Surface passivation of sequentially deposited perovskite solar cells by octylammonium spacer cations

Perovskite solar cells (PSCs) are an emerging PV technology with beneficial optoelectronic properties, such as high absorption coefficients,¹ long diffusion lengths,² tunable bandgap,^3,4 and high charge carrier mobilities.⁵ The Power Conversion Efficiency (PCE) of PSCs has evolved from 3.8%⁶ to values exceeding 25%^7–10 within just a decade. The collaborative research effort of the perovskite community has led to improved device performance and stability bringing the PSC technology one step closer to commercialization. Nonetheless, the reported PCE values are still lower than the theoretical limits,^11–13 while extended lifetime stability performance under accelerated aging tests or realistic outdoor conditions is still under investigation.^14,15

One of the major factors for the rapid evolution of PSCs is their facile solution-based processing performed at low temperatures. However, solution-processed thin film deposition techniques can lead to formation of defects in the bulk of the perovskite film and at the relevant interfaces.^16–18 The presence of these defects results in non-radiative recombination effects that hinder the photovoltaic performance and stability.^19,20 Defect passivation is an efficient method to minimize the impact of traps in the perovskite film and, therefore, suppress the non-radiative recombination, leading to improved PCEs.²¹ Defect passivation strategies include solvent and deposition engineering,^22,23 additives incorporation,^24,25 and surface passivation by using buffer layers at one perovskite/transport-layer interface^26,27 or dual passivation at both perovskite/transport layer interfaces.²⁸ 

Surface passivation of perovskite films is implemented by utilizing various organic materials. Jiang et al. used Phenethylammonium Iodide (PEAI) molecules to passivate perovskite traps in formamidinium lead iodide (α-FAPbI₃) developed on a planar structure. The PCE of optimized PEAI-treated devices was 23.56%. This enhanced PCE value stemmed from the reduced surface recombination.²⁹ Li et al. synthesized a piperazinium iodide molecule for the surface treatment of perovskite films. The passivation treatment suppressed the non-radiative recombination losses, leading to V_oc and PCE improvement from 20.76% to 23.37%.³⁰ Zhu et al. developed a passivation agent, namely 4-tert-butyl-benzylammonium iodide (tBBAI). The tBBAI-treated PSCs exhibited a PCE boost from 21.2% to 23.5%, demonstrating a high fill factor. The tBBAI treatment accelerated the charge transfer, while it reduced the surface recombination. In addition, the PSCs maintained more than 95% of their efficiency under continuous illumination.³¹ 

An alternate route for defect passivation is realized through dimensional engineering. Following this approach, the treatment of 3D perovskite films with various organic spacer cations enables the formation of low-dimensional (2D) perovskite structures caused by the substitution or replacement of the A-site cation, forming layered 2D/3D or columnar 1D/3D mixed perovskites.^32–34 2D perovskite compounds are divided into two main categories: Ruddlesden–Popper (RP)³⁵ and Dion–Jacobson (DJ) perovskites.³⁶ While 3D perovskites are susceptible to humidity,³⁷ the organic interlayer spacers used to form 2D perovskites are often hydrophobic, resulting in an improved PSC lifetime.^38–40 The passivation of 2D structures on the 3D perovskite results in reduced trap-assisted recombination.⁴¹ 

2D perovskites are formed by employing long alkylammonium or aryl ammonium monovalent cations. Among them, some typical spacer cations used to form 2D/3D heterostructures in PSCs include Phenethylammonium (PEA⁺),⁴² Butylammonium (BA⁺),⁴³ Octylammonium (OA⁺),⁴⁴ and Guanidinium (GUA⁺).⁴⁵ Mozaffari et al. used a mixture of n-octylammonium bromide and Guanidinium Bromide (GUABr) as the spacer cations. The resulting 2D/3D heterostructure outperformed the control PSCs (3D perovskite), and the PCE was increased from 21.37% to 23.13%, demonstrating also a much-improved stability. The OABr/GUABr mixture was shown to reduce the trap-assisted recombination at the perovskite/hole transport layer (HTL) interface.⁴⁶ Mahmud et al. investigated the passivation effect of octylammonium spacer cations [Octylammonium Iodide (OAI), Octylammonium Bromide (OABr), and Octylammonium Chloride (OACl)] on a four-cation perovskite in a mesoporous structure. Devices based on OACl showed the best performance compared to the reference device having either OAI or OABr. Additional characterizations revealed that Cl diffuses toward the perovskite bulk, causing simultaneous perovskite surface and bulk passivation.⁴⁷ 

Most of the reports in the literature commonly use OAI spacer cations to form the 2D/3D perovskite heterostructure.^48–52 In this work, we employ OABr spacer cations to study the formation of a 3D/2D heterostructure on a sequentially deposited FA_1−xMA_xPbI₃ perovskite in a planar structure. We show that the OABr-treated PSCs exhibited a PCE enhancement from 20.26% of the control device to 21.40%, while the open circuit voltage (V_oc) is increased by 40 mV on average. PSC devices treated with OAI were also fabricated; however, their performance was inferior compared to OABr-based PSCs. Material and device characterization revealed that the improved V_oc is due to the reduced non-radiative recombination induced by the OABr passivation layer, as shown by Time Resolved Photoluminescence (TRPL) and Transient Photovoltage (TPV) measurements. Furthermore, OABr-treated PSCs retained almost 90% of their initial PCE after 500 h of exposure to high ambient humidity conditions without encapsulation.

The light absorbance of perovskite thin films coated on glass was measured using ultraviolet–visible (UV–vis) spectroscopy, with and without the 2D passivation layers, as shown in Fig. 1(a). All films exhibited an absorbance onset at ∼810 nm that is a characteristic of FAPbI₃, but no considerable difference is reported, which suggests that the effect of passivation on the optical characteristics is insignificant. The optical bandgap energy (E_g) for the same samples was calculated by the Tauc plot [Fig. 1(b)], where a negligible variation between the samples is observed, implying that the 3D perovskite has remained intact after the formation of the 2D passivation layer on top of it.

X-ray diffraction (XRD) measurements were carried out in perovskite layers to study the phase content of thin films, with and without the passivation layers of OAI and OABr (Fig. S1) on top. It should be noted here that due to the very low thickness of the 2D passivation layer compared to the bulk 3D perovskite, conventional XRD is not able to detect the weak signal intensity originating from the ultra-thin 2D passivation layers observed at low angles.³² As can be observed, there are no changes or shifts in the three obtained XRD patterns, confirming that the structural characteristics of the 3D perovskite α-phase are not affected by the passivation layers OAI and OABr. However, in the obtained XRD pattern without a passivation layer, a weak peak is detected close to 11.8°, which corresponds to the undesired δ-phase of FAPbI₃. After the deposition of the 2D passivation layer, it is revealed that the peak intensity of δ-phase is reduced for the case of OAI and becomes fully suppressed for OABr (see the inset). On the other hand, the desired trigonal black α-phase located at 14° is enhanced in the two cases of passivation layer deposition, compared to the pure 3D perovskite layer, which is expected to be favorable for the efficiency of PSC devices.^54,55

A summary of the fitting parameters can be found in Table S1. The average lifetimes of the reference, OAI-treated, and OABr-treated perovskite films were 0.9, 1.2, and 3.1 µs, respectively. The octylammonium salt treatment increases the lifetime compared to the reference, with a threefold increase for the case of OABr. The enhanced lifetimes of the OAI- and OABr-treated perovskite films indicate that the octylammonium salt passivation can reduce non-radiative recombination effects,^62–64 in agreement also with the PL results. The formation of the 2D perovskite layer on top of the 3D perovskite layer passivates the defects on the surface of the 3D perovskite that are one of the main sources of non-radiative recombination.⁶⁵ 

A typical procedure for the formation of the 3D/2D perovskite heterostructure is based on the spin-coating process of the spacer cations on top of the 3D perovskite film, followed by a short annealing step. A common solvent used for these spacer cations is Isopropyl Alcohol (IPA).⁵³ For thin film characterizations, we have deposited the perovskite films with or without passivation treatment on glass. A two-step deposition process was employed for the perovskite films on glass or indium tin oxide (ITO)\tin oxide (SnO₂) nanoparticle thin film used as the electron transport layer. The first step involves spin-coating of a PbI₂ layer in a N₂ glovebox, while an annealing step at 70 °C for 1 min is performed afterward. The organic salt solution of FAI, MAI, and MACl is spin-coated on top of the PbI₂ layer, and to obtain the α-FAPbI₃ phase, an annealing step at 150 °C for 15 min is performed in ambient air. Then, each of the octylammonium spacer cations dissolved in IPA is spin-coated on top of the 3D α-FAPbI₃ perovskite layer, while to complete the 3D/2D perovskite transformation, an annealing step at 100 °C for 5 min is performed. The device fabrication is completed by spin-coating spiro-OMeTAD used as the HTL and by the thermal evaporation of an Au electrode. More details about the experimental procedures can be found in the supplementary material. To test the effect of the 2D passivation layer, we fabricated devices with FA_1−xMA_xPbI₃ as the perovskite and the structure ITO/SnO₂/perovskite/spiro-OMeTAD/Au (where ITO is indium tin oxide and spiro-OMeTAD is 2,2′,7,7′-tetrakis(N,N-dipmethoxyphenylamine)-9,9′-spirobifluorene) [Fig. 2(a)].

First, we tested various concentrations to optimize OAI and OABr aiming at the best efficiency and identified that the optimum precursor concentration is 2 mg ml⁻¹ for OAI and 1 mg ml⁻¹ for OABr (Fig. S2). At higher concentrations, the fill factor (FF) is significantly reduced, which is probably attributed to the increase in charge transport resistance, due to the thicker and semi-insulating 2D layer on top of the perovskite.^53,66,67 Figure 2(b) shows the box-plots of the photovoltaic characteristics of the control and the optimized 2D/3D PSCs for 20 devices of each configuration. Tables S2–S4 show the parameters of each individual fabricated device. Both passivated devices exhibit higher PCEs than the control device, where OABr-based devices exhibit the highest PCE. The improvement in PCE in the case of OABr (as well as OAI) is due to the higher V_oc and the slightly higher short circuit current (J_sc), while OABr-treated PSCs show the highest FF on average compared to the reference and the OAI-treated PSCs. The current density–voltage (J–V) curves for the champion PSC devices are shown in Fig. 3(a), and the photovoltaic parameters for all the devices are shown in Table I. The best performing OABr device has a PCE of 21.40% and a higher V_oc of 1.10 V compared to the control device, which has a PCE of 20.26% and a V_oc of 1.08 V. On the other hand, the OAI device exhibited a PCE of 20.80% and a V_oc of 1.10 V. The OAI device shows the highest J_sc (23.55 mA cm⁻²) compared to control (22.89 mA cm⁻²) and OABr (23.41 mA cm⁻²) devices but, at the same time, has a lower FF of 80% compared to that of the control and OABr devices, which are 82.24% and 82.97%, respectively. The application of the passivation layers led to an improved V_OC for all conditions compared to the reference sample. These results indicate that the successful formation of the 2D/3D heterostructure leads to effective defect passivation and improved device performance of the solar cells.

We then investigated whether the OAI and OABr passivation treatment has any effect on the ion migration properties of the PSCs.⁶⁸ The J–V hysteresis index is shown in Tables S1–S3 for 20 fabricated reference devices, 20 devices with OAI passivation treatment, and 20 devices with OABr passivation treatment. The average hysteresis index as defined in the supplementary material was 0.07, 0.04, and 0.03, for the control, OABr passivated, and OAI passivated PSCs, respectively. These results demonstrate the potential of the octylammonium passivation treatment to mitigate ion migration. Figure 3(b) shows the external quantum efficiency (EQE) spectra for the devices with and without the OABr and OAI passivation layers, as well as the integrated J_sc. The integrated J_sc for the control, OAI-treated, and OABr-treated PSCs was 22.35, 22.68, and 22.47 mA cm⁻², respectively, which agrees with the J–V measurements. The 2D/3D perovskite solar cells have a higher EQE than the control one, due to the improvement in film quality after the passivation treatment. The slight difference between the integrated and the calculated J_sc can be attributed to the bigger active area of the EQE and the frequency of 60 Hz, which may cause a different dynamic response to PSCs.^69,70

The superior performance of the OABr PSCs, compared to the OAI-treated devices, is confirmed by all the structural, optical, and electronic measurements of the fabricated films and devices. The highest PL intensity of the OABr-treated films, along with the better V_OC and FF of the full devices, indicates that OABr provides the most effective passivation of the tested materials, similar to previous studies.⁴⁷ 

To further understand the cause of the improved performance and V_oc of the OABr-passivated PSCs, we performed Transient Photovoltage (TPV) and Photocurrent (TPC) measurements to gain insights into the charge carrier recombination and extraction processes. Figure 4(b) shows the normalized TPV decays for the reference and OABr-passivated PSCs under the same illumination. The photovoltage decay for the case of the passivated perovskite film is slower than the reference. The obtained charge carrier lifetime from a single exponential fitting is 2.37 µs for the control device and 2.80 µs for the passivated one. Figure 4(a) presents the obtained charge carrier lifetime as a function of V_oc, where the charge carrier lifetimes for the OABr-passivated PSCs are higher than the reference device. The raw data for each decay are shown in Fig. S5. The incorporation of OABr as a passivation layer reduces charge carrier recombination dynamics, which is consistent with the 40 mV enhancement of V_oc on average since it is severely affected by non-radiative recombination effects.⁷¹ The TPV results are also in agreement with the steady-steady state and TRPL results, revealing the defect passivation of OABr on the perovskite surface. Several molecules have been used for the surface passivation of perovskite films, showing improved V_oc and charge carrier lifetimes.^72–75 

We then investigated the charge extraction process using high perturbation transient photocurrent measurements with a duty cycle of 0.8. Figure 4(c) demonstrates the extracted charge as a function of current density obtained by the raw data of Fig. S6. Both devices appear to have similar charge extraction capabilities with a slight improvement for OABr-treated PSCs, and therefore, the improved performance mainly originates from the reduced recombination dynamics due to the passivation treatment.

Finally, we evaluated the stability of the reference and the OABr-treated devices by implementing the ISOS D1 protocol. The devices were stored in ambient air (humidity 50%–60%) under dark conditions, without encapsulation. Figure 5 shows the evolution of the PCE for the reference and OABr-treated devices, where the dashed line represents the T₈₀ lifetime. The T₈₀ lifetime for the reference device is about 336 h, whereas the OABr passivated device retains almost 90% of the initial PCE after 500 h of exposure. The degradation of the reference device is evident from Fig. S4, while the OABr devices show no significant visible degradation. The enhanced stability of the passivated devices can be explained by the higher surface hydrophobicity due to the 2D/3D treatment, which prevents humidity insertion while improving the ambient stability, and by the fewer defects at the passivated interface. The stability improvement of the passivated perovskite films is also consistent with the XRD results, where the suppressed δ-FAPbI₃ phase is beneficial for the moisture tolerance of the OABr-treated perovskite films.

In summary, we demonstrated the passivation procedure of perovskite layers in PSCs, by including OABr and OAI as 2D passivation layers between the perovskite and the HTL. This defect passivation strategy enabled a champion device with an efficiency of 21.40% for the case of OABr, exhibiting a V_oc value of 1.10 V. The lifetime stability tests indicated that OABr-treated PSCs have extended device stability with enhanced moisture tolerance, compared to the non-treated PSCs. PL and TRPL measurements confirmed an improved charge carrier lifetime as a result of the non-radiative recombination reduction for the case of the 2D passivated devices. Moreover, the transient analysis of the OABr-passivated PSCs by TPV and TPC measurements revealed the suppression of charge carrier recombination processes that led to the improved performance of the targeted devices. Further enhancements in PSC performance and stability can be achieved by exploring more materials as 2D surface passivation layers.

The supplementary material includes the experimental methods, TRPL fitting parameters, tables with all device parameters, XRD patterns, device optimization statistics, MPP tracking graph, a photograph of the devices after stability testing, and TPV and TPC raw data.

This work was supported by European Union’s Horizon 2020 research and innovation program under Grant Agreement No. 881603—GrapheneCore3. P.M. acknowledges the Institute of Electronic Structure and Laser of the Foundation for Research and Technology-Hellas (IESL-FORTH) for providing access to x-ray diffraction measurements.

The authors have no conflicts to disclose.

M.L. and M.T. contributed equally to this work.

Michalis Loizos: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Methodology (equal); Writing – original draft (equal). Marinos Tountas: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Methodology (equal); Writing – original draft (equal). Panagiotis Mangelis: Data curation (supporting); Formal analysis (supporting); Writing – original draft (supporting). Konstantinos Rogdakis: Conceptualization (supporting); Supervision (supporting); Writing – review & editing (lead). Emmanuel Kymakis: Funding acquisition (lead); Supervision (lead); Writing – review & editing (supporting).

The data that support the findings of this study are available within the article and its supplementary material.