Light/X-ray/ultrasound activated delayed photon emission of organic molecular probes for optical imaging: mechanisms, design strategies, and biomedical applications

Abstract

Conventional optical imaging, particularly fluorescence imaging, often encounters significant background noise due to tissue autofluorescence under real-time light excitation. To address this issue, a novel optical imaging strategy that captures optical signals after light excitation has been developed. This approach relies on molecular probes designed to store photoenergy and release it gradually as photons, resulting in delayed photon emission that minimizes background noise during signal acquisition. These molecular probes undergo various photophysical processes to facilitate delayed photon emission, including (1) charge separation and recombination, (2) generation, stabilization, and conversion of the triplet excitons, and (3) generation and decomposition of chemical traps. Another challenge in optical imaging is the limited tissue penetration depth of light, which severely restricts the efficiency of energy delivery, leading to a reduced penetration depth for delayed photon emission. In contrast, X-ray and ultrasound serve as deep-tissue energy sources that facilitate the conversion of high-energy photons or mechanical waves into the potential energy of excitons or the chemical energy of intermediates. This review highlights recent advancements in organic molecular probes designed for delayed photon emission using various energy sources. We discuss distinct mechanisms, and molecular design strategies, and offer insights into the future development of organic molecular probes for enhanced delayed photon emission.

---

Rui Qu

Rui Qu is currently pursuing his PhD in the School of Chemistry and Chemical Engineering at Nanjing University, under the supervision of Prof. Xiqun Jiang and Prof. Xu Zhen. His current research interest is the development of novel organic molecular probes for bioimaging applications.

Xiqun Jiang

Prof. Xiqun Jiang have been worked at School of Chemistry and Chemical Engineering, Nanjing University since 1986. He received his PhD in Polymer Materials and Engineering from Nanjing University in 1998. His research interests include macromolecule self-assembly, molecular imaging probe, polymer drug delivery systems, and precise modification and biomedical applications of peptide- or protein-based biomaterials.

Xu Zhen

Prof. Xu Zhen received his PhD in Chemistry from Nanjing University in 2014 followed by postdoctoral studies at National University of Singapore (2014–2015) and Nanyang Technological University (2015–2018). He joined Nanjing University as an Associate Professor in 2019. His research interests focus on the development of smart molecular probes and nanomedicines for precision diagnosis and therapy.

---

1. Introduction

Optical imaging, known for its noninvasiveness and high spatiotemporal resolution, serves as a powerful tool in bioimaging.^1–7 Compared to conventional clinical imaging techniques such as magnetic resonance imaging (MRI),^8–10 ultrasound (US) imaging,^11,12 computed tomography (CT)^13,14 and positron emission tomography (PET),^15,16 optical imaging offers superior high-resolution visualization of morphological and functional information.^17,18 This capability significantly enhances accuracy in foundational biological analysis and clinical diagnostics. However, traditional optical imaging techniques, such as fluorescence imaging, require real-time light excitation that often leads to significant background noise in vivo due to autofluorescence from biological tissue or intraperitoneal food.^19,20 As a result, fluorescence imaging often suffers from poor signal-to-background ratios (SBRs) and reduced sensitivity, limiting its effectiveness in various imaging applications.

To counteract the impact of background noise from autofluorescence, a novel optical imaging strategy that captures optical signals after light excitation has been developed.^21,22 This approach relies on molecular probes designed to store photoenergy and release it gradually as photons, resulting in delayed photon emission that minimizes background noise during signal acquisition. Inorganic delayed photon emission, which originated from natural minerals known to glow after sunlight exposure, has been exploited far earlier than organic chromophores. The mechanisms and applications of inorganic delayed photon emission are well-established.^23–25 Drawing inspiration from these well-developed inorganic systems, chemically constructed defects have become essential in organic delayed photon emission. These defects transform absorbed photoenergy into activation energy for emissive excitons, which subsequently emit photons with a delay, without the need for real-time optical excitation, thereby enhancing imaging sensitivity.²⁶ The delayed photon emission behavior of these probes can be observed at various stages from excitation to emission and can be categorized into three mechanisms in organic system: (1) charge separation and recombination, which uses the charge diffusion process to gradually produce exciplex emission;^27,28 (2) generation, stabilization and conversion of the triplet excitons, which utilizes crossing between energy states and additional stabilizing energies in traps to extend excitons lifetime;^29,30 (3) generation and decomposition of chemical traps, where photoenergy is transformed into chemical energy to create metastable emissive intermediates.^31,32 These mechanisms have proven effective in achieving prolonged emission with certain molecular probes, offering significant potential for high SBR imaging in biomedical applications.

Due to limited tissue penetration, utilizing light as the energy source to induce energy storage in defects within deep-seated molecular probes is challenging.^33,34 The inefficient energy delivery by light in deep tissues restricts energy input efficiency, leading to inadequate energy storage and a consequent shallow penetration depth for delayed photon emission. In contrast, X-ray and ultrasound, with their inherent deep tissue penetration capabilities, can serve as alternative energy sources for deep-tissue energy input.^35,36 These modalities convert high-energy photons or mechanical waves into potential energy of excitons or chemical energy for intermediates, facilitating the development of delayed photon emission strategies that achieve the desired depth of tissue penetration.

In this review, we present recent advancements in organic molecular probes designed for delayed photon emission, triggered by energy from diverse sources. We explore the distinct mechanisms underlying the three main delayed photon emission processes and elucidate the design strategies employed in developing these organic molecular probes for biomedical applications. Additionally, we provide insights into the future development of these probes for delayed photon emission, highlighting current challenges and proposing potential methods to enhance the feasibility of these design strategies in biomedical applications.

2. Light-activated delayed photon emission

Delayed photon emission is a photophysical process wherein optical materials capture excitation energy and gradually emit it via an extended radiative decay. Distinct from conventional fluorescence, which is characterized by an electronic transition from singlet excited state (S_n) to the ground state (S₀) within a few nanoseconds,^37–39 delayed photon emission stabilizes excitons in a metastable state throughout the excitation and de-excitation processes, significantly prolonging its lifetime to seconds.⁴⁰ Moreover, the efficacy of delayed photon emission hinges on the formation of energy trap, which result from the synergistic effects of energy input and the structural design of the materials.^41–43

Light, serving as the primary energy source for optical materials, triggers diverse species in excited state that undergo distinct photophysical processes under varying power and environmental conditions.^44–50 These processes facilitate delayed photon emission through the formation and stabilization of excitons and electronic transitions across different energy states. When an optical molecule is activated by light to an excited singlet state, intersystem crossing (ISC) leads the transition to a higher spin multiplicity (excited triplet state), a process inherently spin-forbidden from reverting to the ground state, thereby delaying photon emission.^51,52 In turn, reverse intersystem crossing (RISC) can thermally activate the triplet exciton to overcome the energy gap between singlet state and triplet state (ΔE_ST), enabling singlet exciton formation for delayed emission when ΔE_ST is sufficiently small.^53–55 Moreover, delayed singlet excitons can be generated through triplet–triplet annihilation (TTA) after removal of the excitation source.^56–58

The generation of excitons and subsequent photophysical process are further complicated with the consideration of intermolecular interactions.^59–62 Aggregations and doping not only facilitate the overlap of electronic orbitals to generate stabilized exciton state but also enable charge transfer between separated donors and acceptors.^63–65 Strategically arranged molecular designs in aggregations guide excitons towards newly stabilized triplet states and shield them from excessive non-radiative decay due to structural rotation and oxygen/humidity quenching, thereby extending their lifespan and improving emitting efficiency.^66–69 Additionally, delayed exciplex formation, emerging from ionized exciton recombination, also defers light emission.^70–73 Upon light activation, certain donor–acceptor systems enable charge separation, capturing charges in metastable states. These charges can be gradually recombined under thermal or optical excitation, leading to delayed exciplex emission. Notably, the delayed generation of emissive species is also prevalent in chemical process.^74–77 Advances in chemiluminescence have unveiled classical pathways involving the formation and decomposition of unstable cyclic peroxides, such as 1,2-dioxetane/1,2-dioxetanedione species.²² This process, relying on both a photosensitizer and a chemiluminescence substrate, converts light energy to chemical energy stored in cyclic peroxides, which is gradually released as these peroxides decompose into light-emitting excitons.

Both photophysical and chemical processes underscore the essential role of a metastable excited state with an extended lifetime for emission delay, a principle extensively applied in various applications. We aim to elucidate the various photophysical and chemical mechanisms, alongside their molecular design strategies, for achieving delayed photon emission.

2.1. Light-activated charge separation and recombination

Theoretically, delayed photon emission can be achieved within light-activated charge separation and recombination occurring prior to the formation of singlet and triplet excited states, independent from the radiative decay lifetime. Light-activated charge separation is a photoelectronic process where electrons are separated from their original atoms or molecules to generate electron–hole pairs.⁷⁸ These electrons or holes can be captured in energy traps and, upon thermal stimulation, recombine to form exciplexes in emissive excited states. The duration of delayed photon emission, extending from several hours to days, can be achieved by the efficient trapping and gradual release of these charge carriers, resulting in persistent luminescence.

Leveraging the charge separation and recombination strategy, an advanced persistent luminescence system was developed in 1996 with inorganic materials, utilizing SrAl₂O₄ doped with europium and dysprosium.⁷⁹ These dopant ions play a crucial role in splitting energy levels to facilitate the formation of electron or hole traps. However, replicating this thermally stimulated trap-release cycle poses a significant challenge in organic materials. This difficulty arises from the high binding energy of organic excitons, in both singlet and triplet states, which can reach several hundred millielectron volts (meV), attributed to the characteristics of Frenkel excitons.⁸⁰ In contrast to the relatively low binding energy of inorganic Wannier–Mott excitons, this makes the separation of tightly bound excitons into free holes and electrons upon photoexcitation at room temperature particularly challenging in organic systems. The phenomenon of organic molecule photo-ionization was first reported through sequential two-photon absorption.⁸¹ The initial photon absorption leads to an excited intermediate state, followed by the secondary photon absorption resulting in an ionized state with energy surpassing the ionization potential. The obtained radical cations are stabilized in the rigid environment within polymer matrices, isolated from electrons that undergo accumulation and gradual reverting combination with radical cations, producing exciplexes of the guest molecule for delayed emission. This system has been shown to facilitate long persistent luminescence (LPL) exceeding ten hours at 20 K. Nonetheless, this specific two-photon ionization process necessitates both an intense excitation source and extremely low temperatures.^82,83

Drawing on the principles of charge-separated states in organic photovoltaics,⁸⁴ Chihaya Adachi and colleagues unveiled a blend of organic donor and acceptor molecules in 2017 that demonstrated prolonged photon emission, termed organic long persistent luminescence (OLPL).²⁷ This was achieved through electron–hole pairs generated in a system conducive to charge transfer under weak photo-irradiation (Fig. 1). The proposed mechanism for creating a charge-separated state involves electron transfer between blended donor and acceptor molecules, followed by the accumulation of photo-induced radical cations and anions (Fig. 2a). Once a photo-induced electron transits from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO) of the electron-acceptor, charge transfer occurs, forming donor cations and acceptor anions. The subsequent diffusion of radical anions through charge hopping among acceptors separates the cations–anions pairs, creating a stable charge-separated state (Fig. 1 and 2b). The separated cations–anions pairs recombine to form exciplexes with a distribution of 25% singlets and 75% triplets. Consequently, this leads to the observation of both fluorescence and phosphorescence post-recombination.^85,86 Unlike the two-photon ionization process, the recombined exciplexes exhibit a delayed photon emission linearly dependent on the excitation power with a slope of one, signifying a charge dissociation within the charge transfer state from single-photon absorption.^87,88 To realize long-lived charge-separated states that enhance delayed exciplex emission, several critical factors must be considered within material design strategies. First, the efficiency of photo-induced charge separation is paramount, as excited states may quickly dissipate via radiative or non-radiative decay. Utilization of molecules with strong electron-donating and electron-accepting properties is recommended for constructing the charge separating system. Examples include N,N,N′,N′-tetramethylbenzidine (TMB) as the donor paired with 2,8-bis(diphenylphosphoryl)dibenzo[b,d]thiophene (PPT) as the acceptor, or 4,4′,4′′-tris[phenyl(m-tolyl)amino]-triphenylamine (m-MTDATA) with PPT as the donor–acceptor system.⁷¹ Second, delayed photon emission is susceptible to quenching by air due to the instability of photo-induced radical anions and cations in the presence of water and oxygen. Thus, encapsulating the charge-separated system is crucial to maintain its stability.⁸⁹ Lastly, the duration of photon emission is heavily influenced by the doping concentration of donor. Insufficient doping may result in diminished emission, whereas excessive doping increases the likelihood of recombination between radical cations and anions, thereby reducing the duration of photon emission. Thus, meticulous control over donor-doping concentration is essential to prolong the lifetime of delayed emission.

Fig. 1 Schematic mechanism of charge separation and recombination process for delayed photon release. Upon light irradiation, electrons (grey circles) in acceptor were excited from HOMO to LUMO (i), followed by electron transfer (ii) from HOMO of donor to HOMO of acceptor, thus forming acceptor radical anions. The negative charges undergo diffusion between acceptors (iii), where long-lifetime radical charge pairs are generated. In this mechanism, the light-activated photon release is indeed slow recombination (iv) of radical anions and radical cations to produce exciplexes in both S₁ and T₁ with light emission (v).

Fig. 2 Representative mechanism for charge separation through different types of charge carrier. (a) Schematic presentation of charge separating process for n-type charge carrier. (b) Schematic presentation of charge diffusion through n-type charge carrier and typical chemical structure for donor and acceptor in n-type OLPL system. (c) Schematic presentation of charge separating process for p-type charge carrier. (d) Schematic presentation of charge diffusion through p-type charge carrier and typical chemical structure for donor, acceptor and trap in p-type OLPL system. CT: charge transfer; CS: charge separation; CR: charge recombination.

While charge separating systems can be shielded from quenching environments through methods like crystallization or encapsulation, challenges related to processability and absorption wavelength have significantly restricted the use of these materials in biomedical applications. To address these limitations, Chihaya Adachi et al. introduced a p-type, air-stable, amorphous OLPL model, pioneering the use of radical cation for diffusion over the traditional radical anions in the charge-separating process.⁹⁰ The susceptibility of the TMB:PPT system to air is largely due to the high likelihood of interactions between mobile radical anions and oxygen. Altering the mobile charge can prevent this electron despoilment by oxygen during charge accumulation, thereby diminishing non-radiative decay of separated charges.⁹¹ Cationic electron acceptors were chosen as the electron carriers, owing to their lower LUMO compared to the reduction potential of oxygen (−3.5 eV).⁹² Upon photo-excitation, cationic electron acceptors capture electrons to form neutral radicals, further stabilizing the charge-separation state by diminishing the Coulombic interaction among radicals (Fig. 2c).⁹³ To optimize absorption and emission characteristics, spectra tuning can be achieved by pairing donors and acceptors with a suitable energy gap between the LUMO of acceptor and the HOMO of donor. Additionally, it is crucial to maintain a specific HOMO gap between donors and acceptors to facilitate the formation of a CT excited state. Enhancements via the integration of hole-trapping materials have significantly extended the lifetime of delayed photon emission in such p-type system.⁹⁴ The electron donor (TPBi), cationic acceptor (TPP⁺), and trap (TCTA) molecules are strategically aligned to exhibit a cascading HOMO and LUMO relationship to efficiently diffuse and stabilize the holes (Fig. 2d). For effective hole trapping, the hole-trapping material typically necessitates an energy gap approximately 0.5 eV shallower than the HOMO level of the acceptor.

Chihaya Adachi et al. advanced the development of multi-color OLPL in 2018.⁹⁵ They introduced fluorophores with varying emission properties into the charge-separating matrix to tune the exciplex emission wavelengths. The expansive emission spectrum of the TMB:PPT system overlaps well with a diverse array of fluorescent emitters, facilitating energy transfer from the exciplex to the emitter dopants through Förster energy transfer (FRET). This process significantly enhances color purity, photoluminescence quantum yield, and emission duration. By 2020, the same group unveiled molecular selection strategy for optimizing OLPL within this organic charge-separating system.⁹⁶ They described the energy levels as a combination of the locally excited state of donors/acceptors (LE_D/LE_A) and the charge transfer excited state (CT) within the recombined exciplex. The donor's lowest triplet locally excited state (³LE_D) modulates OLPL emissions by altering the energy gap between the singlet charge transfer state and itself (ΔE (¹CT − ³LE_D)). On the one hand, a minimal ΔE (¹CT − ³LE_D) facilitates the cascade energy state transition from LE to CT and then to charge separation state (CS) (Fig. 3a), while large ΔE (¹CT − ³LE_D) impedes populating of ¹CT and exhausts the absorbed energy through phosphorescence from ³LE_D (Fig. 3b).⁴⁰ On the other hand, after charge recombination, exciplex in ¹CT maintains ISC towards ³LE. A small ΔE (¹CT − ³LE_D) promises efficient exciplex emission in singlet state (Fig. 3c), while large ΔE (¹CT − ³LE_D) can lead rapid ISC from ¹CT to ³LE_A and impede RISC from ³LE_D to ¹CT, which finally cause spin-forbidden radiative decay with low emission efficiency (Fig. 3d).⁹⁷

Fig. 3 Influence of ΔE(¹CT − ³LE_D) on charge separation and exciplex emission. (a) Schematic diagram of charge separating process from LE state to CS state benefited from small ΔE(¹CT − ³LE_D). (b) Schematic diagram of charge separating process from LE state to CS state suppressed by large ΔE(¹CT − ³LE_D). (c) Schematic diagram of exciplex emission process for LPL in CT state benefited from small ΔE(¹CT − ³LE_D). (d) Schematic diagram of exciplex emission process for LPL in CT state suppressed by large ΔE(¹CT − ³LE_D). Red arrow: CS/LPL-favored process; Blue arrow: CS/LPL-unfavored process. LE: localized excitation; LPL: long persistent luminescence.

Tang et al. developed a similar charge-separating system for delayed photo emission in 2020 using a phosphonium bromide salt as a strong electron acceptor and N,N-dimethylaniline (DMA) as the electron donor (Fig. 4a).⁹⁸ The triphenylphosphine cation, previously known for its phosphorescent properties, can initiate D–A charge transfer and act as a neutral electron carrier within a charge separating state with DMA (Fig. 4b). The crystallizing nature and phenyl ring decoration of the phosphonium cation contribute to the stabilization of excited radicals, facilitating delayed emission lasting up to 7 h. This demonstrates the triphenylphosphine-cation-type OLPL as a viable option for cost-effective delayed photon emission applications. In 2022, this type of OLPL evolved into a single-component system through the introduction of a novel class of phosphonium salts that integrate both donor and acceptor functionalities within a single molecule, significantly streamlining the production of OLPL materials (Fig. 4c).⁹⁹ The power law decay of this single component system pointed to the charge separation and recombination mechanism, but underwent a more complicated molecular interaction. Reasonable aggregation emerges as a key contributor to the formation of a dimeric excited state for charge separation and stabilization.¹⁰⁰ Upon photo-excitation, these dimeric excitons may absorb an additional photon to become ionized, producing separated charges via a multiphoton process rather than induction by charge transfer (Fig. 4d). The crystalline and delocalized phenyl environment afford stability to both cationic and anionic radicals within the system, prolonging charge separation periods and thus enabling sustained delayed photon emission.

Fig. 4 Representative mechanism for charge separation in phosphonium-bromide-salt-based OLPL system. (a) Typical chemical structure of donor and phosphonium-bromide-salt acceptor in D–A doping system. (b) Schematic presentation of charge generation and diffusion process for phosphonium-bromide-salt-based D–A doping system. (c) Typical chemical structure of phosphonium-bromide-salt based D–A single component system. (d) Schematic presentation of charge generation and diffusion process for D–A single component system.

2.2. Light-induced generation, stabilization and conversion of triplet excitons

Light activation on optical materials can generate excitons with varying spin multiplicities, undergoing energy decay across different timescales.^101–105 Among these, triplet excited states, often referred as spin-forbidden states, exhibit significant potential for delayed photon emission due to their ability to undergo radiative decay or convert to other emissive states over an extended lifetime.^29,106–109 Due to their unsupported transition dipole moments towards the ground state, triplet excitons demonstrate a slow radiative decay to emit light known as phosphorescence, one of the earliest recognized mechanisms for delayed emission.¹¹⁰ Additionally, non-radiative processes that convert the spin multiplicity of triplet states back to singlet can expedite the delayed radiative decay from singlet excited states, a phenomenon known as delayed fluorescence.^111,112 Conversion of triplet excitons can also occur between excimers within different energy states, which further extends the phosphorescent emission lifetime (Fig. 5).^113,114 Recent advancements in both phosphorescent and delayed-fluorescent materials have provided essential insights into structural design strategies and aggregation configurations that effectively regulates triplet excited states during both photoexcitation and deactivation.¹¹⁵ The extensive application and diverse regulatory strategies underscore the importance of triplet excitons as the most controllable species for achieving delayed photon emission.^116–118 Herein, we aim to provide a comprehensive overview of effective methods for generating, stabilizing, and converting triplet excitons to other emissive states.

Fig. 5 Schematic mechanism for delayed photon emission via generation, stabilization and conversion of triplet excitons. Generation: ISC from S₁ to T_n; Stabilization: suppressing nonradiative decay and quenching; conversion: RISC from T₁ to S₁ and trap/detrap between T and T*. In this mechanism, the delayed photon emission is indeed gradual exhaustion of triplet excitons through radiative decay.

Triplet excitons, which are rarely produced directly from excitation without ionization, are typically generated from the lowest singlet excited state (S₁).¹¹⁹ Upon photoexcitation, molecules in S₀ absorb photons and transit to higher singlet states (S_n), where rapid internal conversion (IC) converts the excitons to S₁ according to Kasha's rule.¹²⁰ These excitons in S₁ can alter their spin multiplicity to form triplet excited states (T_n) through various non-radiative processes, such as singlet fission (SF) and intersystem crossing (ISC).^121–124 Due to stringent energy level distribution requirements of SF that are difficult to achieve in optical materials, ISC is often the more significant factor influencing triplet exciton generation. For effective population of triplet excitons, ISC must occur at a sufficiently rapid rate to compete with fluorescent decay and IC in S₁, given the short lifetime of S₁. The rate of ISC (k_ISC) is a fundamental characteristic of organic molecules, heavily dependent on their electronic configuration and energy level. It can be mathematically represented as:^125–127

In this formula, 〈S|Ĥ_SOC|T〉 denotes the spin–orbit coupling (SOC) between singlet and triplet states, ħ is the reduced Planck's constant, λ represents the total reorganization energy, k_B is the Boltzmann constant, T is the temperature and ΔE_ST is the energy gap between the involved singlet and triplet states.

Two key factors influential in structural design, which can be derived from this formula, are SOC and ΔE_ST (Fig. 6).

Fig. 6 Schematic mechanism and designing strategies for generation of triplet excitons. k_ISC is a key factor that determines population of triplet excitons, where SOC and ΔE_ST are influential in structure designing.

SOC, which describes the relativistic interaction between the spin and momentum degrees of freedom of electrons,^125,128 is often enhanced through the heavy-atom effect. As Z⁴ (Z is the atomic nuclear charge) is one of the coefficients in operator Ĥ_SOC, atoms with large atomic number (such as S,¹²⁹ Se,¹³⁰ Te,¹³¹ Cl,¹³² Br,¹³³ I¹³⁴) can significantly increase SOC due to their contribution to the matrix element 〈S|Ĥ_SOC|T〉 (Fig. 6a). These heavy atoms can be introduced as substitutions on aromatic ring or within ionic groups to pair with charged molecules, enhancing other triplet-forming processes.^135,136 According to El-Sayed's rule, effective SOC can also be achieved in organic molecules through difference in molecular-orbital configurations of S₁ and T_n that enhance orbital overlap.¹³⁷ For instance, S₁ states with a dominant (n, π*) characteristic show a significant difference compared to T_n states which maintain a (π, π*) characteristic. This difference leads to a large matrix element 〈S|Ĥ_SOC|T〉, and consequently improves the k_ISC (Fig. 6b). Thus, introducing n orbitals perpendicular to π orbitals can significantly strengthen the spin–orbit coupling and enhance ISC from singlet to triplet excited states.¹³⁸ In terms of chemical structure design, the involvement of n orbitals in electron distribution of S₁ states can be achieved not only through carbonyl or cyanic groups conjugated to aromatic skeletons but also via heteroatoms (such as O,¹³⁹ S,¹⁴⁰ N,¹⁴¹ P¹⁴²) embedded within the conjugated systems. These structures, rich in lone-pair electrons, convert the initially (π, π*) characterized molecular S₁ states into a hybrid of α_n (n, π*) + β_π (π, π*) states (where α_n + β_π = 1), creating a significant disparity in (n, π*) character between S₁ and T_n states.¹⁴³ Moreover, phosphorescent aromatic molecules substituted with siloxy or arylboronic groups have demonstrated a σ–π orbital hybridization due to the nonplanar confirmation of their triplet states.^144,145 This hybridization, supported by σ–π orbital interaction, contributes to spin–vibronic coupling, further enhancing SOC and the overall efficiency of ISC.

Another widely employed molecular design strategy to enhance ISC focuses on narrowing the energy gap between S₁ and T_n.^146,147 One approach is to bring singlet and triplet states closer by utilizing energy level splitting through intermolecular or intramolecular interactions. These interactions can cause the energy levels of both singlet and triplet states to divide into higher and lower intermediate states, altering the initial equilibrium geometries due to environmental or structural impacts on electron distribution.¹⁴⁸ This results in lower singlet states and higher triplet states, effectively reducing the ΔE_ST. In monomer systems, donor–acceptor (D–A) type molecules facilitate the introduction of intermediate states via charge transfer during the ISC process (Fig. 6c). Upon photoexcitation, D–A molecules involve both intramolecular charge-transfer states (¹CT and ³CT) and localized-excited states (¹LE and ³LE), creating favorable conditions for ISC between these closed intermediate states.¹⁴⁹ Moreover, strong vibronic coupling between adjacent ³CT and ³LE may indirectly couple ¹CT and ³CT to enhance ISC.¹⁵⁰ When considering intermolecular interactions, the overlap of excitonic orbitals among aggregated chromophores becomes a critical factor in inducing energy level splitting. Recent advances in phosphorescent materials have highlighted various intermolecular interactions that rely on specific chemical structures to enhance electronic overlap, including H-aggregation,¹⁵¹ n–π stacking¹⁵² and π–π stacking¹⁵³ (Fig. 6d). The aggregates formed through these interactions provide additional opportunities for monomer electronic orbitals to hybridize, thereby establishing optimal ISC channels. Furthermore, ΔE_ST can be minimized or even eliminated in radical pair excitons due to hyperfine coupling, where the singlet and triplet states of radical pairs are subjected to rapid spin exchanges influenced by internal nuclear spin magnetic fields (Fig. 6e).¹⁵⁴ However, the capability for hyperfine coupling in excited states has been observed in only a few materials, and the underlying mechanisms require further investigation.

It is important to recognize that an excessively small ΔE_ST or overly large SOC can negatively affect the stability of triplet states in ambient environment. When the triplet-to-singlet energy gap is too narrow, thermal activation may facilitate a reversible intersystem crossing, leading to premature depopulation of triplet excitons. Therefore, careful molecular engineering to achieve optimal ΔE_ST and SOC values is crucial for stabilizing long-lasting triplet states, which are essential for effective delayed photon emission.

Theoretically, delayed photon emission, particularly through phosphorescence, can be observed when triplet excitons are populated and undergo slow radiative decay to emit light. However, in practical material fabrication, there exists an energy-consuming competition among radiative decay, non-radiative decay and external quenching in the triplet states, expressed as:^52,115,155

where τ_T represents the lifetime of triplet states; k_r and k_nr are the rate constants for radiative and non-radiative deactivations of triplet exciton, respectively; and k_q is the rate constant for quenching influenced by environmental factors such as oxygen or humidity (Fig. 7). It is evident that a smaller sum of these rate constants results in a longer lifetime for the triplet excitons. Notably, k_r values are relatively low for pure organic molecules due to forbidden spin exchange, with their tuning range being restricted (<10⁰ s⁻¹). In contrast, the values of k_nr + k_q, influenced by intramolecular motion and various triplet quenching processes, are significantly higher (>10² s⁻¹) in conventional aromatic systems. Therefore, it is practical to aim for stabilized triplet excitons with effectively impeded k_nr + k_q.

Fig. 7 Schematic mechanism and designing strategies for stabilization of triplet excitons. Suppression on k_nr and k_q through regulation on intramolecular motion and intermolecular interaction benefits the improvement of τ_Triplet.

In a monomer triplet exciton, k_nr can be expressed as:^156–158

where f_S represents the prohibition factors arising from spin configuration changes during non-radiative decay, f_V is the Franck–Condon factor and f_T is the temperature factor.

f _V is calculated as:

f_V = exp(−αΔE/ħω)

where α is a constant, ΔE is the energy difference between the ν = 0 vibrational level in the T₁ states and the ν = n vibrational level in the S₀ states, and ω represents the energy of an effective nonclassical vibrational mode involved in the nonradiative transition, often identified as the highest frequency vibration mode, such as the C–H stretching vibration in aromatic molecules.¹⁵⁵ Molecular engineering on the triplet excitons can effectively regulate f_S and f_V to decrease k_nr. f_S shares a similar spin–orbital Hamilton operator and integration of the singlet–triplet wave function with k_ISC, suggesting that strategies such as reducing the (n, π*) characteristic and minimizing heavy atom effects could be employed to suppress non-radiative decay. Therefore, a well-designed Ĥ_SOC and a distinct configuration difference between the singlet and triplet excited states are essential to balance the generation and stabilization of triplet excitons. Additionally, a reduction of f_V can be achieved through the deuteration of aromatic hydrocarbons because of a smaller ω in C–D vibration (Fig. 7a). For aggregated triplet excitons, intramolecular motions and diffusional collisions with quenchers are restrained in a highly ordered aggregation, suggesting that structural rigidification through extensive intermolecular interactions can substantially decrease both k_nr and k_q (Fig. 7b).¹⁵⁹

Recently, a variety of molecular design strategies have been employed to create rigid aggregates, including crystallization,^160–162 host–guest complexation^163,164 and clusterization.^165,166 These strategies leverage noncovalent interactions such as hydrogen bonding, halogen bonding and π–π interaction to create materials with stabilized triplet states. However, it is worth noting that concentrated triplet excitons in aggregates can inadvertently lead to additional triplet quenching. In some cases, an amorphous matrix with a high T₁ state energy may be more conducive to preventing triplet excitons from concentration-dependent quenching and exhaustion through triplet–triplet energy transfer (Fig. 7c).¹⁶⁷ Another approach to rigidify the molecular structure for reduced k_nr and k_q is polymerization^168,169 and cross-linking.^170,171 This approach involves the covalent conjugation of chromophores, which imposes spatial constraints on intramolecular vibration or rotation, and promotes a symmetrical dispersion of triplet excitons within the polymer matrix to avoid triplet quenching (Fig. 7d). These designing strategies, guided by the variables in the formula, have been shown to effectively reduce non-radiative decay and quenching in triplet excitons, thereby extending the lifetime of triplet states and enhancing radiative decay for delayed photon emission. Furthermore, the intermolecular or intramolecular interactions between chromophores can be modulated under different electronic states, making packing configurations difficult to predict. Further investigations into the electronic states of electron-donating or electron-accepting sites within the chemical structure of triplet excitons may offer new opportunities to understand precise packing configurations and contribute to the establishment of universal principles for the chemical design of long-lived triplet excitons.¹⁷²

Although significant progress has been made in molecular engineering to regulate the generation and stabilization of triplet excitons,¹⁶³ a key challenge remains in achieving highly efficient delayed photon emission from purely organic triplet excitons due to their spin-forbidden nature.¹⁷³ The inherent spin multiplicity of these excitons limits their emissive properties, making the conversion of triplet excitons to other emissive excited species a viable approach for harnessing the triplet-state energy for delayed emission (Fig. 8).¹¹⁹ The most commonly employed method for this conversion is from triplet to singlet excited states via RISC, which facilitates a process known as thermally activated delayed fluorescence (TADF).^174–176 In TADF emitters, ΔE_ST is a critical factor for efficient triplet-to-singlet conversion, as it significantly influences RISC efficiency. Thus, precise control of ΔE_ST through molecular engineering is a key approach to achieve nearly complete utilization of triplet-state energy for efficient photon emission.

Fig. 8 Schematic mechanism and designing strategies for conversion of triplet excitons. Rational regulation on energy gap between S₁ and T₁ or between T₁ and accelerates the conversion of triplet excitons, which improves the emission efficiency and prolongs the emission lifetime, respectively.

The first structural design for organic materials to thermally overcome ΔE_ST was reported by Adachi et al., who employed steric hindrance to create a significant dihedral angle between donor and acceptor (Fig. 8a).¹⁷⁷ This arrangement spatially separates the highest occupied and lowest unoccupied natural transition orbitals. Originally, this strategy aimed to tune the ratio of singlet to triplet excitons in the electroluminescence process to enhance emission efficiency, without considering the lifetime of delayed photon emission. Recent advancements in research on organic afterglow materials have further refined the triplet-to-singlet conversion process, balancing the RISC efficiency with emission lifetime.¹⁷⁸ Since rapid RISC can undermine the stability of triplet excitons, a novel approach introduced a low-lying triplet state as an exciton trap to extend delayed emission through a cascade exciton transformation from to T₁, and then from T₁ to S₁.¹⁴⁰ To facilitate this multiple process, enhanced ISC and RISC are essential. This was achieved by incorporating active nonbonding electrons and twisted donor–acceptor structure into the molecular design. Additionally, strategies for generating chemical structures conducive to trap creation and establishing shallow trap depth (E_TD) were implemented through selective donor usage and limited π–π stacking, aiming to stabilize the excitons and facilitate their release through thermal energy fluctuation (Fig. 8b). This design strategy synergizes the benefits of energy trapping and spin conversion to achieve efficient organic afterglow, providing a novel approach to delayed photon emission through the conversion of triplet excitons.

In addition to thermal activation for triplet-to-singlet conversion, triplet excitons can also be photoexcited to a charge-separating state, achieving delayed emission through ionization and charge recombination processes.¹⁷⁹ It has been observed that an inefficient TADF compound, when doped in host matrices, exhibited a prolonged T₁ lifetime and consequent large T₁ populations due to its mediate ΔE_ST and rigid molecular structure. This substantial T₁ population is advantageous for photoinduced charge generation via successive two-photon ionization processes. The recombined exciplex emitted light from both the S₁ and T₁ state with an improved ratio, leveraging the reserved TADF feature to facilitate prolonged photon release lasting up to thousands of seconds. However, the specific design strategies to achieve a suitable ΔE_ST for this mechanism remain unclear. Further research is necessary to develop a reliable structure that can form stabilized triplet excitons capable of efficient ionization and charge separation within host matrices.

2.3. Light-activated chemical traps

Numerous methods have been developed to efficiently store light energy for delayed photon emission.^180,181 Beyond radical pairs and stabilized triplet excitons, chemical traps can serve as “optical batteries” by converting absorbed light energy into chemical energy.¹⁸² Chemical traps are typically high-energy chemical intermediates that continuously undergo fragmentation to form exciplexes, ultimately leading to delayed emission. These intermediates are characterized by metastable bonds formed during energy transforming reactions, which are primed to release stored chemical energy through photon generation. Recent advancements in the structural design of chemical traps have primarily focused on oxidized systems featuring high-energy dioxetane, which is commonly produced through the addition of singlet oxygen (¹O₂) to aromatic structures.^183–185 To operationalize this light-activated energy trapping system, two main components are essential: a photosensitizer and a chemical trap precursor.^77,186 Upon photoexcitation, the photosensitizer absorbs light energy and acts as a generator of ¹O₂, initiating the transformation of photon energy into reactive oxygen species (ROS). Subsequently, a chemical trap precursor functions as a singlet oxygen acceptor, undergoing oxidation to form a metastable dioxetane intermediate that store the chemical energy. These intermediates then undergo controlled split, which benefits a long-lasting production of emissive exciplexes to release the stored energy (Fig. 9). Herein, we aim to elucidate the generation process of light-activated chemical traps across various types of chemical trap precursors and explore their underlying mechanisms.

Fig. 9 Schematic mechanism of delayed photon emission via generation and decomposition of chemical traps. Oxygen sensitization was conducted by initiators excited into triplet state. The produced ¹O₂ can reacted with chemical trap precursors to generate chemical traps, which subsequently decompose into emissive species. In this mechanism, the delayed photon emission is indeed gradual release of chemical energy stored in dioxetane intermediates.

Originated from studies on bioluminescent substrates, dioxetane-based chemical intermediates has garnered significant interest as optimal structures for prolonged photon emission.^183,187 Substrates like luciferin and coelenterazine are believed to undergo catalyzed oxygen addition at their deprotonation sites, leading to the closure of peroxy rings through transesterification, and resulting in the formation of carbonyl-substituted dioxetane intermediates.^188–190 The subsequent splitting of these dioxetane rings into excited species, facilitated by CO₂ escape, leads efficient emission, demonstrating a feasible method for converting chemical energy into photon generation. Motivated by this natural light-emitting process via chemiexcitation, light-activated chemical traps for delayed photon emission are predominantly designed to undergo dioxetane formations from chemical trap precursors under photoexcitation, followed by dioxetane decomposition for emissive species (Fig. 10a).^191–194

Fig. 10 Mechanism and designing strategies for chemical trap generation. (a) Schematic presentation of components and mechanism for chemical trap generation. (b) Schematic illustration of electron-releasing group for chemical trap generation. (c) Schematic illustration of thermal stabilization for chemical trap stabilization.

Unlike bioluminescent system, achieving deprotonation without biocatalysis to create reactive carbon species from chemical trap precursors poses significant challenges. One feasible alternative for forming dioxetane intermediates through light activation is to convert the addition of oxygen on reactive carbon species to the addition of ROS on carbon sites that favor peroxidation, typically vinyl bonds. In this process, light-activated singlet oxygen can rapidly react with the vinyl bond through a radical reaction, which can be accelerated by stabilization of the resulting radicals through introduction of electron-releasing group, such as alkoxy group, tertiary amino group and aromatic group, on the vinyl carbon (Fig. 10b).^183,195

The resulting dioxetanes contain high strain energy, as anticipated from dihedral angle comparisons with hydrogen peroxide, which may lead to incident decomposition during the stretching of the O–O bond.^196,197 Consequently, enhancing the thermal persistency of the dioxetane ring is crucial to ensure that these compounds are stable enough for handling at room temperature, suitable for long-term energy storage.¹⁹⁸ The stretching of the O–O bond is heavily dependent on the rotation around the C–C bond in the dioxetane ring. When there is significant twisting of the C–C bond, the two oxygen atoms are isolated, forming a transient diradical that subsequently leads to the cleavage of C–C bond.¹⁹⁹ The introduction of steric hindrance at the axial carbon positions can inhibit this free rotation through mutual obstruction between R₁ and R₃, as shown in Fig. 10c(i). According to this strategy, structural design of equatorial methylenic hydrogens such as adamantylidene^200,201 and bulky alkyl groups,^202,203 has been realized to obtain stable dioxetanes. Additionally, bicyclic dioxetanes with annelated multi-membered rings can form rigid planar structures that resist decomposition through similar mechanism.²⁰⁴ Moreover, involvement of electron-releasing group at the ortho-position of phenyl adjacent to dioxetane also provides steric hindrance against rotation. The extra conjugation effect via ortho-substitution joins the phenyl ring to C–O bond of dioxetane at the potential energy minimum, where energy barriers exist to fix the configuration, thus stabilizing the dioxetane (Fig. 10c(ii)).²⁰⁵ These design strategies featured by modification of “stabilizers”, have been proved to extend the process of delayed decomposition for emissive species, thereby underscoring the significant potential of stabilized dioxetane intermediates to serve as energy traps for delayed photon emission.

Another important consideration for using dioxetanes as chemical traps is whether they meet the energy requirements necessary for chemiexcitation. First, it is essential to confirm that the generation of excited species during the dioxetane decomposition is a thermodynamically feasible process. To explore this, tetramethyldioxetane was chosen as a model compound without functional substitutions to estimate the available energy for chemiexcitation (Fig. 11a).^183,206 The reaction's heat (ΔH₀) and activation enthalpy (ΔH_a) were determined to be −61 and 25 kcal mol⁻¹, respectively. This analysis shows a total energy decrease of 86 kcal mol⁻¹ in the transition state for thermal decomposition, which is sufficient to produce a singlet/triplet excited acetone molecule. Likewise, ΔH₀ for the thermal decompositions of typical dioxetanes has been theoretically calculated to range from 69 to 90 kcal mol⁻¹, while their ΔH_a were measured to vary around 20–30 kcal mol⁻¹. Thus, these dioxetane-based chemical traps can indeed provide sufficient energy for chemiexcitation during thermal decomposition.

Fig. 11 Mechanism and designing strategies for production of emissive species during chemical trap decomposition. (a) Schematic presentation of thermodynamic analysis of decomposition process. (b) Schematic illustration of chemical trap decomposition. (c) Schematic diagram of two decomposition route for dioxetane-based chemical trap. Top: potential curve for thermolysis; bottom: potential curve for CIEEL. (d) Schematic presentation of regulation on LUMO coefficients to avoid quenching in aqueous solution.

Second, the resulting excited species must be emissive to ensure efficient delayed photon emission with minimal non-radiative decay (Fig. 11b). The simplest pathway for dioxetane decomposition through thermolysis typically involves a twisted diradical-like transition state, which includes a crossover of the diradical path to the triplet-excited product path.^207,208 This chemiexcitation process can be described as a transition from the lowest singlet potential surface (S₀) to the lowest triplet surface (T₁), with these surfaces crossing twice during the transition state. The S₀–T₁ spin conversion is efficient at the first crossing point (O–O bond cleavage) but become irreversible at the second crossing point (C–C bond cleavage) due to reduced spin–orbit coupling caused by altered chemical geometry. This spin forbiddance results in the eventual decomposition on the T₁ surface, predominantly yielding triplet-excited carbonyl products that are less emissive than their singlet counterparts (Fig. 11c(i)). Recent advances in catalyzed dioxetane decomposition have revealed a promising method for enhancing the production of singlet-excited carbonyl through chemically initiated charge transfer or electron exchange.^209–211 This process, known as chemically initiated electron exchange luminescence (CIEEL), is facilitated by either intermolecular or intramolecular introduction of electron donor, which induce a charge transfer state in dioxetanes, leading to efficient luminescent emission. This phenomenon is largely due to the excited singlet surface (S₁) sharing similar transient energy to that of the S₀.¹⁸³ Typically, dioxetanes featuring a phenoxide anion exhibit effective energy crossover between S₁ and S₀ as shown in Fig. 11c(ii). The reduced activation energy required to reach transient state in S₀ (E_a1) accelerates the formation of a transient diradical at the crossing point, thus leading direct dioxetane decomposition on the S₁ surface and producing singlet-excited carbonyl.²¹² Given that emissive species are efficiently generated during the CIEEL process, CIEEL-type dioxetanes are recognized as highly promising chemical traps for delayed photon emission. Additionally, the environmental quenching of excited carbonyls by ambient substrates can significantly reduce the excitonic population. It has been shown that the LUMO coefficient on the ester carbonyl moiety of the decomposition product is closely associated with reduced CIEEL emission intensity in aqueous system (Fig. 11d).²¹³ Consequently, structures characterized by D–A configurations based on phenoxide anion are commonly employed in dioxetanes to introduce an electron-withdrawing effect on the easter group during chemiexcitation, thereby enhancing the quantum yield of CIEEL.

Thanks to insightful investigations into chemiexcitation mechanisms and dioxetane design principles, molecular engineering of light-activated chemical traps has advanced significantly in recent years.^{31,77,186,214} This development has led to the creation of several distinctive chemical trap precursors, each with unique dioxetane-forming pathways and splitting kinetics. Tracing back to the first report of persistent luminescence in polyphenylene vinylene (PPV)-based nanoparticles by Rao et al.,¹⁸² the mechanistic investigation and structure design of PPV launched the initial generation of light-activated chemical traps, heralding a new era in pure organic delayed photon emission. The pioneering study on the mechanism of afterglow luminescence in PPV by Pu et al. revealed that PPV and its analogs are characterized by the polymerization of vinylenes in semiconducting polymers (SPs), featuring sp²-hybridized carbon atoms that extend the π-conjugation throughout the polymer chain.³¹ This polymerized structure enhances the electron-releasing capacity, facilitating rapid addition of ¹O₂ on random vinylenes to form dioxetane-based chemical traps (Fig. 12a). To optimize the bright emission of PPV in aqueous solution, both PPV and photosensitizers (silicon 2,3-naphthalocyanine bis (trihexylsilyloxide), NCBS) were encapsulated within water-soluble nanoplatforms. Upon light irradiation, this configuration enables efficient ¹O₂ generation and subsequent PPV-dioxetane formation, followed by thermal decomposition into PPV-aldehyde exciplexes. The interaction between these exciplexes and the photosensitizers redshifts the delayed emission wavelength to the near infrared (NIR) range. Furthermore, molecular integration of PPV analogs with photosensitizers has been achieved through copolymerization of these key components.²¹⁵ This covalent conjugation is designed to reduce the ¹O₂ diffusion distance and facilitate intramolecular energy transfer, thereby speeding up the generation and decomposition of chemical traps, resulting in enhanced delayed NIR photon emission.

Fig. 12 Representative chemical trap precursors for chemical trap generation and their corresponding emissive products. (a) PPV and its analogs. (b) AEE and its analogs. (c) DO/SO and their analogs. (d) Porphyrin and its analogs. (e) Cyanine and its analogs. (f) Thiophene-based semiconducting polymer.

Inspired by the crucial role of vinyl bond in PPV for dioxetane generation, adamantylidene enol ethers (AEEs), originated from chemiluminescence substrates, were first employed as small molecule ¹O₂ acceptor in chemical trap systems by Ding and his coworkers.²¹⁶ They have since become a popular class of chemical trap precursors containing essential dioxetane-favorable moieties. Given the modest ¹O₂ generating capacity of these precursors, the inclusion of photosensitizers is essential to initiate dioxetane formation and decomposition within this system. The electron-donating structures contributed by alkoxy and phenolic groups exhibit strong reactivity towards the generated ¹O₂, forming chemical traps that are further stabilized by the antirotating adamantylidene structure. The subsequent deprotonation of the phenolic hydroxyl group enhances electron-donating behavior, leading to rapid decomposition and the production of emissive carbonyls (Fig. 12b). Versatile substitutions at the ortho position of the phenol group allow for the tuning of emission wavelength and improvements in luminescent quantum yield. Compounds such as acrylic acid,²¹⁷ quinolinecarboxylic acid,²¹⁸ methyl acrylate,²¹⁹ and acrylonitrile²²⁰ serve as electron-withdrawing groups that facilitate the creation of a charge transfer state in chemiexcited species, thus regulating the energy gap and altering electron distribution. To manage the splitting kinetics of this chemical trap, strategies such as intramolecular hydrogen bonding and halogen substitution are employed to regulate the pK_a of the phenolic hydroxyl group, which determines the rate of deprotonation.^221,222 Therefore, AEEs represent promising chemical trap precursors, offering a wide range of possibilities for structural customization to optimize their performance in various applications.

Similar dioxetane formation and decomposition processes are observed in the vinyl bond of 1,4-dioxin and 1,4-oxathiin.^223–225 The saturated substitution of the vinyl carbon with alkoxy and phenyl groups facilitates radical release during ¹O₂ addition and modulates the dioxetane-splitting kinetics through steric hindrance produced by the cyclized alkoxy groups. The p-(dimethylamino) phenyl moiety, serving as an effective electron donor, significantly enhances the chemiexcitation process, leading to highly efficient decomposition for singlet excitons (Fig. 12c).²²⁶ In recent years, Pu et al. have utilized these chemical trap precursors in a generic approach to develop afterglow imaging probes.¹⁸⁶ They observed that the delayed emission intensity from the dioxin ring was significantly higher than that from the oxathiin ring, which can be attributed to enhanced spin–orbit coupling producing inefficient triplet emitters. To improve light-emitting properties, Li et al. transformed the initial phenyl-substituted oxathiin into a fused molecule, incorporating oxathiin as the chemical trap and triphenylamine as the emitter.²²⁷ The incorporation of the triphenylamine group not only enhances the oxidation of oxathiin by ¹O₂, but also increases the population of singlet excitons during dioxetane decomposition. Overall, dioxin or oxathiin rings, through thoughtful structural design, have been proved as exceptional chemical trap precursors that undergo singlet chemiexcitation, resulting in satisfactory brightness for delayed photon emission.

Despite the prevalent chemical trap generation and decomposition in typical alkoxy-substituted aromatic structures, vinyl bonds in certain chromophores, such as porphyrin and cyanine, have also been demonstrated to be reactive acceptors of ¹O₂. Porphyrins, including hematoporphyrin monomethyl ether (HMME), Pyropheophorbide-a (Ppa), tetraphenylporphyrin (TPP) and their derivates, feature a conjugated nitrogen heterocyclic structure where vinyl bonds are situated in an electron-releasing environment.²²⁸ Recent studies by Ding et al. have shown that various ROS (¹O₂, ONOO⁻, ˙OH and O₂˙⁻) can interact with porphyrins to form two-dioxetane-involved intermediates.²²⁹ This chemical trap generation process involves the oxidation of αβ or αγ vinyl bonds located between the conjugated pyrrole rings. Porphyrins are capable of efficiently generating ¹O₂ upon light irradiation, enabling the establishment of a single component chemical trap system. The self-generated ¹O₂ reacts with the vinyl bonds in porphyrins to produce two-dioxetane-involved intermediates which then decompose and facilitate energy transfer to surrounding unreacted porphyrins (Fig. 12d). This chemiexcitation process endows porphyrins and their analogs with long lifetime energy trapping capacity and bright NIR delayed emission, indicating their significant potential as chemical trap precursors. Hemicyanines have been reported to undergo similar light excitation processes, resulting in the formation of dioxetane intermediates followed by chemiexcitation for delayed photon emission.²³⁰ Hemicyanines serve as both ¹O₂ generators and chemical trap precursors, undergoing oxidation at the π-bridge site between donors and acceptors. The resulting hemicyanine-dioxetanes are thermodynamically favorable for decomposition, facilitating energy transfer to emitters occurs for photon generation (Fig. 12e).

Beyond dioxetanes formed from vinyl bonds, high-energy intermediates that produce emissive species can also be formed in thiophene-based SPs such as TTFQx²¹⁴ and PFODBT.²³¹ These newly discovered chemical trap precursors tend to form sulfur-containing intermediates, which are distinctly different from dioxetanes. In this mechanism, ¹O₂ preferentially attacks the C–S bond in thiophene, probably due to an oxidation preference for radical addition on sulfur atom rather than the vinyl bond. The resulting high-energy sulfur-containing four-membered ring spontaneously splits into a (3-oxoprop-1-enyl) ethanethioate-like product through intramolecular charge transfer, thereby generating excited SP species for delayed photon emission (Fig. 12f).²³² Although thiophene-based chemical trap precursors are capable of efficiently generating ¹O₂ on their own, their eventual emission intensity is typically lower than that of PPV-dioxetanes. This phenomenon can be attributed to enhanced ISC for triplet excited decomposition products, which are inherently less emissive.

2.4. Light-activated delayed photon emission for bioimaging

Fluorescent imaging suffers from a limited signal-to-background ratio (SBR) due to autofluorescence from biological tissues during signal acquisition.^22,233,234 Light-activated delayed photon emission overcomes this limitation by separating the signal generation and acquisition steps, thereby avoiding autofluorescence. The continuous emission after the removal of external light excitation allows signal acquisition to occur when autofluorescence is not generated, significantly reducing the background noise in imaging results. This improvement in SBR highlights the great potential of light-activated delayed photon emission in bioimaging. Given that delayed emission can be achieved through various mechanisms and design strategies, we provide a summary of recent advances in bioimaging that utilize these methods described above to achieve light-activated delayed photon emission.

Charge separation and recombination have been proven to be reliable strategies for achieving delayed generation of emissive exciplexes in pure organic materials. However, the implementation of rational doping method and isolated encapsulation is crucial to protect the separated charges from short-lived diffusion and ambient quenching.⁹⁰ Meeting these requirements is particularly challenging when constructing bioimaging probes, which need to possess dynamic nanoscale sizes and operate in complex biological environments filled with water and charge-transfer-capable biomolecules. Although several studies have demonstrated that a charge separating state is achievable in nanomaterials applied in vivo,^235–237 efficient quenching by water has been observed, which rapidly transforms light excitation energy into non-emissive species. This phenomenon may enhance photosensitization but is unfavorable for charge recombination. It should be noted that an exciplex-generating system has been realized in nanoparticles for bioimaging, where both charge separation and recombination occur to produce emissive exciplexes in a biological environment but with short-lived charge diffusion. Li et al. designed a host–guest doping material (M-CH₃) consisting of phenothiazine derivatives as guests and their corresponding dioxide derivatives as hosts.²³⁸ In this host–guest system, charge separation occurs between the matched energy levels of these two moieties via charge transfer in a co-crystalized structure. Subsequent charge recombination produces triplet exciplexes emitting at 500 nm with an improved phosphorescence quantum yield compared to either single component. To ensure the maintenance of efficient charge separating and recombining system, nanocrystal of M-CH₃ was encapsulated in biocompatible amphiphilic copolymer PEG-b-PPG-b-PEG (F127). The obtained nanoparticles exhibited persistent emission in aqueous media up to 25 min after 1 min of irradiation at 365 nm and were successfully applied in pre-irradiated subcutaneous imaging and time-resolved biodistribution imaging in mice. The prolonged emission lifetime was attributed to the similar T₁ energy levels between the acceptor and its corresponding exciplex, which allows triplet excitons to recycle between these two states. Despite significant efforts in developing novel probes and imaging technology for persistent emission, autofluorescence-free bioimaging based on charge separation and recombination remains rare. This is due to the stringent requirements for specific molecular interactions and an inert environment to achieve long-lived charge diffusion necessary for delayed photon emission.

The generation and stabilization of triplet excitons is another effective method for delaying photon generation, enabling autofluorescence-free bioimaging. The emission behavior of triplet excitons, known as phosphorescence, is spin-forbidden and further slowed by additional energetic stabilization, thereby providing a delayed signal after the removal of excitation. In recent years, there have been significant advancements in the chemical design and packing mode regulation of phosphorescent materials. Strategies such as the heavy atom effect, halogen bonding, hydrogen bonding, H-aggregation and π–π stacking have shown promise in improving ISC through enhanced SOC or reduced ΔE_ST. Despite the progress in populating triplet excitons, the consumption of triplet excitons through triplet–triplet annihilation and triplet–polaron interaction is common in biological environments filled with triplet-quenching substrates.^239,240 This exciton depletion process can result in a decrease in phosphorescence intensity due to dominant non-radiative decay, leading to weak imaging signals and a shortened time window for bioimaging. In this regard, appropriate nanoplatform encapsulation is crucial for maintaining efficient generation of triplet excitons and protecting them from environmental quenching. To develop a universal encapsulating method for biocompatible nanoparticles suitable for phosphorescent bioimaging, Pu et al. fabricated carbazole-based organic semiconducting nanoparticles (OSN-T and OSN-B) using two different approaches: top-down and bottom-up (Fig. 13a).²⁴¹ These nanosized engineering approaches utilize opposite size regulation method, where the top-down approach breaks initially packed bulk crystals by mechanical erosion, while the bottom-up approach enables precipitation or crystallization from dispersed molecules. Transient luminescence decay images of OSN-T and OSN-B in phosphate buffer saline (PBS) showed significant differences. The phosphorescence of OSN-T decayed much more slowly than that of OSN-B (Fig. 13b), and OSN-T exhibited higher phosphorescence intensity under the same excitation conditions (Fig. 13c). This result is attributed to the nanocrystal-maintaining fabrication of OSN-T, which retained stronger molecular packing to facilitate the formation of H-aggregates. This specific packing mode helps to avoid ambient quenching and impedes non-radiative molecular motion, thus providing prolonged phosphorescence emission (Fig. 13d). In vitro imaging using the IVIS living imaging system confirmed that the long-lifetime emitting capacity of OSN-T was sufficient to be detected in PBS even 10 s after excitation removal. To validate its in vivo imaging ability, OSN-T was subcutaneously injected and activated in situ with light. The ultralong phosphorescence of OSN-T was easily detected, whereas the signal from OSN-B was weak, further confirming the superiority of the top-down approach in phosphorescent nanoparticle encapsulation for bioimaging in living subjects (Fig. 13e). Additionally, OSN-T was successfully applied in real-time mapping of lymph nodes in living mice. The bright signal detected in the axillary lymph node at 1 h post-intradermal-injection demonstrated a significant advance in pure organic phosphorescent in vivo imaging (Fig. 13f). This landmark success in generating and stabilizing triplet excitons for autofluorescence-free bioimaging suggests that maintaining crystallization should be a key consideration in the fabrication strategies for ultralong phosphorescent bioimaging probes. Several recent studies have adopted this nanonization strategy to ensure a bright phosphorescent signal for in vivo imaging. For instance, Li's group utilized the top-down strategy to maintain intermolecular π–π interactions among halogen-atom-substituted phosphors.¹⁵³ The versatile packing mode with different halogen atom was well-reserved in these water-soluble nanoparticles. Similarly, Ding et al. encapsulated a NIR guest–host phosphorescent doping material in F127 using the top-down method.²⁴² The resulting nanoparticles exhibited a phosphorescence signal capable of penetrating 12.5 mm of chicken breast tissue with a high SBR, indicating the preservation of guest–host doping for bright phosphorescence.

Fig. 13 Top-down strategies for fabrication of phosphorescence bioimaging probe. (a) Schematic illustration of OSN-T and OSN-B synthesized via top-down and bottom-up routes, respectively. (b) Luminescence decay images of OSN-T and OSN-B. (c) Delayed photon emission images in vitro. (d) luminescence decay diagram of OSN-T and OSN-B in vitro. (e) Subcutaneous bioimaging of OSN in vivo. (f) Lymph node mapping of OSN in vivo. Reproduced with permission: Copyright 2017, Wiley-VCH (adapted from ref. 241).

Besides ambient quenching, restricted imaging depth is another concern for emissive triplet excitons, as most pure organic phosphorescent molecules emit in the ultraviolet (UV) or visible range. To address this issue, red-shifting the emission wavelength is an effective strategy for phosphorescent probes, similar to its application in traditional optical materials. Inspired by FRET in fluorescence, Li's and Pu's groups designed a pure organic phosphorescent nanoprobe with NIR emission via intraparticle phosphorescence resonance energy transfer (PRET) (Fig. 14a).²⁴³ This nanoprobe (mTPA-N) involved a phosphorescent molecule (mTPA) as the donor and an NIR fluorescent chromophore (NCBS) as the acceptor. Efficient energy transfer was observed in the afterglow spectrum of mTPA-N nanoparticles, where the original phosphorescent emission peak at 530 nm shifted to 780 nm (Fig. 14b). This red-shifting process can be attributed to the partial spectrum overlap between donor emission and acceptor absorption (Fig. 14c). When applied in in vivo imaging, mTPA-N with NIR phosphorescent signal exhibited stronger afterglow intensity in both subcutaneous imaging and lymph node mapping, as compared to mTPA, demonstrating the significant advantage of the wavelength red-shifting strategy in sensitivity improvement and signal amplification. This approach enhances biological tissue penetration depth, making it highly effective for in vivo imaging. Despite the improved emission, the excitation wavelength of phosphorescent materials is mostly concentrated in the UV range, which is biologically unfriendly to animals and human beings. Regulation of excitation wavelength via molecular engineering is necessary to accelerate the application of existing triplet exciton generation and stabilization strategies in bioimaging. To achieve more suitable excitation and emission wavelengths, Li's and Zhen's groups developed a rational molecular fusion of electron donors and acceptors with different geometries and numbers (Fig. 14d).²⁴⁴ The designed molecule (DTBT) included three critical moieties: triphenylamine as a multi-armed donor, benzothiadiazole as an acceptor and carboxylic acid as the ending group for molecular interaction. Benzothiadiazole groups were conjugated at the para site of triphenylamine to form intramolecular charge transfer through a D–π–A system with linear and branched structures, while carboxylic acid was incorporated to induce hydrogen bonding for enhanced packing. These design strategies benefited both wavelength and lifetime regulation in this phosphorescent system. The obtained chromophores possessed broad absorptions extending to 600 nm and persistent red phosphorescence emission at 635–660 nm after visible light excitation (Fig. 14e and f), making them promise for bioimaging with high SBR and rational penetration depth. Although a slight decrease in emission intensity and lifetime of DTBT was observed after biocompatible encapsulation, the bright signal in aqueous solution with retained high SBR were presented in in vitro imaging (Fig. 14g). Further application of DTBT in subcutaneous tissue imaging and lymph node imaging successfully demonstrated high SBR imaging results after sunlight or mobile phone flashlight excitation, highlighting the importance of excitation and emission properties for phosphorescent systems in bioimaging (Fig. 14h and i). Both nanoencapsulation and wavelength regulation have advanced recently to achieve better performance of triplet excitons in bioimaging. However, it should be noted that light penetration depth reaches its best in the second near infrared (NIR-II) window at about 1300 nm.^245–247 No fundamental principle for the generation of triplet species with both controllable spectra properties and rational exciton stabilization currently guides the molecular engineering for further emission wavelength red-shifting towards NIR-II, along with prolonged emission lifetime. More candidate donor and acceptor skeletons with inherent interacting properties should be explored to promote the application of stabilized triplet excitons in bioimaging.

Fig. 14 Designing strategies for red-shifted emission of phosphorescence bioimaging. (a) Schematic illustration of PRET between mTPA and NCBS. (b) Afterglow spectra of mTPA and mTPA-N. (c) Spectrum overlap between mTPA and NCBS. (d) Chemical structure of s-DTBT, d-DTBT and t-DTBT. (e) Absorption spectra, (f) Phosphorescence spectra and (g) Afterglow images of s-DTBT, d-DTBT and t-DTBT. (h) Subcutaneous bioimaging of s-DTBT in vivo. (i) Lymph node mapping of s-DTBT in vivo. Reproduced with permission: Copyright 2020, Wiley-VCH (adapted from ref. 243). Reproduced with permission: Copyright 2022, Wiley-VCH (adapted from ref. 244).

In contrast to separated charges and stabilized triple excitons, chemical traps are the most widely employed species for realizing delayed photon emission in vivo due to their several advantages in producing emissive excitons during continuous decomposition. Firstly, energy storage in chemical traps relies on oxidative reactions instead of changes in electron configuration. This chemical process is more compatible with the biological environment than photophysical process. Secondly, the delayed emission lifetime for chemical traps is mainly controlled by dioxetane splitting kinetics, which can be fine-tuned by structural design. Although light emission from decomposition product encounters aqueous quenching, similar to exciplexes or triplet excitons, designing strategies for chemical traps have been developed to vastly prevent emitter exhaustion via regulated electron distribution, which is much easier than managing stabilized exciplexes or triplet excitons among complicated quenching substrates. Thirdly, chemical traps have been formulated as biomarker-responsive probes that release photons under particular biomarker stimulation, thus enabling specific visualization of physiological activity and disease progression. Benefiting from an explosion in substrate variety, various kinds of chemical trap precursors, such as PPV, AEE, porphyrin and cyanine, have been widely applied in biomedical applications.

Since the fundamental investigation into chemiexcitation process during polymer degradation of PPV, versatile design strategies have been developed to optimize PPV-based chemical trap system into formulated persistent emitters for bioimaging. Initially, nanoprecipitation was employed to transform PPV into water-soluble nanoparticles without any dopant.³¹ The self-generated ¹O₂ was sufficient to initiate dioxetane generation and decomposition in PPV, resulting in long-lasting luminescence with a half-life of 6.6 min under biological conditions. To achieve more suitable emission wavelength and signal intensity for in vivo imaging, the photosensitizer NCBS was doped into PPV nanoparticles (PPV-NCBS), serving as both efficient ¹O₂ generator and energy transfer acceptor, to realize bright NIR afterglow (Fig. 15a). This optimization significantly improved the afterglow intensity of PPV under 808 nm excitation and redshifted the emission wavelength towards 775 nm (Fig. 15b), enabling in vivo afterglow lymph node mapping and tumor imaging with a much higher SBR than fluorescence imaging (Fig. 15c–e). The construction of activatable afterglow nanoparticles was also explored in this work by designing a biothiol-cleavable amphiphilic polymer conjugated with an afterglow quencher to encapsulate PPV and NCBS. This biothiol-responsive chemical trap system was successfully applied in afterglow imaging of drug induced hepatotoxicity (DIH). Ye et al. also designed an activatable PPV-based chemical trap system (F1²⁺-ANP) through the ingenious combination of PPV, NCBS and an H₂S-responsive quencher (F1²⁺).²⁴⁸ F1²⁺ performed thorough quenching of NIR afterglow due to its strong absorption at 758 nm. Once reacted with H₂S, F1²⁺ converted into F2, resulting in the recovery of ¹O₂ generation and NIR afterglow emission. F1²⁺-ANP was then applied in tumor imaging on mice, detection of H₂S in blood sample, and afterglow imaging of liver tumor tissue in clinically excised liver specimens. Another design for activatable PPV-based afterglow was reported by Tan et al., where the PPV-based chemical trap system was formulated into three moieties: afterglow initiator, afterglow substrates, and responsive energy transfer dyes.²⁴⁹ As biomarker stimulation can cause differentiated signal ratios between afterglow emission before and after energy transfer, a ratiomatric afterglow in vivo imaging of macrophage polarization was conducted using this designed chemical trap system.

Fig. 15 PPV-based chemical trap for bioimaging. (a) Schematic illustration for fabrication of PPV-based chemical trap and mechanism for inner energy transfer. (b) Afterglow spectra of PPV nanoparticle with or without NCBS. (c) Lymph node mapping and (d) tumor imaging of PPV-NCBS in vivo. (e) SBR analysis of imaging results in (d). Reproduced with permission: Copyright 2017, Springer Nature (adapted from ref. 31).

Versatile modifications of PPV-type materials have been explored to improve their biocompatibility and emitting properties for bioimaging applications. Typically, hydrophilic poly (ethylene glycol) (PEG) can be introduced into the hydrophobic PPV backbone as grafting chains to create water-soluble amphiphilic PPV derivatives. Pu et al. firstly reported this polymer brush design, endowing PPV with self-assembling capacity in aqueous solution.²⁵⁰ The grafting density of PEG was fine-tuned to ensure efficient encapsulation of the photosensitizer (NCBS), resulting in smaller nanoparticle sizes for better biodistribution and closer contact between PPV and NCBS for brighter afterglow. Based on this amphiphilic PPV design, various hydrophobic optical agents can be precisely combined with PPV to fabricate multifunctional composites for biomedical applications. Pu et al. reported the design and synthesis of semiconducting polymer nanococktail (SPN_CT) with temperature-monitoring afterglow luminescence for imaging-guided photothermal therapy (PTT).²⁵¹ SPN_CT involved two major functional moieties: amphiphilic PEG-grafted PPV serving as the temperature sensor, and NIR absorbing semiconducting polymer (PCSD) as the photothermal agent (Fig. 16a). As discussed in Section 2.3, the chemiexcitation of dioxetane-based chemical traps exhibits intrinsic temperature sensitivity because external thermal activation can accelerate the transition to the O–O bond-cleaving transient state, resulting in faster dioxetane decomposition. To confirm this phenomenon in a biological environment, SPN_CT in PBS was subjected to afterglow activation for 1 min, followed by photothermal operation for varying lengths of time to reach specific temperatures (Fig. 16b). An excellent linear correlation between afterglow intensity and temperature was observed, indicating the great potential of SPN_CT for temperature-monitoring in vivo (Fig. 16c). Due to its autofluorescence-free imaging ability and temperature-monitoring capacity, SPN_CT produced bright emission in tumor areas for PTT guidance in mice and subsequently monitored PTT temperature to avoid hyperthermia, ensuring biosafety for theronostics. A Similar design was reported by Zhen's group for afterglow imaging-guided adjuvant therapy following breast-conserving surgery (BCS).²⁵² In this work, PEG-grafted PPV was modified with cyclic arginine-glycine-aspartate (cRGD) (PPV–PEG–cRGD) to improve its tumor targeting ability for sensitive tracking of tumor residue and redshifted its emission to NIR window. To approach the penetration depth limit of the optical theranostic system, NIR-II absorbing photothermal agent (PBBTOT) and NIR photosensitizer (NCBS) was incorporated into PPV–PEG–cRGD to create afterglow/photothermal bifunctional polymeric nanoparticles (APPN) (Fig. 16d). The afterglow signal from APPN successfully illuminated residual microscale tumor foci after BCS and guided post-BCS adjuvant NIR-II PTT (Fig. 16e, g and h). Additionally, APPN was also applied in the early diagnosis of breast tumor recurrence (Fig. 16f). Compared to the results of hematoxylin and eosin (H&E) staining and magnetic resonance imaging (MRI), afterglow imaging by APPN performed much earlier detection of in situ recurrence at day 6 post-surgery with significant amplification of SBR (Fig. 16i and j). Beyond combining with photothermal agents for theranostic application, PPV can also be modified as charged species to target charge-favorable biomolecules for specific bioimaging. For instance, Pu et al. copolymerized cationic quaternary-ammonium-brushed PPV with tetraphenylporphyrin to obtain an afterglow semiconducting polyelectrolyte (ASP).²⁵³ ASP can attract quencher-tagged aptamers through electrostatic interactions to form an activatable afterglow nanocomplex. Upon recognition by exosomes, the quencher-tagged aptamer was despoiled, thus activating the afterglow signal in ASP.

Fig. 16 Multifunctional composites based on PPV chemical trap for biomedical theranostics. (a) Schematic illustration for temperature-correlated afterglow realized by SPN_CT. (b) Thermal (left) and afterglow (right) images of tumor-bearing mice with 0 min or 5 min 808 nm laser irradiation. (c) Linear correlation analysis between afterglow intensity and photothermal temperature. (d) Schematic presentation of components in APPN. (e) Schematic illustration of afterglow-guided post-BCS NIR-II PTT. (f) Schematic illustration of early diagnosis of local recurrence by afterglow imaging. (g) Fluorescence and afterglow images of tumor residue post BCS through deep tissue. (h) Representative thermal images of mice receiving afterglow-guided post-BCS NIR-II PTT. (i) Afterglow monitoring of tumor recurrence for 15 days on mice post BCS. (j) SBR analysis of imaging results in (i). Reproduced with permission: Copyright 2018, Wiley-VCH (adapted from ref. 251). Reproduced with permission: Copyright 2023, American Chemical Society (adapted from ref. 252).

AEE initially appeared as a chemiluminescent probe to emit light upon reaction with ¹O₂ in McNeill's work.²⁵⁴ Further development for aqueous compatibility was achieved by Shabat et al., leading to the first chemiluminescent probe with bright persistent emission under physiological conditions.^200,201,217

Since then, AEE has become the most frequently used small molecule chemical trap precursor, with extensive chemical design efforts for optimal emission properties in bioimaging. Inspired by the construction principles of PPV-based chemical traps, Ding et al. first reported an AEE-based bioimaging probe (AGL AIE dots) that comprised AEE as the chemical trap precursor and aggregation-induced emission luminogens (AIEgens) as both initiator and energy transfer acceptor (Fig. 17a).²¹⁶ AGL AIE dots exhibited superior tissue-penetrating ability compared to AEE dioxetane nanoparticles due to their redshifted emission (Fig. 17b). They were successfully applied in imaging-guided cancer surgery on peritoneal carcinomatosis bearing mice. The strong NIR afterglow signal from AGL AIE dots distinctly illuminated sub-millimeter tumor nodules in the abdomen, which were challenging to differentiate using NIR fluorescence imaging, demonstrating the feasibility of AEE-based chemical traps for practical surgical navigation (Fig. 17c). To fully exploit AEE's potential for delayed photon emission in biological applications, Pu et al. reported a generic approach for formulating chemical trap systems for bioimaging with tunable emission wavelength and brightness. This approach relies on intraparticle cascade photoreactions involving three critical components: the afterglow initiator, afterglow substrate, and afterglow relay unit.¹⁸⁶ These components work together to store photoenergy in chemical traps, enabling delayed luminescence after light excitation.

Fig. 17 AEE-based chemical trap for bioimaging. (a) Schematic illustration for fabrication of AEE-based chemical trap (AGL AIE dot) and mechanism for inner energy transfer. (b) In vitro deep tissue afterglow images of AEE-based chemical trap with (right) or without (left) TPE-Ph-DCM. (c) Fluorescence and afterglow images of abdominal cavity before and after tumor resection. Reproduced with permission: Copyright 2019, American Chemical Society (adapted from ref. 216).

On the other hand, with its modifiable structure, AEE can be designed as theranostic nanoplatforms that integrate disease diagnosis and treatment. Pu et al. reported the synthesis of prodrug-caged AEE (APtN) within a chemical trap system to achieve both pharmaceutical effect and diagnostic signals in the tumor microenvironment of living mice.²¹⁹ Under stimulation by tumor-upregulated H₂O₂, the prodrug was cleaved from the phenyl site of AEE, releasing both the active drug and the emissive AEE. A linear correlation was observed between the delayed emission signal and drug concentration (Fig. 18a). To verify the accuracy of H₂O₂-related signal activation and drug release, this theranostic system was applied to mice receiving pre-treatment with BSO to increase H₂O₂ levels and NAC to scavenge H₂O₂ (Fig. 18b). Significant increases and decreases in the delayed emission signal were observed in in vivo imaging, correctly corresponding to the pre-regulated H₂O₂ concentration (Fig. 18c). A similar theranostic nanoplatform was developed by Ding et al., providing self-amplification of diagnostic signals and drug release along with red-shifted emission in the NIR window via energy transfer.²⁵⁵ The recycled signal enhancement in this theranostic system was managed by drug-induced biomarker increases based on immunogenic cell death (ICD). The peroxynitrite (ONOO⁻)-responsive AEE and ONOO⁻-cleavable ICD prodrug (HCPT) was fused in one molecule and doped with NIR photosensitizer ((TPE-DPA₂)-Py). Upon light irradiation, the theranostic probes accumulated in tumor, generating ¹O₂ to evoke ICD, which subsequently recruited neutrophils and T cells to convert the cold tumor to hot, with significantly increased ONOO⁻ levels. The resulting ONOO⁻ led to HCPT release as well as chemical trap activation, accelerating the amplification of ICD and delayed emission (Fig. 18d). This theranostic probe was intravenously injected into 4T1-tumor-bearing mice to evaluate its ICD-activating capacity and pharmaceutical imaging ability (Fig. 18e). At 4 h post-injection, a bright signal illuminated the tumor area and exhibited continuous intensity increases until 8 h post-injection, which was highly consistent with the analysis of drug release and neutrophil infiltration (Fig. 18f). These bioimaging results demonstrated the great potential of AEE in theranostic probe fabrication.

Fig. 18 Theranostic probes based on AEE chemical trap. (a) Schematic illustration for the linear correlation between H₂O₂-induced drug release and afterglow intensity of APtN. (b) Fluorescence and afterglow images of tumor bearing mice receiving regulation on H₂O₂ concentration in tumor. (c) Statistic analysis of imaging results in (b). (d) Schematic illustration for the linear correlation between ONOO⁻-induced drug release and NIR afterglow intensity of AIE/B-AGL-HCPT NPs. (e) Representative afterglow images of mice treated by AIE/B-AGL-HCPT NPs, which received regulation on ONOO⁻ concentration and ¹O₂ generation. (f) Quantitative afterglow intensity of tumor in e. Reproduced with permission: Copyright 2019, Wiley-VCH (adapted from ref. 219). Reproduced with permission: Copyright 2022, Wiley-VCH (adapted from ref. 255).

The first research utilizing porphyrins as light-activated chemical traps in bioimaging was conducted by Ding et al. In this study, Ppa was chosen as a candidate chemical trap precursor due to its excellent delayed emitting properties after laser irradiation and concomitant photoacoustic (PA) imaging ability.²²⁹ To construct a porphyrin-based chemical trap system with adaptive functionality for in vivo imaging, Ppa was conjugated with the supramolecular self-assembling peptide FFGYSAYPDSVPMMS (FFGYSA) to form the Ppa-FFGYSA conjugate. In this conjugate, the FFG unit facilitated β-sheet formation for self-assembly, while the YSAYPDSVPMMS (YSA) sequence served as a tumor-targeting group. The conjugate exhibited an emission maximum at 760 nm and enabled sensitive imaging through 1 cm of chicken breast tissue. To evaluate the feasibility of Ppa-FFGYSA in multi-mode-imaging-guided tumor surgery, a functional transformation from PA imaging to persistent luminescence imaging was performed on orthotopic breast-cancer-bearing mice. At 8 h post-intravenous-injection, the PA signal from Ppa-FFGYSA in the tumor area reached its maximum, guiding subsequent laser irradiation to activate the persistent luminescence. After excitation removal, a highly distinguishable persistent luminescence signal was acquired, showing a 26.9-fold higher SBR compared to that of the fluorescence image, indicating the superiority of the Ppa-FFGYSA conjugate in multi-mode bioimaging. Around the same time, Gao et al. reported a kind of chlorin nanoparticles (Ch-NPs) that emitted delayed emission peaking at 680 nm with a long lifetime up to 1.5 h, where Ce4 was chosen as the optimal chemical trap precursor (Fig. 19a).²⁵⁶ The Ch-NPs were applied on mice to detect intraperitoneal tumors of different sizes and to navigate tumor resection surgery (Fig. 19b). Persistent luminescence images showed a bright signal in the region of interest 2 h post-intravenous-injection (Fig. 19c). At this time point, a laparotomy was performed on the mice, followed by the resection of metastatic microtumors. Persistent luminescence imaging of the original focus and resected tumor were also conducted to confirm the precision of the tumor resection under the imaging guidance by Ch-NPs (Fig. 19d and e).

Fig. 19 Porphyrin-based chemical trap for bioimaging. (a) Schematic illustration for fabrication of NPs-Ce4. (b) Schematic presentation for procedure of afterglow-guided abdominal metastatic tumor resection. (c) Representative afterglow images of mice treated at different time post injection of NPs-Ce4 and the corresponding quantitative afterglow intensity analysis. (d) Afterglow images of mice receiving laparotomy. (e) Bright-field and afterglow images of resected tumor. Reproduced with permission: Copyright 2022, American Chemical Society (adapted from ref. 256).

Following the research revealing the potential of porphyrins in delayed emission for autofluorescence-free bioimaging, cyanine and its analogs have also been reported to possess sufficient delayed photon release detectable in biological environments. Song et al. developed a series of hemicyanine (HD) based chemical trap precursors (MAP) as activatable probes for high-contrast persistent luminescent bioimaging.²³⁰ As described in Section 2.3, photon generation in hemicyanines can originate from dioxetane generation and decomposition at the π bridge, which is highly regulated by the electron-donating ability of the phenolic hydroxyl group. When caged by electron- localizing groups, the delocalization of lone pair electrons at the phenolic hydroxyl site is impeded, leading to the inhibition on chemiexcitation. In this study, HD-based chemical trap precursors were designed as afterglow probes activable by O₂˙⁻ (MAP-O₂˙⁻) and leucine aminopeptidase (MAP-LAP) for in vivo imaging of early DIH (Fig. 20a and c). To verify whether O₂˙⁻ and LAP are sequential biomarkers in DIH, MAP-O₂˙⁻ and MAP-LAP were respectively injected in mice treated with acetaminophen (APAP) treatment. A significant increase in both fluorescence and afterglow signal was observed in the MAP-O₂˙⁻ treated group at 15 min post-APAP-injection, much earlier than that of MAP-LAP (Fig. 20b and d). This imaging result not only demonstrated that O₂˙⁻ is a reliable early biomarker for DIH bioimaging, but also indicated the excellent potential of HD-based chemical trap precursors for persistent luminescent imaging in vivo. Based on a similar structure, Song et al. further developed a molecular fusion of Rhodamine and hemicyanine to enhance ¹O₂ self-generation and delayed emission intensity.²⁵⁷ This novel chemical trap system was pH-sensitive and was employed in afterglow imaging to monitor both upregulation and downregulation of tumor glycolysis.

Fig. 20 Cyanine-based chemical trap for bioimaging. (a) Mechanism for afterglow signal activation of MAP-O₂˙⁻ and fluorescence/afterglow intensity from MAP-O₂˙⁻ activated by different concentration of O₂˙⁻. (b) Fluorescence and afterglow images of mice pre-treated by MAP-O₂˙⁻ at different time post APAP treatment. (c) Mechanism for afterglow signal activation of MAP-LAP and fluorescence/afterglow intensity from MAP-LAP activated by different concentration of LAP. (d) Fluorescence and afterglow images of mice pre-treated by MAP-LAP at different time post APAP treatment. Reproduced with permission: Copyright 2023, American Chemical Society (adapted from ref. 230).

3. X-ray activated delayed photon emission

Light-activated delayed photon emission has garnered significant attention for its application in bioimaging, owing to its long-lifetime emitting properties and the elimination of real-time excitation.^258–261 Despite its great advantages, the penetration limits of light pose a considerable challenge for its use in bioimaging, particularly during the light pre-excitation process.²⁶² Due to the strong biological scattering and absorbance in deep tissue, light energy input encounters vast decrease, leading to inefficient energy storage for effective photon generation.²⁶³ To address this limitation, it is crucial to explore alternative energy sources that can achieve deep-tissue penetration, enhancing the depth and effectiveness of bioimaging applications that rely on delayed photon emission.

In contrast to visible or NIR light, X-ray consists of high-energy photons with extremely short wavelengths that can penetrate through atomic intervals.²⁶⁴ Consequently, X-ray is inherently capable of delivering energy efficiently to targeted areas deep within tissues, enhancing energy storage for delayed emission.^265–267 To harness X-ray activation, an understanding of X-ray absorption and energy transformation is essential for guiding molecular engineering for photon emitters (Fig. 21a).^268–270 It is important to note that X-ray energy is generally too high for valence shell electrons to absorb. While a fraction of high-energy electrons can be generated via Compton scattering, only the inner electrons, which seated in deep potential well, can be activated through the photoelectric effect and subsequently ionized by X-ray. This is due to the large atomic-number (Z)-related attenuation coefficient (μ) of heavy atoms (μ ∝ Z⁴).²⁷¹ This activation process generates electron–hole pairs and initiates a cascade of secondary electron excitations through Auger–Meitner decay or intermolecular Coulombic decay.²⁷² These secondary excitations subsequently relax into a non-equilibrium distribution before photon emission occurs.²⁷³ The energy conversion from X-ray absorption to emitter generation mirrors the processes of charge separation, charge diffusion and charge recombination observed in light-activated photon release. Based on this mechanism, X-ray transducers, known as X-ray scintillators, are designed to generate emissive species under X-ray activation.²⁷⁴ Significant progress has been made in developing strategies for X-ray scintillators, encompassing both organic and inorganic, to achieve delayed radioluminescence.^275–277

Fig. 21 Mechanism and designing strategies for X-ray activated delayed photon emission. (a) Schematic presentation of X-ray-energy transforming process in X-ray absorbing materials. Upon X-ray irradiation, X-ray photon can be absorbed via photoelectric effect and Compton scattering to generate ionized high-energy electrons, which subsequently decay into secondary electrons. Then, the ionized electrons recombine with corresponding holes to give exciplexes in singlet/triplet excited state. (b) Schematic presentation of designing strategies to realize delayed photon emission for X-ray energy transformation from X-ray photon to ionized electron. (c) Schematic presentation of designing strategies to realize delayed photon emission for X-ray energy transformation from ionized electron to triplet/singlet exciton.

X-ray scintillators utilize charge separation and recombination to convert X-ray energy into long-wavelength photon emission through cascaded electron ionization. This energy conversion process can be divided into two steps: energy storage and lifetime prolongation.²⁷⁸ In the first step, converting X-ray energy to ionized electrons, the heavy atom effect is the most critical design principle for both organic and inorganic materials, determining the efficiency of X-ray energy utilization (Fig. 21b(i)).^279,280 To address the inherently weak X-ray absorption in organic materials, heavy atoms such as bromine and iodine can be incorporated into the aromatic structure through covalent bonding and the formation of ionic pairs. Moreover, the lifetime of delayed emission in this process is significantly dependent on prolonged charge diffusion, necessitating the involvement of charge carriers that act as energy traps to stabilize the ionized electrons and holes (Fig. 21b(ii)).²⁸¹ For inorganic materials, X-ray activated delayed photon emission is predominantly governed by two critical steps: (i) the capture of excited charge carriers within discrete defect levels in the bandgap, and (ii) the escape of charges from these traps to form exciplexes that emit light.²⁸² Discrete defect levels within inorganic lattice defects provide controlled energy gaps, facilitating rational escape energies that not only prolong the emission lifetime but also enable thermal activation, enhancing the delayed generation of emitters. Conversely, it is challenging for organic materials to implement thermally activable charge traps since ionized organic molecules tend to be inherently unstable and recombine rapidly. Although design strategies for light-activated charge separation in pure organic systems have seen considerable success, a corresponding principle for X-ray activated charge separation and recombination has not yet been effectively developed in the fabrication of organic materials.

In organic systems, the formation of exciplexes from recombination of X-ray-induced charges typically leads to the generation of excitons with various spin multiplicities, where triplet excitons predominate due to statistic principle (Fig. 21c(i)). This indicates that design strategies effective for light-induced triplet excitons could also be applicable for achieving delayed emission under X-ray activation. Huang et al. firstly reported the design principles for X-ray-activated organic phosphors with exceptional energy transformation capabilities.²⁸³ To address the typically weak X-ray absorption of conventional phosphor skeletons, modifications were made by introducing iodine atoms at different positions on the compound 9,9′-(6-iodop henoxy-1,3,5-triazine-2,4-diyl)bis(9H-carbazole), resulting in three isomers: ortho-ITC, meta-ITC and para-ITC. The effectiveness of the heavy atom effect in enhancing X-ray absorption was confirmed through comparative substitutions of the iodine atom with bromine, chlorine or hydrogen. Among these, the ITCs exhibited the highest absorption coefficients. Upon X-ray excitation, the ITCs displayed significantly enhanced radio-phosphorescence and a higher phosphorescence-to-fluorescence ratio compared to their ultraviolet-excited counterparts. This phenomenon can be attributed to the differences in triplet exciton generation between photoluminescence and radioluminescence processes. Moreover, the presence of heavy atoms and the rigid environment within the ITC crystals are crucial for promoting strong ISC thereby further increasing the population of triplet excitons which enhances the X-ray-excited phosphorescence.

The enhanced generation of triplet excitons in radioluminescence underscores the importance of the packing mode within X-ray phosphorescent crystals for stabilizing triplet exciton and achieving long-lived emission. Effective molecular packing significantly reduces non-radiative decay in anti-motion environment. To explore the optimal packing mode for X-ray-excited phosphorescence, Huang et al. reported a polymorphism-dependent scintillation utilizing phenothiazine-based polymeric phosphors.²⁸⁴ They employed a solvent evaporation method to create different packing mode (T-type and C-type) by changing the binary solvent systems using trichloromethane and tetrahydrofuran with hexane, respectively. In these systems, sulfur atoms function as X-ray absorbing moieties. Their findings revealed that the radioluminescence of the T-type packing configuration showed a significant improvement in radioluminescence lifetime compared to the C-type configuration. This enhancement was attributed to the synergistic effects of strong π–π interactions and hydrogen bonds.

To achieve efficient photon emission under radio-excitation, it is essential to convert X-ray-induced triplet excitons into singlet excitons, given that the electronic transition from triplet to singlet is inherently spin-forbidden. Recently, Yang et al. demonstrated that TADF-type design is a feasible method for optimizing photon generation from X-ray-induced excitons.²⁸⁵ Consistent with the TADF mechanism discussed in Section 2.2, the presence of a small triplet–singlet energy separation in anthracene-like structures was found to enhance radioluminescence intensity and lower the detection limit in X-ray imaging applications. Notably, anthracene-like structures exhibit efficient absorption of X-ray photons within the energy range of 3–50 keV. This characteristic offers significant potential for molecular engineering, allowing for the fusion of extensive conjugation blocks to develop organic TADF-type X-ray scintillators.

Beyond small molecule scintillators, the development of polymeric scintillators has emerged as a promising approach to achieve X-ray-excited phosphorescent emission from amorphous polymers, which is more feasible for material fabrication in practical applications. Huang et al. devised a straightforward strategy to induce radioluminescence using copolymers composed of heavy halogen atoms-substituted chromophores and acrylic acids.²⁸⁶ In this polymeric system, chromophores substituted with heavy halogen atoms are utilized to efficiently absorb X-ray energy. Concurrently, acrylic acid is employed to introduce multiple inter/intra-molecular hydrogen bonds. These hydrogen bonds effectively restrict the molecular motion of the chromophores, thereby minimizing non-radiative transition and enhancing phosphorescence generation.

Given the availability of triplet excitons in organic X-ray-activated systems, the production of ¹O₂ for dioxetane generation from X-ray-excited triplet excitons presents a feasible method for establishing X-ray-activated chemical traps (Fig. 21c(ii)).^287–289 In light of advances in organic radiodynamic materials, Pu et al. introduced an innovative approach with an organic luminophore (IDPA) that integrates a radiosensitizer with an AEE-type chemical trap precursor to develop an X-ray-activated chemical trap system.²⁹⁰ The introduction of an iodine atom onto phenoxy group, along with a sulfur or selenium atom within the acceptor unit, endows IDPAs with remarkable X-ray absorbing capacity, enabling effective ROS production under radioexcitation. The resulting ¹O₂ can then be captured by the AAE structure to form high-energy dioxetane. The decomposition of this dioxetane is further regulated by the delocalizing ability of lone pair electrons on the phenolic oxygen. Furthermore, the versatile design of the acceptors contributes to red-shifting the X-ray-activated emission towards the NIR region, allowing for the fine-tuning of spectral properties suitable for delayed photon emission. This radio-emitter has been demonstrated to undergo an efficient radiodynamic process, emitting radio afterglow that is over 25 times brighter than previously reported inorganic afterglow scintillators.

X-ray activated delayed photon emission represents an innovative imaging strategy that leverages X-ray absorbing moieties, yet it is rarely reported in bioimaging, particularly for pure organic probes. Traditionally, strategies to achieve long-lived emissive species under X-ray excitation have predominantly utilized inorganic materials,²⁸¹ due to the shallow lifetimes of ionized charges and triplet excitons in organic system, which are highly reactive and susceptible to environmental quenching. In contrast to mechanisms such as charge separation and triplet exciton stabilization, X-ray activated chemical traps decouple the delayed emission process into two distinct phases: X-ray-dependent energy storage and structure-dependent photon generation. Notably, the critical step determining the emission lifetime occurs during the decomposition of dioxetane, which is independent of X-ray excitation. This separation of steps offers the advantage of deep-tissue penetration ability afforded by X-ray energy delivery, combined with the chemically controlled emission lifetime stemming from the dioxetane structure, providing a great opportunity for X-ray activated chemical traps in bioimaging. For instance, Pu et al. successfully demonstrated the use of IDPAs in precise theranostics of intracranial glioblastoma, capitalizing on the specific and deep-tissue penetrating imaging properties of this radio probe.²⁹⁰ To evaluate the influence of tissue penetrating capacity on energy delivery, photoafterglow and radioafterglow activation of IDPAs were conducted under different thickness of chicken breast tissue (Fig. 22a). In contrast to the significantly reduced photoafterglow intensity through 1 cm of tissue depth, radioafterglow exhibited a bright signal capable of penetrating up to 15 cm of tissue depth in vitro (Fig. 22b and c), indicating the substantial potential of X-ray-activated chemical traps for deep imaging of orthotopic lesions in organisms. Moreover, the total X-ray dosage for X-ray-activated delayed photon emission falls within the range used in X-ray computed tomography, presenting a low radiation risk.^291,292 To facilitate specific cancer diagnosis and therapy, IDPAs were further modified with a cleavable peptide cage to create a molecular radio afterglow dynamic probe (MRAP), which was responsive to the tumor biomarker cathepsin B (CatB). In addition, cRGD was introduced as a tumor-targeting moiety to enhance biodistribution within tumor (Fig. 22d). In the absence of CatB, the radiodynamic process for radio afterglow in MRAP is quenched due to the reduced electron-donating capability of the phenolic oxygen in the phenoxy-adamantylidene unit, which inhibits intramolecular charge transfer. However, upon cleavage of the peptide by CatB, MRAP exhibited a 7.8-fold enhancement in ¹O₂ generation and a 312-fold signal enhancement in radioafterglow. This tumor targeting and CatB response mechanism provided by the MRAP probe resulted in a sharp radioafterglow signal contrast between U87 cells and 3T3 cells after incubation with MRAP (Fig. 22e and f), confirming its precision for cancer theranostics. When applied in vivo, MRAP demonstrated excellent imaging sensitivity for small glioma tumor tissue within the brain (Fig. 22g), which was further validated by precise detection through colocalization with bioluminescence tumor signal and linear correlation with tumor size (Fig. 22h and i). This example of X-ray activated delayed photon emission offers a general approach to utilize X-ray photon delivery in deep tissue, inducing photophysical energy transformations in biomedical applications.

Fig. 22 X-ray activated delayed photon emission for bioimaging. (a) Schematic of evaluation on deep tissue penetrating capacity of afterglow activated by light and X-ray. (b) Representative afterglow images of IDPA after light or X-ray irradiation at different tissue depth. (c) Quantitative afterglow intensity of imaging results in (b). (d) Chemical structure of MRAP and mechanism for radio afterglow activation of MRAP. (e) Schematic illustration for specific activation of radio afterglow of MRAP in U87 glioma cell. (f) Cellular radio afterglow intensity from MRAP in U87 or 3T3 at different cell counts. (g) Schematic illustration of radio afterglow ultrasensitive tumor imaging exhibiting excellent colocalization with bioluminescence signal. (h) Representative bioluminescence and radio afterglow images of tumor-bearing mice with different tumor sizes. (i) Linear correlation analysis between bioluminescence/radio afterglow signal and tumor size. Reproduced with permission: Copyright 2023, Springer Nature (adapted from ref. 290).

4. Ultrasound-activated delayed photon emission

While X-ray photon delivery can penetrate deeply into tissues, its use is often limited by the potential for incidental damage to normal tissues due to the generation of ionized species.^293,294 In contrast to the electromagnetic-wave-based energy sources such as light or X-ray, ultrasound emerges as a highly promising alternative. As a mechanical wave, ultrasound can penetrate deeper into biological tissues (greater than 10 cm) to conduct energy delivery, offering the superior penetration depth, spatiotemporal controllability and high biosafety over light or X-ray.^295–299

Different from the direct generation of excitons or separated charges under light or X-ray activation, the energy transformation via ultrasound relies significantly on the transmission medium.^300,301 This transformation can manifest as sonoluminescence in water, a phenomenon where spontaneous luminescence is triggered by plasma generation during bubble cavitation in a sound field (Fig. 23a).³⁰² Under varying acoustic pressure amplitudes, gas-filled bubbles induced by the sound field undergo periodic expansions and energy extraction, followed by violent collapses. Once collapsing bubble well attains supersonic velocities, micro shock waves converge at the core of bubble leading to rapid compression. This compression heats the gas to high temperatures, causing ionization to form plasma, which in turn generates light.³⁰³ Although such sonoluminescence is highly restrained in direct imaging application due to its low brightness and short duration, the cavitation process is capable of exciting nearby sonosensitizers to generate reactive species, storing the mechanical energy of ultrasound in the form of chemical energy.^304,305 Another energy transformation strategy for ultrasound is interaction with mechanical-force-responsive materials to induce reactive species production and light emission. One common approach is acoustically mediated piezoelectric stimulation (Fig. 23b).^306–308 High-frequency ultrasound can induce polarization in piezoelectric materials, creating an endogenous electric field that separates electrons and holes, thus transforming ultrasound's mechanical energy to potential energy. The separated charges accumulate on opposite surfaces and engage in oxidation and reduction reactions with substrates in an aqueous solution, leading to the generation of ROS. This conversion process effectively translates mechanical energy into chemical energy. Both organic and inorganic piezoelectric materials have been proved as effective piezocatalysts for ROS generation, underscoring their potential in the fabrication of chemical traps and confirming the broad applicability of piezoelectric materials in ultrasound-activated delayed photon emission. Another viable strategy leverages mechanically stimulated electron escape from energy traps, utilizing ultrasound as the activation energy source to overcome the small gaps between defect levels.^309–311 In this setup, electrons pre-excited to the conduction band are trapped by engineered defects without emission. Upon mechanical perturbation, these trapped electrons are released and undergoes radiative relaxation to generate delayed emission. While this mechanism has been demonstrated in inorganic systems, it presents significant challenge to engineer mechanically-activable trap in organic materials (Fig. 23c).

Fig. 23 Mechanism and designing strategies for ultrasound activated delayed photon emission. (a) Schematic illustration of sonoluminescence from ultrasonic cavitation. (b) Schematic illustration of ROS generation through ultrasound-induced piezocatalyst. (c) Schematic presentation of energy transformation from mechanical wave to light through mechanically activated electron trap. (d) Schematic presentation of energy transformation from mechanical wave to reactive species for chemical trap generation.

Concerning the three involved strategies for delayed photon emission discussed in this review, the generation and stabilization of triplet excitons under ultrasound stimulation seems to be inaccessible. Mechanical perturbation from ultrasound is typically insufficient to excite molecules from the ground state to excited states directly, and the continuous mechanical wave might even disrupt the structural integrity of the well-packed nanocrystals. Similarly, charge separation generally requires a robust and instantaneous energy supply, which mechanical perturbation cannot provide, although it remains feasible through polarization in acoustically mediated piezoelectric stimulation. Conversely, the strategy of delayed charge recombination facilitated by ultrasound-activated detrapping of pre-excited electrons has been proved to be a viable method for achieving ultrasound-activated delayed photon emission. This is particularly evident in inorganic materials doped with ions such as Ag⁺ and Co²⁺ in ZnS.³¹⁰ However, extending this method to pure organic system is highly challenging. The construction of appropriate energy levels through molecular engineering in organic materials remains a complex task. Chemical traps are more compatible with ultrasound energy delivery due to the distinct separation between their energy storage and photon emission. Energy storage involves the accumulative addition of ROS over an extended period, and the lifetime of delayed emission is determined solely by the chemical properties of the traps. Utilizing two primary mechanisms for ROS generation by ultrasound mechanical energy, bubble cavitation induced photocatalysis and acoustically mediated piezocatalysis, various sonosensitizers have been incorporated into chemical trap systems to produce ¹O₂ or ˙OH under ultrasound excitation, facilitating the formation of dioxetanes that promise a continuous supply of emissive species for photon emission (Fig. 23d).^312,313

Ultrasound-activated delayed photon emission, pioneered for deep tissue mechanoluminescence in brain for optogenetic applications by Hong et al.,³¹⁴ has been developed in both organic and inorganic materials for biomedical applications. Unlike mechanically pumped emissions from colloidal inorganic nanoparticles in biological environment, organic emitters activated by ultrasound are primarily designed as chemical traps for bioimaging. These systems use ultrasound energy delivery to generate ROS, facilitating energy storage in dioxetane intermediates. Pu et al. introduced the first fabrication of ultrasound-activated chemical trap systems (SNAPs) that incorporate a sonosensitizer as the initiator to produce ¹O₂ under ultrasound irradiation, which then converts a chemical trap precursor into an active dioxetane substrate.³¹² This substrate's luminescence is persistent and capable of transferring energy back to the sonosensitizer, thereby producing a sonoafterglow at a rational emission wavelength. Initial screening of sonosensitizers and chemical trap precursors was essential to optimize the performance of SNAPs. Following an assessment of various combinations, NCBS as the initiator and dicyanomethylene-4H-benzothiopyran-phenoxyl-adamantylidene (DPAs) as the chemical trap precursor were selected to construct the NCBS/DPAs SNAP for bioimaging applications. The superior tissue penetration of ultrasound-activated delayed emission compared to light-activated systems was confirmed in vitro (Fig. 24a). The NCBS/DPAs SNAP achieved a sonoafterglow that penetrated a tissue depth of 4 cm, surpassing the 2 cm depth achieved by photoafterglow (Fig. 24b and c). To create a biomarker-responsive sonoafterglow signal for cancer immunotherapy monitoring, molecular engineering was applied to cage DPAs with a ONOO⁻ responsive moiety (Pro-DPAs), resulting in SNAP-M (Fig. 24d), which exhibited specific signal activation upon incubation with M1 microphage (Fig. 24e). When tested in vivo, at 36 h post-intravenous injection of SNAP-M in tumor-bearing mice treated with an M1-oriented macrophage-polarizing agent (R848), the tumors displayed a notably brighter sonoafterglow compared to saline-treated mice (Fig. 24f), and sonoafterglow exhibited higher SBR than that of photoafterglow (Fig. 24g). Flow cytometry analysis of the M1 macrophage population further confirmed a close correlation between SNAP-M signal and the level of intra-tumoural M1 macrophage during immunotherapy. Additionally, the SNAP system was expanded into a cancer immunotheranostic system (SCAN) (Fig. 24h), incorporating a methylene blue (MB) derivative caged with a ONOO⁻-cleavable moiety as ONOO⁻-responsive sonosensitizer and a ¹O₂-activable prodrug (Pro-R837). In vivo validation demonstrated the effectiveness of SCAN in conducting the sonoafterglow immunotheranostic cycle, (Fig. 24i and j), highlighting its potential for precise theranostics. Pu group also explored expanding the types of biomarkers detectable in sonoafterglow imaging to macromolecules like enzymes, developing a granzyme B (GZMB)-activatable sonoafterglow nanoprobe (Q-SNAP) for early and accurate in vivo detection of T cells.³¹⁵ Furthermore, Song et al. made strides in integrating the sonosensitizer and chemical trap precursor into a single molecule, creating an ultrasound-activable probe used for deep tissue and tumor foci imaging.³¹⁶

Fig. 24 Ultrasound activated delayed photon emission for bioimaging realized by generic composite of sonosensitizer and chemical trap precursor. (a) Schematic of evaluation on deep tissue penetrating capacity of afterglow activated by light and ultrasound. (b) Representative sonoafterglow, photoafterglow and fluorescence images of SNAP at different tissue depth. (c) SBR analysis of imaging results in (b). (d) Chemical structure of pro-DPAs and schematic mechanism for ONOO⁻-activated sonoafterglow of SNAP-M. (e) Cellular sonoafterglow intensity of SNAP-M in three microphage subtypes under different treatment. (f) Representative photoafterglow and sonoafterglow images of tumor-bearing mice with or without treatment of R848. (g) SBR analysis of imaging results in (f). (h) Schematic components in SCAN and mechanism for ONOO⁻-activated theranostic system. (i) Sonoafterglow images of mice receiving theranostic recycle utilizing SCAN. (j) Quantitative sonoafterglow intensity on tumor in theranostic recycle. Reproduced with permission: Copyright 2023, Springer Nature (adapted from ref. 312).

Tan et al. have introduced a novel ultrasound-activated chemical trap system for bioimaging, utilizing organic piezoelectric materials.³¹³ This system integrates ROS generation and dioxetane formation into a single component by leveraging the piezoelectric effect and piezocatalyzed oxidation. In this research, an initial screening of various luminescent molecules led to the selection of trianthracene derivative-based nanoparticles (TD NPs) due to their superior ultrasound-induced luminescence. Mechanistic investigation revealed that the TD NPs generate polarization charges under ultrasound activation through the piezoelectric effect. This polarization facilitates the production of ¹O₂ and ˙OH via piezocatalysis. The subsequent addition of these ROS with TD molecules leads to the formation of TD–˙OH intermediates or dioxetane intermediates. These intermediates undergo C–C bond rupture after reacting with O₂ or direct cleavage, respectively, thereby releasing stored chemical energy through photon generation (Fig. 25a). The piezoeffect of TD and ¹O₂ generation of TD NPs were confirmed by reproducible voltage output monitoring and ESR spectrum (Fig. 25b and c), respectively. The TD NPs were subsequently employed in various bioimaging applications including orthotopic glioma imaging (Fig. 25d), metastatic tumor diagnosis (Fig. 25e) and lymph node mapping (Fig. 25f). These applications demonstrated bright ultrasound-induced luminescence signals with high SBR across all imaging results. Additionally, the TD NPs were further developed into activable probes by introduction of a luminescence quencher coupled with biomarker-cleavable peptide or ROS-degradable energy transfer acceptor (Fig. 25g). These activatable probes proved successful in monitoring immunotherapy, effectively distinguishing immune response in CT-26 and 4T1 tumor (Fig. 25h) and predicting abscopal response in distant tumor (Fig. 25i).

Fig. 25 US activated delayed photon emission for bioimaging realized by chemical trap precursors with piezocatalytic capacity. (a) Schematic illustration for ultrasound-induced piezocatalysis and subsequent chemiluminescent process in TD NPs. (b) Reproducible voltage output of TD under ultrasound treatment. (c) ¹O₂ ESR spectra of TD NPs after ultrasound treatment. (d) Schematic illustration and representative images of orthotopic glioma imaging (d), metastasis tumor imaging (e) and lymph node mapping (f) on mice treated with TD NPs. (g) Schematic designing of biomarker-responsive TD NPs. (h) Representative images of granzyme B-responsive ultrasound-induced luminescence on tumor-bearing mice for distinguishing immune response (left) and the corresponding quantitative analysis (right). (i) Representative images of granzyme B-responsive ultrasound-induced luminescence on tumor-bearing mice for predicting abscopal response (left) and the corresponding quantitative analysis (right). Reproduced with permission: Copyright 2024, Springer Nature (adapted from ref. 313).

5. Summary and outlook

Delayed photon emission offers a promising strategy to enhance SBR and imaging sensitivity in optical bioimaging by minimizing tissue autofluorescence. This review discusses the distinct mechanisms and versatile design strategies of organic molecular probes that facilitate delayed photon emission, thereby improving resolution and accuracy in biological imaging. This process involves organic molecular probes storing and releasing energy through different metastable high-energy species, such as charge carriers, excitons or chemical intermediates. Upon light irradiation, these probes capture photoenergy within defects, generating high-energy species that yield emissive products with delayed emission in the absence of real-time external light excitation, thus extending emission lifetimes. This delayed photon emission phenomenon occurs through various stages in the life of the excited species. Charge separation and recombination involve the diffusion of charge carriers, transforming excitation energy into the ionization potential of separated radical cations and anions. This process significantly extends the lifetime of exciplex emission, often lasting several hours due to its power-law decay feature. Generation, stabilization and conversion of triplet excitons involve accessible energy states between different spin multiplicity. This process stores energy in stabilized triplet excitons that exhibit spin-forbidden radiative decay, thereby extending the lifespan of these excited species. Subsequently, efficient photon generation occurs through spin conversion from triplet to singlet states. Generation and decomposition of chemical traps, such as dioxetane intermediates, involve the addition of reactive oxygen species through sensitization, followed by thermolysis or CIEEL to result in the production of emissive species, allowing for a chemically controlled delay in photon generation. Extensive mechanistic exploration and the development of design strategies across these three routes for delayed photon emission have facilitated sensitive and precise visualization with high SBR for early diagnostics and prognosis evaluation in bioimaging. Moreover, the penetration depth for imaging deep-seated lesion presents a challenge for light-activated systems. Alternative energy sources such as X-ray and US have superior tissue penetration abilities. X-ray photons are absorbed by heavy atoms, while US mechanical waves can be harnessed by liquid media or piezoelectric materials, both facilitating the generation of charge carriers, triplet excitons and chemical intermediates. These enhancements in energy input not only expand noninvasive methods for generating emissive species in deep tissue but also enhance the clinical potential of delayed photon emission technologies. While the signal output from living body is still constrained by the tissue-penetrating limitations of light, the use of X-ray/ultrasound (US) energy input has significantly improved the imaging depth of delayed photon emission, with reported penetration depths of up to 4 cm.²⁹⁸ This progress suggests a promising potential for delayed photon emission to enable sensitive bioimaging through substantial tissue depth, offering opportunities for clinical application in optical imaging-guided surgical navigation or adjuvant therapy where a high signal-to-noise ratio is crucial.

Despite significant advancements in light, X-ray, and US activated delayed photon emission, the development of design strategies for organic molecular probes capable of energy conversion into long-lifetime emissive species in bioimaging remains in its nascent stages. Challenges include a lack of universal principles and difficulties in integrating emissive species with specific chemical building blocks in complex biological environments. Enhanced investigations into the feasibility of these reactive emitters for bioimaging is critical, and as delineated here.

(1) Emissive species, such as charge carriers, triplet excitons and chemical intermediates, demonstrate high sensitivity to quenching environments rich in oxygen and water. Detailed exploration of the quenching mechanisms is essential to understand how biological substrates influence the generation and sustained emission of these species. This understanding could enable the development of encapsulation strategies that effectively counteract key quenching processes.

(2) As discussed in Section 2.4, there are few reported instances of utilizing charge separation and recombination for delayed photon emission in bioimaging, probably due to the intrinsic quenching properties of such systems in biological environment. Typical charge separating systems depend on donor–acceptor pairs for charge carrier generation, charge diffusion and exciplex emission, all susceptible to energy or electron transfer with environmental quenchers. Introducing additional stabilizers within the donor–acceptor doping system and designing structures that regulate the redox potential of both donor and acceptor can shield the delayed photon emission from both energetic or electric quenching, enhancing the viability of charge separating system for biomedical applications.

(3) Organic triplet excitons, generated through strong SOC and small ΔE_ST, inherently facilitate fast singlet–triplet crossing. The optimization of SOC and ΔE_ST is crucial to balancing the lifetime and emission efficiency. Ensuring an adequate lifetime and efficient spin conversion from triplet to singlet is essential for further stabilization through energy trapping and the subsequent TADF process. This method could harmonize improvements in both lifetime and emission efficiency of triplet excitons. In this sequential process, excimers with specific aggregation patterns and precisely engineered energy states acts as effective traps that enhance triplet excitons stabilization through efficient triplet–triplet energy transfer.

(4) In chemical traps, stabilizers and CIEEL promoters function as antagonistic pairs that influence the decomposition dynamics of dioxetane intermediates. Commonly, the adamantylidene stabilizer paired with the phenoxide anion promoter facilitates persistent emission extended to hours. However, this prolonged emission process often yields low instantaneous brightness, potentially compromising the spatiotemporal resolution in bioimaging applications. Tailoring molecular designs to optimize the choice of stabilizer-promoter pairs could enhance the brightness of delayed photon emissions, thereby improving the utility of chemical traps in bioimaging applications.

(5) X-ray activated inorganic systems achieved delayed emission via electron traps with reproducibility and biological orthogonality, yet they are often hampered by inherent toxicity. Given the demonstrated feasibility of X-ray induced charge separation in organic system, further development of thermally activatable traps for organic molecules could mimic this trapping–detrapping process. This approach could foster the generation of long-lifetime organic charge carriers, presenting a promising avenue for achieving X-ray activated charge separation and recombination for delayed exciplex emission.

(6) The piezoelectric effect activated by US can induce internal polarization within organic piezoelectric materials, which is similar to charge separation process. Developing design strategies that integrate piezoelectric polarization with charge stabilization and gradual charge recombination for exciplex emission, could open new avenues for achieving US-activated delayed photon emission.

(7) Delayed photon emission typically exhibits continuous but low-intensity emission, necessitating long exposure times for signal acquisition, often extending to seconds or minutes. This limitation hinders the application of delayed photon emission in real-time imaging of biological events. Further investigations into the chemical structure of delayed-emitting probes are essential to enhance the brightness of delayed photon emission, which would not only reduce exposure times but also improve the signal-to-noise ratio in deep-tissue imaging.

(8) Achieving delayed photon emission through various emissive species and energy inputs presents challenges in signal activation and acquisition using standard imaging equipment. Additionally, balancing the cost trade-offs between molecular design, which aims to extend the lifespan of emissive species, and the instrumental design, which focuses on minimizing exposure times, is crucial for cost-effective bioimaging. Therefore, advances in instrumental design for activating and capturing delayed photon emission are as important as the development of the probes themselves. Moreover, the poor maneuverability of instruments capable of generating sufficient high-energy photons or mechanical waves is a significant barrier to the clinical adoption of X-ray/US activated delayed photon emission. For instance, surgical navigation and gastrointestinal endoscope require compact and steerable excitation sources, whereas traditional X-ray/US energy sources are too large for these potential clinical applications. Therefore, instrument miniaturization is an essential step towards advancing the clinical application of X-ray/US activated delayed photon emission.

(9) The electron configuration, energy gaps between different states, spin multiplicity, and intermolecular interactions are all critical factors influencing the generation and decomposition rate of emitters, which directly affect their emission lifespan and brightness. However, there have been limited investigation into the emission dynamics of charge carriers and chemical traps. Theoretical calculations and ultrafast spectroscopic techniques provide valuable insights into the excited states of chromophores.^317–319 Further understanding of the singlet-to-triplet ratio in recombined exciplexes and decomposed dioxetanes under various biological conditions can clarify quenching mechanisms and guide the structural design of high-brightness emitters.

We believe that, through leveraging rapid advancements in mechanistic exploration, molecular engineering, and instrumental design, the next generation of organic probes with delayed photon emission properties will possess validated mechanisms and an expanded structural library. These advancements are expected to enhance spatiotemporal resolution and tissue penetration capabilities in bioimaging and promise a seamless translation into clinical applications.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

X. Z. and X. J. thank the National Key R & D Program of China (No. 2022YFB3804600), the National Natural Science Foundation of China (No. 52173130, 52373141, 92163214, and 52333003), the Fundamental Research Funds for the Central Universities (No. 020514380274 and 020514380330) and the Natural Science Foundation of Jiangsu Province (BK20202002) for the financial support.

References

1. M. L. James and S. S. Gambhir, Physiol. Rev., 2012, 92, 897–965.
2. X. Gao, Y. Cui, R. M. Levenson, L. W. K. Chung and S. Nie, Nat. Biotechnol., 2004, 22, 969–976.
3. G. Hong, A. L. Antaris and H. Dai, Nat. Biomed. Eng., 2017, 1, 1–22.
4. B. R. Smith and S. S. Gambhir, Chem. Rev., 2017, 117, 901–986.
5. X. Zhen and K. Pu, Nano Res., 2018, 11, 5258–5280.
6. H. Yin, W. Jiang, Y. Liu, D. Zhang, F. Wu, Y. Zhang, C. Li, G. Chen and Q. Wang, BMEMat, 2023, 1, e12023.
7. Z. Lou, P. Li and K. Han, Acc. Chem. Res., 2015, 48, 1358–1368.
8. C. T. W. Moonen, P. C. M. van Zijl, J. A. Frank, D. Le Bihan and E. D. Becker, Science, 1990, 250, 53–61.
9. P. M. Matthews and P. Jezzard, J. Neurol., Neurosurg. Psychiatry, 2004, 75, 6–12.
10. E. Terreno, D. D. Castelli, A. Viale and S. Aime, Chem. Rev., 2010, 110, 3019–3042.
11. E. Macé, G. Montaldo, I. Cohen, M. Baulac, M. Fink and M. Tanter, Nat. Methods, 2011, 8, 662–664.
12. P. N. T. Wells and H.-D. Liang, J. R. Soc., Interface, 2011, 8, 1521–1549.
13. G. D. Rubin, Radiology, 2014, 273, S45–S74.
14. S. Achenbach, J. Am. Coll. Cardiol., 2006, 48, 1919–1928.
15. J. B. Bomanji, D. C. Costa and P. J. Ell, Lancet Oncol., 2001, 2, 157–164.
16. J. Czernin and M. E. Phelps, Annu. Rev. Med., 2002, 53, 89–112.
17. X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J. Li, G. Sundaresan, A. M. Wu, S. S. Gambhir and S. Weiss, Science, 2005, 307, 538–544.
18. L. Chen, X. Zhen and X. Jiang, Adv. NanoBiomed Res., 2022, 2, 2200097.
19. T. Maldiney, A. Bessière, J. Seguin, E. Teston, S. K. Sharma, B. Viana, A. J. J. Bos, P. Dorenbos, M. Bessodes, D. Gourier, D. Scherman and C. Richard, Nat. Mater., 2014, 13, 418–426.
20. S. He, J. Song, J. Qu and Z. Cheng, Chem. Soc. Rev., 2018, 47, 4258–4278.
21. M. Yang, Z. Zeng, J. W. Y. Lam, J. Fan, K. Pu and B. Z. Tang, Chem. Soc. Rev., 2022, 51, 8815–8831.
22. Y. Jiang and K. Pu, Chem. Rev., 2021, 121, 13086–13131.
23. X. Zhang, H. Suo, Y. Guo, J. Chen, Y. Wang, X. Wei, W. Zheng, S. Li and F. Wang, Nat. Commun., 2024, 15, 6797.
24. L. Yang, S. Gai, H. Ding, D. Yang, L. Feng and P. Yang, Adv. Opt. Mater., 2023, 11, 2202382.
25. S. Wang, R. Liu, J. Li, C. Meng, J. Liu, J. Chen, P. Cheng and K. Wu, Angew. Chem., Int. Ed., 2024, 63, e202403927.
26. X. Yang, G. I. N. Waterhouse, S. Lu and J. Yu, Chem. Soc. Rev., 2023, 52, 8005–8058.
27. R. Kabe and C. Adachi, Nature, 2017, 550, 384–387.
28. X. Wang, M. Dong, Z. Li, Z. Wang and F.-S. Liang, Dyes Pigm., 2022, 204, 110400.
29. Z. An, C. Zheng, Y. Tao, R. Chen, H. Shi, T. Chen, Z. Wang, H. Li, R. Deng, X. Liu and W. Huang, Nat. Mater., 2015, 14, 685–690.
30. L. Gu, H. Shi, M. Gu, K. Ling, H. Ma, S. Cai, L. Song, C. Ma, H. Li and G. Xing, Angew. Chem., Int. Ed., 2018, 57, 8425–8431.
31. Q. Miao, C. Xie, X. Zhen, Y. Lyu, H. Duan, X. Liu, J. V. Jokerst and K. Pu, Nat. Biotechnol., 2017, 35, 1102–1110.
32. G. Ma, M. Dirak, Z. Liu, D. Jiang, Y. Wang, C. Xiang, Y. Zhang, Y. Luo, P. Gong and L. Cai, Angew. Chem., Int. Ed., 2024, 136, e202400658.
33. A. M. Smith, M. C. Mancini and S. Nie, Nat. Nanotechnol., 2009, 4, 710–711.
34. S. Diao, J. L. Blackburn, G. Hong, A. L. Antaris, J. Chang, J. Z. Wu, B. Zhang, K. Cheng, C. J. Kuo and H. Dai, Angew. Chem., Int. Ed., 2015, 54, 14758–14762.
35. P. J. Withers, C. Bouman, S. Carmignato, V. Cnudde, D. Grimaldi, C. K. Hagen, E. Maire, M. Manley, A. Du Plessis and S. R. Stock, Nat. Rev. Methods Primers, 2021, 1, 1–21.
36. Y.-S. Chen, Y. Zhao, C. Beinat, A. Zlitni, E.-C. Hsu, D.-H. Chen, F. Achterberg, H. Wang, T. Stoyanova, J. Dionne and S. S. Gambhir, Nat. Nanotechnol., 2021, 16, 717–724.
37. P. I. H. Bastiaens, A. Squire, P. I. H. Bastiaens, A. Squire, P. I. H. Bastiaens and A. Squire, Trends Cell Biol., 1999, 9, 48–52.
38. M. Y. Berezin and S. Achilefu, Chem. Rev., 2010, 110, 2641–2684.
39. A. J. Bowman and M. A. Kasevich, ACS Nano, 2021, 15, 16043–16054.
40. L. Liang, J. Chen, K. Shao, X. Qin, Z. Pan and X. Liu, Nat. Mater., 2023, 22, 289–304.
41. H. Xiang and Y. Chen, Small, 2019, 15, 1805339.
42. J. Chen, Y. Zhu, C. Wu and J. Shi, Chem. Soc. Rev., 2020, 49, 9057–9094.
43. H. Xu, D. Kim, Y.-Y. Zhao, C. Kim, G. Song, Q. Hu, H. Kang and J. Yoon, Adv. Mater., 2024, 2402806,  DOI:10.1002/adma.202402806.
44. O. P. Dimitriev, Chem. Rev., 2022, 122, 8487–8593.
45. R. Qu, X. Zhen and X. Jiang, CCS Chem., 2022, 4, 401–419.
46. H. Li, S. Zhou, M. Wu, R. Qu, X. Wang, W. Chen, Y. Jiang, X. Jiang and X. Zhen, Adv. Mater., 2023, 35, 2210920.
47. M. Wu, R. Qu, H. Li, L. Chen, X. Zhang, Y. Yuan, W. Chen, X. Jiang and X. Zhen, Nano Today, 2023, 48, 101691.
48. W. Zhou, X. Jiang and X. Zhen, Biomater. Sci., 2023, 11, 5108–5128.
49. W. Zhou, L. Chen, H. Li, M. Wu, M. Liang, Q. Liu, W. Wu, X. Jiang and X. Zhen, ACS Nano, 2024, 18, 19771–19782.
50. X. Liu, H. Li, M. Pang, J. Liu, X. Song, R. He, M. He, X. Jian, C. Ma and H. Deng, View, 2024, 5, 20240023.
51. G. N. Lewis and M. Kasha, J. Am. Chem. Soc., 1944, 66, 2100–2116.
52. S. K. Lower and M. A. El-Sayed, Chem. Rev., 1966, 66, 199–241.
53. K. Goushi, K. Yoshida, K. Sato and C. Adachi, Nat. Photonics, 2012, 6, 253–258.
54. H. Noda, H. Nakanotani and C. Adachi, Sci. Adv., 2018, 4, eaao6910.
55. Y. Wada, H. Nakagawa, S. Matsumoto, Y. Wakisaka and H. Kaji, Nat. Photonics, 2020, 14, 643–649.
56. C. Ye, V. Gray, J. Mårtensson and K. Börjesson, J. Am. Chem. Soc., 2019, 141, 9578–9584.
57. P. Bharmoria, H. Bildirir and K. Moth-Poulsen, Chem. Soc. Rev., 2020, 49, 6529–6554.
58. L. Zeng, L. Huang, J. Han and G. Han, Acc. Chem. Res., 2022, 55, 2604–2615.
59. C. J. Bardeen, Annu. Rev. Phys. Chem., 2014, 65, 127–148.
60. J. Dostál, F. Fennel, F. Koch, S. Herbst, F. Würthner and T. Brixner, Nat. Commun., 2018, 9, 2466.
61. S. Xu and B. Liu, J. Am. Chem. Soc., 2022, 144, 17897–17904.
62. T. J. Kim, S.-h Lee, D. Kwon and J. Joo, J. Mater. Chem. C, 2024, 12, 404–437.
63. F. C. Spano, Acc. Chem. Res., 2010, 43, 429–439.
64. T. Brixner, R. Hildner, J. Köhler, C. Lambert and F. Würthner, Adv. Energy Mater., 2017, 7, 1700236.
65. S. Ma, S. Du, G. Pan, S. Dai, B. Xu and W. Tian, Aggregate, 2021, 2, e96.
66. V. B. Nazarov, V. I. Gerko and M. V. Alfimov, Russ. Chem. Bull., 1996, 45, 969–970.
67. M. Baroncini, G. Bergamini and P. Ceroni, Chem. Commun., 2017, 53, 2081–2093.
68. M. Hayduk, S. Riebe and J. Voskuhl, Chem. – Eur. J., 2018, 24, 12221–12230.
69. Q. Peng, H. Ma and Z. Shuai, Acc. Chem. Res., 2021, 54, 940–949.
70. T. Yamanaka, H. Nakanotani and C. Adachi, Nat. Commun., 2019, 10, 5748.
71. S. Tan, K. Jinnai, R. Kabe and C. Adachi, Adv. Mater., 2021, 33, 2008844.
72. Z. Wang, X. Song, Q. Chen, J. Liang, Z. Tu, C. Wang, Z. Zhang, H. Bi and Y. Wang, Adv. Funct. Mater., 2022, 32, 2206207.
73. K. H. Lee, S. O. Jeon, Y. S. Chung, J.-M. Kim and J. Y. Lee, Adv. Opt. Mater., 2023, 11, 2201687.
74. D. M. Hercules and F. E. Lytle, J. Am. Chem. Soc., 1966, 88, 4745–4746.
75. A. U. Khan and M. Kasha, J. Am. Chem. Soc., 1966, 88, 1574–1576.
76. W. Adam, D. V. Kazakov and V. P. Kazakov, Chem. Rev., 2005, 105, 3371–3387.
77. X. Wang and K. Pu, Chem. Soc. Rev., 2023, 52, 4549–4566.
78. N. Nishimura, Z. Lin, K. Jinnai, R. Kabe and C. Adachi, Adv. Funct. Mater., 2020, 30, 2000795.
79. T. Matsuzawa, Y. Aoki, N. Takeuchi and Y. Murayama, J. Electrochem. Soc., 1996, 143, 2670.
80. S. Xu, R. Chen, C. Zheng and W. Huang, Adv. Mater., 2016, 28, 9920–9940.
81. G. N. Lewis and D. Lipkin, J. Am. Chem. Soc., 1942, 64, 2801–2808.
82. H. Ohkita, W. Sakai, A. Tsuchida and M. Yamamoto, Macromolecules, 1997, 30, 5376–5383.
83. H. Ohkita, W. Sakai, A. Tsuchida and M. Yamamoto, J. Phys. Chem. B, 1997, 101, 10241–10247.
84. J.-L. Brédas, E. H. Sargent and G. D. Scholes, Nat. Mater., 2017, 16, 35–44.
85. Y. Sun, N. C. Giebink, H. Kanno, B. Ma, M. E. Thompson and S. R. Forrest, Nature, 2006, 440, 908–912.
86. F. Li, A. J. Gillett, Q. Gu, J. Ding, Z. Chen, T. J. H. Hele, W. K. Myers, R. H. Friend and E. W. Evans, Nat. Commun., 2022, 13, 2744.
87. G. C. Abell and A. Mozumder, J. Chem. Phys., 1972, 56, 4079–4085.
88. P. Debye and J. O. Edwards, J. Chem. Phys., 1952, 20, 236–239.
89. Z. Lin, R. Kabe, N. Nishimura, K. Jinnai and C. Adachi, Adv. Mater., 2018, 30, 1803713.
90. K. Jinnai, R. Kabe, Z. Lin and C. Adachi, Nat. Mater., 2022, 21, 338–344.
91. H. Usta, C. Risko, Z. Wang, H. Huang, M. K. Deliomeroglu, A. Zhukhovitskiy, A. Facchetti and T. J. Marks, J. Am. Chem. Soc., 2009, 131, 5586–5608.
92. K. Zhou, H. Dong, H.-L. Zhang and W. Hu, Phys. Chem. Chem. Phys., 2014, 16, 22448–22457.
93. S. Fukuzumi, H. Kotani, K. Ohkubo, S. Ogo, N. V. Tkachenko and H. Lemmetyinen, J. Am. Chem. Soc., 2004, 126, 1600–1601.
94. M. Sakurai, R. Kabe, M. Fuki, Z. Lin, K. Jinnai, Y. Kobori, C. Adachi and T. Tachikawa, Commun. Mater., 2021, 2, 74.
95. K. Jinnai, R. Kabe and C. Adachi, Adv. Mater., 2018, 30, 1800365.
96. Z. Lin, R. Kabe, K. Wang and C. Adachi, Nat. Commun., 2020, 11, 191.
97. B. Zhou and D. Yan, Adv. Funct. Mater., 2019, 29, 1807599.
98. P. Alam, N. L. C. Leung, J. Liu, T. S. Cheung, X. Zhang, Z. He, R. T. K. Kwok, J. W. Y. Lam, H. H. Y. Sung, I. D. Williams, C. C. S. Chan, K. S. Wong, Q. Peng and B. Z. Tang, Adv. Mater., 2020, 32, 2001026.
99. P. Alam, T. S. Cheung, N. L. C. Leung, J. Zhang, J. Guo, L. Du, R. T. K. Kwok, J. W. Y. Lam, Z. Zeng, D. L. Phillips, H. H. Y. Sung, I. D. Williams and B. Z. Tang, J. Am. Chem. Soc., 2022, 144, 3050–3062.
100. S. Li, L. Fu, X. Xiao, H. Geng, Q. Liao, Y. Liao and H. Fu, Angew. Chem., Int. Ed., 2021, 60, 18059–18064.
101. T. Chatterjee and K.-T. Wong, Adv. Opt. Mater., 2019, 7, 1800565.
102. B. Liu, V. M. Menon and M. Y. Sfeir, ACS Photonics, 2020, 7, 2292–2301.
103. D. Sasikumar, A. T. John, J. Sunny and M. Hariharan, Chem. Soc. Rev., 2020, 49, 6122–6140.
104. S. Gorgon, K. Lv, J. Grüne, B. H. Drummond, W. K. Myers, G. Londi, G. Ricci, D. Valverde, C. Tonnelé, P. Murto, A. S. Romanov, D. Casanova, V. Dyakonov, A. Sperlich, D. Beljonne, Y. Olivier, F. Li, R. H. Friend and E. W. Evans, Nature, 2023, 620, 538–544.
105. H. Jin, C. Livache, W. D. Kim, B. T. Diroll, R. D. Schaller and V. I. Klimov, Nat. Mater., 2023, 22, 1013–1021.
106. M. A. El-Sayed, Acc. Chem. Res., 1968, 1, 8–16.
107. C. Lévêque, D. Peláez, H. Köppel and R. Taïeb, Nat. Commun., 2014, 5, 4126.
108. R. Feng, X. Yu and J. Autschbach, J. Chem. Theory Comput., 2021, 17, 7531–7544.
109. X. Zhen, R. Qu, W. Chen, W. Wu and X. Jiang, Biomater. Sci., 2021, 9, 285–300.
110. G. Baryshnikov, B. Minaev and H. Ågren, Chem. Rev., 2017, 117, 6500–6537.
111. Y. Tao, K. Yuan, T. Chen, P. Xu, H. Li, R. Chen, C. Zheng, L. Zhang and W. Huang, Adv. Mater., 2014, 26, 7931–7958.
112. Z. Yang, Z. Mao, Z. Xie, Y. Zhang, S. Liu, J. Zhao, J. Xu, Z. Chi and M. P. Aldred, Chem. Soc. Rev., 2017, 46, 915–1016.
113. D. Polak, R. Jayaprakash, T. P. Lyons, L. Á. Martínez-Martínez, A. Leventis, K. J. Fallon, H. Coulthard, D. G. Bossanyi, K. Georgiou, I. I. Anthony, J. Petty, J. Anthony, H. Bronstein, J. Yuen-Zhou, A. I. Tartakovskii, J. Clark and A. J. Musser, Chem. Sci., 2020, 11, 343–354.
114. R. M. Young and M. R. Wasielewski, Acc. Chem. Res., 2020, 53, 1957–1968.
115. W. Zhao, Z. He and B. Z. Tang, Nat. Rev. Mater., 2020, 5, 869–885.
116. X. Yang, G. Zhou and W.-Y. Wong, Chem. Soc. Rev., 2015, 44, 8484–8575.
117. Y. Liu, C. Li, Z. Ren, S. Yan and M. R. Bryce, Nat. Rev. Mater., 2018, 3, 1–20.
118. H. Shi, W. Yao, W. Ye, H. Ma, W. Huang and Z. An, Acc. Chem. Res., 2022, 55, 3445–3459.
119. J. Chen, X. Chen, L. Cao, H. Deng, Z. Chi and B. Liu, Angew. Chem., Int. Ed., 2022, 134, e202200343.
120. A. P. Demchenko, V. I. Tomin and P.-T. Chou, Chem. Rev., 2017, 117, 13353–13381.
121. M. B. Smith and J. Michl, Chem. Rev., 2010, 110, 6891–6936.
122. D. Casanova, Chem. Rev., 2018, 118, 7164–7207.
123. S. M. Sartor, B. G. McCarthy, R. M. Pearson, G. M. Miyake and N. H. Damrauer, J. Am. Chem. Soc., 2018, 140, 4778–4781.
124. C. M. Marian, Annu. Rev. Phys. Chem., 2021, 72, 617–640.
125. C. M. Marian, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2, 187–203.
126. Z. Shuai and Q. Peng, Phys. Rep., 2014, 537, 123–156.
127. A. Lv, W. Ye, X. Jiang, N. Gan, H. Shi, W. Yao, H. Ma, Z. An and W. Huang, J. Phys. Chem. Lett., 2019, 10, 1037–1042.
128. T. J. Penfold, E. Gindensperger, C. Daniel and C. M. Marian, Chem. Rev., 2018, 118, 6975–7025.
129. S. Schott, E. R. McNellis, C. B. Nielsen, H.-Y. Chen, S. Watanabe, H. Tanaka, I. McCulloch, K. Takimiya, J. Sinova and H. Sirringhaus, Nat. Commun., 2017, 8, 15200.
130. L. Xu, G. Li, T. Xu, W. Zhang, S. Zhang, S. Yin, Z. An and G. He, Chem. Commun., 2018, 54, 9226–9229.
131. G. He, W. Torres Delgado, D. J. Schatz, C. Merten, A. Mohammadpour, L. Mayr, M. J. Ferguson, R. McDonald, A. Brown, K. Shankar and E. Rivard, Angew. Chem., Int. Ed., 2014, 53, 4587–4591.
132. R. A. Caldwell, L. D. Jacobs, T. R. Furlani, E. A. Nalley and J. Laboy, J. Am. Chem. Soc., 1992, 114, 1623–1625.
133. J. Wang, C. Wang, Y. Gong, Q. Liao, M. Han, T. Jiang, Q. Dang, Y. Li, Q. Li and Z. Li, Angew. Chem., Int. Ed., 2018, 130, 17063–17068.
134. L. Xiao, Y. Wu, J. Chen, Z. Yu, Y. Liu, J. Yao and H. Fu, J. Phys. Chem. A, 2017, 121, 8652–8658.
135. Z. Mao, Z. Yang, Y. Mu, Y. Zhang, Y.-F. Wang, Z. Chi, C.-C. Lo, S. Liu, A. Lien and J. Xu, Angew. Chem., Int. Ed., 2015, 54, 6270–6273.
136. J. Wang, X. Gu, H. Ma, Q. Peng, X. Huang, X. Zheng, S. H. P. Sung, G. Shan, J. W. Y. Lam, Z. Shuai and B. Z. Tang, Nat. Commun., 2018, 9, 2963.
137. P. Ceroni, Chem, 2016, 1, 524–526.
138. Kenry, C. Chen and B. Liu, Nat. Commun., 2019, 10, 2111.
139. B. Xu, H. Wu, J. Chen, Z. Yang, Z. Yang, Y.-C. Wu, Y. Zhang, C. Jin, P.-Y. Lu, Z. Chi, S. Liu, J. Xu and M. Aldred, Chem. Sci., 2017, 8, 1909–1914.
140. Y. Tao, R. Chen, H. Li, J. Yuan, Y. Wan, H. Jiang, C. Chen, Y. Si, C. Zheng, B. Yang, G. Xing and W. Huang, Adv. Mater., 2018, 30, 1803856.
141. C. Zhou, S. Zhang, Y. Gao, H. Liu, T. Shan, X. Liang, B. Yang and Y. Ma, Adv. Funct. Mater., 2018, 28, 1802407.
142. Y. Takeda, T. Kaihara, M. Okazaki, H. Higginbotham, P. Data, N. Tohnai and S. Minakata, Chem. Commun., 2018, 54, 6847–6850.
143. W. Zhao, Z. He, J. W. Y. Lam, Q. Peng, H. Ma, Z. Shuai, G. Bai, J. Hao and B. Z. Tang, Chem, 2016, 1, 592–602.
144. M. Shimizu, R. Shigitani, M. Nakatani, K. Kuwabara, Y. Miyake, K. Tajima, H. Sakai and T. Hasobe, J. Phys. Chem. C, 2016, 120, 11631–11639.
145. Y. Shoji, Y. Ikabata, Q. Wang, D. Nemoto, A. Sakamoto, N. Tanaka, J. Seino, H. Nakai and T. Fukushima, J. Am. Chem. Soc., 2017, 139, 2728–2733.
146. H. Matsuoka, M. Retegan, L. Schmitt, S. Höger, F. Neese and O. Schiemann, J. Am. Chem. Soc., 2017, 139, 12968–12975.
147. Y. Xiong, Z. Zhao, W. Zhao, H. Ma, Q. Peng, Z. He, X. Zhang, Y. Chen, X. He, J. W. Y. Lam and B. Z. Tang, Angew. Chem., Int. Ed., 2018, 57, 7997–8001.
148. X. Chen, C. Xu, T. Wang, C. Zhou, J. Du, Z. Wang, H. Xu, T. Xie, G. Bi, J. Jiang, X. Zhang, J. N. Demas, C. O. Trindle, Y. Luo and G. Zhang, Angew. Chem., Int. Ed., 2016, 55, 9872–9876.
149. Z. Yang, Z. Mao, X. Zhang, D. Ou, Y. Mu, Y. Zhang, C. Zhao, S. Liu, Z. Chi, J. Xu, Y.-C. Wu, P.-Y. Lu, A. Lien and M. R. Bryce, Angew. Chem., Int. Ed., 2016, 55, 2181–2185.
150. M. K. Etherington, J. Gibson, H. F. Higginbotham, T. J. Penfold and A. P. Monkman, Nat. Commun., 2016, 7, 13680.
151. J. Yuan, S. Wang, Y. Ji, R. Chen, Q. Zhu, Y. Wang, C. Zheng, Y. Tao, Q. Fan and W. Huang, Mater. Horiz., 2019, 6, 1259–1264.
152. S. Tian, H. Ma, X. Wang, A. Lv, H. Shi, Y. Geng, J. Li, F. Liang, Z.-M. Su, Z. An and W. Huang, Angew. Chem., Int. Ed., 2019, 58, 6645–6649.
153. J. Yang, X. Zhen, B. Wang, X. Gao, Z. Ren, J. Wang, Y. Xie, J. Li, Q. Peng, K. Pu and Z. Li, Nat. Commun., 2018, 9, 840.
154. J. T. Buck, A. M. Boudreau, A. DeCarmine, R. W. Wilson, J. Hampsey and T. Mani, Chem, 2019, 5, 138–155.
155. S. Hirata, K. Totani, J. Zhang, T. Yamashita, H. Kaji, S. R. Marder, T. Watanabe and C. Adachi, Adv. Funct. Mater., 2013, 23, 3386–3397.
156. G. W. Robinson and R. P. Frosch, J. Chem. Phys., 1963, 38, 1187–1203.
157. W. Siebrand, J. Chem. Phys., 1967, 46, 440–447.
158. K. Schmidt, S. Brovelli, V. Coropceanu, D. Beljonne, J. Cornil, C. Bazzini, T. Caronna, R. Tubino, F. Meinardi, Z. Shuai and J.-L. Brédas, J. Phys. Chem. A, 2007, 111, 10490–10499.
159. M. Hayduk, S. Riebe and J. Voskuhl, Chem. – Eur. J., 2018, 24, 12221–12230.
160. W. Z. Yuan, X. Y. Shen, H. Zhao, J. W. Y. Lam, L. Tang, P. Lu, C. Wang, Y. Liu, Z. Wang, Q. Zheng, J. Z. Sun, Y. Ma and B. Z. Tang, J. Phys. Chem. C, 2010, 114, 6090–6099.
161. Y. Gong, G. Chen, Q. Peng, W. Z. Yuan, Y. Xie, S. Li, Y. Zhang and B. Z. Tang, Adv. Mater., 2015, 27, 6195–6201.
162. H. Wu, Y. Zhou, L. Yin, C. Hang, X. Li, H. Ågren, T. Yi, Q. Zhang and L. Zhu, J. Am. Chem. Soc., 2017, 139, 785–791.
163. S. Hirata, Adv. Opt. Mater., 2017, 5, 1700116.
164. Z.-Y. Zhang, Y. Chen and Y. Liu, Angew. Chem., Int. Ed., 2019, 58, 6028–6032.
165. Q. Zhou, B. Cao, C. Zhu, S. Xu, Y. Gong, W. Z. Yuan and Y. Zhang, Small, 2016, 12, 6586–6592.
166. H. Zhang, Z. Zhao, P. R. McGonigal, R. Ye, S. Liu, J. W. Y. Lam, R. T. K. Kwok, W. Z. Yuan, J. Xie, A. L. Rogach and B. Z. Tang, Mater. Today, 2020, 32, 275–292.
167. P. F. Jones and S. Siegel, J. Chem. Phys., 1969, 50, 1134–1140.
168. X. Ma, C. Xu, J. Wang and H. Tian, Angew. Chem., Int. Ed., 2018, 57, 10854–10858.
169. Y.-F. Zhang, Y.-C. Wang, X.-S. Yu, Y. Zhao, X.-K. Ren, J.-F. Zhao, J. Wang, X.-Q. Jiang, W.-Y. Chang, J.-F. Zheng, Z.-Q. Yu, S. Yang and E.-Q. Chen, Macromolecules, 2019, 52, 2495–2503.
170. K. Jiang, L. Zhang, J. Lu, C. Xu, C. Cai and H. Lin, Angew. Chem., Int. Ed., 2016, 55, 7231–7235.
171. S. Tao, S. Lu, Y. Geng, S. Zhu, S. A. T. Redfern, Y. Song, T. Feng, W. Xu and B. Yang, Angew. Chem., Int. Ed., 2018, 57, 2393–2398.
172. G.-J. Zhao and K.-L. Han, Acc. Chem. Res., 2012, 45, 404–413.
173. N. Aizawa, Y.-J. Pu, Y. Harabuchi, A. Nihonyanagi, R. Ibuka, H. Inuzuka, B. Dhara, Y. Koyama, K.-I. Nakayama, S. Maeda, F. Araoka and D. Miyajima, Nature, 2022, 609, 502–506.
174. X.-K. Chen, D. Kim and J.-L. Brédas, Acc. Chem. Res., 2018, 51, 2215–2224.
175. R. J. Vázquez, J. H. Yun, A. K. Muthike, M. Howell, H. Kim, I. K. Madu, T. Kim, P. Zimmerman, J. Y. Lee and T. G. Iii, J. Am. Chem. Soc., 2020, 142, 8074–8079.
176. H. Jiang, J. Jin and W.-Y. Wong, Adv. Funct. Mater., 2023, 33, 2306880.
177. H. Uoyama, K. Goushi, K. Shizu, H. Nomura and C. Adachi, Nature, 2012, 492, 234–238.
178. J. Jin, H. Jiang, Q. Yang, L. Tang, Y. Tao, Y. Li, R. Chen, C. Zheng, Q. Fan, K. Y. Zhang, Q. Zhao and W. Huang, Nat. Commun., 2020, 11, 842.
179. W. Li, Z. Li, C. Si, M. Y. Wong, K. Jinnai, A. K. Gupta, R. Kabe, C. Adachi, W. Huang, E. Zysman-Colman and I. D. W. Samuel, Adv. Mater., 2020, 32, 2003911.
180. M. D. Lukin, Rev. Mod. Phys., 2003, 75, 457–472.
181. M. Pelton, Nat. Photonics, 2015, 9, 427–435.
182. M. Palner, K. Pu, S. Shao and J. Rao, Angew. Chem., Int. Ed., 2015, 54, 11477–11480.
183. M. Matsumoto, J. Photochem. Photobiol., C, 2004, 5, 27–53.
184. H. Isobe, Y. Takano, M. Okumura, S. Kuramitsu and K. Yamaguchi, J. Am. Chem. Soc., 2005, 127, 8667–8679.
185. N. Hananya and D. Shabat, Angew. Chem., Int. Ed., 2017, 56, 16454–16463.
186. Y. Jiang, J. Huang, X. Zhen, Z. Zeng, J. Li, C. Xie, Q. Miao, J. Chen, P. Chen and K. Pu, Nat. Commun., 2019, 10, 2064.
187. Y. Su, J. R. Walker, M. P. Hall, M. A. Klein, X. Wu, L. P. Encell, K. M. Casey, L. X. Liu, G. Hong, M. Z. Lin and T. A. Kirkland, Nat. Chem. Biol., 2023, 19, 731–739.
188. O. Shimomura and F. H. Johnson, Proc. Natl. Acad. Sci. U. S. A., 1978, 75, 2611–2615.
189. E. H. White, M. G. Steinmetz, J. D. Miano, P. D. Wildes and R. Morland, J. Am. Chem. Soc., 1980, 102, 3199–3208.
190. J. Li, L. Chen, L. Du and M. Li, Chem. Soc. Rev., 2013, 42, 662–676.
191. O. Green, S. Gnaim, R. Blau, A. Eldar-Boock, R. Satchi-Fainaro and D. Shabat, J. Am. Chem. Soc., 2017, 139, 13243–13248.
192. U. Haris, H. N. Kagalwala, Y. L. Kim and A. R. Lippert, Acc. Chem. Res., 2021, 54, 2844–2857.
193. X. Li, C. Yin, S. S. Liew, C.-S. Lee and K. Pu, Adv. Funct. Mater., 2021, 31, 2106154.
194. C. Chen, X. Zhang, Z. Gao, G. Feng and D. Ding, Nat. Protoc., 2024, 1–27,  DOI:10.1038/s41596-024-00990-4.
195. W. H. Richardson, J. H. Burns, M. E. Price, R. Crawford, M. Foster, P. Slusser and J. H. Anderegg, J. Am. Chem. Soc., 1978, 100, 7596–7600.
196. H. E. O'Neal and W. H. Richardson, J. Am. Chem. Soc., 1970, 92, 6553–6557.
197. W. H. Richardson, J. Org. Chem., 1989, 54, 4677–4684.
198. W. Adam, D. Reinhardt and C. R. Saha-Möller, Analyst, 1996, 121, 1527–1531.
199. M. Vacher, I. Fdez Galván, B.-W. Ding, S. Schramm, R. Berraud-Pache, P. Naumov, N. Ferré, Y.-J. Liu, I. Navizet, D. Roca-Sanjuán, W. J. Baader and R. Lindh, Chem. Rev., 2018, 118, 6927–6974.
200. N. Hananya, A. Eldar Boock, C. R. Bauer, R. Satchi-Fainaro and D. Shabat, J. Am. Chem. Soc., 2016, 138, 13438–13446.
201. O. Green, T. Eilon, N. Hananya, S. Gutkin, C. R. Bauer and D. Shabat, ACS Cent. Sci., 2017, 3, 349–358.
202. M. Matsumoto, H. Suganuma, Y. Katao and H. Mutoh, J. Chem. Soc., Chem. Commun., 1995, 431–432,  10.1039/C39950000431.
203. N. Watanabe, H. Suganuma, H. Kobayashi, H. Mutoh, Y. Katao and M. Matsumoto, Tetrahedron, 1999, 55, 4287–4298.
204. A. L. Baumstark and C. E. Wilson, Tetrahedron Lett., 1981, 22, 4363–4364.
205. M. Matsumoto, H. Murakami and N. Watanabe, Chem. Commun., 1998, 2319–2320.
206. W. H. Richardson and H. E. O'Neal, J. Am. Chem. Soc., 1972, 94, 8665–8668.
207. M. Reguero, F. Bernardi, A. Bottoni, M. Olivucci and M. A. Robb, J. Am. Chem. Soc., 1991, 113, 1566–1572.
208. C. Tanaka and J. Tanaka, J. Phys. Chem. A, 2000, 104, 2078–2090.
209. W. Adam, I. Bronstein, A. V. Trofimov and R. F. Vasil'ev, J. Am. Chem. Soc., 1999, 121, 958–961.
210. J.-C. Tseng and A. L. Kung, J. Biomed. Sci., 2015, 22, 45.
211. J. Liu, J. Huang, X. Wei, P. Cheng and K. Pu, Adv. Mater., 2024, 36, 2310605.
212. Y. Takano, T. Tsunesada, H. Isobe, Y. Yoshioka, K. Yamaguchi and I. Saito, Bull. Chem. Soc. Jpn., 1999, 72, 213–225.
213. M. Matsumoto, T. Sakuma and N. Watanabe, Tetrahedron Lett., 2002, 43, 8955–8958.
214. Y. Wang, G. Song, S. Liao, Q. Qin, Y. Zhao, L. Shi, K. Guan, X. Gong, P. Wang, X. Yin, Q. Chen and X.-B. Zhang, Angew. Chem., Int. Ed., 2021, 60, 19779–19789.
215. D. Cui, C. Xie, J. Li, Y. Lyu and K. Pu, Adv. Healthcare Mater., 2018, 7, 1800329.
216. X. Ni, X. Zhang, X. Duan, H.-L. Zheng, X.-S. Xue and D. Ding, Nano Lett., 2019, 19, 318–330.
217. N. Hananya, O. Green, R. Blau, R. Satchi-Fainaro and D. Shabat, Angew. Chem., Int. Ed., 2017, 56, 11793–11796.
218. H. Yuan, L. Guo, Q. Su, X. Su, Y. Wen, T. Wang, P. Yang, M. Xu and F. Li, ACS Appl. Mater. Interfaces, 2021, 13, 27991–27998.
219. S. He, C. Xie, Y. Jiang and K. Pu, Adv. Mater., 2019, 31, 1902672.
220. C. Chen, H. Gao, H. Ou, R. T. K. Kwok, Y. Tang, D. Zheng and D. Ding, J. Am. Chem. Soc., 2022, 144, 3429–3441.
221. N. Hananya, J. P. Reid, O. Green, M. S. Sigman and D. Shabat, Chem. Sci., 2019, 10, 1380–1385.
222. M. Yang, J. Zhang, D. Shabat, J. Fan and X. Peng, ACS Sens., 2020, 5, 3158–3164.
223. X. Wang, W. Yuan, M. Xu, X. Su and F. Li, ACS Appl. Mater. Interfaces, 2022, 14, 259–267.
224. F. Zhang, M. Xu, X. Su, W. Yuan, W. Feng, Q. Su and F. Li, Chem. – Eur. J., 2022, 28, e202103795.
225. Y. Liu, Y. Li, Y. Wen, X. Su, M. Xu, W. Feng, Q. Liu and F. Li, ACS Mater. Lett., 2021, 3, 713–720.
226. X. Su, Y. Wen, W. Yuan, M. Xu, Q. Liu, C. Huang and F. Li, Chem. Commun., 2020, 56, 10694–10697.
227. X. Su, X. Kong, K. Sun, Q. Liu, Y. Pei, D. Hu, M. Xu, W. Feng and F. Li, Angew. Chem., Int. Ed., 2022, 61, e202201630.
228. V. V. Roznyatovskiy, C.-H. Lee and J. L. Sessler, Chem. Soc. Rev., 2013, 42, 1921–1933.
229. X. Duan, G.-Q. Zhang, S. Ji, Y. Zhang, J. Li, H. Ou, Z. Gao, G. Feng and D. Ding, Angew. Chem., Int. Ed., 2022, 61, e202116174.
230. Y. Liu, L. Teng, X.-F. Lou, X.-B. Zhang and G. Song, J. Am. Chem. Soc., 2023, 145, 5134–5144.
231. S. Liao, Y. Wang, Z. Li, Y. Zhang, X. Yin, S. Huan, X.-B. Zhang, S. Liu and G. Song, Theranostics, 2022, 12, 6883–6897.
232. C. Lu, C. Zhang, P. Wang, Y. Zhao, Y. Yang, Y. Wang, H. Yuan, S. Qu, X. Zhang and G. Song, Chem, 2020, 6, 2314–2334.
233. Z. Guo, S. Park, J. Yoon and I. Shin, Chem. Soc. Rev., 2014, 43, 16–29.
234. J. Wang, Q. Ma, X.-X. Hu, H. Liu, W. Zheng, X. Chen, Q. Yuan and W. Tan, ACS Nano, 2017, 11, 8010–8017.
235. J. Li, X. Liu, Y. Zheng, Z. Cui, H. Jiang, Z. Li, S. Zhu and S. Wu, Adv. Mater., 2023, 35, 2210296.
236. M. Zangoli, A. Cantelli, A. Candini, A. Lewinska, F. Fardella, A. Tino, G. Tommasini, M. Wnuk, M. Moschetta, S. Perotto, M. Lucarini, C. Tortiglione, G. Lanzani and F. Di Maria, J. Phys. Chem. C, 2023, 127, 4672–4683.
237. K.-X. Teng, L.-Y. Niu and Q.-Z. Yang, J. Am. Chem. Soc., 2023, 145, 4081–4087.
238. Y. Wang, H. Gao, J. Yang, M. Fang, D. Ding, B. Z. Tang and Z. Li, Adv. Mater., 2021, 33, 2007811.
239. A. Ligthart, X. de Vries, L. Zhang, M. C. W. M. Pols, P. A. Bobbert, H. van Eersel and R. Coehoorn, Adv. Funct. Mater., 2018, 28, 1804618.
240. R. Coehoorn, P. A. Bobbert and H. van Eersel, Phys. Rev. B: Condens. Matter Mater. Phys., 2017, 96, 184203.
241. X. Zhen, Y. Tao, Z. An, P. Chen, C. Xu, R. Chen, W. Huang and K. Pu, Adv. Mater., 2017, 29, 1606665.
242. F. Xiao, H. Gao, Y. Lei, W. Dai, M. Liu, X. Zheng, Z. Cai, X. Huang, H. Wu and D. Ding, Nat. Commun., 2022, 13, 186.
243. Q. Dang, Y. Jiang, J. Wang, J. Wang, Q. Zhang, M. Zhang, S. Luo, Y. Xie, K. Pu, Q. Li and Z. Li, Adv. Mater., 2020, 32, 2006752.
244. Y. Fan, S. Liu, M. Wu, L. Xiao, Y. Fan, M. Han, K. Chang, Y. Zhang, X. Zhen, Q. Li and Z. Li, Adv. Mater., 2022, 34, 2201280.
245. J. Li, Y. Liu, Y. Xu, L. Li, Y. Sun and W. Huang, Coord. Chem. Rev., 2020, 415, 213318.
246. Z. Lei and F. Zhang, Angew. Chem., Int. Ed., 2021, 60, 16294–16308.
247. J. Mu, M. Xiao, Y. Shi, X. Geng, H. Li, Y. Yin and X. Chen, Angew. Chem., Int. Ed., 2022, 61, e202114722.
248. L. Wu, Y. Ishigaki, Y. Hu, K. Sugimoto, W. Zeng, T. Harimoto, Y. Sun, J. He, T. Suzuki, X. Jiang, H.-Y. Chen and D. Ye, Nat. Commun., 2020, 11, 446.
249. Y. Liu, L. Teng, Y. Lyu, G. Song, X.-B. Zhang and W. Tan, Nat. Commun., 2022, 13, 2216.
250. C. Xie, X. Zhen, Q. Miao, Y. Lyu and K. Pu, Adv. Mater., 2018, 30, 1801331.
251. X. Zhen, C. Xie and K. Pu, Angew. Chem., Int. Ed., 2018, 57, 3938–3942.
252. R. Qu, D. He, M. Wu, H. Li, S. Liu, J. Jiang, X. Wang, R. Li, S. Wang, X. Jiang and X. Zhen, Nano Lett., 2023, 23, 4216–4225.
253. Y. Lyu, D. Cui, J. Huang, W. Fan, Y. Miao and K. Pu, Angew. Chem., Int. Ed., 2019, 58, 4983–4987.
254. L. A. MacManus-Spencer, D. E. Latch, K. M. Kroncke and K. McNeill, Anal. Chem., 2005, 77, 1200–1205.
255. Z. Gao, S. Jia, H. Ou, Y. Hong, K. Shan, X. Kong, Z. Wang, G. Feng and D. Ding, Angew. Chem., Int. Ed., 2022, 61, e202209793.
256. W. Chen, Y. Zhang, Q. Li, Y. Jiang, H. Zhou, Y. Liu, Q. Miao and M. Gao, J. Am. Chem. Soc., 2022, 144, 6719–6726.
257. L. Lei, F. Yang, X. Meng, L. Xu, P. Liang, Y. Ma, Z. Dong, Y. Wang, X.-B. Zhang and G. Song, J. Am. Chem. Soc., 2023, 145, 24386–24400.
258. K. Y. Zhang, Q. Yu, H. Wei, S. Liu, Q. Zhao and W. Huang, Chem. Rev., 2018, 118, 1770–1839.
259. R. Gao, M. S. Kodaimati and D. Yan, Chem. Soc. Rev., 2021, 50, 5564–5589.
260. M. Yang, J. Huang, J. Fan, J. Du, K. Pu and X. Peng, Chem. Soc. Rev., 2020, 49, 6800–6815.
261. Z. Wu, A. C. Midgley, D. Kong and D. Ding, Mater. Today Bio, 2022, 17, 100481.
262. S. Wang, B. Li and F. Zhang, ACS Cent. Sci., 2020, 6, 1302–1316.
263. V. V. Tuchin, J. Biomed. Photonics Eng., 2015, 1, 98–134.
264. J. A. Bearden, Rev. Mod. Phys., 1967, 39, 78–124.
265. H. Chen, M. M. Rogalski and J. N. Anker, Phys. Chem. Chem. Phys., 2012, 14, 13469–13486.
266. L. Song, X.-H. Lin, X.-R. Song, S. Chen, X.-F. Chen, J. Li and H.-H. Yang, Nanoscale, 2017, 9, 2718–2722.
267. Z. Yang, Q. Xu, Q. Chen and H. Yang, Sci. China: Chem., 2024, 67, 1056–1059.
268. W. C. Röntgen, Science, 1896, 3, 227–231.
269. X. Chen, J. Song, X. Chen and H. Yang, Chem. Soc. Rev., 2019, 48, 3073–3101.
270. H. Wu, Y. Ge, G. Niu and J. Tang, Matter, 2021, 4, 144–163.
271. Q. Chen, J. Wu, X. Ou, B. Huang, J. Almutlaq, A. A. Zhumekenov, X. Guan, S. Han, L. Liang, Z. Yi, J. Li, X. Xie, Y. Wang, Y. Li, D. Fan, D. B. L. Teh, A. H. All, O. F. Mohammed, O. M. Bakr, T. Wu, M. Bettinelli, H. Yang, W. Huang and X. Liu, Nature, 2018, 561, 88–93.
272. G. Gopakumar, I. Unger, P. Slavíček, U. Hergenhahn, G. Öhrwall, S. Malerz, D. Céolin, F. Trinter, B. Winter, I. Wilkinson, C. Caleman, E. Muchová and O. Björneholm, Nat. Chem., 2023, 15, 1408–1414.
273. C. Roques-Carmes, N. Rivera, A. Ghorashi, S. E. Kooi, Y. Yang, Z. Lin, J. Beroz, A. Massuda, J. Sloan, N. Romeo, Y. Yu, J. D. Joannopoulos, I. Kaminer, S. G. Johnson and M. Soljačić, Science, 2022, 375, eabm9293.
274. P. Büchele, M. Richter, S. F. Tedde, G. J. Matt, G. N. Ankah, R. Fischer, M. Biele, W. Metzger, S. Lilliu, O. Bikondoa, J. E. Macdonald, C. J. Brabec, T. Kraus, U. Lemmer and O. Schmidt, Nat. Photonics, 2015, 9, 843–848.
275. A. Jana, S. Cho, S. A. Patil, A. Meena, Y. Jo, V. G. Sree, Y. Park, H. Kim, H. Im and R. A. Taylor, Mater. Today, 2022, 55, 110–136.
276. M. Chen, L. Sun, X. Ou, H. Yang, X. Liu, H. Dong, W. Hu and X. Duan, Adv. Mater., 2021, 33, 2104749.
277. H. Chen, M. Lin, Y. Zhu, D. Zhang, J. Chen, Q. Wei, S. Yuan, Y. Liao, F. Chen, Y. Chen, M. Lin and X. Fang, Small, 2024, 20, 2307277.
278. X. Ou, X. Qin, B. Huang, J. Zan, Q. Wu, Z. Hong, L. Xie, H. Bian, Z. Yi, X. Chen, Y. Wu, X. Song, J. Li, Q. Chen, H. Yang and X. Liu, Nature, 2021, 590, 410–415.
279. J.-X. Wang, L. Gutiérrez-Arzaluz, X. Wang, T. He, Y. Zhang, M. Eddaoudi, O. M. Bakr and O. F. Mohammed, Nat. Photonics, 2022, 16, 869–875.
280. J. J. Rehr and R. C. Albers, Rev. Mod. Phys., 2000, 72, 621.
281. P. Pei, Y. Chen, C. Sun, Y. Fan, Y. Yang, X. Liu, L. Lu, M. Zhao, H. Zhang, D. Zhao, X. Liu and F. Zhang, Nat. Nanotechnol., 2021, 16, 1011–1018.
282. H. Suo, X. Zhang and F. Wang, Trends Chem., 2022, 4, 726–738.
283. X. Wang, H. Shi, H. Ma, W. Ye, L. Song, J. Zan, X. Yao, X. Ou, G. Yang, Z. Zhao, M. Singh, C. Lin, H. Wang, W. Jia, Q. Wang, J. Zhi, C. Dong, X. Jiang, Y. Tang, X. Xie, Y. Yang, J. Wang, Q. Chen, Y. Wang, H. Yang, G. Zhang, Z. An, X. Liu and W. Huang, Nat. Photonics, 2021, 15, 187–192.
284. M. Dong, A. Lv, X. Zou, N. Gan, C. Peng, M. Ding, X. Wang, Z. Zhou, H. Chen, H. Ma, L. Gu, Z. An and W. Huang, Adv. Mater., 2024, 36, 2310663.
285. W. Ma, Y. Su, Q. Zhang, C. Deng, L. Pasquali, W. Zhu, Y. Tian, P. Ran, Z. Chen, G. Yang, G. Liang, T. Liu, H. Zhu, P. Huang, H. Zhong, K. Wang, S. Peng, J. Xia, H. Liu, X. Liu and Y. M. Yang, Nat. Mater., 2022, 21, 210–216.
286. N. Gan, X. Zou, M. Dong, Y. Wang, X. Wang, A. Lv, Z. Song, Y. Zhang, W. Gong, Z. Zhao, Z. Wang, Z. Zhou, H. Ma, X. Liu, Q. Chen, H. Shi, H. Yang, L. Gu, Z. An and W. Huang, Nat. Commun., 2022, 13, 3995.
287. K. Lu, C. He, N. Guo, C. Chan, K. Ni, G. Lan, H. Tang, C. Pelizzari, Y.-X. Fu, M. T. Spiotto, R. R. Weichselbaum and W. Lin, Nat. Biomed. Eng., 2018, 2, 600–610.
288. L. Zheng, R. Zhu, L. Chen, Q. Fu, J. Li, C. Chen, J. Song and H. Yang, Nano Res., 2021, 14, 3744–3755.
289. X. Wang, W. Sun, H. Shi, H. Ma, G. Niu, Y. Li, J. Zhi, X. Yao, Z. Song, L. Chen, S. Li, G. Yang, Z. Zhou, Y. He, S. Qu, M. Wu, Z. Zhao, C. Yin, C. Lin, J. Gao, Q. Li, X. Zhen, L. Li, X. Chen, X. Liu, Z. An, H. Chen and W. Huang, Nat. Commun., 2022, 13, 5091.
290. J. Huang, L. Su, C. Xu, X. Ge, R. Zhang, J. Song and K. Pu, Nat. Mater., 2023, 22, 1421–1429.
291. M. I. Balonov and P. C. Shrimpton, Ann. ICRP, 2012, 41, 129–141.
292. W. A. Kalender, Phys. Med. Biol., 2014, 59, R129.
293. P. S. Henshaw, Radiology, 1945, 44, 569–580.
294. H. B. Stone, C. N. Coleman, M. S. Anscher and W. H. McBride, Lancet Oncol., 2003, 4, 529–536.
295. A. L. Klibanov and J. A. Hossack, Invest. Radiol., 2015, 50, 657.
296. J. Ouyang, Z. Tang, N. Farokhzad, N. Kong, N. Y. Kim, C. Feng, S. Blake, Y. Xiao, C. Liu, T. Xie and W. Tao, Nano Today, 2020, 35, 100949.
297. W. Wang, X. Wu, K. W. Kevin Tang, I. Pyatnitskiy, R. Taniguchi, P. Lin, R. Zhou, S. L. C. Capocyan, G. Hong and H. Wang, J. Am. Chem. Soc., 2023, 145, 1097–1107.
298. X. Wang, M. Wu, H. Li, J. Jiang, S. Zhou, W. Chen, C. Xie, X. Zhen and X. Jiang, Adv. Sci., 2022, 9, 2104125.
299. J. Jiang, M. Zhang, T. Lyu, L. Chen, M. Wu, R. Li, H. Li, X. Wang, X. Jiang and X. Zhen, Adv. Mater., 2023, 35, 2300854.
300. K. J. Opielinski and T. Gudra, Ultrasonics, 2006, 44, e295–e302.
301. J. Wu and W. L. Nyborg, Adv. Drug Delivery Rev., 2008, 60, 1103–1116.
302. D. J. Flannigan and K. S. Suslick, Nature, 2005, 434, 52–55.
303. M. P. Brenner, S. Hilgenfeldt and D. Lohse, Rev. Mod. Phys., 2002, 74, 425–484.
304. W. Wang, A. Tasset, I. Pyatnitskiy, H. G. Mohamed, R. Taniguchi, R. Zhou, M. Rana, P. Lin, S. L. C. Capocyan, A. Bellamkonda, W. Chase Sanders and H. Wang, Adv. Drug Delivery Rev., 2022, 186, 114343.
305. R. Canaparo, F. Foglietta, N. Barbero and L. Serpe, Adv. Drug Delivery Rev., 2022, 189, 114495.
306. Q. Zhou, K. H. Lam, H. Zheng, W. Qiu and K. K. Shung, Prog. Mater. Sci., 2014, 66, 87–111.
307. A. Cafarelli, A. Marino, L. Vannozzi, J. Puigmartí-Luis, S. Pané, G. Ciofani and L. Ricotti, ACS Nano, 2021, 15, 11066–11086.
308. Y. Wang, Y. Xu, S. Dong, P. Wang, W. Chen, Z. Lu, D. Ye, B. Pan, D. Wu, C. D. Vecitis and G. Gao, Nat. Commun., 2021, 12, 3508.
309. G. Hong, Science, 2020, 369, 638.
310. S. Jiang, X. Wu, F. Yang, N. J. Rommelfanger and G. Hong, Nat. Protoc., 2023, 18, 1–34.
311. F. Yang, X. Wu, H. Cui, Z. Ou, S. Jiang, S. Cai, Q. Zhou, B. G. Wong, H. Huang and G. Hong, Sci. Adv., 2022, 8, eabo6743.
312. C. Xu, J. Huang, Y. Jiang, S. He, C. Zhang and K. Pu, Nat. Biomed. Eng., 2023, 7, 298–312.
313. Y. Wang, Z. Yi, J. Guo, S. Liao, Z. Li, S. Xu, B. Yin, Y. Liu, Y. Feng, Q. Rong, X. Liu, G. Song, X.-B. Zhang and W. Tan, Nat. Photonics, 2024, 18, 334–343.
314. X. Wu, X. Zhu, P. Chong, J. Liu, L. N. Andre, K. S. Ong, K. Brinson, A. I. Mahdi, J. Li, L. E. Fenno, H. Wang and G. Hong, Proc. Natl. Acad. Sci. U. S. A., 2019, 116, 26332–26342.
315. C. Xu, S. He, X. Wei, J. Huang, M. Xu and K. Pu, Adv. Mater., 2023, 35, 2211651.
316. R. Wu, Z. Yao, Z. Chen, X. Ge, L. Su, S. Wang, Y. Wu and J. Song, Anal. Chem., 2023, 95, 11219–11226.
317. C. Wang, Y. Liu, X. Feng, C. Zhou, Y. Liu, X. Yu and G. Zhao, Angew. Chem., Int. Ed., 2019, 58, 11642–11646.
318. P. Li, Y. Guo, Y. Jia, H. Guan, C. Wang, Z. Wu, S. Sun, Z. Qu, P. Zhou and G. Zhao, Mater. Chem. Front., 2021, 5, 7170–7175.
319. Z. Wu, M. Wang, Y. Guo, F. Ji, C. Wang, S. Wang, J. Zhang, Y. Wang, S. Zhang and B. Jin, J. Phys. Chem. B, 2021, 125, 9572–9578.

---