Electrochimica Acta 443 (2023) 141935 Contents lists available at ScienceDirect Electrochimica Acta journal homepage: www.journals.elsevier.com/electrochimica-acta Reduced hysteresis and enhanced air stability of low-temperature processed carbon-based perovskite solar cells by surface modification Soe Ko Ko Aung a, b, Anuja Vijayan b, Masoud Karimipour b, c, d, Tosawat Seetawan a, Gerrit Boschloo b, * a Department of Physics, Faculty of Science and Technology, Optic Research Laboratory, Center of Excellence on Alternative Energy, Research and Development Institute, Sakon Nakhon Rajabhat University, 47000 Sakon Nakhon, Thailand b Department of Chemistry—Ångström-Laboratory, Physical Chemistry, Box 523, Uppsala University, SE-751 20 Uppsala, Sweden c Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and the Barcelona Institute of Science and Technology (BIST), Building ICN2, Campus UAB, E- 08193 Bellaterra, Barcelona, Spain d Department of Physics, Faculty of Science,Vali-e-Asr University of Rafsanjan, Rafsanjan 77139-36417, Iran A R T I C L E I N F O A B S T R A C T Keywords: Low temperature processed carbon-based perovskite solar cells (C-PSCs) have gained great interest because of Carbon-based perovskite solar cells (C-PSCs) low cost and ease of fabrication. By replacing the Au electrode with carbon, stable solar cells suited for mass- SnO2 production process can be made. However, power conversion efficiencies (PCEs) of C–PSCs still lag behind Hexyl trimethylammonium bromide (HTAB) that of PSCs with Au contact. Surface passivation Carbon electrode Here we explore low temperature (≤150 ◦ C) processed C-PSCs with, where a two-step method is used to prepare mixed-ion lead perovskite films, with tin oxide (SnO2) electron transport layer, poly(3-hexylthiophene- 2,5-diyl) (P3HT) hole transport layer and carbon electrode, resulting in devices with a PCE of 14.0%. Moreover, hexyl trimethylammonium bromide (HTAB) was introduced to improve the interface between perovskite and P3HT. Perovskite grains were remarkably enlarged into micrometer-size and defects were reduced. As a result, a champion PCE of 16.1% was obtained, mainly due to enhanced fill factor from 0.67 to 0.73. The interface modification by HTAB molecule is an effective way to passivate the perovskite defects and facilitate the carrier transport at the perovskite/HTL interface. Unencapsulated devices showed excellent stability over 1500 h stored under ambient air (relative humidity ~ 50 ± 10%). 1. Introduction requires additional dopants, such as 4‑tert-butylpyridine (t-BP), the hygroscopic bis(trifluoromethane)sulfonimide lithium salt (Li-TFSI) and Hybrid metal halide perovskite solar cells (PSCs) are highly inter­ tris (2-(1H-pyrazol-1-yl)− 4‑tert-butylpyridine)-cobalt(III)tris(bis-(tri- esting materials for photovoltaic energy research; Their power conver­ fluoromethylsulfonyl)imide)) (FK209), which lowers the device stability sion efficiencies (PCE) have rocketed from 3.8% to 25.7% in recent years [6,7]. Noble metals contact such as gold (Au) or silver (Ag) are deteri­ [1,2]. These impressive PCEs can be achieved due to their unique op­ orating the device stability due to migration of both halide ions and toelectronic properties including tunable bandgap, high absorption co­ metal ions [8]. Therefore, metal contact are not suited for efficient (~105 cm–1), long charge carrier diffusion length (1>μm), low mass-produced PSC, also because of the relatively high cost and energy exciton-binding energies (~20 meV) and high defect tolerance [3–5]. consumption of the thermal vacuum evaporation process [9]. The device stability is, however, far from ideal and may limit further To solve the drawback of these issues, various types of carbon elec­ industrial and practical exploitation. Poor stability arises from several trodes (CEs) such as carbon black, graphene, graphite and carbon components: The perovskite itself is water-sensitive and can degrade in nanotubes (CNTs) have been applied in perovskite solar cells. These tend the presence of light, and elevated temperature REF. The most to be hydrophobic and can protect the perovskite layer from moisture commonly used hole transport material (HTM), 2,20,7,70- tetrakis-(N, corrosion, improving long term stability, thus contributing to large scale N-di-4-methoxyphenylamino)− 9,90-spirobifluorene (spiroOMeTAD) fabrication [10,11]. Furthermore, they have a suitable work function of * Corresponding author. E-mail address: gerrit.boschloo@kemi.uu.se (G. Boschloo). https://doi.org/10.1016/j.electacta.2023.141935 Received 7 July 2022; Received in revised form 18 January 2023; Accepted 22 January 2023 Available online 23 January 2023 0013-4686/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 Fig. 1. Schematic diagram of (a) fabrication process (b) architecture of the HTAB passivated C-PSC (c) The cross–section SEM images of HTAB passivated C–PSCs, zoomed in on the perovskite layer (left) and complete cross section (right). 5.0 eV, low cost and can be deposited using vacuum-free printing due to interface modification leading to enhanced charge transfer and techniques [12]. Carbon based perovskite solar cells (C-PSC) were first defect passivation in C-PSC [24]. Recently, Zou et al. added 2,3,5,6-tet­ explored without hole transport layer (HTL) in a multi-layer mesoporous rafluoro-7,7,8,8- tetracyanoquinodimethane (F4TCNQ) into the perov­ structure [13]. A typical multi–layer structure is composed of screen­ skite of an undoped P3HT/carbon device, improving the PCE to 15.1% printed layers of TiO2, ZrO2 and graphite / carbon black mixture, which [25]. are annealed at high temperature. Subsequently, perovskite solution is In perovskite solar cells, a high density of defect occurs in the bulk drop casted on the carbon surface to infiltrate through the mesoporous and at the grain boundaries of the polycrystalline perovskite film, which layers, which yielded a PCE from 6.6% [13] to 17.5% [14]. Both metal inevitably deteriorate the device performance. Hence, the defect healing oxide and carbon layer typically require a high temperature (≥ 450 ◦ C) is an effective way to reduce recombination and energy losses at the sintering step which increases the production costs and limits their interfaces. This can be achieved by surface modification using long- applicability on flexible, polymeric substrates, and multi–junction de­ chain alkyl or aromatic ammonium ions [26]. vice structures [15,16]. In contrast, SnO2 electron transport layers Various passivation strategies are developed to improve charge (ETLs) are more attractive in PSCs compared with the traditional TiO2, carrier extraction, suppress non-radiative recombination through the due to high performance and a low temperature deposition process undesired grain boundaries leading to both high efficiency and stable (≤150 ◦ C), good stability and optical transparency, all desirable for PSCs [27,28]. Powerful passivation agents include phenyl­ commercial application [17,18]. The low-cost polymeric hole conductor ethylammonium iodide (PEAI), while methylammonium chloride poly (3-hexylthiophene-2,5-diyl) is a promising candidate for C-PSCs. (MACl) can significantly improve the perovskite crystallinity, enlarge Chu et al. investigated graphene-doped P3HT as HTM, delivering PCE crystal grain, reduce trap state, negligible hysteresis and long-term from 11.1% to 17.8%, which was achieved due to the increase hole stability [29–32]. To eliminate the interface defects, halides passiva­ mobility from 8.3 × 10− 3 to 1.2 × 10− 2 cm2 V–1 s–1 [19]. Undoped P3HT tors, such as adamantylammonium halides (ADAHX, X = Cl–,Br–,I–) used in C-PSCs yielded a PCE of 15.7%, and maintained 70% of initial [33], cetyl trimethyl ammonium bromide (CTAB) [34] were investi­ efficiency at 82 ◦ C for 900 h in air [20]. gated to reduce non-radiative recombination while ammonium group Doping-free P3HT based n–i–p PSCs usually exhibit loss of open also could insert the cationic vacancies of the 3D perovskite which -circuit potential (VOC) due to non-radiative recombination, lower car­ provided the resilience from moisture attack. To date, a high PCE of rier lifetime, and mismatched energy alignment at interface [21–23]. 24.6% was impressively achieved when n-hexyl trimethylammonium Fan et al. reported the dopant-free π- conjugated polymer PBDT(S)-T1 bromide (HTAB) was inserted between perovskite and P3HT after gal­ incorporated with 1,8-diiodooctane (DIO) that showed a PCE of 14.82% lium (III) acetylacetonate (Ga(acac)3) was directly incorporated into the 2 S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 Fig. 2. SEM images of the pristine perovskite films with and without HTAB treatment of different molar concentration (a) 0 mM (b)1 mM (c) 2 mM (d) 3 mM at 200 nm scale bars. HTL [35]. (FF) from 67% to 73% as well as reduced hysteresis from 8.7% to 3.4%. Lee et al. introduced the dual protection of two-dimensional (2D) The PCE of a champion passivated device was boosted to 16.1% from perovskite layer (PEAI) to be good interface between perovskite and 14.0% for reference devices. Storage of the devices, with and without carbon electrode, suppressed the ion migration, improved efficiency up HTAB, did not significantly reduce their performance, both in dry to 15.6%. Lewis-bases approaches in multiple ligand passivators create (relative humidity (RH) ~20 ± 5%) and ambient air (RH ~50 ± 10%). the stronger anchoring effects than monodentate coordination bonds, cationic and anionic vacancies which are passivated at the perovskite 2. Results and discussion surface owing to more heteroatoms [36,37]. Recently, both pyridine and amino group, a bidentate passivator, 2-amino-5-(trifluoromethyl)pyri­ Mixed cation and halide perovskite films with an approximate dine (5-TFMAP) was applied to be stronger bonding between the composition (Cs0.05 FA0.54MA0.41) Pb(I0.98 Br0.02)3, where FA and MA metal-ligand complex and carbon electrode, enhanced efficiency up to stand for formamidinium and methylammonium, respectively, are pre­ 14.96% [38]. pared by a two-step method, [23] followed by surface modification with In this work, we report on low temperature processed C-PSC with HTAB forming an ultrathin 2D perovskite layer [22] (Fig. 1a). The interface modification using HTAB. Passivation effect on perovskite positively charged trimethylammonium group from the HTAB can layer was effectively improved the crystallinity and enlarge grain sizes, replace A-site cations at the perovskite surface or fill A-site vacancies attributed to the more efficient charge transfer, reduced recombination there. The whole schematic structure of planar–type C-PSC is displayed at perovskite/P3HT interface, resulting in drastically raised fill factor in Fig. 1b. ETL for SnO2 layer, perovskite absorber layer with and Fig. 3. Water contact angle measurement of perovskite films without HTAB (a) and with HTAB treatment (b). 3 S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 Fig. 4. (a) XPS survey spectra with and w/o HTAB modification films. (b) I/Br core level XPS spectra of mixed perovskite with and w/o HTAB modified films. XPS core-level spectra of N 1 s of (c) pristine and (d) passivated perovskite films. without (w/o) HTAB modification, P3HT as doping–free hole conductor here. The peak is, however, lower in HTAB-treated films than in pristine was processed by spin–coated method. Finally, carbon paste was films. deposited on the P3HT using doctor blading. The cross-sectional scan­ To understand the elementary composition on the passivated films, ning electron microscopy (SEM) images of C-PSC (see Fig. 1c) demon­ survey and core-level X-ray photoelectron spectroscopy (XPS) spectra of strate that the carbon layer is about 20 μm thick and contains large perovskite films with and w/o HTAB are displayed in Fig. 4a. A change flakes of graphite. The magnified view show that large grains of in the I/Br ratio is confirms addition of Br from HTAB molecules perovskite films are neatly arranged to the normal direction; the thick­ (Fig. 4b). The I/Br atomic ratio decreased from 4.60 to 3.94 in HTAB- ness is about 470 nm. modified perovskite films [40]. Fitting results can be seen in support­ Top view of SEM image of perovskite surfaces with and without ing information, Fig. S2. The N 1 s spectrum for films without HTAB HTAB are shown in Fig. 2. The pristine film shows some unreacted PbI2 display only one peak at 400.9 eV(Fig. 4c), which can be attributed to white grains and several pin–holes. After HTAB treatment (1 and 2 mM), nitrogen in FA [40]. No clear evidence for MA is found, which should full coverage, compact and dense morphologies are obtained, presum­ give a peak for N at 402.5 eV. Upon HTAB treatment, a clear peak at ably due to reaction of PbI2, as well as some of the 3D perovskite, with 402.9 eV appears, which is attributed to the nitrogen in HTAB (Fig. 4d). HTAB to form a 2D perovskite layer on top of the 3D perovskite surface. In addition, Fig. S4 shows no distinct shift peaks are observed at the As a result, the maximum grain size of pristine film (~880 nm), is spin-orbit splitting of Pb 4f7/2, Pb 4f5/2 (~139.6, ~143.5 eV) and I 3d5/2 significantly increased up to 1.5 μm. For 3 mM HTAB treatment more and 3d3/2 (~619.5, ~631 eV) [23,41]. Compared with the pristine film, pin–holes are evident in the SEM top-view. the atomic ratio of I/Pb is slightly increased from 2.63 to 2.89 after the Fig. 3 illustrates the water contact angle tests on untreated and treatment of HTAB. Thus, the elemental analysis of (I+ Br3d)/Pb ratios HTAB-treated films to further confirm the presence of the amphiphilic of pristine and HTAB treated film are 2.75 and 3.07 while (I + C1s)/Pb HTAB molecule. The HTAB-treated perovskite films show an average ratios are 2.70 in the original film and 3.04 in the modified film, contact angle of 60.24◦ , which is higher than pure perovskite film which respectively (see Table S1 and S2). has an average contact angle of 42.13◦ This can be directly related to the Furthermore, HTAB modified perovskite film was determined by apolar nature of the hexyl chains from the HTAB that point upward [39]. energy-dispersive X-ray spectroscopy (EDS), see Fig. S5. From the XRD analysis shows that the perovskite peaks at 14.1º (110) and elemental mapping it is evident that the Br signal from the HTAB is 28.47º (220) are remarkably increased due to HTAB modification, which uniformly distributed on the surface area of the film. For this purpose a suggests enhanced crystallization of perovskite film (supporting infor­ pure FAPbI3 films was used. mation, Fig. S1). There is a significant PbI2 peak at 12.7º remaining in Fig. S6 shows the Raman spectra of perovskite films with and w/o the film, which is a rather common feature in the 2-step method as used HTAB modification. The perovskite peak at 137 cm–1 increased in 4 S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 Fig. 5. (a) UV–Vis absorption spectra of different perovskite films, 1.57 eV estimated from the Tauc plot. Steady-state PL spectra of films with and w/o HTAB, excitation wavelength 470 nm (b) on the perovskite/glass substrate (c) on the perovskite/ITO/glass substrate. magnitude and shifted to higher frequency values (141 cm–1) upon surface traps by HTAB. HTAB modification. This suggests improved film crystallization and due The optical properties of untreated and HTAB-treated films have to interaction between HTAB and the perovskite surface. Moreover, the been studied by UV–Vis. The onset of absorbance spectra is nearly appearance of PL background indicated the successful passivation of identical around 780 nm in Fig. 5a. The optical absorption of HTAB- Fig. 6. (a) Nyquist plot of pristine and HTAB-passivated devices measured at 0.90 V under the dark. (b) Voc decay measurement. Inset shows stabilized open circuit voltage of each type. 5 S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 Fig. 7. (a) Hysteresis J–V curves of champion devices of pristine and HTAB-treated devices. (b) IPCE spectra with integrated JSC derived from IPCE spectra over the standard AM 1.5 G. treated film was slightly higher in the visible range (see inset Fig. 5a). Table 1 The HTAB treatment film did not affected the optical bandgap (Eg) of The key photovoltaic parameters derived from J–V curves of champion and about 1.57 eV calculated from Tauc-plot (Fig. 5a inset). average devices. To study recombination in untreated and HTAB-treated films, steady-state photoluminescence (PL) analysis was performed. The Devices Scan PCE Voc (V) FF (%) Jsc (mA HI (%) cm–2) steady-state PL intensity is clearly increased for passivated perovskite films on glass substrate, indicating reduced non-radiative recombination w/o Reverse 14.04 0.966 67.16 21.64 0.087 HTAB Forward 12.82 0.963 61.60 21.61 (see Fig. 5b). In Fig. 5c, stronger PL quenching of passivated film is Average Reverse 13.35 0.93 65.34 21.89 0.10 observed when it is deposited on the ITO substrate. This suggests ±0.47 ±0.018 ±1.61 ±0.23 ±0.058 improved electron transfer to the ITO substrate, as a result of the With Reverse 16.08 0.984 73.48 22.23 0.034 passivation, which is slightly unexpected since HTAB treatment should HTAB Forward 15.53 0.973 71.83 22.22 mostly affect the perovskite / air interface [38]. It is found that PL peak Average Reverse 15.25 0.97 69.43 22.68 0.056 slightly shifts to shorter wavelengths from 789 nm to 785 owing to ±0.64 ±0.016 ±2.48 ±0.36 ±0.028 passivation effects, see inserts in Fig. 5b,c [42]. To further probe the effect of HTAB treatment, electrochemical A champion PCE of 16.08% is achieved using modification by HTAB, impedance spectroscopy (EIS) was investigated at applied forward bias leading to a remarkable increase in FF up to 73.48%, JSC up to of 22.23 of 0.90 V under the dark. The Nyquist plot in the frequency range is set mA cm–2 and a VOC of 0.984 V. In contrast, the pristine device achieves a from 500 kHz to 1 Hz with a complete semicircle. The equivalent circuit champion PCE of 14.04%, a JSC of 21.64 mA cm–2, a VOC of 0.966 V and a model and fitted parameters extracted from the EIS spectra are shown in FF of 67%. Variation of the concentration of HTAB from 1 mM to 2 mM Fig. 6a and Table S3. The high frequency intercept in the Nyquist plot in isopropanol yielded similar average PCEs (~15.3%) in C-PSCs (see gives the series resistance (Rs) of the device, which originates from the Figure S8 and Table S4). The statistical distribution of 2 mM HTAB resistance of the ITO and the carbon electrode. The inset shows that Rs concentration on the photovoltaics parameters for 10 devices can be amount to about 9 Ω for both types of devices. The semicircle is given by seen in Fig. S9. the geometric capacitance (Cgeom) and the recombination resistance Hysteresis in J–V curves was significantly reduced in the HTAB- (Rrec) of the system [43]. HTAB treatment in perovskite device leads to modified C-PSC devices. The hysteresis index (HI) index was calcu­ an increase of Rrec from 318 Ω to 917 Ω, indicating slower charge lated from forward and reverse scans by the following formula [45]: recombination kinetics at the perovskite/P3HT/carbon interface. The PCEreverse − PCEforward carrier lifetime (τe) of pristine and passivated devices were 18.4 μs and HI = (1) PCEreverse 57.4 μs, respectively, calculated by this equation: τe = Rrec Cgeom. Increased carrier lifetime attributed to the reduced recombination at HI of champion devices is effectively reduced from 8.7% (w/o HTAB) P3HT/carbon interface, improving device performance due to higher FF to 3.4% (with HTAB) device. Moreover, the average modified HI values and PCE. These results are in good agreement with the J–V curves and (0.056 ± 0.028) are effectively lower than that of pristine (0.10 ± VOC decay. 0.058) values. These results are summarized in Table 1. A slight improvement of VOC is found upon HTAB modification. To The incident photon to current efficiency (IPCE) spectra of C-PSCs validate the VOC enhancement, steady-state open circuit voltage (VOC) showed values over 80% in the wavelength of 400–600 nm as shown in traces were monitored under 1 sun illumination [44]. The stabilized VOC Fig. 7b. The integrated JSC with HTAB device (19.39 mA cm–2) was for the passivated device showed higher value of 0.98 V compared with slightly higher than that of 3D device (19.14 mA cm–2), while corre­ the pristine device of 0.96 V (inset Fig. 6b). Furthermore, steady–state sponding measured values from J–V curves for pristine and modified VOC under 1 sun for devices without MAI was stabilized to 0.95 V device are 22.34 mA cm–2 and 22.57 mA cm–2; We ascribe the lower (without HTAB) and 0.99 V (with HTAB), as displayed in Fig. S7. VOC values of integrated JSC due to the very low light intensity of mono­ decay measurements displayed a marked difference for devices with or chromatic light used in IPCE measurements. The trend in integrated JSC without HTAB modification, see Fig. 6b. After initial fast decay for both, value is comparable to that of JSC value from J–V curves. the VOC in HTAB devices decayed slower in the second part, which might To validate the reproducibility of photovoltaic (PV) cells, statistical be related to the slow release of ions accumulated at interfaces. distribution of (2 × 17 cells) pristine and HTAB-modified devices are Fig. 7a shows the J–V curves of C-PSCs in forward and reverse scans. given in Fig. 8. The average PCEs (15.25 ± 0.64%) of passivated devices 6 S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 Fig. 8. Statistical analysis results obtained from 17 cells of each pristine and 1 mM HTAB-modified devices for (a) PCE (b) FF (c) VOC and (d) JSC. All characterization is carried out under 1 sun illumination (AM 1.5 G) at room temperature with an active area of 0.125 cm2. maximum PCE of 13.7% for pristine and 14.79% for HTAB-modified devices (Fig. S10, Table S5), with much lower HI upon HTAB- modification. In addition, the statistical distributions show that PCE and VOC are significantly improved by HTAB modification (Fig. S11). To assess reliability of C-PSCs, steady state power output, devices with and without HTAB were tested by maximum power point tracking (MPPT) as shown in Fig. 9. The HTAB-modified device exhibits a sta­ bilized PCE of 15.0% and current density of 22.38 mA cm–2 with a maximum power point at Vmax of 0.74 V while PCE of pristine device decreased to 12.5% after illumination for 500 s, with a current density of 20.93 mA cm–2 (Vmax = 0.72 V). The stabilized efficiency of the HTAB- modified C-PSC was close to the 15.25% measured from reverse J-V scan. Furthermore, steady-state PCEs were measured for 50 min under 1 sun illumination. Stabilized PCEs of 12.06% (pristine) and 14.35% (passivated) devices were observed under MPPT after 50 min (see Fig. S12). Currently, long-term stability is one of the key challenges for perovskite solar cells. Hence, the stability test was performed for Fig. 9. Stabilized power out measured by maximum power point tracking for unencapsulated C-PSC devices stored in the dark in dry air (RH ~ 20 ± 500 s for un–capsulated devices in ambient air. 5%) or ambient air (RH ~50 ± 10%) at room temperature. No obvious efficiency drop was observed during 1500 h storage (see in Fig. 10). The are significantly higher than that of pristine devices, owing to HTAB device in air displayed an enhancement of efficiency due to improvement of average FF (69.43 ± 2.48%), VOC (0.97 ± 0.016 V) and improvement in FF and JSC (Fig. S13). JSC (22.68 ± 0.36 mA cm–2) listed in Table 1. In the unmodified devices, In addition, when the passivated device is kept in dry air for 1752 h, the average PCE was 13.35 ± 0.47% and all other solar cells parameters VOC is increased from 0.97 to 1.02 V with a slight lower current density were slightly lower too. (from 22.71 to 22.01 mA cm–2) and fill factor (from 70.83% to 68.52%). Similar trends were found for devices prepared without MAI in the However, stabilized efficiency (~15.5%) of these devices is preserved second step of perovskite film preparation. These C–PSCs exhibited a (see Fig. S14). It is noted that co-solvents in the carbon electrode may 7 S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 Fig. 10. Dark storage stability tests without encapsulated devices under dry condition (RH ~20 ± 5%) and ambient air (RH ~50 ± 10%). gradually evaporate without any damage or degradation of perovskite in ambient air. After having cooled down to room temperature (RT), layer during stored in the dry air. The improved interface between they were cleaned by UVO again before use. The perovskite precursor perovskite and P3HT/carbon causes the enhanced Voc. solution was prepared by two-step spin coating method [23]. At first, Thermal stress testing at 82 ◦ C in air for unencapsulated C-PSC de­ 1.3 M PbI2 additive with 5% CsI solution was dissolved in solvent of vices showed moderate stability in this study, see Fig. S15. Devices DMF/DMSO (9:1 v/v). Then, PbI2 solution was deposited on the showed a gradual decrease and 70% of the initial PCE was retained after ITO/SnO2 layer at 2500 rpm for 30 s. Then, it was annealed at 60 ◦ C for 310 h. We believe that remaining PbI2 in the perovskite films (prepared 60 s in the N2 filled glove box. After cooling down, 0.12 M MAI, 0.05 M using a two-step method) is responsible for relatively poor thermal MABr, 0.07 M MACl and 0.23 M FAI mixed in the IPA solution was stability, because in our previous work using C-PSC with MAPbI3 (one- poured on the PbI2 layer and spin–coated at 2300 rpm for 30 s. Then, it step method), 70% of the PCE was maintained after 900 h [20]. was taken out from the N2 glovebox and annealed at 150 ◦ C for 15 min in ambient air (RH ~50 ± 10%). For HTAB interlayer 2D devices, 1 mM of 3. Conclusions hexyl trimethylammonium bromide solution was diluted in IPA solvent and then spin–coated on perovskite layer at 3000 rpm for 30 s and fol­ Hexyl trimethylammonium bromide treatment of mixed-ion perov­ lowed by annealing at 120 ◦ C for 60 s in the N2 filled glove box. For a skite films significantly improves the performance of perovskite solar thin P3HT hole selective layer on all devices, 10 mg of P3HT in 1 mL of cells with P3HT as hole conductor and a carbon contact. This treatment CB was spun coated at 3000 rpm for 30 s. Finally, carbon paste was modifies the perovskite surface, leading to increased grain size and coated on the P3HT layer using doctor blading method, and annealed at reduced non-radiative recombination at the perovskite/P3HT interface, 100 ◦ C for 20 min on a hotplate in ambient air. yielding higher open-circuit potentials and solar cell efficiencies. Importantly, the hysteresis in J-V curves is significantly reduced by the 4.2. Characterizations methods treatment. Unencapsulated devices showed excellent storage stability in ambient air, where no degradation was found during 1500 h. The current density and voltage (J–V) measurements were per­ formed using a Sinus-70 solar simulator (Wavelabs, Leipzig, Germany) 4. Experimental section in the ambient atmosphere under 1 sun illumination (AM 1.5 G) at room temperature. A black metal mask with a circular aperture of 0.125 cm2 4.1. Materials and solar cell preparation was used to define the active area. Reverse (from 1.05 to –0.1 V) and forward scans (from –0.1 to 1.05 V) in all devices were set to a scan Lead (II) iodide (PbI2, 99.99%) and hexyl trimethylammonium bro­ speed of 50 mV/s. Storage stability tests of unencapsulated devices was mide (HTAB, >98.0%) were purchased from TCI. Formamidinium io­ done in the dark in a dry box (relative humidity, RH ~20 ± 5%) or in dide (FAI, >99.99%), methylammonium iodide (MAI > 99.99%), ambient air (RH ~ 40 ± 10%). IPCE spectra were recorded with a xenon methylammonium bromide (MABr>99.99%) and methylammonium lamp (Spectral products ASB-XE-175), a monochromator (Spectral chloride (MACl >99.99%) were ordered from Greatcell Solar. Poly (3- Products CM110) and a LabJack U6 digital acquisition board in the hexylthiophene-2,5-diyl), (P3HT, regioregular, average Mw wavelength range from 350 nm to 850 nm. The light intensity was 50,000–100,000, prod. no. 445703, Sigma-Aldrich), cesium iodide (CsI, calibrated using a certified silicon photodiode (Fraunhofer ISE) prior to 99.99%, Sigma-Aldrich), 2-propanol anhydrous (IPA, 99.5%, Sigma- characterization. The ultraviolet− visible (UV− vis) spectra were Aldrich), N,N-dimethylformamide (DMF, 99.8%, ACROS), methyl sulf­ measured using a Varian Cary 5000 spectrophotometer. The surface oxide (DMSO, 99.7%, ACROS), chlorobenzene (CB, 99.8%, ACROS), morphology and cross–sectional scanning electron microscopy (SEM) carbon paste (Dyenamo, DN–CP01). Tin (IV) oxide, (SnO2, 15% in H2O images of the samples were observed using the Zeiss LEO 1530/1550 colloidal dispersion) was purchased from Alfa Aesar. All purchased microscope. The X-ray diffraction (XRD) patterns were obtained by a chemicals were used without further purification. Siemens D5000 θ− 2θ goniometer with Cu Kα (λ = 1.54051 Å) under Indium tin oxide (ITO, 15 Ω/sq, Optical Filters, 2.4 × 2.8 cm2) glass ambient air condition. The steady state photoluminescence (PL) spectra substrates were first etched using 2 M of hydrochloric (HCl) acid with were recorded on a HORIBA Jobin Yvon Fluorolog station, with exci­ deionized water (DI). And then, followed by sonication in sequence with tation wavelength at 470 nm from the perovskite side. The samples were 2% Hellmanex in water solution, DI, acetone and ethanol for 20 min prepared by with and without HTAB treatment perovskite films onto the each. The cleaned substrates were dried using a stream of air. Then, the ITO glass substrates. Electrochemical impedance spectroscopy (EIS) of patterned ITO was treated with ultraviolet ozone (UVO) for 30 min devices were examined by an Autolab Potentiostat/Galvanostat before deposition ETL. After that, colloidal SnO2 solution was diluted (PGSTA100) under the dark at an applied voltage (0.9 V). Fitting pa­ with DI volume ratio of 4 to 1, then spin coated on the ITO substrate at rameters were collected using FRA 4.5 (Frequency Response Analyzer) 3000 rpm for 30 s. Then, all substrates were heated at 150 ◦ C for 30 min software to obtain from the frequency range between 500 kHz and 1 Hz 8 S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 with an AC perturbation of 10 mV. Open–circuit voltage (Voc) decay and perspectives, Energy Fuels 35 (2021) 18839–18844, https://doi.org/10.1021/ acs.energyfuels.1c02190. recorded using a homebuilt system with a white LED light (Luxeon Star [8] K. Domanski, J.P. Correa-Baena, N. Mine, M.K. Nazeeruddin, A. Abate, M. Saliba, 1 W) and digital acquisition board (National Instrument 6252). X-ray W. Tress, A. Hagfeldt, M. Grätzel, Not all that glitters is gold: metal-migration- photoelectron spectra (XPS) were recorded by using a Physical Elec­ induced degradation in perovskite solar cells, ACS Nano 10 (2016) 6306–6314, tronics Quantera II using AlKα (1486.6 eV) X-rays at a pressure below 7 https://doi.org/10.1021/acsnano.6b02613. [9] F. Zhang, X. Yang, H. Wang, M. Cheng, J. Zhao, L. Sun, Structure engineering of × 10–9 Torr. The binding energy was calibrated against the Fermi level hole-conductor free perovskite-based solar cells with low-temperature-processed by measuring the Au4f core levels of a gold foil mounted on the commercial carbon paste as cathode, ACS Appl. Mater. Interfaces 6 (2014) manipulator and setting the binding energy of Au4f7/2 to 84.0 eV. XPS 16140–16146, https://doi.org/10.1021/am504175x. [10] W. Chen, X. Yin, M. Que, H. Xie, J. Liu, C. Yang, Y. Guo, Y. Wu, W. Que, data analysis was carried out with CasaXPS (H243 UU Site License) A comparative study of planar and mesoporous perovskite solar cells with printable software. carbon electrodes, J. Power Sources. 412 (2019) 118–124, https://doi.org/ 10.1016/j.jpowsour.2018.11.031. [11] M. Que, B. Zhang, J. Chen, X. Yin, S. Yun, Carbon-based electrodes for perovskite CRediT authorship contribution statement solar cells, Mater. Adv. 2 (2021) 5560–5579, https://doi.org/10.1039/ d1ma00352f. S.K.K.A., T.S. and G.B.: devised the research. S.K.K.A. performed [12] H. Zhang, K. Song, L. Zhu, Q. Meng, Back-interface regulation for carbon-based perovskite solar cells, Carbon N. Y. 168 (2020) 372–391, https://doi.org/10.1016/ all experiment unless noted otherwise. A.V. performed SEM, XRD, PL j.carbon.2020.06.065. and UV–vis spectroscopy. M. K. performed and analyzed XPS, EDS and [13] A. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu, J. Chen, Y. Yang, Raman experiments. S.K.K.A. and G.B. wrote the manuscript. G.B. M. Grätzel, H. Han, A hole-conductor-free, fully printable mesoscopic perovskite solar cell with high stability, Science 345 (80) (2014) 295–298, https://doi.org/ supervised the work. 10.1126/science.1254763. [14] X. Chen, Y. Xia, Q. Huang, Z. Li, A. Mei, Y. Hu, T. Wang, R. Cheacharoen, Y. Rong, Declaration of Competing Interest H. Han, Tailoring the dimensionality of hybrid perovskites in mesoporous carbon electrodes for type-II band alignment and enhanced performance of printable hole- conductor-free perovskite solar cells, Adv. Energy Mater. 11 (2021), 2100292, The authors declare that they have no known competing financial https://doi.org/10.1002/aenm.202100292. interests or personal relationships that could have appeared to influence [15] H. Zhou, Y. Shi, K. Wang, Q. Dong, X. Bai, Y. Xing, Y. Du, T. Ma, Low-temperature the work reported in this paper. processed and carbon-based ZnO/CH3NH3PbI3/C planar heterojunction perovskite solar cells, J. Phys. Chem. C. 119 (2015) 4600–4605, https://doi.org/ 10.1021/jp512101d. Data availability [16] G. Yang, H. Tao, P. Qin, W. Ke, G. Fang, Recent progress in electron transport layers for efficient perovskite solar cells, J. Mater. Chem. A. 4 (2016) 3970–3990, https://doi.org/10.1039/c5ta09011c. Data will be made available on request. [17] L. Xiong, Y. Guo, J. Wen, H. Liu, G. Yang, P. Qin, G. Fang, Review on the application of SnO2 in perovskite solar cells, Adv. Funct. Mater. 28 (2018), 1802757, https://doi.org/10.1002/adfm.201802757. [18] C. Altinkaya, E. Aydin, E. Ugur, F.H. Isikgor, A.S. Subbiah, M. De Bastiani, J. Liu, Acknowledgement A. Babayigit, T.G. Allen, F. Laquai, A. Yildiz, S. De Wolf, Tin oxide electron- selective layers for efficient, stable, and scalable perovskite solar cells, Adv. Mater. This work was financially supported from Thailand International 33 (2021), 2005504, https://doi.org/10.1002/adma.202005504. [19] Q.Q. Chu, B. Ding, J. Peng, H. Shen, X. Li, Y. Liu, C.X. Li, C.J. Li, G.J. Yang, T. Cooperation Agency (TICA), Thailand Science Research and Innovation P. White, K.R. Catchpole, Highly stable carbon-based perovskite solar cell with a (TSRI), National Research Council of Thailand (NRCT) from The Royal record efficiency of over 18% via hole transport engineering, J. Mater. Sci. Golden Jubilee (RGJ) Programme, Swedish International Development Technol. 35 (2019) 987–993, https://doi.org/10.1016/j.jmst.2018.12.025. [20] S.K.K. Aung, A. Vijayan, T. Seetawan, G. Boschloo, Improved efficiency of Cooperation Agency (Sida) and International Science Program (ISP) at perovskite solar cells with low-temperature processed carbon by introduction of a Uppsala University under Thai-Swedish Trilateral Development Coop­ doping-free polymeric hole conductor, Sol. RRL. (2021), 2100773, https://doi.org/ eration Programme (grant no. PHD/0101/2560), the Swedish Founda­ 10.1002/solr.202100773. tion for Strategic Research (project no. RMA15–0130), and the Standup [21] M. Stolterfoht, P. Caprioglio, C.M. Wolff, J.A. Márquez, J. Nordmann, S. Zhang, D. Rothhardt, U. Hörmann, Y. Amir, A. Redinger, L. Kegelmann, F. Zu, S. Albrecht, for Energy program for financial support. We thank Dr. Byeong-Jo Kim N. Koch, T. Kirchartz, M. Saliba, T. Unold, D. Neher, The impact of energy for useful discussions. alignment and interfacial recombination on the internal and external open-circuit voltage of perovskite solar cells, Energy Environ. Sci. 12 (2019) 2778–2788, https://doi.org/10.1039/c9ee02020a. Supplementary materials [22] E.H. Jung, N.J. Jeon, E.Y. Park, C.S. Moon, T.J. Shin, T.Y. Yang, J.H. Noh, Efficient, stable and scalable perovskite solar cells using poly(3-hexylthiophene), Nature 567 Supplementary material associated with this article can be found, in (2019) 511–515, https://doi.org/10.1038/s41586-019-1036-3. [23] W. Zhang, L. Wan, S. Fu, X. Li, J. Fang, Reducing energy loss and stabilising the the online version, at doi:10.1016/j.electacta.2023.141935. perovskite/poly (3-hexylthiophene) interface through a polyelectrolyte interlayer, J. Mater. Chem. A. 8 (2020) 6546–6554, https://doi.org/10.1039/d0ta01860k. References [24] B. Fan, T. You, K. Wang, R. Yin, Y. Gao, L. Huo, P. Yin, Additive-assisted interfacial engineering for efficient carbon-based perovskite solar cell incorporated dopant- free polymeric hole conductor PBDT(S)-T1, ACS Appl. Energy Mater. 4 (2021) [1] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, Organometal halide perovskites as 5821–5829, https://doi.org/10.1021/acsaem.1c00631. visible-light sensitizers for photovoltaic cells, J. Am. Chem. Soc. 131 (2009) [25] Y. Zou, W. Yu, Z. Tang, X. Li, H. Guo, G. Liu, Q. Zhang, Y. Zhang, Z. Zhang, C. Wu, 6050–6051, https://doi.org/10.1021/ja809598r. J. Xiao, B. Qu, Z. Chen, L. Xiao, Improving interfacial charge transfer by multi- [2] NREL, Best Research-Cell Efficiency Chart, https://www.nrel.gov/pv/assets/p functional additive for high-performance carbon-based perovskite solar cells, Appl. dfs/best-research-cell-efficiencies.pdf. Phys. Lett. 119 (2021), 151104, https://doi.org/10.1063/5.0061869. [3] B.W. Park, B. Philippe, S.M. Jain, X. Zhang, T. Edvinsson, H. Rensmo, B. Zietz, [26] C. Zhi, Z. Li, B. Wei, Recent progress in stabilizing perovskite solar cells through G. Boschloo, Chemical engineering of methylammonium lead iodide/bromide two-dimensional modification, Appl. Phys. Lett. 9 (2021), 070702, https://doi. perovskites: tuning of opto-electronic properties and photovoltaic performance, org/10.1063/5.0056106. J. Mater. Chem. A. 3 (2015) 21760–21771, https://doi.org/10.1039/c5ta05470b. [27] M.A. Mahmud, T. Duong, J. Peng, Y. Wu, H. Shen, D. Walter, H.T. Nguyen, [4] F. Wang, S. Bai, W. Tress, A. Hagfeldt, F. Gao, Defects engineering for high- N. Mozaffari, G.D. Tabi, K.R. Catchpole, K.J. Weber, T.P. White, Origin of performance perovskite solar cells, Npj Flex. Electron. 2 (2018), https://doi.org/ efficiency and stability enhancement in high-performing mixed dimensional 2D-3D 10.1038/s41528-018-0035-z. perovskite solar cells: a review, Adv. Funct. Mater. (2021), 2009164, https://doi. [5] H.J. Eggimann, J.B. Patel, M.B. Johnston, L.M. Herz, Efficient energy transfer org/10.1002/adfm.202009164. mitigates parasitic light absorption in molecular charge-extraction layers for [28] J. Suo, B. Yang, A. Hagfeldt, Passivation strategies through surface reconstruction perovskite solar cells, Nat. Commun. 11 (2020), https://doi.org/10.1038/s41467- toward highly efficient and stable perovskite solar cells on n-i-p architecture, 020-19268-w. Energies 14 (2021) 4836, https://doi.org/10.3390/en14164836. [6] G. Ren, W. Han, Y. Deng, W. Wu, Z. Li, J. Guo, H. Bao, C. Liu, W. Guo, Strategies of [29] T. Zhu, D. Zheng, J. Liu, L. Coolen, T. Pauporte, PEAI-based interfacial layer for modifying spiro-OMeTAD materials for perovskite solar cells: a review, J. Mater. high-efficiency and stable solar cells based on a macl-mediated grown Chem. A 9 (2021) 4589–4625, https://doi.org/10.1039/d0ta11564a. Fa0.94Ma0.06PbI3 perovskite, ACS Appl. Mater. Interfaces. 12 (2020) [7] G. Tumen-Ulzii, T. Matsushima, C. Adachi, Mini-review on efficiency and stability 37197–37207, https://doi.org/10.1021/acsami.0c09970. of perovskite solar cells with spiro-OMeTAD hole transport layer: recent progress 9 S.K.K. Aung et al. Electrochimica Acta 443 (2023) 141935 [30] F. Zheng, C. Zuo, M. Niu, C. Zhou, S.J. Bradley, C.R. Hall, W. Xu, X. Wen, X. Hao, [38] T. Xu, K. Zou, S. Lv, H. Tang, Y. Zhang, Y. Chen, L. Chen, Z. Li, W. Huang, Efficient X. Hao, M. Gao, T.A. Smith, K.P. Ghiggino, Revealing the role of methylammonium and stable carbon-based perovskite solar cells via passivation by a multifunctional chloride for improving the performance of 2D perovskite solar cells, ACS Appl. hydrophobic molecule with bidentate anchors, ACS Appl. Mater. Interfaces. 13 Mater. Interfaces. 12 (2020) 25980–25990, https://doi.org/10.1021/ (2021) 16485–16497, https://doi.org/10.1021/acsami.1c02218. acsami.0c05714. [39] Y. Liu, Z. Wu, Y. Dou, J. Zhang, T. Bu, K. Zhang, D. Fang, Z. Ku, F. Huang, Y. [31] Q. Jiang, Y. Zhao, X. Zhang, X. Yang, Y. Chen, Z. Chu, Q. Ye, X. Li, Z. Yin, J. You, B. Cheng, J. Zhong, Formamidinium-based perovskite solar cells with enhanced Surface passivation of perovskite film for efficient solar cells, Nat. Photonics 13 moisture stability and performance via confined pressure annealing, J. Phys. Chem. (2019) 460–466, https://doi.org/10.1038/s41566-019-0398-2. C 124 (2020) 12249–12258, https://doi.org/10.1021/acs.jpcc.0c02289. [32] M. Tang, S. Zhang, T.J. Magnanelli, N.V. Nguyen, E.J. Heilweil, T.D. Anthopoulos, [40] T.J. Jacobsson, J.P. Correa-Baena, E. Halvani Anaraki, B. Philippe, S.D. Stranks, M. C.A. Hacker, Unraveling the compositional heterogeneity and carrier dynamics of E.F. Bouduban, W. Tress, K. Schenk, J. Teuscher, J.E. Moser, H. Rensmo, alkali cation doped 3D/2D perovskites with improved stability, Mater. Adv. 2 A. Hagfeldt, Unreacted PbI2 as a double-edged sword for enhancing the (2021) 1253–1262, https://doi.org/10.1039/d0ma00967a. performance of perovskite solar cells, J. Am. Chem. Soc. 138 (2016) 10331–10343, [33] M.M. Tavakoli, W. Tress, J.V. Milić, D. Kubicki, L. Emsley, M. Grätzel, Addition of https://doi.org/10.1021/jacs.6b06320. adamantylammonium iodide to hole transport layers enables highly efficient and [41] N. Yi, S. Wang, Z. Duan, K. Wang, Q. Song, S. Xiao, Tailoring the performances of electroluminescent perovskite solar cells, Energy Environ. Sci. 11 (2018) lead halide perovskite devices with electron-beam irradiation, Adv. Mater. 29 3310–3320, https://doi.org/10.1039/c8ee02404a. (2017), 1701636, https://doi.org/10.1002/adma.201701636. [34] S. Wang, Y. Zhu, C. Wang, R. Ma, Interface modification by a multifunctional [42] P. Wang, J. Wang, X. Zhang, H. Wang, X. Cui, S. Yuan, H. Lu, L. Tu, Y. Zhan, ammonium salt for high performance and stable planar perovskite solar cells, L. Zheng, Boosting the performance of perovskite solar cells through a novel active J. Mater. Chem. A. 7 (2019) 11867–11876, https://doi.org/10.1039/c9ta02631b. passivation method, J. Mater. Chem. A. 6 (2018) 15853–15858, https://doi.org/ [35] M.J. Jeong, K.M. Yeom, S.J. Kim, E.H. Jung, J.H. Noh, Spontaneous interface 10.1039/c8ta05593a. engineering for dopant-free poly(3-hexylthiophene) perovskite solar cells with [43] Q. Hu, E. Rezaee, M. Li, Q. Chen, C. Li, S. Cai, H. Shan, Z.X. Xu, P3HT with Zn efficiency over 24%, Energy Environ. Sci. 14 (2021) 2419–2428, https://doi.org/ (C6F5)2 as p-type dopant for the enhanced performance of planar perovskite solar 10.1039/d0ee03312j. cells, Sol. RRL. 4 (2020), 1900340, https://doi.org/10.1002/solr.201900340. [36] H. Zhang, Y. Wu, C. Shen, E. Li, C. Yan, W. Zhang, H. Tian, L. Han, W.H. Zhu, [44] K. Lee, J. Kim, H. Yu, J.W. Lee, C.M. Yoon, S.K. Kim, J. Jang, A highly stable and Efficient and stable chemical passivation on perovskite surface via bidentate efficient carbon electrode-based perovskite solar cell achieved: via interfacial anchoring, Adv. Energy Mater. 9 (2019), 1803573, https://doi.org/10.1002/ growth of 2D PEA2PbI4 perovskite, J. Mater. Chem. A. 6 (2018) 24560–24568, aenm.201803573. https://doi.org/10.1039/c8ta09433k. [37] Z. Wu, M. Jiang, Z. Liu, A. Jamshaid, L.K. Ono, Y. Qi, Highly efficient perovskite [45] S.A.A. Shah, M.H. Sayyad, J. Sun, Z. Guo, Hysteresis analysis of hole-transport- solar cells enabled by multiple ligand passivation, Adv. Energy Mater. 10 (2020), material-free monolithic perovskite solar cells with carbon counter electrode by 1903696, https://doi.org/10.1002/aenm.201903696. current density–voltage and impedance spectra measurements, Nanomaterials 11 (2021) 48, https://doi.org/10.3390/nano11010048. 10