Structurally Tunable Two-Dimensional Layered Perovskites: From Confinement and Enhanced Charge Transport to Prolonged Hot Carrier Cooling Dynamics

Abstract

Two-dimensional (2D) layered metal halide perovskites are potential alternatives to three-dimensional perovskites in optoelectronic applications owing to their improved photostabilities and chemical stabilities. Recent investigations of 2D metal halide perovskites have demonstrated interesting optical and electronic properties of various structures that are controlled by their elemental composition and organic spacers. However, photovoltaic devices that utilize 2D perovskites suffer from poor device efficiency due to inefficient charge carrier separation and extraction. In this Perspective, we shed light on confinement control and structural variation strategies that provide better parameters for the efficient collection of charges. The influence of these strategies on the exciton binding energies, charge-carrier mobilities, hot-carrier dynamics, and electron–phonon coupling in 2D perovskites is thoroughly discussed; these parameters highlight unique opportunities for further system optimization. Beyond the tunability of these fundamental parameters, we conclude this Perspective with the most notable strategies for attaining 2D perovskites with reduced bandgaps to better suit photovoltaic applications.

Hybrid organic–inorganic metal halide perovskites have become increasingly popular as efficient active layers for utilization in various optoelectronic and X-ray detection devices.¹⁻⁹ In just 10 years of research and development, the efficiency of perovskite-based devices has increased from 3.8%⁴ to above 25%,¹⁰ thus representing the fastest growing photovoltaic technology to date. These scientific breakthroughs indicate an inviting vista of commercialization for perovskite-based photovoltaic (PV) devices, not only because of their high photoconversion efficiency and low-temperature, large-scale, solution processability¹¹⁻¹⁴ but, perhaps more importantly, because of the earth-abundant availability of their raw materials and the vast tunability of the perovskite structures, bandgaps, and compositions. However, the lack of long-term device stability under ambient operation conditions remains a fundamental challenge that requires significant effort and investigation. The instability of three-dimensional (3D) perovskites is manifested through surface restructuring and degradation exacerbated by water and humidity exposure, and according to theoretical density functional theory (DFT) studies, this phenomenon is related to the low formation energy of 3D perovskites (Figure 1a).¹⁵

Figure 1.

(a) Schematic diagram showing the energetics of 2D metal halide perovskite formation and stability based on (PEA)₂(MA)_n−1Pb_nI_3n+1 with different numbers of layers (n values), where PEA is phenylethylammonium and MA is methylammonium. The figure shows the evolution of perovskite dimensionality from 2D (n = 1) through quasi-2D (1 < n < ∞) to 3D (n = ∞) accompanied by a reduction in stability due to the lower formation energy from their respective precursors.¹⁵ (b) Color tunability in perovskite films on the substrate based on the number of perovskite layers, whereby a smaller n results in larger E_g (optical values shown) and higher E_b.²⁰ (c) Schematic depicting the energy diagram of multiple quantum wells in 2D layered perovskites. The organic layers (green) form a barrier for carrier transport, and E_g is tunable depending on the composition and n. (d) Structures of monolayer 2D perovskites (n = 1), which are largely dependent on the choice of the organic spacer cation (monovalent versus divalent).⁵¹ (e) Structure of Ruddlesden–Popper and Dion–Jacobson phases (1 < n < ∞) that utilize monoamine organic barriers (L) and diamine organic barriers (L′), respectively.⁵¹ Images are reproduced from refs (a) (15) with slight modification, copyright 2016 American Chemical Society; (b) (20) with slight modification, copyright 2015 American Chemical Society; and (d and e) (51) with slight modification and permission from Elsevier, copyright 2018.

Experimental and theoretical studies have shown that the introduction of bulky organic ions/spacers that are not incorporated into the 3D perovskite lattice can generate structures with reduced dimensionalities and improved stabilities.¹⁵⁻¹⁸ The enhanced photostability and chemical stability of two-dimensional (2D) perovskites compared to their 3D counterparts are related to the higher formation energies of 2D perovskites, which is supported by additional forces (e.g., van der Waals and hydrogen bonding) that reduce molecular desorption (Figure 1a),^15,19 in addition to the improved resistivity toward moisture, which is achieved by employing hydrophobic organic spacers.^16,20,21 Therefore, 2D perovskites are often integrated into perovskite PVs as stabilizer components because of their longer device lifetime and stability. In fact, the current longest lifetime reported for perovskite PVs, which is approximately one year, has been achieved by 2D/3D interface engineering.²² However, the reduced dimensionality of perovskites confers enhanced exciton binding energies because of confinement through naturally integrated quantum well structures,^23,24 thus posing an extra challenge for their utilization in PV devices. Instead, numerous successes have been achieved for the incorporation of 2D perovskites in photonic devices that require large exciton binding energies and high radiative recombination rates, such as light-emitting diodes and lasers.²⁵⁻³¹ Furthermore, the presence of heavy metals induces spin–orbit coupling with resultant Rashba band splitting, which is of interest for spintronic applications.³²⁻³⁴

Importantly, efforts to optimize 2D perovskites for PV applications have not ceased,^16,20,35 especially after promising breakthrough reports by the Mohite group^36,37 that revealed enhanced power conversion efficiencies through the hot-casting deposition approach and disclosed new electronic states (i.e., layer-edge states) that prolong the survival of the separated charge carriers in layered 2D perovskite structures. In fact, the distinct metal-like conducting features reported at the layer-edge of 2D perovskites provide a different dimension for enhancing the performance of next-generation photovoltaics.³⁸ The vast potential for tuning the optical and electronic properties of 2D perovskites by varying their elemental composition (e.g., the choice of metals, halides, and small organic cations), organic spacers, and number of layers (i.e., thickness) holds considerable promise for further system optimization. Additionally, charge-carrier collection can be supported through optimized deposition and process engineering,^36,39−41 additive engineering (e.g., SCN^– and MACl),⁴²⁻⁴⁴ and doping approaches⁴⁵⁻⁴⁸ that favor vertical growth and preferential out-of-plane charge transport. In fact, the current certified power conversion efficiency (PCE) record for 2D perovskite solar cells in conventional structures is 16.6%,³⁹ which is still much lower than that for 3D perovskite solar cells. This was achieved by fabricating narrow-bandgap methylammonium-based 2D perovskite thin films with decreased exciton binding energies due to the relatively higher dielectric constant of the organic spacer methylammonium compared to those of conventional films such as butylammonium.³⁹ Essentially, varying the dielectric constant of the organic spacer not only improves free carrier extraction but also sheds light on the possibility of extracting hot carriers in carefully designed 2D perovskites.⁴⁹ In particular, utilizing polarized organic spacers with large dielectric constants prolongs the lifetime of the highly energetic hot carriers in 2D metal halide perovskites well beyond that of conventional 3D perovskites, which presents a new route toward exceeding the Shockley–Queisser limit⁵⁰ in PV efficiency upon proper hot carrier extraction.⁴⁹

In this Perspective, we highlight possible approaches for capturing solar energy more efficiently with structurally tunable 2D metal halide perovskites by focusing on the fundamental photophysics that dictate the efficiency of the materials as light absorbers and highlight pathways of further system optimization. In comparison, the fundamental photophysics within 3D metal halide perovskites has been widely studied and reviewed;⁵²⁻⁶⁰ however, the complex nature of the excitonic properties in 2D perovskites still requires more consideration.⁶¹⁻⁶³ Here, we emphasize modes of controlling confinement in 2D lead halide perovskites through structural variations that impact their exciton binding energies, charge-carrier mobilities, hot-carrier extraction, and electron–phonon coupling. Furthermore, we shed light on strategies for attaining 2D perovskites with reduced bandgaps, which are of interest to solar cell researchers who intend to explore more stable and efficient perovskites.

Structure and Confinement. 2D metal halide perovskites can be derived from 3D perovskites by partially or fully replacing the small cation (e.g., CH₃NH₃⁺ (MA)) with a larger organic cation, which disrupts the 3D perovskite structure with steric hindrance according to Goldschmidt’s tolerance factor.⁶⁴ Thus, structural slicing along specific crystallographic planes is achieved such that the metal-halide octahedron (inorganic layer) remains connected along only two axes, forming a layered material that stacks via weak van der Waals interactions (Figure 1a,d,e). The empirical formula of fully cation-replaced 2D perovskites is L₂MX₄, where L is the large monovalent organic cation/spacer (i.e., R-NH₃⁺, where R is a long-chain alkyl or an aromatic group), M the metal cation (e.g., Pb²⁺, Sn²⁺), and X the anion (e.g., Cl^–, Br^–, I^–). Examples of monovalent organic spacers (L) include butylammnium (BA), allylammonium (ALA), ammonium ethanol (AE), ammonium propanol (AP), and 2-phenylethylammonium (PEA). Alternatively, divalent organic cations of a diamine (i.e., ⁺NH₃-R-NH₃⁺) could be utilized to produce perovskite with a formula of L′MX₄ (Figure 1d). Examples of divalent organic spacers (L′) include 4-(aminomethyl)piperdinium (4-AMP), 3-(aminomethyl)pyridinium (3-AMPY), 1,4- phenylenedimethanammonium (PDMA), and 2,2-(ethylenedioxy)-bis(ethylammonium) (EDBE). Quasi-2D perovskite structures, namely, Ruddlesden–Popper (RP) and Dion-Jacobson (DJ) perovskites, constitute partial cation replacement whereby their layered structures are synthetically controlled by adjusting the ratio between the spacer cation and a smaller organic cation (A), which is accommodated between perovskite octahedra. RP perovskites have the formula L₂A_n–1M_nX_3n+1, where n is an integer specifying the number of metal cation layers between the organic chains (also known as perovskite thickness), based on the molar ratio of the precursors.^21,29,36,65 DJ perovskite phases are distinguished from RP perovskite phases by having a divalent organic spacer cation (L′), smaller interlayer spacing, lower bandgap energies, and an undisplaced inorganic layer stacking with a general formula of L′A_n–1M_nX_3n+1 (Figure 1e).⁶⁶⁻⁷³

Generally, perovskite thickness can be determined by X-ray diffraction measurements; for example, if every layer of an inorganic sheet is sliced (n = 1), pure 2D structures with L₂MX₄ or L′MX₄ formulas are achieved, while other integers (1 < n < ∞) are indicative of multilayered quasi-2D perovskites whose properties fall between those of 2D and 3D perovskites by combining layered stacking and polarizable central cation-templated cubic structures.⁶³ These low-dimensional perovskites are bulk materials that are confined into quantum wells due to the potential barriers imposed by the organic spacers (Figure 1c).⁷⁴ The value of n defines the degree of quantum confinement, where large n values indicate a small perovskite bandgap (E_g, Figure 1b). Increasing n in 2D perovskites allows E_g to approach values provided by their 3D counterparts, which are of better interest for PV applications whose optimal single-junction E_g is ∼1.1–1.4 eV.^20,50,74 In addition to quantum confinement, the high dielectric constant between the organic barriers and inorganic lead-halide wells strengthens 2D confinement by the image charge effect, which is also known as dielectric confinement.^75,76 Thus, enhanced quantum and dielectric confinements in 2D perovskites lead to exceptionally large bandgaps and exciton binding energies (E_b of approximately a few hundred millielectonvolts) based on the utilized organic spacer and the resultant number of layers.⁷⁵⁻⁷⁸ To this end, it is essential to explore and decipher the role of the organic spacers in controlling the dielectric confinement, exciton binding energies, and charge carrier dynamics in 2D and quasi-2D perovskites to reveal possibilities for reducing the exciton binding energy and enhancing charge carrier dissociation and conductivity.

Charge Carrier Dynamics and Mobility. The conductivity of perovskites, regardless of their dimensionality, is a complex phenomenon that depends on several factors, including in- and out-of-plane distortions, X–M–X angle, M–X distances, exciton binding energy, doping, additives, and vacancies.^8,44,79,80 Initial reports that investigated the utilization of 2D metal halide perovskites in PV devices revealed PCEs close to only 5%.^16,20 This relatively poor efficiency was attributed to the inhibition of out-of-plane charge transport by the organic cation spacers that act as insulating layers between the conducting inorganic slabs.³⁶ The hot-casting approach was developed as a solution for producing thin films of near-single-crystalline quality with the crystallographic planes of the inorganic perovskite component having a preferential out-of-plane alignment with respect to the contacts in planar solar cells, thus facilitating charge transport and improving PCEs beyond 12%.³⁶ Solvent engineering, utilizing additives, and metal doping are alternative approaches that could favor vertical growth of 2D perovskites (Figure 2a), improve crystallinity, increase grain size, and reduce defect states, thus enhancing carrier mobility and PCEs.⁴⁰⁻⁴⁸

Figure 2.

(a) Schematic illustration comparing random orientation (left) to a vertically oriented quasi-2D perovskite (right). (b) Schematic of the pump–probe (time-resolved) spectroscopy techniques that are often utilized to reveal charge carrier dynamics of various samples including perovskites (top) along with a time scale sketch for exciton generation and dynamics (bottom). (c) Schematic showing the trend in diffusion length, carrier density, and recombination upon decreasing n. Adapted and modified from a study on (PEA)₂(MA)_n−1Pb_nI_3n+1 as a function of changing the PEA-to-MA ratio.⁸⁷ (d) Charge-carrier diffusion lengths (collected at a charge carrier concentration of 10¹⁴ cm^–3) plotted versus %PEA.⁸⁷ (e) Monomolecular trapping rates plotted versus %PEA as determined from time-resolved PL measurements.⁸⁷ (f) Effective bimolecular rate constant versus %PEA based on OPTP measurements.⁸⁷ Panels c–f are reproduced from ref (87), where panel c is slightly modified; copyright 2016 American Chemical Society.

Pump–probe experiments, in which a pump pulse resonantly excites the sample, followed by a probe pulse that detects the pump-induced dynamic changes within a controlled time delay, are often utilized to understand the charge carrier dynamics and transport mechanism in perovskites (Figure 2b).^74,81−83 Typically, carrier dynamics in perovskites, like other semiconductors, can be described through the equation d[N(t)] = −k₁N(t) – k₂N²(t) – k₃N³(t), where N(t) is the carrier density at a pump–probe delay of t and k₁, k₂, and k₃ denote the monomolecular, bimolecular, and trimolecular recombination rate constants, respectively. Monomolecular recombination is generally assigned to defective nonradiative (trap) recombination channels, whereas bi- and trimolecular recombinations reflect intrinsic recombination pathways such as direct electron–hole (radiative) recombination and Auger recombination. However, the actual assignments in perovskite samples are usually more complicated due to the varying exciton binding energies for each sample, in addition to the simultaneous presence of both excitons and free-charge carriers.³⁴ Upon analyzing the carrier dynamics of 2D perovskites of different thicknesses, n = 1 samples often display faster recombination rates (in the picosecond range) and decreased carrier mobilities compared to higher-n samples that instead promote free carrier formation and slower charge carrier recombination rates (i.e., longer lifetime of approximately nanoseconds), as shown in Figure 2b,c.^34,84,85

Time-domain ab initio studies have shown that multilayered perovskites exhibit slower recombination rates due to enhanced elastic electron–phonon scattering that causes rapid loss of quantum coherence between the HOMO and LUMO, thus counteracting and suppressing the effect of their smaller energy bandgaps and higher-frequency phonons that should instead accelerate recombination.⁸⁶ Herz et al. combined optical-pump tetrahertz-probe (OPTP), time-resolved photoluminescence (T-PL), and grazing-incidence wide-angle X-ray scattering (GIWAXS) measurements to reveal the effects of changing the perovskite thickness (ME-to-PEA ratio) in PEA₂MA_n–1Pb_nI_3n+1 thin films on the orientation, carrier mobility, recombination rates, and charge carrier diffusion lengths of perovskite samples.⁸⁷ Overall, increasing the PEA% (i.e., lowering the n values) allowed the thin films to become more preferentially oriented until the crystallites were aligned parallel to the substrate for the 100% PEA film.⁸⁷ However, the trends in optoelectronic properties upon increasing PEA% were a complex interplay of different factors.⁸⁷ Among all perovskite samples, the 50% PEA sample demonstrated the highest charge-carrier mobility of 11 cm² V^–1 s^–1 and the maximum charge-carrier diffusion length of 2.5 μm, which were related to the balance between the preferential orientation of crystallites in the film and the decreased trapping rate (Figure 2d,e).⁸⁷ For samples with PEAs greater than 50%, the effect of increasing the exciton binding energy (E_b) outweighed the benefits of preferential orientation, and the effective mobility decreased significantly in accordance with a reduction in the photon-to-charge ratio (φ).⁸⁷ The monomolecular recombination rates (k₁) were affected by two counteracting trends whereby an initial decrease was recorded upon increasing PEA content (lower n values) due to the trap passivation effect, followed by an increase as excitonic recombination was enhanced (Figure 2e). This trade-off between these counteracting trends leads to the intermediate 50% PEA film exhibiting the highest charge-carrier diffusion length.⁸⁷ As for bimolecular and Auger recombination rates for PEA₂MA_n–1Pb_nI_3n+1 films, an increase was noted until the 67% PEA fraction, followed by a decrease for the 100% PEA film (n = 1) (Figure 2f).⁸⁷ These results were attributed to the initial increase in k₂ and k₃ for 0–67% PEA samples upon increased quantum confinement, followed by a decrease in the photon-to-charge branching ratio (φ) upon increased propensity for exciton formation, which counteracts the increase in the recombination constants for the 100% PEA samples.⁸⁷ Another interesting study on bromide-based perovskites with PEA ligands, i.e., PEA₂MA_n–1Pb_nBr_3n+1 (n = 1–∞), showed that quasi-2D phases with large n values (∞ > n > 40) could exhibit improved PV performance while having relatively low carrier mobility, as inferred from charge extraction measurements and ab initio calculations.⁸⁸ In fact, the reduced mobility contributes positively to having higher open-circuit voltage (V_OC) in quasi-2D phases (∞ > n > 40) compared to their 3D counterparts through favoring better charge carrier extraction and less recombination at the selective contacts.⁸⁸

While low-n 2D perovskites could exhibit unfavorable carrier-transport parameters, as shown above, their conductivity and PV performance can still be improved through careful organic cation design, additives, and doping that influence perovskite orientation, layer stacking, and trap densities.^44,45,89 An investigation of a series of n = 1 layered perovskites of the form (aromatic-O-linker-NH₃)₂PBI₄, where the aromatic moiety is naphthalene, pyrene, or perylene, has revealed enhanced out-of-plane conductivity upon using aromatic cations whose energy levels better match those of the inorganic layers (i.e., perylene and pyrene) and whose intramolecular hydrogen bonding and supramolecular π-stacking are enhanced.⁸⁹ Additionally, varying the linker moiety between ethyl, propyl, and butyl groups had a significant influence on the π-stacking interactions across gaps, which in turn controlled out-of-plane conductivity, such that higher conductivity was noted upon having edge-to-face arrangement in the crystal structure.⁸⁹ These observations suggest that future efforts to improve conductivity must focus on how to control the special arrangements of electronically active cations within layered perovskite crystal.⁸⁹

A theoretical study by Frost et al.,⁹⁰ which suggested that increasing the dipole moment of the organic cation through fluorination could improve charge separation and increase carrier lifetime by polarizing the crystal lattice, has inspired several researchers to explore polarizable spacers for enhancing charge carrier transport.^19,91,92 One example reported that fluorine substitution on the para position in PEA (F-PEA) demonstrates a direct impact on the intermolecular packing, as well as the electronic interactions of (F-PEA)₂MA_n–1Pb_nI_3n+1 perovskite, thus leading to improved charge transport, elongated charge carrier lifetime, and significantly reduced trap density in comparison to the pristine PEA-based analogue perovskite.⁹¹ In fact, the desired out-of-plane (intersheet) transport in 2D (n = 1) perovskite thin films of (F-PEA)₂PbI₄ showed a 7-fold improvement in mobility compared to the nonfluorinated analog, as determined by time-resolved microwave conductivity (TRMC) measurements.⁹¹ Analysis of the TRMC measurements for quasi-2D (n = 5) perovskites yielded peak mobility values of ∼7 cm² V^–1 s^–1 for (PEA)₂MA₄Pb₅I₁₆ and ∼9 cm² V^–1 s^–1 for (F-PEA)₂MA₄Pb₅I₁₆, which along with improved carrier lifetime, supported an increase in PCE from ∼9.6% to ∼13.6% upon fluorination without using any additives and through room-temperature fabrication (without hot-casting).⁹¹

Alternatively, employing divalent organic spacers (L′) is expected to shorten the interlayer distance and improve charge carrier transport in 2D DJ perovskites relative to 2D RP perovskites that employ monovalent ligands such as PEA (Figure 1e).^67,72,93 Nevertheless, the charge carrier lifetime studies for the 2D layered (aminomethyl)piperdinium (4-AMP) perovskite series (n = 1–4) are in the range of 0.1–0.3 ns, which are comparable to those of perovskites with PEA ligands.⁶⁶ Kanatzidis and co-workers employed other aromatic spacers that are similar in size to the aliphatic ligand AMP but offer increased rigidity, shorter interlayer distances, charge delocalization, and decreased dielectric mismatch between the inorganic perovskite and the spacer, namely, through the (aminomethyl)pyridinium (AMPY) spacer.⁷¹ The photoluminescence lifetimes of the 3-AMPY-based series were longer than those of the 4-AMPY and AMP series, which is an indication of slower carrier recombination and improved carrier transport for the 3-AMPY-based 2D perovskite.⁷¹ This reveals that not only does the nature of the spacer control the perovskites’ optoelectronic properties through dielectric confinement but also even the position of the functional groups within the organic spacers plays an essential role in defining the structural variations (i.e., eclipsed, distorted, distances, angles, etc.) in the layered perovskite, which greatly affects the optoelectronic and carrier transport properties.⁷¹ A recent DFT study showed that introducing electron-withdrawing or electron-donating molecules within the organic spacers in 2D perovskites leads to the formation of localized states, in either the organic or the inorganic part, which can be tuned independently.⁷⁸ Hence, the inclusion of novel organic cations in the organic layers offers the possibility of improving the charge transport in the organic part by engineering π-stacking.⁷⁸

Carrier Dissociation and Exciton Binding Energies. The reduced dimensionality in 2D and quasi-2D perovskites significantly influences the dissociation of photoexcited carriers. Several studies have shown that within a homologous series of 2D perovskites, large n values decrease the exciton binding energies (E_b) due to reduced quantum confinment.^78,84,94Figure 3a depicts an example of the fraction of collected free charge as a function of temperature for RP perovskite with the (BA)₂(MA)_n−1Pb_nI_3n+1 (n = 1–5) formula, as obtained from TRMC measurements.⁸⁴ Clearly, more free charges are collected for higher n values upon decreasing their E_b from 370 meV (n = 1) to 80 meV (n = 4).⁸⁴ Even higher n values (n = 6 and 7) of the series were more recently explored, revealing more efficient exciton dissociation while contributing to optimized bandgaps and effective dielectric constants for PV applications (Figure 3b,c).⁶⁵ Importantly, exciton stability does not arise from dimensional confinement alone, but rather, the organic spacers play an important role in modulating the dielectric properties of the perovskite material and controlling the E_b.⁷⁵⁻⁷⁸

Figure 3.

Factors affecting carrier dissociation and exciton binding energies (E_b). (a) Fraction of free charges as a function of temperature for RP perovskite (BA)₂(MA)_n−1Pb_nI_3n+1 (n = 1–4) obtained from TRMC measurements.⁸⁴ (b) Optical bandgap as a function of n values (red line) and the corresponding E_b versus n values, which are calculated from the analytical model of 2D perovskites (blue line).⁶⁵ (c) Effective dielectric constant (ε_eff) as a function of n values based on the model proposed by Ishihara and co-workers,⁹⁹ illustrating that as the thickness of 2D perovskites is increased, the dielectric properties of the materials asymptotically approach those of 3D perovskites.⁶⁵ (d) E_b and the Bohr radius versus the dielectric constant of the organic group in 2D perovskites as predicted by the image charge model. The calculated data for (PEA)₂PbI₄ (square) and (EA)₂PbI₄ (diamond) are indicated with their lattice structures in the inset.⁹⁵ (e) Effect of I₂ intercalation represented through slabs of (IC₆H₁₂NH₃)₂[PbI₄] and (IC₆H₁₂NH₃)₂[PbI₄]·2I₂ and their corresponding calculated high-frequency dielectric constants perpendicular to the direction of layer propagation (ε_∞,⊥).⁹⁷ (f) Band diagrams of (BA)₂PbI₄ with and without lead vacancy V_Pb, whose modification affects E_b.⁹⁸ Images are reproduced with permission (https://creativecommons.org/licenses/by/4.0/) from refs (a) (84), copyright 2017 American Chemical Society; (b and c) (65) with slight modification, copyright 2019 Proceedings of the National Academy of Sciences; (d) (95), copyright 2018 Springer Nature; and (e) (97), copyright 2017 Royal Society of Chemistry. Panel f is reproduced with slight modifications from ref (98), copyright 2019 American Chemical Society.

Theoretical and experimental studies have shown that employing organic barriers of large dielectric constants yields low dielectric-confined 2D perovskites with improved photoexcited carrier dissociation, enhanced charge carrier mobility, and reduced E_b values.^39,73,95 For example, when organic spacers with high dielectric constants, such as ammonium ethanol AE (ε_AE ≈ 37), were compared to conventional PEA spacers (ε_PEA ≈ 3.3) in 2D perovskites, the Bohr radius describing the mean distance between the electron and hole in the exciton increased because of the enhanced dielectric screening effect, and accordingly, E_b was estimated to decrease from ∼250 meV for (PEA)₂PbI₄ to ∼13 meV for (AE)₂PbI₄ (Figure 3d).⁹⁵ Another recent study aimed to develop solely MA-based 2D perovskites to benefit from the spacer’s relatively higher dielectric constant MA (ε_MA ≈ 11) compared to the conventional bulky cations of lower dielectric constants such as PEA and BA (ε_BA ≈ 4).³⁹ Vapor-fumigation technology enabled the development of MA-based 2D perovskite thin films whose E_b values were significantly reduced in comparison to those of BA-based 2D perovskites (from 510 to 172 meV) while also enhancing the tunneling probability of the carriers by 4 orders of magnitude because of the smaller layer spacing, as confirmed by XRD.³⁹ This has led to a certified record PCE of 16.6% for MA₂PBI₄-based PV devices with superior long-term stability under illumination and exposure to environmental conditions.³⁹ Interestingly, if the dielectric constant of the organic spacer is sufficiently large, E_b is predicted to be significantly reduced, approaching those in 3D perovskites.⁹⁵ In accordance with the studies that suggest employing fluorinated spacers to improve the charge carrier transport and mobility,^19,91,92 a positive influence was also noted toward achieving low dielectric-confined 2D perovskites of reduced E_b upon fluorination.⁹⁶ It is reported that within a series of fluorous organic cations, longer cations favor reducing the perovskites’ bandgap and its E_b while offering enhanced stabilities due to the improved hydrophobicity with fluorine atoms.⁹⁶

In comparison, Karunadasa and co-workers managed to reduce the E_b value of RP 2D perovskites by employing targeted postsynthesis modification, which involves the intercalation of a highly polarizable molecule, namely, iodine, into the organic layers.⁹⁷ In particular, the E_b of the starting (IC₆H₁₂NH₃)₂PbI₄ was estimated to be ∼230 meV, which is lower than those of typical 2D perovskites because of the polarizability of organoiodines, and it was further reduced to ∼180 meV upon intercalation with iodine molecules.⁹⁷ This reduction in E_b was consistent with the enhancements in dielectric constant profiles in both the inorganic and organic layers upon I₂ intercalation (Figure 3e).⁹⁷ Although the I₂-intercalated perovskites were metastable, this study proved again that enhanced dielectric constants lead to reduced dielectric confinement of the excitons in the inorganic layers, supported by the decrease in E_b upon I₂ intercalation.⁹⁷

Quasi-2D DJ perovskites were also examined for E_b reduction and for improved optoelectronic properties owing to the smaller degree of distortion in their crystal structure.^68,69,73,93 Experimental investigations of 1,4-phenylenedimethanammonium (PDMA)-based DJ perovskites, i.e., (PDMA)(MA)_n−1Pb_nI_3n+1, disclosed that these perovskites transform gradually from the confinement structure (for small n values of 1 and 2) to the nonconfinement structure in larger n values (n ≥ 3), which is attributed to the squeezing of PDMA spacers by the octahedral layers to a narrow spacing ∼3.4 Å.⁶⁹ Consequently, the nonconfinement structure displayed E_b of ∼69 meV which expedite exciton dissociation and interlayer charge transport.⁶⁹ This is in agreement with recent DFT calculation on ethylenediamine cation (EDE)-based 2D DJ perovskites, whereby n ≥ 3 samples offered enhanced stability, smaller effective masses, larger dielectric constants, and lower E_b in comparison to their conventional 3D counterpart.⁷³ In fact, E_b of (EDA)Cs_n–1Pb_nI_3n+1 perovskites (n ≥ 3) falls between 29 and 46 meV, and their enhanced stability is related to the strong I–H interaction of diamine cations with a shortened interlayer distance of ∼3.5 Å.⁷³ These studies emphasize the active role played by the organic cation spacer in controlling the properties of 2D perovskites and highlight DJ perovskites as unique alternatives for improved charge carrier separation and stability.

Defect engineering is another feasible approach for controlling the E_b of 2D perovskites. Through first-principles calculations, Lu et al. predicted the influence of many defects, among which only neutral Pb vacancies (V_Pb) in I-rich environments resulted in a drastic decrease in the E_b of atomically thin 2D perovskites.⁹⁸ In particular, when a Pb atom is removed from the lattice of (BA)₂(MA)_n−1Pb_nI_3n+1 (n = 1 or 2), nearby I atoms become electron deficient, which leaves the highest occupied I 5p level vacant, as shown in Figure 3f.⁹⁸ Thus, excitation no longer occurs from the I 5p to Pb 6p orbitals but rather between the two I 5p orbitals themselves, which results in a reduction in E_b.⁹⁸ For instance, (BA)₂PbI₄ revealed a decrease in its E_b from 340 to 110 meV upon introducing V_Pb at a concentration of ∼1.3 × 10²⁰ cm^–3.⁹⁸ Further, V_Pb’s are shallow defects that attract electrons and repel holes, thereby assisting in providing spatially separated and delocalized charge carriers with minimized recombination rates.⁹⁸

Electron–Phonon Coupling and Hot Carrier Relaxation. Notably, in addition to yielding stable excitons with high binding energies, the confinement in 2D perovskites often facilitates exciton interaction with phonons, spins, and defects.¹⁰⁰ The interaction between charge carriers and phonons (i.e., lattice vibrations) is based on the electron–phonon or hole–phonon coupling strength. Because of the polarity of perovskite crystals, carrier scattering is usually dominated by coupling to longitudinal optical (LO) phonons,⁶³ which is evaluated by the Fröhlich coupling parameter (α).⁴⁹ This presents an intrinsic pathway for exciton relaxation, which influences the charge-carrier mobilities, transport, and recombination. Surely, the molecular structure of the organic spacer controls the crystal rigidity and thus the exciton–phonon coupling strength.^101−103 An example was reported by Sargent and co-workers showing that single crystals of 2D perovskites with BA spacers possess electron–phonon coupling that is 2-fold stronger than those employing PEA ligands based on temperature-dependent PL studies and deformation potential (D) extraction.¹⁰¹ Importantly, both samples had comparable defect densities, and thus, the reduced electron–phonon coupling for the more rigid PEA-based perovskite was held accountable for the observation of slower nonradiative decay and higher PL quantum yields.¹⁰¹ For 2D perovskites that have strong distortions in their inorganic lattice, the exciton–phonon couplings lead to low-energy self-trapped excitons (STEs) that are caught as a small polaron in the lattice distortion field (Figure 4a), favoring broadband emissions with large Stokes shifts.^80,104−106 In transient absorption (TA) experiments, STEs can be distinguished from defect trapping because STE signatures appear on a wavelength-dependent subpicosecond or picosecond time scale.¹⁰⁶ Additionally, STEs typically have longer lifetimes when compared to free exciton emission; for instance, free exciton emission in DJ perovskite (EDBE)PbI₄ had a lifetime of ∼10 ps compared to ∼1–3 ns for white light emission from STEs.¹⁰⁷

Figure 4.

(a) Schematic illustration of large polaron formation and long-range distortions upon adding an electron to an ionic lattice.¹⁰⁵ (b) Schematic of the carrier relaxation process in 2D perovskites upon the above band-edge excitation.¹¹⁰ (c) Schematic of the optical (top) and acoustic (bottom) phonon vibrations coupled with hot carrier relaxation, where the vertical dotted lines represent the equilibrium position of the atoms in the perovskite crystal lattice.¹¹⁰ (d) Hot carrier temperature as a function of delay times for PEA-based 2D perovskite single crystals of different thicknesses.¹¹⁰ (e–g) Schematics of the hot carrier relaxation and electron–hole recombination processes upon high-energy excitation of 2D perovskites that utilize organic spacers of different dielectric constants, along with schematics of their single crystal interlayer distances.⁴⁹ Images are reproduced from refs (a) (105) with slight modification, copyright 2018 American Chemical Society; (b–d) (110) copyright 2019 American Chemical Society; and (e–g) (49) with slight modification, copyright 2019 American Chemical Society.

Furthermore, electron–phonon coupling controls the cooling process of hot carriers that embrace high energies (>E_g of perovskite material).^49,108,109 In fact, upon above band-edge excitation, the generated hot carriers can relax through nonradiative phonon emission in the form of heat, mainly through the decay of LO phonons into coherent acoustic phonon oscillations (Figure 4b,c).^110,111 Slowing phonon relaxation allows the carriers to stay “hot” for a longer time through what is known as the “hot phonon bottleneck effect”.¹⁰⁸ The prolonged lifetime of the highly energetic hot carriers may facilitate their proper extraction and offer a unique opportunity toward utilizing solar energy more efficiently. 3D metal halide perovskites such as MAPbI₃ were reported to have ∼400 fs hot carrier cooling time, which is adjusted upon cation, halogen, and heavy metal variations.^56,112−114 On the other hand, 2D perovskites offer an additional variable parameter by organic cation tuning. Compared to their 3D counterparts, 2D perovskites have shown encouraging initial results by offering reduced phonon velocity and cross-plane propagation lengths due to the weak van der Waals bonding between the organic spacers and due to the large acoustic impedance mismatch between alternating layers of the perovskite and the bulky organic cations.¹¹¹ Furthermore, the organic spacers in 2D perovskites can enhance the upconversion of confined acoustic phonons to optical phonons, which presents a mode of obtaining long-lived hot carriers.¹⁰⁸ Moreover, tuning the number of layers and varying the nature of the organic spacers in 2D perovskites allows control of the generation and propagation of coherent longitudinal acoustic phonons (CLAPs), which are crucial for thermal conductivity devices.¹¹⁰ For example, a recent study employed femtosecond transient reflectance (TR) spectroscopy on PEA-linked perovskites (n = 1–3), which revealed slower hot carrier cooling, lower LO phonon emission, and decreased CLAP velocity upon increasing the number of layers, which is mainly attributed to the increased mass of inorganic layers and their coupling with the organic cations, along with a decrease in E_b and reduced electron–phonon coupling (Figure 4d).¹¹⁰ Furthermore, increasing the dielectric constant of the organic spacers slows hot carrier relaxation.^49,110 Recent TR spectroscopic studies and time-domain ab initio calculations on (PEA)₂PbI₄, (AP)₂PbI₄, and (AE)₂PbI₄ RP perovskites have proven that the influence of organic spacers extends beyond controlling the perovskite’s interlayer distances and exciton binding energies to manage Coulomb interactions and electron–LO phonon couplings based on their dielectric constants.⁴⁹ Accordingly, the hot carrier cooling dynamics decreased by a factor of 3 for (AE)₂PbI₄ (ε_AE ≈ 37, α ≈ 3.82) compared to (PEA)₂PbI₄ (ε_AE ≈ 3.3, α ≈ 4.29), which was attributed to stronger Coulomb interaction screening, weaker electron–LO phonon coupling, and reduced nonradiative internal conversion within the conduction bands (Figure 4e–g).⁴⁹ These studies demonstrate that careful perovskite design through controlling the number of layers and the nature of the organic spacers can prolong hot carrier thermalization and possibly facilitate hot carrier extraction through interface engineering to execute PVs that operate beyond the Shockley–Queisser limit (i.e., ∼33% for a single-junction solar cell).⁵⁰

Outlook for Optimizing 2D Perovskites for Photovoltaics. Variations in the organic cation length, aromaticity, rigidity, and dielectric constant offer routes for tuning exciton binding energies and charge carrier dynamics. As such, understanding the charge carrier dynamics in 2D perovskites is crucial not only for the fundamental knowledge of this promising class of materials but also for optimizing the desired optoelectronic devices. The reported reduction in exciton binding energies and elongation of hot carrier lifetimes upon employing low dielectric-confined 2D perovskites offer exceptional promise for effective utilization in photovoltaic devices.^39,49 Future research should explore the possible contribution of these designs to the overall PCE in devices. Furthermore, it is essential to examine other compositions that also provide matched energy levels between the organic and inorganic layers and favor supramolecular π-stacking and optimized structures for enhanced conductivity.^71,89 In addition, metal doping is an effective approach for modulating the fundamental properties of 2D perovskites, and it offers plenty of room for studying the interaction of electrically, optically, and magnetically active dopants with quantum confinement effects.^47,115−118

In addition to hot carrier extraction, another process that allows utilizing solar energy more efficiently is multiple exciton generation (MEG) or carrier multiplication,^119,120 whereby a single high-energy photon produces multiple electron–hole pairs instead of merely dissipating its excess energy as heat. This behavior has been observed in 3D metal halide perovskites¹²¹ and recently studied in 2D perovskites as well.¹²² Resonant impulsive stimulated Raman spectroscopy (RIRS) investigations on (PEA)₂PbI₄ and (BA)₂PbI₄ 2D RP perovskites have shown the coexistence of multiple excitons with distinct lattice couplings to low-frequency phonons that are separated by 35 meV; some excitons coupled to vibrations in the crystal plane, while others coupled to out-of-plane motion, hence providing insight into the possibility of producing multiple excitons in 2D perovskites.¹²² Forthcoming research should explore the possibility of proper extraction of multiple carriers in 2D perovskite devices by reducing Auger recombination rates, which are influenced by several factors, including the band structure, the presence of phonons, confinement, and the concentration of impurities.⁸⁷

Quasi-2D perovskites, especially those with large n values, provide access to intermediate bandgaps and exciton binding energies that approach the optimal standard values of 3D perovskites. However, the synthesis of 2D perovskites with large thicknesses is challenging, as they are less thermodynamically favorable than low-n perovskites. Accordingly, much progress is still needed to achieve the controllable synthesis of pure monodisperse phases with large n values. Careful selection of the organic spacer cation could greatly influence the distribution in quantum well widths of large-n 2D perovskites and provide flatter energy landscapes with reduced carrier trapping.^68,74,123 On the other hand, optimizing the bandgaps of monolayered and small-n 2D perovskites to achieve efficient utilization in single-junction solar cells is another fundamental challenge that continues to attract the attention of researchers in the solar cell community.¹²⁴ In terms of composition, the steric size of the organic cation largely affects the perovskite bandgap by influencing the M–X–M bond angles and M–X bond lengths.^125−127 Halide mixing is an alternative approach that is commonly utilized for tuning the bandgap, in addition to utilizing metals with smaller ionic radii that favor reducing the perovskite bandgap.^128−132 Importantly, the extent of hydrogen and halogen bonding at the organic–inorganic interface should also be considered for optimal bandgap manipulations.¹³³ Efforts have also extended beyond traditional synthetic designs toward applying external stimuli such as temperature and pressure to modify perovskite structures and hence their optical and electronic properties.^134−136 For instance, a long-term bandgap narrowing from 1.94 to 1.78 eV was achieved in (BA)₂(MA)₂Pb₃I₁₀ (n = 3) after compression to 26 GPa and then decompression to ambient pressure.¹³⁴ This bandgap reduction is explained by the narrowing of Pb–I–Pb bond angles and the shortening of Pb–I bond lengths, which influence the overall overlapping electron density wave function.¹³⁴ Notably, considerable bandgap reduction was also reported upon applying a moderate pressure of 3.5 GPa on (PEA)₂PbI₄ with full reversibility and immediate responses to pressure application.¹³⁵ As such, 2D metal halide perovskites are unique “malleable” semiconductors whose properties can be drastically modified with small changes in temperature and pressure, providing plenty of room for material design and property engineering.¹³⁶ Conversely, it is noteworthy to mention that wide-bandgap 2D perovskites with E_g ≈ 1.9–1.7 eV remain of interest for the fabrication of tandem solar cells.¹³⁷

Summary. The simple processability, improved stability, and additional structural versatility of 2D metal halide perovskites in comparison to their 3D counterparts place them at the center of current research for high-efficiency solar cells and optoelectronic devices. In particular, the compositional flexibility of layered 2D perovskites beyond Goldsmidt’s tolerance factor presents a unique path toward employing different organic spacers of variable quantities to tune the optical, electronic, and carrier-transport properties. Among the fundamental properties that dictate the efficiency of 2D perovskites are their bandgap, exciton binding energies, charge-carrier mobilities, hot-carrier extraction, electron–phonon coupling, and carrier multiplication. The use of ultrafast spectroscopies to study the properties of 2D perovskites in their excited state is valuable for understanding these materials systematically upon varying their composition and layer thickness. The performance of 2D perovskites as active layers in solar cells can then be enhanced through alignment engineering by careful selection of the hole transport layer, electron transport layer, and/or scaffold within the device architecture. It is thus expected that deciphering the fundamental properties and improving the performance of these tunable systems as their complexity increases are feasible through continued collaborative efforts from multidisciplinary researchers.

The authors declare no competing financial interest.