Conjugated polyelectrolyte hole transport layer for inverted-type perovskite solar cells

Organic–inorganic hybrid perovskite materials offer the potential for realization of low-cost and flexible next-generation solar cells fabricated by low-temperature solution processing. Although efficiencies of perovskite solar cells have dramatically improved up to 19% within the past 5 years, there is still considerable room for further improvement in device efficiency and stability through development of novel materials and device architectures. Here we demonstrate that inverted-type perovskite solar cells with pH-neutral and low-temperature solution-processable conjugated polyelectrolyte as the hole transport layer (instead of acidic PEDOT:PSS) exhibit a device efficiency of over 12% and improved device stability in air. As an alternative to PEDOT:PSS, this work is the first report on the use of an organic hole transport material that enables the formation of uniform perovskite films with complete surface coverage and the demonstration of efficient, stable perovskite/fullerene planar heterojunction solar cells.

. Real images of perovskite films coated on PEDOT:PSS and CPE-K illustrating visible degradation as a function of air exposure time. Average temperature and humidity were 20 ± 3 °C and 40 ± 10% for air stability, respectively.

Preparation of perovskite precursor solution.
Lead chloride (PbCl 2 ) was purchased from Sigma-Aldrich and used without purification.
Methylammonium iodide (MAI) was synthesized using synthetic routes in previous literature 1 . MAI and PbCl 2 with molar ratio of 3:1 were dissolved in N,N-dimethylformamide (DMF) at concentration of 40 wt.% and this solution was stirred at 60 °C for 6h in nitrogenfilled glovebox.

Film preparation and characterization.
We prepared perovskite films on different substrates for various measurements by using same procedures that were used for optimum devices. SEM images were obtained using FEI XL40 Sirion FEG digital scanning microscope. XRD measurements were carried out using a Bruker, D8 ADVANCE at a scan rate of 2.4° min -1 . UV-vis absorption was measured using a OLIS 14 spectrophotometer. AFM images were obtained using a Asylum MFP-3D standard system AFM microscope in a tapping mode. Contact angle measurements were carried out using DSA 100 (KRUSS, Germany).

Supplementary figure and description
We employed four conjugated polyelectrolytes (CPEs) with different anionic polymer backbones and counter ions. Chemical structures and a variety properties of these CPEs were listed in Supplementary Table 1. To compare the effect of different CPEs on device performance, we fabricated ipero-SCs using CPEs as the HTL. Supplementary Fig. 1a and   Fig. 1b).
Supplementary Fig. 2 shows absorption spectra of PEDOT:PSS and CPE-K before and after washing them with DMF. DMF is a solvent for dissolving two perovskite precursor materials, MAI and PbCl 2 . Absorption spectrum of PEDOT:PSS was unchanged after washing with DMF, whereas we observed 30% decrease in optical density (OD) of CPE-K.
Although washing process with DMF slightly removed CPE-K film, this layer was still existed on ITO substrate without complete washing out. Absorption spectra of perovskite film on PEDOT:PSS and CPE-K were shown in Supplementary Fig. 3. The film on CPE-K exhibited slightly higher OD than that of the film on PEDOT:PSS owing to higher absorption of CPE-K in visible wavelength region. Absorption difference between perovskite films on PEDOT:PSS and CPE-K (Inset of Supplementary Fig. 3) was consistent with absorption difference between PEDOT:PSS and CPE-K (Supplementary Fig. 2).
To investigate the influence of different substrates on perovskite film morphology, we first performed AFM measurements for PEDOT:PSS and CPE-K. In spite of small differences in topography and phase images, both films exhibited smooth surface with rootmean-square (rms) roughness of 1.0 nm (Supplementary Fig. 4). However, different morphology was clearly seen in perovskite films spin-coated on PEDOT:PSS and CPE-K ( Supplementary Fig. 5). Perovskite film on CPE-K was uniform with complete surface coverage and rms roughness of 14.7 nm (Supplementary Fig. 5a), whereas film on PEDOT:PSS exhibited uneven surface with rms roughness of 15.6 nm and large number of voids between crystal boundaries (Supplementary Fig. 5b). This implies that CPE-K results in more uniform perovskite film with higher surface coverage than PEDOT:PSS. We also studied surface energy of PEDOT:PSS and CPE-K by performing contact angle measurements. Both film exhibited extremely low contact angles to DMF below 3° ( Supplementary Fig. 6) (Fig. S8a), whereas the devices with CPE-K exhibited higher J SC and FF than those of the devices with PEDOT:PSS (Supplementary Fig. 8b and 8c). As a result, the device with CPE-K yielded higher average PCE (11.20%) than that of the device with PEDOT:PSS (9.37%). This high device efficiency was attributed to improved perovskite film morphology ( Fig. 2) and efficient hole transport from perovskite to ITO anode (Fig. 3) by CPE-K HTL.
Supplementary Fig. 9 presents visible degradation of perovskite films on PEDOT:PSS and CPE-K as a function of air exposure time. We tested film stability in ambient air condition with average temperature of 20±3° and humidity of 40±10%. We observed the same dark-brown films on both PEDOT:PSS and CPE-K. After 24 h, although visible degradation occurred in the film on CPE-K, degradation rate of film on PEDOT:PSS was much faster than that of film on CPE-K. After 180 h, perovskite film on CPE-K still maintained dark brown color, whereas film on PEDOT:PSS was completely changed to yellow solid, indicating decomposition of perovskite phase into PbI 2 by the presence of water 3 . These results reveal that acidic nature of PEDOT:PSS has bad influence on perovskite film stability and thus is detrimental for device stability in air condition (Fig. 5d).