Two-dimensional Ruddlesden–Popper layered perovskite for light-emitting diodes

Metal halide perovskites have undergone rapid development since their first employment as photoactive layers by Kojima and co-workers in 2009.¹ Their emergence as promising emitters in perovskite light-emitting diodes (PeLEDs) is attributed to their superior properties, such as high photoluminescence quantum yield (PLQY), narrow light emission full width at half maxima (FWHM), tunable exciton binding energy, balanced charge carrier mobility,^2,3 as well as relatively low cost and facile large-scale manufacturing.^4–7 Actually, a layered perovskite emitter was applied in PeLEDs as early as 1994; however, the electroluminescence emission was only achieved at liquid-nitrogen temperature.⁸ In 2014, Tan and co-workers explored three-dimensional (3D) perovskites (CH₃NH₃PbI_3−xCl_x and CH₃NH₃PbBr₃) as emitters to realize PeLEDs at room-temperature,⁹ indicating that the research of PeLEDs is on track.

The external quantum efficiency (EQE) of the PeLEDs rockets over 20% in 5 years, which reinforces the belief that PeLEDs are a good alternative to conventional organic light-emitting diodes (OLEDs).^10–14 Although the improvement of the EQE is overwhelming, researchers found that the intrinsically vulnerable nature to moisture and oxygen of perovskite materials caused by their structural characteristics and the severe trap-assisted non-radiative recombination due to the poor film quality are critical factors limiting PeLEDs to industrialization.^15–18 

Two-dimensional Ruddlesden–Popper (2DRP) perovskites with a layered structure show better stability without sacrificing much EQE compared to their 3D counterparts, which offers a new strategy for fabricating PeLEDs with high efficiency and long-term stability.^19–21 The 2DRP perovskites take the formula of A′₂A_n−1B_nX_3n+1, where A′ is a large aliphatic or aromatic ammonium cation (R–NH₃⁺, such as butylammonium, phenethylammonium, etc.), A is a small covalent cation (Cs⁺, MA⁺, and FA⁺), B is a divalent metal that can adopt an octahedral coordination (Pb²⁺, Sn²⁺, Ge²⁺, etc.), X is a halogen (Cl⁻, Br⁻, and I⁻), and n is the number of inorganic sheets.^22–24 Compared with the 3D perovskite structure (AMX₃), the distance between the inorganic sheets in 2DRP perovskites can be tuned by incorporating various large organic cations (R–NH₃⁺).²⁵ The formation of the layered perovskite can be considered as a 3D perovskite being cut into thin slices with different values of n (n = 1, 2, 3, …, ∞) in different crystallographic directions.^26–30 In a layered structure, the bandgap is affected by the number of inorganic sheets; therefore, the 2DRP perovskites with different n values can be prepared by adjusting the stoichiometric ratio of the perovskite precursor solution.³¹ A multiple quantum well structure is formed with the inorganic sheets as wells and the organic layers as barriers (Fig. 1).^31–33 Benefiting from the structural characteristics of the 2DRP perovskites, the energy transfer becomes faster and more efficient, the defect-induced non-radiative recombination is effectively suppressed, and the PLQY is improved,^27,34–36 suggesting that the 2DRP perovskites are well-suited to PeLEDs.

In this perspective, we start with the advantages of the 2DRP perovskites over their 3D counterparts. We then discuss how the introduced organic molecules interact with the octahedral structure of the perovskites (such as the increased binding energy due to stronger interactions and the hydrophobicity of the end of the organic molecular chain) to improve the stability of the devices and the carrier transport in terms of layered structures. In addition, the contributions to the improvement of the PLQY and EQE are also introduced. Finally, we summarize the progress of the stability and EQE of PeLEDs in recent years, as well as the challenges that still need to be overcome, and propose the potentials of PeLEDs as a new generation light source of LEDs.

In this part, we will describe the unique characteristics of the 2DRP perovskites in terms of their excellent stability, improved luminescence efficiency, good film morphology, and tunable exciton binding energy.

Stability has always been a bottleneck restricting the development of perovskite-based optoelectronic devices.^{18,21,37–39} The 2DRP perovskites with a layered structure show a good stability from the unique structural characteristics and intermolecular interactions compared to their 3D analogs.^40–42 Here, we discuss the improvement of the stability by analyzing the perovskite structures after inserting the organic layers.

Moreover, both experimental and theoretical studies have shown that the lower activation energy of hybrid organic–inorganic perovskites leads to the inherent existence of ion migration.^48–51 Yang et al. reported that the ionic conductivity is higher than the electronic conductivity in MAPbI₃ (an activation energy of ≈0.4 eV).⁴⁸ The large ionic transference number contributes to a high native ionic disorder and substantial trapping of the electronic carriers. Yuan et al. concluded that the conductivity of MAPbI₃ in the dark consists of two regions: the first assigned to electronic carriers (an activation energy of ≈40 meV) and the second assigned to ionic migration (an activation energy of ≈0.4 eV).⁵² The similar activation energies of these results present the evidence for ion migration. For perovskites, migration barriers are 0.1–0.6 eV for iodine vacancies and 0.5–0.9 eV for MA⁺ migration.^53–55 The presence of a large amount of ion migration in the perovskite film will accelerate the degradation of the perovskite films, where vacancy defects and non-radiative recombination centers are formed by the degradation of the transport layer and corrosion of the electrode.⁵⁶ Lin et al. performed transient and steady-state PL and photoresponsivity analysis on perovskite thin films, and the results show that the formation of the 2DRP perovskite can effectively suppress the ion migration and reduce non-radiative recombination centers.⁵¹ A 2DRP (BA)₂(MA)₃Pb₄I₁₃ film is formed when introducing the organic n-butylammonium (BA) spacers, and the ion migration activation energy increases to 41 ± 7 meV compared to those of its 3D counterparts with 0.19 eV in the dark [Fig. 2(c)]. The introduction of the large organic cations may contribute a special geometry of crystals in 2DRP perovskites, blocking the ion migration in the films.

The competition between the trap-assisted non-radiative recombination/Auger recombination and the bimolecular radiative recombination exists in the entire luminescence process of the PeLEDs. In 3D perovskites, the small exciton binding energy (only 9–60 meV) leads to the fact that most excitons decompose into free charges at room temperature, and their bimolecular recombination is even slower than the charge diffusion.^57–60 The severe trap-assisted non-radiative recombination results in the low PLQY and EQE of 3D PeLEDs.^60–63 In the 2DRP perovskite with layered structures, the organic layers act as insulating barrier layers, which can enhance the electron–hole Coulomb interaction and manifest the quantum confinement effect.⁶⁴ Therefore, this perovskite multiple quantum well can experience rapid energy transfer and substantially avoid exciton quenching, achieving effective radiation recombination.⁶⁵ Therefore, the layered 2DRP perovskite materials show a high PLQY compared to those of their 3D counterparts.^31,66

Yuan et al. achieved an appropriate energy level distribution by adjusting the stoichiometry of the precursor solutions to ensure that the carriers were effectively injected into the lowest energy levels, thereby, improving the radiation recombination rates.²⁷ Eventually, they obtain the PeLEDs with an EQE of 8.8%. Zhao et al. used the 2DRP perovskite (NMA)₂(FA)Pb₂I₇ as an emitter, improving the EQE of the PeLEDs to 20.1%.⁶⁷ The perovskite–polymer bulk heterostructure thin films with or without charge-transport contacts exhibit a PL decay tail lifetime of ∼1.5 µs, which indicates low trap densities in these structures.^68,69 The improved EQE and near-unity PLQY (∼96% under 100 mW cm⁻²) originate from the suppressed non-radiative recombination. These outstanding device performances are attributed to the fast energy transfer in the perovskite multiple quantum wells, avoiding exciton quenching and leading to effective radiative recombination.

In addition to fast and efficient energy transfer, the increase in the bimolecular recombination rate of the 2DRP perovskites is another factor that affects the EQE. The values of carrier recombination rates in the 3D and 2DRP perovskites were compared. Xing et al. observed that the PL lifetime of the carriers in the 2DRP perovskites is about two orders of magnitude shorter than that in the 3D perovskites at a low excitation intensity [Figs. 3(a)–3(d)].⁶⁰ The rapid recombination in the 2DRP perovskites can also inhibit the trap-assisted non-radiative recombination and lead to a high PLQY at a low excitation intensity.

The pinholes in traditional 3D perovskite thin films lead to large leakage currents. In addition, these pinholes cause Joule heating under electrical stress, resulting in a rapid decrease in EQEs in the high current density regime. In contrast, the interaction between large organic cations and perovskites in 2DRP perovskite structures can promote the formation of dense films.^35,70 Therefore, the 2DRP perovskite thin films exhibit high coverage and low roughness (Fig. 4).³¹ It should be noted that Cao et al. demonstrated that discontinuous 2DRP perovskite platelets (not pinholes) can still realize high-performance PeLEDs due to a thin organic layer filling in the gaps between the perovskite platelets.¹⁰ The perovskite films are still pinhole-free, flat, and compact. In fact, the morphology is significantly improved in their work.

Compared to 3D perovskites, the binding energy of 2DRP perovskites is large due to the quantum confinement effect, leading to the formation of excitons rather than free carriers.⁷¹ Generally, the quantum confinement effect increases as the thickness of the inorganic layer decreases, which results in the tunable exciton binding energy to achieve the best optical performance and the desired color gamut of the PeLEDs.⁷² As shown in Figs. 5(a)–5(c), the recombination of injected electrons and holes is enhanced for a large exciton binding energy, contributing to the improved performance of PeLEDs.^44,73–75 In polymorphic mixed perovskites, electron–hole enrichment exists in the large n-phase because of the slow hole transport rate, and the large n-phases act as the radiative recombination centers to achieve high EQE PeLEDs.^76–78 

The harsh working conditions of the first PeLEDs based on 2D (PEA)₂PbI₄ reported in 1990s limit their practical applications.⁸ The first near-infrared (NIR) 2DRP LED with high efficiency was successfully fabricated at room-temperature by Yuan et al. in 2016.²⁷ In the same year, Wang et al. reported a strategy by introducing 1-naphthylmethylamine iodide (NMAI) into the inorganic octahedron structure as a separation layer to form a multiple quantum well structure which improves the recombination rate of electrons and holes.³¹ The emergence of these achievements usher the development of the LEDs based on 2DRP perovskites.^79,80 In this part, we will review the development of 2DRP PeLEDs in terms of interfacial engineering, component engineering, and light output coupling engineering.

The surface condition of the perovskite films is one of the key factors affecting the performance of PeLEDs. The loss of efficiency in PeLEDs is usually ascribed to the non-radiative recombination which is due to the point defects in the lattice and trap states at the surface and grain boundaries of the perovskites.⁸¹ The energy level barrier between the contact layers and the perovskite emissive layer is an important factor to improve the device performance.⁶³ Therefore, interfacial engineering has been widely considered as an effective method to inhibit the non-radiative recombination and, thereby, improve the performance of the obtained PeLEDs.^82,83

Wang et al. reported that the polyethyleneimine (PEI) interlayer between the electron transporting layer and the perovskite emissive layer can afford a low work-function cathode for efficient electron injection into the perovskites [Figs. 6(a) and 6(b)].⁸⁴ Gao et al. demonstrated the critical role of the interfacial reaction on achieving high-efficiency PeLEDs by modifying the electron injection layer with an ultra-thin polyethylenimine ethoxylated (PEIE) layer [Figs. 6(c) and 6(d)].⁸⁵ You and co-workers achieved high-performance 2DRP PeLEDs by the surface passivation to reduce the non-radiative recombination on the perovskite surface or grain boundaries [Figs. 6(e) and 6(f)].⁷⁶ 

The perovskite family has been greatly enriched by the introduction of large organic cations. The large organic cations include aliphatic ammonium, aromatic amines, and the amines with the functional groups. The distribution of n values, carrier transport, and film stability are affected by the size and spatial configuration of large organic cations and their interactions between the large organic cation and the perovskite structure.^86–89 

Wang et al. improved the morphology of the perovskite thin film by introducing 1-naphthylmethylamine iodide (NMAI) into the precursor solutions.³¹ They point out that the lower bandgap regions generating electroluminescence are effectively confined by the layered perovskite with higher bandgaps, contributing to a very efficient radiative decay. The PeLEDs with an EQE of 11.7% are prepared based on this system [Fig. 7(a)].³¹ In the same year, Yuan and co-workers add the bulky cation phenylethylammonium (PEA = C₈H₉NH₃) into the precursor solutions to form the 2DRP perovskites. Dynamic studies show that the layered structures enable a more efficient radiative recombination and overcome the trap-mediated non-radiative recombination [Fig. 7(b)].²⁷ Xiao et al. reported a solution process to form highly uniform and smooth perovskite films by adding n-butylammonium halides (BAX, X = I, Br) into the perovskite precursor solutions [Fig. 7(c)].⁷⁰ The PeLEDs based on BAX-incorporated layers show greatly improved device performance and stability. Xu et al. compared the difference of large organic cations with different terminal groups (such as –NH₂, –O, and –N) when they are added into the precursor solutions. They demonstrate that when the –O and –NH₂ coexist, the non-radiative recombination is effectively reduced, and they obtain high-efficiency near-infrared PeLEDs with a peak EQE of 21.6% [Fig. 7(d)].¹⁴ 

Cao and co-workers reported that the perovskites can spontaneously form the submicrometer-scale structure by introducing amino-acid additives into the perovskite precursor solutions (Fig. 8).¹⁰ The submicrometer-scale structure can efficiently extract light from the device and retain wavelength- and viewing-angle-independent electroluminescence, as well as effectively passivate perovskite surface defects and reduce the non-radiative recombination. Eventually, the PeLEDs with an EQE of 20.7% are achieved based on this new structure.

In this perspective, we have listed the inherent advantages of the PeLEDs based on 2DRP perovskites over their 3D counterparts. Although the PeLEDs based on the 2DRP structure exhibit the improvement of stability and EQE, there are still some limitations for their commercial applications.

First, the stability of devices is still one of the most critical factors. Either the intrinsically vulnerable nature to moisture and oxygen or the heavy ion migration and the thermal degradation under the continuously operating voltage will restrict their long-term and stable work. Although the large organic molecule in the 2DRP structure has efficiently protected perovskite films from moisture and oxygen, the device operation stability is still limited by the defect associated ion migration. Another instability factor for the PeLEDs is related to possible instability of the charge transport layers. The all-inorganic charge transport layers such as ZnO and NiO offer better stability, and they have often been used to improve the device stability.^9,10,90–92 Stability issues can be resolved via proper encapsulation, but it is the intrinsically vulnerable nature of metal halide perovskite materials which holds top priority to be solved.

Second, although the record efficiency of PeLEDs has exceeded 20%, the pure-blue and white PeLEDs still need further studies since the record EQEs are not over 5%.^93–96 Recently, researchers have made great progress in sky-blue PeLEDs with an EQE over 10%, which proves that there is much more potential for pure-blue perovskite materials.⁹⁷ Effective charge transport can improve the EQE, which, however, would lead to the excess carrier accumulation, thereby resulting in charge quenching. Moreover, the development of the pure-blue emitter with a less quantum confinement effect is important to reduce the Auger recombination.^98,99

Third, as it is known that the use of Pb is harmful to the human body and the environment, the highly efficient lead-free perovskite emitters with different emissive wavelengths are highly demanding. The 2DRP perovskites have shown a prospect to enhance the performance of Sn-based NIR PeLEDs.¹⁰⁰ However, both the PLQY and charge transport of Sn-based perovskites are more inferior than those of the Pb-based perovskites. Moreover, the current lead-free perovskites for NIR PeLEDs are mainly focused.^101–104 Achieving efficient, stable, and non-toxic PeLEDs with different visible emissions is vital to prompt their application in lighting and displays.

Finally, in terms of commercial applications, the preparation of large-area devices is an indispensable link.¹⁰⁵ At present, perovskite-based large-area photovoltaic devices are concentrated on solar cells; few studies, however, have been made on large-scale PeLEDs. Therefore, it is necessary to pay more attention to the fabrication techniques for large-scale production. Although challenges still exist, PeLEDs will usher in a new era when the efficient and stable red, green, and blue LEDs are achieved, and the technology to prepare large-area flexible displays can be realized at low cost.

This work was financially supported by the National Key R&D Program of China (Grant No. 2017YFB1002900), the Natural Science Foundation of China (Grant Nos. 51972172, 61705102, and 91833304), Jiangsu Specially-Appointed Professor program, Young 1000 Talents Global Recruitment Program of China, and “Six talent peaks” Project in Jiangsu Province China.