Metal Halide Perovskites for Violet and Ultraviolet Light Emission

Abstract

Emissive metal halide perovskites (MHPs) have emerged as candidates for next-generation optoelectronics due to their sharp color purity, inexpensive processing, and bandgap tunability. However, the development of violet and ultraviolet light-emitting MHPs has lagged behind due to challenges related to material and device stability, charge carrier transport, tunability into the ultraviolet spectrum, toxicity, and scalability. Here, we review the progress of both violet and ultraviolet MHP nanomaterials and light-emitting diodes, including materials synthesis and device fabrication across various crystal structures and dimensions (e.g., bulk thin films, 2D thin films, nanoplatelets, colloidal nanocrystals, and more) as well as lead-free platforms (e.g., rare-earth metal halide perovskites). By highlighting several pathways to continue the development of violet and ultraviolet light-emitting MHPs while also proposing tactics to overcome their outstanding challenges, we demonstrate the potential of violet and ultraviolet MHP materials and devices for important applications in public health, 3D printing, nanofabrication, and more.

Introduction

Metal halide perovskites (MHPs) have emerged as a promising semiconductor material for a variety of applications such as solar cells, − photodetectors, photocatalysis, lasers, , and light-emitting diodes (LEDs) ,− due to their favorable electrical, optical, and magnetic properties accessible through inexpensive processing methods. In particular, perovskite LEDs (PeLEDs) have strong potential for next-generation LED technologies due to their sharp color purity, high luminance, bandgap tunability, defect tolerance, and potential for low-cost fabrication. ,,,− Following the rapid development of perovskite solar cells, PeLED efficiencies have likewise experienced significant improvements over the past decade, with external quantum efficiencies (EQEs) now exceeding 32% for red and green PeLEDs − and 26% for blue PeLEDs, compared to 0.1% for some of the first green room-temperature PeLEDs in 2014. ,

Despite the success of red, green, and blue PeLEDs, studies on violet (defined here as between 400 and 435 nm) and ultraviolet (UV) (<400 nm) emissive MHPs have lagged behind (Figure a) due to challenges related to material and device stability, charge carrier transport, tunability into the UV spectrum, toxicity, and scalability. For example, violet and UV (V/UV) light-emitting perovskites like CsPbCl₃ can be difficult to process in solution due to the poor concomitant solubility of CsCl and PbCl₂. Additionally, many reports of V/UV PeLEDs demonstrate short device lifetimes on the order of seconds to minutes. − However, light emission in the V/UV range is crucial for many applications beyond lighting and display technologies, motivating the need for further research to improve these emitters. Solving these current challenges could unlock low-cost, scalable V/UV PeLEDs for applications such as biosensing, 3D printing, micro/nanofabrication, public health (sterilization ,− and water purification), optical communications, and more − (Figure b).

1.

Progress of perovskite light-emitting diodes (PeLEDs). (a) The number of PeLED publications each year sorted by violet/ultraviolet, blue, green, and red PeLEDs. To systematically quantify the research disparity across the emission spectrum of PeLEDs, a bibliometric analysis was performed using the Web of Science Core Collection (https://www.webofscience.com/wos/woscc/smart-search) on February 3, 2025. We utilized a rigorous Boolean search strategy that intersected device definitions with specific spectral indicators, employing the “Topic” field for material identification and the “Title” field for strict chromatic filtering. The data set was generated using the base string Topic: (perovskite*) AND Topic: (“light-emitting diodes”*) refined by AND Title: (color*), where the color variable was iteratively substituted with blue*, green*, red*, or violet* to capture relevant variations. Note: only publications with a listed color are represented and thus this does not fully represent the overall field of PeLED publications. (b) Summary of the applications of light in both the visible and ultraviolet energy regimes. Adapted with permission from ref Copyright 2023 Elsevier Inc.

We highlight some of these promising applications in Figure . From Figure a, Shur et al. illustrate a UV LED fluorimeter to measure the fluorescence from particles dispersed in water or air. The use of UV LEDs within fluorimeters can lead to sharper contrast by inducing a stronger fluorescence response from the sample. Moreover, UV LEDs can be integrated into nanofabrication processes, including photolithography. In Figure b, Feng et al. develop AlGaN UV micro-LEDs to enable maskless photolithography. Specifically, they construct microdisplays made of UV micro-LEDs and transfer the image on a photoresist-coated wafer, showcasing the possibilities for UV micro-LEDs to reduce time and expenses incurred by the semiconductor industry. Lastly, UV light can deactivate microbes (e.g., bacteria and viruses) and thus improve public health, broadly. From Figure c, Bhattarai et al. show the main mechanisms in which UV light can inflict damage on microbes, including thymine dimer formation, protein cross-linking, oxidative damage, protein structure disruption, and enzyme inactivation. The UV–C regime (100–280 nm) is particularly attractive since these emission wavelengths are known to inactivate harmful viruses. , For instance, 270–280 nm light can greatly reduce Escherichia coli populations with doses on the order of a few mJ mL^–1.

2.

Applications of V/UV materials and LEDs. (a) Schematic of a UV LED fluorimeter. Adapted with permission from ref Copyright 2010 IEEE. (b) AlGaN UV micro-LED displays (top) used to pattern photoresist-coated silicon wafers and corresponding maskless photolithography images (bottom). Adapted with permission under a Creative Commons CC BY-NC-ND 4.0 license from ref Copyright 2024 Springer Nature. (c) Five UV light-induced microbial deactivation mechanisms. Adapted with permission under a Creative Commons CC BY 4.0 license from ref Copyright 2024 The Authors.

Traditionally, V/UV light emission for these applications has been achieved by fluorescent mercury lamps, which are inefficient, bulky, and slow to turn on and off. To replace sources like mercury lamps, V/UV LEDs have become increasingly common as they offer a low-power, compact solution with quick turn on and off speeds and the ability to tune the emission spectrum. Currently, commercial V/UV LEDs are based on III–V material systems such as InGaN/GaN/AlGaN materials grown with metal organic chemical vapor deposition (MOCVD). However, MOCVD is a costly and complex process that requires high temperatures and involves toxic and pyrophoric precursor gases like trimethylaluminum (Al­(CH₃)₃) and trimethylgallium (Ga­(CH₃)₃), which significantly contribute to safety concerns and manufacturing complexity. Furthermore, the requirement of lattice-matched substrates and precise thermal budgets elevates the cost and scalability challenges of producing high-performance V/UV emitters.

Compared to fluorescent mercury lamps and III–V LEDs, the MHP material system discussed here can be fabricated using simpler methods via solution-based processing or thermal evaporation, paving the way for low-cost, nontoxic, scalable V/UV PeLEDs. However, emissive MHPs are understudied at this important V/UV spectrum regime, and innovations in materials and devices are needed to achieve such short emission wavelengths. As evident from the current state-of-the-art highlighted in this review, most work on emissive MHP materials and devices focuses on wavelengths greater than 400 nm, with only a few reports of PeLEDs with peak emission between 380 to 400 nm. ,− To unlock the full range of applications in areas like public health, shorter wavelength UV–C emission is required. While there are some material systems with demonstrated emission peaks as low as 265 nm such as Pr³⁺-based double perovskites, further innovations in materials and device engineering are required to enable effective charge transport and device stability for PeLEDs in this challenging UV–C regime.

Thus, in this review, we survey current research and provide a perspective on challenges, emerging trends, and future directions to continue pushing the development of V/UV light-emitting MHPs. We first explore the fundamentals of the MHP structure that dictate light emission. We then discuss the strategies employed to achieve V/UV emission in MHPs through compositional engineering of the halide (X-site), B-site, and A-site as well as dimensional control. We survey the current progress on a variety of material systems that demonstrate V/UV photoluminescence (PL) across different crystal structures and dimensions (e.g., three-dimensional (3D), two-dimensional (2D), one-dimensional (1D), or zero-dimensional (0D) crystal structures in thin films, nanoplatelets, colloidal nanocrystals, and more) as well as lead-free platforms (e.g., rare-earth metal halide perovskites). We then discuss LED device architectures and the different charge injection strategies utilized to achieve V/UV electroluminescence (EL). By reviewing the current state-of-the-art performance of V/UV PeLEDs, we highlight the promising work conducted so far and the challenges and limitations that motivate areas of further research. We then propose tactics to overcome some of the outstanding challenges and suggest several pathways to continue the development of V/UV light-emitting MHPs. Lastly, we present an outlook on the potential impacts of improved V/UV light-emitting MHPs and their corresponding PeLEDs.

Fundamentals of Metal Halide Perovskites for Light Emission

The standard 3D perovskite structure is based on the formula ABX₃, where A and B are cations of different sizes and X is an anion. For traditional MHPs, the A-site is typically an organic or inorganic cation (e.g., methylammonium (MA⁺), formamidinium (FA⁺), Cs⁺), the B-site is a divalent metal (e.g., Pb²⁺, Sn²⁺), and the X-site is a halide anion (e.g., I^–, Br^–, Cl^–). However, due to the vast parameter space of available materials that influence both composition and lattice structure, there are an increasing number of alternative structures that are typically referred to as perovskites or perovskite derivatives.

As shown in Figure a, the standard 3D ABX₃ perovskite structure can be modified to form unique materials by changing the relative amounts of halides at the X-site or by partially or fully replacing cations at the A- or B-sites. The choices in chemical composition for these sites, along with material synthesis and processing parameters, can alter the crystal structure and geometric structure of the material, affecting its electronic and optical properties. Beyond the traditional inorganic or organic cations at the A-site, the crystal structure dimensionality can be controlled by employing large, bulky organic cations like phenethylammonium (PEA⁺) to form layered 2D (PEA₂BX₄) or quasi-2D (PEA₂(ABX₃) _n−1BX₄) crystal structures. , At the B-site, partial doping of lead halide perovskites with ions like Ni²⁺ and Cd²⁺ has been shown to improve the luminescent properties of these materials. , Alternatively, fully replacing the divalent metal (e.g., Pb²⁺) with a trivalent metal such as Sb³⁺ or Bi³⁺ can form materials with an A₃B₂X₉ structure. , Other trivalent metals such as Ce³⁺ can form structures with a different stoichiometry, A₃BX₆. The B-site can also be occupied by a monovalent metal such as Cu⁺, which has been demonstrated in different crystal structures such as A₂BX₃ or A₃B₂X₅. , As evident from the variety of possible crystal structures that arise from changing the A- or B-site cations, the structure formed not only depends on the charge and size of the new cations, but also on the properties of the other ions in the structure that influence the bonding between the atoms and the resulting crystal structure. The choices in chemical composition for these ionic sites, along with material synthesis and processing parameters, can alter the crystal structure and geometric structure of the material, affecting its electronic and optical properties. These choices enable access to a wide array of perovskite, double perovskite, and perovskite-inspired structures, all of which will be broadly included in this review.

3.

Tunable composition and structure, defect tolerance, and tunable emission of MHPs. (a) MHPs can be engineered by tuning the chemical composition at the A-site, B-site, or X-site. These changes, along with other factors like the synthesis method, can result in different crystal structures across 0D, 1D, 2D, and 3D dimensionalities as well as geometric structures across quantum dots, nanowires, nanoplatelets, and thin film morphologies. Structures in the “Composition” section were made with VESTA. Structures in the “Crystal Structure” and “Geometric Structure” sections were adapted with permission from ref Copyright 2018 Royal Society of Chemistry. (b) Location of trap states with respect to the band edges for III–V (e.g., GaAs)/II–VI (e.g., CdTe) semiconductors (left) and MHPs (right). The deep trap states within the bandgap in III–V and II–VI semiconductors can severely impair optoelectronic device performance. Conversely, MHPs have shallow trap states, making them a more defect-tolerant material. Adapted with permission under a Creative Commons CC BY-NC 3.0 license from ref Copyright 2019 Royal Society of Chemistry. (c) Tunable PL emission from CsPbX₃ (X = Cl, Br, I) nanocrystals. Adapted with permission from ref Copyright 2015 American Chemical Society.

For example, these derivatives of the standard 3D ABX₃ structure tend to have a reduced-dimensional (RD) crystal structure like those shown in Figure a (e.g., 2D layers in PEA₂PbBr₄, 1D chains in Rb₂CuBr₃, or 0D octahedra in Cs₃CeBr₆). ,,, There are two main types of 2D perovskite crystal structures: Ruddlesden–Popper structures with offset layers separated by two monoammonium ligands and Dion−Jacobson structures with compact, aligned layers separated by one diammonium ligand. In both types, the alternating layers of organic ligands and 2D perovskite octahedra create a multiple quantum well structure with a high exciton binding energy. There are also two main types of 1D crystal structures, face-sharing chains and edge-sharing chains, where the perovskite structures are confined in two dimensions and form isolated 1D nanowires. When confined in three dimensions, 0D isolated perovskite octahedra are formed. In many of these RD structures, such as 0D Cs₃CeBr₆ or 2D PEA₂PbBr₄, the strong confinement of excitons can lead to higher exciton binding energies and more effective radiative recombination. ,

Along with the dimensionality of the crystal structure, the dimensionality of the geometric structure (i.e., nanoscale morphology) can be controlled by multiple factors including the composition or synthesis method. As shown in Figure a, the geometric structure can be composed of a thin film (3D), nanoplatelets (2D), nanowires (1D), or quantum dots (0D), which may or may not match the dimensionality of the crystal structure. Thus, it is important to distinguish between crystal structure and geometric structure dimensionality when discussing dimensionality effects. For example, a bulk CsPbCl₃ thin film could have a 3D crystal structure and 3D geometric structure. Alternatively, many of the studies reviewed here synthesize CsPbCl₃ nanocrystals (NCs) that are geometrically confined in three dimensions and thus form a 0D geometric structure with side lengths typically on the order of 10 nm. In the MHP literature, the terms NC and quantum dot (QD) are sometimes used interchangeably, but there is a critical distinction. QDs are NCs that are small enough such that their particle size is on the order of the excitonic Bohr radius, which depends on the material system. In this regime, the physical confinement of charges causes QDs to exhibit quantum confinement effects, unlike larger bulk-like NCs. Thus, while QDs are NCs, not all NCs are QDs, and readers should critically evaluate studies to determine if short-wavelength emission derives from quantum confinement or other effects. In cases where materials do exhibit quantum confinement effects, these quantum confinement effects blue-shift the emission to higher energies, making dimensional control one of the key levers toward achieving V/UV emission. , Thus, in addition to 3D ABX₃ perovskites, RD metal halide crystal and geometric structures will also be considered for this review.

Along with tunable structural properties, defect tolerance within MHPs increases their viability for optoelectronic devices like solar cells or LEDs. Unlike III–V or II–VI semiconductor materials (e.g., GaAs and CdTe) which have deep trap states, MHPs typically have shallow traps near the band edges (Figure b). The formation of deep trap states within the bandgap can impair device performance due to increased nonradiative recombination and immobilization of charge carriers, but since the defect states of MHPs are typically near or within the conduction and valence bands, the bandgap for MHPs is relatively unobstructed and free of traps. Thus, MHPs are not a defect- or trap-free material, but the energetic location of the defect states makes MHPs much more defect tolerant as compared to adjacent semiconductor classes. However, for Cl-based perovskites such as violet-emitting CsPbCl₃, halide vacancies can form deeper defects within the bandgap, leading to reduced photoluminescence quantum yield (PLQY) compared to their bromide or iodide counterparts (e.g., CsPbBr₃ or CsPbI₃). Thus, as discussed further in this review, defect passivation is particularly important in the context of V/UV light-emitting MHP materials and devices.

Another promising characteristic of MHPs for light emission is their tunable bandgap and peak emission wavelength that can be altered by modifying the composition of the halide anion (X-site). For lead halide perovskites, it is well-known that increasing halide electronegativity (Cl^– > Br^– > I^–) lowers the energy of the valence band maximum (VBM) but does not significantly alter the conduction band minimum (CBM), resulting in a higher bandgap. , Therefore, by sweeping the halide from iodide to bromide to chloride in CsPbX₃ NCs, the emission can be tuned across the visible spectrum (700 to 410 nm) as shown in Figure c. ,−

The optical properties of MHPs can be further modified by engineering at the A-site or B-site in the ABX₃ perovskite structure. Changing the A-site cation can expand, contract, or distort the lattice structure, leading to changes in bandgap. For example, in lead iodide (APbI₃) systems, increasing the size of the A-site cation (Cs⁺ < MA⁺ < FA⁺) leads to smaller bandgaps. However, this effect is typically less pronounced because the electronic structure of halide perovskites is primarily influenced by the B-site metal and X-site halide. , Changes at the B-site can alter the valence and conduction bands, as well as the octahedral factor (i.e., the ratio between the ionic radii of the B-site and X-site ions), which can influence the crystal structure and bandgap. For example, when replacing lead (Pb²⁺) with tin (Sn²⁺), the VBM and CBM shift upward, consistent with the smaller electronegativity of tin. The VBM shifts upward more than the CBM due to a larger shift in the B-site s level compared to the p level, leading to an overall reduction of the bandgap for tin-based perovskites compared to their lead analogs. However, Tao et al. reported that there are some exceptions to this trend (e.g., FASnBr₃, MASnCl₃, and FASnCl₃) where other effects like lattice distortions could play a larger role, leading to larger bandgaps for the tin-based compounds.

Overall, along with their varied structural compositions, beneficial defect tolerance, and tunable optical properties, emissive MHPs show promise for a plethora of applications due to their high color purity, high PLQY, direct bandgap, and potential for low-cost fabrication. For more details on the general characteristics of MHPs and further explanation of the fundamentals and mechanisms governing light emission in MHPs, the authors direct the reader to these other reviews. ,− , In the following section, we will introduce the specific strategies utilized to achieve V/UV emission by engineering the composition, crystal structure, and geometric structure of MHP and MHP derivative materials.

Strategies for Achieving Violet and Ultraviolet Emission

Enabled by the tunable nature of MHPs, various strategies targeting their composition and structure have been implemented to achieve wide-bandgap violet (defined here as having a luminescence emission peak between 400 and 435 nm) or UV (<400 nm) emission. Starting with the common cesium lead halide (CsPbX₃) material system, this review explores strategies to create material systems for V/UV emission by altering the composition of the halide (X-site), B-site, or A-site, which can induce changes in the crystal or geometric structure. We then further explore the dimensional control of the geometric structure through synthetic and processing parameters. The materials discussed are categorized by their crystal structure. Table and Figure summarize the ABX₃ perovskite materials with a 3D crystal structure, such as CsPbX₃ NCs. Within our dimensionality framework, these colloidal NCs exhibit a 3D crystal structure and a confined 0D geometric structure. While most references discussed refer to these materials broadly as NCs, they are also sometimes referred to as QDs because the NCs are confined in three dimensions and small enough to demonstrate quantum confinement effects, depending on the exciton Bohr radius of the material system. However, as discussed in the previous section, readers should be careful to note the requirement for quantum confinement for classification as a QD. For example, many of the CsPbCl₃ NCs reviewed here are not referred to as QDs because the excitonic Bohr radius for CsPbCl₃ (on the order of ∼5 nm) is smaller than the typical size of NCs. There is also growing interest in exploring lead-free or RD systems that break from the standard ABX₃ structure. Table and Figure summarize these perovskite or perovskite derivative materials (e.g., double perovskite or RD crystal structures) which represent a variety of crystal and geometric structures.

1. Optical Properties of V/UV Light-Emitting Lead Halide Perovskite NCs with the Standard 3D ABX₃ Crystal Structure .

Perovskite	PL Peak	fwhm	PLQY	Abs. Peak	reference
CsPbCl3	405 nm
CsPbCl3	404 nm
CsPbCl3	390 nm		1%
CsPbCl3	420 nm
CsPbCl3	405 nm	12 nm	10%
CsPbCl3	410 nm		3.8%
CsPbCl3 (Ce-doped)	410 nm, 430 nm		24.3%
CsPbCl3	408 nm	11 nm	65%
CsPbCl3 (Cd-doped)	406 nm	10–12 nm	98%	400 nm
CsPbCl3	402 nm	14 nm	50%
CsPbCl3 (Ni-doped)	406 nm		96.5%
CsPbCl3	404 nm	11 nm	60%	394 nm
CsPbCl3 (La/F-doped)	410 nm			397 nm
CsPbCl3	406.1 nm	10 nm	77.1%
CsPbCl3 (Cu-doped)	403 nm	14 nm	60%
CsPbCl3	405 nm		97.2%	402 nm
CsPbCl3	405 nm		50%
CsPbCl3 (Cu-doped)	403 nm	12 nm	12.3%
CsPbCl3 (Mg-doped)	405 nm		79%
CsPbCl3	410 nm	13 nm
CsPbCl3 (Zn-doped)	409 nm	10–15 nm	88.7%	407 nm
CsPbCl3	404 nm	12 nm	71%	402 nm
CsPbCl3	404 nm	9 nm	80%
CsPbCl3 (Mg-doped)	402 nm		75.8%	385 nm
CsPbCl3 (Cd-doped)	381 nm		60.5%
CsPbCl2Br	431 nm	12 nm	87%
CsPbCl3	406 nm	10 nm	70%
CsPbCl3 (Sr-doped)	410 nm	12 nm	82.4%
CsPb(Cl/Br)3	405 nm	24 nm
CsPbCl3	405 nm	10.6 nm	87%	402 nm
CsPbCl3	408 nm	11 nm	42.5%	393 nm
CsPbCl3 (Ce-doped)	411 nm		65%
CsPbCl3	409 nm
MAPbCl3	428 nm			400 nm
MAPbCl3	404 nm	15 nm	5%
MAPbCl3	397 nm	11 nm	3.3%
MAPbCl3	405 nm	9 nm		400 nm
FAPbCl3	400 nm	13 nm	1.2%
FAPbCl3	407 nm	16 nm	2%

^aValue estimated from figure.

^bEntries are left blank when values are not reported in a reference.

4.

Optical properties of lead-based V/UV light-emitting MHPs. (a) Absorption and PL spectra of CsPbCl₃ (350 nm excitation wavelength). Inset shows a photograph of CsPbCl₃ NCs under UV illumination. Adapted with permission under a Creative Commons CC-BY-NC-ND 4.0 license from ref Copyright 2021 American Chemical Society. (b) Absorption and PL spectra of undoped and Ni-doped CsPbCl₃ NCs. Insets show the corresponding photographs of NC solutions under UV illumination. Adapted with permission under a Standard ACS Author Choice/Editors’ Choice usage agreement from ref This is an unofficial adaptation of an article that appeared in an ACS publication. ACS has not endorsed the content of this adaptation or the context of its use. Copyright 2018 American Chemical Society. (c) PL spectra of CsPbCl₃ and CsPbCl₃:Cd²⁺ NCs with different doping content of Cd²⁺. Adapted with permission from ref Copyright 2021 John Wiley and Sons.

2. Optical Properties of Lead-Free or Reduced-Dimensional (RD) V/UV Light-Emitting MHP or Perovskite Derivative Materials That do Not Follow the Standard 3D ABX₃ Structure .

Perovskite	PL Peak	fwhm	PLQY	Abs. Peak	Crystal Structure	ref
Cs4CuIn2Cl12	381 nm	73 nm	1.7%	273 nm	2D
Rb2CuBr3	385 nm	54 nm			1D
Rb2CuCl3	395 nm	52 nm	100%		1D
Rb2CuBr3	385 nm		98.6%	300 nm	1D
Rb2CuCl3	397 nm	50.2 nm	99.4%	276 nm	1D
K2CuBr3	391 nm	61.4 nm	86.98%		1D
K2CuCl3	386 nm	52 nm	98.8%		1D
K2CuBr3	388 nm	54 nm	55%		1D
K2CuCl3	392 nm	54 nm	96.58%		1D
K2CuBr3	383 nm	49 nm	79.2%		1D
Cs3Bi2Br9	410 nm	48 nm	19.4%	396 nm	2D
Cs3Bi2Cl9	393 nm	59 nm	26.4%		1D
FA3Bi2Br9	437 nm	65 nm	52%	404 nm	2D
FA3Bi2Cl9	399 nm	63 nm	39%
MA3Bi2Br9	423 nm	62 nm	12%	376 nm	2D
MA3Bi2Cl9	360 nm	50 nm	15%		1D
MA3Bi2Br9	422 nm	62 nm	13.5%	376 nm	2D
MA3Bi2(Cl0.33Br0.67)9	422 nm	41 nm	54.1%	388 nm
MA3Bi2(Cl0.5Br0.5)9	412 nm		19.1%	366 nm
MA3Bi2(Cl0.67Br0.33)9	399 nm		22.4%	353 nm
MA3Bi2Cl9	370 nm		24.7%	333 nm	1D
MA3Bi2Br9	400 nm		50.1%	376 nm	2D
MA3Bi2Br6Cl3	379 nm		35.4%	355 nm
Cs3Sb2Br9	410 nm	41 nm	46%		2D
Cs3Sb2Br9	409 nm		51.2%	368 nm	2D
Cs3CeBr6	400 nm
Cs3CeBr7	410 nm
Cs3CeCl6	385 nm
Cs3CeCl7	375 nm
Cs3CeBr6	391 nm, 421 nm		91.2%		0D
Cs3CeI6	430 nm, 470 nm		71.4%		0D
Cs3CeBr2I4	410 nm, 450 nm				0D
Cs3CeI6	424 nm, 470 nm		74.6%		0D
Cs3CeCl6	380 nm, 409 nm		58.83%		0D
Cs3CeBr6	392 nm, 421 nm			357 nm	0D
Cs3CeCl6·3H2O	334 nm, 356 nm		93.54%	305 nm	1D
Cs3CeCl6	380 nm, 410 nm		91.82%	365 nm	0D
Cs3CeCl6·3H2O	373 nm, 406 nm		∼100%		1D
Rb3CeI6	427 nm, 468 nm		51%
(DFPD)4CeBr7	400 nm, 430 nm	30–40 nm	98%		0D
(DFPD)4CeCl7	375 nm, 400 nm	30–40 nm	∼100%		0D
(DFPD)CeCl4·2MeOH	353 nm, 368 nm	30–40 nm	95%		1D
Cs3MnBr5 (Ce-doped)	387 nm, 419 nm	22.5 nm, 30.5 nm
Cs2ZnCl4 (Ce-doped)	370 nm		∼100%	225 nm	0D
Cs2ZnBr4 (Ce-doped)	342 nm, 367 nm		97%		0D
Cs2NaCeCl6	370 nm, 410 nm		6%	340 nm	3D
Cs2NaPr0.99Ce0.01Cl6	265 nm, 276 nm, 302 nm, 370 nm, 410 nm				3D
Cs2NaPrCl6	265 nm, 276 nm, 302 nm		14%	245 nm	3D
PEA2PbBr4	405 nm				2D
PEA2PbCl4	350 nm				2D
PEA2PbBr4	407 nm		26%	404 nm	2D
PEA2PbBr4	410 nm		11%	400 nm	2D
PEA2PbBr4	408 nm	11.6 nm	15–25%		2D
PEA2PbBr4	437 nm			433 nm	2D
PEA2PbBr4	406 nm	8 nm	8%	402 nm	2D
PEA2PbBr4	407 nm				2D
PEA2PbCl1Br3	394 nm				2D
PEA2PbCl2Br2	384 nm				2D
PEA2PbCl1Br3	393 nm	14 nm	1%	382 nm	2D
PEA2PbCl1Br3	401 nm	16.82 nm		385 nm	2D
PEA2EuCl4	401 nm	29 nm	7.73%	350 nm	2D
MAPbBr3	436 nm	<25 nm			2D
BA2PbBr4	406 nm		26%		2D
BA2PbBr4	406 nm	15.9 nm		400.5 nm	2D
EA2CdBr4	388 nm	32 nm			2D
(5-MBI)PbBr3	430 nm		34.7%	297 nm	1D
[BAPrEDA]PbCl6·(H2O)2	392 nm	73 nm	21.3%	300 nm	0D
MA2CdBr4	415 nm	94 nm			0D
CdCl2-4HP	416 nm		63.55%		2D
Rb3InCl6 (Cu-doped)	398 nm	54 nm	95%		0D
T-PbBr2	423 nm	20 nm	16%	408 nm	2D
B-PbBr2	413 nm	19 nm	53%	398 nm	2D
Cs2AgIn0.9Bi0.1Cl6	390 nm		36.6%	367 nm	3D
KMgF3	145 nm, 165 nm
BaLiF3	160 nm, 180 nm, 225 nm

^aValue estimated from figure.

^bEntries are left blank when values are not reported in a reference.

5.

Optical properties of V/UV light-emitting bismuth halide, cerium halide, manganese halide, and 2D lead halide perovskites or perovskite derivatives. (a) Absorption and PL spectra (left) of MA₃Bi₂(Cl, Br)₉ perovskite QDs with different levels of chloride content. Photograph (right) of corresponding QD solutions under UV illumination. Adapted with permission from ref Copyright 2018 American Chemical Society. (b) PL spectra of Cs₃CeBr₆, Cs₃CeBr₂I₄, and Cs₃CeI₆. Insets show the corresponding photographs of cerium halide perovskite films under UV illumination. Adapted with permission under a Creative Commons CC BY-NC 4.0 license from ref Copyright 2022 The Authors. (c) Absorption (left) and PL spectra (right) of undoped and Ce-doped Cs₃MnBr₅ samples. Insets show the corresponding photograph of pristine and 0.2 mmol Ce-doped Cs₃MnBr₅ samples under daylight and under UV illumination. Note: inset text size has been increased for readability. Adapted with permission from ref Copyright 2023 Springer Nature. (d) PL spectra of PEA₂PbBr₄, PEA₂PbCl₁Br₃ and PEA₂PbCl₂Br₂. Inset shows the corresponding photographs of PEA₂PbX₄ (X = Br or Cl) thin films under UV illumination, from left to right: PEA₂PbBr₄, PEA₂PbCl₁Br₃, PEA₂PbCl₂Br₂, PEA₂PbCl₃Br₁, and PEA₂PbCl₄. Adapted with permission from ref Copyright 2023 SPIE, the international society for optics and photonics.

Halide Composition

Bandgap-tunable CsPbX₃ colloidal NCs have demonstrated emission across the visible spectrum (410 to 700 nm), with violet emission achieved by employing chloride at the X-site to form CsPbCl₃ NCs. A summary of representative studies to date can be found in Table , which details the optical properties, including PL and absorption when reported, of lead halide perovskite NCs with emission in the V/UV range. Early studies on CsPbX₃ NCs across the visible spectrum reported narrow PL emission line widths (10 to 40 nm) and high PLQYs as high as 95%, ,− but performance of CsPbCl₃ NCs with wide-bandgap emission in the V/UV range lagged behind that of red/green/blue CsPbI₃ or CsPbBr₃ emitters. Akkerman et al. reported that CsPbBr₃ NCs with a slight addition of chloride through an anion exchange reaction achieved a PLQY of 95% with emission around 495 nm, but further replacement of bromide anions with chloride yielded a PLQY as low as 1% with emission at 390 nm. Since then, considerable efforts have been made to improve the PLQY of violet-emitting CsPbCl₃ NCs, and various techniques such as doping of metals (e.g., Ni²⁺, Cd²⁺) into the crystal lattice structure have been utilized to demonstrate near-unity PLQYs in CsPbCl₃ NCs. ,,

As shown in Figure a, Zhang et al. measured CsPbCl₃ NCs (undoped and synthesized using a PhPOCl₂ precursor) with a sharp excitonic absorption peak around 402 nm and narrow PL emission around 404 nm (fwhm of 12 nm). This relatively small Stokes shift (∼15 meV) suggests that emitted photons come from exciton recombination. The direct radiative emission of photons is further supported by other studies on CsPbCl₃ NCs that demonstrate that photoluminescence excitation (PLE) measurements align closely with the absorption spectra. Note that changes in the synthesis or composition, such as through doping at the B-site discussed in the next section, can alter the optical properties of the CsPbCl₃ NCs, with reported PL emission peaks as low as 381 nm (Cd²⁺-doped CsPbCl₃ NCs) to as high as 430 nm (Ce³⁺-doped CsPbCl₃ NCs) in different studies.

Similar to the effects observed for CsPbX₃, employing smaller halides also blue-shifts the PL emission in some of the other metal halides summarized in Table . In 2D phenethylammonium lead halide (PEA₂Pb­(Br/Cl)₄) thin films, we showed that the PL emission can be blue-shifted from 407 nm (PEA₂PbBr₄) to 394 nm (PEA₂PbCl₁Br₃) to 384 nm (PEA₂PbCl₂Br₂) by increasing the amount of chloride relative to bromide. Increasing the relative amount of chloride at the X-site causes a blue-shift in PL across a variety of other metal halide systems as well, such as MAPbX₃, , FAPbX₃, , MA₃Bi₂X₉, ,, FA₃Bi₂X₉, Cs₃CeX₆, , and (DFPD)₄CeX₇, which are discussed in more detail in the following sections.

Unlike most of the materials discussed, replacing bromide with chloride can lead to a red-shift in PL in some lead-free material systems in Table . For example, copper halide systems such as Rb₂CuCl₃ and K₂CuBr₃ demonstrate shorter wavelength PL for the bromide compounds compared to chloride compounds. These copper halides have 1D crystal structures. For example, K₂CuBr₃ is made up of 1D chains made up of corner sharing [CuBr₃]^2– tetrahedra separated by K⁺ cations, and the same type of structure is seen for the Rb-based copper halides as well. Creason et al. reported Rb₂CuBr₃ with PL centered at 385 nm and Rb₂CuCl₃ with PL centered at 395 nm, and then in a later study, Creason et al. reported K₂CuBr₃ with PL centered at 388 nm and K₂CuCl₃ with PL centered at 392 nm. Similar studies from other groups on Rb₂CuX₃ and K₂CuX₃ systems show the same trends. − The demonstrated red-shift in emission when replacing bromide with chloride, contrary to the trends for most other MHP systems, can be attributed to a combination of changes in the bandgap associated with structural deformation and a strong excitonic effect arising from self-trapped exciton (STE) states in these materials.

B-Site Engineering

Lead is commonly employed at the B-site cation for MHP optoelectronic devices, ,− but there is increasing interest to partially or fully replace the lead in MHPs due to lead toxicity concerns and limited performance of lead systems in the V/UV range. While lead systems demonstrate high performance in the red/green/blue emission ranges, − moving to wider bandgap emission requires further defect passivation, dimensional control/confinement, or higher energy electronic state transitions that can be accessed by utilizing other metals.

One successful strategy to increase the PLQY within the MHP semiconductor class, broadly, has been metal doping at the B-site using metals such as Ni²⁺, Cd²⁺, Cu⁺/Cu²⁺, Zn²⁺, Mg²⁺, Sr²⁺, or Mn²⁺. ,,,,,,,,, Table shows a variety of different studies in CsPbCl₃ NCs employing different doping strategies to achieve high PLQY V/UV emission. As shown in Figure b, Yong et al. utilized Ni²⁺ doping via a modified hot-injection method to replace Pb²⁺ with Ni²⁺ at the B-site, which improved the short-range order of the lattice and eliminated halide vacancy defects, improving the PLQY from 2.4% to 96.5%. Mondal et al. demonstrated a room temperature postsynthetic treatment of weakly emitting CsPbCl₃ NCs with CdCl₂, which improved the PLQY from 3% to 96% (and to 98% when starting with initial samples with a higher PLQY of 32%). The postsynthetic treatment led to facile exchange of Pb²⁺ by Cd²⁺ at the B-site without changing the size and structure of the lattice. The authors showed that this postsynthetic treatment suppressed nonradiative carrier trapping centers and introduced shallow energy levels facilitating radiative recombination, leading to the enhanced PLQY without affecting the PL peak position (406 nm) and spectral width. In a different study where Cd²⁺ was incorporated into the NCs during synthesis as opposed to a postsynthetic treatment, Zhang et al. synthesized Cd-doped CsPbCl₃ NCs that demonstrated a blue-shift in emission with increasing Cd²⁺ content as shown in Figure c. The authors state that this blue-shift in emission is likely due to an increase in bandgap caused by lattice contraction from the smaller radius of Cd²⁺ compared to Pb²⁺.

Along with partially replacing Pb²⁺ through doping the B-site in CsPbCl₃ NCs, making entirely lead-free systems with other metals such as copper (Cu⁺, Cu²⁺), bismuth (Bi³⁺), or antimony (Sb³⁺) has proven successful for pushing emission deeper into the UV range as shown in Table . Liu et al. reported lead-free Cs₄CuIn₂Cl₁₂ double perovskite NCs with 381 nm emission and a PLQY of 1.7%. Copper has also been used in the K₂CuX₃ and Rb₂CuX₃ systems discussed in the previous section. , Bi³⁺ is an appealing candidate to replace Pb²⁺ because it is less toxic, isoelectronic to Pb²⁺, and more stable than other commonly explored alternatives like Sn²⁺. Leng et al. reported one of the first MA₃Bi₂X₉ perovskite QDs, achieving 360 nm emission with a 15% PLQY for MA₃Bi₂Cl₉ and 423 nm emission with a 12% PLQY for MA₃Bi₂Br₉. In another study, Leng et al. further improved upon these QDs by utilizing a Cl-passivation process that greatly improved the PLQY, achieving emission centered at 370 nm with a 24.7% PLQY for MA₃Bi₂Cl₉ and centered at 422 nm with a 54.1% PLQY for Cl-passivated MA₃Bi₂Br₉ (33% Cl). Note that even after the addition of 33% Cl content to the MA₃Bi₂Br₉ QD, the PL peak position remains unchanged, although the absorption peak red-shifted from 376 to 388 nm as shown in Figure a. The authors attribute the unchanged PL peak to reduced surface defects after Cl introduction, combined with the effect of size increase. However, after a further increase of Cl content past 33%, the authors observed the expected trends of a blue-shift of the PL (and the absorption peak). Wu et al. produced MA₃Bi₂Br₉ QDs with an even lower PL emission wavelength of 400 nm via a cryo-bonding ligand-assisted reprecipitation (Cb-LARP) method, achieving a PLQY of 50.1%. The authors then demonstrated tunability of the halide to further push emission into the UV range with MA₃Bi₂Br₆Cl₃ QDs emitting at 379 nm with a PLQY of 35.4%. As discussed further in the next section on A-site engineering, the A-site cation can also be altered to create FA₃Bi₂X₉ or Cs₃Bi₂X₉ perovskite QDs. , Similar to Bi³⁺, antimony (Sb³⁺) forms a Cs₃Sb₂Br₉ perovskite structure and is another alternative to Pb²⁺ at the B-site. Zhang et al. demonstrated violet-emitting Cs₃Sb₂Br₉ perovskite QDs with PL centered at 410 nm and a PLQY of 46%, and Ma et al. similarly demonstrated PL centered at 409 nm and a PLQY of 51.2%. While Cs₃Sb₂Br₉ QDs have not been heavily studied, the study by Ma et al. is one of the few lead-free demonstrations of violet EL (408 nm) in a PeLED, which will be discussed further in the section titled, “State-of-the-Art Violet and Ultraviolet PeLED Performance”.

The other lead-free material system that has been used to demonstrate V/UV EL in PeLEDs incorporates the rare-earth element cerium (Ce³⁺), which has bright violet emission from the Ce-5d to Ce-4f transition. Wang et al. produced thermally evaporated Cs₃CeBr₆ thin films in which [CeBr₆]^3– octahedra are isolated from each other by Cs⁺ cations in a 0D crystal structure. Due to this 0D structure and the localized f-orbitals in Ce³⁺, there is small band dispersion, which increases the exciton binding energy and enables high PLQYs. The films exhibited a doublet PL emission peak at 391 and 421 nm and a PLQY of 91.2%. In this study, Wang et al. also reported the first demonstration of V/UV EL from a Cs₃CeBr₆-based PeLED. Similar to lead halide systems, the Cs₃CeX₆ material system has demonstrated shifts in emission with halide tunability. Figure b shows the PL spectra of Cs₃CeBr₆ with doublet emission at 391 and 421 nm, Cs₃CeBr₂I₄ with doublet emission around 410 and 450 nm, and Cs₃CeI₆ with doublet emission around 430 and 470 nm. The emission blue-shifts with increasing bromide content, and studies on cerium chloride systems have demonstrated emission even further into the UV range. ,,,

Cerium-based perovskite derivatives synthesized using solution processing also enable V/UV emission. Aiming to create a simple solution-based method that is cost-effective and environmentally friendly, Dutta et al. used a water-based synthesis to produce spin-coated Cs₃CeBr₆ thin films, reporting doublet PL emission at 392 and 421 nm and also showing EL with a corresponding PeLED. Wang et al. developed an antisolvent vapor-assisted crystallization approach for preparing (DFPD)₄CeCl₇ and (DFPD)­CeCl₄·2MeOH single crystals. The (DFPD)₄CeCl₇ exhibited doublet PL emission peaks at 375 and 400 nm with a PLQY approaching 100% and (DFPD)­CeCl₄·2MeOH exhibited doublet PL emission peaks at 353 and 368 nm with a PLQY of about 95%. They also reported (DFPD)₄CeBr₇ with PL emission at 400 and 430 nm and a PLQY of 98%, demonstrating the tunability of emission through the halide site.

Cerium has also been used as a dopant to achieve violet emission, both in the CsPbCl₃ NCs discussed previously , and in other lead-free material systems. For example, Dutta et al. reported a water-assisted synthesis of Ce³⁺-doped Cs₃MnBr₅ powder that exhibited doublet PL at 387 and 419 nm. As shown in Figure c, the undoped Cs₃MnBr₅ has no emission in the V/UV range, but when the Mn²⁺ at the B-site is doped with Ce³⁺, the characteristic doublet emission from Ce³⁺ appears. Cerium doping was also utilized in studies by Wen et al. and Zhu et al. that incorporated Ce³⁺ into Cs₂ZnCl₄ and Cs₂ZnBr₄ hosts, respectively, to achieve UV emission centered around 370 nm with near-unity PLQY. ,

Along with cerium, the electronic transitions in the rare-earth element praseodymium (Pr³⁺) are conducive to wide-bandgap emission, enabling even shorter wavelength UV emission. In a study on lanthanide double perovskite NCs such as Cs₂NaPrCl₆ and Cs₂NaCeCl₆, Saghy et al. showed emission in the UV–C range from Cs₂NaPrCl₆ NCs with peaks at 265, 276, and 302 nm. In comparison, Cs₂NaCeCl₆ NCs exhibited peaks at 370 and 410 nm. These emission peaks in the Cs₂NaPrCl₆ NCs correspond to different electronic transitions in praseodymium, and because emission in the UV–C range is rare for perovskite emitters, this highlights the potential for future studies on praseodymium-based metal halide emitters.

A-Site Engineering

While CsPbCl₃ NCs have been the most widely researched lead halide perovskites for blue-violet emission, the A-site Cs⁺ can be replaced by another monovalent cation such as FA⁺ or MA⁺, changing the properties of the material. Recall that increasing the size of the A-site cation (Cs⁺ < MA⁺ < FA⁺) typically leads to smaller bandgaps. Imran et al. demonstrated results that follow this trend for both iodide and bromide systems: CsPbI₃, MAPbI₃, and FAPbI₃ NCs exhibited emission peaks at 691, 730, and 764 nm, and CsPbBr₃, MAPbBr₃, and FAPbBr₃ NCs exhibited emission peaks at 512, 527, and 531 nm. Das and Samanta also report similar trends for APbI₃ and APbBr₃ NCs. Interestingly, the trend breaks for APbCl₃ NCs in both studies (see Table ), as they show CsPbCl₃, MAPbCl₃, and FAPbCl₃ NCs with emission peaks at 408, 404, and 407 nm, respectively (Imran et al.), and 406, 397, and 400 nm, respectively (Das and Samanta). The reasons for this difference are not explicitly explored in these studies, but the results demonstrate that simply changing cation size does not always induce the expected bandgap trends, as there are a variety of competing factors influencing the structural and optical properties of a material. It should also be noted that the PLQY varies drastically between different cation choices, particularly for the lead chloride NCs. For example, Imran et al. reported a PLQY of 65% for CsPbCl₃ NCs but PLQYs of only 5% and 2% for MAPbCl₃ and FAPbCl₃, respectively, and similar numbers are reported by Das and Samanta (70% versus 3.3% and 1.2%).

Different A-site cations have been utilized in other lead-free, V/UV-emitting MHP systems as shown in Table . For the copper halide systems discussed in the previous sections, rubidium (Rb⁺) and potassium (K⁺) were used in place of cesium (Cs⁺) to achieve violet emission. Cs-based copper halides like Cs₃Cu₂X₅ do not emit in the violet range, with the lowest achievable wavelength of emission centered around 445 nm for Cs₃Cu₂I₅ (recall that larger halides lead to lower emission wavelengths, unlike most other MHPs). But when Cs⁺ is replaced by Rb⁺ or K⁺, the materials with the smaller A-site cations (K⁺ < Rb⁺ < Cs⁺) form different perovskite structures that emit in the violet range, with emission wavelengths reported as short as 383 nm for K₂CuBr₃.

For Bi-based QDs, Leng et al. reported all-inorganic Cs₃Bi₂Cl₉ and Cs₃Bi₂Br₉ QDs emitting at 393 and 410 nm, respectively. In a separate study, Leng et al. also demonstrated MA₃Bi₂Cl₉ and MA₃Bi₂Br₉ QDs emitting at 360 and 423 nm, respectively. Shen et al. showed FA₃Bi₂Cl₉ and FA₃Bi₂Br₉ QDs emitting at 399 and 437 nm, respectively. Unlike the clear trend between halide size (Cl^– < Br^– < I^–) and emission peak, there is no clear trend between A-site cation size (Cs⁺ < MA⁺ < FA⁺) and emission peak, suggesting that additional factors influence the bandgap in these Bi-based QD systems. Similar to the lead halide systems, tuning the A-site, B-site, or X-site can all change the bandgap, but the trends may differ depending on the complementary atoms at the other sites and the dimensionality of the material system.

Dimensional Control

Utilizing quantum confinement through RD geometric structures is another strategy to achieve wide-bandgap violet emission. Utilizing a bulky organic cation such as PEA⁺ at the A-site creates a 2D layered Ruddlesden–Popper perovskite structure (e.g., PEA₂PbX₄). Within our dimensionality framework, these materials have a 2D crystal structure because they consist of perovskite octahedra layers sandwiched between PEA layers. They are also 2D in their geometric structure because they form nanoplatelets, but the morphology can change slightly depending on how the material is synthesized. , In PEA₂PbBr₄, the 2D crystal structure forms repeating quantum wells, tightly confining excitons with high binding energies and leading to narrow, violet PL centered at 407 nm that was used to create one of the first violet PeLEDs (EL centered at 410 nm) by Liang et al. in 2016. Multiple other studies on PEA-based lead halide systems listed in Table have demonstrated the promise of using RD systems to push emission into the UV range, ,,,,,− with PL demonstrated as short as 384 nm for PEA₂PbCl₂Br₂ as shown in Figure d. Lead-free systems are also attractive for achieving deeper UV emission, and we demonstrated 401 nm emission for 2D PEA₂EuCl₄ nanoplatelets. Furthermore, other A-sites and B-sites have been used in studies on 2D MHPs and other RD structures with emission into the UV range such as ligand-confined MAPbBr₃, BA₂PbBr₄, EA₂CdBr₄, and more shown toward the end of Table .

Beyond nanoplatelets, 0D geometric structures like QDs with sizes on the order of the excitonic Bohr radius are common for V/UV emission because the additional quantization and confinement enables higher energy emission. For example, Leng et al. showed that a MA₃Bi₂Br₉ single crystal (3D geometric structure) exhibited broad PL emission with a peak at 550 nm and fwhm of around 100 nm. Then, when synthesizing MA₃Bi₂Br₉ QDs (0D), they observed that the PL emission peak shifted to 423 nm with a fwhm of 62 nm, demonstrating a significant blue-shift in emission and a sharper peak compared to the bulk 3D structure.

Overall, the full list of materials in Tables and demonstrate the current state-of-the-art in V/UV light-emitting perovskite and perovskite derivative materials. Starting with the common CsPbX₃ material system, violet emission can be achieved by employing chloride at the X-site to produce CsPbCl₃ NCs. Then, compositional engineering can either alter the CsPbCl₃ structure through doping or by fully replacing the A-, B-, or X-site atoms to form new crystal structures, which are typically lower-dimensional. Importantly, to move into the UV range, we can utilize a wide range of lead-free systems employing B-site metals such as Cu⁺, Bi³⁺, Sb³⁺, Ce³⁺, Mn²⁺, Zn²⁺, Pr³⁺, Eu²⁺, Cd²⁺, and In³⁺. Lastly, dimensional control of the geometric structure through synthetic and processing parameters can form structures like nanoplatelets or QDs, in which quantum confinement effects enable higher energy emission. Additional studies on these materials and exploration of new materials can improve the current state-of-the-art and push emission further into the UV range.

Device Architecture and Engineering

While the optical properties of MHPs are highly tunable, achieving efficient device operation relies heavily on the ability to inject, transport, and confine charges effectively. Thus, PeLEDs typically employ a structure in which the perovskite or perovskite derivative emissive layer is sandwiched between two wide-bandgap transport layers: a hole transport layer (HTL) and an electron transport layer (ETL). ,− These transport layers facilitate charge injection and confine carriers within the emissive region. , They are broadly categorized as organic or inorganic, with inorganic materials generally preferred due to their higher chemical and thermal stability within V/UV devices. − This is particularly critical for V/UV PeLEDs, which operate at comparatively higher voltages, because resulting Joule heating accelerates structural degradation and induces EQE roll-off, making thermally stable transport layers essential for mitigating device deterioration. , Common inorganic and organic transport layers used in PeLEDs, including their respective energy levels, are shown in Figure a. For efficient PeLED operation, the transport layer energy levels must align with the energy levels of adjacent layers to inject charge carriers into the emissive perovskite layer and confine them within the perovskite to induce radiative recombination. Specifically, the ETL should possess a deep valence band to prevent hole leakage and nonradiative recombination, while the HTL should have a shallow conduction band to prevent electron leakage. Consequently, the operating voltage and efficiency of PeLEDs depend on the energy-level offsets, carrier mobilities, layer thicknesses, and interface quality of each layer. ,,−

6.

Direct charge injection-based device architectures for V/UV PeLEDs. (a) Summary of energy level and electron/hole mobilities of widely used ETLs and HTLs for PeLEDs. Note: material name text size has been increased for readability. Adapted with permission from ref Copyright 2021 IOP Publishing Ltd. (b) Device structure schematic and (c) corresponding energy band diagram of Cs₃CeBr₆ V/UV PeLEDs. (d) Normalized EL spectra of Cs₃CeBr₆ PeLEDs at different applied voltages. Note: the x-axis was changed to “Wavelength (nm)” from “Voltage (V)”. Insets show the (top left) photograph of a working Cs₃CeBr₆ PeLED and (bottom left) EL spectra of the PeLED operated at different voltages. Note: the x-axis of the bottom left inset was changed to “Wavelength (nm)” from “Voltage (V)”. Adapted with permission from ref Copyright 2021 American Chemical Society. Energy band diagrams for PEA₂PbCl₁Br₃ UV PeLEDs using (e) a single ETL (TPBi) device structure or (f) a dual ETL (TmPyPB/TPBi) device structure. (g) EL spectra of the TPBi and TmPyPB/TPBi devices show pure 399 nm UV emission from the dual ETL device structure. Adapted with permission from ref Copyright 2024 The Authors.

One persistent challenge that hampers the performance of V/UV PeLEDs is the inefficient injection and transport of charge carriers from the electrodes into the perovskite emissive layer. This issue commonly arises due to (i) energy-level misalignment between the transport and emissive layers, and (ii) imbalance in the rate of electron and hole injection. Improper band alignment at either interface can cause carrier accumulation and exciton quenching, significantly lowering EQE. Moreover, when one carrier type is injected more efficiently than the other, charge imbalance occurs, leading to accumulation of excess carriers within the transport layers. This imbalance also promotes nonradiative pathways such as interfacial exciton quenching and Auger recombination, thereby suppressing EQE.

Energy-level misalignment becomes even more challenging in deep-UV PeLEDs because the wide-bandgap perovskites used in these devices typically possess deeper valence bands than their visible emitting counterparts, which increases the hole-injection barrier for conventional organic hole-transport materials. As such, cascaded interlayers with stepwise energy-level alignment may be needed to facilitate charge injection. However, if these added layers introduce extra resistance, they can increase the turn-on voltage and promote Joule heating, potentially compromising device stability. − Thus, alternative approaches are needed to improve hole injection. One potential strategy that could be explored further is the use of transition-metal oxides (TMOs), such as molybdenum oxide, tungsten oxide, etc., as hole-injection layers. These materials are attractive because they possess high work functions (∼6–7 eV), strong hole selectivity, and wide bandgaps that limit parasitic UV absorption relative to organic transport layers. , Meyer et al. showed that for OLEDs, MoO₃ can enable interfacial charge generation through electron transfer from the HOMO of the adjacent organic layer into the low-lying CBM of the oxide, thereby leaving behind holes in the organic layer and facilitating hole injection. A related interfacial charge-transfer mechanism could potentially be exploited in deep-UV PeLEDs, where TMOs may help lower the effective hole-injection barrier. However, although TMOs are promising for deep-UV hole injection, most of the current V/UV PeLED studies surveyed in this review still utilize organic HTLs (e.g., PEDOT:PSS). Further research is required to evaluate TMOs as HTLs for V/UV PeLEDs. In particular, the TMO and MHP interface must be investigated and engineered carefully because interfacial redox chemistry and defect formation can degrade performance and stability. ,

Furthermore, a practical route to mitigate charge imbalance in PeLEDs is to adjust the thickness of transport layers or insert additional interlayers to achieve balanced injection rates at the perovskite interfaces. For optimal device performance, the ETL/HTL thickness must strike a balancethin enough to facilitate transport, yet thick enough to effectively block opposite carriers. As seen in Figure b,c, Wang et al. introduced a 1 nm Al₂O₃ interlayer between the ETL and the perovskite emissive layer which helped suppress exciton quenching. They also utilized HAT-CN as an electron-blocking/hole-injecting layer and TCTA as a hole-transport interlayer to further improve hole injection and confinement. The introduction of these layers suppressed carrier quenching and yielded an improved V/UV PeLED with a doublet emission at 391 and 421 nm (Figure d) and maximum EQE of 0.46%. In our previous work, we demonstrated that adding an additional ETL (Figure e,f) improved charge injection by fine-tuning the energy-level alignment between the perovskite and electrode, leading to a singular EL peak at 399 nm (Figure g) and maximum EQE of 0.16%. Similar strategies targeting transport layer engineering have been employed across V/UV PeLEDs studies in the literature as reviewed further in the next section.

State-of-the-Art Violet and Ultraviolet PeLED Performance

To overview the status of V/UV PeLEDs, we categorize the previous reports as reduced-dimensional (RD) MHPs, 3D MHPs, or rare-earth MHPs. We focus our comparison of these different PeLED studies on their reported EQEs and EL peak wavelengths. In the PeLED field, device performance often includes photometric quantities such as luminance (cd m^–2) that are based on the photopic luminosity function which describes the wavelength-dependent sensitivity of the human eye. However, this function decreases sharply in the V/UV range, so it is best practice for studies to report radiometric quantities that do not depend on the sensitivity of the human eye such as radiance (W sr^–1 m^–2), EQE, or spectral power distribution (and hence EL peak wavelength). The definition for EQE and EL peak wavelength is consistent across studies and is thus used as the primary metric to summarize PeLED performance in this review.

Focusing first on RD MHPs, we examine the reports related to layered perovskites (e.g., reports where alternating organic (e.g., bulky ligand) and inorganic layers (e.g., metal halide octahedra) are present). By synthesizing phenethylammonium lead bromide (PEA₂PbBr₄) nanoplatelets coupled with a solvent vapor annealing process, Liang et al. demonstrated 410 nm violet PeLEDs with an EQE of 0.04%. To improve device performance, they employed transport layer engineering by optimizing the thickness of their TPBi ETL and consequently observed a more symmetric carrier flux and improved device EQE. Similarly, we altered the solvent annealing procedure with water additives and incorporated dual electron transport layers to fabricate violet 408 nm PEA₂PbBr₄ and UV 399 nm PEA₂PbCl₁Br₃ (Figure a,b) PeLEDs with champion EQEs of 0.41% and 0.16%, respectively. , By optimizing the HTL through varying the concentration of Poly-TPD, Ni et al. demonstrated tunable EL between 394 and 406 nm in PEA₂PbBr₃Cl PeLEDs and achieved EQEs as high as 2.41%. The different Poly-TPD concentrations altered the film morphology and surface roughness of the perovskite layer, with the smoothest film (RMS ≈ 5.7 nm) exhibiting the highest EQE of 2.41% at an emission wavelength of 401 nm. These RD MHPs can also be fabricated from other deposition techniques, and using an electric-field-deposition technique, Deng et al. showcased 408.8 nm PEA₂PbBr₄ PeLEDs with an EQE of 0.31%.

7.

Reduced-dimensional (RD) V/UV PeLEDs. (a) Current density–voltage−radiance characteristics of the champion PEA₂PbCl₁Br₃ UV PeLED. Inset shows the device structure of the dual ETL PEA₂PbCl₁Br₃ UV PeLED. (b) EQE−current density characteristics of the champion PEA₂PbCl₁Br₃ UV PeLED. Adapted with permission from ref Copyright 2024 The Authors. (c) Current density–voltage–luminance characteristics of the champion Cs₃Sb₂Br₉ violet PeLED. Inset shows a photograph of the PeLED operated at 7.0 V. (d) EQE−voltage characteristics of the champion Cs₃Sb₂Br₉ violet PeLED. Inset shows the device structure of the Cs₃Sb₂Br₉ violet PeLED. Adapted with permission from ref Copyright 2019 American Chemical Society.

In the V/UV range, utilizing bulky PEA ligands has been the most common strategy to form RD MHPs, but there have also been some reports on other types of RD MHP materials for V/UV PeLEDs. Sun demonstrated centimeter scale 2D BA₂PbBr₄ (BA = butylammonium) PeLEDs with an EL peak of 406 nm and EQE of 0.083%. In order to quantum confine MAPbBr₃ MHPs, Kumar et al. incorporated oleic acid and octylamine into the synthesis procedure which then led to 2D MAPbBr₃-based PeLEDs with EL peaks as short as 432 nm with a corresponding EQE of 0.004%. Moving away from lead, violet PeLEDs with emission at 408 nm were made from layered lead-free Cs₃Sb₂Br₉ quantum dots fabricated by Ma et al. (Figure c,d). To improve charge injection, they incorporated poly­(ethylenimine) (PEI) into the ETL to lower the cathode work function, facilitating improved electron injection and leading to a significant increase in EQE up to 0.206%. Lastly, the fabrication of a nonlayered 1D (5-MBI)­PbBr₃ (5-MBI = 5-methylbenzimidazole) PeLEDs by Cheng et al. led to a 430 nm EL peak with and EQE of 0.009%. Thus, RD PeLEDs have successfully enabled the development of V/UV PeLEDs, but more work is needed to improve device performance, including maximum EQEs and tunability into the UV spectrum.

Next, we summarize the developments of 3D MHP V/UV PeLEDs. Sadhanala et al. employed halide engineering within the MAPb­(Br _x Cl_3–x ) system which resulted in a 425 nm MAPbCl₃ PeLED but low corresponding EQE that could not be measured. Along the same lines, 408.8 nm MAPbCl₃ PeLEDs were fabricated using both spin-coating and thermal evaporation of perovskite precursors by Sun et al. with an EQE of 0.036%. Moving onto inorganic 3D MHPs, Bai et al. thermally evaporated CsPbCl₃ PeLEDs with EL at 410 nm. Similarly, Zhang et al. thermally evaporated CsPbCl₃ double-layer films (i.e., one unannealed CsPbCl₃ film and one annealed CsPbCl₃ film) which led to 412 nm PeLEDs. By incorporating an atypical Mg_0.71Zn_0.29O electron transport layer with thermally evaporated CsPbCl₃ emitters, Zhang et al. demonstrated 409 nm PeLEDs. Alternatively, Zhang et al. synthesized CsPbCl₃ nanocrystals to be integrated into violet 405 nm PeLEDs with an EQE of 0.18%. In many of these studies, the EQEs of devices are either low or not reported, motivating further work to improve the emissive properties of the MHP materials through strategies like doping. By introducing Mg²⁺ dopants into CsPbCl₃ nanocrystals, Hu et al. achieved 402 nm PeLEDs with an EQE of 0.1%. Analogously, Wang et al. doped Ce³⁺ ions into CsPbCl_3–x Br _x nanocrystals and demonstrated 414 nm PeLEDs with an EQE of 0.84% (Figure a,b). While both 3D and RD MHPs have demonstrated the capability to move toward and into the UV spectrum, the majority of these PeLEDs are dominated by lead halides, which raise significant toxicity concerns. To make this technology more commercially viable, alternative MHP material classes should be considered.

8.

3D and rare-earth metal halide V/UV PeLEDs. (a) Current density–voltage–luminance characteristics of both the control and modified (i.e., champion) Ce³⁺-doped CsPbCl₃ violet PeLEDs. Inset shows a photograph of a representative PeLED under operation. (b) EQE−voltage characteristics of both the control and modified (i.e., champion) Ce³⁺-doped CsPbCl₃ violet PeLEDs. Inset shows the device structure of the Ce³⁺-doped CsPbCl₃ violet PeLEDs. Adapted with permission from ref Copyright 2024 John Wiley and Sons. (c) Current density–voltage−radiance characteristics of the champion Cs₃CeBr₆ V/UV PeLED. (d) EQE−voltage characteristics of the champion Cs₃CeBr₆ V/UV PeLED. Inset shows the device structure of the Cs₃CeBr₆ V/UV PeLED. Adapted with permission from ref Copyright 2021 American Chemical Society.

Rare-earth metal halide PeLEDs are particularly attractive given their potential to address toxicity concerns from lead halide PeLEDs. Specifically, Ce³⁺ has strongly localized f-orbitals which paired with the allowed 5d → 4f electronic transitions make Ce³⁺-based halide perovskites a strong candidate for high-energy light emission. Additionally, the 5d → 4f excited-state lifetime for Ce³⁺ is short (17 ns in CeBr₃), which could suggest enhanced luminous efficiency and yield reduced LED degradation. Cerium is also known to be abundant and nontoxic, as the Earth’s cerium crust abundance is 0.006 wt %, comparable to that of lead (0.010 wt %). Regarding cerium halide PeLEDs, Yang et al. fabricated thermally evaporated Cs₃CeI₆ PeLEDs with a doublet EL peak at approximately 430 and 470 nm and EQE of 7.9%. By replacing cesium with rubidium, Du et al. fabricated thermally evaporated Rb₃CeI₆ PeLEDs with a doublet EL peak at 427 and 468 nm with an EQE of 0.67%, showcasing the effects of A-site engineering within cerium halide PeLEDs. Moving toward bromide-based perovskites, Guo et al. demonstrated thermally evaporated Cs₃CeBr _x I_6–x PeLEDs with tunable doublet EL peaks as short as 411 nm with a corresponding EQE of 0.93% for Cs₃CeBr₂I₄. Fully eliminating iodide, Wang et al. fabricated thermally evaporated Cs₃CeBr₆ PeLEDs with a doublet EL peak at 391 and 421 nm (Figure c,d). As discussed in the prior section, careful design of the ETL and HTL layers in this study enabled V/UV PeLEDs with an EQE of 0.46%, a relatively high value for this wavelength range. However, device performance could likely be improved even further with additional balancing of charge injection. As Wang et al. discuss in their paper, further improvements in hole injection and transport are needed given the higher electron mobility compared to hole mobility in Cs₃CeBr₆. Moving away from thermal evaporation, Dutta et al. spin coated Cs₃CeBr₆ thin films which resulted in PeLEDs with a doublet EL peak at 392 and 421 nm and champion EQE of 0.44%. In this study, they found that mismatched carrier mobilities caused exciton recombination at the transport layer/perovskite interface. They resolved this by inserting a thin PVK layer acting as both an electron-blocking and hole-injection layer, coupled with SPPO13 as the electron-injection layer and a 5 nm LiF interlayer to minimize quenching at the cathode side. Thus, similar to the thermally evaporated PeLED study by Wang et al., this solution-processed PeLED study by Dutta et al. demonstrates the effectiveness of interfacial modification to improve the balance of carriers in PeLEDs. Beyond the Cs₃CeBr₆ considered in these studies, employing chloride at the halide site could enable even deeper UV emission. Sun et al. spin coated Cs₃CeCl₆·3H₂O thin films which resulted in UV PeLEDs with a singular EL peak at 380 nm and champion EQE of 0.13%. Thus, rare-earth metal halide PeLEDs show great potential to enable nontoxic and efficient UV PeLEDs. Looking forward, additional strategies, such as halide composition modification toward Cl-based rare-earth metal halide PeLEDs and opportunities for dimensionality control should be explored to push deeper into the UV spectrum. Table summarizes the V/UV RD, 3D, and rare-earth metal halide PeLEDs discussed here while Figure illustrates the EQE performance of V/UV RD, 3D, and rare-earth metal halide PeLEDs across EL peak wavelength.

3. Summary of V/UV Metal Halide PeLEDs.

Perovskite	EL Peak	EQE	Type	Fabrication	reference
PEA2PbBr4	410 nm	0.04%	RD	Spin-coating
PEA2PbBr4	408 nm	0.41%	RD	Spin-coating
PEA2PbBr3Cl	399 nm	0.16%	RD	Spin-coating
PEA2PbBr3Cl	406 nm	1.63%	RD	Spin-coating
PEA2PbBr3Cl	401 nm	2.41%	RD	Spin-coating
PEA2PbBr3Cl	394 nm	1.96%	RD	Spin-coating
PEA2PbBr3Cl	394 nm	0.90%	RD	Spin-coating
PEA2PbBr4	408.8 nm	0.31%	RD	Electric-field-deposition
BA2PbBr4	406 nm	0.083%	RD	Spin-coating
MAPbBr3	432 nm	0.004%	RD	Spin-coating
Cs3Sb2Br9	408 nm	0.206%	RD	Spin-coating
(5-MBI)PbBr3	430 nm	0.009%	RD	Spin-coating
MAPbCl3	427 nm	N/A	3D	Spin-coating
MAPbCl3	408.8 nm	0.036%	3D	Spin-coating/Thermal Evaporation
CsPbCl3	410 nm	N/A	3D	Thermal Evaporation
CsPbCl3	412 nm	N/A	3D	Thermal Evaporation
CsPbCl3	409 nm	N/A	3D	Thermal Evaporation
CsPbCl3	405 nm	0.18%	3D	Spin-coating
Mg2+: CsPbCl3	402 nm	0.1%	3D	Spin-coating
Ce3+: CsPbCl3	414 nm	0.84%	3D	Spin-coating
Cs3CeI6	∼430 & ∼470 nm	7.9%	Rare-earth	Thermal Evaporation
Cs3CeI6	430 and 470 nm	3.5%	Rare-earth	Thermal Evaporation
Cs3CeBr2I4	411 nm and unreported	0.93%	Rare-earth	Thermal Evaporation
Cs3CeBr6	391 and 421 nm	0.46%	Rare-earth	Thermal Evaporation
Rb3CeI6	427 and 468 nm	0.67%	Rare-earth	Thermal Evaporation
Cs3CeBr6	392 and 421 nm	0.44%	Rare-earth	Spin-coating
Cs3CeCl6·3H2O	380 nm	0.13%	Rare-earth	Spin-coating

9.

Summary of V/UV PeLED EQE performance across EL peak wavelength. Data points are sorted by MHP type (i.e., reduced-dimensional (RD), 3D, and rare-earth metal halide perovskites) as well as single or double EL peaks. All PeLEDs shown are plotted according to their shortest (i.e., primary) EL peak since most of the rare-earth metal halide PeLEDs result in a double EL emission peak. Illustrated from Table . Each point is annotated with its corresponding underlined reference number.

Challenges and Limitations of Violet and Ultraviolet PeLEDs

Efficiency Roll-Off

Most PeLEDs exhibit a phenomenon known as efficiency roll-off where the EQE decreases as the current density increases past a certain threshold. V/UV PeLEDs are not immune to this effect, as evidenced from Figures and . While the causes of this effect can vary, it has been attributed to Auger recombination, , unbalanced charge carrier injection, , and Joule heating. , Given the comparatively lower PLQYs and wider bandgaps of the emissive V/UV perovskite layer as well as the limited optimized transport layers, efficiency roll-off must be carefully addressed as it pertains to V/UV PeLEDs.

To address detrimental Auger effects as well as Joule heating, the use of phosphonic acid modifiers for hole charge carrier injection (e.g., 2PACz) provides a viable alternative to traditional hole transport layers (HTLs). For instance, Shen et al. showed that 2PACz and t-Bu-2PACz can limit both harmful Auger and Joule heating effects in green perovskite QD LEDs. Compared to a typical control device with a PEDOT:PSS/PVK HTL, they found that the devices with 2PACz-based HTLs had a lower turn-on voltage. Although the nominal energy level diagram of PEDOT:PSS/PVK and the perovskite QDs suggests that the control devices should have more facile charge injection, Shen et al. hypothesize that current is limited by barriers other than the HTL/QD interface such as the interface between PEDOT:PSS and PVK. Thus, they state that charge injection imbalance in the control devices leads to more charge accumulation (and thus more Joule heating) compared to the 2PACz-based PeLEDs. Furthermore, by performing current density-dependent PL measurements on both 2PACz-based and control PeLEDs, they showed that Auger effects and/or Joule heating are greatly suppressed for 2PACz-based PeLEDs. As a result, the 2PACz-based PeLEDs reach maximum luminances as high as 373,000 cd m^–2 while prolonging the onset of efficiency roll-off at a 1000-fold higher current density as compared to control PeLEDs. Additionally, other 2PACz derivatives can yield low-voltage operation in PeLEDs, which could greatly enhance the operation of V/UV PeLEDs considering their inherently higher voltage operation. Wang et al. showed that integrating Br-2PACz into green 3D PeLEDs yields turn-on voltages as low as 1.9 V. The authors argue that Br-2PACz facilitates hole charge carrier injection as well as passivating interfacial defects between the perovskite and ITO layers. Lastly, the incorporation of 2PACz derivatives in V/UV PeLEDs can address unbalanced charge carrier injection because they induce a lower VBM energy level ,− (e.g., between −5.5 and −6.0 eV for Br-2PACz) as compared to traditional HTLs (e.g., −5.1 eV for PEDOT:PSS). This lower VBM is better aligned with the wide bandgap UV perovskite layer’s VBM, and thus these materials hold the potential to enhance hole charge carrier injection. Overall, 2PACz and its derivatives deserve further investigation within V/UV PeLEDs given their potential to address Auger effects, Joule heating, and unbalanced charge carrier injection.

For Joule heating effects, it is important to note the generally low thermal conductivity and diffusivity of many MHPs, so several techniques have been shown to enhance thermal management during PeLED operation. For example, Li et al. showed that coupling red light-emitting perovskite nanocrystals treated with diphenylphosphoryl azide (DPPA) and employing thermally conductive sapphire substrates greatly decreased Joule heating effects and efficiency roll-off. Specifically, DPPA increases charge injection into the nanocrystals by modifying the nanocrystals’ coordination with insulating carbon-chain ligands while replacing glass substrates with sapphire (K = 46 W m^–1 K^–1) yields effective heat dissipation, as evidenced by a 28% decrease in device operation temperature. Furthermore, Zhao et al. incorporated doped charge transport layers, additional heat spreaders and sinks, and optimized device geometries to address Joule heating within infrared PeLEDs. First, by doping both charge transport layers, efficiency roll-off was reduced while achieving EQEs similar to undoped PeLEDs. Specifically, POPy₂ (ETL) was n-doped with [RuCp*Mes]₂ while Poly-TPD (HTL) was p-doped with F₄-TCNQ. Next, integrating graphite (K = 1300 W m^–1 K^–1) or polycrystalline diamond heat spreaders (K > 2000 W m^–1 K^–1) with copper heat sinks also significantly reduced efficiency roll-off, especially at current densities above 500 mA cm^–2 where Joule heating can induce rapid roll-off in PeLEDs. Lastly, by adjusting the PeLED device area to a line-shape geometry (i.e., 4 μm × 1 mm), heat dissipation was improved, yielding additional reductions to efficiency roll-off. Altogether, addressing Joule heating effects within the emissive layer and corresponding PeLED device stack layers can lessen efficiency roll-off and should be carefully considered for V/UV PeLEDs.

One of the largest obstacles limiting the efficiency in V/UV PeLEDs is balanced charge carrier injection. Many of the conventional charge transport layers used within PeLEDs (Figure a) are not optimized for wide-bandgap MHPs capable of V/UV light emission. In particular, the reliance on PEDOT:PSS for hole injection is problematic for V/UV PeLEDs given its misaligned energy levels. While phosphonic acid modifiers modifiers provide a promising solution to circumvent this obstacle, modifications to PEDOT:PSS itself can also improve hole charge carrier transport. For instance, Lu et al. fabricated an ultrathin (6.9 nm) PEDOT:PSS HTL using a water stripping method and consequently achieved EQE enhancements in 3D, quasi-3D, and quasi-2D PeLEDs. Notably, this ultrathin PEDOT:PSS HTL achieves a deeper energy level that is better aligned with V/UV MHPs. Introducing additives into PEDOT:PSS can also achieve similar effects. Incorporating additives like sodium polystyrenesulfonate and guanidinium iodide within PEDOT:PSS improved charge transport within PeLEDs and perovskite solar cells, respectively, while also achieving lower energy levels as compared to pure PEDOT:PSS. While modifications to PEDOT:PSS could yield more efficient V/UV PeLEDs, broader exploration of charge transport layers that can support efficient injection into V/UV PeLEDs is needed.

Operational Stability

PeLED operational lifetime is arguably the most important obstacle to overcome to ensure their commercialization. Within visible PeLEDs, green PeLEDs have achieved half lifetimes (T₅₀) as high as 520 h with an initial brightness of 1000 cd m^–2, while deep-blue PeLED stability has demonstrated a T₅₀ as high as 59 h at an initial brightness of 100 cd m^–2. Unfortunately, the operational stability of V/UV PeLEDs lags behind deep-blue PeLEDs. In Figure a,b, PEA₂PbCl₁Br₃ UV PeLEDs achieve a T₅₀ of 0.43 s while inorganic Ce-doped CsPbCl₃ violet PeLEDs demonstrate a T₅₀ of 240 s. From Figure c,d, lead-free violet PeLEDs based on Cs₃CeI₆ and Cs₃Sb₂Br₉ emitters achieved a T₅₀ and T₉₀ of 45 min and 6 h, respectively. While the superior operational stability displayed by the lead-free V/UV MHPs as compared to their lead-based counterparts is promising from a toxicity perspective, significant improvements are still needed to realize this promising technology. The poor operational stability for MHPs, including V/UV MHPs, likely stems from ion migration effects, poor material stability, and lack of defect passivation.

10.

Operational stability of V/UV PeLEDs. (a) Normalized EL evolution in time of a PEA₂PbCl₁Br₃ UV PeLED operated at a constant current bias of 10 mA/cm². The resulting half-lifetime is 0.43 s. Adapted with permission from ref Copyright 2024 The Authors. (b) Normalized EL evolution in time of both the control and modified (i.e., champion) Ce³⁺-doped CsPbCl₃ violet PeLEDs operated at a constant voltage bias of 6 V. The resulting half-lifetime of the modified (i.e., champion) PeLED is 240 s. Adapted with permission from ref Copyright 2024 John Wiley and Sons. (c) Normalized EL evolution in time of both the Si₃N₄-based (i.e., unoptimized electron blocking layer) and Al₂O₃-based (i.e., optimized electron blocking layer and resulting champion PeLED) Cs₃CeI₆ violet PeLEDs operated at a constant voltage bias of 5 V. The resulting half-lifetimes of the Si₃N₄-based and Al₂O₃-based Cs₃CeI₆ violet PeLEDs are ∼18 and ∼45 min, respectively. Adapted with permission under a Creative Commons CC BY-NC 4.0 license from ref Copyright 2022 The Authors. (d) Normalized EL evolution in time of a Cs₃Sb₂Br₉ violet PeLED operated at a constant voltage bias of 7 V. The resulting T₉₀ lifetime (i.e., intensity decay of 10%) is 6 h. Adapted with permission from ref Copyright 2019 American Chemical Society.

Given the mixed electronic-ionic nature of MHPs and low activation energies of the constituent ionic species, ion migration significantly contributes to the poor operational stability of PeLEDs. During operation, the voltage bias imposed on the PeLED can induce a large electric field within the emissive MHP layer and cause ions to redistribute themselves within the device stack. One of the more common consequences of this effect is the migration of halide anions toward the metal electrodes because the activation energy of halide anions tends to be lower than other constituent ions within the MHP. Given the higher voltage biases required to operate V/UV PeLEDs, ion migration effects are likely enhanced and thus must be carefully considered in order to improve the operational stability within this energy regime. As mentioned previously, B-site doping can improve the PLQY of MHPs, but they can also address harmful ion migration effects in PeLEDs. Futscher et al. demonstrated how Mn²⁺ doping can decrease ion migration effects within PeLEDs and consequently improve operational stability due to a twofold increase in the activation energy of mobile halide anions. Strategies to limit ion migration, including those related to compositional engineering and defect passivation, should be further explored within V/UV PeLEDs.

Compositional engineering, particularly at the A-site, has been demonstrated to enhance the operational stability of PeLEDs. For example, Cao et al. incorporated 5-aminovaleric acid (5AVA), a large cation introduced at the A-site, into red FAPbI₃ PeLEDs to prolong their lifetime to a T₅₀ of ∼20 h operated at a current density of 100 mA cm^–2. Additionally, Jiang et al. developed a rubidium–cesium alloyed deep blue PeLED and reported an increase in the T₅₀ lifetime from 5.2 to 14.5 min as a result of rubidium incorporation at the A-site. Moreover, while organic A-site cations, like MA⁺, can yield high PeLED EQEs, their poor thermal stability can also lead to reduced stability as a result of operation-induced degradation mechanisms. − Thus, moving toward inorganic MHPs could improve the operational stability of V/UV PeLEDs. Lastly, Figure suggests that lead-free MHPs could lead to better operational stability within V/UV PeLEDs and should be further investigated.

Defect Passivation

Effective defect passivation can improve the emissive properties of MHP materials and boost the operational stability of PeLEDs. As discussed in earlier sections, wide-bandgap chloride-based perovskites can suffer from deeper defects within the bandgap compared to their bromide- or iodide-based counterparts. Thus, compositional engineering strategies such as B-site doping have been employed to suppress chloride vacancy defects and produce high PLQY violet emitters like Ni²⁺-doped CsPbCl₃ NCs. Alongside compositional engineering, realizing efficient and operationally stable V/UV emission requires advanced defect passivation strategies to neutralize deep trap states, suppress ion migration, and withstand extreme photochemical environments. Key chemical approaches involve Lewis acid–base coordination, particularly utilizing phosphonic acids and phosphine oxides; ,− these agents offer optimized orbital alignment to heal undercoordinated lead clusters and chloride vacancies without suffering from the rapid, ultraviolet-induced deprotonation that severely degrades traditional ammonium-based ligands. For example, Zhang et al. reported synthesis of CsPbCl₃ NCs using phenylphosphonic dichloride (PhPOCl₂) as a chloride precursor. In their surface chemistry investigations, they determined that this method promoted a surface with a PbCl _x -rich shell capped with the nonprotonated L-type capping ligand oleylamine (OLA) with strong covalent binding. As a result, trap states are less likely to form on the surface of these PhPOCl₂-derived NCs compared to other common synthesis approaches that form a halide-rich surface with protonated ligands that are prone to desorption from the NC surface, leading to chloride vacancy defects. Phosphine oxide additives have also improved PeLED stability, particularly within thermally evaporated PeLEDs. Given the comparatively higher use of thermal evaporation for depositing V/UV MHPs as compared to visible light-emitting MHPs, this additive class system is particularly well-poised for V/UV PeLEDs but has not yet been extensively studied in the V/UV range. Thus, we can look to PeLED studies in other wavelength ranges for potential techniques to apply in the V/UV. For instance, Li et al. showed that coevaporating triphenylphosphine oxide (TPPO) with CsBr and PbBr₂ precursors improved the T₅₀ lifetime of green thermally evaporated PeLEDs from 25 to 78 min with an initial brightness of 1600 cd m^–2. This is largely attributed to the surface passivation of CsPbBr₃ via TPPO. Similarly, thermally evaporating tris­(trifluoromethyl)­phosphine oxide (TFPPO) with quasi-2D MHP precursors yielded a green PeLED with an enhanced T₅₀ lifetime of 125 min, as compared to 34 min for non-TFPPO PeLEDs. However, our previous work and similar investigations have shown that phosphine oxide additives can also lead to operational stability impairments. Thus, further investigation is warranted before universal incorporation of phosphine oxides within PeLED technologies, including those in the V/UV range.

Additionally, zwitterionic molecules − provide cooperative, dual-action passivation by simultaneously binding to cationic lead sites and anionic halide vacancies via coordinate and hydrogen bonding, which drastically suppresses nonradiative recombination and locks the fluid ionic lattice. For example, Guo et al. fabricated near-infrared FAPbI₃ PeLEDs and incorporated sulfobetaine 10 (SFB10) additives to yield record stability T₅₀ lifetimes of 11,539 and 32,675 h operated at current densities of ∼5.0 mA cm^–2 and ∼3.2 mA cm^–2, respectively. These authors attribute the long device lifetime to SFB10s ability to bind with the FA⁺, Pb²⁺, and I^– ions at the perovskite grain boundaries, which in turn provides surface passivation of the MHP and mitigates ion migration of Pb²⁺ and I^– ions. In another study, Zhang et al. utilize zwitterion 3-aminopropanesulfonic acid (APS) as an additive to simultaneously passivate deep and shallow level defects in FAPbI₃ via coordinate and hydrogen bonding. The defect passivation led to suppressed nonradiative recombination and ion migration and enabled near-infrared PeLEDs with a peak EQE of 19.2% and a half-lifetime of 43 h. Furthermore, inorganic engineering using alkali metals, halides, and pseudohalidessuch as synergistic treatments with KSCN and KCl, or dual-vacancy passivation via LiClworks to physically fill vacant lattice sites, modulate crystallization kinetics, and sterically hinder detrimental ion migration. Finally, integrating these molecular and ionic treatments with two-dimensional heterostructure engineering or robust zwitterion-functionalized polymeric matrices creates targeted energy funnels and physical barriers that substantially improve both the EQE and the long-term operational stability of V/UV PeLEDs.

Toxicity and Scalability

As V/UV PeLEDs are integrated into applications that interface with humans, such as those related to public health and sanitization, addressing toxicity concerns from these materials is critical. So far, lead-free V/UV PeLEDs have largely been limited to either Sb³⁺- or Ce³⁺-based emitters, while Table showcases lead-free V/UV MHP emitters employing Cu⁺, Bi³⁺, Sb³⁺, Ce³⁺, Mn²⁺, Zn²⁺, Pr³⁺, Eu²⁺, Cd²⁺, and In³⁺ halide perovskites. This discrepancy between the limited lead-free V/UV PeLEDs demonstrations and the numerous previously studied emitters suggests that more investigations are needed to integrate these materials into functional PeLEDs. Some of the obstacles related to the integration of these materials within PeLEDs include mismatched energy level alignment leading to poor charge injection, nonuniform thin film formation, and low thin film PLQY. Nevertheless, many of the strategies outlined in this section are equipped to address these obstacles and should help lead to an expanded lead-free V/UV PeLED class. Furthermore, to push deeper into the UV spectrum, both Pr³⁺- and Gd³⁺-based halide perovskites deserve additional investigations. Specifically, Gd³⁺ demonstrates strong PL emission at 313 nm due to the spin-allowed ⁶P_7/2 → ⁸S_7/2 transition. , Similarly, Pr³⁺ displays three major PL peaks at 265, 276, and 302 nm due to the electronic transitions from the 4f¹5d¹ excited state to 4f² levels. Overall, there are substantial opportunities to study lead-free V/UV PeLEDs and venture toward even shorter wavelength emission.

Additionally, the pursuit of industrially compatible fabrication methods to promote the scalability of V/UV PeLEDs should be prioritized. While laboratory-scale spin-coating techniques have dominated MHP deposition within PeLEDs, thermally evaporated PeLEDs are quickly progressing as a promising alternative to increase their scalability. As previously noted, this also extends to V/UV PeLEDs. There are several scalability advantages that thermally evaporated PeLEDs have compared to spin-coated PeLEDs. First, since many commercial organic LEDs (OLEDs) utilize thermal evaporation in their production lines, transitioning PeLEDs to thermal evaporation can reduce capital investment. Moreover, insights from thermally evaporated OLEDs could be directly applied to PeLEDs in industrial settings. Second, the use of quartz crystal microbalances as well as clean and consistent vacuum environments can help ensure the repeatability and reliability of thermally evaporated PeLEDsincluding those with larger device areas as compared to spin-coated PeLEDs. Lastly, since thermal evaporation can process many types of substrates (e.g., glass, silicon, flexible plastics, etc.), there are substantial opportunities for heterogeneous integration of thermally evaporated PeLEDs. While spin-coated PeLEDs may be largely limited to laboratory-scale demonstrations, other solution-processing deposition techniques should be explored to access the strong potential of solution-processed V/UV MHPs. For example, blade-coating, a mass-fabrication solution-processing technique, has been used to fabricate large-area PeLEDs within the visible range. , Chu et al. blade-coated red MAPbI₃ PeLEDs and achieved a large device area of 28 cm² with uniform EL emission across the entire device area, showcasing the possibilities of scalable large-area PeLEDs with solution-processing techniques. Altogether, both thermal evaporation and mass-fabrication solution-processing techniques are promising deposition processes that could improve the scalability of V/UV PeLEDs.

Emerging Trends and Future Directions

Novel Material Systems

As the field moves to address the toxicity of lead and the challenges of accessing deep-UV emission, lead-free double perovskites (A₂B′B″X₆) have emerged as a leading solution. Specifically, sodium-lanthanide double perovskite nanocrystals (Cs₂NaLnCl₆) represent a versatile platform for wide-bandgap optoelectronics (Figure a). Unlike lead-halide systems, which rely on band-edge emission, these materials utilize the localized intraconfigurational f → f or d → f transitions of trivalent lanthanide ions (Ln³⁺) to achieve precise spectral control. Recent work has demonstrated that by varying the lanthanide dopant, emission can be tuned across the entire UV–vis–NIR spectrum, including the difficult-to-access UV–C range; for instance, praseodymium-based Cs₂NaPrCl₆ nanocrystals have shown sharp emission peaks at 265 and 276 nm with PLQYs reaching 14%, while cerium-based Cs₂NaCeCl₆ enables efficient UV−A emission at 370 nm. These materials benefit from large bandgaps (>5 eV) that suppress nonradiative recombination and high environmental stability due to their all-inorganic nature.

11.

Emerging materials and device considerations for V/UV PeLEDs. (a) Absorption, PL, and PL excitation spectra of Cs₂NaLnCl₆ (Ln = Pr, Ce, Tb, Sm, Eu, and Yb) lanthanide double perovskite nanocrystals. Adapted with permission from ref Copyright 2023 John Wiley and Sons. (b) PL spectra of various 2D hybrid MHPs. The emissive 2D sheets are as follows: (i) (C₄H₉NH₃)₂PbCl₄, (ii) (C₄H₉NH₃)₂PbBr₄, (iii) (C₄H₉NH₃)₂PbI₄, (iv) (C₄H₉NH₃)₂PbCl₂Br₂, (v) (C₄H₉NH₃)₂PbBr₂I₂, and (vi) (C₄H₉NH₃)₂(CH₃NH₃)­Pb₂Br₇. Inset shows the corresponding optical PL images of the six 2D hybrid MHPs. Scale bars are 2 μm for (i) to (v) and 10 μm for (vi). Adapted with permission from ref Copyright 2015 American Association for the Advancement of Science. (c) (left) Device structure with an assembled carbon nanotube (CNT) network on SiO₂/Si and patterned source contact. The device operates by applying AC voltage between the metal grid on the CNT network (source) and Si backgate (gate). (right) Uniform EL is achieved with various types of emitters including Ru­(bpy)₃(PF₆)₂ (red emitter), CdSSe/ZnS alloyed quantum dots (520 nm, green emitter), and PFO (blue emitter). Devices are measured with a 15 V 100 kHz square wave. Scale bar is 0.1 mm. Adapted with permission from ref Copyright 2020 Springer Nature. (d) Schematic diagram of encapsulation significance in thin film optoelectronic technologies. Adapted with permission from ref Copyright 2021 John Wiley and Sons.

Concurrently, the exploration of low-dimensional perovskites (2D and 0D) continues to offer a robust strategy for enhancing UV emission efficiency (Figure b). By reducing the dimensionality of the perovskite lattice to atomically thin 2D sheets or isolated 0D octahedra, dielectric and quantum confinement effects are dramatically amplified. This results in substantially increased exciton binding energies, which are critical for maintaining high radiative recombination rates at room temperature in wide-bandgap materials. While early work focused on 2D lead-halide sheets like PEA₂PbBr₄, current trends are converging toward low-dimensional lead-free systems. The Cs₂NaLnCl₆ double perovskites described above are structurally 3D but electronically 0D, with optically active [LnCl₆]³⁻ octahedra isolated from one another by intervening Cs⁺ ions and [NaCl₆]⁵⁻ octahedra. This zero-dimensional isolation minimizes charge carrier diffusion to defect sites, providing a “defect-tolerant” lattice that can sustain high PLQY without the need for the rigorous surface passivation required in 3D nanocrystals. Future research lies in optimizing the charge transport layers for these highly confined, often insulating, low-dimensional materials to translate their exceptional photophysical properties into efficient EL.

Encapsulation

For V/UV PeLEDs to transition from the lab to real-world applications, advanced encapsulation is required. As depicted in Figure d, the degradation pathways for these devices are multifaceted, involving not just moisture and oxygen intrusion, but also mechanical damage from environmental stressors and the potential leakage of toxic decomposition products (such as lead or organic volatiles). Traditional encapsulation methods often fail to address all these vectors simultaneously.

Emerging strategies are therefore moving beyond simple epoxy barriers to multifunctional architectures that integrate mechanical robustness with chemical inertness. These next-generation encapsulants employ graded layers or hybrid organic–inorganic structures that provide mechanical rebound protection against physical stress while simultaneously serving as a chemical barrier.

Crucially, for perovskite devices, the encapsulation must also function “internally” to prevent the escape of volatile decomposition gases and toxic lead leakage, ensuring both device longevity and environmental safety. Developing UV-transparent, hydrophobic, and mechanically robust encapsulation materialscapable of withstanding the higher photon energies of V/UV light without degradingwill be critical for the deployment of PeLEDs in harsh environments.

Alternative Device Architectures

As mentioned previously, the development of efficient V/UV PeLEDs is fundamentally hindered by the dual challenge of injecting both electrons and holes into wide-bandgap perovskites. As the bandgap widens to enable UV emission (>3.0 eV), the band edges shift significantly: the VBM drops to very deep energy levels, while the CBM moves closer to the vacuum level. This creates severe energy mismatches with standard charge transport materials. Finding stable HTLs with sufficiently deep ionization potentials and ETLs with appropriate electron affinities to match these V/UV perovskites simultaneously is a critical bottleneck. Consequently, V/UV devices often suffer from unbalanced carrier injection or require complex multilayer architectures.

To bypass these injection constraints entirely, alternating current electroluminescence (AC-EL) offers a compelling alternative architecture. As demonstrated in recent work by Zhao et al. (Figure c), AC-driven devices operate on a field-induced carrier generation mechanism rather than continuous charge injection from external electrodes. In this architecture, a gate voltage creates transient, steep band bending at the interface between a source contact (e.g., a carbon nanotube network) and the emissive layer. This field-induced bending thins the injection barrier sufficiently to allow bipolar carrier tunneling directly into the emitter, regardless of the intrinsic band alignment of either the valence or conduction bands. During one-half-cycle of the AC drive, electrons tunnel in. During the next, holes tunnel in, facilitating radiative recombination without the need for perfectly aligned ETLs or HTLs. This mechanism enables EL across the entire spectrumfrom infrared (0.13 eV) to ultraviolet (3.3 eV)using a single, generic device structure. For V/UV perovskites, this is particularly advantageous as it eliminates the need to identify exotic transport materials. Furthermore, AC operation inherently mitigates the detrimental effects of ion migrationa phenomenon exacerbated by constant DC electric fields in halide perovskites. By periodically reversing the bias, ion accumulation at interfaces is minimized, potentially extending the operational lifetime of V/UV devices significantly while simplifying the device stack.

Another effective approach to enhance carrier injection and suppress interfacial recombination is embedding insulating interlayers at the perovskite interfaces. Shi et al. demonstrated this strategy by incorporating LiF on both sides of their perovskite layer and analyzed this structure with two other structures, Device A (ITO/PEDOT:PSS/perovskite/Bphen/LiF/Al), Device B (ITO/PEDOT:PSS/PVK/perovskite/Bphen/LiF/Al), and Device C (ITO/LiF/perovskite/LiF/Bphen/LiF/Al), all emitting in the green spectral regime. Device C exhibited markedly higher EQEs5.53%, 2.36%, and 2.99% for FAPbBr₃, MAPbBr₃, and CsPbBr₃, respectivelycompared to Devices A and B, which showed significantly lower values. This improvement stems from reduced leakage currents, as evidenced by distinct percolation pathways (P, PE, P′, and PE′) in Devices A and C (Figure a). In Device A, hole injection was hindered due to a large interfacial barrier, whereas in Device C, tunneling through the insulating layers enabled improved injection while simultaneously blocking opposite-carrier leakage (Figure b). Time-resolved PL further confirmed that these insulating layers mitigate interfacial quenching. However, the thicknesses of these insulating layers should be such that the carriers tunnel through but do not leak out of the perovskite layer into the transport layers. Luo et al. showed this by analyzing different thicknesses of LiF as the insulating layers on both sides of a deep-blue CsEuBr₃ PeLED. They observed that the tunneling current was highly dependent on LiF thickness. Optimal performance was achieved using a 10 nm LiF layer on the hole-injection side and a 15 nm layer on the electron-injection side, resulting in a maximum EQE of approximately 6.5%. In addition to enhancing charge balance, the insulator-perovskite-insulator (IPI) structure can enhance device operational stability by limiting the migration of ions from the perovskite layers into the transport layers. As shown in Figure c, the perovskite layer was sandwiched between thin LiF layers in a fully inorganic device, resulting in green EL independent of voltage bias, consequently achieving a maximum brightness of 156,155 cd m^–2, an EQE of 11.05%, and half-lifetime of about 255 h at an initial luminance of 120 cd m^–2. Thus, given the difficulty of finding suitable ETL/HTL pairs that can yield efficient V/UV PeLEDs, IPI holds the potential to circumvent this obstacle by taking advantage of tunneling instead of charge injection.

12.

Alternative insulator-based LED device architectures. (a) The working mechanisms of (top) device A (ITO/PEDOT:PSS/perovskite/Bphen/LiF/Al) and (bottom) device C (ITO/LiF/perovskite/LiF/Bphen/LiF/Al; insulator-perovskite-insulator (IPI) device structure). (b) Energy band diagram analyses for device A (left) and device C (middle and right). (left) PE region (current density through the crystal grains) in device A has a thicker energy barrier for hole injection. (middle) PE′ region (current density through the crystal grains) in device C has a thinner energy barrier for hole injection, where (right) P′ region (current density through the gaps) in device C has a thick LiF layer that prevents tunneling of charge carriers and effectively prevents leakage current. Adapted with permission from ref Copyright 2018 John Wiley and Sons. (c) Device structure schematic of a fully inorganic IPI PeLED. Adapted with permission from ref Copyright 2020 John Wiley and Sons. (d) Device structure schematic for TFEL devices. (e) EL mechanism of TFEL devices: (1) tunnel injection, (2) carrier acceleration, (3) luminescence center excitation, (4) radiative recombination and light emission. Adapted with permission from ref Copyright 2022 American Chemical Society.

Incorporating thin-film electroluminescent (TFEL) phosphor centers between insulating layers (Figure d) is another device-engineering strategy that could help fabricate V/UV PeLEDs. TFEL phosphors are mainly synthesized by integrating lanthanide ions into inorganic host materials, such as perovskites and other inorganic lanthanide compounds (ILC), which circumvents the solubility problems of lanthanide-based MHPs in commonly used solvents. This structure has been studied extensively within the LED community. , Figure e demonstrates that when a sufficiently high electric field is applied, electrons tunnel through into the emissive layer and are subsequently accelerated, leading to impact excitation of lanthanide ions within the phosphor centers. These excited lanthanide ions then relax radiatively, resulting in light emission. , Van Haecke et al. demonstrated this mechanism using a red-emitting Ca_0.5Sr_0.5S/Eu TFEL phosphor, reporting a luminance of 80 cd m^–2 at 40 V above the threshold voltage with an optimal Eu doping concentration of 0.1 mol %. They observed that at higher Eu²⁺ concentrations, Eu–Eu interactions became stronger, increasing the density of shallow trap states and quenching both the PL and EL of their devices. Although TFEL devices can achieve reasonably high luminance, their high turn-on voltages, large charge-injection barriers, and comparatively low efficiencies have largely limited their use in modern optoelectronic applications.

Conclusion and Outlook

The rapid ascent of MHPs has redefined the landscape of solution-processable optoelectronics, yet the extension of this success into the V/UV spectral regimes remains a formidable frontier. As detailed in this review, the challenges governing V/UV PeLEDs are fundamentally distinct from their visible counterparts, necessitating a departure from standard composition and device engineering strategies. To bridge the performance gap between current V/UV PeLEDs and commercial requirements, the field must navigate a strategic evolution in both materials science and device architecture (Figure ).

13.

Evolutionary pathway of MHP materials and corresponding device architectures toward high-performance UV light emission.

Material Evolution: From Band-Edge to Activator-Based Emission

While 3D bulk lead halide perovskites (e.g., CsPbCl₃) have provided a foundational understanding of violet emission, their utility diminishes rapidly as the target wavelength shifts from UV−A into the UV–B and UV–C regions. The ”bandgap engineering” approachrelying solely on chloride-bromide alloyingfaces intrinsic limitations regarding phase segregation and defect intolerance. The evolutionary pathway forward points decisively toward dimensional reduction and rare-earth incorporation. Low-dimensional (2D and quasi-2D) structures utilize quantum confinement to sharpen emission line widths and enhance exciton binding energies, critical for overcoming thermal quenching. However, to access the deep UV (UV–C), the field is shifting toward lead-free double perovskites (e.g., Cs₂NaLnCl₆) or 0D perovskites (e.g., Cs₃LnCl₆) utilizing lanthanides (e.g., Ce³⁺, Pr³⁺, Gd³⁺). By exploiting the localized f–f or d−f transitions of these activators, rather than relying on band-edge recombination, researchers can bypass the defect-sensitivity of the host lattice and achieve precise spectral control down to 265 nm.

Device Architecture: Overcoming the Injection Bottleneck

High-efficiency UV EL is currently limited by charge injection, an inherent property of the standard DC p-i-n architecture. As the perovskite bandgap widens toward the UV–C region, the valence-band maximum deepens significantly. Identifying stable hole transport layers that align energetically with these deep levels remains elusive. Consequently, future engineering must pivot toward architectures that fundamentally circumvent this energetic impedance. Promising alternatives include light-emitting transistors and AC-EL configurations. AC-EL structures sandwich the emitter between insulators, utilizing field-induced carrier generation rather than direct charge injection. This effectively eliminates the need for ohmic alignment with transport layers. Similarly, field-emission concepts utilize high electric fields to achieve ballistic electron injection directly into the conduction band. However, while these devices bypass current roadblocks, they introduce new challenges regarding operating voltage, high-field stability, and fabrication complexity. Therefore, the field urgently requires architectures designed specifically for the harsh realities of UV operation. Future research must explore radical departures from the standard visible-LED stack to withstand UV degradation and achieve the paradigm shift required for commercialization.

Stability and Scalability: The Path to Commercialization

The transition from lab to industry requires overcoming stability and scalability challenges. UV PeLEDs suffer a unique degradation from high-energy photons, accelerating the decomposition of organic cations and transport layers. This necessitates developing UV-stable, inorganic charge transport layers and robust encapsulation techniques to mitigate ion migration and toxic materials leakage. Furthermore, scalable deposition methods like thermal evaporation or blade-coating are essential for integration into large-area applications such as photolithography and sterilization arrays.

While V/UV PeLEDs are less mature than red, green, and blue devices, the fundamental roadblocks are identifiable and surmountable. By synergizing low-dimensional material design, rare-earth doping, and novel injection architectures, MHPs are positioned to disrupt the UV optoelectronics market. They offer a versatile, low-cost alternative to III–V semiconductors and mercury lamps for critical applications in sensing, nanofabrication, public health, and more.

Glossary

Vocabulary

Metal Halide Perovskites (MHPs)

A broad semiconductor class primarily characterized by the formula for 3D ABX₃ materials where A is a monovalent cation (e.g., Cs⁺ and MA⁺), B is a divalent metal cation (e.g., Pb²⁺ and Eu²⁺), and X is a halide anion (e.g., I^–, Br^–, and Cl^–). Derivations from this ABX₃ crystal structure that employ these ionic octahedra, resulting in 2D, 1D, and 0D crystal structures are classified as metal halide perovskite derivatives

Violet Light-Emitting Perovskite

A metal halide perovskite that emits light between 400 and 435 nm

Ultraviolet Light-Emitting Perovskite

A metal halide perovskite that emits light below 400 nm

Photoluminescence Quantum Yield (PLQY)

The ratio of the number of photons emitted to the number of photons absorbed by a material. A metric to quantify the efficiency of photoluminescence

Light-Emitting Diode (LED)

A semiconductor device that emits light when electric current flows through the device. Light is produced by electroluminescence via electron–hole recombination

External Quantum Efficiency (EQE)

The ratio of the number of photons emitted from the LED to the number of electrons injected into the LED. A metric to quantify the efficiency of electroluminescence

†.

S.F., M.H. and T.K.C. contributed equally.

The authors acknowledge support from a Beckman Center Technology Development Grant.

This manuscript has been posted as a preprint on arXiv: Sebastian Fernández;* Manchen Hu;* Tyler K. Colenbrander;* Divine Mbachu; Daniel N. Congreve. Metal Halide Perovskites for Violet and Ultraviolet Light Emission. 2026, 2601.21270. arXiv. https://arxiv.org/abs/2601.21270 (accessed May 9, 2026).

The authors declare no competing financial interest.