditions on planar heterojunction perovskite solar cell current−voltage characteristics. A range of scan rates from 0.3 to 0.011 V/s was set. If scanned at 0.3 V/s it is possible to measure current-voltage curves without significant hysteresis [3]. The references indicate that hysteresis may depend on the scan rates, while the appropriate scan rate is different in different experiments. Thus, the high hysteresis obtained in this work is mainly due to the employed fast scan rate.

4. Conclusions

In summary, we investigated the effect of modulating spin-coating rate of TiO2 precursor for mesoporous layer on hysteresis of solar cells with polar CH3NH3PbI3 perovskite thin film. Firstly, the 3D surface topography AFM images illustrate the polarization behaviors and local hysteresis phenomenon of the CH3NH3PbI3 films. The result demonstrates that the CH3NH3PbI3 perovskite films have a scale of vertical deformation under the DC bias voltage and AC voltage groups, and the deformation scale is large at the corresponding position when the DC bias voltage increases. Secondly, we investigated the effect of spin-coating rate of TiO2 precursor for mesoporous layer on the PFM phase orientation of CH3NH3PbI3 films. It is suggested that the CH3NH3PbI3 perovskite films which applied the positive DC biases voltage show a tendency to 0° phase angle, while the negative DC biases voltage shows a tendency to −180° phase angle. Moreover, the CH3NH3PbI3 perovskite film which applied +6 V and −6 V bias with the spin-coating rate of the mp-TiO2 precursor for 1500 rpm shows the most promising tendency to 0° and −180° phase angle, respectively. Thirdly, a weak polar hysteresis loop relation between the polarization and the external electric field of FS and RS for CH3NH3PbI3 film indicates that the CH3NH3PbI3 perovskite could potentially be a polar material. Finally, the HI for the 1500 rpm mesostructured solar cell shows relatively low hysteresis compared with the 3000 rpm mesostructured solar cell and the planar structured solar cell. Our experiments bring novel routes for the research of the CH3NH3PbI3 perovskite material and CH3NH3PbI3-based perovskite solar cells. On the basis of our results, it may find its way into reducing the hysteresis and investigating the polar nature for CH3NH3PbI3 perovskite material.

Competing Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

This work was partially supported by Project of Natural Science Foundation of China (91233201 and 61376057), Project of Natural Science Foundation of Beijing (Z160002), the Opened Fund of the State Key Laboratory on Integrated Optoelectronics (no. IOSKL2016KF19), Beijing Key Laboratory for Sensors of BISTU (KF20161077203), and Beijing Key Laboratory for Photoelectrical Measurement of BISTU (GDKF2013005).