Naveen Kumar
Tailor
ab,
Sanchi
Monga
c,
Saurabh K.
Saini
de,
Mahesh
Kumar
ef,
Saswata
Bhattacharya
*c and
Soumitra
Satapathi
*ab
aDepartment of Physics, Indian Institute of Technology Roorkee, Roorkee 247667, India. E-mail: soumitrasatapathi@gmail.com
bCenter for Sustainable Energy, Indian Institute of Technology Roorkee, Roorkee 247667, India
cDepartment of Physics, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India. E-mail: saswata@physics.iitd.ac.in
dCSIR-National Physical Laboratory, Dr K.S. Krishnan Marg, New Delhi, 110012, India
eAcademy of Scientific and Innovative Research (AcSIR), Ghaziabad, 201002, India
fInnovation Management Directorate, Anusandhan Bhawan, New Delhi 110001, India
First published on 16th April 2025
The interaction between carriers and photons in halide perovskites gives rise to intriguing phenomena in their excited states. In particular, bismuth halide perovskites exhibit behavior that extends beyond free carriers, involving excitons and polarons. Here, we report the steady state and excited state dynamics in the lead-free A3Bi2I9 [A = FA (formamidinium), MA (methylammonium), Cs (cesium)] perovskite derivatives. The A3Bi2I9 system exhibits strong excitonic peaks in the absorption spectra because of defect-related direct-bound excitons. The emission from self-trapped excitons influenced by carrier-phonon coupling and exciton–exciton interactions results in broad photoluminescence spectra. The low-energy photo-induced absorption (PIA-L) band below the bandgap energy is attributed to band gap renormalization (BGR) and the formation of self-trapped excitons (STSs) through electron-acoustic phonon coupling. Hot carrier cooling results in a transient absorption response and the occupation of modified band edge states. The interplay between BGR and polaron formation plays a crucial role in determining the amplitude of PIA-L during the cooling process. We observe that the carrier dynamics in the A3Bi2I9 system are mostly dominated by localized excitons and small polarons. This study enhances our understanding of the fundamental processes governing their optoelectronic behavior and paves the way for their further utilization in advanced device applications.
Bismuth halide perovskites exhibit low dimensional lattice connectivity and quantum confinement effect, which can significantly influence their lattice dynamics and photophysical characteristics.9,10 The relationship between the structural and photophysical properties of A3Bi2I9 perovskites has spurred investigations into the dynamics of charge carriers in these materials after photoexcitation. The zero-dimensional lattice connectivity of A3Bi2I9 leads to quantum confinement of charge carriers. Ghosh et al. examined the limitations of Cs3Bi2I9 as a photovoltaic absorber material and identified weak interactions between [Bi2I9]3− bioctahedra as a major barrier to achieving high solar cell efficiency.11,12 Scholz et al. explored exciton dynamics in the MA3Bi2I9 film deposited on mesoporous TiO2.13 Earlier, we investigated the polaron-mediated photoconduction and phonon-mediated hot carrier cooling dynamics in the Cs3Bi2I9 single crystalline system.14,15 We also studied the hot carrier relaxation in A3Bi2I9 polycrystalline films and found that the nature of the A-site cation significantly influences the hot carrier cooling time.16
Despite the considerable interest generated by bismuth halide perovskite materials and extensive theoretical and experimental research aimed at uncovering their photophysical processes, key aspects such as electron-lattice coupling, hot carrier cooling dynamics, and polaron formation are still not fully understood. Additionally, the origins of the observed photoinduced absorption at both the high-energy and low-energy sides of the bleaching band in transient absorption features remain elusive. Moreover, it is imperative to develop a comprehensive understanding of how different cation substitutions impact carrier relaxation dynamics and photoinduced effects.
In this work, we have explored the steady-state and excited-state dynamics in lead-free A3Bi2I9 (A = FA, MA, Cs) perovskite nanocrystals using the steady state and transient optical spectroscopy technique and investigated how the nature of the A-site cation can influence these characteristics. These nanocrystals were synthesized using the antisolvent precipitation technique and characterized by X-ray diffraction (XRD), transmission electron microscopy (TEM), and steady-state optical spectroscopy techniques. Furthermore, we have excited the three perovskite NCs above the band edge (350 nm) with the same excitation power and observed that the relaxation of hot carriers occurs most slowly in the inorganic Cs3Bi2I9 perovskite. The significant variation in hot carrier cooling with different cation compositions is due to the interactions between the various cations (A = FA, MA, Cs) and the Bi-I frameworks. Moreover, we observed distinct features in the form of high-energy positive bands (photoinduced absorption, PIA-H), negative bands (ground state bleaching, GSB), and low-energy broad positive bands (PIA-L) for all three nanocrystals. We found the bandgap renormalization in the presence of hot carriers, and polaron-mediated activation of forbidden transitions are mainly responsible for the observed transient absorption features. These findings enhance our understanding of the fundamental processes governing their optoelectronic behavior and pave the way for their potential utilization in advanced device applications.
Furthermore, to gain detailed insights into the optical and electronic properties of these materials, we compute the electronic band structure and atom-projected partial density of states (pDOS) using density functional theory (DFT).9,11,26 We find that all these materials are indirect band gap semiconductors, with indirect (direct) band gaps of 1.81 (1.92) eV, 1.79 (1.89) eV, and 1.78 (1.82) eV for Cs3Bi2I9, MA3Bi2I9, and FA3Bi2I9, respectively (Fig. 2a–c).11,12,27–30 The difference between the indirect and direct band gaps is minimal, on the order of 0.1 eV. Additionally, we observe a decreasing trend in the band gap with increasing size of the A-site cation, consistent with experimental observations (Fig. S5, ESI†). From the pDOS plots, we observe that for all A3Bi2I9 materials, the valence band edges are primarily contributed by the I (p) orbitals, while the conduction band edges are predominantly composed of hybridized Bi (p) and I (p) orbitals (Fig. 2d–f). We note that the A-site cation does not directly contribute to the band edges (Fig. S6, ESI†). However, it significantly affects the band dispersion. We estimate the effective mass of electrons for these perovskites by the parabolic fitting of the lowest conduction band edges. The effective masses are found to be 1.64 me, 4,55 me, and 2.02 me for Cs3Bi2I9, MA3Bi2I9, and FA3Bi2I9, respectively. In the case of organic A-site cations, we observe a decrease in the band dispersion of the lowest lying conduction band, indicating smaller electron mobility in organic perovskites compared to inorganic Cs3Bi2I9.
Moreover, we observe excitonic peaks in the optical spectra computed theoretically using many-body perturbation theory (G0W0/BSE). The first peak obtained through the G0W0 method represents the electronic band gap of a material, incorporating many-body effects. However, it does not account for the formation of electron–hole pairs (i.e., excitons). To include electron–hole interactions, we solved the model Bethe–Salpeter equation (mBSE) and determined the optical spectra (Fig. S7, ESI†). The shift in the optical spectra, resulting in a lower optical band gap compared to the electronic band gap, indicates the presence of excitons in these materials. The difference between the electronic (G0W0) and optical (mBSE) band gaps provides an estimate of the exciton binding energy (EB). We observe that EB (Cs3Bi2I9) < EB (MA3Bi2I9) < EB (FA3Bi2I9). The larger dielectric screening of the Coulomb interaction between electrons and holes in the inorganic system results in smaller exciton binding energy compared to the organic systems. However, there is a very small difference in the static electronic dielectric contribution of these materials, with values of 4.0, 3.8, and 3.4 for Cs3Bi2I9, MA3Bi2I9, and FA3Bi2I9, respectively, resulting in closely lying values of EB.
Furthermore, we measured the photoluminescence (PL) spectrum of all three NCs with different excitation wavelengths as shown in Fig. 3. The Cs3Bi2I9 NCs display emission peaks ∼405 nm (3.06 eV) and ∼455 nm (2.72 eV) accompanied by a broad emission spanning the range of 500 to 650 nm when excited within the wavelength range of 350 to 400 nm. Subsequent excitation at higher wavelengths only resulted in the observation of broad emission in the range of 500 to 650 nm. In the case of MA3Bi2I9 and FA3Bi2I9, emission peaks are observed at ∼408 nm (3.06 eV), ∼438 nm (2.72 eV), and ∼462 nm (with a small shift of 2–3 nm with changing MA to FA) with a broad emission from 500 to 650 nm in response to excitation wavelengths from 350 nm to 400 nm. The occurrence of emission peaks at ∼405 and ∼455 nm, at energies higher than the lowest-energy broad emission, is quite unusual. To confirm the origin of these high-energy peaks, we perform measurements of the excitation spectrum by analyzing the emission at all observed wavelengths. Remarkably, the obtained photoexcitation (PLE) spectrum matches well with the absorption spectrum. We find that these high energy emission peaks occur because of electron transitions from the ground 1S0 to the excited states of Bi3+ in the isolated [Bi2I9]3− clusters. The broad emission spanning from 500 nm to 650 nm can be attributed to excitonic emission. The broad emission profile displays a characteristic shape with a relatively sharp cutoff on the high-energy side, followed by a long exponential tail extending towards lower energies.31 This unique behavior clearly indicates photoluminescence associated with an exponential tail in the density of states.32 The extended low-energy tail in the PL spectrum corresponds to the emission of localized excitons (LE) and is broad, existing at lower energies.32,33 Previous observations in bismuth halide systems have reported the presence of self-trapped exciton (STE) photoluminescence at room temperature.34 The carrier-phonon coupling effect induced by phonon-assisted absorption and recombination processes is strong in these indirect-bandgap nanocrystals, leading to the formation of self-trapped excitons.35–38 Consequently, due to the presence of an indirect band structure calculated from DFT calculations, intervalley scattering processes are expected in these systems.39 Given these insights, a more comprehensive investigation into the interplay between the indirect band structure, phonon interactions, and self-trapped excitons is warranted in future studies to gain a deeper understanding of the fundamental photophysical properties of bismuth halide nanocrystals.
In our systems, a large number of excitons are formed after photoexcitation. These excitons are primarily localized in nature, characteristic of Frenkel excitons, which possess high binding energy.13,36,40 The GSB band position is located around the excitonic peak in the absorption spectra for all three systems, ∼2.49 eV for Cs3Bi2I9 (∼0.15 eV width), ∼2.38 eV for MA3Bi2I9 (∼0.19 eV width) and ∼2.43 eV for FA3Bi2I9 (∼0.19 eV width) as presented in Fig. 4b, d and f, respectively. The origin of GSB is attributed to the band-filling effect and the width of GSB is determined by the hot carrier's distribution and exciton–exciton interaction.41–43 Localization of excitons in this system can lead to exciton–exciton interaction and scattering, which results in the width of the GSB band at initial delays.42,44 The excitons in A3Bi2I9 perovskites exhibit strong confinement within the isolated 0D (Bi2I9)3− dimers. The bleach observed near the exciton resonance, along with two PIA bands flanking the bleaching peak, is characteristic of exciton band broadening caused by carrier-exciton scattering following photoexcitation.17,18,45 The observed shift in the GSB band at earlier delay is due to hot carrier occupancies in the density of states and then relaxation towards low energy levels of the conduction band. During the initial delay times following photoexcitation, the broadening of the exciton bands can be primarily attributed to the presence of non-relaxed hot carriers; after that, the carriers gradually relax towards the indirect band edge of the perovskite, and at this stage their contribution to the scattering effects becomes more significant.21,45 In addition, coulombic interaction due to A-cation and rotational dynamics of organic MA and FA cations significantly alters the density of states distribution and band dispersion, which affects the width and position of the GSB band and decay kinetics.46,47 We observe the GSB band broadness increase in the case of MA3Bi2I9 and FA3Bi2I9 cations because of the increased disorder in the density of states and enhanced exciton–exciton scattering. The vibrational modes of these organic cations enhance the involvement of lattice phonons and hence increase the contribution from the indirect band edge, which increases the broadness of the GSB band.23,24
The GSB kinetics comparison is shown in Fig. 5a. By analyzing the GSB kinetics, we observe a slower initial decay in Cs3Bi2I9, suggesting that the relaxation of hot carriers occurs at a slower rate. Conversely, in FA3Bi2I9, the initial decay is found to be faster, indicating a rapid relaxation process for hot carriers. One contributing factor to this distinct kinetics can be attributed to the size disparity among the organic cations present in the nanostructures. The FA cation, being considerably larger than the MA and Cs cations, promotes a stronger interaction with the perovskite Bi-I framework. As a result, the motion of the FA cation induces larger changes in the wave functions, consequently leading to more significant couplings.21,22,48 This increased interaction highlights a stronger distribution of excited states' charge on the FA cation compared to Cs and MA. Moreover, previous studies have demonstrated that FA-based perovskites exhibit a smaller bulk modulus, indicating a softer material, in contrast to MA and Cs-based perovskites.21–23 This difference is primarily attributed to the interaction between the organic cations and the Bi-I framework, which can alter the lattice phonons within hybrid perovskites when compared to the purely inorganic Cs3Bi2I9 structure. In the case of MA3Bi2I9 and FA3Bi2I9, the modes of libration and torsion of the organic cations can couple with the Bi-I modes, which are essential for nonadiabatic coupling, ultimately leading to relaxation. Specifically, the phonon modes associated with the organic cations exhibit significant overlap with the Bi-I modes, enabling their active participation in the relaxation of hot carriers via carrier-phonon interactions. Consequently, the bismuth halide perovskites based on organic cations (MA and FA) demonstrate faster hot carrier relaxation than their counterparts based on Cs cations.
Furthermore, the low energy PIA (PIA-L) is observed in all three systems. Previously, it is mentioned that in low-dimensional materials, the broad positive absorption below the bandgap energy is typically associated with STEs and polaron formation.49–52 From the steady-state absorbance and photoluminescence features, we also observe the presence of self-trapped excitons in the A3Bi2I9 system with large electron–phonon coupling. In lead-halide perovskites, large polaronic features typically emerge (Fröhlich polaron) due to long-range interactions within the 3D connected lattice structure. However, in our A3Bi2I9 system, short-range interactions dominate, with electron-acoustic phonon coupling occurring through the deformation potential, a consequence of the spatial isolation between [Bi2I9]3− bioctahedra (0D electronic connectivity).42 The deformation potential in the lattice facilitates the formation of small polarons, which act as self-trapped excitons and lead to charge carrier trapping. Additionally, band gap renormalization in the presence of hot carriers can also contribute to the rise of this PIA-L. It is understood that hot carriers occupy fewer states at the newly lowered band edge compared to cooler carriers, resulting in an initial TA signal at the renormalized band edge, caused by newly created optical transitions due to the altered band edge states. The observed bandgap renormalization is likely attributable to an amplified contribution from photo-induced lattice fluctuations.53 Therefore, we attribute the occurrence of the PIA-L band to the simultaneous interplay of BGR and polaron formation features in the A3Bi2I9 system.
The decay kinetics of this PIA-L band are presented in Fig. 5b, which shows faster decay in the case of MA3Bi2I9 than FA3Bi2I9 and Cs3Bi2I9 exhibits slower decay of PIA-L kinetics, indicating that PIA-L decay is significantly influenced by the A-site cation size and nature. Previously, it was demonstrated in the case of lead-halide perovskite that the MA cation exhibits strong overall H-bonding because of a larger dipole on MA+ than FA+.54,55 Similarly, in our case, we state that MA3Bi2I9 has strong H-bonding capability and a large dipole moment.19 The Cs3Bi2I9 has no H-bonding capability. Interestingly, the strength of hydrogen bonding (MA > FA > Cs) is found to be inversely correlated with the decay lifetime of PIA-L. This suggests that if the dominance of hydrogen bonding strength was to govern the kinetics of PIA-L, a clear anticorrelation would be observed, indicating that weaker hydrogen bonding would result in slower decay of the PIA-L signal. As we discussed above the disorder induced by the organic cation significantly influences the exciton–phonon coupling; hence polaron formation is altered by the nature of the organic cations. In addition, bandgap renormalization also depends on the occupied density of states and band dispersion, which is influenced by the orientational dynamics and disorder of the organic cations. Therefore, MA3Bi2I9 exhibits faster decay of PIA-L. Furthermore, the dynamics of polaron formation and the kinetics of PIA-L can also be influenced by ionic migration. Ionic species, carrying charge, can significantly impact lattice fluctuations. Consequently, their migration within the material can have implications for the formation of polarons and affect the decay kinetics of PIA-L. In our previous study, it was demonstrated that FA3Bi2I9 exhibits lower levels of ionic migration compared to MA3Bi2I9.19 This difference in ionic migration suggests that the interference from polarons may be more pronounced in MA3Bi2I9, leading to a faster decay of the PIA-L signal. However, it is important to note that since A3Bi2I9 materials possess zero-dimensionality, the influence of ionic migration on excited state dynamics may not be as significant. This assumption requires further experimental and theoretical verification to confirm its validity.
Furthermore, we have observed PIA-H in all three-systems centered at ∼2.61 eV in Cs3Bi2I9, ∼2.58 eV in MA3Bi2I9, and ∼2.62 eV in FA3Bi2I9. Previously, the origin of this high-energy PIA is attributed to the photoinduced refractive index change.56 Scholz et al. proposed that this PIA-H could be associated with the Stark effect, which corresponds to the second derivate of the absorption spectra.13 In a report, Rosi et al. attributed the origin of the observed PIA-H in CsPbBr3 QDs to the light-induced activation of parity-forbidden exciton transitions facilitated by the formation of symmetry-breaking polarons.57 They elucidated that under photoexcitation, there is strong activation of the forbidden transitions, resulting in a distinct and well-defined induced absorption signal located between the two dipole-allowed excitonic transitions. This activation occurs due to a significant perturbation generated by the lattice-distorting polaron, which is formed when an exciton is excited. The strong perturbation can modify the selection rule for exciton transitions. The A3Bi2I9 system exhibits strong quantum confinement due to isolated (Bi2I9)3− dimers (0D lattice connectivity). Previously, we have demonstrated the polaron-mediated features and strong exciton–phonon coupling because of the large deformation potential.14,15,58,59 Therefore, we attribute the origin of PIA-H in the A3Bi2I9 system to the parity-forbidden exciton transitions mediated by the symmetry-breaking polarons. The cationic effect can alter the position and intensity of this band. The decay kinetics of this band are shown in Fig. 5c and Fig. S10 (ESI†) and extracted components are shown in Table 1. The first two components are shorter in the case of MA3Bi2I9 because of the high orientation and strong dipole moment of the MA cation, which induces strong lattice deformation. All inorganic Cs3Bi2I9 exhibit longer decay of this PIA-H band suggesting less lattice deformation.
Perovskite | Bands (eV) | A 1 (%) | τ 1 (ps) | A 2 (%) | τ 2 (ps) | A 3 (%) | τ 3 (ps) |
---|---|---|---|---|---|---|---|
Cs3Bi2I9 | 2.61 (PIA-H) | 87 | 25.4 | 13 | 344 | ||
2.18 (PIA-L) | 89 | 26.8 | 11 | 724 | |||
MA3Bi2I9 | 2.58 (PIA-H) | 88 | 0.26 | 10 | 136 | 2 | >500 |
2.10 (PIA-L) | 47 | 0.49 | 17 | 13 | 36 | 133 | |
FA3Bi2I9 | 2.62 (PIA-H) | 56 | 0.48 | 36 | 20.7 | 8 | >500 |
2.14 (PIA-L) | 16 | 1.24 | 62 | 30.3 | 22 | 1400 |
Furthermore, we compare the GSB, PIA-L, and PIA-H kinetics for each system (Fig. 5d–f). By scrutinizing these kinetics, we observe that PIA-L and PIA-H exhibit similar patterns of decay kinetics patterns with variations in lifetime components. The observation of similar kinetics between PIA-L and PIA-H provides compelling evidence supporting the formation of small polarons and the occurrence of light-induced activated forbidden transitions mediated by polarons within these systems. The formation of small polarons, induced by lattice distortions caused by photoexcitation, plays a crucial role in facilitating the activation of forbidden transitions and subsequently contributes to the observed kinetics. Moreover, we found an intriguing relationship between the kinetics of GSB and PIA-L. The kinetics of GSB is found to reflect those of PIA-L, indicating a connection between these two phenomena. This correspondence lends support to the notion of bandgap renormalization, which involves a modification of the effective bandgap due to interactions between excited carriers and the surrounding lattice. Additionally, the kinetics of GSB provides insight into the relaxation of hot carriers, further emphasizing the interplay between electronic and lattice dynamics within the A3Bi2I9 system.
Material | ε ∞ −1 | λ |
---|---|---|
Cs3Bi2I9 | 0.27 | 0.69 |
MA3Bi2I9 | 0.26 | 1.21 |
FA3Bi2I9 | 0.29 | 1.33 |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5tc00498e |
This journal is © The Royal Society of Chemistry 2025 |