Quantum Yield and Stability Improvement of Two‐Dimensional Perovskite Via a Second Fluorinated Insulator Layer

Aiming at the low luminous‐efficiency of two‐dimensional (2D) perovskite, its quantum yield (QY) and stability are effectively improved by a fluorinated second insulator layer 3,4‐difluoroaniline (F2PA) outside of the first insulator 2‐(2,4‐difluorophenyl) ethylamine (FPEA). The QY of perovskite with two insulator layers is improved by 3.4‐time in contrast with its counterpart with one insulator, which originates from the defect suppression and reduction of exciton Bohr radius by the second layer. The optical stability has extended 1.5‐time by the introduced second layer, as an isolation role for inner decomposed products to volatilize and for outer water molecules to penetrate. The perovskite with two insulator layers retained 90% of its initial optical properties in relative humidity 90%–95% atmosphere for 600 h, meanwhile only 50% for its counterpart with one layer. Two‐more fluorine with stronger hydrophobicity can account for this feature. Adding a second insulator layer is an effective strategy to engineer 2D perovskites and push forward its applications in optoelectronic devices.


Quantum Yield and Stability Improvement of Two-Dimensional Perovskite Via a Second Fluorinated Insulator Layer
Pengfei Ma, Yuan Zhuang, Hui Lin, Guanjun You, Gongjie Xu,* and Bin Cai* DOI: 10.1002/admi.202202003 2D RPP for optoelectronic devices, such as solar cells, LEDs, microlasers, and photodetectors. [9][10][11][12] However, the photoluminescence quantum yield (PLQY) of 2D RPP with one octahedral layer (n = 1) is only ≈0.5%, [13][14][15] much less than that of the quantum nanocrystal counterpart, which can achieve to 95%. [16] The reported value for 2D RPP also achieved to 6%, [15] and 9%, [17] even near 100%. [18] It should be noted that the mainstay of the above radiative process is the self-trapped exciton, [17][18] rather than the free exciton in 2D RPPs. According to the defect theory in low-dimensional perovskites, [19] the surface sites of interstitial and vacancy defects form deep levels, which are detrimental to radiative transition. [20] So far, Li et al. improved PLQY by encapsulating graphene on (BA) 2 PbI 4 (where BA stands for n-butyl amine) surface to reduce quantum confirmed Stark effect, [14] which realized 28-time enhancement of PLQY. Dhanabalan et al. obtained 2-3 times PLQY improvement of 2D RPP by mechanically exfoliating flakes owing to the low-defects feature. [21] Pang et al. achieved high-efficiency skyblue LEDs by rearranging low dimensional phases in quasi-2D perovskites via sodium ion doping, which can obviously reduce the n = 1 phase, forming other small-n phases with higher PLQY, to enhance the external quantum efficiency. [22] Gong et al. realized 79% efficiency from exfoliated Br-based RPPs via controlling crystal rigidity and electron-phonon interactions. [23] M Abdelhamied et al. obtained near 3-time improvement of (PEA) 2 PbI 4 (where PEA stands for phenethylamine) PLQY via polymer encapsulation and passivation. [24] In fact, the strategy to increase the PLQY of n = 1 2D RPP L 2 PbI 4 is still limited. [14] Here, we reported a strategy to improve the PLQY and stability of 2D RPP L 2 PbI 4 via adding a second organic insulator layer, as illustrated in Figure 1a. The fluorinated cation (2-(4-fluorophenyl) ethylamine (FPEA)) was employed as the intrinsic organic layer as left panel of Figure 1b, which was proved to be effective for stability but with partial sacrifice PLQY. [25] For the second layer, 3,4-difluoroaniline (F 2 PA), marked by dashed boxes in Figure 1, was utilized for the three following goals. 1) Stronger stability performance is expected due to the introduction of more-fluorinated insulators. 2) The smaller mobile freedom of halide means less vacancies, which would produce detrimental deep levels for higher PLQY. [20,26] 3) The low permittivity of the fluorinated second insulator leads Aiming at the low luminous-efficiency of two-dimensional (2D) perovskite, its quantum yield (QY) and stability are effectively improved by a fluorinated second insulator layer 3,4-difluoroaniline (F 2 PA) outside of the first insulator 2-(2,4-difluorophenyl) ethylamine (FPEA). The QY of perovskite with two insulator layers is improved by 3.4-time in contrast with its counterpart with one insulator, which originates from the defect suppression and reduction of exciton Bohr radius by the second layer. The optical stability has extended 1.5-time by the introduced second layer, as an isolation role for inner decomposed products to volatilize and for outer water molecules to penetrate. The perovskite with two insulator layers retained 90% of its initial optical properties in relative humidity 90%-95% atmosphere for 600 h, meanwhile only 50% for its counterpart with one layer. Two-more fluorine with stronger hydrophobicity can account for this feature. Adding a second insulator layer is an effective strategy to engineer 2D perovskites and push forward its applications in optoelectronic devices.

Introduction
Two-dimensional (2D) Ruddlesden-Popper perovskite (RPP), sometimes called 2D perovskite in short, has a general formula L 2 A n−1 B n X 3n+1 , where L is organic amine cation, A is cation (MA + , FA + , or Cs + ), B is metal cation (Pb 2+ , Sn 2+ ), X is halide ion (Cl − , Br − or I − ), [1,2] and n represents the number of octahedral [BX 6 ] 4− inorganic layers. [3][4][5][6] This type of perovskite has remarkably higher moisture resistance, the structural flexibility [7] and narrow absorption, [8] and large binding energies originating from the reduced Coulomb screening effect due to the low-dielectric constant of organic part. [9] Based on these advantages, many efforts were devoted to developing www.advmatinterfaces.de to a smaller exciton Bohr radius, which implies a higher capability to accommodate a higher exciton density, [27] as shown in the right panel of Figure 1b. By this strategy, we effectively improved the PLQY of the 2D RPP (F 2 PA) 2 /(FPEA) 2 PbI 4 to 0.78%, which is 3.4-time higher than that of its counterpart (FPEA) 2 PbI 4 . The PLQY was measured by an integration sphere in the form of powder, rather than in solution or singlecrystalline. [23] Meanwhile, the stability for optical and humidity was also demonstrated effectively by this interface engineering.

Preparation and Structures of 2D Perovskite with Two-Layer Insulators
The synthesis of (F 2 PA) 2 /(FPEA) 2 PbI 4 consists of two steps. The first step was to prepare (FPEA) 2 PbI 4 crystals by the solventantisolvent method as references. [7,25] The second step was to add F 2 PA (1:1 molar ratio with FPEA) into previously obtained samples. The details of samples synthesis can be referred to the chemical's synthesis in the Experimental part of Supporting Information. Three samples (PEA) 2 PbI 4 , (FPEA) 2 PbI 4 , and (F 2 PA) 2 /(FPEA) 2 PbI 4 were synthesized and studied in this work. The prepared three samples were shown in Figure S1, Supporting Information, in the visible and ultraviolet field, where the (F 2 PA) 2 /(FPEA) 2 PbI 4 had a better performance under the ultraviolet excitation.
For the structural changes, we performed X-ray diffraction (XRD), as shown in Figure 2a. All three powder samples have similar diffraction peaks at almost the same angles, which demonstrates the same phase Ruddlesden-Popper perovskite. [4,13,25] The diffraction angles 2θ of (PEA) 2 PbI 4 are 10.64°, 16.02°, 21.44°, and 26.9° also in Table S1, Supporting Information, corresponding to Miller index (004), (006), (008) and (0010), respectively, which is the same with reported one. [28] Two changes are notable for (FPEA) 2 PbI 4 and (F 2 PA) 2 / (FPEA) 2 PbI 4 . First, the diffraction intensity decreased and full width at half maximum (FWHM) broadened in Figure 2a since the low-affinity of fluoride makes it hard to form largesize crystals, which can be seen from their reduced quality of microplate crystals in Figure S1, Supporting Information, and thin-film states by Figure S2 in Supporting Information, consequently, diffraction intensity decreases as the fluorine quantity increases. The relative crystallinity for (PEA) 2 PbI 4 , (FPEA) 2 PbI 4 , and (F 2 PA) 2 /(FPEA) 2 PbI 4 is 62.73%, 58.32%, and 48.73% respectively, details in Table S2, Supporting Information. Second, all diffracted angles have shifted slightly to a lower position for (FPEA) 2 PbI 4 and (F 2 PA) 2 /(FPEA) 2 PbI 4 in comparison with that of (PEA) 2 PbI 4 , as summarized in Figure 2b. This phenomenon originates from the insertion of the second insulator F 2 PA, which leads to a little adjustment of the position of cation FPEA + relative to the octahedral framework, consequently a larger spacing d hkl of the plane with Miller index (hkl). Here the distance along the c-axis direction of the octahedral structure is enlarged by the adjusted cation. Therefore, the smaller diffraction angles were obtained according to Bragg's law 2d hkl sinθ = mλ, where θ is the diffraction angle, m is an integer, λ is the wavelength of the incident X-ray. The calculated data (Table S1 in Supporting Information for details) are consistent with the above conclusions. [4] For the large-size crystal growth of (F 2 PA) 2 /(FPEA) 2 PbI 4 , further efforts are needed to be devoted for developing another method.
About the position of the inserted F 2 PA relative to the FPEA, here we list three reasons to confirm in a way like that in Figure 1a. From the synthesis view, when the F 2 PA is added, the FPEA is already existed in (FPEA) 2 PbI 4 as an organic insulator, therefore the stirring procedure plays a role of F 2 PA distribution around the original (FPEA) 2 PbI 4 . From the bonding view, the charge (Huckel) of N in FPEA and F 2 PA is −0.289 e and +0.108 e respectively (calculated from ChemOffice, listed in Table S3 of Supporting Information), which means that FPEA could form FPEA + after protonation, then as an L constitute in RPP L 2 PbI 4 , but F 2 PA can't owing to its positive attribute, also verified in experiment. What we cannot ignore is that F (−0.115 e, listed in Table S3, Supporting Information) in FPEA   Table S3, Supporting Information) in F 2 PA can combine by a hydrogen bond, which was marked by the dash line in Figure 1a between F in FPEA and the H attached to N in F 2 PA. From the above structural data and the following optical properties, the insertion of the second organic layer produces structural alternation and bandgap modification. Thus, the second organic insulator have been successfully introduced into the 2D RPP.
For the changes in morphology, we performed scanning electron microscopy (SEM) characterization for microplate samples by self-assembly (details in the Experimental part), shown in Figure S2a-c, Supporting Information. From (PEA) 2 PbI 4 , (FPEA) 2 PbI 4 , to (F 2 PA) 2 /(FPEA) 2 PbI 4 , the microplate surface becomes rough, which is also obvious in their thin-film state ( Figure S2d-f, Supporting Information). This phenomenon originates from the extremely-low polarity of fluorinated molecules, which can effectively reduce the surface energy of compounds, and fluorinated alkanes exhibit very poor affinity with other materials. [29] Meanwhile, the energy dispersive spectroscopy (EDS) for (FPEA) 2 PbI 4 and (F 2 PA) 2 /(FPEA) 2 PbI 4 was obtained ( Figure S3 in Supporting Information for details), and the mappings for elements F, Pb and I of these two samples were presented in Figure 2c-d. The data of atomic fraction was listed in Table S4, Supporting Information, where the ratio of F, Pb, and I is ≈1.36:1.00:4.83 for (FPEA) 2 PbI 4 , and it is 2.51:1.00:5.94 for (F 2 PA) 2 /(FPEA) 2 PbI 4 . Although there exists a deviation from the stoichiometric ratio of 6:1:4 for (F 2 PA) 2 /(FPEA) 2 PbI 4 due to the inaccurate EDS in SEM for F with small-mass, it is obvious that the relative quantity of F becomes 2.51, in comparison with 1.36 of (FPEA) 2 PbI 4 . In order to investigate the bonding of different elements in perovskites, X-ray photoelectron spectroscopy (XPS) for the above three samples were conducted, and N 1s and F 1s spectra were shown in Figure 2e-f respectively, while other data was given in Figure S4 in Supporting Information. According to chemical shifts of photoelectron, [30] from the spectrum of N 1s, the binding energy peak of -NH 3 + shifts from 401.31 eV in (FPEA) 2 PbI 4 to 401.44 eV in (F 2 PA) 2 /(FPEA) 2 PbI 4 , originating from the introduction of more strong-electronegativity element F in F 2 PA. Besides, the peak locates at 398.97 eV in (F 2 PA) 2 /(FPEA) 2 PbI 4 , which can be ascribed to -NH 2 in agreement with the reported value. [31] For the binding energy change of F 1s, the value shifts from 686.74 eV in (FPEA) 2 PbI 4 to 685.51 eV in (F 2 PA) 2 /(FPEA) 2 PbI 4 , since for the F in FPEA for the latter forms a hydrogen bond with chemically reducing -NH 2 in F 2 PA, referred to Table S3, Supporting Information. By integration of every element in XPS, the relative atomic ratio was also obtained, given in Table S4, Supporting Information. Because the performed spot in powder is randomly selected, inhomogeneity of leads to a large discrepancy with the ideal case.

Optical Performance of 2D Perovskite with Two-Layer Insulators
The optical properties of the three samples were shown in Figure  3. From the normalized absorption, the peaks of (PEA) 2 PbI 4 , (FPEA) 2 PbI 4 , and (F 2 PA) 2 /(FPEA) 2 PbI 4 are 514 nm, 517 nm, and  Figure 3a. And the photoluminescence (PL) peaks located at 520 nm, 521 nm, and 524 nm correspondingly in Figure 3b. The red-shift of absorption and PL peaks originates from the reduced Coulomb screening. [25,27] The polarizability of fluorinated molecules is smaller than its counterpart, leading to the smaller effective dielectric constant. [32] For the two samples, (FPEA) 2 PbI 4 and (F 2 PA) 2 /(FPEA) 2 PbI 4 , due to more F introduction in the insulator of the latter, its effective constant ε 2 is smaller than that of the former ε 1 , consequently, the reduced Coulomb screening, as the Figure 3c depicted. Therefore, the binding energy of excitons in 2D perovskite is increasing as the F content in insulators increases. Thus, for the same energy band, the increasing binding energy indicates smaller transition energy, namely the photon energy of absorption and PL. [25,33] Other analysis, like lattice strain [34] induced by F 2 PA, probably is applicable here. According to the following relationship a* ∝ ε r × a H , where a*, ε r , and a H is effective Bohr radius of an exciton, relative dielectric constant, and Bohr radius of hydrogen atom respectively, [35] therefore, the effective Bohr radius of exciton in (F 2 PA) 2 /(FPEA) 2 PbI 4 becomes smaller than that in (FPEA) 2 PbI 4 , as shown in Figure 3c and also in Figure 1b.
For an optimal ratio of two insulators between FPEA and F 2 PA, a set of PLs was obtained in Figure 4a. The ratio 1:1 is the best option, which also indicates that every F 2 PA layer lies exactly outside every FPEA layer. Besides, as the component of F 2 PA increases, the PL peak shows a little red-shift, which is consistent with the results of Figure 3. The intensity of the blue curve with ratio of 1:1 is ≈3.3-time of that of the black one without F 2 PA. To verify the direct improvement of quantum efficiency by F 2 PA, we performed the PLQY of the three samples (PEA) 2 PbI 4 , (FPEA) 2 PbI 4 and (F 2 PA) 2 /(FPEA) 2 PbI 4 , which are 0.27%, 0.23%, and 0.78% respectively, shown in Figure 4b and details in Figure S5, Supporting Information. The introduction of the second insulator F 2 PA into (FPEA) 2 PbI 4 has improved its PLQY up to 3.4 times. The improvement by the second insulating layer can be ascribed to the reduction of halide vacancies, which limits the QY of 2D perovskite systems. [20] Maybe the reduced self-absorption ( Figure S6 in Supporting Information) in (F 2 PA) 2 /(FPEA) 2 PbI 4 also contributes to this improvement. By comparing the radiative recombination ratio of three samples, we can get a conclusion in more details. The PL lifetimes of excited excitons, obtained by time-correlated single photon counting technique (TCSPC), were 0.30 ns, 0.33 ns, and 0.42 ns respectively (Table S5, Supporting Information, for details) for the same pumping power in Figure 4c. The longer lifetime may originate from the reduced Bohr radius of the exciton. Along with the PLQY in Figure 4b, according to the equation [36] K r = η/τ, and K nr = K r /η-K r , where K r , η, τ, and K nr is radiative recombination rate, PLQY, PL lifetime, and non-radiative recombination rate respectively. The calculated data was given in Table S6, Supporting Information, in detail. In Figure 4d, the K nr (black squares) of (F 2 PA) 2 /(FPEA) 2 PbI 4 is the lowest, and the K r (red squares) is the highest, so is the ratio of K r / K nr (blue squares). To further confirm the influence of the second insulator on the energy band structure, an ultraviolet photoelectron spectroscopy (UPS) was performed (HeI 21.22 eV), shown in Figure 4e and Figure S7, Supporting Information. According to Figure S7b, Supporting Information, we obtained the work functions of three samples as 5.17 eV, 4.87 eV, and 4.72 eV, and the highest occupied states (HOSs) as 1.24 eV, 0.98 eV, and 0.88 eV respectively from Figure 4e. Therefore, we can calculate the valence band maximums (VBMs) of the three samples as 6.41 eV, 5.85 eV, and 5.60 eV respectively. As schematically depicted in Figure 4f, the VBM reflects the reduction of defects, which lies within the valence band or near the edge. [37] This indicated the PLQY improvement of 2D perovskites by the second insulator from the view of energy band. Besides, the F 2 PA "shell" layer can effectively reduce the possibility of I − migration out of octahedral structure [PbI 6 ] 4− , namely forming halide vacancy. In summary, the second insulator makes the 2D perovskite form smaller crystals, in which the region of nonradiative recombination by vacancy defect becomes smaller as well as the effective exciton Bohr radius. This result is similar to that the PLQY of perovskite quantum dots is larger than that of bulk crystals.
From a view of carrier dynamics, we can also analyze the role of the second insulator F 2 PA. The PL lifetimes of (PEA) 2 PbI 4 , (FPEA) 2 PbI 4, and (F 2 PA) 2 /(FPEA) 2 PbI 4 , were obtained also by TCSPC at different pumping power at 200, 300, and 400 µW, shown in Figure 5a-c respectively. The lifetime data was calculated via double exponential fitting, given in Table S7, Supporting Information. And the obtained data was summarized in Figure 5d, where two features should be noted. One is that PL lifetimes of all samples decrease as the pumping power increases. This result corresponds to the Auger recombination process. [38][39] The other one is that the PL lifetime of (F 2 PA) 2 /(FPEA) 2 PbI 4 is longer than those of (PEA) 2 PbI 4 and (FPEA) 2 PbI 4 at whatever pumping power. One reason for this result is the reduction of defects, such as I − vacancy (also shown in Figure 4f), suppressed by the second insulator F 2 PA. Besides, as demonstrated in Figure 1b and Figure 3c, the effective Bohr radius of exciton in (F 2 PA) 2 /(FPEA) 2 PbI 4 becomes smaller due to the reduced effective dielectric constant, indicates that the probability of exciton-exciton annihilation decreases. In another word, for the same pumped condition, the intrinsic quantum well of (F 2 PA) 2 /(FPEA) 2 PbI 4 could accommodate more excitons as depicted in Figure 1b.
It's reasonable to expect a better stability performance of (F 2 PA) 2 /(FPEA) 2 PbI 4 due to the introduction of the low-affinity F. To verify this, optical and humidity stability were performed by monitoring the PL peak intensity as time evolved, as shown in Figure 6. First, the range ≈365 nm of mercury lamp was employed as the light source of continuous irradiation. And the times of various PLs, which decay to 1/e of their initial values, are 12, 23, and 35 min respectively for (PEA) 2 PbI 4 , (FPEA) 2 PbI 4 and (F 2 PA) 2 /(FPEA) 2 PbI 4 films in Figure 6a. In another word, the improvement of optical stability by the second insulator layer is ≈3-time and 1.5-time respectively in comparison with (PEA) 2 PbI 4 and (FPEA) 2 PbI 4 . During the light-induced degradation of (PEA) 2 PbI 4 , [40] the generation rate of the I − vacancies increases by the light irradiation due to the intrinsic soft-lattice nature, [26] leading to the distorting octahedral PbI 4 2− and deprotonation of PEA + , furthermore breaking down into PbI 2 and releasing of the volatile decomposed gases HI and PEA. The fluorinated insulator can alleviate this problem in (FPEA) 2 PbI 4 here as reported ones, [20,25] because it has an obvious hydrophobic feature, which can prevent the invasion of water molecules. And the introduction of the second fluorinated insulator F 2 PA with two-more F enhances further this alleviation effect. From the perspective of degradation kinetics, the shell layer formed by F 2 PA can prevent the volatilization of decomposed gases PEA and HI from the inner octahedral core, consequently, it has better optical performance.
We also recorded the evolution of PLs of the three samples every three days when they were placed in a relative humidity (R.H.) 90%-95% condition, shown in Figure 6b. To avoid the occasionality of collected data, the average of randomly selected 10-20 spots from the samples was used to depict then their The measured PLQY of the three samples. c) The PL lifetimes for the three samples under the same pumping power 300 µW. d) Radiative (red, the first right axis), nonradiative transition rates (black, left axis), and their ratios (blue, the second right axis) for the three samples. e) UPS of the three samples in the low-energy end, which reflects the highest occupied state, and the energy of valence band maximum (VBM). f) Schematic diagram of the VBMs for these three samples, where "×" denotes the location of defects.
www.advmatinterfaces.de performance every time, and the initial values were regarding as normalized states. As time evolved, the PLs of (PEA) 2 PbI 4 and (FPEA) 2 PbI 4 gradually decreased, and after 600 h, only 20% and 50% of their respective initial values were retained. The better performance of (FPEA) 2 PbI 4 than (PEA) 2 PbI 4 originates from the better hydrophobicity of FPEA than that of PEA. [20,25] The introduction of the second insulator layer is to enhance this improvement due to two-more F, where (F 2 PA) 2 /(FPEA) 2 PbI 4 kept its PL 90% of its initial value after 600 h exposure to R.H. 90%-95% atmosphere. The obvious feature of the evolved PL of (F 2 PA) 2 /(FPEA) 2 PbI 4 is its large fluctuation, which originates from its inhomogeneity of the film, also shown in Figure S2f, Supporting Information. Some of the PL intensities even over the original value, this can be credited to the annealing of the sample, which corresponds to the defects repair, also reported previously. [41] In a word, the second insulator layer with twomore fluorine can prevent from the penetration of water molecules, so that it can effectively enhance the humidity stability.

Conclusion
In summary, we proposed an interfacial strategy to improve the quantum yield and stability performance of 2D perovskite  www.advmatinterfaces.de (FPEA) 2 PbI 4 via a second fluorinated insulator layer F 2 PA. By utilizing the hydrogen bond between the ammonium of F 2 PA and the fluorine of FPEA, 2D perovskites with two organic insulator layers were formed and verified by various ways. The PL peak is a little red-shifted, which originates from the reduced Coulomb screening effect due to the low-dielectric constant of the fluorinated layer. The PLQY of (F 2 PA) 2 /(FPEA) 2 PbI 4 was improved by 3.4-time, in contrast with that of (FPEA) 2 PbI 4 . The second insulator can suppress the defects' generation, which were illustrated by the longer lifetime and reduced nonradiative recombination rate. The time of (F 2 PA) 2 /(FPEA) 2 PbI 4 by light-induced degradation has extended 1.5-time than that without F 2 PA. The second layer plays as an isolation role, which prevents the decomposed volatile gas out and the detrimental water molecules. Furthermore, in R.H. 90%-95% atmosphere for 600 h, (F 2 PA) 2 /(FPEA) 2 PbI 4 kept its PL 90% of the initial value, while (FPEA) 2 PbI 4 kept 50% for comparison. originates from the hydrophobicity of fluorinated chemicals. Thus, adding a second insulator layer is an effective strategy to improve the quantum yield and stability of 2D perovskites and it provides heuristic ideas for other RPP systems.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.