1 Introduction

Recently, hybrid perovskite solar cells (PSCs) have become one of the most important developments in the field of photovoltaic energy. They have attracted much attention due to their great power conversion efficiency (PCE) which has increased from 3.9 to 25.2% in 10 years [1,2,3,4,5,6]. Hybrid halide perovskites have become a promising materials in the solar field due to their good advantages including high optical absorption coefficient, long carrier diffusion length, high carrier mobility and tunable band structure [7,8,9,10,11,12,13]. The degradation and stability of perovskite solar cells depend on several parameters, among which layers composition plays an important role. It was reported that the incorporation of chloride (Cl–) in the starting materials could trigger perovskite crystallization and stabilize the MAPbI3 lattice [14, 15]. Furthermore, using mixed-halide perovskite as an active layer can be useful to control the bandgap and to increase perovskite resistance to decomposition [16]. Consequently, the conversion efficiencies of photovoltaic devices were improved with MAI and PbCl2 precursors.

Furthermore, many factors can influence the efficiency of the perovskite solar cells. In this circumstance, many efforts have been developed to improve the photovoltaic performance and to enhance the stability of the perovskite cells, as well as the perovskite deposition process [17,18,19], the annealing treatment [20, 21], the solvent engineering [15, 22], doping of perovskite material [23, 24], the perovskite solar cells structure [20, 25, 26], use of various electron and hole transport materials [27, 28] and doping of transport materials layer deposition processes [29].

Generally, standard perovskite solar cells are manufactured using titanium oxide (or other metallic oxides) as the electron transport layer (ETL) and 2,2′-7,7′-tetrakis (N, N-di-p-methoxyphenylamine)-9.9′-spirobifluorene (spiro-OMeTAD) as a hole transport layer (HTL) [30]. However, titanium oxides require high treatment temperature of titanium oxide and the spiro-OMeTAD suffers from very high cost and low stability. Indeed, this organic HTL material requires a complicated synthesis in several stages and is generally expensive which limits further the development of perovskite solar cells, especially in flexible substrates.

As it was cited in the previous paragraph, optimizing the interface layer is an important step in perovskite solar cell efficiency improvement. It was recommended to select interface transport layers with high charge carrier’s extraction. Recently, it has been demonstrated that the Tin(IV) oxide “SnO2” appears to be a better ETL for PSCs [31] compared to Titanium oxide (IV)“TiO2” [3] or Zinc oxide “ZnO” [32] due to its high mobility and low trap density.

A detailed study of the annealing treatments of the SnO2 (ETL layer) is demonstrated in this study. Three thermal annealing temperatures are used: 200, 220 and 240 °C. The nature of the HTL layer was also investigated. Two-hole transport materials, P3HT and PTAA, were used.

Different characterization tools were used like the XRD diffraction to study the crystallinity of the materials, the absorption allowing the study of the material absorption capacity, which affects directly the short-circuit current, and the photoluminescence PL to study the mechanism of recombination in the material. A complete device optimization method based on the combination of these different characterizations was used to improve the performance of these devices.

2 Experimental section

2.1 Materials and solar cell fabrication

ITO-coated glass substrates from Visiontek were cleaned sequentially by a performing ultrasonic treatment in acetone, isopropanol and deionized water for 5 min. A tin (IV) oxide (SnO2) nanoparticle colloidal suspension with a thickness of 25 nm was spin-coated at 4000 rpm for 40 s on cleaned ITO after UV–ozone treatment for 30 min and dried at different heat treatments (200, 220 and 240 °C) for 20 min in air. For the preparation of the perovskite compounds, methylammonium iodide and lead (II) chloride were dissolved in the glovebox in N. N-dimethylformamide (DMF) at a (3:1)molar ratio, respectively, for (MAI: PbCl2) and then stirred at room temperature overnight in the glove box. After that, a solution of the perovskite was spin-coated on the SnO2 layer at 4000 rpm for 40 s and then annealing at 100 °C for 35 min. A perovskite layer with thickness of 230 nm was obtained. Next, the hole transport layer HTL (P3HT (30 mg/ml), PTAA (10 mg/ml)) was spin-coated at 2000 rpm for 30 s. Finally, to complete the whole solar cells, a thin layer of Au (100 nm) was evaporated on top of the HTL layer. The illustration of the elaborated CH3NH3PbI3-xClx hybrid perovskite cell device is shown in Fig. 1.

Fig. 1
figure 1

a Structure of the device and b schematic energy-level diagram

2.2 Materials and device characterization

Perovskite films crystallinity was measured using X-ray diffraction (AerisPANalytical, Benchtop X-ray diffractometer) equipped using CoKα radiation at 40 kV and 15 mA. The UV–Vis spectra were measured by a Hitachi UV-2600 spectrophotometer in the wavelength range of 300 to 900 nm with an integrating sphere. Steady-state photoluminescence spectra were obtained with a 4500-fluorescence spectrometer, with a 450 W monochromatic xenon lamp source and a monochromator (selected wavelength λexc = 500 nm). The amount of charge transfer can be evaluated by calculating the parameter η, which defines the transfer efficiency (quenching efficiency), by using the following known expression [33]:

$$\eta \, = \,1\,{ - }\,I/I_{0}$$

where I0 and I are, respectively, the integrated PL intensities of the device in the absence and presence of the HTL layer.

The current density–voltage (JV) curves of solar cells were measured on a Keithley 2402 measure unit under AM1.5 G illumination at 1000 W m–2 using an Oriel 9403A-SR1.

3 Results and discussion

Perovskite crystallization is strongly dependent on ETL nature and surface state [34, 35]. Therefore, ETL annealing treatment might affect perovskite crystallinity and device efficiency. In this work, we investigate the annealing effect of the ETL layer (SnO2) on the properties of MAPbI3-xClx films after different annealing temperatures: 200, 220 and 240 °C for 20 min in air.

3.1 Effect of annealing temperature of SnO2

3.1.1 Crystallinity

The XRD patterns of MAPbI3-xClx, (MA = CH3NH3) on top of the SnO2 layers annealed at different temperatures (200, 220 and 240 °C) are shown in Fig. 2. The XRD patterns of the perovskite layers are approximately the same. Two dominant perovskite peaks appear at 17.11° and 34.03° and are attributed to the (110) and (220) planes, respectively, and another one at 37,85° which corresponds to the (310) planes of tetragonal MAPbI3 [36,37,38]. We also note the presence of one peak at 2θ = 18.85°, corresponding to the (100) plane and attributed to the pure chlorinated perovskite MAPbCl3 [39]. Its presence has already been reported for starting formulations involving MAI and PbCl2 in the same molar ratio (3:1) as ours (refs). In addition, it was noted that the diffraction peak at 12.44° could be attributed to the PbI2 lattice plane.

Fig. 2
figure 2

X-ray diffraction patterns of MAPbI3-xClx films deposited on ITO / SnO2 substrates for different SnO2 annealing temperatures: 200, 220 and 240 °C, (*) peaks attributed to ITO.

According to this diffractogram, we can also identify peaks, which are attributed to the ITO, which indicates the passage of the beam through the whole thickness. Since all measurements are made on films with the same thickness, (SnO2 (25 nm) and MAPbI3-xClx (230 nm)), diffraction peak intensities could be compared. Or, in a more rigorous way, intensity ratio such as PbI2/MAPI was used for comparison between the different conditions.

As shown in Fig. 2, we can notice that the SnO2/MAPbI3-xClx film with a layer of SnO2 annealed at 220° C has the highest intensities of the planes (110) and (220).

Furthermore, from the variation of FWHM (full width at half maximum) in the plane (110), extracted from the X-ray diffraction patterns which are illustrated in Table 1, we can notice a low value of FWHM obtained for this same annealing condition (220 °C), suggesting an increase of the crystal sizes [40]. This increase in crystal size can be deduced from the Debye–Scherrer equation [41] and the variations of the perovskite crystal size obtained after different annealing temperatures of SnO2 layer are summarized in Table 1.

$$L\, = \,K \, \lambda /\beta \, \cos \theta$$
(1)

where L is the average crystallite size, λ is the X-ray wavelength in nanometer (nm), β is the peak width of the diffraction peak profile at half maximum height resulting from small crystallite size in radians and K is a constant related to crystallite shape, normally taken as 0.9.

Table 1 Variation of (110) peaks Full Width at Half Maximum (FWHM) and crystal size with the annealing temperature of SnO2

Another striking fact pointing at the 220 °C condition is the ratio of PbI2 relative to the amount of MAPI for the different annealing temperatures, with a much larger amount of PbI2 for the 220 °C condition (Fig. 3) with a ratio of 0.006, 0.022 and 0.003 for 200° C, 220° C and 240° C, respectively. The beneficial effect of PbI2 as defect passivation has already been mentioned in the literature [42].

Fig. 3
figure 3

Variation of intensity ratios of PbI2/MAPI with the annealing temperature of SnO2

3.1.2 Optical band gap and photoluminescence

In order to investigate the annealing effect of the electron transport layer (SnO2) on the optical properties of MAPbI3-xClx films, we have used UV–visible absorption spectra on the perovskite so-called ‘half-cells’ (glass/ITO/ETL/Perovskite). Figure 4 shows the UV–visible absorption spectra of the SnO2/perovskite films with different annealing temperatures of the ETL layer.

Fig. 4
figure 4

Absorption spectra of MAPbI3-xClxwith different annealing temperatures of SnO2 layers (200 °C, 220 °C and 240 °C)

The UV–Vis spectra of all samples show a typical MAPbI3-xClx absorbance, with an onset absorption at 740 nm [13, 34, 43].

The absorption of the MAPbI3-xClx perovskite film at 740 nm peak after various SnO2 layer annealing temperatures was found to be unchanged. Nevertheless, a small additional absorption from 300 to 500 nm was observed for the SnO2 film annealed at 220 °C. This additional absorption may be due to PbI2 which is in accordance with the previous XRD measurement.

Steady-state photoluminescence (PL) spectroscopy was then conducted to evaluate the charge transfer efficiency between perovskite absorber and the SnO2 ETLs. Figure 5 shows the PL spectra of ITO/SnO2/MAPbI3-xClx films. In these experiments, all samples exhibit an emission peak at 760 nm (radiative recombination from the valence band to conduction band of perovskite), obtained after excitation at 500 nm and originating from MAPbI3 [44].

Fig. 5
figure 5

Steady-state photoluminescence (PL) spectra of ITO /SnO2/MAPbI3-xClx substrates under different annealing temperatures of SnO2 layer

When looking at the three annealing conditions, it is observed that the variation of the thermal annealing of SnO2 (200, 220 and 240 °C) has induced a change in photoluminescence intensity without a change in the emission peak position. This shows that the gap remains unchanged for the three temperatures. However, a strong quenching of perovskite emission is clearly observed with annealing treatment of the SnO2 layer at 220 °C.

This decrease in the PL intensity indicates the good transfer of charges and the fast transportation of electrons from perovskite films to SnO2 layers [45]. Moreover, this PL quenching efficiency is proof of the high crystallinity of the perovskite film, suggesting an efficient electron transfer from perovskite to the ETL.

3.1.3 Photovoltaic performance

Figure 6 shows the photocurrent density–voltage (JV) characteristics of the best PSCs with the configuration ITO/SnO2/MAPbI3-xClx /P3HT/Au, elaborated from different annealing temperatures of SnO2 and measured under AM 1.5 G irradiation (100 mW.cm−2). The photovoltaic parameters of the devices are summarized in Table 2. The devices prepared with SnO2 layer and annealed at 220 °C show the highest photovoltaic performance with an open-circuit voltage (VOC) of 0.97 V, a short-circuit current (JSC) of 20.33 mA/cm2, a fill factor (FF) of 61% and a power conversion efficiency (PCE) of 12.1%. All these parameters are significantly better than those of PSC with SnO2 layer, annealed at 200 °C (JSC  = 19.41 mA/cm2, VOC  = 0.89 V, FF = 55%, PCE = 10.3%) and 240 °C (JSC  = 19.45 mA/cm2, VOC  = 0.90 V, FF = 62%, PCE = 11.1%).

Fig. 6
figure 6

JV characteristics of the best PSCs fabricated from different annealing temperatures of SnO2

Table 2 Photovoltaic parameters extracted from the champion and average solar cells fabricated from different annealing temperatures of SnO2

The device with the SnO2 layer annealed at 200 °C has the worst curve with a low FF. This poor FF is linked to low shunt resistance and high series resistance. The high PL obtained for such condition, indicating the poor charge transfer from perovskite to ETL, which could be linked to this higher series resistance.

The enhancement of the JSC for annealing at 220 °C can be attributed to the slight enhancement of the absorption around 490 nm as observed previously in Fig. 4.

The improvement of FF and VOC at 220 °C could be attributed to the improved quality of the perovskite film and the low electron agglomeration in the SnO2 layer, which greatly reduces the recombination rate. Thus, the annealing temperature of SnO2 at 220 °C appears to be an optimum for the operation of the cells with the highest absorption, which leads to the best JSC.

After fixing the optimum annealing temperature of the ETL layer at 220° C, we have moved to study the influence of the HTL layer on the performance of the perovskite solar cells. A series of MAPbI3-xClx perovskites solar cells were elaborated with different HTL layers: PTAA and P3HT.

3.2 Effect of the nature of HTL

3.2.1 Ultraviolet–Visible Absorption

To further investigate the effect of the HTL layer, we performed UV–visible absorption spectroscopy. As shown in Fig. 7, all films show a typical absorbance spectrum of MAPbI3-xClx. The absorption spectrum of the P3HT deposited on the ITO/SnO2/ PVK substrate shows an additional absorption, in the zone ranging from 450 to 650 nm, which is due to the absorption of the P3HT material (Fig. 7b [46]).

Fig. 7
figure 7

a Absorption spectra of the ITO/SnO2 / MAPbI3-xClx/P3HT and ITO/SnO2/MAPbI3-xClx/PTAA structures. b Absorption spectrum of P3HT and absorption spectrum of PTAA

3.2.2 Photoluminescence

To further elucidate the process of the photo-generated electron–hole separation and injection by using different HTL layers, steady-state photoluminescence spectra were performed and are displayed in Fig. 8, which shows the PL spectra of ITO/SnO2/PK, ITO/SnO2/PVK/P3HT and ITO/SnO2/PVK/PTAA structure.

Fig. 8
figure 8

Photoluminescence spectra for the two HTL deposited on ITO/SnO2/MAPbI3-xClx substrate

We note a blue shift for the ITO/SnO2/PVK/P3HT film and the presence of a second peak with a low-intensity emission around 650 nm. The 650 nm contribution is attributed to the P3HT emission because the P3HT can also absorb at 500 nm and it emits between 600 and 800 nm. The blue-shifted emission at 725 nm could be explained by a decrease in the crystallinity of the film that was well confirmed by the measure XRD (Fig. 1 information support).

The sample-based on PTAA shows a higher degree of PL quenching than the P3HT, indicating a good separation efficiency of charge carriers and an increase in the crystallinity [34].

A quenching of perovskite PL emission was observed after the deposition of the HTL layer (P3HT or PTAA), with a quenching efficiency η of 64.84% and 68.90% for the P3HT and PTAA, respectively.

Furthermore, these results are explained by the difference in the energy level of the highest occupied molecular orbital between CH3NH3PbI3-xClx and PTAA, which was lower than that with P3HT as shown in Fig. 1. Further, the exciton at the CH3NH3PbI3-xClx/P3HT interface could be easily recombined that due to the poor electronic conductivity of P3HT, which is a consequence of the higher barrier holes collection of P3HT [34].

3.2.3 Photovoltaic performance

In order to investigate the charge transport efficiency of the cells, JV characteristics of the best PSCs with different HTL layers (P3HT versus PTAA) were investigated. The photocurrent–voltage (JV) curves measured under simulated AM 1.5 sunlight conditions are shown in Fig. 9. The corresponding photovoltaic parameters are listed in Table 3. The high efficiency of 13.25% was obtained using the PTAA as HTL layer, with an open-circuit voltage (VOC) of 0.89 V, a current density (JSC) of 22.87 mA/cm2, a fill factor (FF) of 64%, and a series resistance of 4.28 Ω.cm2. The improvement of FF and VOC could be attributed to the enhancement of the perovskite film quality and low interface loss [47]. These results are in good agreement with the PL spectra and XRD measurements (Fig. 1 support information).

Fig. 9
figure 9

Current–voltage density characteristics of the best NIP cells under illumination for different HTL materials

Table 3 Photovoltaic parameters extracted from the average solar cells fabricated from different HTL materials

The performance parameters of the device based on CH3NH3PbI3-Clx/PTAA have also the highest JSC values than those of the device using P3HT. This increase of the short-circuit current is due to the elimination of the recombination pathways owing to an of perovskite film with good crystallinity [48].

Furthermore, in order to understand both enhanced JSC and FF, the series resistance (RS) and the shunt resistance (RSh) are directly extracted from the J–V curve using the reverse of slope of JV curve [49,50,51].

The PSCs based on PTAA have lower series resistances, while higher shunt resistance was obtained due to the reduced charge injection barrier between the perovskite and gold electrode [52].

3.3 Hysteresis effect

Figure 10 shows the JV plots of the optimum standard cell with the structure: ITO/SnO2/MAPbI3-xClx/PTAA/Au, in forward and reverse scan directions and the extracted parameters are summarized in Table 4. This cell achieves a PCE of 13.3% with a VOC of 0.89 V, a JSC of 22.87 mA/cm2, and a FF of 64.57% under forwarding voltage scan. When the cell is measured in the reverse forward, a slight decrease in yield to 12.5% has been achieved with a VOC of 0.92 V, a JSC of 21.94 mA/cm2, and an FF of 61.90%. A hysteresis behavior observed was the consequence of different physical effects. It can be probably due to the nature of the interfaces between the perovskite layer and the charge transport layers (ETL and HTL). Moreover, it can be attributed to ion accumulation [53, 54].

Fig. 10
figure 10

JV plots of the best solar cell with forward and reverse scan

Table 4 Photovoltaic performance of PSC based on MAPbI3-xClxwith forward and reverse scan

4 Conclusions

Due to outstanding material properties, organic–inorganic halide perovskites have attracted tremendous attention in the photovoltaic field. In this contribution, MAPbI3-xClx perovskite solar cells using SnO2 as an electron transport layer were elaborated. The effect of the annealing temperature of the SnO2 layer was investigated. We reveal that the best annealing treatment of the SnO2 is 220 °C, resulting in a highly efficient carrier’s transportation with fewer defects at the interface SnO2/perovskite. We have also demonstrated that the performance of the device is also affected by the energy level of the HTL. The results show that the high PCE of the devices is obtained using the PTAA as an HTL, corresponding to the good charge transfer to HTL. With this approach, device efficiency could be raised from 10 to 13%.