Enhancing Perovskite Solar Cell Performance through Surface Engineering of Metal Oxide Electron-Transporting Layer

: Perovskite solar cells have gained increasing interest in recent times owing to the rapidly enlarged device e ﬃ ciency and tunable optoelectronic properties in various applications. In perovskite solar cells, interface engineering plays an important role in determining the ﬁnal device e ﬃ ciency and stability. In this study, we adopted TiCl 4 treatment to reduce the surface roughness of the metal oxide layer and improve the perovskite ﬁlm quality to obtain better device performance. After proper TiCl 4 treatment, the e ﬃ ciencies of TiCl 4 –TiO 2 - and TiCl 4 –ZnO-based devices were signiﬁcantly enhanced up to 16.5% and 17.0%, respectively, compared with those based on pristine TiO 2 and ZnO (13.2% and 10.2%, respectively).

In this study, to investigate the mechanism of TiCl 4 and the role it plays with TiO 2 and phenyl-C61-butyric acid methyl ester (PCBM) in PSCs, we investigated the effects of TiCl 4 treatment on the quality of perovskite thin film and the corresponding device performance. Results showed that, after TiCl 4 treatment, the perovskite film quality was significantly enhanced; moreover, the charge transfer and extraction became more efficient. Finally, compared to pristine TiO 2 -and ZnO-based devices (13.2% and 10.2%), the efficiencies of both TiCl 4 -TiO 2 -and TiCl 4 -ZnO-based devices were significantly improved up to 16.5% and 17.0%, respectively, with simultaneously enhanced open-circuit voltage (V oc ), short-circuit current density (J sc ), and fill factor (FF).

Results and Discussion
Surface morphologies of ETLs with and without TiCl 4 treatment were investigated by tapping-mode atomic force microscopy (AFM) (Dimension Icon AFM, Bruker, Billerica, MA, USA). The surface roughness of the TiO 2 surface decreased from 18.0 to 13.9 nm after TiCl 4 treatment ( Figure 1). After depositing the PCBM layer, the surface roughness was maintained without much change. This indicates that the surface roughness is mainly determined by TiCl 4 treatments. These properties may affect perovskite crystallization and formation as well as the crystal size. The energy levels of TiO 2 /PCBM electron-transporting layers with and without TiCl 4 treatment were investigated by ultraviolet photoelectron spectroscopy (UPS) (Escalab 250Xi, Thermo Scientific, Waltham, MA, USA). In Figure 2, it can be seen that the TiO 2 /PCBM work function decreased from 4.18 to 4.09 eV after TiCl 4 treatment. The decreased work function would result in better energy level alignment and more efficient electron transfer between the conduction band of perovskite and the Fermi level of ETL. Moreover, the potential difference between two electrode contacts, such as TiO 2 and spiro-OMeTAD, was also enlarged due to the decreased work function, hence increasing the V oc of the device [33].
In order to illustrate the effect of TiCl 4 treatment on the properties of perovskite thin film, the thin film morphologies of the perovskites upon ETLs with and without TiCl 4 treatment were investigated. As shown in Figure 3, the perovskite crystal size was enhanced, indicating a favorable crystal growth process on the TiCl 4 -treated surface, which may have been caused by the low surface roughness value. The improved crystalline quality was also confirmed by the XRD patterns. As shown in Figure 4, strong diffraction intensity at around 14.6 • , 28.9 • , and 32.3 • were observed and assigned to (110), (220), and (310) crystal planes of tetragonal perovskite phase, respectively. The diffraction pattern at 13.1 • originated from PbI 2 because of the excess PbI 2 over CH 3 NH 3 I. These results are consistent with those previously reported [8,9]. Moreover, the diffraction intensity of the perovskite thin films with additional TiCl 4 treatment exhibited slight enhancement. This indicates that the thin film crystallinity was increased, resulting in more efficient charge transport and collection.   The film absorption spectra of perovskite films deposited on TiO 2 and TiCl 4 -treated ETLs were investigated by UV-Vis spectroscopy. Both films exhibited high absorption intensity over the entire UV-visible range, and the absorption band edge was around 1.61 eV ( Figure 5). Compared with pristine TiO 2 ETL, the absorption of the perovskite film based on TiCl 4 -treated TiO 2 ETL was significantly enhanced, indicating enhanced crystallinity of the resulting thin film.
To further investigate the photophysical properties of the thin films, steady-state photoluminescence (PL) and time-resolved PL (TR-PL) measurements of perovskite thin films were further investigated. The perovskite film based on TiCl 4 -treated TiO 2 ETL showed larger PL intensity decay compared to the pristine TiO 2 -based perovskite film (Figure 6), indicating a faster charge transfer process between TiO 2 and perovskite thin films [30]. Furthermore, the PL lifetimes of perovskite films on glass were also measured ( Figure 7). By fitting the TR-PL curves with biexponential function, two decay parts-fast decay and slow decay-could be obtained [14]. The average lifetimes derived from the fitting curves were around 25.7 and 29.8 ns for perovskite films with and without TiCl 4 treatment, respectively. The decreased PL lifetime is related to the efficient charge carrier transfer induced quenching process [27]. This is essential for efficient charge extraction and collection of the device.  The effect of TiCl 4 treatment on PSC device performance was studied based on a planar structure with a configuration of FTO/TiO 2 /PC 60 BM/CH 3 NH 3 PbI 3−x Cl x /spiro-OMeTAD/Ag. Figure 8 exhibits the current density − voltage (J − V) curves of PSC devices with and without TiCl 4 treatment, and Table 1 summarizes the corresponding device parameters. In Figure 8, a PCE of 13.2% (V oc of 1.04 V, J sc of 19.1 mA/cm 2 , and FF of 0.66) can be seen for the device with pristine TiO 2 . By comparison, all the parameters, i.e., V oc , J sc , and FF, for the device with TiCl 4 treatment were enhanced simultaneously, resulting in an improved average PCE of 16.5% (V oc of 1.08 V, J sc of 22.4 mA/cm 2 , and FF of 0.68). The device series resistance (R s ) and shunt resistance (R sh ) were also calculated to further understand the improvement in performance. Compared to the device without TiCl 4 treatment, the R s value of the device with TiCl 4 treatment decreased to 3.7 Ω·cm 2 , while the R sh increased to 6.9 kΩ·cm 2 . The decreased R s and increased R sh are responsible for the J sc and FF enhancement [7]. It needs to be mentioned that, when the high-temperature TiO 2 was replaced with the low-temperature ZnO layer, the device also showed similar improvement. The ZnO with TiCl 4 treatment revealed an improved PCE of 17.0% with a V oc of 1.08 V, J sc of 23.2 mA/cm 2 , and FF of 0.67, which was much higher than those of the device with pristine ZnO.   The device hysteresis behavior is also a significant parameter to determine the final performance of the device. It can be affected by many factors, such as carrier trapping, ion motion, charge accumulation at interfaces, and so on [45][46][47][48]. The device hysteresis behaviors under different scan directions were measured (Figure 9). The device did not exhibit serious hysteresis behavior before and after TiCl 4 treatment, indicating that TiCl 4 has less effect than PCBM, which plays a major role in determining the final hysteresis behavior.
In order to further investigate the mechanism involved in performance enhancement, transient photocurrent and photovoltage measurements of perovskite solar cells were performed. Figure 10a shows the transient photocurrent decay of perovskite devices measured at short-circuit condition. After TiCl 4 treatment, the devices exhibited faster decay with shorter lifetime (1.02 µs) compared to the pristine TiO 2 -based device (1.78 µs). This indicates that the device with TiCl 4 treatment possessed more efficient charge transfer process. The transient photovoltage was used to determine the charge recombination process (Figure 10b). It can be clearly seen that the device with TiCl 4 treatment exhibited much longer lifetime (2.32 ms) compared to the pristine TiO 2 -based device (1.66 ms). This indicates that the charge recombination process was efficiently suppressed. Hence, the J sc and FF exhibited significant improvement.

Conclusions
In conclusion, we investigated the effect of TiCl 4 treatment on perovskite thin film formation and PSC device performance. The results showed that TiCl 4 treatment had a beneficial effect on the properties of perovskite thin film. It could enhance thin film crystallinity and improve charge transfer and extraction. Finally, the PCE was enhanced from 13.2% to 16.5% for devices based on TiO 2 ETL and from 10.2% to 17.0% for devices based on ZnO ETL. This is important for understanding interfacial treatment and further improves device efficiency and stability.