Pengfei
Wu
ab,
Dewang
Li
*cd,
Shirong
Wang
*ab and
Fei
Zhang
*ab
aSchool of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, China. E-mail: wangshirong@tju.edu.cn; fei_zhang@tju.edu.cn
bCollaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin 300072, China
cInternational Campus of Tianjin University, Joint School of National University of Singapore and Tianjin University, Fuzhou 350207, China. E-mail: lidewang@tju.edu.cn
dSchool of Materials, Science and Engineering Tianjin Key Laboratory of Composite and Functional Materials, Tianjin University, Tianjin 300350, China
First published on 6th March 2023
Guanidinium (GA+) has been proven as the multifunctional cation in halide perovskite solar cells (PSCs), attracting considerable attention due to its improved power conversion efficiencies (PCE) and operational stability. This review introduces the elementary properties of the GA cation and GA-based perovskite. Then, we summarize the development history of utilizing guanidinium materials in PSCs, including composition engineering, additive engineering, and interface engineering, accompanied by their optimization mechanisms. Moreover, we propose a perspective on the potential opportunities in this popular field and challenges to the future design of guanidinium-based PSCs with enhanced performance.
To overcome the abovementioned drawbacks, bulky or long-chain organic cations such as butylammonium (C4H9NH3+ [BA+]), phenethylammonium (C8H9NH3+ [PEA+]), and octylammonium (CH3(CH2)7NH3+ [OA+]) could be introduced to passivate the various trap states and enhance the stability of the perovskite.9–11 These cations demonstrate significantly improved operational stability under multiple environments (e.g., moisture, thermal, and illumination). Unfortunately, their positive effects on the photovoltaic performance are not satisfying and are primarily limited by the hindered carrier transport because of the relatively low conductivity.12,13 Additionally, these cations hardly form and diffuse into the 3D perovskite crystal structure due to the large ionic radius.
Among various kinds of cations, GA+ is substantially larger in ionic radius (278 pm) as compared to the commonly used MA+ (217 pm) and FA+ (253 pm) cations while smaller than BA+ (393 pm) and PEA+ (437 pm). Still, it does not form a 3D perovskite structure as the sole A-cation in terms of the t value (see detailed information in Table 1).14,15 Regardless of the phase of the GA+-containing perovskites, positive interactions are usually found endowed by GA+, as long as it could be introduced in the system. GA has nearly zero dipole character, which has been proven to be effective in suppressing bias-induced ionic mobility and hysteresis characteristics.16,17 In addition, GA+ can offer a lone pair of electrons to specifically passivate the imperfections in the surface and bulk perovskite films, i.e., ionic vacancies, uncoordinated Pb, and/or Pb clusters.18 Most importantly, owing to the plentiful ammonium groups in the molecular structure, GA+ could provide more hydrogen bonds to coordinate with the surrounding halides, especially undercoordinated I− ions firmly, thus increasing the formation energy of halide vacancies and improving the perovskite structural stability.19,20 As a consequence, the charge carrier lifetime and photovoltaic performance, such as the cell's open circuit voltage (VOC) and fill factor (FF) are expected to be ameliorated.
| A-site cations | Formula | Radius (pm) | t | μ e (Debye) |
|---|---|---|---|---|
| Note: the t is calculated based on the chemical formula APbI3; the ionic radius of Pb and I are 220 pm and 119 pm, respectively. “—” means the absence of relative information elsewhere. | ||||
| Potassium | K+ | 137 | 0.74 | 10.82 |
| Ammonium | [NH4]+ | 146 | 0.76 | 0 |
| Rubidium | Rb+ | 152 | 0.78 | 11.48 |
| Cesium | Cs+ | 167 | 0.84 | 11.69 |
| Methylammonium | [(CH3)NH3]+ | 217 | 0.91 | 2.17 |
| Azetidinium | [(CH2)3NH2]+ | 250 | 0.98 | 2.51 |
| Formamidinium | [NH2(CH)NH2]+ | 253 | 0.99 | 0.60 |
| Dimethylammonium | [(CH3)2NH2] + | 272 | 1.02 | 1.80 |
| Ethylammonium | [(C2H5)NH3]+ | 274 | 1.03 | 4.49 |
| Guanidinium | [C(NH2)3]+ | 278 | 1.04 | 0 |
| Tetramethylammonium | [(CH3)4N]+ | 292 | 1.07 | 0 |
| Ethylenediammonium | [NH3(CH2)2NH3]22+ | 333 | 1.15 | — |
| Butylammonium | [(C4H9)NH3]+ | 393 | 1.28 | — |
| Phenylethylammonium | [C2H4(C6H5)NH3]+ | 431 | 1.35 | 1.5 |
| Butane-1,4-diammonium | [NH3(CH2)4NH3]22+ | 457 | 1.41 | — |
| Hexane-1,6-diammonium | [NH3(CH2)6NH3]22+ | 582 | 1.67 | — |
The GA cation-based material was first introduced in the dye/liquid electrolyte interface as an additive by Jeanbourquin et al.24 to improve the performance of dye-sensitized solar cells (DSSCs) in 2014. The additive could eliminate the interfacial dye–I2 complexes, increasing the collection efficiency and thus the photocurrent. Inspired by this, in 2016, Marco et al.25 also added the GA-based additive firstly into the perovskite precursor to suppress nonradiative recombination defects of iodide vacancies in the bulk and iodine species at grain boundaries, resulting in superior performance with PCEs of more than 17% and a minor VOC loss of 0.43 eV. Encouraged by the progress, GA-based materials seemed to be good candidates for PSCs, which have acquired rapid development in the past few years. In this passage, representative endeavors on the GA+ species for PSC developments have been summarized, which could be categorized into three strategies of compositional engineering, additive engineering, and interface engineering, among which the first one could be further classified into 3D compositional or 2D alternating-cation-interlayer (ACI) engineering types according to the perovskite matrix (Fig. 1). Although there have been reports of further PCE performances of over 24%, their simultaneous long-term stabilities against thermal, illumination or other harsh conditions have yet to be presented.20,26 Therefore, how to simultaneously achieve both excellent efficiency and stability of the device would be the significant challenge in the future for GA-based PSCs. This systematic review could help us identify the fundamental mechanisms and the origins of the existing problems. Conversely, it further opens access to rational designs to overcome the challenges and fabricate more stable and efficient PSCs.
![]() | ||
| Fig. 1 (a) The schematic diagram of strategies for the GA+ cation in the PSCs. (b) The development of GA+-based PSCs. | ||
![]() | ||
| Fig. 2 Schematic representation of the halide perovskite with different dimensionalities from 3D to 0D at the structural level (top) and by morphology (bottom). | ||
When t is in the range of ∼0.8–1.0, a stable 3D perovskite structure can be formed. The 2D perovskite is preferred for inclusion in the case of t > 1,27 which generally adapts a formula of (A′)m(A)n−1BnX3n+1. Here, A′ is structurally featured with divalent (m = 1) or monovalent (m = 2) cations, and n indicates the number of layers consisting of [BX6]4− octahedra. 1D perovskite micro/nanocrystals generally present a well-defined fiber or ribbon-like morphology. Distinct from the uniform 3D growth, the self-assembly of the precursors into a 1D nanostructure reacting in a particular solution or vapor environment, as well as the ordered texture aligned by external forces such as casting and evaporation techniques or the periodically confined nano-templates, consists of the two prerequisites for practical 1D perovskite applications.28 0D perovskites have the formula
consisting of A′′—a monovalent organic or inorganic cation, and isolated BX6 octahedral anions or separated clusters from each other by the surrounding inorganic or organic cations without certain composition and periodic bands' formation. Therefore, the t of 0D perovskite often deviates from 1.29
Within the 3D framework, the inorganic layers of the [BX6]4− octahedra are connected via sharing each of their six corners, while the A cations occupy voids between each layer.30 In the case of B and X, respectively, representing pure Pb2+ and I−, the notorious B site oxidation and phase segregation problem can be avoided. When the A site is Cs+, MA+, FA+, and their alloys, t falls typically in the range of 0.8–1, which ensures a 3D cubic/tetragonal structure. Compared to cations with moderate radii in the 3D perovskites, the A-site in a 2D perovskite occupied by large radii organic cations such as PEA+ and BA+ would result in t > 1. They cannot fit into the voids between the octahedral units. After insertion, the A-site cations expand the van der Waals gaps, making the inorganic layers discrete, along with the broadened band gaps.31
Different from MA+ and FA+, GA+ is highly symmetric and non-polarized, and the lattice can be stabilized by its abundant hydrogen bonds. However, because of the large cation radii leading to the high lattice stress, the 3D [PbI6]4− octahedra networks break into the 1D ribbon (Fig. 3a) of the GAPbI3 form with edge-sharing octahedra or GA2PbI4, which is a 2D layer structure (Fig. 3b) with corner-sharing octahedra.32 The crystallographic data of GAPbI3 and GA2PbI4 are summarized, as shown in Fig. 3c and d, and Table 2. In the GA–Pb–I system, GA in the unit cell is approximately orientated along the (11
) plane and connects with six I ions to generate N–H⋯I hydrogen bonds within the PbI6 octahedra, in the most stable structure. Furthermore, compared to MAPbI3 and FAPbI3 perovskites, the GA-based perovskite has more N–H⋯I hydrogen bonds in the unit cell (the number of N–H⋯I hydrogen bonds in MAPbI3, FAPbI3, and GAPbI3 are 3, 4, and 6, respectively) (Fig. 3e).33 Different conditions could regulate the formation of diverse dimensional GA–Pb–I perovskites. A solution process, such as spin-coating, is expected to deposit the 1D GAPbI3 and 2D GA2PbI4 perovskite film with a stoichiometric ratio of PbI2 and GAI.
![]() | ||
| Fig. 3 SEM images of (a) GAPbI3 and (b) GA2PbI4 (high-resolution SEM image as inset). Crystal structure of (c) 1D GAPbI3 and (d) 2D GA2PbI4. Reproduced with permission.32 Copyright 2021, ACS Publishing Group. (e) PbI6 inorganic framework and N–H⋯I hydrogen bonds in MAPbI3, FAPbI3, and GAPbI3. Reproduced with permission.33 Copyright 2018, ACS Publishing Group. (f) UV-vis absorption and (g) steady-state photoluminescence spectra of MAPbI3, FAPbI3, GAPbI3, and GA2PbI4. Reproduced with permission.36 Copyright 2016, Wiley-VCH. | ||
| Chemical formula | GAPbI3 | GA2PbI4 |
|---|---|---|
| Formula weight | 641.93 | 834.97 |
| Crystal system | Monoclinic | Orthorhombic |
| Space group | Pnma | P21/n |
| a (Å) | 11.9866(5) | 9.2440(3) |
| b (Å) | 4.4740(2) | 26.9511(11) |
| c (Å) | 20.8652(10) | 12.7155(3) |
| α (deg) | 90 | 90 |
| β (deg) | 90 | 91.482(2) |
| γ (deg) | 90 | 90 |
| Volume (Å3) | 1118.96(9) | 3166.82(17) |
| Z | 4 | 8 |
| ρ calc (g cm−3) | 3.81 | 3.503 |
The photophysical property of the GA-based perovskite, in contrast to MAPbI3 and FAPbI3, has an apparent blue shift of the absorption, which is also ascribed to the disruption of the coherent 3D PbI6 octahedra by the large size of the GA cation.34 Thereby, GA-based perovskite, for example, GAPbI3, provides a higher band gap (Eg) of 2.25 eV.35 The photoluminescence spectra of both GAPbI3 and GA2PbI4 also show the apparent shifts toward higher energy, which is in agreement with the results of the absorption spectra36 (Fig. 3f and g). Due to the wide bandgap, the GA-based perovskite could not generate a sufficient visible light-induced carrier. Moreover, GAPbI3 and GA2PbI4 usually exhibit poor uniformity in film processing, which makes them ineffective for direct use as photoactive layers in PSCs. However, it would become effective in the case of GA as a supplementary composition, as introduced by the above-mentioned three strategies in Fig. 1.
Besides the improved photovoltaic performance, the long-time thermal and photo instabilities are also of vital concerns. Jodlowski et al.43 incorporated the GA+ into the MAPbI3 perovskite film, forming the stable 3D perovskite structure. The SEM results showed that when x < 25%, the grain size was increased without unwanted phase segregation, whereas higher amounts of GA+ (x > 25%) tended to form the 1D GAPbI3 perovskite (Fig. 4a and b). Besides the improved film quality, the employed GA+ offered more H bonds within the inorganic framework to stabilize the 3D perovskite structure, resulting in improved efficiency and resistance to thermal and light corrosion.
![]() | ||
| Fig. 4 (a) Normalized XRD patterns of perovskite films with different amounts of GA. (b) Top-view SEM images of MA1−xGAxPbI3 with x = 0, 0.14, 0.30, and pure GAPbI3. Reproduced with permission.43 Copyright 2017, Nature Publishing Group. (c) XRD data for perovskite films with different GAI doping. Reproduced with permission.44 Copyright 2022, Wiley-VCH. (d) Schematic illustration of the halide ion distribution in the GAxCs1−xPbIBr2 lattice in response to a low-power laser (1.27 mW). (e) Normalized PCEs of CsPbIBr2- and GA0.1Cs0.9PbIBr2-based PSCs under continuous illumination for 8 h with the AM 1.5G solar spectrum in the ambient environment (∼55% relative humidity) without encapsulation. Reproduced with permission.45 Copyright 2022, Wiley-VCH. | ||
Moreover, GA+ has a significant effect on the inhibition of ion migration, which is responsible for the hysteresis and poor stability of the device.48,49 Ferdani et al.50 reported that MAPbI3 with 5% GA+ substitutions had higher activation energy (Ea = 0.78 eV) for iodide ion conduction than that of pristine MAPbI3 (Ea = 0.44 eV). Furthermore, Minussit et al.49 found that GA+ could slow down the relaxation process of space charge, which was two orders of magnitude longer, resulting in retarded ion migration. Meanwhile, incorporating GA+ has an impact on the low-temperature phase transitions of MAPbI3. When MA was substituted by 2 mol% GA+, the β (tetragonal) to α (cubic) phase transition happened under a reduced temperature of 278 K.
Wang et al.44 added GAI in the triple halide-mixed CsPbTh (Th = either I, Br, or Cl). The GA composition guaranteed the stabilization of the crystal lattice, achieving enhanced thermal stability (Fig. 4c). After that, the optimal device was obtained by PbAc2 introduction as a nucleating additive, accompanied with GCA spin-coated for interface passivation. Eventually, it delivered a PCE of 19.37% with a high FF of 82.7% and VOC of 1.17 V, and the operational stability was also dramatically improved. A more profound mechanism of the I− ion migration suppression was proposed, suggesting that the strengthened hydrogen bonds between the GA+ and the [PbI6]4− octahedron were responsible for reducing the local lattice distortion, which was the origin of ion migration45 (Fig. 4d). The retardation of light-induced ion migration via GA+ incorporation could retain over 80% under 8 h continuous illumination, showing superior light stability to the pristine CsPbIBr2 (Fig. 4e).
Hysteresis has raised the concern of accurate performance measurement in practice, while its origin remains an open question. GA+ participation could inhibit hysteresis, with diverse interpretations of the mechanisms. Wang et al.53 added GAI in a FA/Cs-based perovskite to regulate the energy alignment and extend the PL lifetime. Consequently, the low content of GA+ in the FA/Cs-based perovskite induced a slight lattice expansion, but could reduce interfacial charge accumulation, thus removing the J–V hysteresis while improving PCE (Fig. 5a and b). On the other hand, Li et al.54 explored the MA0.8FA0.1GA0.1PbI3 system, and concluded that the GA+ incorporation into the perovskite structure could build the potential barrier of iodide ion migration and improve the perovskite crystalline, achieving the best PCE of over 22% and 19.18% for an active area of 14 cm2 (Fig. 5d and e).
![]() | ||
| Fig. 5 Dark J–V curves in logarithm scaled current representation at different scan rates for (a) the control device and (b) the device with 2% GAI. Reproduced with permission.53 Copyright 2019, Wiley-VCH. Temperature-dependent ion conductivity of the (d) MAPbI3 (control) and (e) MA0.9GA0.1PbI3 (MGPbI3). Reproduced with permission.54 Copyright 2022, Wiley-VCH. (c) J–V curves of quadruple-cation CsMAFAGA-based device at 85 °C. (f) Schematics of GA+ dynamic incorporation in vacuum-deposited, quadruple cation CsMAFAGA perovskite films. Reproduced with permission.55 Copyright 2022, ACS Publishing Group. | ||
Due to the limited number of thermal sources and complexity of the Cs/MA/FA triple-cation-based perovskite, it is difficult to insert the GA+ into the lattice to form the CsMAFAGA quadruple-cation perovskite via solution process and conventional vacuum process. However, Susic et al.55 realized the uniform alloying by sublimation of FA and GA in the same thermal source during vacuum deposition, given that the sublimation properties of FAI and GAI were similar. The GA+ initially accumulated at the grain boundaries. However, upon subsequent annealing, it progressively penetrated the crystal and induced a lower-symmetry phase perovskite structural change to relieve the strain stress. Ultimately, stabilization of the cubic phase with a larger unit cell volume was observed (Fig. 5c and f). The CsMAFAGA-based device could maintain the perovskite structure, and no parasitic phase (PbI2, δ-CsPbI3, and δ-FAPbI3) formed at 85 °C after 37 days, demonstrating no degradation or segregation after stabilization by GA cations. Simultaneously, it kept over 75% original PCE at the 85 °C for about a month.
Although the GA+ in A-site compositional engineering plays an important role in the double-, triple- and even multiple-cation-based perovskite system for the ion migration inhibition, perovskite structure stabilization, charge defects passivation, and band structure change, it inevitably increases the complexity of the perovskite system for processing. In addition, introducing a high concentration of GA+ (more than 20%) could facilitate low-dimensional perovskite impurities, such as 1D GAPbI3 and 2D GA2PbI4 rather than uniform alloying in 3D structure, which would be imperfections in the interface and grain boundaries. The limited doping concentration becomes the bottleneck in composition engineering, so other guanidinium-involving approaches for perovskite modification would be the alternative for performance enhancement.
. Here, A′ is structurally featured with bulky or long-chain monovalent organic cations for the 2D RP-perovskite, while divalent interlayer organic cations are suitable for the 2D DJ-perovskite. Unfortunately, the expanded interspaces filled with such bulky organic cations result in increased exciton binding energy and charge transfer barriers between the inorganic parts. The enlarged bandgap also brings the loss of optical absorption. Theoretically, it is hard to achieve a satisfying photovoltaic performance in the case of the normal 2D perovskite as a photoactive layer. Thus, partially replacing GA with a smaller molecule, forming the 2D ACI perovskite with a narrower band gap, has drawn considerable attention. 2D ACI follows the formula (A′)(A)nBnX3n+1, where the slightly larger A′ cation (e.g., GA+) and smaller A cation (e.g., MA+) form an alternating arrangement in the interlayer space. So far, Cs, MA, FA, and otheres are typically regarded as the substitutable spacer to build 2D ACI perovskites together with GA cations. The crystallographic data of the ACI perovskites reported are presented in Table 3.
| GAMAPbI4 | GAMA2Pb2I7 | GAMA3Pb3I10 | GACsPbI4 | GACsPbBr4 | GACs2Pb2Br7 | GAFAPbI4 | |
|---|---|---|---|---|---|---|---|
| Formula wt | 807.0 | 1426.9 | 2046.9 | 907.79 | 719.83 | 1299.66 | 819.95 |
| Temperature (K) | 293 | 300 | 298 | ||||
| Color | Yellow | Dark red | Black | Red | Yellow | Yellow | Brown |
| Morphology | Plate | None | None | None | Platelet | ||
| Crystal system | Orthorhombic | Monoclinic | |||||
| Space group | Imma | Bmm2 | Imma | Pnnm | Imma | Cmmm | C2/m |
| Volume (Å3) | 1507.7(2) | 2508.2(3) | 3473.8(5) | 2897.3(5) | 1226.2(2) | 2003.2(7) | 4678.7(7) |
| Unit cell dimensions (Å) | a = 6.4286(6) | a = 6.3729(4) | a = 6.3520(7) | a = 12.7426(13) | a = 6.0613(6) | a = 6.0029(12) | a = 26.948(2) |
| b = 12.4577(9) | b = 12.5435(8) | b = 12.4378(9) | b = 8.5776(19) | b = 11.5868(12) | b = 28.773(6) | b = 12.8189(11) | |
| c = 18.8258(13) | c =31.376(2) | c = 43.970(3) | c = 12.2390(13) | c = 17.4591(18) | c = 11.598(2) | c = 14.4080(12) | |
| Z | 4 | 8 | 4 | 4 | 12 | ||
| ρ calc (g cm−3) | 3.5551 | 3.7788 | 3.9138 | 4.162 | 3.899 | 4.309 | 3.492 |
| μ (mm−1) | 19.349 | 22.017 | 23.395 | 22.607 | 29.671 | 34.311 | 18.710 |
| F(000) | 1392 | 2424 | 3464 | 3056 | 1240 | 2208 | 4224 |
| Crystal size (mm3) | 0.1111 × 0.0225 × 0.0167 | 0.0511 × 0.179× 0.0100 | 0.0733 × 0.0372× 0.0193 | 0.3 × 0.07 × 0.02 | 0.2 × 0.08 × 0.02 | 0.46 × 0.08 × 0.04 | 0.32 × 0.06 × 0.04 |
| θ range (deg) | 3.27–29.23 | 2.6–29.27 | 2.78–29.29 | 3.876–63.122 | 4.218–62.904 | 3.512–62.718 | 3.006–52.738 |
| Index ranges | −8 ≤ h ≤8 | −8 ≤ h ≤8 | −8 ≤ h ≤ 8 | −18 ≤ h ≤ 17 | −8 ≤ h ≤ 8 | −8 ≤ h ≤ 8 | −33 ≤ h ≤ 33 |
| −17 ≤ k ≤ 16 | −17 ≤ k ≤ 17 | −17 ≤ k ≤ 16 | −26≤ k ≤ 26 | −17 ≤ k ≤ 16 | −41 ≤ k ≤ 41 | −16 ≤ k ≤ 16 | |
| −25 ≤ l ≤ 25 | −42 ≤ l ≤ 42 | −57 ≤ l ≤ 60 | −17 ≤ l ≤ 17 | −25 ≤ l ≤ 24 | −16 ≤ l ≤ 16 | −18 ≤ l ≤ 18 | |
| Reflections collected | 7172 | 11 855 |
12 119 |
33 586 |
7142 | 11 917 |
21 674 |
| Independent reflections | 784 [Rint = 0.1277] | 2595 [Rint = 0.0563] | 1810 [Rint = 0.1261] | 4767 [Rint = 0.0637, Rσ = 0.0432] | 1094 [Rint = 0.0406, Rσ = 0.0295] | 1814 [Rint = 0.0473, Rσ = 0.0370] | 5009 [Rint = 0.0386, Rσ = 0.0319] |
| Completeness to θ | 99% | 99% | 99% | None | None | None | None |
| Data/restraints/parameters | 784/3/2 | 2595/9/82 | 1810/5/59 | 4767/18/109 | 1094/0/40 | 1814/0/58 | 5009/0/144 |
| Goodness-of-fit on F2 | 3.18 | 2.19 | 2.80 | 1.004 | 1.041 | 1.035 | 1.017 |
| Final R indexes [I ≥ 2σ(I)] | R 1 = 0.0575 | R 1 = 0.0555 | R 1 = 0.0597 | R 1 = 0.0389 | R 1 = 0.0223 | R 1 = 0.0321 | R 1 = 0.0409 |
| wR2 = 0.1356 | wR2 = 0.1099 | wR2 = 0.1254 | wR2 = 0.0722 | wR2= 0.0519 | wR2 = 0.0772 | wR2 = 0.0945 | |
| Final R indexes [all data] | R 1 = 0.0677 | R 1 = 0.0898 | R 1 = 0.0863 | R 1 = 0.0835, | R 1 = 0.0261 | R 1 = 0.0407 | R 1 = 0.0552 |
| wR2 = 0.1361 | wR2 = 0.1126 | wR2 = 0.1270 | wR2 = 0.0888 | wR2 = 0.0535 | wR2 = 0.0823 | wR2 = 0.1016 | |
| Largest diff. peak/hole (e Å−3) | 1.84/−2.31 | 5.69/−1.65 | 3.32/−3.20 | 1.36/−1.55 | 0.84/−0.88 | 2.21/−1.52 | 1.87/−1.27 |
![]() | ||
| Fig. 6 Crystal structures of (a) (MA)n+1PbnI3n+1 and (GA)(MA)nPbnI3n+1 (right), n = 1, 2 and 3. Reproduced with permission.56 Copyright 2017, ACS Publishing Group. (b) (GA)(Cs)nPbnX3n+1 (n = 1 and 2; X = I and Br) Reproduced with permission.57 Copyright 2017, ACS Publishing Group; (c) GAFAPbI4. Reproduced with permission.34 Copyright 2018, ACS Publishing Group. | ||
In summary, the critical feather of 2D ACI is the alternating-cation interlayer configuration, which is distinguished from the DJ and RP phases, leading to chemical differences.
(I) The octahedral arrangement in the structures. Due to the bulky or long-chain monovalent organic cation, the RP phase has half octahedral displacement between the adjacent 2D perovskite slabs. For DJ phase, divalent cations are vertically connected with 2D perovskite slabs. In contrast, an alternating cation arrangement is found in the ACI phase, which induces octahedral rotation in the crystal structure.
(II) The interlayer distance (the distance between the terminal I ions and the connecting organic ammonium). The RP phase has the largest interlayer distance58 because of the bilayer structure of the monovalent cation pairs. Compared with the RP phase, the DJ phase is shorter on account of the monolayer structure of its divalent organic cation.59 Regarding the ACI phase, the small cations (e.g., MA and GA) form the tightly interacted interlayer, resulting in the lowest interlayer distance.56
(III) The interlayer interaction. Both the van der Waals gap and hydrogen bonding are present in the RP phase. Nevertheless, the strong interaction with adjacent 2D perovskite slabs weakens the van der Waals gap. As in the ACI phase, the van der Waals gap and hydrogen bonding are retained, but the hydrogen bonding is the main interaction. Therefore, the elimination or reduction of the van der Waals gaps may increase the structural stability of 2D.60
| Device structure | ACI formula | Strategy | PCE (%) | Stability | Ref. |
|---|---|---|---|---|---|
| PVSK: perovskite. RH: relative humidity. RT: room temperature. MPP: maximum power point. | |||||
| ITO/PEDOT: PSS/PVSK/PC61BM/Al | (GA)MA3Pb3I10 | None | 7.26 | None | 56 |
| ITO/PEDOT:PSS/PVSK/PC61BM/BCP/Ag | (GA)MA3Pb3I10 | Solvent engineering: DMSO | 14.7 | T 88 = 240 d in the air with 30–40% RH at 25 °C | 65 |
| FTO/TiO2/PVSK/spiroOMeTAD/Au | (GA)MA3Pb3I10 | Additive engineering: MASCN | 10.12 | T 80 = 30 d in air with 20% RH at 30 °C; T50 = 3200 min in N2 | 66 |
| ITO/PEDOT:PSS/PVSK/PCBM/LiF/Al | (GA)MA3Pb3I10 | None | 16.65 | T 80 = 2400 h in N2 | 67 |
| FTO/TiO2/PVSK/spiroOMeTAD/Au | (GA)MA3Pb3I10 | Additive engineering: MACl | 18.48 | T 95 = 131 d in the air with 30–40% RH at 25 °C; T79 = 65 h at 80 °C; T57 = 200 h at MPP in the air with 55% RH at 60 °C | 64 |
| ITO/PEDOT:PSS/PVSK/C60/BCP/Ag | (GA)MA4Pb4I13 | Solvent engineering: DMSO | 12.8 | None | 62 |
| ITO/PEDOT:PSS/PVSK/PC61BM/BCP/Ag | (GA)MA4Pb4I13 | Post-treatment: GASCN and MACl | 15.27 | T 75 = 480 h in air with 50% RH at 25 °C; T70 = 300 h in air with 50 ± 5% RH at 25 °C | 63 |
| FTO/TiO2/PVSK/spiroOMeTAD/Au | (GA)MA5Pb5I16 | Solvent engineering: DMSO | 19.18 | T 95 = ∼123 d in the air with 30 ± 10% RH at 25 ± 5% °C; T80 = ∼60 h in an inert atmosphere at 80 °C; T80 = ∼240 h at 60%–70 RH | 61 |
| FTO/TiO2/PVSK/spiroOMeTAD/Au | (GA)MA5Pb5I16 | Post-treatment: amphoteric imidazolium iodide | 22.26 | T 93 = ∼1200 h in air with 25% RH at 25 °C; T55 = ∼60 h in an inert atmosphere at 85 °C | 26 |
Zhang et al.65 optimized the crystallization process of the 2D ACI perovskite by solvent engineering–adopting a mixed solvent of dimethylformamide (DMF) and dimethyl sulfoxide (DMSO) at the volume ratio of 1
:
10. With the assistance of in situ grazing-incidence X-ray scattering (GIWAXS) characterization, it was found that 2D GA2PbI4 was the crucial intermediate phase, which provided the scaffold for the ACI perovskite growth, and dominated the distribution of GAMA3Pb3I10 (n = 3) with less of other n-value-phases and 1D GAPbI3 impurities. Additionally, the hot-casting method was used instead of traditional anti-solvent processing to effectively reduce the defect density. Consequently, the high-quality ACI perovskite film was deposited, yielding a greatly improved PCE of 14.7%.
Additive engineering also helps improve the crystallinity of the 2D GAMA3PbI10 (n = 3) perovskite films. Zhang et al.66 introduced MASCN (0.3 molar ratio relative to PbI2) in the perovskite precursor. Compared to the pristine 2D ACI perovskite film, the grain size in the MASCN-doped 2D ACI films was clearly enlarged, which was accompanied by high crystallinity. Eventually, the PCE of the device with the MASCN additive was boosted from 9.05% (non-doped) to 10.12%. Suitable additives also controlled the distribution of the targeted ACI quantum wells (QWs, i.e., the n number).67 Luo et al.64 effectively promoted the charge transfer of 2D ACI (n = 3) by adding MACl in the precursor solution. The optimized thickness distribution of the QWs decreased the carrier-populating time to 0.8 ps, which enabled highly efficient transport/extraction in PSCs (Fig. 7a and b). Meanwhile, the morphology of the 2D ACI became smoother with the increased grain size. After optimization, the MACl-treated device delivered a record PCE of 18.48% (Fig. 7c).
![]() | ||
| Fig. 7 The morphology of the 2D ACI film: (a) the top-view SEM images of the ACI films without/with MACl treatment (top); atomic force microscopy (AFM) images of the ACI films without/with MACl treatment. (b) Transient absorption (TA) spectra at different delay times for the ACI films without/with MACl treatment. (c) J–V curves under the reverse voltage scan of the best device ACI films without/with MACl treatment. Reproduced with permission.64 Copyright 2019, Wiley-VCH. (d) Trap density (ntrap) and mobility (μ) of the GA(MA)nPbnI3n+1 (n = 5) perovskite films with ImI treatment. (e) J–V curves under reverse- and forward-scan of PSCs with ImI treatment. (f) Long-term stability of PSCs without/with 2 mg mL−1 ImI treatment at RT and 25% RH for 50 days. Reproduced with permission.26 Copyright 2022, ACS Publishing Group. | ||
:
v = 1
:
10) to increase the GAMA5Pb5I16 (n = 5) ACI formation. The stable intermediate phases can slow down the interposition of ions, which in turn retarded the crystallization of the ACI phases. Herein, the DMSO co-solvent led to a new intermediate solvate phase of MA2Pb3I8·2DMSO, whose solvation was immediately deprived along with the transition into ACI phases. The following thermal treatment completed the decomposition of intermediates and ACI crystallization with good film uniformity and surface coverage, leading to reduced charge recombination loss. Eventually, the unencapsulated PSCs realized a PCE of 19.8%.
Besides solvent engineering, post-treatment benefits the quasi-2D ACI phase arrangement. Zhang et al.63 proposed a sequential post-treatment method of modifying the GAMA4Pb4I13 ACI film with GASCN and MACl. GASCN could improve the quality of the perovskite films through the coordination bonding between SCN− and Pb2+, resulting in dense and well-aligned crystal grains. The sequential MACl treatment can suppress the charge defects on the perovskite surface and grain boundaries, which increased the electron/hole mobility. The PSCs based on sequential post-treatment demonstrated the best PCE of 15.27%, which was much higher than that of the pristine device (8.26%). Park et al.26 employed a post-treatment reagent of amphoteric imidazolium iodide (ImI) on the top of the ACI n = 5 perovskite film. ImI post-treatment effectively reduced the vacancies/excess defects of the perovskite (Fig. 7d). Consequently, lower trap density and higher charge mobility were achieved, yielding a record-high PCE of 22.26% among the ACI-based 2D PSCs (Fig. 7e and f).
No research studies have been focused on GA-based ACI PSCs with further increased n, perhaps because of their negligible structural and optical differences from the GA-doped 3D species. Despite additive and solvent engineering optimization, the parasitic 1D GAPbI3 or 2D GA2PbI4 within the 3D perovskite matrix during ACI crystallization could not be eliminated. Their insufficiency in light capture, the role of the carrier trap center, as well as the resulting nonuniformity in the film morphologies are the critical barrier to achieving highly efficient and stable ACI-based devices. More technologies are needed to obtain the highly pure 2D ACI phases.
| Additive | Perovskite component | Device configuration | PCE (%) | Stability | Ref. |
|---|---|---|---|---|---|
| GABr | FAxMAyCszPbI3−yBry | FTO/SnO2/PVSK/spiro-OMeTAD/Au | 20.29 | T 90 > 60 d in the air at 25 °C with 25% RH; T70 > 60 d in the air at 80 °C with 40% RH; T75 > 5 h in the air under 1 sun illumination | 72 |
| FAPbBr3 | ITO/SnO2/PVSK/spiro-OMeTAD/Au | 8.92 | T 90 > 1024 h in the air at RT with 50% RH | 75 | |
| FA0.7MA0.3Pb0.7Sn0.3I3 | ITO/EMICPEDOT:PSS/PVSK/S-acetylthiocholine chloride/C60/BCP/Ag | 20.63 | T 85 > 1000 h in the air at RT with 50–60% RH; T90 > 24 h in the air at 80 °C with 50–60% RH | 76 | |
| GA0.05Cs0.15FA0.8Pb(I0.85Br0.15)3 | FTO/SnO2/PVSK/spiro-MeOTAD/Au | 21.33 | T 80 > 8 h in N2 at MPP under 1 sun illumination | 77 | |
| GASCN | (FASnI3)0.6(MAPbI3)0.4 | ITO/PEDOT:PSS/PVSK/C60/BCP/Ag | 20.5 | T 80 = 220 h in N2 at MPP | 78 |
| (FASnI3)0.6(MAPbI3)0.4 | ITO/PEDOT:PSS/PVSK/C60/BCP/Ag | 22.1 Tandem: 25.4 | T 82 = 1830 h in N2 at 30–35 °C under 0.8 light soaking; T90 = 500 h in air | 79 | |
| PEA0.1FA0.9SnI3 | ITO/PEDOT:PSS/PVSK/PC61BM/poly-ethylenimine/Ag | 10.06 | T 91 = 720 h in N2 at 25 °C; T45 = 15 h in the air at 25 °C with 25% RH; T55 = 2 h in air at 25 °C with 60% RH | 80 | |
| FASnI3 | ITO/NiOx/PVSK/ICBA/BCP/Ag | 13.79 | T 90 = 1200 h in N2 | 81 | |
| GACl | CsMAFA/Pb/(I0.87Br0.13)3 | FTO/TiO2/PVSK/spiro-OMeTAD/Au | 17.9 | None | 82 |
| MAPbI3 | FTO/TiO2/ZrO2/PVSK/carbon layer | 14.35 | None | 73 | |
| GABF4 | MAPbI3 | ITO/NiOx/PVSK/PC61BM/Ag | 20.62 | T 81 = 45 d in the air at RT with 50–60% RH | 83 |
| GuaSO3NH2 | Cs0.05(MA0.12FA0.88)0.95Pb(I0.88Br0.12)3 | FTO/TiO2/PVSK/spiro-MeOTAD/Au | 21.66 | T 90 = 750 h in the air at RT with 25% RH | 84 |
| GAPbI2Br | CsPbI2Br | ITO/ZnO/PVSK/spiro-MeOTAD/Au | 14.34 | T 94 = 1000 h in N2 | 85 |
NH) in the GA+ cation, these two differently charged groups could be used to eliminate both positively charged (uncoordinated Pb2+) and negatively charged (I vacancy) defects to reduce nonradiative recombination (Fig. 8a). More noticeably, the grain size in the perovskite film with the GABr additive was increased from 600 nm to about 1 μm (Fig. 8b). The high-quality crystallization contributed to the reduced surface charge defects, and the mechanisms of morphology modulation were attributed to the incorporation of bulky GA+ into the perovskite lattice. Last, the band alignment between the perovskite and HTL became better. In contrast to the pristine device, the 15 wt.% GABr-doped devices showed a higher built-in potential (Vbi), which generated a strong driving force for intrinsic charge separation. These multiple factors together corresponded to the minimized VOC loss. The FAPbBr3-based wide band-gap device (Eg ∼ 2.3 eV) with GABr reached the PCE of 8.92% with a high VOC of 1.639 V.
![]() | ||
| Fig. 8 (a) The schematic mechanism of GABr modulation on FAPbBr3-based PSCs. (b) Top-view SEM images of the control (left) and 15% GABr film (right) annealed at 110 °C for 15 min. Reproduced with permission.75 Copyright 2022, Elsevier. (c) 2D grazing-incidence X-ray diffraction (GIXRD) patterns of the perovskite films without (left) and with 12% GABr (right), respectively. (d) The SEM top images of the perovskite films without and with 12% GABr, respectively (the scale bar is 2 μm). Reproduced with permission.76 Copyright 2020, Elsevier. | ||
Despite the great promise of the Sn-Pb binary perovskite for developing low band-gap PSCs (Eg = 1.2–1.3 eV), it suffers from the enormous challenge of Sn2+ oxidation, leaving Sn vacancies. The defect elimination caused by Sn2+ oxidation could also be achieved by GABr addition, as reported by Zhou et al.76 In 2D GIXRD, the strongly dispersed secondary blots and bright rings in the perovskite film with 12% GABr additive demonstrated that the crystalline realms retained good orientation (Fig. 8d). Apart from the improved crystal quality and purity, GA+ was beneficial to the structural stability and restriction of the ion migration. Moreover, Br anions in the additive could act as a substituent on the X site to increase the tolerance factor and bandgap of the perovskites compared to those of the iodine-based ones.77,85 As a result, the device with 12% GABr obtained a PCE of 20.63% with negligible J–V hysteresis and superior environment and thermal stability.
![]() | ||
| Fig. 9 (a) Time-resolved photoluminescence of the perovskite film with GASCN. J–V curves of (b) the champion single PSC; the inset shows the SPO efficiency near the maximum power point (MPP), and (c) the 2-T tandem cell. The inset shows the EQE spectra of the top and bottom subcells, respectively. Reproduced with permission.78 Copyright 2019, AAAS. (d) Top-view SEM photograph of the perovskite film with PEAI and GASCN additives. J–V curves of (e) the single champion device with PEAI and GASCN additives along with the SPO efficiency, and (f) all-perovskite tandem PSCs along with the SPO. Reproduced with permission.79 Copyright 2022, Nature Publishing Group. | ||
Recently, PEASCN and GASCN additives have been employed in the Sn–Pb-based perovskite precursor.79 The binary cation additives helped form the quasi-2D (PEA)2GAPb2I7 deposited at the surfaces and grain boundaries to suppress more charge traps than that caused by Sn2+ oxidation (Fig. 9d). The 2D additive engineering endowed the Sn–Pb-based single-junction cells and two-terminal tandems with efficiencies of 22.1% and 25.5%, respectively (Fig. 9e and f).
Given that the O and N atoms of H2N–SO3(SM) could bond with uncoordinated Pb2+ because of the lone electron pairs in O and N, Liu et al.84 introduced GASM into the [Cs0.05(MA0.12FA0.88)0.95Pb(I0.88Br0.12)3] perovskite precursor to passivate charge traps at the surface and grain boundary effectively. With the dual-function synergy effect of GA+ and SM, the grain size of the perovskite film was enhanced, and the PL intensity of perovskite film became stronger. The optimal device yielded the best PCE of 21.66%, and kept 90% of its original efficiency after 800 h storage in ambient conditions.
Similar to GA-based dopants in A-site composition engineering, the content of GA-based additives is usually less than 20% (molar ratio vs. Pb species). A high concentration of GA-based additives would cause the formation of the impure phases, and the disruption of high-quality crystallization. Moreover, the anions could influence the GA incorporation in the perovskite lattice, which expands the lattice constant. A detailed mechanism of the synergistic effect of the GA+ cation and anion on the perovskite is still unclear, and requires further research.
| Interface engineering layer | 3D perovskite layer | Device structure | PCE (%) | Stability | Ref. |
|---|---|---|---|---|---|
| PVSK: perovskite; RH: relative humidity; Tx is when PCE degrades to x% of the initial value. | |||||
| GACl | MAPbI3 | ITO/PTAA/PVSK/PC61BM/LiF/Al | 18.65 | None | 96 |
| GACl | MAPbI2.7Cl0.3 | FTO/TiO2/PVSK/spiroOMeTAD/Ag | 18.46 | T 95 > 22 d in N2 | 91 |
| GABr | CsPbI3 | FTO/TiO2/PVSK/PTAA/Au | 18.02 | T 92 > 30 d at 20–30% RH | 107 |
| GABr | FA0.8Cs0.2Pb(I0.7Br0.3)3 | ITO/PTAA/PVSK/C60/BCP/Ag | 18.19 | T 72 > 70 h at RT and > 50% RH in air | 74 |
| GABr | (FA0.95PbI2.95)0.85(MAPbBr3)0.15 | ITO/PTAA/PVSK/PC61BM/C60/BCP/Cu | 21.51 | T 95 = 500 h at 85 °C in N2 | 94 |
| GAI | MAPbI3 | ITO/SnO2/PVSK/spiroOMeTAD/Au | 18.54 | T 85 = 30 d in dry condition; T50 = 15 d in 60 ± 10% RH. | 52 |
| GAI | FAPbI3 | FTO/TiO2/PVSK/spiroOMeTAD/MgF2/Au | 24.36 | None | 95 |
| GASCN | MAPbI3 | FTO/TiO2/PVSK/spiroOMeTAD/Au | 21.26 | T 80 = 1000 h in the air at 20% RH | 97 |
| 1D GAPbI3 | (FA0.9Cs0.07MA0.03Pb(I0.92Br0.08)3 | FTO/TiO2/PVSK/spiroOMeTAD/Au | 21 | T 80 = 550 h in N2 at MPP under RT | |
| 2D GA2PbI4 | CsPbI2Br | FTO/SnO2/PVSK/spiroOMeTAD/Au | 15.6 | T 88 = 60 d in the air at 25 °C with 25% RH; T76 = 30 d in the air at 80 °C with 25% RH; T90 = 5 h in the air at MPP | 103 |
| 2D GA2PbI4 | n–i–p: Cs0.05FA0.85MA0.1PbI3 | n–i–p: ITO/SnO2/PVKS/spiroOMeTAD/Au | n–i–p: 20.1 | T 90 > 30 d in N2 | 99 |
| p–i–n: Cs0.3FA0.6DMA0.1PbI2.4Br0.6 | p–i–n: ITO/NiOx/PVKS/PC61BM/BCP/Ag | p–i–n: 22.9 | |||
| 2D GA2PbI4 | FA1−xMAxPbI3−xBrx | FTO/SnO2/PVKS/spiroOMeTAD/Au | 22.53 | T 90 > 30 d in the air with 25% RH; T80 > 30 d in the air with 75 ± 5% RH; T80 > 30 d in N2 under light illumination | 98 |
Luo et al.94 spin-coated the multi-functional GABr on the surface of (FA0.95PbI2.95)0.85(MAPbBr3)0.15 perovskite film in the inverted PSCs. With the GABr passivation, the bright spots on the surface and boundaries disappeared, indicating that the PbX2 impurity was removed (Fig. 10a and b). Moreover, the GABr additive made the Fermi levels of the perovskite films shift toward the vacuum (400 mV), which was higher than the valence band upshift (200 mV), indicating more n-type nature after GABr treatment (Fig. 10c). The more n-type perovskite film could result in a larger fraction of occupied versus vacant traps, and a reduction in the rate of trap-assisted recombination. Additionally, Br enrichment was observed near the interface after the addition, leading to a wider bandgap (80 meV) in this region. It is energetically favorable for one or both charge carriers to reside predominantly within the bulk of the film, away from the surface defects. The device with GABr passivation delivered a PCE of 21.51% with a record VOC of 1.21 V, resulting from the suppression of trap-induced recombination (Fig. 10d).
![]() | ||
| Fig. 10 (a) Top view SEM images of the perovskite film with/without GABr passivation (upper panel). XRD patterns (lower panel) of the perovskite film with/without GABr treatment (α: black perovskite phase; β: non-perovskite phase). (b) Schematic energy-level diagrams of the control and GABr-treated films. (c) J–V curves under the reverse scan of devices with/without GABr passivation. Reproduced with permission.94 Copyright 2018, AAAS. (d) Top view SEM images of the perovskite film with/without binary GAI/tBPMAI treatment. (e) The time-of-flight secondary-ion mass spectrometry (ToF-SIMS) of the binary GAI/tBPMAI-treated perovskite film. (f) The forward/reverse scan J–V curves of the best devices with/without the binary GAI/tBPMAI treatment. Reproduced with permission.95 Copyright 2021, Royal Society of Chemistry. | ||
Sometimes, GA-based passivation agents were simultaneously post-treated along with other salts for more progress. Zhang et al.95 addressed the complete passivation from the interfaces to bulk in the perovskite film by combining GAI with 4-tert-butyl-phenylmethylammonium iodide (tBPMAI) as the passivation agent. Given the difference between GAI and tBPMAI in ionic radius, GA could diffuse into the perovskite lattice. At the same time, the tBPMAI gathered on the surface of the perovskite, simultaneously modulating defects in perovskites from bulk to interfaces, which were confirmed by ToF-SIMS (Fig. 10f). More interestingly, it showed a positive effect on crystallization modulation after the post-treatment of this composite layer. The perovskite grains were regularly arranged, leading to the compact and smooth morphology of the perovskite (Fig. 10e). Finally, the device with individual GAI or tBPMAI treatment showed a low PCE of 21.79%. In contrast, the device with binary passivation obtained the champion PCE of 24.36% with a lower voltage loss of 0.388 V (Fig. 10g).
Better crystal growth with fewer trap states also can be regulated by the passivator itself and its solvent. Yang et al.97 selected GASCN dissolved in chlorobenzene (CB) anti-solvent to passivate the MAPbI3 surface effectively. The GA cation could stabilize the black phase, anchor the lattice ion, and inhibit their migration due to the hydrogen bond formation. Results from different anti-solvents such as ethyl acetate (EA), i-propanol (IPA), and cyclohexane (CH) were also discussed here. Compared with other anti-solvents, CB with low polarity could cause minor damage to the perovskite grain and slow down the crystallization process. As a synergistic result of GASCN and CB, the grain size was dramatically increased. This binary strategy increased the PCE from 20.14% to 21.26%, and the thermal and moisture stability were enhanced.
The functions of constructing a GA-based interface (defect passivation, energy level alignment, and crystallization regulation) are similar to that of GA-based additives. However, this approach induces two individual layers, different from the doping behavior of additives in the perovskite layer during additive engineering. So, the limits of the GA molar ratio (vs. overall A-site cation content) considering the possible interference on perovskite crystallization are not involved here. However, the limits lie in its low solubility in CB. Generally, IPA is chosen to dissolve the GA-based agents, but as mentioned before, IPA could erode the surface of the perovskite film owing to the relatively high polarity. Searching for a low-polarized anti-solvent that is compatible with GA salts or exploring soluble GA salts in CB should be the future task for managing GA-contained interfaces.
Similar to the dual optimization scheme introduced previously,95 Zhang et al.98 treated the perovskite with GAI and GACl to realize surface and bulk modification. Theoretical calculations proposed that among GAX salts, the Cl doping system exhibited the highest formation energy of iodine vacancy and the longest excited state lifetime. Therefore, GACl was expected to be an appropriate passivator for bulk doping, and the outperformed PCE of GACl (20.07%) was higher than others (17.71% for GABr and 18.01% for GAI). However, the smallest ionic radius of Cl−, showing the lowest activation energy barrier for X ion migration, infers the need to suppress its bulk diffusion to the surface. Co-doping with GAI was suggested to form a stable capping layer owing to its larger activation energy barrier (Fig. 11a and b). After co-doping, GAI was inserted into the 3D perovskite lattice forming the 2D GA2PbI4 layer for surface passivation (Fig. 11c). The incorporation of large GA cations also modulated the lattice tensile strain towards minimizing defect centers. Benefiting from the strategy for reducing nonradiative recombination centers from both bulk and interfacial regions and the energy level alignment after 2D layer insertion, the PCE of the device with a 2D/3D perovskite structure exceeded 22.53%. Moreover, the PCE maintained more than 90% after 30 days at 40% R.H. due to the enhanced hydrophobicity of the 2D layers.
![]() | ||
| Fig. 11 (a) Illumination of the halogen ion interstitial (Cli−/Bri−/Ii−) migration from the perovskite bulk (defined as initial state, IS) to the surface (defined as the final state, FS), along with (d) the energy profile. (b) The activation energy barrier of the halogen ion. (c) J–V curves under the reverse scan of devices with/without 2D perovskite passivation. Reproduced with permission.98 Copyright 2022, Wiley-VCH. (d) The average binding energy of different cations with NiOx surface by DFT. (e) Structural model and corresponding schematics of the crystallization process for films with varying concentrations of GA on NiOx. (f) The forward/reverse scan J–V curves of the champion devices with/without 2D perovskite passivation. Reproduced with permission.99 Copyright 2021, Wiley-VCH. | ||
The 2D perovskite layer not only passivates the 3D perovskite layer, but also shows the great potential of passivating buried side. Chen et al.99 introduced the 2D GA2PbI4 passivation layer in the inverted PSCs between NiOx HTL and the perovskite layer. The GA in 2D GA2PbI4 could connect with NiOx by forming hydrogen bonds, suppressing the migration of oxygen vacancies on the NiOx surface. In addition, the symmetric structure of GA promoted its accumulation onto the NiOx surface to suppress non-radiative recombination (Fig. 11e). It was special when other cations such as n-butylammonium (BA), ethylammonium (EA), and dimethylammonium (DMA) were also added to the precursor. It was found that DMA and EA could not form the 2D perovskite. Furthermore, the 2D BA-based perovskite phase was not present in the NiOx/perovskite interface, but was located in bulk. The authors ascribed the unique phenomenon to the GA cation's higher binding energy and polarity action (Fig. 11d). Finally, the inverted device boosted the PCE up to 22.9% with an increase of 65 mV in VOC (Fig. 11f).
It should not be ignored that in some cases, other low-dimensional phases would form when a GA-contained solution is spin-coated on top of a pre-formed 3D perovskite layer, which also shows a good passivation effect and interfacial band alignment. For instance, Alharbi et al.106 utilized solid-state magic angle spinning (MAS) NMR to verify the coexistence of 1D GAPbI3 and 2D GA2PbI4 on the surface of FA0.93Cs0.07PbI3. Similarly, Chavan et al.104 spin-coated GAI/IPA on the surface of a 3D perovskite film and detected the 1D δ-FAPbI3-like phase. Empirically, the formation of the GA-based 2D perovskite passivation interface is probably accompanied by the complicated mixture of the original GA-based precursors, 3D perovskite constitutes, or other 2D species with different QWs. However, it does not mean that there are positive contributions from every impurity, or in many cases. Moreover, the inefficiency in charge transport across the organic interlayer and the light conversion due to the wider bandgap becomes intrinsic drawbacks of 2D materials. Delicate control for pristine or specific composites in 2D interlayers, as well as the design of innovative 2D materials with better intrinsic properties, should be valued in the future.
(1) Choosing the proper anion in GA salts for improving the performance of PSCs. Nowadays, GA-based salts have been widely used in the field of PSCs, and the anion is another factor for defect passivation, energy alignment, and structure stabilization. For example, Cl− could tune the crystallization process, and Br− could change the charge distribution and bandgap of the perovskite surface. SCN− and BF4− could effectively passivate defects such as undercoordinated Pb ions in the bulk and interface due to lone pair electrons. As a result, a suitable anion is a key aspect to eliminate charge defects.
(2) Developing a more moisture-resistant GA-based perovskite. Due to the lack of long hydrophobic chains or ring structures, the GA-based perovskite film degrades faster than DJ or RP perovskites under moisture and thermal stresses. Therefore, it should be taken into consideration to introduce the new GA-based derivatives for long-term operation stability, but the relevant papers are limited.108,109 In addition, it is an alternative way to take the place of GA by long-chain cations as the interlayer cations phase to endow more hydrophobic properties, thereby increasing moisture repellence and stability for ACI perovskites.110 Few reports are applying monovalent cations (e.g., Cs and FA) to replace MA. Regarding the effect of steric hindrance on the perovskite structure, the cations with appropriate ionic radius are expected to be better candidates to alternate with GA than MA.
(3) Constructing the new 2D GA-based ACI capping layer on top of the 3D perovskite. Although GA2PbI4 was shown to passivate the surface defects of the 3D perovskite and prevent moisture from damaging the perovskite structure, its poor charge mobility inhibits efficient device operation. The 2D ACI layer with large n is expected to be a competitive alternative as the capping layer due to the high carrier mobility. It is noted that when the 2D ACI perovskite is deposited on the surface of 3D perovskite, the formation of multiple-perovskite phases would be induced, including (GA)x(MA)1−xPbI3, GA2PbI4, or other GA-based ACI with different quantum wells (n), which are possible trap sites on the surface. More precise control towards n should be valued in the future.
(4) Further investigating the photophysical behavior in GA-based perovskites, especially ACI systems. The mechanisms of charge carrier dynamics, phase degradation, and ion migration in 2D ACI have not been systematically studied. A comprehensive understanding of these evolutions would provide critical guides for improving photovoltaic performances. Moreover, it would help ACI materials to become competitive candidates for application in other semiconducting fields, including photodetectors, light-emitting diodes, or field-effect transistors.
(5) Application in other photoelectric devices. In addition to PSCs, the GA-based perovskite materials demonstrated excellent performance in the area of the optoelectrical device including light-emitting diodes (LEDs) and photodetectors.111,112 Similar to solar cells, the GA could treat the surface of nanocrystals or colloidal quantum dots113,114 to passivate the defects and provides lattice-stabilizing effect via hydrogen bonds, reducing non-radiative recombination. Further research studies are still needed to define the “insight” mechanism.
| This journal is © the Partner Organisations 2023 |