High Performance Planar Structure Perovskite Solar Cells Using a Solvent Dripping Treatment on Hole Transporting Layer

: Perovskite solar cell e ﬃ ciency is not only related with material properties, but also a ﬀ ected by the interface engineering that used in perovskite solar cells. The perovskite ﬁlm / electrode interface properties play important roles in charge carrier extraction, transport, and recombination. To achieve better interface contact for the device operation, proper interlayers or surface treatment should be applied. In this study, we applied a poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) interlayer with a solvent / solution washing treatment as the hole transport layer. It showed that by the solvent / solution treatment, the PEDOT:PSS ﬁlm conductivity was signiﬁcantly enhanced, and hence, the charge carrier transfer e ﬃ ciency was e ﬃ ciently improved, and the device short-circuit current density was enlarged. Finally, the device e ﬃ ciency signiﬁcantly increased from 14.8% to 16.2%.


Results and Discussion
By spin coating the PEDOT:PSS precursors with different coating speeds, the pristine PEDOT:PSS films with different thicknesses were formed. It was surprised that the PEDOT:PSS thickness has a significant effect on the final device performance. From Figure 1 and Table 1, it can be found that the device open circuit voltage (V oc ) and J sc significantly increased upon increasing the coating speed of PEDOT:PSS HTL. Hence, the efficiency of solar cell devices was enhanced. It is clear to see that the J sc gradually increased upon increasing the coating speed, indicating that the enhanced shunt resistance/series resistance (R sh /R s ) ratio and/or decreased device resistance upon decreasing the PEDOT:PSS layer thickness (Table 1). To further increase the device performance, enhancing the PEDOT:PSS layer conductivity is necessary, and moreover, to enable a large thickness range of the interlayer is beneficial for printed devices.
Coatings 2019, 9, x FOR PEER REVIEW 2 of 10 hydrophobic properties or using inorganic NiO HTL to replace PEDOT:PSS film become more important [23,25,26,[31][32][33][34]. Previously, in organic solar cells and perovskite solar cells, some polar solvents including ethylene glycol (EG), dimethyl sulfoxide (DMSO), dopamine, F4-TCNQ, dimethyl formamide (DMF), imidazole, etc. have been used as additives to enhance the PEDOT:PSS conductivity and tune the work-function [35][36][37][38]. Among them, DMSO exhibits superior advantages due to its good dielectric constant, dipolar moment, and relatively higher boiling point. Hence, between PEDOT:PSS and DMSO molecules, strong dipole-dipole or dipolar-charge interactions can be easily formed, which can enhance the film conductivity and charge carrier mobility. Previously, some studies have been performed to investigate the PEDOT:PSS work-function and morphology effects on the perovskite thin film morphologies and corresponding device performance [32,33,39,40]. Even though the DMSO has been widely used in perovskite solution precursor or as an additive in PEDOT:PSS solution, there are few researches about direct solvent treatment on PEDOT:PSS film to optimize film quality and device performance.
In this study, we studied the solvent/solution dripping effects on the PEDOT:PSS and perovskite film properties. By employing the solvent/solution washing treated PEDOT:PSS as HTL, high performance perovskite solar cell devices were achieved. The PCE was enhanced to 16.2% which was much higher than that of the device with pristine PEDOT:PSS film (14.8%). After a series of characterization measurements, it was found that the significantly increased device short circuit current density (Jsc) was due to the improved conductivity of the PEDOT:PSS film.

Results and Discussion
By spin coating the PEDOT:PSS precursors with different coating speeds, the pristine PEDOT:PSS films with different thicknesses were formed. It was surprised that the PEDOT:PSS thickness has a significant effect on the final device performance. From Figure 1 and Table 1, it can be found that the device open circuit voltage (Voc) and Jsc significantly increased upon increasing the coating speed of PEDOT:PSS HTL. Hence, the efficiency of solar cell devices was enhanced. It is clear to see that the Jsc gradually increased upon increasing the coating speed, indicating that the enhanced shunt resistance/series resistance (Rsh/Rs) ratio and/or decreased device resistance upon decreasing the PEDOT:PSS layer thickness (Table 1). To further increase the device performance, enhancing the PEDOT:PSS layer conductivity is necessary, and moreover, to enable a large thickness range of the interlayer is beneficial for printed devices.    To improve the conductivity of PEDOT:PSS HTLs, a DMSO solvent treatment was taken during the PEDOT:PSS spin coating step. For the as-prepared PEDOT:PSS film (~30 nm) without solvent washing treatment, the conductivity was about 1.7 × 10 −3 S cm −1 [34]. For the PEDOT:PSS film with DMSO solvent washing treatment, the conductivity increased to 0.7 S cm −1 , which is similar with the literature reported that the polar organic solvents could improve the PEDOT:PSS conductivity [32]. The conductivity enhancement mechanism has been reported previously [32]. Some PSS chains could be washed away from PEDOT:PSS films by DMSO solvent. Therefore, the ratio of PEDOT was increased, leading to higher conductivity.
With the DMSO solvent washing treatment, the PSS chains could be removed from PEDOT:PSS, which could be confirmed by XPS spectra (Escalab 250Xi, Thermo Scientific, Waltham, MA, USA). The S 2p has two states: one is from S in PSS (166-172 eV), and another one is from S atoms in PEDOT (162-166 eV). The ratio of S sp XPS intensity in PEDOT to that in PSS is significantly enhanced after treating with DMSO solvent (Figure 2) [41]. The relatively increased PEDOT ratio is contributed to the increased conductivity.  To improve the conductivity of PEDOT:PSS HTLs, a DMSO solvent treatment was taken during the PEDOT:PSS spin coating step. For the as-prepared PEDOT:PSS film (~30 nm) without solvent washing treatment, the conductivity was about 1.7 × 10 −3 S cm −1 [34]. For the PEDOT:PSS film with DMSO solvent washing treatment, the conductivity increased to 0.7 S cm −1 , which is similar with the literature reported that the polar organic solvents could improve the PEDOT:PSS conductivity [32]. The conductivity enhancement mechanism has been reported previously [32]. Some PSS chains could be washed away from PEDOT:PSS films by DMSO solvent. Therefore, the ratio of PEDOT was increased, leading to higher conductivity.
With the DMSO solvent washing treatment, the PSS chains could be removed from PEDOT:PSS, which could be confirmed by XPS spectra (Escalab 250Xi, Thermo Scientific, Waltham, MA, USA). The S 2p has two states: one is from S in PSS (166-172 eV), and another one is from S atoms in PEDOT (162-166 eV). The ratio of S sp XPS intensity in PEDOT to that in PSS is significantly enhanced after treating with DMSO solvent (Figure 2) [41]. The relatively increased PEDOT ratio is contributed to the increased conductivity. The effect of PSS chain removal from PEDOT:PSS on PEDOT:PSS film surface morphologies was investigated by atomic force microscopy (AFM) techniques (Dimension Icon AFM, Bruker, Billerica, MA, USA). Figure 3 shows the topography images of a 2 × 2 µm 2 area of pristine and DMSO solvent treated PEDOT:PSS films. The root mean square (RMS) roughness of pristine PEDOT:PSS thin film was 1.21 nm, indicating a quite smooth surface. After DMSO solvent washing treatment, crystal grains became clearer and the film surface became rougher, resulting in an increased root mean square (RMS) roughness of 2.92 nm. The increased RMS is mainly induced by the PSS chains removal and PEDOT chains aggregation. The increased conducting PEDOT-rich regions lead to superior charge-transport pathways for the treated films. Thus, the PEDOT:PSS film with solvent washing treatment exhibited higher conductivity.  Figure 3 shows the topography images of a 2 × 2 µm 2 area of pristine and DMSO solvent treated PEDOT:PSS films. The root mean square (RMS) roughness of pristine PEDOT:PSS thin film was 1.21 nm, indicating a quite smooth surface. After DMSO solvent washing treatment, crystal grains became clearer and the film surface became rougher, resulting in an increased root mean square (RMS) roughness of 2.92 nm. The increased RMS is mainly induced by the PSS chains removal and PEDOT chains aggregation. The increased conducting PEDOT-rich regions lead to superior charge-transport pathways for the treated films. Thus, the PEDOT:PSS film with solvent washing treatment exhibited higher conductivity.  The perovskite thin films deposited on pristine and DMSO solvent washing treated PEDOT:PSS were studied by scanning electron microscopy (SEM, JSM-7800F SEM, Japan). As exhibited in Figure  4, the perovskite film based on solvent washing treated PEDOT:PSS had similar crystal morphology with that of the perovskite film based on pristine PEDOT:PSS. All the perovskite films are uniform with large regular crystallites, and they fully cover the underlying layer. The favorable thin film uniformity and coverage indicate a homogenous nucleation of the perovskites on the substrates. The crystal size increased possibly due to the relatively hydrophobic nature of solvent washing treated PEDOT:PSS, which is beneficial for device performance [42]. In order to illustrate the effect of solvent treatment, we have measured the X-ray diffraction (XRD, Bruker D8 Advance XRD instrument, Billerica, MA, USA) and UV-Vis of the perovskites without or with solvent treatment. Figure 5a presents the UV-vis absorption spectra of perovskites and showed that the film with solvent washing treated PEDOT:PSS exhibited higher light absorption than pristine film. The thin film crystallinity was investigated by XRD, and the XRD patterns of perovskite films with/without solvent treatment are shown in Figure 5b. The diffraction peaks around 14.21°, 28.51°, and 31.88° are assigned to the (110), (220), and (310) crystal planes, respectively, which indicates that the perovskite film is formed with good quality. Meanwhile, the PbI2 peak around 12.6° appeared, and the exceed PbI2 could efficiently passivate the perovskite grain boundaries and reduce trap density as reported [16,43]. Moreover, the solvent washing treated PEDOT:PSS produced higher crystallinity of treated film than that of pristine film. This improved film crystallinity has positive effect on charge transport and charge extraction.  The perovskite thin films deposited on pristine and DMSO solvent washing treated PEDOT:PSS were studied by scanning electron microscopy (SEM, JSM-7800F SEM, Japan). As exhibited in Figure 4, the perovskite film based on solvent washing treated PEDOT:PSS had similar crystal morphology with that of the perovskite film based on pristine PEDOT:PSS. All the perovskite films are uniform with large regular crystallites, and they fully cover the underlying layer. The favorable thin film uniformity and coverage indicate a homogenous nucleation of the perovskites on the substrates. The crystal size increased possibly due to the relatively hydrophobic nature of solvent washing treated PEDOT:PSS, which is beneficial for device performance [42]. In order to illustrate the effect of solvent treatment, we have measured the X-ray diffraction (XRD, Bruker D8 Advance XRD instrument, Billerica, MA, USA) and UV-Vis of the perovskites without or with solvent treatment. Figure 5a presents the UV-vis absorption spectra of perovskites and showed that the film with solvent washing treated PEDOT:PSS exhibited higher light absorption than pristine film. The thin film crystallinity was investigated by XRD, and the XRD patterns of perovskite films with/without solvent treatment are shown in Figure 5b. The diffraction peaks around 14.21 • , 28.51 • , and 31.88 • are assigned to the (110), (220), and (310) crystal planes, respectively, which indicates that the perovskite film is formed with good quality. Meanwhile, the PbI 2 peak around 12.6 • appeared, and the exceed PbI 2 could efficiently passivate the perovskite grain boundaries and reduce trap density as reported [16,43]. Moreover, the solvent washing treated PEDOT:PSS produced higher crystallinity of treated film than that of pristine film. This improved film crystallinity has positive effect on charge transport and charge extraction.  The perovskite thin films deposited on pristine and DMSO solvent washing treated PEDOT:PSS were studied by scanning electron microscopy (SEM, JSM-7800F SEM, Japan). As exhibited in Figure  4, the perovskite film based on solvent washing treated PEDOT:PSS had similar crystal morphology with that of the perovskite film based on pristine PEDOT:PSS. All the perovskite films are uniform with large regular crystallites, and they fully cover the underlying layer. The favorable thin film uniformity and coverage indicate a homogenous nucleation of the perovskites on the substrates. The crystal size increased possibly due to the relatively hydrophobic nature of solvent washing treated PEDOT:PSS, which is beneficial for device performance [42]. In order to illustrate the effect of solvent treatment, we have measured the X-ray diffraction (XRD, Bruker D8 Advance XRD instrument, Billerica, MA, USA) and UV-Vis of the perovskites without or with solvent treatment. Figure 5a presents the UV-vis absorption spectra of perovskites and showed that the film with solvent washing treated PEDOT:PSS exhibited higher light absorption than pristine film. The thin film crystallinity was investigated by XRD, and the XRD patterns of perovskite films with/without solvent treatment are shown in Figure 5b. The diffraction peaks around 14.21°, 28.51°, and 31.88° are assigned to the (110), (220), and (310) crystal planes, respectively, which indicates that the perovskite film is formed with good quality. Meanwhile, the PbI2 peak around 12.6° appeared, and the exceed PbI2 could efficiently passivate the perovskite grain boundaries and reduce trap density as reported [16,43]. Moreover, the solvent washing treated PEDOT:PSS produced higher crystallinity of treated film than that of pristine film. This improved film crystallinity has positive effect on charge transport and charge extraction.   The photoluminescence (PL) spectra of perovskite films based on pristine and solvent washing treated PEDOT:PSS were also studied ( Figure 6a). Compared to pristine PEDOT:PSS based perovskite film, the solvent washing treated PEDOT:PSS based perovskite film exhibited more efficient PL quenching, indicating that charge transfer between solvent washing treated PEDOT:PSS and perovskite is more efficient. This is beneficial for efficient charge transfer and charge extraction. Time-resolved PL (TR-PL) measurement was also carried out to investigate the charge carrier dynamic behavior regarding to the solvent washing treatment. It can be seen that the average PL lifetime decreased from 47.4 ns to 36.7 ns ( Table 2). The decreased PL lifetime indicates that an efficient charge transfer induced quenching process occurred based on solvent treatment.  The solvent washing treatment effect on perovskite solar cell performance was also studied as well. The device adopted a configuration of ITO/PEDOT:PSS/Perovskite/PCBM/BCP/Ag, and the perovskite layer was composed by MAI, PbI2, and PbCl2 (The details can be seen in supplementary materials) [44]. Table 3  The photoluminescence (PL) spectra of perovskite films based on pristine and solvent washing treated PEDOT:PSS were also studied (Figure 6a). Compared to pristine PEDOT:PSS based perovskite film, the solvent washing treated PEDOT:PSS based perovskite film exhibited more efficient PL quenching, indicating that charge transfer between solvent washing treated PEDOT:PSS and perovskite is more efficient. This is beneficial for efficient charge transfer and charge extraction. Time-resolved PL (TR-PL) measurement was also carried out to investigate the charge carrier dynamic behavior regarding to the solvent washing treatment. It can be seen that the average PL lifetime decreased from 47.4 ns to 36.7 ns ( Table 2). The decreased PL lifetime indicates that an efficient charge transfer induced quenching process occurred based on solvent treatment. The photoluminescence (PL) spectra of perovskite films based on pristine and solvent washing treated PEDOT:PSS were also studied ( Figure 6a). Compared to pristine PEDOT:PSS based perovskite film, the solvent washing treated PEDOT:PSS based perovskite film exhibited more efficient PL quenching, indicating that charge transfer between solvent washing treated PEDOT:PSS and perovskite is more efficient. This is beneficial for efficient charge transfer and charge extraction. Time-resolved PL (TR-PL) measurement was also carried out to investigate the charge carrier dynamic behavior regarding to the solvent washing treatment. It can be seen that the average PL lifetime decreased from 47.4 ns to 36.7 ns ( Table 2). The decreased PL lifetime indicates that an efficient charge transfer induced quenching process occurred based on solvent treatment.  The solvent washing treatment effect on perovskite solar cell performance was also studied as well. The device adopted a configuration of ITO/PEDOT:PSS/Perovskite/PCBM/BCP/Ag, and the perovskite layer was composed by MAI, PbI2, and PbCl2 (The details can be seen in supplementary materials) [44]. Table 3   The solvent washing treatment effect on perovskite solar cell performance was also studied as well. The device adopted a configuration of ITO/PEDOT:PSS/Perovskite/PCBM/BCP/Ag, and the perovskite layer was composed by MAI, PbI 2 , and PbCl 2 (The details can be seen in Supplementary Materials) [44]. Table 3 summarizes the parameters of the devices, and Figure 7 exhibits the device current density-voltage (J-V) characteristics. The pristine PEDOT:PSS layer exhibited a J sc of 19.8 mA/cm 2 , a V oc of 0.98 V, a fill factor (FF) of 0.76, and an overall PCE of 14.8%. After DMSO solvent washing treatment, the J sc significantly increased to 22.6 mA/cm 2 , V oc had almost no change, and FF slightly decreased to 0.73. Hence, the overall PCE was enhanced to 16.2%. The performance enhancement is mainly resulting from J sc increment. The increased film conductivity and perovskite film absorption all contributed to the J sc enhancement. This can be further confirmed by incident photon-to-electron conversion efficiency (IPCE) measurements. The DMSO solvent washing treated PEDOT:PSS based device showed improvement in IPCE spectra from 350 nm to 750 nm. This method is also applied to other solvent system like DMF. As shown in Table 1, the DMF dripping treatment improved the device PCE to 15.5% with an increased J sc of 20.9 mA/cm 2 .   The device Rs and Rsh were extracted from the J-V curves to further understand the solvent washing treatment effects. As can be seen from Table 3, the device with pristine PEDOT:PSS had a Rs value of 4.2 Ω cm 2 and a Rsh value of 1.5 kΩ cm 2 . After solvent washing treatment, Rs decreased to 2.1 Ω cm 2 , while Rsh also decreased to 1.1 kΩ cm 2 . The decrease in Rs is ascribed to the PEDOT:PSS conductivity enhancement [45]. While decreased Rsh is because of the rougher surface of PEDOT:PSS, which might increase the leakage current.
Hysteresis behaviors of different devices were also performed by measuring the J-V characteristics in both forward and reverse directions, as shown in Figure 8. Compared to pristine PEDOT:PSS based device, solvent washing treated device exhibited slightly larger hysteresis behavior which may be due to its rough surface caused by PEDOT chains aggregation and/or trapping/detrapping of charge carriers due to interface defects [32]. The device R s and R sh were extracted from the J-V curves to further understand the solvent washing treatment effects. As can be seen from Table 3, the device with pristine PEDOT:PSS had a R s value of 4.2 Ω cm 2 and a R sh value of 1.5 kΩ cm 2 . After solvent washing treatment, R s decreased to 2.1 Ω cm 2 , while R sh also decreased to 1.1 kΩ cm 2 . The decrease in R s is ascribed to the PEDOT:PSS conductivity enhancement [45]. While decreased R sh is because of the rougher surface of PEDOT:PSS, which might increase the leakage current.
Hysteresis behaviors of different devices were also performed by measuring the J-V characteristics in both forward and reverse directions, as shown in Figure 8. Compared to pristine PEDOT:PSS based device, solvent washing treated device exhibited slightly larger hysteresis behavior which may be due to its rough surface caused by PEDOT chains aggregation and/or trapping/detrapping of charge carriers due to interface defects [32].
Coatings 2019, 9, x FOR PEER REVIEW 6 of 10 enhancement is mainly resulting from Jsc increment. The increased film conductivity and perovskite film absorption all contributed to the Jsc enhancement. This can be further confirmed by incident photon-to-electron conversion efficiency (IPCE) measurements. The DMSO solvent washing treated PEDOT:PSS based device showed improvement in IPCE spectra from 350 nm to 750 nm. This method is also applied to other solvent system like DMF. As shown in Table 1, the DMF dripping treatment improved the device PCE to 15.5% with an increased Jsc of 20.9 mA/cm 2 .  The device Rs and Rsh were extracted from the J-V curves to further understand the solvent washing treatment effects. As can be seen from Table 3, the device with pristine PEDOT:PSS had a Rs value of 4.2 Ω cm 2 and a Rsh value of 1.5 kΩ cm 2 . After solvent washing treatment, Rs decreased to 2.1 Ω cm 2 , while Rsh also decreased to 1.1 kΩ cm 2 . The decrease in Rs is ascribed to the PEDOT:PSS conductivity enhancement [45]. While decreased Rsh is because of the rougher surface of PEDOT:PSS, which might increase the leakage current.
Hysteresis behaviors of different devices were also performed by measuring the J-V characteristics in both forward and reverse directions, as shown in Figure 8. Compared to pristine PEDOT:PSS based device, solvent washing treated device exhibited slightly larger hysteresis behavior which may be due to its rough surface caused by PEDOT chains aggregation and/or trapping/detrapping of charge carriers due to interface defects [32].