Multistate photoswitches beyond on and off Authors: Yang Xu, Yuqi Tang, Quan Li Published: 2026 ABSTRACT AbstractMolecular photoswitches are compounds that undergo isomerization between at least two distinct states upon light irradiation, accompanied by pronounced changes in their physicochemical properties. The incorporation of photoswitches into material systems enables light-triggered switching of material properties. Multistate photoswitches are attracting considerable attention because of their manipulation of light-controllable states and wider application when compared to traditional binary photoswitches. In this Perspective, we discuss the rapidly advancing field of multistate photoswitches, exploring the evolution of the field, fundamental principles, key design strategies, representative examples, diverse emerging applications, as well as the current challenges and future research directions.Subject terms: Optical materials, Synthetic chemistry methodology, Photochemistry, Information storage Introduction Owing to their significant application potential, the design and fabrication of stimuli-responsive smart materials attracted substantial research interest1. Photoswitches are often incorporated into responsive material systems to enable light-controlled functionalities and make use of the high spatiotemporal precision, remote controllability, and noninvasive nature of light as stimulus2. Molecular photoswitches are characterized as compounds that undergo reversible isomerization between at least two distinct states upon light irradiation and, upon isomerization, show pronounced changes in their physicochemical properties3. Consequently, photoswitches were widely used across diverse fields, including photoswitchable optoelectronic devices4,5, photocatalysis6,7, information storage8,9, and biological applications10,11. Established molecular photoswitches such as azobenzenes (AB), spiropyrans (SP), diarylethenes (DTE), acylhydrazones (AH), dimethyldihydropyrenes (DHP), dihydroazulenes (DHA), and fulgides (FG) are widely employed, with an increasing research effort on emerging classes including hemiindigoids (HI), donor-acceptor Stenhouse adducts (DASA), and other photoswitches12–14. Conventional binary photoswitches are confined to two discrete states, typically described as “on” and “off”. The development of materials with advanced functionalities for complex applications relies on photoswitches possessing more than two states, termed multistate photoswitches. The increased state diversity and potential for synergistic effects in these systems offer enhanced versatility and functional complexity when compared to conventional binary photoswitches, and broaden considerably their field of application15. However, the design of efficient multistate photoswitches faces major challenges, such as balancing the integration of additional functionalities with the preservation of core photoswitching performance. Creating diverse photoresponsive states typically requires intricate synthetic modifications, which often detrimentally impact essential switching properties such as quantum yield, fatigue resistance, or absorption profiles, and potentially lead to diminished performance or complete loss of switching capability16. Additionally, achieving orthogonal and reversible control over multiple states remains difficult. Independently and reversibly populating specific states by using distinct wavelengths of light is challenging, and unwanted cross-talk between states frequently hinders precise targeting of specific states. Finally, the photochemical mechanisms governing multistate switches are inherently more complex than those commonly observed in bistable systems. Understanding the detailed reaction pathways, identifying intermediate states, and elucidating the stepwise transformation kinetics are critical to the development of multistate photoswitches. These challenges may be one reason why despite their potential, reports on high-performance multistate photoswitches are scarce. Currently, research focuses primarily on molecular design and exploration of these photoswitches, while exploring practical applications remains in its infancy. Given the immense application potential of multistate photoswitches, a perspective which discusses the state-of-the-art of this field as well as the critical challenges which currently impede the development of the field is a welcome addition to the literature. Here we provide a thorough exploration of this rapidly advancing field, examine the development trajectory, define characteristics and key design strategies, and discuss representative examples and emerging applications of multistate photoswitches. This perspective first provides a general overview of multistate photoswitches, before systematically discussing different pathways to achieve multistate behavior. We then highlight emerging applications of multistate photoswitches and conclude with a summary and perspective on future developments. This perspective aims to provide a cohesive overview of the current research landscape and to catalyze further innovation within this field. Typical features and development of multistate photoswitches This section provides an overview of the characteristic features and development of multistate photoswitches, along with a discussion of the difference between binary photoswitches and multiphotochromic systems. Conventional binary photoswitches are limited to toggling between only two states: “on” and “off”. The logic is simple: one stimulus results in one of the two states. In contrast, multistate photoswitches can access three or more distinct states. This expanded functionality arises from either the intrinsic ability of a photoswitch to access multiple distinct states, or the integration of photoswitches with other photoswitching units and complementary stimulus-responsive units (Fig. 1a). The advantages of multistate photoswitches are evident, as a greater number of states enables more complex control and multiplexed changes in all kinds of material properties (Fig. 1b). State control in multistate photoswitches is characterized as multi-path or sequential control, whereby the system transitions through multiple well-defined states upon exposure to distinct stimuli or a sequence of stimuli, following a specific order or pathway. Specifically, multi-path control involves branching responses, which means that from a single starting state, different states can be accessed either simultaneously or selectively, allowing “shortcuts” that bypass intermediate states. Sequential control is more restrictive, requiring a fixed sequence of steps to reach a particular state that cannot be accessed directly from other states. In this context, orthogonal switching is defined as the ability to address each state independently without altering the others. A path-independent version of orthogonal switching enables access to any other state regardless of the starting point. Contrary to multistate photoswitches based on single molecules, multiphotochromic systems are typically constructed using different molecular photoswitches, which are physically blended or co-loaded onto a carrier to enable multistate control17,18. Multiphotochromic systems generally operate via parallel or combinatorial control, wherein multiple independent stimuli can simultaneously act on different components of the system and enable the Characteristics and common photoswitching core structures of multistate photoswitches (PS). a Schematic comparison of conventional binary photoswitches and multistate photoswitches. b Schematic of property control in single-molecule-based multistate photoswitches. c Structures and isomerization mechanisms of photoswitching cores commonly used in the construction of multistate photoswitches. activation of distinct states or behaviors in a non-interfering manner. Notably, multicomponent mixtures of photoswitches is a straightforward strategy to realize orthogonally addressable systems and to avoid the synthetic complexity of covalent linking of multiple photoswitching units and are therefore gaining popularity19,20. However, in certain multiphotochromic systems, particularly those involving either physical mixtures or co-loading onto a carrier, challenges such as inhomogeneous dispersion, difficulties in optimizing doping ratios, and mutual interference among photoswitch molecules may still arise. In contrast, multistate control based on a single photoswitch molecule typically offers simpler and more precise composition control along with straightforward optical manipulation. The fundamental differences between binary photoswitches, multiphotochromic systems, and multistate photoswitches are systematically summarized in Table 1. Recent research increasingly focuses on single-molecule multistate photoswitches. The development of such systems, however, places higher demands on the molecular photoswitching core, which must exhibit stable and reversible isomerization, must allow easy chemical modification, and must retain its switching functionality after functionalization. This Perspective centers on single-molecule multistate photoswitches, while binary photoswitches and multiphotochromic systems are discussed only in a comparative manner. Table 1Comparison of binary photoswitches, multiphotochromic systems, and multistate photoswitchesFeatureBinary photoswitchesMultiphotochromic systemsMultistate photoswitchesComponentsOneTwo or moreOneStatesTwoThree or moreThree or moreCarrier requirementTypically not requiredOften requiredTypically not requiredInput stimuliLight or heatLight, heat and other stimuliLight, heat and other stimuliDesign strategyMolecular modulationModular assemblyPrecision molecular engineeringState control logicTwo-state switchingParallel/combinatorialMuti-path/sequential control Comparison of binary photoswitches, multiphotochromic systems, and multistate photoswitches Currently, several important classes of photoswitching core skeletons are successfully employed in the construction of multistate photoswitches. Representative examples include AB21, HI22, AH23, DTE24, DHA25, FG26, and SP derivatives27. Moreover, emerging photoswitches such as DHP and DASA derivatives are increasingly reported as framework for the development of multistate photoswitchable systems (Fig. 1c)28,29. Classifying them according to their isomerization mechanisms, AB, HI derivatives, and AH are representative examples for a category of photoswitches which rely on E-Z isomerization involving double-bond rotation. These types of photoswitches typically exhibit broad applicability, facile synthesis and modification, as well as tunable response wavelengths, and are therefore typically employed in the design of multistate photoswitches. The second major category of photoswitches is based on electrocyclic reaction mechanisms, with DTE and DHP serving as representative examples. These photoswitches typically enable bidirectional optical control, where forward and reverse isomerization can be selectively induced using two distinct wavelengths of light. This bidirectional switching capability facilitates efficient, rapid, and active control in multistate configurations, making them a preferred choice for advanced multistate photoswitch designs. Diarylethenes represent typical P-type photoswitches, meaning they operate under complete photochemical control without thermal relaxation, and offer distinct advantages for applications. Moreover, recent studies on DHP derivatives demonstrated quantitative photoconversion between two photostationary states, accompanied by a bathochromic shift of the response wavelengths into the visible (VIS) and near-infrared (NIR) regions, and highlighted their potential as multistate photoswitching systems30,31. A third category operates through more complex switching mechanisms, typically involving multistage isomerization processes, as those observed in DASAs and SPs. DASAs represent an emerging class of photoswitches with negative photochromism, in which the colored isomer converts to a colorless metastable isomer upon irradiation with light, whose modular synthesis and tunable spectral absorption make them attractive components for multistate photoswitch design. Furthermore, DASAs respond readily to long-wavelength light which distinguishes them from other photoswitches such as AB and enables orthogonal control in multiple photoswitch systems. Additionally, their sensitivity to stimuli such as pH further confers multistimuli responsive characteristics, expanding their potential application in advanced photoswitchable materials. SP-based photoswitches are extensively studied for the design of multistate photoswitchable systems, owing to their ability to coordinate metal ions, their response to pH and mechanical forces, and their ability to function as photoacids for light-controlled proton release. Furthermore, photoswitches with complex isomerization mechanisms enable multistage structural control by manipulating distinct steps of their isomerization pathways, thus expanding the scope and diversity of the multistate photoswitch family. Multistate photochromism is fundamentally linked to photochemical reactions involving sequential multiphoton absorption. From this perspective, multistate photoswitching can be conceptualized as a sequence of photochemical events that are analogous to stepwise two-photon or multiphoton absorption processes32,33. Building on this concept, extensive studies explored multiphotochromic reactions, particularly those based on multistep photochromic processes in imidazole dimer systems34,35. These systems serve as prototypical models for studying sequential photochemical transformations and state-selective control. Despite the structural diversity, robust performance, and orthogonal compatibility with nonphotomediated stimuli, the development of synthetic strategies and practical applications for single-molecule multistate photoswitches remains at an early stage. Design strategies and representative examples of multistate photoswitches This section discusses the design strategies employed to achieve multistate behavior and highlights representative examples of multistate photoswitches. In recent years, considerable efforts have been made to develop efficient and stable multistate photoswitches36. Current multistate photoswitches can be classified into three main categories: (1) multistate singular photoswitches with inherent multiisomeric states, (2) composite or multimer photoswitches, and (3) composites of photoswitches with secondary stimulus-responsive moieties. The first category, multistate singular photoswitches with inherent multi-isomeric states, possesses more than two accessible isomeric forms, enabling state multiplicity without the need for multiple covalently linked units. The design of such systems presents unique challenges, as it requires precise molecular design and modulation. The realization of the second and third categories typically involves the integration of multiple photoswitching units or the combination of photoswitches with other stimulus-responsive components. One key challenge is the efficient coupling of these functional elements while preserving the intrinsic photoswitching properties, thus enabling precise and multiplexed control of the overall behavior of the multistate photoswitching system. Current design strategies face several constraints, such as exciton coupling. When a chromophore absorbs photon energy and transitions from the ground state to an excited state, the altered electron distribution generates a transition dipole moment (μ). In systems containing multiple chromophores, the electronic excitation becomes delocalized across the molecule, a phenomenon known as exciton coupling37. Depending on whether the transition dipole moments adopt parallel or collinear orientations, this coupling can result in either a blueshift or redshift of the absorption maximum (Fig. 2a)38. Consequently, compared with isolated chromophores, oligomeric molecules with several photoactive components often exhibit significantly different photophysical behaviors. Another significant constraint involves steric effects, where geometrical constraints or macrocyclic strain can hinder complete isomerization (Fig. 2b)39. Under substantial spatial confinement, photoswitches often lose their ability to isomerize or exhibit only partial isomerization. However, these constraints can be strategically exploited to enhance control over photoswitching properties, e.g. by steering photoisomerization in a more directional way, or by adding more states through cyclization40–44. In addition, energy transfer among multiple units within a photoswitch molecule also constitutes a limiting factor in achieving efficient control over multiple states45. Generally, the design strategies for multistate photoswitches are categorized according to several distinct approaches, including multistate singular photoswitches with inherent multi-isomeric states, composite or multimer photoswitches, and composites of photoswitches and secondary stimulus-responsive moieties. Schematic of electronic and steric effects in the multistate photoswitch design. a Schematic of the underlying mechanism of exciton coupling. b Schematic of steric hindrance and macrocyclic strain effects on multistate photoswitches. GS: Ground state; ES: Excited state. One strategy involves multistate singular photoswitches with inherent multi-isomeric states, referring to molecules in which a single central photoswitch unit can access more than one isomeric form. This approach generally encompasses two distinct scenarios. In singular photoswitches, multistate photoswitching is achieved by stabilizing metastable intermediates along stepwise photoisomerization pathways through molecular engineering (Fig. 3a). The selective population of specific isomers can be controlled by modulating the relative stability and energy barriers between these states (Fig. 3b). Representative photoswitches employing this strategy include, but are not limited to, DTEs46, FGs47, chromenes3, and DASAs (Fig. 3c)48–50. For instance, DASA-1 (Fig. 3d) achieves a three-state control by incorporating substantial steric hindrance which stabilizes the intermediate state and enables distinct and addressable photostationary states (Fig. 3d)51. Design strategy and examples of multistate singular photoswitches with inherent multi-isomeric states. a Schematic of potential energy surfaces during isomerization of a photoswitch. b Two types of multistate singular photoswitches with inherent multi-isomeric states. The first type involves a photoswitch that undergoes isomerization through multiple stages, transitioning from isomer A to isomer B and subsequently to isomer C. The second type involves a photoswitch that, upon isomerization, generates a new photoswitch capable of undergoing further isomerization. c Representative photoswitches undergoing multistep isomerization processes. d Isomerization pathway and three states of photoswitch donor-acceptor Stenhouse adduct (DASA-1). e Isomerization pathway and three states of the aza-diarylethene. Another powerful strategy uses different types of correlated or uncorrelated bond rotations within a single molecule to achieve a high number of addressable states. These systems can be controlled either by fully light-driven processes or by a combination of light and thermal stimuli52–54. Earlier studies achieved eight- and nine-state switching in composite multimers or multi-stimuli systems55,56. This concept is further advanced in recent works utilizing similar principles to achieve control over multiple states57–60. In addition, certain photoswitches generate new photoisomerizable units upon initial light activation, enabling sequential structural transformations to access multiple states. One representative case is the zwitterionic aza-diarylethene system, which reversibly forms and cleaves a C=N bond while simultaneously creating an isomerizable double bond, ultimately demonstrating three-state photoswitching behavior (Fig. 3e)61. Another strategy to achieve multistate behavior involves composite or multimer photoswitches, in which at least two individual photoswitch units are interlinked. These systems can be broadly divided into two categories: those with fully individual combinations of photoswitch units, and those adopting fused forms that share common connecting parts. In the former case, the constituent units may be either identical (homo-multimers) or different (hybrids), and their photoresponses can be either interfering or isolated depending on the degree of electronic coupling. Dimers and multimers of classical photoswitches, such as AB, SP, and DTE, have also been extensively reported28,62–64. Historically, pioneering work on multistate switching using DTE systems was reported by Irie and Branda, who demonstrated that DTE dimers can access multiple states through sequential photocyclization reactions65,66. Mitchell and coworkers reported the foundational example of a dimeric dihydropyrene switch representing one of the first systems capable of accessing three distinct photochemical states67. Although covalent linkers remain the predominant strategy for conjugating multiple photoswitch units, this methodology often leads to compromised photoresponsiveness. Consequently, the construction of dimeric and multimeric architectures that fully retain their light-driven isomerization capability is crucial. When multiple identical units are connected via electronically coupled linkers, the photoinduced transformation of an individual unit inevitably perturbs the photoisomerization behavior of adjacent units because of intramolecular excited-state energy transfer. One efficient approach involves reducing the π-conjugation between two photoswitching units to minimize their electronic coupling. As demonstrated by the AB dimer AB-4, the introduction of sterically demanding groups induces the largest dihedral angle among the series, resulting in the weakest coupling and thus enabling more complete photoisomerization (Fig. 4a)68. The substitution position of the linker exerts a profound influence on the performance of multistate photoswitches. For example, meta-substituted linkers can efficiently attenuate electronic conjugation between photochromic units44. The meta-linked AB dimer AB-6 exhibits more pronounced multistate photoswitching behavior compared to its para- and ortho-substituted analogues (Fig. 4b)69,70. Furthermore, appropriate macrocyclic strain can facilitate isomerization in macrocycle-based multistate photoswitches. Since their introduction by Rau et al. in 1982 with azobenzenophanes71, namely cyclic compounds in which azobenzene units are linked into a ring via bridging structures, biphotochromic systems garnered significant interest. Over several decades, research expanded to systems with multiple covalently linked ABs, where cooperative effects, intramolecular coupling, and other emergent photochemical behaviors were experimentally characterized32,72. Design strategy and examples of composite or multimer photoswitches. a Design strategy for reducing π-conjugation with representative examples. b The meta-rule applied in structural design and examples. c Macrocycle-based design approaches and examples. d Asymmetric substitution strategies and examples. e Nonconjugated linker design and examples. f Schematic of the principle and examples of constructing multistate photoswitches via metal coordination. AB: azobenzene; DTE: dithienylethene; CC closed closed, OO open open, CO closed open, OC open closed. Based on previous studies, compound AB-9 demonstrates enhanced multistate photoisomerization properties compared to its counterparts, which is achieved by introducing appropriate steric hindrance and macrocyclic strain (Fig. 4c)73. Sequential isomerization in dimeric and multimeric systems can also be realized through asymmetric substitution, where distinct spectral response windows of individual units enable multistate control via wavelength selectivity. A representative example is compound NBD-1, in which four distinct states can be addressed through sequential irradiation at three different wavelengths (Fig. 4d)74. Alternatively, the use of nonconjugated linkers presents another viable design strategy. Although the individual photoswitching units behave relatively independently during isomerization, this approach nevertheless enables the development of high-performance multistate photoswitches in numerous cases. For instance, compound hydrazone-1 achieves molecular steganography in multistate photoswitchable hydrazones by incorporating an isosorbide-based linker (Fig. 4e)75. Metal-coordination complexes offer a versatile platform for constructing composite or multimer photoswitches by coordinating identical or distinct photochromic units. This strategy typically involves functionalizing photoswitches with coordinating ligands (e.g., pyridine derivatives), followed by metal-ion-directed assembly using Fe, Cu, Pd, Pt, or Zn ions76. The metal centers act as structural spacers between photochromic units, mitigating electronic perturbation of individual photoswitches and frequently enabling stepwise photoisomerization processes or full photoisomerization (Fig. 4f)77. As exemplified by compound DTE-1 (Fig. 4f), a system formed by coordinating two different DTE units to a ruthenium center enables orthogonal optical control over four distinct states78. In contrast to dimers and multimers constructed from identical photoswitches, hybrid photoswitches integrate distinct photochromic units within a single molecular framework. This design leverages the wavelength-selective responses of different photoswitches to achieve orthogonal control (Fig. 5a). Diverse combinations of molecular photoswitches are reported in the literature79–83. Photoswitches with well-separated response wavelengths are used in the design of hybrid multistate systems, which allow addressing different states individually. Currently, research on developing hybrid photoswitches follows primarily three developmental directions. The first involves transitioning from nonconjugated to conjugated linkers. While conjugated linkers facilitate coupled photoresponse and enable synchronized multi-property changes rather than independent transformations, they present a considerably greater synthetic challenge. For instance, the hybrid system AB-SP-1 achieves multistate photoswitching through nonconjugated connections but generally lacks sophisticated photophysical and photochemical properties (Fig. 5b)84. In contrast, the hybrid photoswitch AB-SP-2 (Fig. 5b) features chromophores that share a common benzene ring and exhibits wavelength-selective photocontrol over both photophysical and Design strategy and examples of composite multistate photoswitches based on hybrid photoswitches. a Design and structural schematics of hybrid photoswitches b Examples of hybrid photoswitches from nonconjugated to conjugated linkers. c Examples of hybrid photoswitches from ultraviolet to visible and near-infrared light response. d Examples of hybrid photoswitches from partial to all-photoswitching. AB azobenzene, SP spiropyran, FG fulgide, DTE diarylethene, DASA donor-acceptor stenhouse adduct, DHP dimethyldihydropyrene. photochemical characteristics. This demonstrates the advantage of electronic communication between chromophores85. A second direction focuses on multistate photoswitches driven by VIS and NIR light. VIS and NIR light are preferred for practical applications due to their deeper tissue penetration, minimal photodamage, and higher energy efficiency. By contrast, ultraviolet (UV) irradiation poses significant limitations: it carries the risk of inducing biological damage and undesirable side reactions, which can compromise switching durability. These advantages make VIS and NIR light not only indispensable for biological applications but also highly favorable for a broader range of material and device applications86,87. Currently, most multistate photoswitches still require UV light for state transitions. A representative example is compound FG-DTE-1, in which both photochromic units necessitate UV excitation (Fig. 5c)88,89. Some progress has been made in systems where one photoswitch utilizes UV light while the other operates entirely under VIS light. For instance, the DASA-AB-1 (Fig. 5c) system employs a DASA unit to control phase transitions and an AB unit to regulate binding with α-cyclodextrin90. Nevertheless, composite or multimer photoswitches where both units can be selectively addressed using only VIS or NIR light remain rarely reported in the literature91. The third direction is all-photonic switching composite or multimer photoswitches, which provide direct and instantaneous optical control. This capability has drawn considerable research attention to photoswitchable units that undergo bidirectional isomerization at distinct wavelengths, including diarylethenes, dimethyldihydropyrene, and azobenzene derivatives. A representative example is compound DHP-AB-1, a molecular dyad integrating dimethyldihydropyrene and azobenzene units, which demonstrates full optical switching functionality applicable to logic operations (Fig. 5d)92. Molecular motors represent a class of systems that are inherently multistate in nature93,94. These molecules undergo sequential, unidirectional rotations or translations through multiple well-defined states, and can be understood within the same conceptual framework discussed above95,96. On one hand, single-molecule motors operate via sequential photochemical and thermal steps, demonstrating intrinsic multi-isomeric states within a single molecular scaffold97,98. On the other hand, composite molecular motors employ architectures where multiple units work together in a concerted way to achieve multistate behavior99,100. The design principles, control, and emergent functions of these molecular motors offer valuable insights for the broader field of multistate photoswitches. Another widely adopted strategy involves coupling photoswitches with complementary stimuli-responsive units to achieve multistate transitions driven by multiple stimuli beyond light alone101,102. In these systems, light functions as the primary triggering stimulus, whereas secondary stimuli subsequently modulate the states, which differs fundamentally from gated photoswitching mechanisms103,104. One primary approach involves chemical stimuli, which encompass responses to metal ions, humidity, solvents, and specific chemical substances105. Metal ion responsiveness can be achieved by incorporating crown ethers or functional groups with metal-coordination sites into the photoswitch architecture. For instance, a light and Ca²⁺-responsive organic-peptide hybrid, designated SP-1, was reported (Fig. 6a)106. In addition, crown-ether-modified diarylethene incorporated into lecithin-Ca²⁺ gels enables reversible Ca²⁺ binding upon optical stimulation, leading to a fully reversible gel-to-sol transition that can be controlled by light107. However, this concept suffers from certain limitations, particularly the requirement for external chemical additives to trigger transitions of different states, leading to the gradual accumulation of chemical byproducts over multiple operation cycles. Furthermore, chemical stimulation often adversely affects the recyclability of molecular photoswitches and imposes strict constraints on molecular design and compatibility with the matrix. Future research efforts will focus on exploring clean chemical fuels and developing strategies to mitigate the accumulation of chemical waste. Design strategies and examples of composites of photoswitches and secondary stimulus-responsive moieties. a Examples of photoswitches with chemical stimuli. b Examples of photoswitches with pH responsiveness. c Examples of thermally responsive photoswitches. d Examples of redox-responsive photoswitches. e Examples of photoswitches with other photoresponsive units. f Examples of other stimuli-responsive photoswitches. a Adapted with permission from ref. 106. Copyright Zinah Hilal Khaleel et al. (2025) Proc Natl Acad Sci USA, licensed under CC BY-NC-ND 4.0 (https://creativecommons.org/licenses/by-nc-nd/4.0/.). Changes were made. pH variation is another important stimulus, whereby addition of acids or bases modulates the protonation and deprotonation processes of molecular photoswitches and enables multistimuli responsiveness in multistate systems. For instance, diaryl-hemiindigo HI-1 exhibits quadruple-state responsiveness to light, pH, and thermal stimuli (Fig. 6b)108. This capability stems from its nitrogen-containing protonation site, which introduces pH-dependent behavior into the inherent photochromic framework of the photoswitch. Importantly, light-driven proton generation and consumption strategies were developed, such as the use of photoacids to achieve pH regulation109. This approach replaces chemical additives with clean optical stimuli and not only avoids chemical waste accumulation but also significantly improves control precision and operational simplicity. Although both pH modulation and photoacids were explored in multistate photoswitching systems, they present opportunities as well as challenges. Variations in acidity or alkalinity often considerably impact the structural stability of molecular photoswitches. Moreover, the presence of multiple protonation sites in complex multistate systems makes precise regulation of pH difficult in practice. Additionally, despite the remarkable advantages of photoacids, currently available systems exhibit limited ranges of pH variation, and the synthesis of strongly acidic yet reversible photoacids remains challenging. Future research will need to focus on the design and synthesis of integrated systems that combine broad-range photoacids with pH-responsive photoswitches, representing a promising direction for advancing this field. Thermal responsiveness represents another widely utilized stimulus and is characterized by its inherently tunable nature and straightforward implementation. Numerous thermally responsive polymers and matrices were developed for smart material applications. However, integrating thermal response units directly into molecular photoswitches to achieve dual photothermal control remains challenging. The presence of thermal relaxation in most photoswitches renders them inherently sensitive to temperature, which consequently complicates the precise control of thermal responsiveness. Therefore, achieving efficient multistate photoswitching necessitates judicious molecular design and a sufficient thermal operating window to ensure stable population of addressable states. For instance, 1H-2-benzo[c]oxocins have been shown to undergo reversible interconversion with dihydro-4H-cyclobuta[c]isochromenes under light and thermal stimulus. Upon heating to 120 °C for 72 h, the system further transforms into a third state identified as dihydronaphthalenes (Fig. 6c)110. A recent example reported a chameleon aggregation-induced emission luminogen exhibiting six distinct yet tunable thermal and photoswitchable states111. Besides, macrocycles have been demonstrated to exhibit multi-step thermal state switching41,112. For systems requiring rapid, efficient operation with minimal interference, photons and electrical potentials represent ideal stimuli. Both light and applied electrical potential can be delivered remotely with minimal physical perturbation, operate at speeds far exceeding diffusion-limited chemical or thermal stimuli, and the stimuli can be removed just as rapidly as they can be applied. When redox-responsive units are integrated with photochromic molecules, they can alter photochromic properties and further confer novel functionalities to the system. Electrochemical oxidation or reduction, typically employed to toggle between oxidized and reduced states via electron transfer, enables such control. A prominent example is the overcrowded alkene-1, which exhibits multistate switching among anti-folded, syn-folded, and twisted conformations driven by both redox and optical inputs (Fig. 6d)113. One key challenge is the design of molecules that are responsive to both photons and electrons and possess sufficiently long-lived states to allow addressing these states sequentially with complementary stimuli. A distinct strategy for expanding the functionality of photoswitches involves their integration with additional photoactive units that undergo irreversible photoreactions, such as photocaging or introducing photochemically cleavable moieties. Multiphoto stimulation presents unique advantages: it enables simultaneous and orthogonal stimulation with rapid implementation, avoids accumulation of chemical waste or dilution effects, and eliminates the need for additional chemicals, electrical inputs, or other physical components. One common strategy involves linking photoswitches with other photoactive units, such as photocaging groups or photochemically reactive moieties. A related study reported a 2-nitrobenzyl-photocaged SP system, SP-2 (Fig. 6e), in which only cis-trans isomerization occurs prior to uncaging. Upon irradiation at 254 nm, the photocleavage reaction generates free SP, enabling subsequent ring-opening isomerization (Fig. 6e)114. Building on this approach, a subsequent work introduced a sterically hindered adamantane group to construct SP-3 (Fig. 6e), which, after uncaging, can be switched using VIS light115. In addition, a combination of a biradical and an AB motif was independently addressed116. Current research priorities focus on shifting response wavelengths into the NIR region and expanding the operational wavelength range of photoswitches or photoresponsive modules, thus enhancing the potential for selective multiwavelength optical control. Additionally, other forms of stimuli include electrical fields, magnetic fields, mechanical forces, and changes to the local microenvironment. These stimuli can also be incorporated into the design of multistate photoswitches (Fig. 6f). These examples demonstrate that photoswitches can be successfully integrated into systems responsive to a wide range of stimuli, producing multifunctional switches capable of sophisticated behaviors. Multistate photoswitches have already found promising applications in areas such as molecular machines117, molecular logic gates118, and optical memories119. Given their substantial versatility, their full potential across emerging fields remains to be realized. It is therefore crucial to advance the development of reliable and precisely controllable multistate photoswitches grounded in a thorough understanding of their isomerization mechanisms and design principles, while actively exploring their implementation in emerging application fields. Emerging applications of multistate photoswitches In this section, we present an overview of emerging applications of multistate photoswitches. With a growing understanding of the design strategies and working mechanisms of multistate photoswitches, their exploration in emerging fields expanded considerably. Owing to their multistate characteristics, these systems enable advanced functionalities and exhibit enhanced performance features when compared to conventional binary photoswitches. Although research on multistate photoswitches is still in its infancy, they show potential across diverse applications such as energy storage120, information storage and anti-counterfeiting encryption121, molecular machines and soft actuators122, and biomedical applications123. The development of environmentally benign energy storage systems will be important for addressing climate change and ensuring reliable energy supply. Molecular photoswitches have garnered considerable attention in the field of molecular solar thermal (MOST) energy storage, which uses photoisomerization to capture solar energy and release it on demand (Fig. 7a)124–126. In such systems, photons are absorbed to drive the transformation of molecular switches into metastable states of higher energy127. An ideal MOST photoswitch should meet several conditions: strong absorption within the VIS region to match the solar spectrum, high conversion efficiency during the photoisomerization process, large energy density, and a metastable isomer with sufficient thermal stability and long half-life128. Photoswitches are currently widely employed in MOST energy storage. However, several limitations persist. As the photon energy decreases with increasing wavelength, bathochromic shifts in the absorption generally lead to reduced energy density. Additionally, high molecular weight often constrains the gravimetric energy density. Multistate photoswitches represent a promising strategy for overcoming these limitations. By integrating multiple photoisomerizable units within a single molecule, such systems enable stepwise isomerization that not only broadens spectral overlap with sunlight but also enhances molar energy density while maintaining relatively low molecular weight129. Studies indicate that the linkage pattern and bridging units between photoswitches considerably influence both the half-life of the metastable state and the photoconversion efficiency, with meta-substituted linkers often providing optimal electronic decoupling and performance. The multistate approach not only increases energy density but also enables active, precise, and on-demand management of energy storage and release processes. Multistate photoswitches based on AB, norbornadiene, and DHA are increasingly explored for advanced MOST applications. Multistate photoswitches also offer distinct advantages in the field of information storage, anti-counterfeiting, and encryption. Schematic of emerging applications for multistate photoswitches. a Schematic of multistate photoswitches for molecular solar thermal energy storage applications. b Schematic of multistate photoswitches for information storage and anti-counterfeiting encryption applications. c Schematic of applications in molecular machines and actuators using multistate photoswitches. d Schematic of biomedical applications using multistate photoswitches. The concept of applying photochromism to computer data storage was first proposed by Hirshberg in the 1950s130, and since then, molecular photoswitches have been widely explored for information handling and encryption131. At its core, information storage and anti-counterfeiting involve the processing of data. Molecular logic gates are characterized by their nanoscale dimensions, low power consumption, and high performance, and present an attractive platform for next-generation information technologies, enabling writing, erasing, and authentication through the manipulation of molecular states118,132,133. Molecular photoswitches, which serve as fundamental components of logic gates, can alter their states in response to chemical, electrical, or optical inputs, thereby generating corresponding outputs. These molecules, with their multiple addressable states, can act as operational elements in logic systems to perform diverse functions89,134. Owing to their state diversity, multistate photoswitches can be used to construct logic gates with sophisticated functionalities, enabling higher information storage capacity and more secure encryption and anti-counterfeiting135–137. Moreover, their specific responses to the sequence, duration, wavelength and irradiation allow for multidimensional encryption and dynamic authentication and open up new avenues for advanced data security (Fig. 7b). Multistate photoswitches are demonstrating growing potential in advancing molecular machines117,138 and soft actuators139,140, overcoming limitations inherent in conventional binary systems constrained to only two states141. These multistate configurations enable sophisticated motion control and diverse operational modalities, permitting the design of molecular machines that execute complex tasks through sequential optical inputs. In practical applications such as actuators, this capability translates to continuous and tunable deformation where the actuator can be precisely maintained at any intermediate state between fully extended and fully contracted configurations (Fig. 7c) by adjusting the irradiation wavelength or intensity. Moreover, the integration of multiple functional units within a single molecular architecture not only simplifies energy input but also considerably increases the mechanical responses, allowing distinct types of motion to be selectively triggered by specific wavelengths. Key parameters, including the bending angle and twisting direction, can be accurately regulated through controlled irradiation intensity and duration, as evidenced by systems that exhibit programmable deformation with quantifiable stress outputs. These characteristics establish multistate photoswitches as particularly valuable components for developing advanced photoresponsive materials with enhanced functionality in nanoscale machinery and adaptive soft robotics. Furthermore, multistate photoswitches demonstrate substantial potential in biomedical applications, particularly in the emerging field of precise photopharmacology142–145. Photopharmacology advanced rapidly through the photocontrol of drug activity based on the distinct physiological activities of different photoswitch isomers146–149. Multistate photoswitches offer even greater potential in this field, as they leverage the distinct pharmacological activities associated with multiple states to achieve accurate optical control over drug “dosing.” Through selective irradiation, a single drug molecule can be driven into multiple activity states, thus achieving gradient regulation of biological effects (Fig. 7d). This capability is crucial for investigating dose-dependent effects, mimicking dynamic changes in endogenous signals, and enabling refined therapeutic interventions. Furthermore, multistate photoswitches hold significant promise for controlling protein conformations and regulating gene expression150–152. Their application in drug delivery systems offers unique advantages, including remote-controlled drug release at target sites with high spatiotemporal precision153. Compared with conventional binary photoswitches, the state diversity of multistate systems facilitates graded drug release or coordinated release of combination therapies, substantially enhancing both the efficiency and accuracy of drug delivery processes. Beyond the above-discussed applications, multistate photoswitches show significant promise in emerging fields such as catalysis and electronics through their tunable optoelectronic properties154,155. In catalytic systems, these molecular switches enable dynamic regulation of reaction pathways and efficiency via light-driven interconversion between distinct states, facilitating real-time control over catalytic activity and selectivity. For electronic applications, photoswitches exhibit optically addressable conductivity states, supporting their implementation in diverse electronic devices where dynamic electrical characteristics are essential. Summary and outlook This Perspective explores comprehensively the burgeoning field of multistate molecular photoswitches, highlighting their defining characteristics, core design strategies, representative examples, and emerging applications. Compared to their traditional bistable counterparts, multistate photoswitches offer unparalleled versatility and functional complexity due to their increased state diversity and potential for synergistic effects. This positions them as critical components for developing next-generation stimuli-responsive materials. Although considerable progress has been made in molecular design, mechanistic studies, and applications of multistate molecular photoswitches, the field remains in its nascent stage. On the basis of current challenges and application demands, we propose the following perspectives: (1) Leveraging artificial intelligence/machine learning to optimize the design process: The strategic integration of artificial intelligence and machine learning can accelerate the design and optimization of multistate photoswitches, enabling predictive modeling of molecular properties and switching behaviors; (2) Deepening the mechanistic understanding; Fundamental advances in theoretical computations and molecular simulations are essential to elucidate photoreaction pathways, state energy barriers, and spectral properties, and will help to guide the rational design of multistate photoswitches; (3) Expanding the photoswitch toolbox: Efforts towards developing novel photoswitching mechanisms beyond traditional isomerization (e.g., conformational photoswitches), alongside engineering advanced photoswitch motifs with intrinsically well-separated isomer absorption bands, will facilitate modular synthesis of multistate photoswitches; (4) Advancing material integration and device engineering: Innovative strategies for incorporating multistate photoswitches into functional matrices must preserve switching efficiency and state integrity. Concurrently, prototyping next-generation devices that leverage multistate capabilities, such as multilevel memories, programmable soft actuators, and adaptive catalysts, is critical. Addressing scalability and processability barriers remains paramount for real-world implementation; (5) Expanding the applications to complex biological systems: Realizing the transformative potential of multistate photoswitches in biomedicine necessitates overcoming specific hurdles. This requires the concerted effort of designing biocompatible, water-soluble photoswitches that exhibit bioorthogonal reactivity to ensure safe and specific operation within biological systems. Concurrently, achieving precise spatiotemporal control over multistep switching processes in complex and crowded cellular environments is important. Successfully addressing these foundational challenges will unlock the ability to develop sophisticated systems that can perform complex tasks such as multistage drug release, sequential activation of therapeutic agents, or advanced biosensing with multiplexed readouts, thus harnessing the full power of multistate switching for intricate biological control. The field of multistate molecular photoswitches stands at a promising frontier, poised for transformative growth. Through interdisciplinary collaboration and technological innovation, these sophisticated molecular systems are anticipated to revolutionize responsive materials, information technologies, and biomedical applications, ultimately unlocking new dimensions of control and functionality at the molecular level. The future of multistate photoswitches appears exceptionally bright, heralding an era of unprecedented molecular-scale engineering. Acknowledgements The authors thank the financial support by the National Natural Science Foundation ofChina (22505027), the Jiangsu Innovation Team Program, the Fundamental Research Funds for the Central Universities, the Basic Research Program of Jiangsu (BK20251322), the China Postdoctoral Science Foundation (2025M771164), the Postdoctoral Fellowship Program of the CPSF (GZC20250793), and the Jiangsu Funding Program for Excellent Postdoctoral Talent (2025ZB618). Author contributions Y.T. and Q.L. conceived and designed the content of the perspective. Y.X., Y.T., and Q.L. participated in the discussion of all figures and details. Y.X. and Y.T. wrote the original manuscript, and Y.T. and Q.L. revised the manuscript. All authors agree with the final version of the manuscript. Peer review Peer review information Nature Communications thanks Subhajit Bandyopadhyay and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Yuqi Tang, Email: yqtang@seu.edu.cn. Quan Li, Email: quanli3273@gmail.com. References