1 Introduction

Low-dimensional materials are nanocrystals with at least one dimension within the nanoscale range (1–100 nm). For the last few decades, they have been increasingly attracting interest as they exhibit unique properties. They are also referred to as artificial atoms because the density of their electronic states can be widely and easily tuned by adjusting the crystal composition, size, shape, and so on. A typical example of the effect of size on crystal properties is quantum dots (QDs), whose three dimensions are all in the nanoscale range. The bandgaps could be easily tuned by adjusting the dot size [1]. A variety of optoelectronic applications have been explored for low-dimensional perovskites such as photovoltaics, light sensors, light-emitting diodes (LEDs), owing to the strong light–matter interaction [2]. Unlike traditional semiconductor nanomaterials, low-dimensional halide perovskites can be prepared at low cost and by solution-processable techniques. They have demonstrated superior optical, magnetic, dielectric, electrical, and optoelectronic properties [3–5] and have been becoming a hot research topic in new semiconductor materials and optoelectronic devices.

The origin of “Perovskite” is calcium titanate (CaTiO3), which was discovered by Russian mineralogist L.A. Perovski in 1839. After that, perovskite was defined as a class of compounds that have the same crystal structure as CaTiO3, known as the perovskite structure [7]. The general chemical formula for pure perovskite compounds is ABX3 (Fig. 1a) where “A” and “B” are two cations of dissimilar size, and X is an anion that binds to both. Chalcogenide perovskites (AMO3) are formed from divalent AII (Mg2+, Ca2+, Sr2+, Ba2+, Pb2+) and tetravalent MIV (Ti4+, Si4+, Fe4+) elements with O2− as the chalcogenide anion. Halide perovskite (AMX3) represents a large collateral series of the perovskite family, and it is reasonable to divide them roughly into alkali–halide perovskite and organometal halide perovskite. The first category is mainly formed from the monovalent cations (Li+, Na+, K+, Rb+, Cs+; aliphatic or aromatic ammonium) and the divalent MII (Be2+, Mg2+, Ca2+, Sr2+, Ba2+, Zn2+, Ge2+, Sn2+, Pb2+, Fe2+, Co2+, Ni2+) with X representing halogen anions (F−, Cl−, Br−, I−). Oxide-based perovskites have been extensively studied, owing to their superior ferroelectric, magnetic, and superconductive properties [8]. The first halide-based perovskite structure was observed in cesium lead halides (CsPbX3) by Moller in 1958. Their photoconductive properties could be tuned through varying halide components to achieve different spectral responses. The first appearance of the organic cation, methylammonium (MA), in halide perovskites was studied by Weber and Naturforsch in 1978 [9, 10].

Fig. 1
figure 1

Adapted image reproduced from NREL [6]

Perovskite structure and solar cell efficiencies. a Perovskites possess the general crystal structure ABX3. The most prevalent perovskite in optoelectronic devices is MA lead trihalide, for which A = CH3NH3, B = Pb, and X = Cl, Br or I. b Best research cell efficiencies.

Full size image

In the 1990s, Mitzi [11] focused on layered organic–inorganic halide perovskites that featured strong excitonic characteristics and demonstrated applications in thin-film transistors and LEDs. After Miyasaka et al.’s pioneered work in sensitized solar cells, the hybrid perovskite began its debut in photovoltaics in 2006 [12]. In 2009, a power conversion efficiency (PCE) of 3.8% was achieved by replacing Br with I [13]. In 2011, Park and coworkers achieved an efficiency of 6.5% by employing perovskite QDs as the sensitizers [14]. Later, Snaith et al. reported an efficiency of 10.9% utilizing solid electrolyte as a hole transport layer (HTL) [15]. As the study progressed, the superior properties of perovskites unfolded. Ambipolar carrier transport enlightened and promised the intensive employment of planar heterojunction devices [16–18]. This sparked an enormous development in the hybrid lead perovskite solar cells, which obtained a record PCE up to 22.1% in just 6 years using a low-cost solution method (Fig. 1b). In addition to PV applications, low-dimensional perovskite crystals with specific morphologies and ultrasmall size evoked new research interest in optoelectronic applications.

Compared with the hot research into halide perovskite thin-film solar cells, the low-dimensional perovskites hold great potential and just open a small angle. Thus, the present review mainly focuses on low-dimensional halide perovskite to summarize their synthesis methods, optoelectronic applications, and development outlook. By analogy to the human lifetime, the present stage of research into low-dimensional halide perovskites is just in its infancy. The high quantum yield, narrow full-width-at-half-maximum (FWHM), and tunable emission color of low-dimensional perovskite materials make them bright prospects for novel optoelectronic devices. A variety of monochromatic LEDs have been fabricated at room temperature utilizing inorganic cesium lead halide perovskite QDs as the color conversion layer, which have exhibited the best perovskite LED performance so far. Simultaneously, a hybrid perovskite nanowire photodetector (PD) with a photoconductive gain approaching 1014 electrons per photon and a responsivity approaching 6.0 × 105 A W−1 has been developed. Single-crystal lead halide perovskite nanowire lasers exhibited ultralow lasing thresholds (220 nJ cm−2) and high quality factors (Q ∼ 3600) at a charge carrier density of ~1.5 × 1016 cm−3. As hybrid perovskite researchers investigate more deeply, many amazing properties are expected to be uncovered beyond photovoltaics.

In this review, we present a thorough treatment of the recent developments in the fundamental material properties of low-dimensional perovskites (QDs, nanowires, and nanosheets). We mainly focus on the synthesis, unique properties, and notable breakthroughs of perovskites, with optoelectronic applications in photovoltaics, LEDs, PDs, and lasers. Finally, the near-future challenges and the potential directions of this exciting research area are forecasted.

2 Crystal Structure and Tunability of Halide Perovskite

Perovskites, with general formula AMX3, are a well-known class of inorganic materials with widespread deployment in ferro- and piezoelectric, magnetoresistive, semiconducting, and catalysis applications. The rich diversity is attainable owing to the multitude of large bivalent cations occupying the A-site (e.g., Ca2+, Sr2+, Ba2+), the smaller tetravalent cations (e.g., Ti4+, Zr4+) at the M site, and oxygen anions located at the X-site. Halide perovskites are different from their classical chalcogenide counterparts with monovalent moieties introduced at the A-site, divalent metal cations at the M site, and typically halides at the X-site. The AMX3 hybrid perovskite structure is formed by a three-dimensional (3D) network, with A-site cations occupying the cavity between four adjacent corner-sharing MX6 metal halide octahedra (e.g., M = Pb2+, Sn2+, Ge2+, Cu2+, Eu2+, Co2+ and X = Cl−, Br−, I−). The probability of forming the perovskite structure can be estimated using the Goldschmidt tolerance factor (t) and the octahedral factor (µ) [12, 19, 20]. Here, t is based on the ionic radii (r) of the A, M, and X, constituenting in \(t = \left( {r_{A} + r_{X} } \right)/\sqrt 2 \cdot \left( {r_{M} + r_{X} } \right)\), and µ is defined as r M/r X . According to the tolerance factor, only the incorporation of small cations results in perovskite formation (t ~ 1) because it is empirically found that cubic perovskites can form from 0.80 < t < 0.90 and 0.40 < µ < 0.90 [12, 20–22]. This implies that the large ionic radii of Pb (1.19 Å) and the halides (e.g., iodide 2.20 Å) limit the ionic radius of the monovalent A cation to 2.9 Å. Therefore, only two or three C–C or C–N bonds or inorganic cations such as Cs+ (1.88 Å) are expected to fit in the 3D hybrid perovskite structure [22].

One of the major advantages over traditional inorganic oxide is the low energy barrier to halide perovskite formation. A crystalline phase can be readily obtained by merely mixing and grinding the precursor salts at room temperature. Although this method suffers from a lack of precise experimental control, it exemplifies the ease with which the cations can diffuse into the inorganic framework. Typically, halide perovskites are synthesized via wet-chemistry routes, allowing mixing at a molecular level, and resulting in materials with a pure phase. By carefully controlling the reaction conditions (e.g., temperature, solvent, ligands), halide perovskites of various morphologies (0D to 3D) and sizes (ranging over six orders of magnitude) can be prepared. For example, CH3NH3PbBr3 single crystals with a size of 5 × 5 × 2 mm3 are obtained within a growth period of several hours by exploiting its lower solubility in solvents at elevated temperatures [23]. CH3NH3PbBr3 nanoparticles (NPs) [24–27], layered sheets [28, 29], and nanowires (NWs) [30, 31] could be prepared via tuning the synthesis strategy.

These examples briefly demonstrated the robust synthesis processes with which a wide scale of nanostructured perovskites can be synthesized. In recent years, nanomaterials have become more and more interesting, as the novel physical properties are only observed at the nanoscale range, in contrast to their larger-scale counterparts. Fine control over synthesis conditions (precursor concentration, reaction temperature, choice of ligands, etc.) could introduce new physical properties, such as quantum size effects [32] or anisotropic growth [33], to fulfill future application requirements.

3 Synthesis and Fundamental Properties of Low-Dimensional Halide Perovskites

3.1 Quantum Dots

QDs, combining unique optical and electrical properties and solution-processed advantages, have been studied intensively for decades. Here, we will cover the evolution of halide perovskite QDs and devices for light-emission applications. We will disclose the physical and chemical characteristics and analyze the rich diversity in composition and structure.

To date, most reports of hybrid perovskite NPs have utilized the distinct strategy of ligand-assisted reprecipitation (LARP) method. This method could fabricate high-luminescence and color-tunable colloidal CH3NH3PbI3 QDs with an absolute quantum yield (QY) up to 70% (Fig. 2). Polar solvents, capable of dissolving the inorganic lead and ammonium halide salts, are injected into a nonpolar “poor” solvent in the presence of coordinating ligands to stabilize the newly formed particles. It is noted that these syntheses are conducted at low temperature (<80 °C) [25–28, 32, 33, 35, 36]. Alternatively, CsPbX3 NPs could also be prepared using a hot injection method at higher temperatures ranging 140–200 °C under N2 atmosphere [37–40]. Here, the presence of oleic acid (OA) and oleylamine (OAm) ligands helps to inhibit crystal growth, passivate surface defects, and contribute to colloidal stability. This method has been successfully employed in the fabrication of high-quality inorganic semiconductor NPs [41]. One of the remarkable features of the hybrid perovskites is that no additional surface passivation is necessary to achieve high photoluminescence quantum yields (PLQYs), in contrast to traditional semiconductor QDs passivation, as the dangling bonds play a negligible role in the photoluminescence (PL) emission [37, 39].

Fig. 2
figure 2

Adapted image reproduced with permission of Ref. [27, 34]

Synthesis of CH3NH3PbBr3 QDs. Schematic illustration of the reaction system (a) and precursors (b). c Optical image of colloidal CH3NH3PbBr3 solution. d Schematic illustration for the formation of the CH3NH3PbBr3 QDs.

Full size image

The PLQY of QDs was gradually optimized by reaction and ligand selection. Utilizing long-chain ammonium bromide ligands, colloidal CH3NH3PbX3 cubic NPs (~6 nm) produced using the LARP method [28] exhibited bright-green PL emission at 527 nm, with a PLQY of approximately 20%. The increased reaction temperatures (~120 °C) yielding equal quantum efficiencies and PL emission wavelengths [42] showcased the rigidity and high reproducibility of the above method. The synthesis was further optimized by increasing the organic/inorganic precursor ratios. This could narrow the PL emission (FWHM = 30 nm at 520 nm) and significantly improve PLQY values to a new record of 83% [25]. Quantization effects were observed for 1.8- to 3.6-nm-sized CH3NH3PbBr3 NPs [27, 43]. The highest PLQY ~93% was recorded from the supernatant phase of centrifugation [25, 28].

Despite the high PLQY and colloidal stability of methylammonium halide NPs (>5 months of storage in air under dark conditions [25]), a major shortcoming arises from its instability in polar solvents. To overcome this limit, Vybronyi et al. synthesized CH3NH3PbX3 NPs without using polar solvents [44]. Although the precipitate displayed lower quantum efficiencies (25–50%) than previously reported CH3NH3PbBr3 NPs, it demonstrates an alternative synthesis route without the use of polar solvents. Other approaches involved the formation of PbS/CH3NH3PbX3 core–shell NPs via ligand-exchange reactions [45, 46]. Light emission associated with PbS/CdS NPs and PbS/CH3NH3PbI3−x Cl x core–shell NPs has also been reported [47].

Beyond the methylammonium halide perovskite materials, color-tunable full-inorganic CsPbX3 perovskite NPs (4–25 nm in diameter) utilizing a hot injection method in a temperature range of 140–200 °C (Fig. 3) have been reported [37]. The resulting NPs exhibited high PLQY of 50–90% and narrow emission linewidths of 1242 nm. Owing to the large Bohr radius calculated for CsPbCl3 (5 nm), CsPbBr3 (7 nm), and CsPbI3 (12 nm), quantum confinement effects could be conveniently observed [37]. From transient absorption spectroscopy analysis, it was determined that the high PLQY arose from negligible electron–hole trapping pathways [48] and an average PL lifetime of 1–29 ns [34, 40, 48].

Fig. 3
figure 3

Adapted image reproduced with permission of Ref. [37]

Colloidal CsPbX3 perovskite NCs (X = Cl, Br, I) exhibit size- and composition-tunable bandgaps covering the entire visible spectral region with narrow and bright emission: a Colloidal solutions in toluene under UV lamp (λ = 365 nm) excitation. b Representative PL spectra. c Typical optical absorption and PL spectra. d Time-resolved PL decays for all samples shown in c except CsPbCl3.

Full size image

3.2 Nanowires

The high absorption coefficient enables it as an amazing absorber, which was first applied in dye-sensitized solar cells (DSSCs) to replace organic dye by Kojima et al. [13], Horvath et al. [49]. The further investigation shows that they not only play the role of the light absorbers, but also can be viewed both as electron- and hole-transporting media, owing to their ambipolar charge transport character [50]. In perovskite film synthesis, a nonporous homogenous perovskite film must be deposited in order to avoid shunting in planar devices. However, films produced by conventional spin coating methods were found to be comprised of large CH3NH3PbI3 microwhiskers and many uncovered void areas [51]. Dendrite crystal growth implies that organolead iodide perovskite (OIP) exhibits preferential growth. Inspired by that phenomenon, the synthesis of OIP nanowires (NWs) was investigated and implemented in our group [52]. In addition to our work, Swiss scientists Endre Horvá et al. pioneered the synthesis of OIP NWs by a simple slip coating method during a similar period [53]. The above two works opened the new research fields of halide perovskite NWs. The key synthesis methods and novel properties are summarized in chronological order.

3.2.1 One-Step Evaporation-Induced Self-Assembly (EISA) Growth

For the one-step method, the precursor solution (MAI and PbI2 in DMF) was cast on substrates, and then, the evaporation of the solvent resulted in supersaturated crystal growth with the c-axis parallel to the surface at room temperature [54]. The low symmetry (tetragonal I4/mcm) of OIP and strong intermolecular interactions may allow a preferred growth into NW morphology [55, 56]. For a high substrate temperature (>120 °C), it tended to grow homogeneously into NPs and suppress the preferential growth. Subsequent thermal annealing (~80 °C) can help to convert the NWs precursor into perovskite NWs.

In most cases, self-assembled NWs by solution methods tend to distribute in a macroscopically random manner on substrates. The uniform distribution rather than NW bundles was implemented by UV–ozone treatment, owing to the improvement in surface DMF infiltration (Fig. 4d). The selective area growth and patterning of perovskite NWs were also implemented by UV–ozone treatment assisted by a shadow mask in our first CH3NH3PbI3 NW work [58] (Fig. 4e). In addition, Endre Horvá et al. utilized two glass microscope slides sandwiching saturated MAPbI3 DMF solution to shape an ultrathin MAPbI3 precursor solution in order to control the NW distribution. Combined variations of solvent concentration, temperature, fluidphilicity/phobicity, and sliding speed were utilized to control the crystallization kinetics, yielding tunable NW width, height, and length (Fig. 4a, b).

Fig. 4
figure 4

Adapted image reproduced with permission of Ref. [49, 52, 57–59]

One-step growth of perovskite NWs. a, b Slip coating fabrication process of filiform halide perovskite and their optical image. c, f Roll-to-roll printing for perovskite NWs thin film. d EISA method to fabricate aligned perovskite NWs. e Selective area deposition of perovskite NWs. g SEM image of NWs webs. f Transmittance spectra of NWs webs evolved with the precursor concentrations. Inset shows a photograph of semitransparent NWs networks from 30 wt% concentration. h Transmittance evolved with OTP precusor concentrations. i, j Full-inorganic CsPbBr3 NWs synthesized by one-step method and their typical optical absorption and PL spectra (j).

Full size image

Our further work created a network-like NW web, which could satisfy the cross-linking and uniform NW distribution requirements to overcome the obstacles to NW applications [52]. In contrast to our previous drop-casting style, we adopt a spin coating method with suitable rotary speed and an annealing technique to obtain perovskite NW webs (Fig. 4g). All the NWs welded to each other without opening the end. The transparency of NWs webs could be facilely tuned by the precursor concentrations (Fig. 4h). The low-temperature fabrication process and web geometry promise high application potential in transparent and flexible optoelectronics. The usual one-step growth methods often achieve large diameter NWs over 200 nm. Yang and coworkers added a surfactant solvent into the precursor to tune the NW crystallization kinetics [59] (Fig. 4i).

The application of low-dimensional NWs needs large-scale growth to meet the wide application fields. The blade coating method enables the synthesis of aligned single-crystalline OIP microwire (MW) arrays in terms of high yield and wafer size capability. Thus, Jie et al. first applied the doctor blade coating technique for large-scale and aligned NW growth [60]. It is well known that the highest efficiency deposition technique is printing. The low deposition temperature and one-step solution method promise potential applicability by printing methods. Recently, Yang et al. developed a large-scale roll-to-roll microgravure printing technique for perovskite NWs synthesis [61] (Fig. 4f). By systematic deposition recipe optimization, perovskite NW thin film was deposited on PET substrates (Fig. 4c).

3.2.2 Two-Step Method Growth

Similar to perovskite thin-film fabrication, researchers developed a two-step method in order to face the specific requirements. Without using a novel technique such as electrospinning, MAPbI3 NW film was conveniently formed by two-step spin coating technology. They simply coated the PbI2 layer with an isopropanol solution of MAI in the presence of a small amount of polar aprotic solvent [30]. The NW diameters ranged from 30 to 200 nm, much smaller than those obtained from a one-step solution method. The locally dissolved PbI2 may serve as a preferential site for reacting with MAI to grow a 1D structure, like a liquid catalyst cluster model [62]. Time-resolved fluorescence spectroscopy confirmed that charge separation at ETL/perovskite was faster for 1D NWs than for 3D nanocubes, because of the lager surface area of the former structures (Fig. 5c). The conductivity of NW film was enhanced by a factor 1.3–1.6, indicating better connectivity pathways and apparently increased mobility (Fig. 5d).

Fig. 5
figure 5

Adapted image reproduced with permission of Ref. [30, 63]

Two-step method for perovskite NW growth. a SEM images of vertically aligned CsPbBr3 NWs with a rectangular cross section. b HRTEM and FFT images of a CsPbBr3 NWs. c Fluorescence decay kinetics of MAPbI3 NWs. d Conductivity improvement in the in-plane perovskite NWs. e SEM image of CH3NH3PbBr3 nanorod array. f, g Typical absorption and PL spectra of CH3NH3PbBr3 nanorod arrays.

Full size image

Similarly, Yang’s group replaced the second spin coating step with immersion in CH3NH3Br IPA solution [30]. The lead acetate substrate could support vertical perovskite nanorod arrays (Fig. 5e). The NW absorption onset was at approximately 530 nm with a bandgap of 2.3 eV [64] (Fig. 5f). The room-temperature PL of the CH3NH3PbBr3 NWs peaked at 534 nm with a narrow FWHM of 26 nm under 325-nm HeCd laser excitation (Fig. 5g). The temperature-dependent phases of the CH3NH3PbBr3 NWs underwent multiple phase transitions at 236.3 (to a tetragonal phase), 154.0 (to another tetragonal phase), and 148.8 K (to an orthorhombic phase), which may account for the observed redshift [65]. In comparison with bulk crystal PL, the CH3NH3PbBr3 nanorod arrays had a 13% contribution of the fast component to the steady-state PL, which indicated the high quality of the single-crystalline nanorods.

3.2.3 Chemical Vapor Transport (CVT) Method

This instrument is the same as a tube furnace for traditional chemical vapor deposition [66]. The reaction sources of PbX2 and CsX (X = Cl, Br, or I) powders were placed inside a quartz tube reactor [66]. The Si substrate was positioned at a distance from the source. The temperatures of the powder sources and Si substrate were set at 570–600 and 350–380 °C, respectively. The inert carrier gas was utilized to transport the source vapor for deposition on the Si substrate (Fig. 5a, b). Apart from the above-mentioned research, Yi-bing Cheng et al. also utilized the CVT method to synthesize perovskite MWs [67].

3.2.4 Template Method

According to the lead ionic precursor NWs and the later morphology-retaining reaction, Zhang’s group introduced a new lead NW template as the lead source [68]. The reaction between Pb(NO3)2 and l-cysteine aqueous solution with the assistance of ethanolamine could yield the Pb precursor NW templates [69]. The solid hybrid Pb-containing NWs further transformed into porous CH3NH3PbBr3 perovskite NWs by the addition of CH3NH3Br and HBr 2-propanol solution (Fig. 6a–c).

Fig. 6
figure 6

Adapted image reproduced with permission of Ref. [68, 70, 71]

Template method for perovskite NWs: Pb-containing precursor template a method for perovskite NWs (b, c). AAO template for perovskite NWs. d Photograph of a ~9 × 9 cm2 NW arrays. e Cross-sectional SEM images of perovskite NWs. Scale bar is 500 nm. f Average τ c for different diameter NWs. g Extension of carrier lifetime as the decrease in NW diameter. h, i Nanofluidic channel template growth.

Full size image

The traditional universal templates, anodized aluminum oxide (AAO) templates, were also applied for the synthesis of perovskite NWs [70]. Oriented cylindrical nanopores constructed the uniform vertical perovskite nanowire array templates. This method combined the AAO template and perovskite one-step solution method. Perovskite precursor DMF solution filled into the pores of AAO. The following thermal annealing led to perovskite recrystallization and NW formation at the bottoms of the pores (Fig. 6d, e). The advantage of the present AAO template method is the fine-tuning of the NW dimensions (diameter and length) by controlling the AAO anodizing recipes [66, 72–74].

Different from the vertical templates, traditional planar electronic fabrication techniques such as photolithography, electron beam lithography could help to fabricate the 1D channel as the planar growth template. The perovskite NWs were fabricated with the assistance of open nanofluidic channel templates [30] (Fig. 6h, i). The synthesis could be guided and visualized in real time and underwent a metastable solvatomorph formation in polar aprotic solvents. The superior advantages were the precise controlling of sizes, cross-sectional shapes, aspect ratios, and orientations, which had not been achieved by other deposition methods.

3.3 Two-Dimensional Perovskite

Two-dimensional (2D) perovskites such as nanosheets, nanoplatelets, and microdisks (MDs) have recently shown high PLQY [75–77]. These 2D perovskites are promising candidates for a variety of applications in nanoelectronics, nanophotonics, and photovoltaics [32, 78]. For instance, Zhao and Zhu [79] prepared MAPbI2Br nanosheets with a 1.8-eV bandgap through a thermal decomposition process from a precursor containing PbI2, MABr, and MACl. The planar solar cells based on the compact layer of MAPbI2Br nanosheets achieved a PCE of ~10%.

Xiong’s group successfully grew well-defined polygonal CH3NH3PbX3 (X = Cl, Br, I) nanoplatelets by a chemical vapor method [24, 80]. As shown in Fig. 7, PbX2 platelets were first prepared on muscovite mica using van der Waals epitaxy in a vapor transport chemical deposition system. Subsequently, these PbX2 platelets were placed downstream in a quartz tube to react with MAX vapor by a gas–solid heterophase reaction for conversion into perovskite nanoplates. Figure 8 shows the crystal structure and morphologies of perovskite nanoplatelets. Interestingly, the thickness of PbI2 correlated with OIP platelets by a factor of 1.81, which was in good agreement with their lattice constant ratio along the c-axis. This work offers a reliable method to control the thickness of perovskite platelets. Later, the same group readily applied these perovskite nanoplatelets to fabricate near-infrared solid-state lasers, which exhibited low thresholds and wide mode tunability [24].

Fig. 7
figure 7

Adapted image reproduced with permission of Ref. [76]

a Single-crystalline organometal halide perovskite nanoplatelets were synthesized by a one-pot method. Inset is the photograph of the achieved colloidal perovskite nanoplatelet solution. b TEM images and selected area diffraction pattern of MAPbBr3 nanoplatelets. c X-ray diffraction of MAPbBr3 nanoplatelet film. d PL spectra of MAPbBr3 nanoplatelets in toluene solution and solid state. Inset, the optical images of MAPbBr3 nanoplatelets in toluene solution and in solid-state thin film under ambient light (left) and UV irradiation (right). e Schematic drawing of the nanoplatelet synthesis setup using a home-built vapor transport system.

Full size image
Fig. 8
figure 8

Adapted image reproduced with permission of Ref. [24]

Chemical vapor-deposited methylammonium lead halide perovskite nanoplatelets. a Schematic figure of crystal structures of CH3NH3PbI3−a X a (X = I, Br, Cl). Morphological and electronic band edge characterizations of CH3NH3PbI3 (b), CH3NH3PbI3−a Br a (c), and CH3NH3PbI3−a Cl a (d) nanoplatelets.

Full size image

Recently, Liao et al. [81] fabricated single-crystalline CH3NH3PbBr3 square MDs-based microlasers by using a one-step solution self-assembly method. That approach was similar to anti-solvent vapor-assisted crystallization method [82]. The obtained square MDs had smooth outer surfaces and sharp edges and displayed an absorption peak at 535 nm and an emission peak at 545 nm. Their four side faces constituted a built-in whispering-gallery mode microresonator with a quality factor as high as 430. By partial replacement of Br with Cl, the lasing wavelength can be effectively tuned in the green-light range from 525 to 557 nm.

In a more recent study, 2D MAPbBr3 nanoplatelets with nearly single unit cell thickness and submicron lateral dimensions were prepared by a colloidal synthesis method [29]. Those 2D nanoplatelets exhibited a single and sharp excitonic absorption featured at 431 nm, which blueshifted by 0.5 eV from that of the 3D bulk perovskite phase. This large blueshift was a clear evidence of one-dimensional quantum confinement. A similar colloidal synthetic method was used for full-inorganic CsPbBr3 nanoplatelets (Fig. 9), which exhibited narrow PL and strong excitonic absorption [83]. Recently, Yu et al. demonstrated that the crystallinity of monoclinic CsPbBr3 was much lower than that of tetragonal CsPb2Br5, and thus, they reported a new type of highly luminescent perovskite-related pure tetragonal CsPb2Br5 nanoplatelets synthesized by a facile precipitation reaction. Moreover, a new kind of pure phase perovskite nanosheets can be obtained by utilizing cation exchange and anion exchange [84]. In the tetragonal CsPb2Br5 structure, one layer of Cs ions is sandwiched between two layers of Pb–Br coordination polyhedrons, which have an obviously different crystal structure from the CsPbBr3 perovskite. Jiang et al. synthesized CsPb2Br5 nanosheets by using a solution-phase method. They found that the tetragonal CsPb2Br5 nanosheets were formed via an oriented attachment of CsPbBr3 nanocubes with an orthorhombic structure [85]. Jang et al. [86] synthesized MAPbBr3 nanoplates using octylamine as the capping ligand. The composition of nanoplates was tunable by a simple halide exchange reaction. Furthermore, Yang et al. reported the solution growth of atomically thin, uniform, and square-shaped 2D hybrid perovskites of (C4H9NH3)2PbBr4 [87]. Different from conventional 2D materials, the (C4H9NH3)2PbBr4 sheet exhibited an unusual lattice constant expansion, which led to a slightly shifted band edge emission relative to the bulk counterpart. Those 2D crystals also displayed high PL quantum efficiency and color tunability through halide substitution and thickness variation.

Fig. 9
figure 9

Adapted image reproduced with permission of Ref. [75]

Reaction temperature influence on colloidal CsPbBr3 synthesis. 8- to 10-nm nanocubes were formed at 150 °C (a), 20-nm nanoplates from 130 °C (b) reaction and 90 °C, c reaction yielded several hundred nanometer scale lamellar nanostructures. Scale bar is 50 nm. d Absorption (solid lines) and emission (dashed lines) spectra of NPLs and nanocubes for comparison. e Colloidal solution of anion-exchanged NPLs in hexane under UV illumination (λ = 365 nm). f Inorganic perovskite PL peaks evolved with their anions.

Full size image

It is interesting to note that perovskite nanosheets were surprisingly obtained from electrospraying the precursor solution into a mixed bath of toluene (as anti-solvent) and oleylamine (for intercalation) [88]. Recently, single crystals of OIP nanoplates with well-defined facets were grown via a dissolution–recrystallization path from PbI2 (or PbAc2) films [89]. These 2D perovskite nanostructures displayed strong room-temperature PL and long carrier lifetime.

4 Low-Dimensional Perovskite Optoelectronic Applications

4.1 Light-Emitting Diodes

The key parameter for high-performance light-emitting materials is the PLQY, defined as the ratio of emitted to absorbed photons. On the LED level, the radiant efficiency (RE) or wall plug efficiency (WPE) is the electrical-to-optical energy conversion efficiency described as:

$${\text{RE}}\left( {{\text{or}}\;{\text{WPE}}} \right) = {\text{EQE}} \times {\text{LE}} = \eta_{\text{injection}} \times {\text{IQE}} \times \eta_{\text{extraction}} \times \left( {L/IV} \right),$$

where the external quantum efficiency (EQE) is defined as the ratio of emitted photons to the number of electrons injected into the device and the luminous efficacy (LE) is the ratio of emitted photons to the energy injected by the source (IV). The ultralow amplified spontaneous emission (ASE) thresholds of organometallic halide perovskites provide a strong impetus for light-emission applications.

The extraordinary enhancement of the PLQY in the past 2 years has now placed perovskites on a par with the best-in-class solution-processed semiconductors. Different strategies, such as synthesis of low-dimensional layered perovskites, or increasing spatial confinement, have been applied to maximize PLQYs. The high PLQYs combined with the compositional flexibility of these materials render perovskites robust technological candidates. They are distinguished by: (1) high color purity (FWHM ~20 nm) [37] (Fig. 10), irrespective of the crystallite size [32, 91]; (2) bandgap tunability covering the entire spectrum of visible light in layered perovskites [92, 93]; and (3) low to moderate ionization energy to form stable functional interfaces [94].

Fig. 10
figure 10

Adapted image reproduced with permission of Ref. [90]

a–c Emissions of QLED devices utilizing different QD sizes. d The EL (solid line) and the PL spectra (dashed line) of samples shown in (a–c). e CIE coordinates of the three color QLEDs (circular) compared to the NTSC color standards (star).

Full size image

Organic–inorganic lead-based perovskites were first investigated for light-emitting applications before they were used as absorbers in DSSC devices. Perovskite LEDs were first demonstrated in the 1990s using organic–inorganic perovskites with layered two-dimensional structures. However, electroluminescence from radiative recombination of injected electrons and holes was only observed at liquid nitrogen temperatures. Recently, perovskite light-emitting diodes (PeLEDs) of the low-dimensional metal halide perovskites MAPbI3−x Cl x , MAPbBr3, and MAPbI3−x Br x have been demonstrated. These devices operated at room temperature under substantial current densities and brightness. The internal quantum efficiencies reached 3.4%, and the emissions were located at near-infrared and green (Fig. 11). Comparable emission has also been demonstrated in an inverted structure by tuning the work function of the conventional PEDOT:PSS hole injection layer. However, the above two devices require high charge densities for efficient radiative recombination. Enhanced device emission and quantum efficiency would be obtained by removing or filling material trap states and optimizing film quality, similar to halide perovskite solar cells.

Fig. 11
figure 11

Adapted image reproduced with permission of Ref. [96]. (Color figure online)

Device characteristics of visible PeLEDs. a Absorption (black), normalized EL (solid line in green), and PL (dashed line in green) spectra of CH3NH3PbBr3 perovskite. EL spectrum of mixed halide perovskite is shown in red. b Luminance (black) and current density (red) versus voltage characteristics of the green PeLED. c EQE versus voltage characteristics of the green PeLED. d EQE versus current density of the green PeLED.

Full size image

In addition to possessing maximal external luminescence, a good light-emitting device requires efficient carrier injection and radiative recombination. In particular, balancing the electron and hole injection by tailoring the interfaces helps to focus the radiative recombination within the desired emitter layer. In this respect, metal halide perovskites can meet the charge transport and charge injection requirements with high carrier mobility and optimum interface formation with most employed contacts. In PeLEDs, the emitter layer may comprise 3D, layered, or nanostructured perovskites sandwiched between electron and hole transport layers (Fig. 12). The EQE of PeLEDs has rapidly progressed from less than 1% to over 8% in fewer than 2 years. In addition, the CE and luminance are approaching the QLED and OLED top ranges. Tae-Woo Lee and coworkers used an organic small molecule, 2,2′,2″-(1,3,5-benzinetriyl)-tris(1-phenyl-1-H-benzimidazole), as an additive to further reduce perovskite grain size. The perovskite exciton diffusion length was decreased by spatial confinement in uniform MAPbBr3 nanograins. Simultaneously, they prevented the formation of metallic lead (Pb) atoms that cause strong exciton quenching through a small increase in methylammonium bromide (MABr) molar proportion. They eventually boosted the current efficiency (CE) of PeLEDs with a simple bilayer structure to 42.9 cd A-1 [95]. The representative results of PeLEDs are summarized in Table 1. Further breakthroughs in PeLED performance mainly include material design, optimized charge injection layers, and emission tunability to yield high color quality, purity, and white-light emission.

Fig. 12
figure 12

Adapted image reproduced with permission of Ref. [90]

Illustration of multilayer PeLED device. a Device structure. b Cross-sectional TEM image of multiple layers with distinct contrast. Scale bar is 50 nm. c Flat-band energy level diagram.

Full size image
Table 1 Overview of selected representative results of PeLED [36, 76, 88–93, 95]
Full size table

4.2 Solar Cells

Perovskite NPs were originally used as sensitizers in DSSCs. Our group first reported perovskite NW solar cells utilizing the same device structure as thin-film solar cells (Fig. 13a, b) [52]. Compared to the contemporary champion efficiency (∼19.3%) of OIP thin-film solar cells [4, 78, 97], the obtained J sc values were much lower because of the thicker film required to avoid shorting of the inter-NWs void. However, for specific fields, the non-full fill of substrate by NWs may find suitable applications such as semitransparent solar cells, building-integrated photovoltaics. The elementary results were greatly improved by the group of Nam-Gyu Park [63]. In order to avoid the void-induced shorting effect, they developed a two-step method to grow denser and smaller-diameter NWs as the absorber (Fig. 13c, d). Perovskite NWs with a mean diameter of 100 nm showed faster carrier separation in the presence of the hole-transporting layer and higher lateral conductivity than their 3D counterparts. The best performing device exhibited a photocurrent density of 19.12 Ma cm−2, voltage of 1.052 V, and fill factor of 0.721, leading to a PCE of 14.71% with small I–V hysteresis (Fig. 13e, f).

Fig. 13
figure 13

Adapted image reproduced with permission of Ref. [52, 63]

Perovskite NW solar cells: a, b SEM image of NWs from a one-step method and its corresponding J–V curve. Two-step-method-grown NWs and corresponding device performances. c, d Top-view and cross section of NWs films. e, f J–V and IPCE curves of NW solar cells. The scale bar is 1 μm.

Full size image

There are many reviews discussing about halide perovskite thin-film solar cells. Thus, we only list the key breakthroughs in chronological order. Miyasaka and coworkers were the first to report CH3NH3PbBr3 solar cells with a PCE of 2.2% in 2006 [98]. In 2009, they improved the PCE to 3.8% by replacing bromine with iodine. Subsequently, Park and colleagues optimized a titania surface and substituted DMF solvent by γ-butyrolactone, yielding an efficiency of triiodide cells to 6.5% in 2011 [14]. Unfortunately, the device stability had not significantly improved. Until 2012, Kanatzidis and coworkers utilized the p-type solution-processable perovskite fluorine-doped CsSnI3 as a solid HTL in a solid-state DSSC [99]. This was the first time that a perovskite material has been used as the HTL with efficiencies up to 10.2%. The real breakthrough of perovskite stability was obtained by the Gratzel and Park groups [5]. They utilized MAPbI3 as a light harvester combined with the solid hole conductor 2,2′,7,7′, tetrakis(N,N-dimethoxy-phenyl-amine)-9,9′-spirobifluorene (spiro-MeOTAD) on mesoporous TiO2 leading to a PCE of 9.7% and dramatically improving the device stability compared to CH3NH3PbI3-sensitized liquid junction cells. From the above-mentioned research [5, 99], it can be concluded that OIP had the excellent capability to achieve ambipolar charge transport. At the same time, a significant study reported by Seok, M. Gratzel, and coworkers [100] boosted the PCE to 12%.

In the middle of 2013, there were two important reports published in the journal Nature, which accelerated the study of perovskite solar cells to a new level [4, 16]. Both of them had achieved the PCE over 15% by using new, modified perovskite processing methods. One defined as “sequential deposition” produces a device structure of FTO/cp-TiO2/mp-TiO2/MAPbI3/spiro-OMeTAD/Au [102] and the other developed by H. Snaith and coworkers deposited a high-quality MAPbI3−x Cl x film via dual-source vacuum deposition in a planar heterojunction (PHJ) perovskite solar cell, as shown in Fig. 14. This FTO/cp-TiO2/MAPbI3−x Cl x /spiro-OMeTAD/Ag-based device achieved a PCE 15.4% [4]. Up to now, to the best of our knowledge, Seok and coworkers maintain the record certified PCE of 22.1% (Fig. 14). Simultaneously, all-inorganic perovskite solar cells have been reported. Liu and coworkers for the first time discovered how to make CsPbBr3 perovskite solar cells out of QDs and achieved 6.7% conversion efficiency. Recently, Joseph M. Luther and coworkers implemented a low-temperature synthesis of α-CsPbI3 perovskite. The highest PCE of all-inorganic perovskite solar cells reached 10.77% [103].

Fig. 14
figure 14

Adapted image reproduced with permission of Ref. [101]

SEM, J–V, and EQE measurements for perovskite solar cells. a Cross-sectional SEM image of the device. The comparison of SEM surface images of FAPbI3-based layer formed on mp-TiO2 by IEP (b) and conventional method (c). d J–V curves of best device measured in reverse and forward modes, and e EQE spectra for best device and integrated J SC.

Full size image

As the PCEs of perovskite solar cells are very sensitive to the morphology, composition, uniformity, and solution process, among others, the fabrication of perovskite absorbers must be accurately controlled. The main methods (Fig. 15) include one-step deposition [104] and two-step sequential deposition [4] methods based on solution processing, vacuum deposition [16, 105], and vapor-assisted solution processes [18]. In addition, a fast deposition crystallization (FDC) procedure was developed to yield highly uniform perovskite film consisting of microsized crystals, as illustrated in Fig. 15.

Fig. 15
figure 15

Adapted image reproduced with permission of Ref. [18, 105–107]

a One-step deposition. b Two-step sequential deposition. c Dual-source coevaporation. d Sequential vapor deposition. e Vapor-assisted solution growth process.

Full size image

4.3 Photodetectors

PDs have a wide range of commercial and scientific applications in imaging, optical fiber communications, spectroscopy, and biomedical applications [108–110]. The commercially available PDs are commonly made by SiC, Si, and HgCdTe for UV, visible, and infrared detection, respectively [110–114]. Endre Horvá et al. first constructed two-probe detectors based on a few perovskite NWs for 633-nm light detection [49]. Under laser illumination, the light-generated electron–hole pairs caused an increase in the conductance of the material up to a factor of 300 (Fig. 16a, b). The calculated responsivity was estimated to be 5 mA W−1, comparable (10 times higher) to the value that has been achieved with the first prototypes of those 2D materials [115, 116]. The response time showed that rise and decay times for the on–off current are less than 500 μs, ∼104 faster than the state-of-the-art PDs made of 2D TMD materials [117, 118]. Considering the nonuniformity of random NW PDs, our group utilized aligned NWs arrays to construct conductive PDs and systematically investigated the PD performances [52]. Monolayer NW-based PDs exhibited a response time of ∼0.3 ms, a responsivity of 1.3 A W−1, and a detectivity of 2.5 × 1012 Jones, which are superior to recently reported PD performances based on perovskite thin films and inorganic NWs (Fig. 16c, d).

Fig. 16
figure 16

Adapted image reproduced with permission of Ref. [49, 52, 60]

Perovskite NW photodetectors: slip coating-based NWs PDs. a Dark and laser-illuminated I–V curves. b Time-resolved photoresponse. EISA method-based device performances. c I–t curves. d High-resolution scan to one cycle of I–t curves. Doctor blading-based NW PDs. e Wavelength-dependent photoresponsivity of the CH3NH3PbI3 MW array-based photodetectors. f Photocurrent versus light intensity curves. g Variation of dark current/photocurrent with day of photodetector based on OTP MW arrays. h Perovskite MW PD arrays for light source mapping. NW web-based PDs for imaging. i–j Flexibility performances. k–m NW network PD arrays for imaging. n, o Encapsulation could help to improve the device stability.

Full size image

The above studies were all based on individual PDs. However, the practical applications of PDs (for example, in imaging) require large number of pixels to construct target images. Therefore, our group first utilized perovskite PD arrays and utilized them to preliminarily implement imaging systems [58] (Fig. 16i–o). We developed a modified one-step method to implement the new NW morphology of NW webs. The PD arrays were applied in imaging and successfully obtained clear mapping of the light source signal. The network PD arrays showed superior flexibility, which could withstand wide angle (20°–80°) bending over a large number (~10,000) of cycles. Moreover, the sandwich structure (PET/OTP/PMMA) could greatly improve the device stability and the network PD arrays easily stayed intact for one month of storage in air. Our PD responsivity was further optimized by Jie’s group [60] (Fig. 16e–h). Large-scale PD arrays were fabricated by doctor blading, and the corresponding responsivity was improved to 13.5 A W−1. The representative results of perovskite PDs are summarized in Table 2.

Table 2 Overview of selected representative results of perovskite PDs [49, 52, 54–57, 119]
Full size table

4.4 Lasing

Semiconductor NW lasers, owing to their ultracompact physical sizes, highly localized coherent output, and efficient waveguiding, are promising building blocks for nanoscale photonic and optoelectronic devices [120]. Each NW can serve as a waveguide along the axial direction and the two end facets form Fabry–Perot cavities for optical amplification. One of the main obstacles limiting the potential applications of semiconductor NW lasers is the high-threshold carrier density required for lasing. The amazing solar cell performance is ascribed to the long carrier lifetimes and diffusion lengths (µm range) [121–124]. These properties combined with high fluorescence yield and wavelength tunability [96, 124, 125] render lead halide perovskites as ideal materials for lasing. Song Jin et al. teamed up with X–Y. Zhu to develop a surface-initiated solution growth strategy to synthesize high-quality single-crystal NWs. NWs below the lasing threshold (P Th) showed uniform intensity from the whole NWs, and those above P Th demonstrated strong emission with spatial interference at the two end facets [126]. They exhibited a quality factor 3600, which was more than an order higher than that of the state-of-the-art GaAs–AlGaAs core–shell NW laser operating at a temperature of 4 K [127]. In addition to single-mode lasing, they also observed multiple lasing modes from perovskite NWs. For comparison, the QYs from OIP polycrystalline thin films were <10% for spontaneous emission (SPE) and ∼15% for amplified SPE [125]. For another comparison, the state-of-the-art GaAs–AlGaAs core–shell NWs laser has a carrier lifetime of ∼440 ps and a QY of ∼0.4% [127]. These comparisons suggest that the unique lasing performance may be ascribed to the exceptionally low trap density in their single-crystal NWs. A major advantage of lead halide perovskites for lasing application was the broad wavelength tunability based on composition stoichiometry [125]. They further demonstrated the color tunability from near-infrared to blue in single-crystal perovskite NWs. By simply mixing different amounts of methylammonium iodide and bromide or bromide and chloride in the precursor solution, they successfully synthesized single-crystal NWs of MAPbBr y I3−y and MAPbCl x Br3−x alloys with various stoichiometries (Fig. 17a–f). Room-temperature-tunable NW lasers were also implemented by Xiong’s group, who adopted organic–inorganic perovskite NWs as the Fabry–Perot cavities to emit lasers [30].

Fig. 17
figure 17

Adapted image reproduced with permission of Ref. [67, 119, 128, 129]

Perovskite NWs lasers. a–f SEM image and nanolaser performances from MAPbX3 NWs. g A set of dark-field images of a IOP MW waveguide. Scale bar is 10 µm. h, i FAPbX3 NW laser performances. j–l Full-inorganic halide perovskite NW lasers.

Full size image

Considering the photo- and thermal stabilities of MAPbX3 NW lasers, Song Jin et al. reported a low-temperature solution growth of single-crystal NWs of formamidinium lead halide perovskites (FAPbX3) that feature redshifted emission and better thermal stability compared to MAPbX3 [129]. They demonstrated optically pumped room-temperature near-infrared (∼820 nm) and green lasing (∼560 nm) from FAPbI3 and FAPbBr3 NWs with low lasing thresholds of several microjoules per square centimeter and high quality factors of approximately 1500–2300. More remarkably, the FAPbI3 and MABr-stabilized FAPbI3 NWs displayed durable room-temperature lasing under ∼108 shots of sustained illumination, greatly exceeding the stability of MAPbI3 (∼107 laser shots). They also demonstrated tunable NWs lasers in a wider wavelength region from FA-based lead halide perovskite alloys through cation and anion exchange (Fig. 17h–i).

Since organic–inorganic lead halide perovskite materials were still limited by their instability. Full-inorganic halide perovskites, cesium lead halides, offered a robust candidate without sacrificing emission tunability or ease of synthesis. Thus, Yang et al. reported the low-temperature, solution-phase growth of cesium lead halide NWs showing low-threshold lasing and high stability [119]. Upon optical excitation, Fabry–Perot lasing occurred in CsPbBr3 NWs with an onset of 5 μJ cm−2 with the NW cavity displaying a maximum quality factor of 1009 ± 5. Lasing under constant, pulsed excitation could be maintained for over 1 h, the equivalent of 109 excitation cycles. Wavelength tunability in the green and blue regions and excellent stability promised those NW lasers as attractive for device applications (Fig. 17j–l). The light–matter interactions of NW lasers were investigated by Jae Kyu Song et al. [66]. The reduced mode volume of the NWs suggested strong light–matter interactions with an exciton–polariton model. The representative results of perovskite nanolasers are summarized in Table 3.

Table 3 Representative results of the ASE/lasing from perovskite gain media [76, 119, 123–126, 130–132]
Full size table

5 Summary and Outlook

In summary, we have reviewed recent developments of structure and growth of low-dimensional halide perovskites and their applications for high-performance optoelectronic devices. Their remarkable intrinsic properties and high degree of structural flexibility allow the potential beyond academic curiosity and industrial interest. In particular, the high QY, narrow emission, high exciton binding energies, color tunability, and facile solution processing at low temperatures have been outlined. Beyond applications as active layers for emitters and optical gain materials, low-dimensional halide perovskites may also be incorporated in currently existing technologies as a candidate to replace phosphors. As the field grows from its infancy stage, various kinds of heterojunctions, architectures, and processing conditions are probable to promote rapid development. For the routes toward commercialization, the key issues such as the optoelectronic properties (e.g., excitation dynamics, nonradiative recombination), working principles, and degradation mechanisms need better understanding. Owing to the successes in photovoltaics, it is reasonable to expect that the research in other perovskite optoelectronics will intensify and usher in a bright future for PDs, lasers, and light-emitting devices (Table 4).

Table 4 Stability improvement strategies for perovskite [103, 130–136]
Full size table

For low-dimensional halide perovskite synthesis, the perovskite formation processes play a paramount role in determining their final device performances. They can be prepared by a variety of techniques via the competition between in situ transformation and dissolution–crystallization mechanisms. Moreover, the use of capping ligands and solvent engineering can help to tailor the shape of perovskite crystals. Emerging applications of these randomly distributed perovskite units require the achievement of controllable alignment in order to match the planar microelectronic fabrication techniques.

Beyond PV applications, low-dimensional perovskite materials with high crystallinity, emission efficiency, and benign defects enable the fabrication of LEDs, lasers, PDs, and other optoelectronic/microelectronic devices. Furthermore, their high atomic number and high density may find applicability to high-energy radiation detection. Additionally, perovskites exhibit interesting ferroelectric properties, potentially due to the free rotation of the polar organic species, which further bolsters their potential for switchable electronics and memory devices. Despite the novel properties of low-dimensional perovskites, several crucial challenges, including toxicity and instability, limit wide industrial application. Limited success has been reported for the replacement of Pb with environmental friendly elements. Nontoxic Sn shares relatively similar properties to Pb in hybrid perovskites and is theoretically expected to yield more efficient performances. However, Sn-based perovskite solar cells experience more serious instability due to the naturally favorable oxidation of Sn(II), which inevitably requires advanced encapsulation techniques. Apart from the toxicity, Pb-based perovskites experience inherent instability under long-term operation, which is even more urgent to address. The degradation mechanisms of perovskites upon exposure to thermal, moisture, UV, and mechanical conditions require a reliable strategy in order to improve material stability. Encouraging results have been demonstrated through interface modifications between transport materials and electrodes [121]. Optimized devices have realized stable working times over 1000 h. However, considering the durability requirements of 20 years operation for solar cells, there is still much work required to improve the intrinsic properties of halide perovskites and more stable device structures. From an academic research perspective, there have been great success for the synthesis and applications of halide perovskites. Further research efforts are expected to focus on: (1) deep understanding of the structure–property relationships of the entire hybrid material system to guide the rational design and careful manipulation of the optoelectronic properties and (2) introducing new ideas to break the trade-off between the halide perovskite novelties and the working stability. Under the intensely driven research, the future development of environmentally friendly and reliably working perovskite devices is hopeful.