Open Access Article
This Open Access Article is licensed under a Creative Commons Attribution-Non Commercial 3.0 Unported Licence

Collective motion of methylammonium cations affects phase transitions and self-trapped exciton emission in A-site engineered MAPbI3 films

Chia-Hsun Yeha, Wen-Yu Chengbc, Tai-Che Choua, Yi-Chun Liua, Chia-Wei Changa, Yu-Sheng Chend, Chih-Hsing Wanga, Shih-Chang Wengd, Ian D. Sharp*bc, Pi-Tai Chouae and Chang-Ming Jiang*ae
aDepartment of Chemistry, National Taiwan University, 10617 Taipei, Taiwan. E-mail: cmjiang@ntu.edu.tw
bWalter Schottky Institute, Technical University of Munich, 85748 Garching, Germany. E-mail: sharp@wsi.tum.de
cPhysics Department, TUM School of Natural Sciences, Technical University of Munich, 85748 Garching, Germany
dNational Synchrotron Radiation Research Center, 300092 Hsinchu, Taiwan
eCenter for Emerging Materials and Advanced Devices, National Taiwan University, 10617 Taipei, Taiwan

Received 17th June 2025 , Accepted 25th July 2025

First published on 28th July 2025


Abstract

Hybrid organic–inorganic halide perovskites are celebrated for their exceptional optoelectronic properties and facile fabrication processes, making them prime candidates for next-generation photovoltaic and optoelectronic devices. By incorporating larger organic cations at the A-site, a novel class of ‘3D hollow perovskites’ has been developed, exhibiting enhanced stability and tunable optoelectronic properties. This study systematically explores the structural, phase transition, and photophysical characteristics of {en}MAPbI3 thin films with varying ethylenediammonium (en2+) content. The incorporation of less polar en2+ expands the perovskite unit cell, prolongs carrier lifetimes, and disrupts MA+ dipole–dipole interactions, thereby lowering the tetragonal-to-orthorhombic phase transition temperature. Temperature-dependent photoluminescence studies reveal that en2+ incorporation reduces the intensity and Stokes shift of self-trapped exciton emission at low temperatures, which are attributed to the diminished collective rotational dynamics of MA+ cations. These findings underscore the critical role of A-site cation dynamics in modulating phase stability and excitonic behaviour within hybrid halide perovskites, deepening our understanding of the interplay between organic cations and the inorganic framework and highlighting the potential of 3D hollow perovskites for stable and tunable optoelectronic applications.


Introduction

Hybrid organic–inorganic halide perovskites, such as methylammonium lead iodide (MAPbI3), have emerged as promising materials for next-generation optoelectronic devices due to their exceptional properties, including long carrier lifetimes, large absorption coefficients, tuneable band gaps, and intrinsic defect tolerance.1–5 These attributes have propelled the power conversion efficiencies (PCEs) of single-junction perovskite solar cells to beyond 25%.6–8 However, the inherent instability of halide perovskites under ambient conditions, including due to the volatility of organic cations, remains a critical challenge for large-scale deployment.9–11 Moreover, efforts to replace toxic Pb2+ with Sn2+ exacerbate these instability issues and lead to much lower achievable PCEs.12

To address these limitations, a new family of “3D hollow perovskites,” described by the general formula (L)x(A)1−xBX3, has recently been introduced. In these systems, larger organic cations (L), such as ethylenediammonium (en2+) and hydroxyethylammonium (HEA+), partially replace methylammonium (MA+) or formamidinium (FA+) at the A-site, with B = Pb2+ or Sn2+, and X = Br or I.13–23 These materials exhibit enhanced air and photostability, alongside desirable photovoltaic performance characteristics, in both Pb- and Sn-based systems. For example, mixed-halide perovskite such as Cs0.17FA0.83Pb(I0.6Br0.4)3 often suffers from halide segregation under continuous illumination, but the addition of en2+ significantly suppresses the mobility of halide vacancies and improves chemical stability.17 Similarly, replacing 10% propylenediammonium and trimethylenediammonium into the A-site of FASnI3 has been reported to increase the power conversion efficiency from 2.53% to 5.85% and 5.53%, respectively.20

Importantly, the incorporation of these larger organic cations modifies the perovskite crystal structure in two key ways: (1) by expanding the unit cell and, in some cases, inducing phase transitions, as governed by the Goldschmidt tolerance factor,24,25 and (2) through the creation of vacancies at the B-site and halide positions to accommodate the larger cations, forming a robust 3D hollow framework.14,16,18,19 These structural changes not only improve material stability but also allow for tuning of the optoelectronic properties, including by widening optical bandgaps and extending carrier lifetimes, despite the introduction of additional vacancy defects.15,18,19 Nevertheless, the fundamental mechanisms by which these larger cations influence carrier dynamics and excitonic behaviour remain elusive.

The dynamic behaviour of A-site cations, including their anisotropic sizes and roles as hydrogen bond donors, plays a pivotal role in determining the structural and optoelectronic properties of hybrid halide perovskites.26–28 Despite the absence of covalent bonding between organic and inorganic components, A-site cations, such as MA+, significantly influence phase transitions and excitonic behaviour through their rotational dynamics and dipole interactions. For example, MAPbI3 undergoes distinct phase transitions–from cubic to tetragonal to orthorhombic phases–as the temperature decreases, as reflected by the tilting of [PbI6] octahedra.29–32 At low temperatures, the diminished rotational degrees of freedom of MA+ result in organized domains defined by dipole–dipole interactions between MA+, manifesting in the antiparallel arrangement of MA+ dipoles between adjacent MAPbI3 layers.19 These dipole–dipole interactions, alongside hydrogen bonding between MA+ and I, drive lattice distortions and phase transitions. Moreover, hybrid halide perovskites comprising other organic cations show similar behaviour. For example, neutron scattering studies have revealed that the isotropic orientation of formamidinium (FA+) in the high-temperature cubic phase of FAPbI3 becomes frozen into an ordered arrangement in the low-temperature tetragonal phase, signifying the interplay between organic cation rotational dynamics and the inorganic framework.33 Recent studies have also demonstrated that the structural dynamics in hybrid halide perovskites facilitate self-trapped exciton (STE) formation through strong electron–phonon coupling, a characteristic feature of highly ionic halide perovskites.34–37 Unlike free excitons, which recombine comparatively rapidly, STEs are stabilized by lattice distortions, resulting in prolonged carrier lifetimes and broadband emission.37–40 However, the extent to which A-site cation rotational dynamics influence STE formation and phase stabilities is still poorly understood.

Motivated by these challenges, this study investigates the structures, phase transitions, and photophysical properties of 3D hollow {en}MAPbI3 thin films with varying en2+ content. The much lower static dipole moment of en2+ compared to MA+ disrupts long-range dipole–dipole interactions among A-site cations, which alters phase transition temperatures and carrier recombination dynamics. Using temperature-dependent photoluminescence (PL) and X-ray diffraction (XRD), we reveal how en2+ incorporation suppresses the tetragonal-to-orthorhombic perovskite (TP-to-OP) phase transition and reduces the intensity and Stokes shift of STE emission at low temperatures. These insights bridge the gap between structural dynamics and carrier behaviour in hybrid halide perovskites, offering a pathway for optimizing these compounds for photovoltaic and light-emitting diode applications.

Experimental methods

Synthesis of {en}MAPbI3 films

All films were prepared on amorphous SiO2 substrates (fused silica, 1 × 1 cm2) using a one-step spin-coating procedure in a nitrogen-filled glove box (Fig. S1). For pristine MAPbI3, a precursor solution was prepared by dissolving 0.8 mmol of methylammonium iodide (CH3NH3I, >99%, TCI) and 0.8 mmol of lead(II) iodide (PbI2, 99.999%, Alfa Aesar) in 1.00 mL of a 4[thin space (1/6-em)]:[thin space (1/6-em)]1 (v/v) DMF/DMSO mixture. {en}MAPbI3 films were synthesized by partially replacing CH3NH3I with ethylenediammonium iodide (C2H8N2·2HI, >98%, Sigma-Aldrich) at the specified stoichiometric ratio to achieve en2+ substitution of 5.0, 7.5, 10.0, and 15.0%.

The fused silica substrates were cleaned sequentially in an ultrasonic bath with deionized water (10 min, twice), acetone (20 min), and isopropanol (20 min). After drying, first with a nitrogen stream and then in an oven at 80 °C for 8 h, the substrates were treated in a UV-ozone cleaner for 20 min. The spin-coating process was performed using 10 μL of precursor solution at 4000 rpm for 30 s, with an acceleration rate of 800 rpm s−1. At a time of 3 s before the spin-coating process ended, 100 μL of chlorobenzene (anhydrous, 99.9%, Sigma-Aldrich) anti-solvent was dropped onto the rotating substrate. The as-coated films were then annealed on a hotplate at optimized temperatures for 10 min (see SI for details). A protective layer was subsequently applied by spin-coating 15 μL of a 10 mg mL−1 poly(methyl methacrylate) (PMMA) solution in chlorobenzene at 5000 rpm for 35 s (800 rpm per s acceleration rate), followed by annealing at 80 °C for 10 min.

Morphological and structural characterization

The surface morphologies of MAPbI3 and {en}MAPbI3 films were examined using a field emission scanning electron microscope (JSM-7600F, JEOL) in secondary electron detection mode. Room-temperature powder X-ray diffraction (PXRD) patterns were recorded using a PANalytical X'Pert PRO diffractometer equipped with a Cu anode (λ = 1.54178 Å) over a θ–2θ range of 5–50°. Temperature-dependent XRD measurements were performed at Beamline 09A of the Taiwan Photon Source, National Synchrotron Radiation Research Center. For this measurement, the thin-film sample was placed in a closed-cycle helium cryostat mounted on the Eulerian cradle of a 9-circle diffractometer, and a scintillation detector on the 2θ arm was used for XRD measurement using an X-ray energy of 12 keV (λ = 1.0332 Å). The sample temperature was controlled between 20–300 K, with a stabilization time of 15 min at each temperature point.

Steady-state and time-resolved optical characterization

Ultraviolet-visible (UV-Vis) transmission spectra of MAPbI3 and {en}MAPbI3 films were recorded over the wavelength range of 400–1200 nm using a spectrophotometer (UH-5700, Hitachi) with a bare fused silica substrate as the reference. Photoluminescence (PL) spectra were measured using a fluorimeter (FLS 980, Edinburgh Instruments) equipped with a photomultiplier tube as the detector and a 510 nm pulsed laser diode as the excitation source. The temporal evolution of the PL intensity was analysed using the time-correlated single photon counting (TCSPC) technique. The instrument response function, obtained with aluminium foil at the sample mount, was deconvoluted from the time-resolved PL traces.

For low-temperature PL measurements, thin-film samples were mounted in a closed-cycle helium cryostat (CS202-DMX-1AL, Advanced Research Systems, Inc.) evacuated by a turbomolecular pump. A thermocouple installed on the cold finger of the cryostat was connected to a temperature controller (Model 350, Lake Shore), which was used to adjust the sample temperature between 10–300 K using an integrated resistive heater.

Results and discussion

Structural analysis of {en}MAPbI3 films

The structural properties of {en}MAPbI3 thin films with 0–15% en2+ incorporation were first investigated via powder X-ray diffraction (PXRD). As shown in Fig. 1a, all diffraction peaks below 10% en2+ loading can be indexed to the tetragonal perovskite structure (TP; space group I4/mcm). Two trends were observed with increasing en2+ loading: (1) a monotonic shift of all diffraction peaks toward smaller 2θ angles (Fig. 1b), indicating unit cell expansion due to the incorporation of larger en2+, consistent with prior reports of {en}MAPbI3 (ref. 30, 32 and 36) and similar hollow hybrid halide perovskites;16 and (2) reduced diffraction intensities and slightly increased peak widths. To further assess structural disorder, we plot the full width at half maximum (FWHM) of the MAPbI3 (110) diffraction peak as a function of en2+ content (Fig. 1c). A steady increase in FWHM from 0.07° for pristine MAPbI3 to 0.14° for the 15% en2+ film support a trend of increasing lattice disorder. While the reduced diffraction intensities may also arise from variation in film thickness – as suggested by the progressively decreasing optical density in UV-Vis absorption spectra (Fig. 2a) – the increased FWHM values provide direct evidence of disrupted long-range crystallographic order in {en}MAPbI3.
image file: d5na00599j-f1.tif
Fig. 1 (a) PXRD patterns of {en}MAPbI3 films with different en2+ loadings. (b) Enlarged PXRD patterns around the MAPbI3 (110) diffraction peak as enclosed by the grey dashed box in panel (a). (c) The change of the full width at half maxima of the tetragonal MAPbI3 (110) reflection with en2+ loadings.

image file: d5na00599j-f2.tif
Fig. 2 (a) UV-Vis absorption spectra, (b) room-temperature steady-state PL spectra, and (c) time-resolved PL traces measured from {en}MAPbI3 films with different en2+ loadings. The solid lines in panel (c) represent the fitting results using a bi-exponential model convoluted with the instrument response function.

Notably, partial substitution of MA+ with the larger en2+ cation can give rise to additional low-angle diffraction peaks below 2θ = 10° (Fig. S2). Such features are often attributed either to the formation of quasi-2D MAPbI3 domains, commonly observed when long-chain organic cations are introduced during synthesis,21,24 or to transient, solvent-containing intermediate phases. In our case, these low-angle peaks were largely suppressed through careful optimization of the spin-coating and annealing conditions. The absence of both low-angle features and PbI2 impurity peaks in the PXRD patterns (Fig. S2) confirms the successful preparation of 3D hollow {en}MAPbI3 thin films. Additionally, SEM reveals that all samples exhibited compact, continuous films on fused silica substrates, with film thicknesses of approximately 230 nm (Fig. S3 and S4).

Interestingly, the (211) diffraction peak of tetragonal MAPbI3 weakens at 10% en2+ loading and disappears entirely at 15% en2+ loading. While the PXRD peak broadening in our thin films prevents full resolution of the (110)/(220) doublets and precludes reliable unit cell refinement, the observed disappearance of the (211) reflection combined with photophysical trends (vide infra) supports the assignment of a tetragonal-to-cubic phase (CP; space group Pm[3 with combining macron]m) transition with increasing en2+ incorporation. This transition aligns with the prediction from the Goldschmidt tolerance factor and previous reports,16,41 indicating that incorporating larger en2+ cations stabilizes the cubic phase at room temperature.

Photophysical properties and carrier dynamics

The UV-Vis absorption spectra of {en}MAPbI3 films (Fig. 2a) exhibit a systematic blueshift in the absorption onset with increasing en2+ loading, indicative of a widening optical bandgap. Room-temperature PL spectra (Fig. 2b), recorded under 510 nm ( = 2.43 eV) excitation, show a single emission peak attributed to the radiative recombination of free carriers.42 This PL peak shifts from 1.58 eV for pristine MAPbI3 to 1.65 eV for the 15% en2+ sample, reflecting bandgap widening due to lattice expansion. The bandgap modification arises from the reduced energy of the Pb 6s–I 5p anti-bonding valence band maximum (VBM) as the unit cell expands, while the conduction band minimum (CBM), having non-bonding character, remains unaffected.36,43 These results confirm that en2+ incorporation not only modifies the structural properties of MAPbI3 but also tunes its optoelectronic properties.

Time-correlated single-photon counting (TCSPC) measurements provide insights into carrier dynamics of {en}MAPbI3 films (Fig. 2c). PL decay traces were fitted using a bi-exponential model:

 
image file: d5na00599j-t1.tif(1)
where A0 is the background level, Ai is the weighting factor for each decay channel, and τi is the decay time. Here, τ1 represents Schottky–Read–Hall recombination mediated by defect states, while τ2 corresponds to band-to-band radiative recombination of free carriers at room temperature.42,44 The fitting results (Table 1) reveal a significant increase in τ2 with en2+ loading, except for the 15% en2+ sample, where the trend reverses. Concurrently, the ratio A1/(A1 + A2) decreases from 0.80 for pristine MAPbI3 to 0.62 for the 10% en2+ sample, indicating suppression of defect-mediated non-radiative recombination pathways in 3D hollow {en}MAPbI3. The intensity-weighted average PL lifetime (τavg) was calculated as.
 
image file: d5na00599j-t2.tif(2)
Here, τavg increased from 30.5 ns for pristine MAPbI3 to 172.0 ns for the 10% en2+ film, consistent with prolonged carrier lifetimes observed in structurally similar {en}FASnI3, {en}FA0.5MA0.5Pb0.5Sn0.5I3, and {en}MASnI3 films.14,18,19 At higher en2+ loadings, however, the emergence of the cubic MAPbI3 phase disrupts these trends. It has been reported that the CP phase of MAPbI3 exhibits lower electrical conductivity and more pronounced defect-mediated recombination than the TP phase,29 aligning with shortened τ1 and increased A1/(A1 + A2) ratio in films with 15% en2+ loading. Accordingly, τavg decreased significantly to 127.0 ns for the 15% en2+ film.

Table 1 Fitting results for the time-resolved PL traces monitored at the respective peak PL emission energies of {en}MAPbI3 films with varying en2+ loadings
Samples PL peak (eV) τ1 (ns) A1 τ2 (ns) A2 τavg (ns) A1/(A1 + A2)
MAPbI3 1.58 16.5 ± 0.2 0.082 ± 0.001 49.6 ± 0.7 0.020 ± 0.001 30.5 0.80
5% en2+ 1.60 29.4 ± 0.6 0.188 ± 0.002 136.9 ± 1.1 0.084 ± 0.001 102.0 0.69
7.5% en2+ 1.61 46.9 ± 0.9 0.197 ± 0.003 148.4 ± 1.5 0.096 ± 0.003 108.5 0.67
10% en2+ 1.62 23.7 ± 0.4 0.071 ± 0.001 200.9 ± 1.3 0.043 ± 0.001 172.0 0.62
15% en2+ 1.65 21.7 ± 0.4 0.157 ± 0.002 185.6 ± 1.8 0.033 ± 0.001 127.0 0.83


Phase transition behaviours

Temperature-dependent PL spectra of {en}MAPbI3 thin films (Fig. 3) reveal redshifts in emission peaks as the temperature decreases from 300 K. Such a redshift has been previously observed and can be attributed to a unit cell contraction that elevates the Pb–I anti-bonding energy levels at the VBM.44 For pristine MAPbI3, a sudden blueshift in the PL at 150 K corresponds to the TP-to-OP transition, during which the tilting of [PbI6] octahedra weakens the interaction between Pb2+ and I. To provide further evidence for the TP-to-OP transition, temperature-dependent XRD was conducted on a pristine MAPbI3 film (Fig. S5). Here, the (110) diffraction of TP and the (101) diffraction of OP were fitted using a Voigt function, and their peak positions were plotted as functions of temperature. In general, the diffraction peaks shifted to lower 2θ angles as the temperature increased, consistent with the assignment that the expanded unit cell is responsible for the PL redshift with decreasing temperature. In addition, a distinct change in the slopes of the diffraction peak positions over the 130–170 K range verified the TP-to-OP phase transition, consistent with the temperature-dependent PL measurement. Finally, increasing en2+ loading was found to lower the phase transition temperature observed by PL to 130 K, 120 K, and 100 K for films with 5%, 10%, and 15% en2+, respectively, suggesting enhanced TP stability due to en2+ incorporation.
image file: d5na00599j-f3.tif
Fig. 3 Temperature-dependent PL spectra of (a) pristine MAPbI3 and {en}MAPbI3 films with (b) 5%, (c) 10%, and (d) 15% en2+ loading. The tetragonal-to-orthorhombic phase transition temperatures are marked with dashed lines.

Based on the temperature-dependent PL and XRD results, we find that the TP-to-OP phase transition temperature decreases as the en2+ loading in 3D hollow perovskite increases. We hypothesize that this phenomenon originates from the smaller molecular dipole moments of en2+ cations compared to MA+. Indeed, prior computational studies on MAPbI3 have revealed that dipole–dipole interactions give rise to an ordered arrangement of MA+ cations, which is the primary driving force for the TP-to-OP phase transition at low temperatures.27 The random incorporation of the less polar en2+ cation at the A-site would disrupt such collective motion of MA+ and, consequently, require a further reduction of the lattice temperature to induce MA+ ordering and to initiate the transition to the OP phase.

Electron–phonon coupling in {en}MAPbI3

The full widths at half maximum (FWHMs) of PL emission peaks in the TP temperature range were analysed to provide insight into the coupling between electrons and longitudinal optical (LO) phonons, as well as the intrinsic disorder, in {en}MAPbI3 (Fig. 4). In the weak electron–phonon coupling regime, the temperature-dependent PL linewidth (Γ) can be described by:
 
image file: d5na00599j-t3.tif(3)
where Γ0 is the intrinsic linewidth, and ELO and γLO are the energy and coupling constant associated with the LO phonons, respectively.45,46 Here, ELO was fixed as 11.5 meV, an experimental value previously extracted from MAPbI3.45 Fitting results (Table 2) reveal consistent 17–22 meV electron–phonon coupling strengths across all films. However, Γ0 can be seen to increase progressively with en2+ cation loading, consistent with the greater lattice disorder in 3D hollow {en}MAPbI3 films with a higher extent of en2+ incorporation.

image file: d5na00599j-f4.tif
Fig. 4 The change of PL linewidths of {en}MAPbI3 films as a function of temperature. The dotted lines represent the fitting results obtained in the weak electron–phonon coupling limit, as described in the main text.
Table 2 List of intrinsic PL linewidths Γ0 and electron–phonon coupling strengths γLO for {en}MAPbI3 films with different en2+ cation loadings. In all cases, a value of ELO of 11.5 meV was used, as described in the text
Samples Γ0 (meV) γLO (meV)
MAPbI3 16.5 ± 0.5 17.3 ± 0.4
5% en2+ 21.1 ± 0.2 16.7 ± 0.1
10% en2+ 24.0 ± 0.3 17.4 ± 0.2
15% en2+ 23.9 ± 0.3 22.0 ± 0.3


Self-trapped exciton (STE) formation

Importantly, temperature-dependent PL spectra reveal the emergence of broadband emission far below the TP-to-OP transition temperature, especially for the pristine MAPbI3 film (Fig. 3). A comparison of 10 K PL spectra obtained from {en}MAPbI3 reveals a blueshift of this feature with increasing en2+ cation content (Fig. S6). Furthermore, as the en2+ cation loading increases in 3D hollow {en}MAPbI3, the temperature at which the broadband emission feature appears becomes lower, and the Stokes shifts become reduced. The large Stokes shifts of these emission features preclude their origin as being from the recombination of free excitons, which typically exhibit binding energies of < 20 meV in MAPbI3 single crystals and thin films.47,48 Rather, these low-temperature PL features in hybrid halide perovskites have been widely attributed to one of two mechanisms: defect-related (donor–acceptor pair transition) or self-trapped exciton emission.31,37,49–52

To differentiate these radiative recombination pathways, excitation power-dependent PL measurements were performed for the pristine MAPbI3 film at 160 K (Fig. 5a and b), where free carrier recombination dominates, and at 10 K (Fig. 5c and d), where the broadband emission with a large Stokes shift dominates. In both cases, the integrated PL intensities exhibited a linear relationship with the excitation power density. As defect-related emission is expected to saturate at higher excitation powers due to finite defect concentrations, which was not observed here, we conclude that the broad low-temperature emission originates from STEs in the OP phase rather than radiative recombination at defect states. Further evidence of this STE origin is the monotonic blueshift of the 10 K PL peak with increasing excitation power density, which is not expected for PL associated with free carriers, free excitons, or defects. For example, the PL peak of pristine MAPbI3 shifted from 1.45 eV to 1.48 eV as the excitation power density increased from 0.7 mW cm−2 to 6.4 mW cm−2 (Fig. 5d), a trend similarly observed for the 7.5% en2+ film (Fig. S7). We note that the 1015–1016 cm−3 excitation carrier densities used in this study are well below the 1017–1018 cm−3 regime where the band-filling effect is typically observed in MAPbI3,53,54 and no blueshift is seen in the free exciton emission at higher temperatures (Fig. 5a). Therefore, we attribute the blueshifting of the low-temperature broadband emission to the reduction of the STE domain size at higher excitation powers, rather than to band-filling.


image file: d5na00599j-f5.tif
Fig. 5 Steady-state PL spectra measured from a pristine MAPbI3 film at (a) 160 K and (c) 10 K using different excitation powers from a 510 nm pulsed laser. The integrated PL emission intensities and the emission peak positions for free carriers (FC) and self-trapped exciton (STE) are plotted as a function of the excitation power in (b) and (d), respectively.

The formation of STEs involves the interaction between free excitons (FEs) and the surrounding lattice. We posit that the driving force for STE formation primarily arises from the collective motion of MA+ cations surrounding FEs, rather than the distortions in the inorganic [PbI6] framework. Just below the TP-to-OP phase transition temperature, FE emission dominates the PL spectra, with MA+ dipoles aligned within the same plane but antiparallel to the adjacent MAPbI3 layers, as indicated by prior molecular dynamics simulations.27 At lower temperatures, MA+ cations surrounding the FE realign to accommodate the excitonic dipole moment, subsequently inducing distortions in the corner-sharing [PbI6] octahedra and stabilizing STE (Fig. 6a). This process resembles the reorganization of polar solvent molecules around charged species to reach a new equilibrium configuration with lower free energy. Greater displacement of the collective nuclear coordinates corresponds to stronger stabilization, manifesting as larger Stokes shifts for STE emission relative to FE emission (Fig. 6b).


image file: d5na00599j-f6.tif
Fig. 6 (a) Illustrative diagram for the formation of a self-trapped exciton at low temperatures through the collective rearrangement of methylammonium cations in the orthorhombic phase of MAPbI3. (b) Potential energy diagram for the sequential evolution from the free carrier (FC) excited state to the free exciton (FE) state and, finally, to the self-trapped exciton (STE) state.

Under low excitation power, each FE can couple to a greater number of MA+ cations, leading to more extensive lattice distortions and larger Stokes shifts. Conversely, higher excitation powers generate larger densities of FEs, limiting the extent of lattice distortion and resulting in smaller energy stabilization. This behaviour explains the excitation power-dependent PL peak position at low temperatures for MAPbI3 and {en}MAPbI3 films. Furthermore, the differences in PL spectral profiles obtained with different excitation sources, as shown in Fig. S8, support this interpretation. For both pristine MAPbI3 and the 7.5% en2+ film, excitation with a xenon lamp source (monochromated at 510 nm, continuous wave) produced significantly more redshifted PL spectra compared to a pulsed laser source due to the significantly lower excitation power density (∼0.15 mW cm−2) provided by the xenon lamp.

In 3D hollow MAPbI3 films, the suppression of STE formation is primarily attributed to disrupted MA+ dipole–dipole interactions rather than to significant changes in intrinsic electron–phonon coupling strength. While the fitted electron–phonon coupling parameter γLO remains nearly constant up to 10% en2+ incorporation (Table 2), the introduction of less polar en2+ cations interrupts the collective MA+ dynamics that are critical for stabilizing STEs. This manifests as smaller Stokes shifts and less favourable competition with FE emission. Additionally, since STE formation is driven by MA+ motion, higher en2+ content requires lower temperatures for STE emission to appear in {en}MAPbI3, consistent with our experimental observations. At 90 K, the PL spectrum of pristine MAPbI3 can be deconvoluted into three Gaussian peaks, corresponding to one FE and two STE emissions (Fig. 7a).


image file: d5na00599j-f7.tif
Fig. 7 Steady-state PL spectra of (a) pristine MAPbI3 and (d) {en}MAPbI3 with 7.5% en2+ loading, measured at 90 K using 510 nm excitation. For each sample, the time-resolved PL traces were monitored at two wavelengths to track the dynamics of (b and c) FE and (e and f) STE emissions, respectively.

The presence of two distinct STE peaks suggests the existence of multiple local minima in the exciton-lattice potential energy surface, as illustrated in Fig. 6b. These may correspond to different configurations of MA+ dipole orientations and [PbI6] octahedral distortions surrounding the self-trapped STE. Such variation in stabilization energies could also arise from heterogeneity in local lattice environments, especially in polycrystalline films. While the precise structural origins of the dual STE states remain to be resolved, this phenomenon is consistent with the interplay between organic dipole reorganization and inorganic lattice relaxation that governs STE formation.

TCSPC measurements were conducted to measure the temporal evolution at 780 nm and 850 nm (Fig. 7b and c). Here, the FE emission exhibited a much faster decay than the STE emission. More importantly, the rise of STE emission coincided with the fast decay component of the FE emission on the few-nanosecond timescale, signifying the collective rotational dynamics of MA+ cations. For the {en}MAPbI3 film with 7.5% en2+ loading, the steady-state PL spectrum at 90 K also consists of one FE and two STE emission features, though individual components are blue-shifted compared to the pristine MAPbI3 (Fig. 7d). Similarly, the decay of the PL trace at 770 nm, with a major contribution from FE emission, matches the rise time of the STE-dominant PL trace at 870 nm (Fig. 7e and f). Nevertheless, we note that the broad STE emissions prevent the complete deconvolution of the FE emission kinetics, especially in 3D hollow {en}MAPbI3 films with smaller Stokes shifts for the STE emission.

We also note that the broad low-energy MAPbI3 emission observed at low temperatures has previously been attributed to donor–acceptor (DAP) transitions,31,51 which can also exhibit a spectral blueshift with increasing excitation powers – consistent with our observations. However, we found that incorporation of en2+ into MAPbI3 leads to a blueshift of both the orthorhombic-phase free exciton PL and the broadband emission (Fig. 3 and S6). For DAP transitions associated with point defects, the donor and acceptor levels are less likely to change in the 3D hollow MAPbI3 structure. In contrast, prior molecular dynamics simulations suggest that MA+ disorder in the orthorhombic phase can give rise to low-temperature broadband PL that blueshifts with increasing excitation power,55 supporting our assignment of this feature to STE emission. Overall, the A-site cation engineering strategy employed here provides a unique means to modulate MA+ dipole–dipole interactions, thereby tuning both the phase transition behaviour and the photophysical properties of hybrid halide perovskites.

Conclusions

This study investigates the structural and optical properties of {en}MAPbI3, a 3D hollow halide perovskite, focusing on the effects of incorporating less polar en2+ cations at the A-sites. The larger size of en2+ compared to MA+ leads to unit cell expansion, optical bandgap widening, increased lattice disorder, and suppressed non-radiative recombination. However, at higher en2+ loading, {en}MAPbI3 transforms to the cubic perovskite phase at room temperature, making it more susceptible to defect-mediated carrier recombination. The tetragonal-to-orthorhombic phase transition temperature of MAPbI3 decreases with increasing en2+ content, underscoring the role of MA+ dipole–dipole interactions in driving structural changes in the inorganic [PbI6] framework. Additionally, this work identifies the low-temperature broadband PL of MAPbI3 as originating from self-trapped exciton (STE) emission, with the emission peak blueshifting under higher excitation powers. Free excitons induce the realignment of the surrounding MA+ dipole to achieve a more stable nuclear configuration, resulting in broadband emission with a large Stokes shift. The introduction of en2+ at the A-site disrupts MA+ interactions, reducing the extent of STE stabilization. These findings highlight the critical role of A-site cation dipole dynamics in governing phase transitions and excitonic properties in hybrid halide perovskites. By understanding and controlling these dynamics, {en}MAPbI3 emerges as a promising material with a tuneable optical bandgap and suppressed defect-mediated recombination compared to its parent compound, making it suitable for optoelectronic applications that require stable and tuneable emission properties.

Author contributions

C.-H. Yeh: investigation, formal analysis, visualization, writing – original draft. W.-Y. Cheng: investigation, formal analysis, methodology. T.-C. Chou: methodology, supervision. Y.-C. Liu: methodology, supervision. C.-W. Chang: investigation, methodology. Y.-S. Chen: methodology. C.-H. Wang: methodology, supervision. S.-C. Weng: methodology, resources, supervision. I. D. Sharp: conceptualization, project administration, writing – review & editing. P.-T. Chou: resources, supervision, writing – review & editing. C.-M. Jiang: conceptualization, funding acquisition, project administration, writing – review & editing, data curation.

Conflicts of interest

There are no conflicts to declare.

Data availability

The following data supporting this article have been included as part of the SI.

Fig. S1: Schematic diagram of the spin-coating procedure for synthesizing {en}MAPbI3 films. Fig. S2: PXRD patterns of {en}MAPbI3 films synthesized using different annealing temperatures. Fig. S3 and S4: SEM images of {en}MAPbI3 films. Fig. S5: Temperature-dependent (110) diffraction peak position of pristine MAPbI3. Fig. S6: 10 K PL spectra of {en}MAPbI3 films. Fig. S7: 10 K PL spectra of 7.5% en2+ sample measured using different excitation powers. Fig. S8: Comparison of 10 K PL spectra of pristine MAPbI3 and {en}MAPbI3 with 7.5% en2+ loading used either a xenon lamp or a 510 nm pulsed laser as the excitation source. See DOI: https://doi.org/10.1039/d5na00599j.

Acknowledgements

This research was financially supported by the National Science and Technology Council under Grant No. MOST 110-2113-M-002-027-MY3, and by the Office of Research and Development, National Taiwan University (NTU) under Grant No. 113L893305. I. D. S. acknowledges support from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy – EXC 2089/1-390776260 and from TUM. solar in the context of the Bavarian Collaborative Research Project Solar Technologies Go Hybrid (SolTech). We thank C.-Y. Chien and S.-J. Ji of the Instrumentation Center at NTU for their assistance in SEM operation.

References

  1. T. Chen, W.-L. Chen, B. J. Foley, J. Lee, J. P. Ruff, J. P. Ko, C. M. Brown, L. W. Harriger, D. Zhang and C. Park, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 7519–7524 CrossRef CAS PubMed.
  2. M. A. Green, Y. Jiang, A. M. Soufiani and A. Ho-Baillie, J. Phys. Chem. Lett., 2015, 6, 4774–4785 CrossRef CAS PubMed.
  3. J. Xu, A. Maxwell, M. Wei, Z. Wang, B. Chen, T. Zhu and E. H. Sargent, ACS Energy Lett., 2021, 6, 4220–4227 CrossRef CAS.
  4. Y. Zhou, I. Poli, D. Meggiolaro, F. De Angelis and A. Petrozza, Nat. Rev. Mater., 2021, 6, 986–1002 CrossRef.
  5. G. W. Kim and A. Petrozza, Adv. Energy Mater., 2020, 10, 2001959 CrossRef CAS.
  6. M. J. Jeong, C. S. Moon, S. Lee, J. M. Im, M. Y. Woo, J. H. Lee, H. Cho, S. W. Jeon and J. H. Noh, Joule, 2023, 7, 112–127 CrossRef CAS.
  7. Q. Jiang, J. Tong, Y. Xian, R. A. Kerner, S. P. Dunfield, C. Xiao, R. A. Scheidt, D. Kuciauskas, X. Wang and M. P. Hautzinger, Nature, 2022, 611, 278–283 CrossRef CAS PubMed.
  8. Y. Zhao, F. Ma, Z. Qu, S. Yu, T. Shen, H.-X. Deng, X. Chu, X. Peng, Y. Yuan and X. Zhang, Science, 2022, 377, 531–534 CrossRef CAS PubMed.
  9. B. Zhang, Y. Liao, L. Tong, Y. Yang and X. Wang, Phys. Chem. Chem. Phys., 2020, 22, 7778–7786 RSC.
  10. A. F. Akbulatov, S. Y. Luchkin, L. A. Frolova, N. N. Dremova, K. L. Gerasimov, I. S. Zhidkov, D. V. Anokhin, E. Z. Kurmaev, K. J. Stevenson and P. A. Troshin, J. Phys. Chem. Lett., 2017, 8, 1211–1218 CrossRef CAS PubMed.
  11. B. w. Park and S. I. Seok, Adv. Mater., 2019, 31, 1805337 CrossRef PubMed.
  12. L. Lanzetta, N. Aristidou and S. A. Haque, J. Phys. Chem. Lett., 2020, 11, 574–585 CrossRef CAS PubMed.
  13. K. Jayanthi, I. Spanopoulos, N. Zibouche, A. A. Voskanyan, E. S. Vasileiadou, M. S. Islam, A. Navrotsky and M. G. Kanatzidis, J. Am. Chem. Soc., 2022, 144, 8223–8230 CrossRef CAS PubMed.
  14. W. Ke, C. C. Stoumpos, M. Zhu, L. Mao, I. Spanopoulos, J. Liu, O. Y. Kontsevoi, M. Chen, D. Sarma, Y. Zhang, M. R. Wasielewski and M. G. Kanatzidis, Sci. Adv., 2017, 3, e1701293 CrossRef PubMed.
  15. I. Spanopoulos, I. Hadar, W. Ke, P. Guo, E. M. Mozur, E. Morgan, S. Wang, D. Zheng, S. Padgaonkar, G. N. Manjunatha Reddy, E. A. Weiss, M. C. Hersam, R. Seshadri, R. D. Schaller and M. G. Kanatzidis, J. Am. Chem. Soc., 2021, 143, 7069–7080 CrossRef CAS PubMed.
  16. I. Spanopoulos, W. Ke, C. C. Stoumpos, E. C. Schueller, O. Y. Kontsevoi, R. Seshadri and M. G. Kanatzidis, J. Am. Chem. Soc., 2018, 140, 5728–5742 CrossRef CAS PubMed.
  17. L. Grater, M. Wang, S. Teale, S. Mahesh, A. Maxwell, Y. Liu, S. M. Park, B. Chen, F. Laquai, M. G. Kanatzidis and E. H. Sargent, J. Phys. Chem. Lett., 2023, 14, 6157–6162 CrossRef CAS PubMed.
  18. W. Ke, I. Spanopoulos, Q. Tu, I. Hadar, X. Li, G. S. Shekhawat, V. P. Dravid and M. G. Kanatzidis, J. Am. Chem. Soc., 2019, 141, 8627–8637 CrossRef CAS PubMed.
  19. W. Ke, C. C. Stoumpos, I. Spanopoulos, L. Mao, M. Chen, M. R. Wasielewski and M. G. Kanatzidis, J. Am. Chem. Soc., 2017, 139, 14800–14806 CrossRef CAS PubMed.
  20. W. Ke, C. C. Stoumpos, I. Spanopoulos, M. Chen, M. R. Wasielewski and M. G. Kanatzidis, ACS Energy Lett., 2018, 3, 1470–1476 CrossRef CAS.
  21. A. Leblanc, N. Mercier, M. Allain, J. Dittmer, V. Fernandez and T. Pauporté, Angew. Chem., Int. Ed., 2017, 56, 16067–16072 CrossRef CAS PubMed.
  22. J. Lu, L. Jiang, W. Li, F. Li, N. K. Pai, A. D. Scully, C. M. Tsai, U. Bach, A. N. Simonov and Y. B. Cheng, Adv. Energy Mater., 2017, 7, 1700444 CrossRef.
  23. C.-M. Tsai, Y.-P. Lin, M. K. Pola, S. Narra, E. Jokar, Y.-W. Yang and E. W.-G. Diau, ACS Energy Lett., 2018, 3, 2077–2085 CrossRef CAS.
  24. Y. Fu, M. P. Hautzinger, Z. Luo, F. Wang, D. Pan, M. M. Aristov, I. A. Guzei, A. Pan, X. Zhu and S. Jin, ACS Cent. Sci., 2019, 5, 1377–1386 CrossRef CAS PubMed.
  25. X. Li, M. Kepenekian, L. Li, H. Dong, C. C. Stoumpos, R. Seshadri, C. Katan, P. Guo, J. Even and M. G. Kanatzidis, J. Am. Chem. Soc., 2022, 144, 3902–3912 CrossRef CAS PubMed.
  26. A. Al-Kahtani, S. Tabassum, I. Raya, I. Khlewee, S. Chupradit, A. Davarpanah, M. Elveny and S. Ali, Coatings, 2021, 11, 1341 CrossRef CAS.
  27. S. Maheshwari, M. B. Fridriksson, S. Seal, J. Meyer and F. C. Grozema, J. Phys. Chem. C, 2019, 123, 14652–14661 CrossRef CAS PubMed.
  28. A. A. Koegel, E. M. Mozur, I. W. Oswald, N. H. Jalarvo, T. R. Prisk, M. Tyagi and J. R. Neilson, J. Am. Chem. Soc., 2022, 144, 1313–1322 CrossRef CAS PubMed.
  29. A. Bonadio, F. P. Sabino, A. L. Freitas, M. R. Felez, G. M. Dalpian and J. A. Souza, Inorg. Chem., 2023, 62, 7533–7544 CrossRef CAS PubMed.
  30. A. Halder, Y. Rakita, D. Cahen and S. K. Sarkar, J. Phys. Chem. Lett., 2020, 11, 1473–1476 CrossRef CAS PubMed.
  31. W. Kong, Z. Ye, Z. Qi, B. Zhang, M. Wang, A. Rahimi-Iman and H. Wu, Phys. Chem. Chem. Phys., 2015, 17, 16405–16411 RSC.
  32. Y. Li, H. Song, X. Ma, J. Liu, C. Ge, D. Liu and S. Liu, J. Phys. Chem. C, 2023, 127, 12613–12620 CrossRef CAS.
  33. C. Yu, Y. Kawakita, T. Kikuchi, M. Kofu, T. Honda, Z. Zhang, Z. Zhang, Y. Liu, S. F. Liu and B. Li, J. Phys. Chem. Lett., 2024, 15, 329–338 CrossRef CAS PubMed.
  34. S. Maheshwari, S. Patwardhan, G. C. Schatz, N. Renaud and F. C. Grozema, Phys. Chem. Chem. Phys., 2019, 21, 16564–16572 RSC.
  35. E. M. Mozur, M. A. Hope, J. C. Trowbridge, D. M. Halat, L. L. Daemen, A. E. Maughan, T. R. Prisk, C. P. Grey and J. R. Neilson, Chem. Mater., 2020, 32, 6266–6277 CrossRef CAS.
  36. J. H. Lee, J.-H. Lee, E.-H. Kong and H. M. Jang, Sci. Rep., 2016, 6, 21687 CrossRef CAS PubMed.
  37. S. Li, J. Luo, J. Liu and J. Tang, J. Phys. Chem. Lett., 2019, 10, 1999–2007 CrossRef CAS PubMed.
  38. Z. Xu, X. Jiang, H.-p. Cai, K. Chen, X. Yao and Y. Feng, J. Phys. Chem. Lett., 2021, 12, 10472–10478 CrossRef CAS PubMed.
  39. J. Tan, D. Li, J. Zhu, N. Han, Y. Gong and Y. Zhang, Nanoscale, 2022, 14, 16394–16414 RSC.
  40. H.-C. Chu, C.-M. Hung, H.-C. Huang, S.-C. Weng, B.-H. Lin, S. Yang, Y.-H. Wu, K.-H. Chang, J.-J. Shyue and P.-T. Chou, J. Phys. Chem. C, 2024, 128, 20441–20450 CrossRef CAS.
  41. C. Wu, K. Chen, D. Guo, S. Wang and P. Li, RSC Adv., 2018, 8, 2900–2905 RSC.
  42. L. Q. Phuong, Y. Yamada, M. Nagai, N. Maruyama, A. Wakamiya and Y. Kanemitsu, J. Phys. Chem. Lett., 2016, 7, 2316–2321 CrossRef CAS PubMed.
  43. Q. Chen, N. De Marco, Y. Yang, T.-B. Song, C.-C. Chen, H. Zhao, Z. Hong, H. Zhou and Y. Yang, Nano Today, 2015, 10, 355 CrossRef CAS.
  44. A. A. Baloch, F. H. Alharbi, G. Grancini, M. I. Hossain, M. K. Nazeeruddin and N. Tabet, J. Phys. Chem. C, 2018, 122, 26805–26815 CrossRef CAS.
  45. A. D. Wright, C. Verdi, R. L. Milot, G. E. Eperon, M. A. Pérez-Osorio, H. J. Snaith, F. Giustino, M. B. Johnston and L. M. Herz, Nat. Commun., 2016, 7, 11755 CrossRef PubMed.
  46. J. A. Steele, P. Puech, M. Keshavarz, R. Yang, S. Banerjee, E. Debroye, C. W. Kim, H. Yuan, N. H. Heo and J. Vanacken, ACS Nano, 2018, 12, 8081–8090 CrossRef CAS PubMed.
  47. L. M. Herz, Annu. Rev. Phys. Chem., 2016, 67, 65–89 CrossRef CAS PubMed.
  48. V. D’innocenzo, G. Grancini, M. J. Alcocer, A. R. S. Kandada, S. D. Stranks, M. M. Lee, G. Lanzani, H. J. Snaith and A. Petrozza, Nat. Commun., 2014, 5, 3586 CrossRef PubMed.
  49. F. Ambrosio, E. Mosconi, A. A. Alasmari, F. A. S. Alasmary, D. Meggiolaro and F. De Angelis, Chem. Mater., 2020, 32, 6916–6924 CrossRef CAS.
  50. S. Krishnamurthy, R. Naphade, W. J. Mir, S. Gosavi, S. Chakraborty, R. Vaidhyanathan and S. Ogale, Adv. Opt. Mater., 2018, 6, 1800751 CrossRef.
  51. J. M. Urban, T. H. T. Nguyen, G. Chehade, A. Delteil, G. Trippé-Allard, G. Delport, E. Deleporte, J.-P. Hermier and D. Garrot, Phys. Status Solidi RRL, 2023, 17, 2300005 CrossRef CAS.
  52. C. Yang, Q. Wei, Y. Gong, M. Long, G. Zhou, G. Xing and B. Wu, J. Phys. Chem. Lett., 2023, 14, 10046–10053 CrossRef CAS PubMed.
  53. J. S. Manser and P. V. Kamat, Nat. Photon., 2014, 8, 737–743 CrossRef CAS.
  54. M. Saba, F. Quochi, A. Mura and G. Bongiovanni, Acc. Chem. Res., 2016, 49, 166–173 CrossRef CAS PubMed.
  55. M. I. Dar, G. Jacopin, S. Meloni, A. Mattoni, N. Arora, A. Boziki, S. M. Zakeeruddin, U. Rothlisberger and M. Grätzel, Sci. Adv., 2016, 2, e1601156 CrossRef PubMed.

This journal is © The Royal Society of Chemistry 2025
Click here to see how this site uses Cookies. View our privacy policy here.