Diammonium‐induced Dion‐Jacobson 2D/3D wide‐bandgap perovskite solar cells with enhanced efficiency and stability

Abstract

Surface passivation by two‐dimensional (2D) organic spacers provides promising avenue to substantially enhance the photovoltaic performance and structural stability of perovskite solar cells (PSCs). Herein, a diammonium organic salt of 2,2′‐(Ethylenedioxy) diethyl ammonium diiodide (EDBEI₂) is used to passivate surface defects with spontaneously established Dion‐Jacobson (DJ)‐type two‐dimensional/three‐dimensional (2D/3D) perovskite structure. The 2D DJ perovskite (EDBEPbI₄) is formed on the surface of 3D WBG film, which passivates surface defects and reduces non‐radiation recombination with better‐matched energy‐level alignment. Consequently, the modified device shows an excellent power conversion efficiency of 18.85% with large open‐circuit voltage of 1.28 V. Besides, such 2D/3D perovskite exhibits better stability, which maintains 89% of its original efficiency after 1000 h under a relative humidity of 30% for the unencapsulated EDBEI₂‐treated device. These results give a suggestive strategy to enhance the efficiency and stability of WBG PSCs, potentially, providing a feasible way for achieving the commercialization of perovskites.

INTRODUCTION

To date, power conversion efficiencies (PCEs) of the perovskite solar cells (PSCs) is up to 25.7%,¹ and the long-term durability of PSCs is the major obstacle to commercialization. Wide-bandgap (WBG) PSCs have virtually all the instability characteristics of normal-bandgap PSCs, such as invasion of oxygen, moisture, and light illumination.^2–4 Besides, the instability issues resulting from the external environment, WBG PSCs often suffer from severe phase segregation caused by the result of separate I and Br diffusions.^5,6 The polycrystalline perovskite films prepared by solution methods have numerous defects at grain boundaries, surface, and perovskite bulk, which act as trapping sites that cause strong non-radiative recombination.⁷ Furthermore, the defect at the grain boundary or surface also acts as a channel for ion migration and oxygen and water invasion.⁸ Such detrimental defect will lead to nonreversible decomposition and phase segregation, and thus the device instability. To overcome the harmful instability, long-chain organic spacer cations were introduced into 3D perovskites to form ultra-stable 2D or quasi-2D perovskites.⁹ There are two types of quasi-2D perovskite structure: the Ruddlesden-Popper (RP) phase and the Dion-Jacobson (DJ) phase. The chemical formula of the RP perovskite is (A′)₂(A)_n−1Pb_nX_3n+1, A′ is a larger monoamine organic salt, A represents a small monovalent spacer cation, X represents the halide anion, and n refers to the number of inorganics [PbX₆]⁴⁻ layers.¹⁰ The DJ phase shows the chemical formula of (A′′)(A)_n−1Pb_nX_3n+1, where A′′ is a diammonium molecule.¹¹ Compared with 3D perovskites, the 2D perovskites show high formation energy, excellent moisture, thermal stability, and reduced ion migration.^12–14 However, the efficiency of 2D PSCs is restricted by the quantum well structure,¹⁵ lower carrier mobility, and high dielectric constant,¹⁶ which makes the efficiency lag behind that of 3D PSCs.

Recently, 2D/3D-based devices have been proposed to balance the high-efficiency and long-term stability. Numerous 2D/3D heterostructure is based on RP phase with monovalent ammonium, such as butylammonium iodide (BAI),¹⁷ cyclohexylethylammonium iodide (CEAI),¹⁸ and phenethylammonium iodide (PEAI).¹⁹ On the one hand, the composition of the RP phase interacts via weak van der Waals forces,²⁰ whereas the DJ phase has more stable skeleton due to the alternating hydrogen bonding between the organic spacer cation and two adjacent [PbI₆]⁴⁻ layers.²¹ On the other hand, diammonium spacers leave a relatively low degree of freedom for interlayer displacement, which helps to increase the stability of crystal structure.^22,23 Therefore, DJ-type 2D perovskite is more conducive to improving device stability. For instance, Zhang et al.²⁴ reported a metastable DJ 2D perovskite with a reduced transport energy barrier that can maximize the out-of-plane hole transport, the triple-cation-mixed-halide PSC maintains 90% of its original efficiency after 1000 h illumination. Zhong et al.¹⁰ applied 2,2′-(ethylenedioxy)bis(ethylamine) to establish DJ-type 2D/3D perovskite to passivate defect at the grain boundary and improve moisture stability, the encapsulated device retains 82% in relative humidity (RH) of 50 ± 5% for 1560 h aging.¹⁰ Li et al.²⁵ demonstrate a localized DJ-type 2D/3D PSC that yields a PCE of 20.1% and maintains 86% of the original efficiency after 1300 h under high humidity (70% RH). Therefore, the DJ-type 2D/3D cells have provided promising strategy to improve the optoelectronic and structural properties of the PSCs.

Herein, we construct a DJ-type 2D/3D heterostructure by using 2,2′-(Ethylenedioxy) diethylammonium diiodide (EDBEI₂) as the passivation layer in the WBG perovskite community. The EDBEI₂-based DJ 2D perovskite can passivate defects and act as an interfacial layer to protect the penetration of oxygen and water. When exposed to the high RH or elevated temperature condition, such passivated perovskite shows better stability than that of its 3D counterparts. The EDBEI₂-based devices maintain 89% of the initial PCE after 1000 h aging to a RH of ~30%, whereas the control devices only retain 71% of their original PCEs in the same environment. Moreover, the 2D/3D perovskite delivers a champion PCE of 18.85% with high open-circuit voltage (V_OC) of 1.28 V, short-circuit current density (J_SC) of 18.77 mA cm⁻², and fill factor (FF) of 78.5%, which is due to the reduced defect density and good energy-level alignment after EDBEI₂ passivation. Our work demonstrates that constructing a DJ-type 2D/3D perovskite heterostructure contributes to achieving high-efficiency and stability WBG PSCs.

RESULTS AND DISCUSSION Device performances

To study the EDBEI₂ passivation on the photovoltaic performance of WBG PSCs, various concentrations (1, 3, and 5 mg⋅ml⁻¹) of EDBEI₂/ isopropanol(IPA) solution deposited on top of the resulting WBG perovskite film and annealed at 100°C for 5 min, then the EDBEI₂-passivated film was prepared. The composition of WBG perovskite in this work is Cs_0.2FA_0.8Pb(I_0.7Br_0.3)_3. The calculated optical bandgap (E_g) is ~1.75 eV. The device structure is shown in Figure 1A: with an n-i-p structure of ITO/SnO₂/WBG perovskite/spiro-OMeTAD/Ag. According to the photovoltaic parameters depicted in Figure 1B and C, the V_OC shows a linear upward trend with the increase in passivation concentration, and the FF also improved compared to unpassivated devices. Therefore, the enhancement in V_OC and FF, especially the V_OC, lead to the improved PCE. The current density–voltage (J-V) curves and photovoltaic parameters of the optimal PSCs processed with EDBEI₂ passivation are shown in Figure 1D and Table 1.

FIGURE 1. (A) Device structure of the EDBEI2-treated PSCs. (B) V OC distribution scatter plots. (C) FF distribution scatter plots. (D) J-V curves of the champion devices with and without EDBEI2 treatment. (E) J-V curves of the control and EDBEI2-treated devices. (F) Stabilized PCE and JSC of the EDBEI2-treated device

TABLE 1 Photoelectronic parameters of the champion devices

[Table omitted. See PDF]

^aForward scan.

^bReverse scan.

Among these concentrations, 3 mg ml⁻¹ is the optimal concentration that resulted in the best PCE. Although 5 mg ml⁻¹ shows higher V_OC, the short-circuit current density (J_SC) and FF start to trend down so that the overall performance is slightly worse than the 3 mg ml⁻¹ one, which could be the quantum effect of the too thick 2D layers.²⁶ As presented in Figure 1E and Table 1, the control device delivers a PCE of 16.33% with a V_OC of 1.20 V, a J_SC of 18.57 mA⋅cm⁻², and an FF of 72.92%. The EDBEI₂-based device yields a champion PCE of 18.85% with a high V_OC of 1.28 V, a J_SC of 18.85 mA cm⁻², and a high FF of 78.50%. At the same time, the integrated current density obtained from the external quantum efficiency (EQE) are 17.72 and 18.1 mA cm⁻² for the control and EDBEI₂-treated devices, respectively (Figure S1), which agrees with the J_SC from the corresponding J-V curves. The slight difference may be due to the instability of the outside environment during the test process. Moreover, it is obvious that the hysteresis reduced significantly after EDBEI₂ treatment. The hysteresis index (HI) can be used to evaluate the passivation effect and suppressed ion migration, which can be calculated as HI = (PCE_RS − PCE_FS)/PCE_RS. The control device delivers a hysteresis index (HI) of 6.3%, whilst the EDBEI₂-based PSC is 1.7%, indicating the reduced trap density and ion migration after EDBEI₂-passivated. The stabilized current density and PCE of the EDBEI₂-based PSC were characterized at the V_max of 1.02 V. As shown in Figure 1F, a stabilized PCE of ~18.40% is obtained, demonstrating desirable stability after EDBEI₂ passivation.

Surface property of WBG perovskite

The scanning electroscopic microscopy (SEM) images are illustrated in Figure 2A. The control sample exhibits a smooth and hole-free morphology. Obviously, the EDBEI₂-passivated film (1 mg⋅ml⁻¹) shows a rod-like grain on top of the WBG perovskite film. And a large change in morphology is observed when the concentration reached 3 mg ml⁻¹ or 5 mg ml⁻¹, which is due to the in-situ formation of EDBEI₂-induced 2D perovskite. Meanwhile, the 2D perovskite slowly grows from the grain boundary to the 3D perovskite surface as the concentration increases.²⁷ The surface roughness of the film is also decreased by the EDBEI₂ passivating, as seen in Figure S2. X-ray diffraction (XRD) is employed to confirm the existence of 2D perovskite. As shown in Figure 2B, a peak at 5.2° assigned to (020) plane of EDBEPbI₄ perovskite was measured when the passivation concentration reaches 3 mg ml⁻¹, demonstrating the DJ-type 2D/3D perovskite heterostructure was formed.²⁸ Figure 2D shows the crystal structure of the DJ-type 2D/3D perovskite structure, such heterostructure can not only improve the stability, but also facilitate the charge transfer and increase the photovoltaic performance. Moreover, we can obtain the exciton absorption peak at ultraviolet–visible (UV–vis) absorption in Figure 2C, which further proves the formation of 2D perovskite.²⁸ And there is no obvious change in the absorption spectra, indicating the negligible change in the light absorption.

FIGURE 2. (A) SEM images of the perovskite films. (B) XRD patterns. (C) UV–vis spectrum. (D) Schematic diagram of the crystal structure

The XPS measurement was used to probe interactions between the perovskite and organic molecular of EDBEI₂. Unless stated otherwise, all characterization results are based on EDBEI₂-treated perovskite films fabricated by 3 mg ml⁻¹ EDBEI₂/IPA. As seen in Figure 3A, both the films have O1s peak. And the O1s peak at 532.4 eV appeared in the control sample due to absorption from the air,²⁹ whilst the peak at 532.7 eV in the EDBEI₂-treated sample belongs to the symbol from the metal-oxygen bond.^29,30 As shown in Figure 3B, the peak of Pb4f shifts to lower binding energy due to the interaction between uncoordinated Pb²⁺ and the carbon and oxygen chain.¹¹ Similarly, as shown in Figure 3C, the N1s shifts to higher binding energy, which is attributed to the formation of NH···I hydrogen bonds.¹⁰ The C1s spectrum shows clear molecular signal in Figure 3d, which are assigned to CO, CN, and CC bonds, respectively.³¹ The CO bond for the control film comes from the moisture from the air. The EDBEI₂-based film is not detected, indicating EDBEI₂ passivation has good moisture stability that protects from air corrosion.

FIGURE 3. XPS spectrum of the control and EDBEI2-passivated films: (A) O1s, (B) Pb4f, (C) N1s, and (D) C1s

Energy band structure

Kelvin probe force microscopy (KPFM) is conducted to characterize the surface potential (SP) of the WBG perovskite films after EDBEI₂ treatment. Seen from Figure 4A to 4D, the values of SP for the perovskite films are 540, 580, 600, and 800 mV, respectively, which is ascribed to the low surface trap density after EDBEI₂ passivation.^32,33 It is reported that SP at the grain boundary of the 3D perovskite film is different from that of the interior grains.^34,35 The SP difference is diminished in the whole EDBEI₂-treated perovskite film, indicating the defects are passivated,^36,37 thus resulting in improved fill factor. Furthermore, the surface work function of the film can be calculated from the average SP. The calculation formula is V_SP = (Φ_tip−Φ_sample)/−e,³³ where Φ_tip is the work function of the AFM tip, Φ_sample is the work function of the tested sample. Based on the formula, we calculate that the work functions are 4.66 eV and 4.60 eV for control and EDBEI₂-based samples, respectively. We also used ultraviolet photoelectron spectroscopy (UPS) to further measure the change of energy-level after EDBEI₂ passivation. The work function can be calculated as the equation of E_F = −21.22 eV−E_cutoff, in which E_cutoff is the cutoff edge for secondary electrons, and E_F is the Fermi level. Therefore, as seen in Figure 4E, the work function is calculated to be −4.70 eV and − 4.32 eV with the valance band maximum of −6.20 eV and − 5.82 eV for 3D WBG and EDBEI₂-treated perovskite, respectively (Figure 4F). Such results lead to the energy-level at the perovskite surface close to spiro-OMeTAD (−5.22 eV) (Figure 4G), which is beneficial to hole transportation and reduces the carrier recombination. As a result, the V_OC and FF are improved.

FIGURE 4. KPFM images of the EDBEI2-treated perovskite films: (A) control, (B) 1 mg ml−1, (C) 3 mg ml−1, and (D) 5 mg ml−1. UPS spectrum: (E) secondary electron cut-off spectrum, and (F) binding energy range valence spectra. (G) Energy-level alignment of the PSC

Interfacial carrier dynamics

The TRPL measurement is employed to characterize charge recombination kinetics in WBG perovskite films. Figure 5A shows the TRPL decay for control and EDBEI₂-treated films. The corresponding parameters are fitted by the bi-exponential function I_PL = A₁ exp(−t/τ₁) + A₂ exp(−t/τ₂),⁷ where τ₁ is the fast decay lifetime, which is related to the radiative recombination, τ₂ represents the trap-assisted non-radiative recombination. According to the fitting result, the values of τ₂ for the control and EDBEI₂-treated films are 110.32 and 240.41 ns, respectively, indicating the reduced trap density and suppressed non-radiative recombination. A hole-only device with a structure of ITO/PEDOT:PSS/perovskite/spiro-OMeTAD/Ag was fabricated to quantitatively evaluate the reduced defect density in the perovskite film. The trap-state-density (N_t) can be calculated by the space-charge-limited current (SCLC) method, the corresponding dark J-V curves are depicted in Figure 5B and C. According to the equation of N_t = 2εε₀V_TFL/qL² (q is the elementary charge, L is the thickness of the perovskite film, ε is the relative dielectric constant of perovskite, and ε₀ is the vacuum permittivity),^38,39 the calculated trap density for EDBEI₂-treated film is 1.8 × 10¹⁶ cm⁻³, which is lower than that of the control sample (2.7 × 10¹⁶ cm⁻³). Figure 5D depicts the linear relationship between light intensity and voltage to study the recombination mechanism. The ideality factor reflects the defect-assisted recombination.³¹ The control device has a value of 1.93, whilst the EDBEI₂-treated sample is 1.55, leading to reduced defect-assisted recombination after EDBEI₂ passivation. Such results suggest that the EDBEI₂ passivation can efficaciously passivate the surface defects, which corresponds to the TRPL results. To further understand the charge transfer kinetics after EDBEI₂ passivation, electrochemical impedance spectroscopy (EIS) measurement is performed under dark conditions. The fitted circuit model is shown in Figure 5E. Specifically, the resistance (R_s) is caused by some layers, such as the ITO layer. The charge transfer resistance (R_ct) and transport chemical capacitance (C_trans) belong to the high-frequency region. The low-frequency region is related to the recombination resistance (R_rec) and the recombination chemical capacitance (C_rec) feature in the PSCs.⁴⁰ According to the fitted results shown in Table S2, the R_tr of the EDBEI₂-treated sample is 20.06 Ω, while the control sample is 33.22 Ω, indicating that 2D/3D perovskite is conducive to effective hole transmission improving the fill factor. Meanwhile, the R_rec of the EDBEI₂-treated devices is 90 563 Ω, much higher than that of 3D devices (78 795 Ω), which is beneficial to inhibiting charge recombination and facilitating charge transfer at the perovskite/HTL interface in the EDBEI₂-treated perovskite. Such result is also attributed to the increased V_OC after EDBEI₂ passivation. The capacitance–voltage (C-V) measurement is further used to evaluate the relationship between the improved V_OC and built-in potential (V_bi). As depicted in Figure 5F, the V_bi can be obtained from the Mott-Schottky plot.⁴¹ The EDBEI₂-treated device delivers a higher V_bi of 1.05 V, whereas the control device is 0.94 V. The higher V_bi is beneficial to facilitating charge separation and collecting photogenerated carriers at the perovskite/hole interface, hence, both the V_OC and FF in the EDBEI₂-passivated PSCs are improved.

FIGURE 5. (A) TRPL spectrum. (B, C) Space-charge-limited current curves for the hole-only control and EDBEI2-treated devices. (D)Light-intensity-dependent VOC curves for the hole-only control and EDBEI2-treated devices. (E) EIS plots. (F) Mott-Schottky plots of both control and EDBEI2-treated devices

Stability of the EDBEI₂-treated WBG perovskite

Structural stability is another crucial challenge for high-efficiency WBG PSCs. 2D perovskite shows intrinsic stability on account of the protection of the hydrophobic spacer cations. Therefore, to investigate the humidity stability after EDBEI₂ passivation, the water contact angle at the RH of 30% ± 10% after 1000 h aging is tested in Figure 6A and B. The contact angle of EDBEI₂-treated film is 47.52°, whereas the control film shows a smaller contact angle of 37.07°, indicating improved hydrophobicity after EDBEI₂ passivation due to the advantage of in-situ formed DJ-type 2D perovskite. Additionally, to further explore the desirable stability to prevent water penetration, the unencapsulated devices in a confined space at 30% ± 10% RH were tested. Seen from the PCE attenuation diagram in Figure 6C, the EDBEI₂-treated cells retain 89% of their original efficiencies after 1000 h aging, whilst the control devices only maintain 71% of their original efficiency in the same environment. In addition to the enhanced humidity stability, thermal stability also exhibits good result. All the perovskite films and unencapsulated solar cells are annealed at the hot plate at 85°C in the N₂ atmosphere. As seen in Figure 6D and E, the control film is slightly decomposed after aging for 250 h at 85°C, resulting in a PbI₂ phase in the XRD pattern, whilst the 2D/3D film is relatively stable in the same condition. As a result, the 2D/3D devices show better thermal stability than the control devices, which experiences only 28% PCE loss at high-temperature condition (seen in Figure 6F). The above analysis demonstrates that such DJ-type 2D/3D perovskite heterostructure can lead to remarkable durability of the WBG PSCs.

FIGURE 6. Water contact angles of both films exposed to the RH with 30% ± 10% for 1000 h: (A) control film, (B) EDBEI2-film. (C) PCE attenuation diagram of unencapsulated devices at 30% ± 10% RH. (D–E) XRD patterns of both films after 85°C aging in N2 atmosphere. (F) PCE evolution of unencapsulated devices exposed to 85°C in N2 atmosphere

CONCLUSIONS

A DJ-type 2D/3D perovskite heterostructure using 2,2′-(Ethylenedioxy) diethylammonium diiodide (EDBEI₂) as the surface passivating agent is successfully fabricated. A DJ-type 2D perovskite (n = 1) was in-situ formed on top of the 3D WBG perovskite film, which effectively passivates surface defects and suppresses the non-radiative recombination. A combination of UPS, Mott-Schottky, and EIS analysis reveals improved arrangement of energy-level alignment and reduced defect density, which contributes to the effective charge transfer in the EDBEI₂-treated PSCs. As a result, the EDBEI₂-treated device delivers a PCE of 18.85% with a high V_OC of 1.28 V, J_SC of 18.77 mA⋅cm⁻², and FF of 78.50%. Importantly, the robust DJ-type 2D/3D cell delivers enhanced stability in the humid and thermal atmosphere, which retains 89% of its initial efficiency after storage for 1000 h under a RH of ~30%. This work provides important insight into designing DJ-type 2D/3D heterostructure by using diammonium spacer cations to fabricate durable and high-efficient WBG PSCs.

ACKNOWLEDGMENTS

The authors acknowledge support from the Sichuan Science and Technology Program (2021YFH0090), and the Graduate Student Scientific research Innovation Fund of Southwest Petroleum University (2021CXZD27).

CONFLICT OF INTEREST

The authors declare no conflict of interest.