Through-space interactions of optoelectronic materials

Abstract

In molecular science, theories based on covalent through-bond conjugation (TBC) serve as the foundational framework for designing efficient organic functional materials (OFMs). However, while TBC has been extensively established, through-space interaction (TSI) has recently emerged as an equally crucial electronic interaction governing the properties of OFMs, particularly in systems with partially or fully non-conjugated architectures. Nevertheless, the absence of quantitative structure–property relationships and a systematic summary continues to hinder the development of general design strategies for non-conjugated OFMs with tailored optoelectronic characteristics. Herein, this review presents a comprehensive overview of TSI in optoelectronic materials, beginning with its history, development, and current perspective. From the perspectives of luminescence and electronic properties, the working mechanisms, properties, manipulation strategies, and advanced applications of TSI are comprehensively summarized with typical examples, mainly including clusteroluminescence, thermally activated delayed fluorescence, room-temperature phosphorescence, and charge transport. Based on the current achievements and challenges, perspectives for the future development of TSI and related optoelectronic materials are also discussed. This review will facilitate the rational design of TSI-based optoelectronic materials and advance new photophysical theories as a supplement to the well-established TBC-based theories for next-generation functional materials.

---

Jianyu Zhang

Prof. Dr Jianyu Zhang is a tenure-track professor at Zhejiang University. He obtained his PhD degree from HKUST with Prof. Ben Zhong Tang. Then, he was a postdoc at HKUST and a research fellow with Prof. Ben L. Feringa at RUG under the Marie Skłodowska-Curie Actions. His interests are organic luminescent materials, through-space conjugation, and computational chemistry. He has published more than 120 papers (50 as first or corresponding author) in Nat. Photon., Chem, Nat. Commun., J. Am. Chem. Soc., Angew. Chem. etc. He has received many awards, including the ACS PMES Future Faculty Award, ACS PHYS Young Investigator Award, and CAS Future Leaders.

Zuping Xiong

Dr Zuping Xiong is a postdoctoral researcher at HKUST in Hong Kong. He received his PhD degree from Zhejiang University. His research spans through-space interactions, photoswitching, supramolecular chemistry, etc. He has led several important research projects and published 40 papers (18 first/co-first authored), such as Nat. Rev. Chem., J. Am. Chem. Soc., Chem, Angew. Chem. etc. He is a three-time recipient of the China National Scholarship and has been selected as the 2025 Lindau Nobel Laureate Meeting Young Scientist, the 2025 CAS Future Leader, the 2024 RSC excellent student, among other distinctions.

Jacky W. Y. Lam

Prof. Dr Jacky W. Y. Lam received his PhD from HKUST in 2003 under the supervision of Prof. Ben Zhong Tang. He currently serves as a Research Professor in the Department of Chemistry at HKUST. His research focuses on the design and synthesis of novel polymers and AIE molecules for high-tech applications. He has published >800 papers, with total citations exceeding 106 000 and an H-index of 154. He has authored over 30 book chapters and obtained more than 30 authorized patents. Since 2014, he has been recognized as a “Highly Cited Scientist” by Clarivate Analytics in the area of Chemistry.

Jing Zhi Sun

Prof. Dr Jing Zhi Sun has been a professor at Zhejiang University since 2006. His research interests include design and synthesis of functional polyacetylenes, mechanism study and application exploration of aggregation-induced emission, as well as electronic processes in organic semiconductors. He received the Excellence in Teaching Award of Zhejiang University, the Outstanding Teaching Contribution Award of the Faculty of Engineering in Zhejiang University, and was nominated for the Yongping Teaching Award. He has published >230 papers with >9000 citations and holds 16 granted Chinese invention patents. He received the National Natural Science Award of China (1st Class) in 2017.

Haoke Zhang

Prof. Dr Haoke Zhang is a tenure-track assistant professor at Zhejiang University. His research interests are focused on photophysical chemistry, polymer physics and chemistry, and computational chemistry. He received his MSc and PhD degrees from ZJU and HKUST under the supervision of Professor Ben Zhong Tang. He has published more than 160 papers in Nat. Photon., Nat. Rev. Chem., Nat. Commun., Chem., J. Am. Chem. Soc., Angew. Chem. and Adv. Mater., etc., with >10 000 citations. He has been awarded the “Young Scientist Award” by the European Materials Research Society and “Emerging Investigator” by Poly. Chem., Mater. Chem. Front., Aggregate, and Luminescence.

Ben Zhong Tang

Prof. Dr Ben Zhong Tang is the X.Q. Deng Presidential Chair Professor and Dean of the School of Science and Engineering at CUHK, Shenzhen. He has published >2000 papers with >240 000 citations. He has delivered >500 invited talks at international conferences and granted >300 patents. He is currently Editor-in-Chief of Aggregate. He is mainly engaged in materials science, macromolecular chemistry, and biomedical theranostics. He coined the concept of aggregation-induced emission (AIE), and his labs are spearheading global AIE research. He has been listed as a Highly Cited Researcher in both areas of Chemistry and Materials Science since 2014.

---

1. Introduction

Materials science has been a cornerstone in the advancement of human society, driving transformative changes across modern technologies.^1–4 Over the past century, the emergence of novel materials has continually revolutionized industries and daily life. Taking advantage of their lightweight nature, various categories, and scalable processability, organic functional materials (OFMs) have attracted widespread attention from both academia and industry.^5–9 Among different types of OFMs, optoelectronic materials occupy a critically important position due to their broad application fields, including energy, information communication and display, illumination, electronics, biomedicine, and environmental technologies.^10–13 Besides, they serve as the physical foundations for many advanced technologies, such as aerospace, energy conversion, smart construction, and the low-carbon industry.^14–19 Since the initial discovery and subsequent investigation of the photoelectric effect in the late 19th to early 20th century,²⁰ scientists have never stopped exploring optoelectronic functional materials. As a result, many fundamental theories of optoelectronics and effective regulating strategies for organic optoelectronic materials have been established, which have guided their development for more than 100 years and continue to drive innovation today.

From the perspective of molecular science, electronic structure plays a decisive role in governing the physicochemical properties of materials. Electronic delocalization, in particular, is considered essential for achieving excellent performance in organic optoelectronic systems.^21–23 Theoretical frameworks based on through-bond conjugation (TBC) have long been considered the cornerstone for designing these materials and manipulating their optoelectronic properties, as they emphasize the interactions and delocalization of electrons within covalently conjugated structures (i.e., linked π-systems).^24–29 Over the past several decades, TBC-based theories have been extensively validated by countless examples and have led to the development of numerous high-performance optoelectronic materials.³⁰ For instance, extended π-electronic systems are known to endow molecules with versatile properties, such as high luminescence efficiency, redshifted absorption and emission wavelengths, stable organic radicals, and high charge-carrier mobility.^31–36 Unsurprisingly, the majority of modern optoelectronic materials are designed based on TBC-based principles, typically employing polycyclic aromatic hydrocarbons and conjugated polyene interconnected via covalent bonds. On the other hand, noncovalent interactions have received comparatively less attention in the context of optoelectronic material design, primarily due to their weaker interaction strength and poor stability relative to covalent bonds.^37,38 Although noncovalent interactions are indispensable in supramolecular systems, molecular recognition, and self-assembly, their limited ability to significantly alter electronic structures has restricted their direct and wide utility in optoelectronic property modulation.^39–43 Consequently, they are rarely employed as primary design strategies for tuning optoelectronic performance, despite their importance in other areas of materials science.

The conventional wisdom that equates efficient optoelectronic properties with extensive covalent conjugation has, however, been challenged by recent discoveries. Many non-conjugated small molecules, supramolecules, and macromolecules without extended π systems or even aromatic rings have garnered significant research interest recently due to their simple chemical structures, abundant sources (such as natural compounds or simple synthetic routes), and excellent optoelectronic properties rivaling those of conjugated systems. Critically, their properties cannot be adequately explained by traditional TBC-based theories, particularly in the absence of extended conjugation or even aromatic segments. Breakthrough studies have revealed that noncovalent through-space interaction (TSI) of electrons, referring to electronic couplings and orbital overlap through spatial proximity, may play an equally or even more critical role than TBC in determining their optoelectronic properties.^44–46 Remarkably, such through-space electron delocalization can enable non-conjugated luminogens to achieve efficient visible-light and near-infrared (NIR) luminescence comparable to conventional conjugated systems.^47–49 Similarly, through-space σ–σ interactions between neighboring non-conjugated cyclohexanethiols can facilitate charge transport comparable to π–π interactions.⁵⁰ Furthermore, intramolecular O⋯S interactions acting as noncovalent conformational locks can induce planar π-conjugated-like structures, thereby enhancing optoelectronic performance in organic semiconductors.^51–53 In addition, apart from π-ring-based aromatic systems, through-space σ-ring is also experimentally proved to modulate aromaticity and achieve double aromaticity within a single molecule.⁵⁴

Although still in its infancy, these discoveries collectively underscore the fundamental importance and far-reaching scientific value of noncovalent TSI in governing the luminescence and electronic properties of optoelectronic materials, paving the way for advanced applications. Several pioneering studies have further confirmed that TSI-driven electron delocalization and spatial orbital overlap in non-conjugated molecules can mimic the behaviors of electron conjugation within conventional TBC-based compounds.^55–58 Despite these promising advances, the field faces a major challenge: the absence of a quantitative structure–property relationship and a unified theoretical framework for TSI. This knowledge gap results in a fundamental disconnect between non-conjugated molecular structures and optoelectronic properties, hindering precise control over the strength and nature of TSI and stalling the formulation of general design rules for tailoring non-conjugated OFMs with predictable and controllable properties.

Herein, this review aims to provide a fundamental introduction and systematic summary of electronic TSI in the context of organic optoelectronic materials (Fig. 1). First, the history and development of TSI, from its early roles in NMR and chemical reactions to its emerging significance in optoelectronics, are traced. In addition, some typical examples, working mechanisms, manipulation strategies, and advanced applications of TSI for four types of optoelectronic properties, including clusteroluminescence (CL), thermally activated delayed fluorescence (TADF), room-temperature phosphorescence (RTP), and charge transport, are systematically summarized. These discussions encompass diverse material systems, including small molecules, supramolecular assemblies, and polymers with different electronic features. Finally, a comprehensive summary of current achievements, challenges, and potential directions for the future development of TSI and related optoelectronic materials is discussed. More importantly, beyond summarizing existing knowledge, this review hopes to establish clear structure–property relationships for TSI, offer practical design principles for engineering non-conjugated molecules, and propose a unified photophysical framework that is complementary to conventional TBC theories. We envision that this work will not only facilitate the rational design of TSI-based optoelectronic materials but also inspire new theoretical paradigms for next-generation functional materials with transformative applications.

Fig. 1 Outlines of optoelectronic materials based on through-space interactions discussed in this review.

2. History and development of through-space interactions

2.1. History

The recent prominence of TSI in governing noncovalent electronic interactions within organic optoelectronic materials belies a longer history of the concept, which first emerged in the field of nuclear magnetic resonance (NMR) technique during the 1950s.⁵⁹ Initially, TSI described the phenomenon of through-space nuclear spin–spin coupling, particularly observed in small organic and organometallic compounds.^60–63 Saika and Gutowski first observed unusual multiplet signals in the NMR spectrum of a fluorine-containing compound. They attributed this to the spin polarization of nonbonded electron pairs of two spatially proximate fluorine atoms.⁶⁴ This concept was later extended to complex biological molecules of protein as demonstrated by Feeney and co-workers.⁶⁵ This kind of nonbonded nuclear spin–spin coupling was also presented by a simple but meaningful perturbation model proposed by Mallory and coworkers (Fig. 2a), identifying its origin from overlap interactions between lone-pair orbitals.^66,67 In their model compounds, the constrained molecular geometry enforces near-parallel alignment and short distances (e.g., F⋯F and F⋯N, respectively) between specific bonds (e.g., C–F/C–F and C–F/C=N), facilitating significant internuclear overlap of the respective lone-pair orbitals. Although it does not lead to net bonding stabilization due to the occupied bonding and antibonding orbitals (Fig. 2b), this kind of through-space orbital overlap indeed provides an adequate pathway for transmitting spin information between the coupled nuclei, yielding measurable nuclear spin–spin coupling constants.

Fig. 2 Typical examples of through-space nuclear spin–spin interactions in NMR spectroscopy. (a) The orbital diagram, corresponding spin–spin coupling constants, and (b) energy-level diagram of nonbonded fluorine–fluorine interaction and fluorine–nitrogen interaction generated by the overlap of long-pair orbitals. Reproduced with permission from ref. 60. Copyright 2014, American Chemical Society. (c) Visualization of the nuclear spin–spin interaction between two phosphorus atoms via coupling energy density. Isosurfaces are shown with blue and red colors corresponding to the positive and negative function values, respectively. Reproduced with permission from ref. 71. Copyright 2003, John Wiley & Sons, Inc.

Later, many scientists proposed several theoretical considerations to complement and refine the above simple model, as well as offer the visualization approaches of such coupling pathways.^68–70 For instance, Malkina and Malkin introduced a model based on real-space functions in three-dimensional space, which regarded indirect spin–spin coupling as the energy splitting between two states with parallel and antiparallel nuclear spins.⁷¹ In this model, these energies were expressed as an integral over an “energy density” called coupling energy density (CED), which was equal to the reduced coupling constant. As shown in Fig. 2c, for cis-diphosphinoethylene, the CED analysis indicated that the TSI between the two phosphorus nuclei dominates over the through-bond pathway, as evidenced by the dominating positive CED (blue region) located at the phosphorus atoms, as well as the across-space strong negative CED (red region) contribution between them. Similarly, Soncini and Lazzeretti developed analysis methods based on the Fermi contact coupling density and the Fermi hole density maps, which suggested that the overlap between other types of nonbonding electron pairs could also promote an efficient coupling pathway for transmission of the spin coupling.^72,73

While TSI between heavy atoms (e.g., F and P) with high electron density is more readily observed, the detection and interpretation of through-space spin–spin coupling between hydrogen atoms presents a greater challenge.⁷⁴ A significant example was reported in 2018 by Dračínský and Malkina et al., who identified the through-space coupling between two specific hydrogen atoms of pyrene-fused oxa[7]helicene (Fig. 3).^75,76 Using long-range COSY NMR experiments combined with a two-dimensional J-resolved method, they confirm the existence of indirect through-space coupling. Apart from the strongest coupling between A9 and A11 (as shown in Fig. 3a), weaker interactions between the tert-butyl hydrogen atoms and the hydrogen atoms B9 and B11 are also observed, representing the hitherto-unprecedented observation of hydrogen–hydrogen interactions formally separated by 18 covalent bonds (Fig. 3b). With the help of the coupling deformation density, these two through-space coupling pathways could be visualized (Fig. 3c and d). Similar experimental and computational results of through-space J coupling of hydrogen atoms were also reported by Bouř et al., which highlighted the strong dependence between distance/conformation of molecules and such coupling.⁷⁷ Although these NMR-based TSI do not directly dictate optoelectronic properties, they are fundamentally linked to electronic structures and through-space orbital overlap, providing a sensitive spectroscopic signature for interactions that typically manifest bonded interactions (although the magnitude is much weaker).

Fig. 3 Through-space nuclear spin–spin interactions of hydrogen atoms. (a) Chemical structure of the pyrene-fused oxa[7]helicene. (b) The long-range COSY spectrum of the pyrene-fused oxa[7]helicene in CDCl₃. (c) and (d) The visualization of coupling pathways for (c) J(H_t-Bu–H_B9) and (d) J(H_t-Bu–H_B11), respectively. Reproduced with permission from ref. 76. Copyright 2018, The Royal Society of Chemistry.

Beyond NMR spectroscopy, the conceptual framework of TSI has been proven crucial for understanding orbital energy levels and chemical reactivity.⁷⁸ A landmark contribution came in 1971 from Roald Hoffmann (theoretical chemist who won the 1981 Nobel Prize in Chemistry), who presented a detailed introduction of orbital interactions through space and through bonds.⁷⁹ Hoffmann illustrated that when two functional groups approach closely enough for direct spatial overlap of orbitals, the interaction could be simply analyzed via the energy levels with the help of perturbation theory. For example, norbornadiene with two mirror planes of symmetry shows a significant overlap of two π orbitals beneath the molecular skeleton (Fig. 4).⁸⁰ As a result, the symmetric combination (SS) of orbitals is stabilized at lower energy, while the antisymmetric combination (SA) is placed at higher energy, with a splitting between SS and SA of 0.85 eV. One consequence of this TSI is the formation of quadricyclane, a common photochemical reaction of norbornadienes, which is promoted after photoexcitation.^81–83 Similar orbital stabilization via TSI was identified in other systems, such as the 7-norbornenyl cation and various.⁸⁴ At the same time, Hoffmann also pointed out the role of indirect through-bond coupling of orbitals, which involves σ bonds and may operate over surprisingly long distances. This theoretical perspective of TSI is of great significance to understanding the stability, reactivity, and stereoelectronic effects of many organic molecules.

Fig. 4 The energy level of orbitals of norbornadiene without (left) and with (right) TSI. The symmetric (S) or antisymmetric (A) features concerning the mirror planes are indicated. Reproduced with permission from ref. 79. Copyright 1971, American Chemical Society.

During the past several decades, the term TSI has been broadly associated with various noncovalent interactions, such as π–π stacking, hydrogen bonding, and others, and are widely applied in the areas of supramolecular chemistry, catalysis, and materials science. Many review articles have discussed these interactions within their specific research areas.^38,85–89 Different from the fundamental viewpoint of Roald Hoffmann, people pay less attention to the energy level of coupled orbitals but utilize these noncovalent interactions to analyze some properties of specific structures or materials. For example, the synergistic and competitive effects of hydrogen bonding and π–π stacking is critical for determining the three-dimensional structure and stability of DNA (Fig. 5), which in turn governs the physiological and biological functions of organisms.^90–92 Apart from stable-state properties, noncovalent π stacking also affects the transition-state binding in catalysis.^93–95 A representative study by Hunter and co-workers reported the alkylation of pyridine in a supramolecular zipper complex (Fig. 6), where a π-interaction forms between the pyridine ring (ring 3) and an adjacent aromatic ring 2. By modulating the electronic effect of ring 2 with electron-donating and withdrawing groups, the binding constant between the two rings in the transition state of alkylation is regulated, which further influences the energy barrier and reaction rate.⁹⁶

Fig. 5 The tridimensional structure of DNA with non-covalent hydrogen bonds and π–π stacking for its stability. Reproduced with permission from ref. 90. Copyright 2018, Elsevier.

Fig. 6 The conversion of pyridine to methyl pyridinium with the help of non-covalent interaction and the corresponding transition-state binding energies (ΔG_TS) of rings 2 and 3. Reproduced with permission from ref. 93. Copyright 2017, Springer Nature.

Although the history and term of TSI can be traced back to the last century, the abovementioned analysis of TSI mainly focuses on its influence on the ground-state or transition-state properties of molecules. Besides, since the strength of these TSI is much weaker compared to covalent bonds, they have been considered to be auxiliary effects to stabilize or enhance the intrinsic characteristics of materials during the past decades, instead of changing electronic structures or promoting new characteristics. On the other hand, for organic optoelectronic materials, the design paradigm has been overwhelmingly dominated by the TBC-based principles for decades. This TBC-centric approach is deeply rooted in its proven ability to delocalize electrons effectively along the molecular backbone, leading to desirable properties. Within this well-established framework, TSI is largely relegated to secondary roles. Their perceived role is to support and optimize the properties inherently endowed by the covalent conjugated structure, not to create them. Consequently, a significant gap emerges that the potential of TSI to act as a primary design element, capable of fundamentally altering electronic structures and generating novel optoelectronic phenomena in its own right, remains largely overlooked and underdeveloped.

2.2. Through-space conjugation and through-space charge transfer

During the past ten years, the publications and related citations of TSI-based luminescence/electronic phenomenon and materials have witnessed an accelerated development (Fig. 7a). With the rapidly increasing number of optoelectronic compounds without traditionally conjugated structures, attention has begun to change from TBC to TSI to better understand some unconventional photophysical phenomena from these materials. In terms of their electronic transition behaviors, TSI could currently be divided into two categories, including through-space conjugation (TSC) and through-space charge transfer (TSCT). Both of them are based on the through-space electronic communication without covalent bonds. Specifically, TSC arises primarily from the direct wavefunction overlap of spatially proximate orbitals, leading to the formation of a delocalized electronic system independent of traditional covalent bonds. Similar to the LE state, TSC often focuses on a limited area of the molecular skeleton without an obvious and long-range transition of electrons. Non-conjugated systems based on isolated aromantic rings or aliphatic polymers without typical electron-donating or electron-withdrawing units always show TSC. In parallel-oriented systems, TSC between π-orbitals (e.g., π-stacked aromatics or layered [n]cyclophanes) always results in characteristic spectral shifts such as bathochromically shifted absorption and emission, and an enhanced molar absorptivity compared to the isolated subunits.^97–100 It is noteworthy that, in some unparallel systems, TSC is always observed in the excited state when electrons are phase-matched, which promotes redshifted photoluminescence (PL) and excitation spectra with negligible influence on absorption spectra. In contrast, TSCT is an electronically asymmetric process involving the transfer of electron density from an electron-rich donor to an electron-deficient acceptor unit, which are spatially proximate but not necessarily conjugated through-bond.^101–103 Compared with the TBC-based system with partial charge transfer (CT), namely TBCT, TSCT without covalent bonds can realize total separation between the highest-occupied molecular orbital (HOMO) and the lowest-unoccupied molecular orbital (LUMO), and achieve complete charge transfer, creating a significant change in dipole moment upon photoexcitation.¹⁰⁴ Since the oscillator strength of the CT transition is much smaller than that of LE, the corresponding absorption is weak or negligible, resulting in a large Stokes shift.^105,106 Although TSCT is originally proposed in TADF materials with conjugated donor and acceptor fragments connected by saturated linkers or with large steric hindrance, it is also verified to play a critical role in modulating luminescence behaviors of non-conjugated donor- and acceptor-based systems. Therefore, in the following section, the intrinsic features, working mechanism, and modulating strategies of TSC and TSCT will be discussed in different optoelectronic systems with typical examples.

Fig. 7 (a) The number of publications and citations related to the optoelectronic phenomenon and materials and through-space interaction (TSI), according to Web of Science core collection. (b) The schematic diagram of through-bond conjugation (TBC), through-space conjugation (TSC), as well as through-space charge transfer (TSCT).

3. Through-space interactions for clusteroluminescence

Fluorescence is the light emission process that occurs when a molecule absorbs photons of high energy and subsequently relaxes to its ground state (S₀) from the lowest singlet state (S₁) by radiating photons of lower energy.¹⁰⁷ Conventional fluorescent luminogens typically possess planar and rigid molecular structures. According to the well-established photophysical theories based on TBC, achieving highly efficient luminescence in molecules appears to typically require extensive π/n-electron delocalization and associated CT mechanisms occurring via covalent bonding within molecular frameworks. Previous studies have demonstrated that these approaches represent well-established and successful methodologies for developing organic luminogens with fluorescence. By contrast, scientific interest has been relatively limited regarding non-conjugated molecular structures that appear poorly emissive, as well as bulk materials derived from such systems during the past decades.

In recent years, however, accumulating evidence points to a striking departure from conventional design rules that a variety of non-conjugated matter, ranging from natural proteins and polysaccharides to small and ostensibly “dark” molecules exemplified by bovine serum albumin, starch, poly(ethylene glycol) and xylitol, can exhibit pronounced blue luminescence in the solid state despite the absence of classical TBC or extended π-systems.¹⁰⁸ This phenomenon highlights the role of TSI and environment-enabled emissive pathways (Fig. 8a). Interestingly, the historical roots of emission from non-conjugated substances extend more than four centuries to Francis Bacon, who noted that solid sugar emits light when crushed, thereby providing an early documentation of mechanoluminescence in a system lacking conventional TBC (Fig. 8b).¹⁰⁹ Subsequent reports, including the long-wavelength emission of isotactic polystyrene by Lumry et al. in 1963 and the luminescence properties of poly(amidoamine) investigated by Tucker et al. in 2001, also described such unconventional luminescence from non-conjugated polymers.^110–113 Nevertheless, these early observations failed to garner significant interest largely due to their non-emissive nature in solution, weak emission intensity in the solid state, ambiguous underlying mechanisms, and contradictions with conventional optoelectronic theories. Moreover, many researchers initially attribute the observed luminescence to impurities rather than intrinsic material properties. Consequently, such unconventional emission in non-conjugated systems remained largely unexplored until recently. A turning point came in 2013 when Tang, Yuan, and colleagues systematically reinvestigated the unusual luminescence behavior of various non-conjugated polymers and small molecules.^114,115 They noted that while these materials were typically non-emissive as isolated molecules in dilute solutions, they exhibited pronounced visible-light emission (most within the blue region) upon aggregation. To rationalize this behavior, they introduced the concept of clustering-triggered emission (CTE) and termed the resulting emission clusteroluminescence (CL).¹¹⁶ Building on this discovery, Tang's group further investigated tetrahydropyrimidines in 2015 and proposed TSC as the key mechanism driving CL.^117,118 This breakthrough stimulated renewed interest and accelerated research into this once-overlooked class of luminescence phenomena.^119–127

Fig. 8 The phenomena of clusteroluminescence (CL) and the development of the CL study. (a) The emergence of CL from non-emissive (macro)molecules in dilute solutions to emissive clusters in the solid state. (b) The important milestones in the historical development of CL research. Reproduced with permission from ref. 128. Copyright 2025, Springer Nature.

Since luminogens with CL properties (CLgens) show a similar photophysical effect to luminogens with AIE effects (AIEgens), CL can be viewed as a typical manifestation within the broader AIE family. However, CLgens also show some special features that are different from traditional luminogens with the ACQ effect (ACQgens) and AIEgens (Fig. 9).^129,130 (1) Approach of electronic delocalization. CLgens rely solely on TSI, which is closely related to their non-conjugated structures. In contrast, ACQgens with planar and conjugated structures rely on TBC, and traditional AIEgens bearing multiple rotors and twisted conformations produce luminescence mainly through TBC and sometimes supplemented by TSI.^131–133 (2) Emitting species. As the TSI of non-conjugated fragments only forms in the clustered state, the emitting species of CLgens could be monomers (with intramolecular TSI) or multimers (with intermolecular TSI). In contrast, the emitting species of ACQgens and AIEgens are usually single molecules. (3) Emission upon aggregation. Due to the planar and conjugated conformation, ACQgens show intrinsic single-molecule emission in dilute solution but suffer from the quenching effect in the aggregate state. In contrast, AIEgens are weakly emissive in dilute solution, but show intensified emission upon aggregation due to the restriction of intramolecular motions (RIM).¹³⁴ Specifically, due to their dependence on clustering for emission, CLgens are totally non-emissive in the isolated state, but their emission could be activated upon forming aggregates with stabilized TSC. (4) Absorption and excitation spectra. TSC interactions are significantly weaker than covalent TBC, endowing CLgens with low oscillator strength and making them difficult to detect via absorption spectroscopy. Therefore, this results in mismatched absorption and excitation spectra for CLgens.¹³⁵ In contrast, ACQgens and AIEgens display matched absorption and excitation spectra, as is typical for traditional luminogens. (5) Emission spectra and dependence of excitation. Unlike TBC, TSC is highly sensitive to the size, stability, molecular conformation, and electronic coupling of clusters, which may result in multiple clusters within the same CLgen and correspondingly multiple emitting species.¹³⁶ According to Kasha's rule, radiative decay occurs from the lowest excited state of each species, endowing CLgens with excitation-dependent and multi-peaked CL, with emission wavelengths and quantum yields varying by excitation wavelength.^137–139 In comparison, TBC-based ACQgens and AIEgens usually exhibit excitation-independent and single-peak emission. (6) RTP. Many CLgens contain electron-rich heteroatoms (e.g., O, N, S, and P) and mixed compensation of n- and π-electrons, which promote efficient intersystem crossing (ISC) between singlet and triplet states and facilitate RTP in CLgens.^140–144 However, RTP is rare in pure hydrocarbon systems of ACQgens and AIEgens.¹⁴⁵ The above characteristics and working mechanism define CLgens as a distinct class of luminescent materials, attracting much attention from researchers in chemistry, photophysics, and materials science.

Fig. 9 (a) Typical features and comparison of luminogens showing aggregation-caused quenching (ACQgens), aggregation-induced emission (AIEgens), and clusteroluminescence (CLgens) effects; TBC = through-bond conjugation, TSI = through-space interaction. (b) The diagrammatic illustrations of photophysical characteristics of CLgens; PL = photoluminescence, FL = fluorescence, RTP = room-temperature phosphorescence. Reproduced with permission from ref. 128. Copyright 2025, Springer Nature.

In this section, we categorize CLgens into three distinct types based on their dominant electronic characteristics, including n-electron systems, π-electron systems, and π/n-electron hybrid systems (noting that mixed-electron systems also occur in practice, with multiple electronic interactions operating concurrently). For each category, the typical examples, working mechanisms, strategies for modulating their CL properties, and potential applications will be discussed. Additionally, a few radical-involved systems with non-conjugated structures are included in the discussion, as their luminescence originates from the TSI of excitons.

3.1. n-Electron systems

TSI has been established as a fundamental mechanism for CL. Electron-rich heteroatoms with highly directional orbitals can effectively form TSI between lone electron pairs, thus endowing some nonaromatic compounds with good CL properties.¹⁴⁶ Some nonaromatic small molecules and polymers bearing multiple heteroatoms without any π electrons are typical CLgens.¹⁴⁷ A representative oxygen-based example is poly(ethylene glycol) (PEG), which consists solely of repeating ether units (Fig. 10a).¹⁴⁸ It is widely recognized that the C and O atoms in PEG carry a partial positive and negative charge, respectively, creating strong electrostatic interactions between adjacent O and C atoms. Such O⋯C interactions form an interconnected network in the aggregate state that rigidifies the PEG backbone and facilitates the formation of stable through-space O⋯O interactions with an average intrachain O⋯O distance of approximately 2.9 Å, which is notably shorter than the sum of their van der Waals radii (3.04 Å).¹⁴⁹ Thus, the formed TSI between O atoms results in a bright blue CL of solid PEG with an absolute photoluminescence quantum yield (PLQY, Φ) of 8.1% and multiple emission peaks at 378 nm, 392 nm, and 420 nm (Fig. 10b).¹⁵⁰ Subsequent time-resolved measurement at such peaks exhibits varying fluorescent lifetimes of 3.88 ns, 3.51 ns, and 4.26 ns, respectively, indicating their fluorescent nature and excitation-dependent properties (Fig. 10c). Similarly, the commercially available PEO–PPO–PEO triblock copolymer of F127 shows blue CL and Φ of 5.8% in the aggregate state under 312 nm UV irradiation.¹⁴⁸ It also exhibits low-temperature phosphorescence lasting for over 10 s at 77 K. On the other hand, xylitol, as an example of natural small molecules with multiple oxygen atoms, exhibits a similar CTE effect.¹⁵¹ It is non-emissive in the dilute solution but gradually becomes emissive with the increased concentration. Its crystalline sample also shows excitation-dependent emission, although the Φ is much lower than those of PEG and F127.

Fig. 10 Clusteroluminescence of nonaromatic oxygen clusters. (a) The chemical structures of PEG, F127, and xylitol, and photographs of their solid powders, taken under 312 nm UV irradiation. Their absolute photoluminescence quantum yields (Φ) are indicated. (b) The photoluminescence spectra of solid-powder PEG under different excitation wavelengths. (c) The fluorescent lifetimes of solid-powder PEG under excitation wavelengths of 378 nm, 392 nm, and 420 nm, respectively. Reproduced with permission from ref. 148. Copyright 2018, John Wiley & Sons, Inc.

Beyond oxygen-based systems, nitrogen atoms bearing lone pairs can also engage in TSI and give rise to CL. Stiriba et al. reported unprecedented blue CL from polyethylenimines (PEI) and investigated the influence of polymer topology (Fig. 11).¹⁵² Hyperbranched PEI demonstrates weak blue fluorescence with a Φ of 0.01 and a short lifetime of 3.2 ns. In contrast, its linear counterpart shows more efficient emission characteristics, exhibiting a four-fold higher quantum yield (Φ = 0.04) and a significantly longer fluorescence lifetime of 5.7 ns. For polymers of comparable molecular weight, the superior CL efficiency of linear PEI over its hyperbranched analogue indicates that a hyperbranched architecture is not a prerequisite for efficient TSI and CL. This finding redirects the study of CTE in synthetic polymers from a focus on architectural variations to an exploration of specific chemical structures. Furthermore, the CL properties of PEIs are readily tunable through chemical methylation, oxidation, as well as changing the external pH environment.

Fig. 11 Chemical structures and clusteroluminescence of hyperbranched and linear polyethylenimines. Reproduced with permission from ref. 152. Copyright 2007, John Wiley & Sons, Inc.

Moving beyond polymers composed solely of nitrogen or oxygen atoms, numerous structures incorporating mixed heteroatoms have also been demonstrated to exhibit intrinsic CL.^153,154 With amide, amine, and alkyl groups, poly(amidoamine) (PAMAM) dendrimers were first reported by Tomalia et al. in 1985, which were later reported by several scientists to emit weakly intrinsic CL, primarily attributed to (n,π*) transition originating from their amide functionalities.^112,155,156 In addition, hyperbranched polysiloxane (HBPSi) is also a typical example, which has electron-rich nitrogen and oxygen atoms, as well as silicon atoms.^157–159 Yan et al. investigated the effect of chain length between electron-rich atoms on the CL of HBPSi via a facile synthetic method (Fig. 12a).¹⁶⁰ The resulting products all show blue CL in the concentrated solution or solid state under 360 nm UV irradiation, due to the spatial electronic delocalizations driven by the strong hydrogen bond and amphiphilicity. Excitation wavelength-dependent luminescence is also observed, like in other CL materials (Fig. 12b). Notably, a distinct redshift in emission was observed with decreasing chain length, resulting from the reduced distances between electron-rich moieties and the consequent enhancement of electronic TSC. The CL efficiency of these systems can be further optimized. For instance, by introducing vinyl groups into the HBPSi skeleton, the absolute Φ could be further increased up to 43.9% among the reported silica-containing hyperbranched polymers. This enhancement is likely due to the synergistic effect of (π,π*) and (n,π*) transitions from vinyl and carbonyl groups, respectively.¹⁶¹ With the intrinsic CL and good biocompatibility, some modified HBPSi have been utilized in biological applications, including cell imaging, controlled drug delivery, and reactive oxygen species scavenging.^162–166 Although many HBPSi systems have been reported, their optimized CL wavelengths remain confined to the blue spectral region.

Fig. 12 Clusteroluminescence of hyperbranched polysiloxanes (HBPSi). (a) The synthetic route to hyperbranched polysiloxanes with different alkane-chain lengths. (b) The photoluminescence spectra of HBPSi-5C in a water solution with a concentration of 90 mg mL⁻¹, which shows the excitation-dependent emission. Reproduced with permission from ref. 160. Copyright 2023, American Chemical Society.

Following the pure n-electron system, a series of oxygen and sulfur-based dendrimers were reported by Wang and Tang et al. in 2021.¹⁶⁷ Three different generations from G1 to G3 with crowded internal structures were synthesized (Fig. 13a). They all show the non-conjugated feature with a very short absorption wavelength of 200–250 nm but visible-light CL with emission peaks at 425–470 nm in aqueous solutions by 365 nm excitation, with a remarkably low detectable concentration of just 0.01 mg mL⁻¹. With the increased concentration, they display enhanced emission intensity with gradually decreased critical cluster concentration from G1 to G3. This inverse relationship indicates that higher-generation architectures form emitting clusters more readily, which is attributed to their more crowded and compact structures. In addition, they all show excitation-dependent emission, and the higher-generation dendrimers have a bathochromic shift of the optimal excitation due to the overall larger clusters (Fig. 13b). Owing to the non-conjugated skeleton and hydroxyl-terminated peripheries, these dendrimers show good biocompatibility and water solubility. Thus, G3 is selected as a representative example and further applied in cell imaging. Unlike conventional single-color bioprobes, the excitation-dependent CL of G3 enables multi-channel imaging within cells using different excitation wavelengths (405 nm, 488 nm, 543 nm, and 633 nm, respectively), demonstrating its utility as a versatile optical material for advanced bioimaging (Fig. 13c).^168,169

Fig. 13 Clusteroluminescence of oxygen and sulfur-based pure n-electron dendrimers. (a) Chemical structures of oxygen and sulfur-based dendrimers with different molecular sizes (G1, G2, and G3). (b) The excitation-dependent photoluminescence spectra of G3 in aqueous solution under different excitation wavelengths. (c) The confocal laser scanning microscopy images of 4T1 cells after 6 h of incubation with G3. The excitation wavelengths are 405, 488, 543, and 633 nm, respectively, and the image colors are not related to the observed colors. Reproduced with permission from ref. 167. Copyright 2021, Springer Nature.

While organic radicals are typically considered luminescence quenchers, certain radical species themselves can serve as emissive centers in carefully designed systems.^170–172 Such radical-based luminescence remains relatively rare and has been largely confined to π-conjugated systems that stabilize unpaired electrons through extensive TBC.^173–175 In 2022, Tang and Ye et al. reported the first example of non-conjugated radicals with CL properties based on pure n electrons (Fig. 14).¹⁷⁶ The non-conjugated radical polymer, namely poly(4-glycidyloxy-2,2,6,6-tetramethylpiperidine-1-oxyl) (PGTEMPO), could be simply synthesized from stable radical monomers (GTEMPO) via ring-opening polymerization (Fig. 14a). The synthesized PGTEMPO exhibits the same absorption spectra as its monomer, suggesting its non-conjugated structure after polymerization (Fig. 14b). In stark contrast to the non-emissive monomer, solid-state PGTEMPO exhibits a distinct red CL centered at around 635 nm with a PLQY of 1.3% and a lifetime of 0.2 ns, which is even detectable in the concentrated THF solution of 100 mM. Mechanistic investigations attribute this unusual red emission to the nitroxide radicals, whose electron lone pairs engage in dense intra- and inter-chain TSC in the aggregate state, thereby enabling the CL.¹⁷⁷ Meanwhile, the authors also pointed out that polymerization was a requisite for achieving the TSC of radicals and turning on the intrinsic CL of this system, which could not be realized by simply doping monomers in commercial polymers. With the sensitive response of radicals to antioxidants, as a proof of concept, PGTEMPO is also utilized as a fluorescent sensor for antioxidant detection and in vitro vitamin C mapping.

Fig. 14 Clusteroluminescence of TEMPO-based radical polymer. (a) The synthetic route from radical monomer (GTEMPO) to radical polymer (PGTEMPO). (b) The absorption spectra of GTEMPO and PGTEMPO, and photographs of their solid samples under daylight. (c) The photoluminescence spectra of GTEMPO and PGTEMPO, and photographs of their solid samples under 510–550 nm excitation. Reproduced with permission from ref. 176. Copyright 2022, The Royal Society of Chemistry. (d) The diagram of intra-/inter-chain through-space conjugation between radical groups within the polymer. Reproduced with permission from ref. 177. Copyright 2025, John Wiley & Sons, Inc.

3.2. π-Electron systems

Owing to the large oscillator strength and high transition probability of (π,π*) transition, π-electron CL systems exhibit better luminescence performance (e.g., broader color range and higher efficiency) and higher controllability compared to pure n-electron systems. The fundamental units of pure π-electron systems are isolated aromatic rings, such as benzene and naphthalene rings, and even biphenyl.^178,179 These isolated aromatic rings can only emit invisible light (≤380 nm) under UV irradiation, but compounds with isolated rings connected by electronically “insulating” alkyl segments can form TSC and further produce visible-light emission, indicating their CL properties. In addition, these π-electron systems can easily form regular packing, facilitating the growth of well-defined single crystals that are ideal for mechanistic studies and property modulation. In this section, π-electron CL systems will be introduced from multi-arylmethanes to multi-arylpropanes, as well as some special compounds and foldamers, focusing on their working mechanisms and potential structure–property relationships.

Poly(phenylene methylene)s (PPM) is an early-reported non-TBC system that can produce blue-green fluorescence with high efficiency (Φ = 41%).^180,181 From the conventional perspective of TBC, PPM and its methyl-substituted derivatives should not emit visible light, as their benzene rings act as discrete chromophores. Thus, this initially led to speculation that the emission might originate from oxidized anthracene side-products.¹⁸² Nevertheless, the slightly tilted cofacial conformation of the two aromatic rings (energetically the most favored compared with orthogonal or perfect cofacial conformation) and their close proximity of π electrons was further verified to endow them with bright luminescence in the solid state (Fig. 15a and b). This kind of electronic interaction between two π-systems separated by a non-conjugated unit (such as –CH₂–) has been termed homoconjugation by IUPAC, which is a special type of TSC (Fig. 15a).^183–185 Similar luminescence performance is also observed in the simplified model of diphenylmethane (DPM), although its emission wavelength is blue-shifted to about 349 nm (Fig. 15c and d).¹⁸⁶ The distance of two carbon atoms connected by one methylene unit (≈2.6 Å) is shorter than the sum of their van der Waals radius (3.4 Å), resulting in the overlap of p-orbitals of two carbon atoms. In addition, in polymeric PPM, the continuous homoconjugation along the chain enhances electron delocalization, resulting in a red-shifted CL emission spanning 400–500 nm compared to the monomeric DPM. Notably, the same PL behavior and homoconjugation are also observed in poly(phenylene sulfide) systems, where the connected methylene fragments between two benzene rings are replaced by sulfur atoms.^187–189

Fig. 15 Clusteroluminescence of poly(phenylene methylene)s (PPM). (a) Chemical structures of PPM and its methyl-substituted derivatives, the illustration diagram of homoconjugation, and the photographs of solid PPM under daylight (left) and UV irradiation (right). Reproduced with permission from ref. 185. Copyright 2017, Swiss Chemical Society. (b) The photoluminescence spectra of PPM excited at 388 nm (solid black line) and 438 nm (solid green line). The solid black line comprises a superposition of two parts: blue-light emission (dashed blue line) and green-light emission (dashed green line). Reproduced with permission from ref. 180. Copyright 2017, John Wiley & Sons, Inc. (c) The absorption and excitation spectra of diphenylmethane (DPM), which is the simplest example of PPM. (d) The photoluminescence spectra of DPM under different excitation wavelengths. Reproduced with permission from ref. 186. Copyright 2023, John Wiley & Sons, Inc.

While homoconjugation, as a special type of TSC, is widely recognized in small-molecule systems, its role and manifestation in polymeric materials remain comparatively underexplored. To better understand the structure–property relationship and bridge the two systems, a series of oligo(phenylene methylene)s (OPM) with repeating units from two to seven were synthesized and studied (Fig. 16a).¹⁹⁰ Photophysical studies reveal a distinct chain-length dependence. OPM[4]-OPM[7] exhibit strong TSC and blue CL at 440 nm with a high absolute Φ from 27% to 40%. In contrast, OPM[2] and OPM[3] are weakly emissive with Φ of 5% and 11%, respectively (Fig. 16b). Combined thermal measurements and theoretical calculations reveal that the outer DPM units (with larger dihedral angles) in the OPM chain are more flexible and more twisted than the inner ones (with smaller dihedral angles), implying that the efficient CL in OPM[4]–OPM[7] primarily arises from their inner structural segments, which are absent in the shorter oligomers (Fig. 16c). Furthermore, the study demonstrates that the enhancement of TSC with chain length is not unlimited, as evidenced by the fact that the blue CL of OPM[4]–OPM[7] at 440 nm is almost the same as that of PPM. This saturation behavior indicates that a continuous and stable TSC pathway can be established within a finite chain length. Consequently, these findings challenge the conventional assumption that high molecular weight is essential for efficient luminescence in non-conjugated polymers.¹⁹¹ Instead, they highlight that well-defined oligomers containing optimized inner segments for stable TSC are both necessary and sufficient for achieving high-efficiency, long-wavelength CL.

Fig. 16 Clusteroluminescence of oligo(phenylene methylene)s (OPM[n], n = 2–7). (a) Chemical structures of OPM[n] with different numbers of repeating units (from 2 to 7) and photographs of their solid samples under UV irradiation. (b) The absolute quantum yields of these OPM[n]. (c) The mechanism diagram of OPM[n] with different lengths of the main chain, which alters the dihedral angles of two adjacent benzene rings, the strength of intramolecular motion, and TSC. Reproduced with permission from ref. 190. Copyright 2024, John Wiley & Sons, Inc.

Building on the foundational role of homoconjugation in simple systems, diphenylmethane (DPM), as the smallest non-conjugated CL emitter, serves as an excellent model for probing intricate electronic interactions and CL mechanisms.¹⁸⁶ As discussed above, apart from locally excited TSC, TSCT is another important noncovalent electronic behavior of TSI. Under this consideration, a non-conjugated donor–acceptor (D–A)-modified DPM (i.e., DMA-CN-DPM) was designed, utilizing N,N-dimethylamino and cyano as electron donor and acceptor units, respectively (Fig. 17).¹⁰⁴ Its photophysical behavior is systematically compared with that of its structural analogue DMA-CN-BP, which features a more planar conformation favoring TBCT. Interestingly, these two compounds exhibit largely different responses to the polarity of solvents, displaying solvatochromic properties. With the gradually increased polarity from hexane to acetonitrile, the emission wavelength of TSC-based DMA-CN-DPM is greatly redshifted from dark purple to yellow, showing an obvious solvent effect. In contrast, the emission color of DMA-CN-BP only changes from purple to blue. Based on the analysis of electron structure, it is revealed that TSCT can promote complete charge separation between HOMO and LUMO, endowing non-conjugated luminogens with high sensitivity to environmental change (Fig. 17c). However, DMA-CN-BP with a comparatively planar conformation and TBCT display limited charge separation and low responsibility. These findings suggest that TSCT is not only a mechanism for tuning emission color but also an effective design strategy for luminogens with TADF, which requires complete charge separation.^192–194

Fig. 17 The different solvent effects of luminescence based on through-space charge transfer (TSCT) and through-bond charge transfer (TBCT). (a) Chemical structure of DMA-CN-DPM with non-conjugated donor and acceptor groups, photographs of DMA-CN-DPM in different solvents, and the plot of emission wavelength versus the solvent-polarity scale (E_T^N). (b) Chemical structure of DMA-CN-BP with conjugated donor and acceptor groups, photographs of DMA-CN-BP in different solvents, and the plot of emission wavelength versus E_T^N. (c) The illustration diagram of the TSCT and TBCT channels, respectively. Reproduced with permission from ref. 104. Copyright 2024, Springer Nature.

To better illustrate the photophysical processes of TSC and CL, the propeller-shaped triphenylmethane (TPM) and its derivatives with electron-donating and withdrawing groups were studied, which exhibited crowded steric hindrance and a C₃-symmetric conformation.¹⁹⁵ The basic TPM shows absorption at 264 nm in both THF solution and solid states belonging to these isolated benzene rings, which is unmatched by its excitation spectra, like other CLgens (Fig. 18a). The emergence of CL could be first observed in THF/water mixtures. In dilute solution, only invisible emission from benzene rings located at 285 nm could be observed (Fig. 18b). With the increased water fraction, the spectra shows a newly emergent emission peak at around 400 nm, which reaches 17-fold higher than that in dilute solution, showing its typical AIE effect (Fig. 18c). In the solid state, TPM exhibits two emission peaks at 288 nm and 402 nm, respectively. The former arises from the isolated benzene-ring emission while the latter results from the formed intramolecular TSC and the corresponding CL (Fig. 18d). The frontier molecular orbital of LUMO shows this noncovalent overlap of electrons, which could be regarded as a complex homoconjugation at the central of the C₃-symmetric TPM, providing direct evidence of electronic TSC (Fig. 18e). Through theoretical calculation and reorganization energy analysis, the authors provide a clear picture of photophysical processes of TPM in different environments. In the dilute solution, excitons rapidly relax to the TBC-based emission channel after photoexcitation, resulting in the short-wavelength emission at 285 nm. Meanwhile, TSC also forms but is highly associated with nonradiative decay due to dynamic intramolecular motions, which diminish the long-wavelength emission. In both aggregate and solid states, due to the restricted intramolecular motions, the formed TSC in the excited state can be stabilized, resulting in a new radiative decay channel. Thus, some excitons relax via the TBC-based emission channel, and other excitons can produce blue-color emission via the stabilized TSC-based channel (Fig. 18f). Excitingly, the electronic effect on TSC and CL is also introduced using the potential energy surfaces, as shown in Fig. 18g. With the gradually increased electron-donating ability, TPM derivatives display inward charge transfer and concentrated TSC at the central part of their molecular skeleton, which stabilizes their excited-state geometry and increases electron density, resulting in enhanced strength of TSC and CL efficiency. In contrast, TPM derivatives with electron-withdrawing groups show outward charge transfer, which decreases the central electron density and destabilizes TSC, finally quenching the long-wavelength CL. While this work establishes a rational strategy for tuning TSC and CL in non-conjugated systems via electronic structure, the achieved luminescence efficiency remains below 10%, highlighting an area for further optimization.

Fig. 18 Clusteroluminescence of triphenylmethane (TPM). (a) The absorption and excitation spectra of TPM in solution and solid states. (b) Photoluminescence (PL) spectra of TPM in THF/water mixtures with different water fractions (f_w). (c) Plots of relative PL intensity (I/I₀) versus f_w at different emission wavelengths. Concentration = 10⁻⁴ M, λ_ex = 270 nm, and I₀ = intensity at f_w = 0%. (d) PL spectra of TPM and the photograph of its crystalline samples under the 365 nm UV irradiation. (e) The frontier molecular orbitals of TPM, which show the through-space conjugation (TSC) of electrons at the central part of the molecular skeleton in the LUMO. (f) The potential energy surface of TPM illustrating its dual emission from through-bond conjugation and TSC stabilized by restricted intramolecular motion in the crystalline state. (g) The charge-transfer effect and proposed potential energy surfaces for the clusteroluminescence of triphenylmethanes with different electron-donating and electron-withdrawing groups. ICT, inward charge transfer; OCT, outward charge transfer; +, stabilization/increment; −, destabilization/decrement. Reproduced with permission from ref. 195. Copyright 2021, American Chemical Society.

It is important to note that the efficiency of CL primarily depends on two factors, including the strength of TSC and the stability of the corresponding conformation. The former is largely influenced by the degree of electron communication, while the latter is determined by the rigidity of the molecular skeleton and the surrounding environment.¹²⁸ To investigate the above factors, one method is to replace the building units with different flexibility. By introducing rigid naphthalene rings with extended TBC-based conjugation, a series of trinaphthylmethane (TNM) isomers with different connecting positions was designed (Fig. 19).¹⁹⁶ These four isomers show similar photophysical behaviors to TPM that the long-wavelength emission from TSC could be observed when forming aggregates, but their luminescence efficiency (PLQY = 16–61%) is much higher than that of the TPM series. Besides, the connecting position of the naphthalene units plays a decisive role in modulating their photophysical properties (Fig. 19a). In particular, only the intrinsic emission of 222-TNM from isolated naphthalene rings is observed in the solid state, and no long-wavelength emission peak is detected. For 122-TNM, it displays the long-wavelength emission peak at 375 nm, while its intensity is lower than the intrinsic emission at 343 nm. A clear enhancement is observed in 112-TNM and 111-TNM, where the long-wavelength CL band at 380 nm and 393 nm, respectively, becomes the dominant emission and shows fine vibrational structures. The above results indicate that the strength of TSC is gradually enhanced with the increased number of 1-naphthalene units, which reaches the maximum in 111-TNM. Single-molecule analysis reveals that 222-TNM exhibits the most flexible skeleton due to its weak intramolecular steric hindrance. As the number of 1-naphthalene units increases, the molecular rigidity is progressively enhanced, leading to greater stabilization of the excited-state geometry and more effective TSC. Meanwhile, single-crystal structures and intermolecular interaction analysis indicate that the flexibility of formed aggregates also increases from 222-TNM to 111-TNM, which may help realize a suitable conformation of TSC after photoexcitation. Thus, this work proves the independence of the intrinsic rigidity of the molecular skeleton and the flexibility of aggregates, and suggests that the ultrastrong TSC mainly originates from significant intramolecular electronic overlaps.

Fig. 19 Clusteroluminescence parameters and structure–property relationship of trinaphthylmethanes (TNM) with different connecting positions. (a) Photoluminescence spectra of different TNM solids. (b) With the increased number of 1-naphthalene units, the rigidity of molecular skeletons and flexibility of aggregates increase, resulting in the increased strength of through-space conjugation and redshifted emission wavelength. Reproduced with permission from ref. 196. Copyright 2024, Springer Nature.

Building on the structure–property relationship established with naphthalene-based systems, Tang and colleagues extended their investigation to more flexible building blocks by replacing benzene rings with biphenyl units, synthesizing a series of tri(biphenyl)methanes (TBPM) to further elucidate the role of conformational flexibility (Fig. 20).¹⁹⁷ Apart from the intrinsic emission from biphenyl units at around 310 nm, these three compounds also show long-wavelength CL due to the TSC of biphenyl units. Notably, the flexibility of the biphenyl linkers allows for diverse molecular conformations in the crystalline state, leading to markedly different photophysical behaviors.^198–200 For para-substituted p-TBPM, the loose arrangement indicates the flexible skeleton, which can form TSC in the excited state but is also unable to restrict intramolecular motions. As a result, p-TBPM can produce CL at 365 nm with a moderate PLQY of 76% and a full width at half maximum (FWHM) of 53 nm. For symmetric ortho-substituted o-TBPM, the crowded subunits enhance electron overlap, theoretically favoring strong TSC and yielding the longest CL wavelength (390 nm) among the series. However, this structural crowding also brings an energy barrier that hinders relaxation to the conformation required for strong TSC. Additionally, weak intermolecular interactions further destabilize the formed TSC. Consequently, it only produces a broad emission peak (FWHM > 70 nm) with a very low quantum yield (Φ = 4%). Finally, meta-substituted m-TBPM displays an unexpected asymmetric and rigid conformation in crystals, which enables strong electronic communication and significantly stabilizes the formed TSC. In particular, multiple intermolecular interactions create a fixed environment to suppress molecular motions and reduce vibrational energy levels. These combined effects lead to excitation-independent and narrowband CL with an FWHM of 40 nm and a remarkable PLQY of 100%.

Fig. 20 Clusteroluminescence parameters and single-crystal conformation of three tri(biphenyl)methanes with different connecting positions. p-TBPM and o-TPBM show symmetric and flexible molecular skeletons, which result in comparatively low emission efficiency or large full width at half maximum (FWHM). In contrast, m-TBPM with an asymmetric and rigid skeleton exhibits the highest quantum yield with a narrowed FWHM of resulting clusteroluminescence. Reproduced with permission from ref. 197. Copyright 2024, Springer Nature.

While the above multi-arylmethanes have been proven to produce CL due to the formation of TSC, it is undeniable that polyarylmethanes can easily form free radicals under certain conditions, thereby producing extrinsic long-wavelength luminescence.^201,202 Similar to PPM as discussed above, non-conjugated poly(diphenylmethane) (PDPM) with one more benzene substitution is another example of polymeric CL materials.⁴⁹ Apart from its CL within the blue and yellow region, near-infrared (NIR) emission is also observed in PDPM due to the formed radicals (Fig. 21a and b). In concentrated THF solution, PDPM shows excitation-dependent emission peaks within the range of 705–890 nm with the increased excitation wavelength. Electron paramagnetic resonance (EPR) spectra prove the existence of radicals, which could be detected once the concentration is above 0.1 M and are stable for several hours under ambient conditions or photoirradiation in the powder state (Fig. 21c). Although the g factor of 2.006 indicates the benzoquinone-based radical species, intra-/inter-chain TSC of these radicals is also embedded and stabilized by the PDPM polymer-chain matrix. As a result, PDPM can produce full-color emission from blue to NIR under different excitation wavelengths, which is due to the mixed short-range TSC (for blue emission) and long-range (for yellow emission) of PDPM, as well as the in situ generated stable radicals (for red/NIR emission). It is worth noting that tetraphenylmethane and its polymer form (namely poly(triphenylmethane)) are not good candidates for CL, which show short-wavelength emission within the blue-color region and low efficiency (PLQY ≤ 4%). This may be largely due to the inability of electrons to form an effective and stable TSC between the four benzene rings. In addition, a similar redshift of CL caused by radicals is also reported in the poly(methyl acrylate) system, which could be utilized for information loading, rewriting, and multifunctional coating with photochromic features.²⁰³

Fig. 21 Clusteroluminescence of poly(diphenylmethane) (PDPM) with stable radicals. (a) The synthetic route of PDPM with radicals and a diagram of through-space conjugation within the polymeric skeleton. (b) The excitation-dependent photoluminescence spectra of the synthesized PDPM, which extend into the near-infrared region. (c) The electron paramagnetic resonance spectra of PDPM powder under different UV irradiation times. Reproduced with permission from ref. 49. Copyright 2024, John Wiley & Sons, Inc.

Beyond TPM and its derivatives, tetraphenylethane (s-TPE) is another famous CLgen reported early in 2017.⁵⁷ In THF/water mixtures with a low water fraction, s-TPE only exhibits a peak at 290 nm assigned to benzene emission, while a marvelous long-wavelength emission at 460 nm is observed when forming aggregates with a water fraction ≥70% (Fig. 22a). This newly formed emission peak is the result of intramolecular TSC and restricted intramolecular motions, whose intensity is more than 350 times higher than that in pure THF solution. In contrast, the trimethyl-substituted derivative s-TPE-TM demonstrates weaker TSC, as reflected in its shorter-wavelength emission at 390 nm and significantly lower intensity under the same conditions (Fig. 22b). In particular, s-TPE shows bright sky-blue CL at 467 nm with a high Φ of 69% while s-TPE-TM is weakly emissive at 397 nm with an Φ of 9% (Fig. 22c). Frontier molecular orbital analysis reveals that s-TPE achieves effective spatial homoconjugation between two benzene rings on the same carbon atom, leading to a reduced energy gap of 3.44 eV. s-TPE-TM, however, shows poor electronic overlap due to steric hindrance from the methyl groups, preventing efficient TSC and resulting in a larger energy gap of 4.96 eV and consequently weak emission (Fig. 22d). These findings clearly demonstrate that steric hindrance brought by methyl units can change the excited-state conformation of compounds and finally alter the orientation and strength of TSC and corresponding CL. As highlighted by David Schilter, an editor of Nature Reviews Chemistry, these materials are not formally π-conjugated, but TSC sees them fluoresce at longer wavelengths than one might predict.²⁰⁴ It is this work that opened a new and rapidly expanding research direction focused on non-conjugated materials and their clusteroluminescence properties during the past years.

Fig. 22 Clusteroluminescence of tetraphenylethane (s-TPE). (a) The photoluminescence (PL) spectra of s-TPE in THF/water mixtures with different water fractions (f_w) and the plot of relative PL intensity (I/I₀) versus f_w at different emission wavelengths. Concentration = 1 × 10⁻⁴ M and I₀ = PL intensity in pure THF solution. (b) The PL spectra of s-TPE-TM in THF/water mixtures with different f_w and the plot of relative PL intensity (I/I₀) versus f_w at different emission wavelengths, concentration = 1 × 10⁻⁴ M. (c) The illustration of through-space interaction of electrons between four isolated benzene rings and photographs of crystalline s-TPE and s-TPE-TM taken under 365 nm UV irradiation. (d) The calculated energy levels and frontier molecular orbitals of s-TPE and s-TPE-TM. Reproduced with permission from ref. 57. Copyright 2017, American Chemical Society.

Although the orientation of non-conjugated fragments has been proven to affect TSC, the underlying mechanism is still unclear. The connecting units between non-conjugated fragments are an approach to regulating the conformation of CLgens. Under this consideration, a series of tetraphenylalkanes (TPAs) was synthesized and compared, where two diphenylmethyl units were separated by the alkyl chain with different carbon atoms from one to seven (Fig. 23a).²⁰⁵ Interestingly, this system exhibits the first instance of the excited-state odd–even effect of organic luminescence properties, which is occasionally observed in liquid crystals and interface engineering with ground-state features.^206–208 Specifically, TPAs with an even number of alkyl carbons show bright sky-blue CL (432–463 nm) and high PLQY (43–68%), while TPAs with odd-numbered alkyl chains are only weakly emissive (350–383 nm) with negligible PLQY (≤ 8%) (Fig. 23b). Combined experiments and theoretical calculations reveal that the photophysical behavior of TPAs is strongly influenced by the alkyl chain. When the length of the alkyl chain is no more than five carbon atoms, even-numbered alkyl chains induce a staggered conformation of two benzene rings at two terminal carbons, which results in an effective electronic coupling (similar to J-aggregate) and enhanced CL (Fig. 23c). In contrast, odd-numbered alkyl chains promote a face-to-face TSC (similar to H-aggregate), facilitating efficient nonradiative decay.^209–211 On the other hand, when the alkyl chain contains six or seven carbon atoms, intermolecular interactions are dominant. The single crystal of C6-TPA reveals a staggered intermolecular TSC between two molecules with a short C–C distance of around 3.4 Å, which is conducive to long-wavelength luminescence. However, C7-TPA fails to yield single crystals suitable for X-ray diffraction, indicating that the consistent orientation of the terminal benzene rings may hinder molecules from approaching each other closely and forming intermolecular TSC.

Fig. 23 Clusteroluminescence of non-conjugated tetraphenylalkanes (TPAs) with different lengths of alkyl chain. (a) Chemical structures, quantum yields, and luminescence photographs (taken under 365 nm UV irradiation) of TPAs. (b) The plot of the maximum emission wavelength of TPAs versus the number of carbon atoms. (c) An illustrated diagram of the working mechanism and the excited-state odd–even effect in through-space interaction. Reproduced with permission from ref. 205. Copyright 2023, American Chemical Society.

A similar structure-dependent luminescence phenomenon is also noticed in the system of multi-aryl-substituted propane, where substitutions provide a steric hindrance to alter their excited-state geometries and photophysical behavior.²¹² TPP (also known as C3-TPA in Fig. 23), Me-TPP, and Ph-TPP all show a non-conjugated nature as their absorption maximum is shorter than 280 nm (Fig. 24a). In the crystalline state, the intrinsic peak belonging to isolated benzene rings dominates the emission of TPP with a low PLQY of 8%, attributed to ineffective TSC caused by a face-to-face molecular conformation and inherent skeletal flexibility (Fig. 24b). By adding one methyl into the skeleton, methyl substituted Me-TPP displays an enhanced molecular rigidity, but the enhancement of the TSC between two vicinal isolated benzene rings is limited due to the increased distance, resulting in excitation-dependent emission around 420 nm and a high PLQY of 42%. The replacement of a methyl unit by a benzene ring not only increases molecular rigidity but also promotes a staggered conformation of two vicinal isolated benzene rings, promoting efficient TSC and excitation-independent CL at 410 nm with a high PLQY of 38%. The above examples clearly support that precise control over the relative orientation of non-conjugated chromophores is crucial for designing efficient CLgens, providing a clear structural guideline for manipulating their photophysical properties.

Fig. 24 Clusteroluminescence of multi-aryl-substituted propane. (a) The absorption spectra of TPP, Me-TPP, and Ph-TPP, which indicate their non-conjugated nature of molecular skeletons. (b)–(d) The excitation-dependent photoluminescence spectra of (b) TPP, (c) Me-TPP, and (d) Ph-TPP in the crystalline state, and their photographs taken under 365 nm UV irradiation. Reproduced with permission from ref. 212. Copyright 2023, Chinese Chemical Society.

Tetraphenylethylene (TPE), the star molecule of AIE luminogens, exhibits only weak ultraviolet emission in dilute solution, but its aggregate state produces a significantly enhanced visible blue luminescence around 460 nm (Fig. 25a).^213–217 Its visible emission was thought to originate from TBC-based electronic delocalization, but the excited-state energy level of TBC does not match it. To illustrate the controversial concern, Tang et al. systematically studied the spectroscopy and theoretical simulation of TPE and phenylethylene derivatives.⁴⁶ With a progressively decreased number of phenyl rings from TPE to phenylethylene derivatives, they generally display blueshifted emission wavelength. Especially, v-DPE and g-DPE with two phenyl rings in vicinal (v) and geminal (g), respectively, show different behaviors in dilute solutions at 298 K and 77 K. They exhibit primary emission peaks in the ultraviolet region at room temperature (RT), but these peaks persist in low-temperature (LT) solutions with markedly enhanced intensity. Notably, v-DPE develops a distinct shoulder peak at 425 nm under the LT condition, which is negligible at RT (Fig. 25b). On the other hand, g-DPE only shows a pronounced LT shoulder peak at 332 nm (Fig. 25c). This demonstrates that the redshifted emission of TPE can occur via a newly formed excited-state species even with just two adjacent phenyl rings in the molecular skeleton. However, trans-positioned v-DPE (namely t-DPE) maintains identical emission peaks from RT to LT with no visible-range enhancement, which confirms that inter-phenyl interactions are the dominant driver of the bathochromic emission observed in TPE systems from solution to aggregate states. The comparison of v-DPE and g-DPE reveals that v-DPE exhibits vicinal excited-state through-space interaction (ESTSI) and thus exhibits lower LUMO energy, smaller bandgap, and more red-shifted emission.

Fig. 25 Clusteroluminescence of tetraphenylethylene (TPE). (a) The photoluminescence spectra of TPE in THF/water mixtures with different water fractions (f_w) and its solid-state luminescence photograph. (b) and (c) The photoluminescence spectra of (b) v-DPE and (c) g-DPE in the THF solution with a concentration of 1 × 10⁻⁵ M at 298 K (RT) and 77 K (LT). (d) The molecular orbitals for the electronic transition at S₁ geometries of v-DPE and g-DPE. (e) The illustration diagram of excited-state through-space interactions (ESTSI) of TPE in the solution state and solid state. (f) ESTSI systems with different numbers of carbon atoms (n = 1, 2, 3) between the two benzene rings with coupling. Reproduced with permission from ref. 46. Copyright 2022, American Chemical Society.

Accordingly, it can be concluded that TPE exhibits only weak ultraviolet emission in the dilute solution originating from TBC (Fig. 25e). This is attributed to active intramolecular torsion, which promotes ultrafast nonradiative decay of ESTSI excitons via conical intersection points. In contrast, aggregation into the solid state effectively stabilizes ESTSI conformation of TPE via RIM and steric antagonism between vicinal and geminal benzene ring pairs. These effects collectively stabilize electron delocalization and enhance through-space electronic conjugation, resulting in dramatically intensified blue emission in the aggregate state. To provide a general picture of ESTSI for multi-aryl-substituted alkanes, different types of through-space orbital coupling are realized according to different numbers of carbon atoms between phenyl rings (Fig. 25f).

In fact, TSC is not exclusive to non-conjugated structures but also occurs in certain twisted conjugated systems, where the electronic conjugation mode can be precisely tuned by the degree of molecular twist. Terphenyl isomers with different connecting positions are typical examples to illustrate this trend (Fig. 26a).¹³¹ The para-connected terphenyl (p-TPh) is the traditionally TBC-based compound, where its frontier molecular orbitals of HOMO and LUMO are equally delocalized on the skeleton via covalent bonds. In comparison, by changing the connecting position to the ortho, o-TPh becomes twisted with a fjord-type geometry, enabling significant TSC between two terminal benzene rings (Fig. 26b). This fundamental difference in electronic structure directly governs their luminescence behavior. Due to the comparatively planar skeleton, p-TPh shows the ACQ phenomenon where its emission at 340 nm gradually declined in the DMF/water mixtures with the increased water fraction (Fig. 26c). Conversely, o-TPh is non-emissive in dilute solution but shows largely increased emission intensity when forming aggregates and stabilized TSC. The observed redshifted luminescence at 370 nm of o-TPh also suggests that TSC is a more effective approach than TBC to increase the strength of electron delocalization within molecules with the same subunits and molecular weight, highlighting its potential advantage in modulating optoelectronic properties.

Fig. 26 The transformation from through-bond conjugation (TBC) to through-space conjugation (TSC). (a) Chemical structures of p-TPh and o-TPh. The former with planar conformation shows an aggregation-caused quenching (ACQ) effect, while the latter with a twisted fjord-type structure exhibits the aggregation-induced emission (AIE) effect. (b) The frontier molecular orbitals of p-TPh without TSC and o-TPh with TSC in LUMO. (c) and (d) The photoluminescence spectra of (c) p-TPh and (d) o-TPh in N,N-dimethylformamide/water mixtures with different water fractions (f_w). Reproduced with permission from ref. 131. Copyright 2024, Chinese Chemical Society.

According to the traditional design strategies for AIEgens, small apolar aromatic groups (e.g., phenyl rings) are frequently incorporated into fluorophore structures only for their steric effects to restrict intramolecular rotations, thereby enhancing aggregate-state luminescence.^220–222 However, recent studies have revealed that these aromatic subunits can also engage in TSC, leading to unconventional photophysical properties in appropriately designed twisted molecular systems. In 2017, McGonigal et al. reported a series of molecular rotors bearing three, five, six, and seven phenyl units.²¹⁸ Beyond the conventional intermolecular dimer emission commonly observed in AIEgens, these systems also exhibited distinct emission from intramolecular dimers, which could be systematically modulated by varying temperature and solvent viscosity (Fig. 27a). Among these molecular rotors, Ph₇C₇H bearing seven phenyl units is a special one. When its dilute solution in 2-MeTHF is gradually cooled from 200 K to 90 K, a progressive hypsochromic shift of the emission maximum occurs from 2.84 eV (200 K) to 3.32 eV (90 K), accompanied by distinct changes in peak shape (Fig. 27b). The low-energy emission band appeared broad and featureless, while the high-energy band exhibited a clear vibronic fine structure, suggesting contributions from two distinct excited-state species under these conditions. DFT calculation and single-crystal structure reveal that the flexible sp³ center of the Ph₇C₇H skeleton facilitates conformational reorganization, enabling the formation of an intramolecular aromatic dimer in the excited state through nonplanar distortion of the central cycloheptatriene ring. As a result, the long-wavelength emission is observed at 200 K. In contrast, rotors with three or five phenyl rings, constrained by more rigid central cores, cannot achieve comparable intramolecular interactions or tunable emission. This work also hints that the degree of conformational flexibility of these twisted molecules is a double-edged sword, which should be sufficient flexibility for subunits to come into contact while maintaining a sterically crowded environment to minimize nonradiative decay.²²³

Fig. 27 Through-space aromatic interactions of molecular rotors. (a) Schematic representation of two processes for intramolecular and intermolecular dimer formation. (b) The photoluminescence (PL) spectra of Ph₇C₇H in 2-MeTHF solution (concentration = 10 µM) at different temperatures, which shows two different conformations and corresponding luminescence states. Reproduced with permission from ref. 218. Copyright 2017, American Chemical Society. (c) The single-crystal structures and the distances of a specific phenyl–phenyl non-covalent interaction (NCI) of Ph₇C₇H under high pressure. (d) The normalized absorption and PL spectra of Ph₇C₇H under different pressures, which display a redshifted wavelength with increased pressure. Reproduced with permission from ref. 219. Copyright 2023, American Chemical Society.

In addition, this kind of through-space electronic interaction of intramolecular dimers in Ph₇C₇H could be further modulated in the solid state through the application of high pressure (Fig. 27c).^{219,224–226} With the increased external pressure from 0.15 GPa to 2.61 GPa, the distance between the T-shaped (edge-to-face type) phenyl rings becomes shorter, concurrently enhancing the electronic coupling within intermolecular staggered dimers. Correspondingly, the absorption and emission of the solid Ph₇C₇H under ambient conditions are the same as those in the solution state, which are obviously redshifted under a pressure of 3.06 GPa caused by both enhanced intra- and intermolecular dimers (Fig. 27d).

Featuring a distinctive stacked architecture, [2.2]paracyclophane ([2.2]pCp), comprising two co-facially aligned aromatic rings bridged at their para positions by ethylene linkers, has been thought to show TSC.^229–233 Despite extensive investigation in optoelectronics, the nature and extent of TSC between its aromatic rings remain subjects of debate, primarily due to the nearly perfect π-orbital overlap in its symmetric structure.^234–236 Nevertheless, inspired by this folded topology, a new class of TBC-based foldamers containing a TPE core is synthesized and has been proven to show strong TSC (Fig. 28a).^227,237,238 X-ray crystallographic analysis confirms that the prototypical foldamer, (Z)-o-BPTPE, adopts a cis-isomeric configuration, where two phenyl rings adopt a nearly parallel stacked arrangement, exhibiting 50% plane overlap with inter-ring distances around 3.2 Å (Fig. 28b). This arrangement, facilitated by covalent TBC frameworks and a deliberately offset stacking between phenyl groups, enables strong TSC, as visually evidenced by spatially delocalized electron density in the LUMO+1 orbital (Fig. 28c). It is worth noting that the auxiliary phenyl units on the side with spatial repulsion should function as a critical driving force to bring about and stabilize the parallel arrangement of the central phenyl rings.²³⁹ Accordingly, unlike traditional TPE-based AIEgens that are non-emissive in the solution state, (Z)-o-BPTPE shows blue emission at 492 nm in dilute THF solution with a high efficiency of 45% due to the RIM mechanism. Its emission intensity is further enhanced, and a slight hypsochromic shift occurs after forming aggregates in the THF/water mixtures, consistent with the aggregation-enhanced emission (AEE) effect.^240–243

Fig. 28 Through-space conjugation (TSC) and luminescence of foldamers. (a) Chemical structures and (b) single-crystal structure of (Z)-o-BPTPE featuring TSC. (c) The frontier molecular orbitals (FMOs) of (Z)-o-BPTPE, which show typical TSC between folded structures in LUMO+1. Reproduced with permission from ref. 227. Copyright 2016, Elsevier. (d) Chemical structures and FMOs of folded tetraphenylethene derivatives. (e) The absorption and photoluminescence spectra of a π-stacked oligo-p-phenylene, f-4Ph(Me). Reproduced with permission from ref. 228. Copyright 2015, American Chemical Society.

In addition, adding a pair of vinyl-bridged oligo-p-phenylene groups into the skeleton maintains the folded and cis-form conformation, wherein the oligo-p-phenylene groups align closely in an offset and face each other in a roughly parallel manner (Fig. 28d).²²⁸ For these foldamers, their absorption spectra usually show two main peaks, the strong short-wavelength absorption attributed to the high-energy transition associated with TSC from S₀ to S_n, while the weak long-wavelength absorption belongs to the TBC-based molecular skeleton with S₀–S₁ transition. Interestingly, with the increased length of oligo-p-phenylene groups, the short-wavelength band becomes stronger and redshifted, but the long-wavelength absorption bands are weakened, indicating the dominant influence of TSC on their photophysical properties. According to Kasha's rule, fluorescence typically occurs from the lowest excited state of S₁.²⁴⁴ In these systems, however, the large oligo-p-phenylene groups can show a large Stokes shift by virtue of intramolecular energy transfer from the stacked phenyl rings (for absorption) to the central alkenyl-linked fragment (for emission).²⁴⁵ For instance, f-4Ph(Me) displays the maximum absorption peak at 295 nm and the maximum emission peak at 498 nm, indicating a large Stokes shift of 203 nm (Fig. 28e), which is rarely observed in pure hydrocarbon luminogens without electronic donor–acceptor interactions.²⁴⁶ It is also observed that the rotational motion of the unstacked phenyl rings plays a critical role in inducing redshifted PL and large Stokes shifts, supported by the substantially smaller Stokes shift (113 nm) observed in a rigid folded terphenyl dimer lacking such rotatable groups.²⁴⁷ Thus, in these special foldamers, a balance between structural rigidity imparted by π-stacked oligo-p-phenylenes and conformational flexibility provided by unstacked phenyl rings is important to achieve high luminescence efficiency and large Stokes shifts.²⁴⁸ In addition, it is noteworthy that, apart from efficient photoluminescence, this kind of foldamers with TSC are also beneficial for the charge recombination process and stabilize the diradical state, achieving photocatalytic oxidative coupling and hydrogen evolution.²⁴⁹

The folded architecture inherently confers axial chirality to these foldamers, while the integrated donor–acceptor interactions within their scaffold impart stimuli-responsive properties.^250,251 By connecting two biaryl arms to the 9,10-positions of phenanthrene, Zhao et al. designed a series of foldamers with axial chiral folded conformation and high-polarizability substituents.²⁵² Taking the cyano-substituted ap-Pn-PPCN for example, it shows anti-parallel arrangement stabilized by strong TSC, which is conducive to the separation of the axial chiral enantiomers (i.e., (R)-ap-Pn-PPCN and (S)-ap-Pn-PPCN as shown in Fig. 29b). Circular dichroism (CD) spectra indicate the enantiomers of them with the Cotton effect and almost the same shape of peaks in different solvents (Fig. 29c). However, in the PL spectra, with the increase in solvent polarity, ap-Pn-PPCN shows a newly emergent peak at the long-wavelength region, probably suggesting multiple emission states or vibrational-state distribution changes in different solvents (Fig. 29d). Interestingly, further spectroscopic study discloses their distinct dual circularly polarized luminescence (CPL) features in dilute solutions. That is, in low-polarity n-hexane, a single CPL peak is observed at 360 nm with a dissymmetry factor (g) of ±2.10 × 10⁻³. As solvent polarity increases, a new CPL band emerges at 425 nm with an opposite sign and shows significant intensification, while the original peak at 360 nm exhibits slight attenuation (Fig. 29e and f). Furthermore, in high-polarity DMF, the g values for the 425 nm peaks reach +3.26 × 10⁻³ for (R)-ap-Pn-PPCN and −4.70 × 10⁻³ for (S)-ap-Pn-PPCN, respectively. Both PL and CPL results simultaneously indicate the different emissive species in the excited state of this foldamer, which are highly dependent on solvent polarity. Theoretical calculations reveal that there are two local minimum points associated with two different folded secondary structures on the potential energy surface (Fig. 29a). The transformation between the two secondary structures is closely related to the torsion angles between the biaryl arms and central phenanthrene core. High-polarity solvents can decrease the energy barrier between these two secondary structures, facilitating the balance of them and dual CPL with opposite signs, while low-polarity solvents impose a higher barrier that restricts conformational interconversion.

Fig. 29 Circularly polarized luminescence (CPL) of chiral foldamers. (a) Chemical structure of ap-Pn-PPCN and diagram of the excited-state conformation transformation for dual CPL. (b) Schematic illustration of the chiral axis and substituent groups (1–4) in the foldamer. (c) The circular dichroism (CD) spectra of enantiomers of ap-Pn-PPCN in different solvents. (d) The photoluminescence spectra of ap-Pn-PPCN in different solvents at room temperature. (e) The CPL spectra of (R)-ap-Pn-PPCN in different solvents. (f) The CPL spectra of (S)-ap-Pn-PPCN in different solvents. Reproduced with permission from ref. 252. Copyright 2024, John Wiley & Sons, Inc.

From homoconjugation within diphenylmethane to TSC within multi-aryl-substituted alkanes and some TBC-based compounds with twisted conformations, these examples collectively show the electronic structures of TSC between separated phenyl rings, which play a vital role in regulating their molecular conformation and photophysical properties. However, the emission wavelengths in these examples fall in the blue–green region, underscoring the intrinsic limitations of π-electron systems and, in turn, constraining compatibility with conventional TBC-based luminogens and potential applications. Despite this constraint, with well-defined conformation and ordered packing of crystalline structures, these compounds provide a platform to disclose the structure–property relationship between TSC and CL. Such systematic insight is indispensable for advancing the fundamental understanding of these long-overlooked noncovalent electronic interactions and the emergent luminescence they enable.

3.2. n/π-Electron hybrid systems

Although purely n-electron and π-electron systems have been demonstrated to exhibit unconventional luminescence due to the formation of TSI, their performance is often limited by short emission wavelengths and low PLQY. By strategically integrating the complementary features of both systems, n/π-electron hybrid architectures present a promising route toward overcoming these limitations and achieving enhanced luminescence performance.²⁵³ In such hybrid systems, the two components operate synergistically. High electron density, wider controllability of orbital orientation, and various cluster forms of heteroatoms can collectively promote and enhance the strength of TSI, which may redshift the emission wavelength.²⁵⁴ Meanwhile, isolated double bonds or benzene rings can increase the rigidity and ordered arrangement of the whole skeleton, stabilizing the excited-state conformation and decreasing non-radiative decay.^255,256

Under this consideration, Xiong et al. reported a series of diarylmethanes with only two isolated heteroatomic rings (Fig. 30).⁴⁷ First, o-2Ox, o-2Th, and o-2Pr bearing furan, thiophene, and pyrrole rings are introduced to investigate the effect of different heterocyclic rings (Fig. 30a–c). Although it has a non-conjugated nature, pure o-2Ox oil with intramolecular π/n–π TSI can emit visible light at 550 nm, showing typical CL properties. However, crystalline o-2Th and o-2Pr only exhibit π–π TSI without the involvement of lone pairs, resulting in weak and short-wavelength emission at 420 nm and 350 nm, respectively. It is also noteworthy that hydrogen bonding may be a negative factor in destroying intra/intermolecular interactions of lone pairs, further weakening TSI and quenching CL.²⁵⁷ In addition, by gradually increasing the number of heteroatoms, three non-conjugated CLgens (namely, o-1Py–1Ph, o-2Py, and o-2Md) with different numbers of nitrogen atoms are constructed, and the effect of n–n TSI is validated (Fig. 30d–f). Interestingly, the oily nature of these three emitters endows them with enhanced molecular mobility and conformational flexibility, which may facilitate the formation of multiple excited-state conformations. With only one pyridine ring, o-1Py–1Ph only shows a single emission peak at 534 nm, which could be attributed to the n–π TSI in the aggregate state due to the negligible contribution of pure π–π TSI for the long-wavelength emission (Fig. 30d). As a comparison, the PL spectra of o-2Py exhibits excitation-dependent emission ranging from 580 to 612 nm, indicating multiple excited-state conformations and the participation of n–n TSI (Fig. 30e). Surprisingly, o-2Md demonstrates a board emission peak 700 nm with a high PLQY of 25%, where two pyrimidine units with four lone pairs can easily and efficiently form strong n–n TSI (Fig. 30f). Accordingly, the diagram of the mechanism for CL of these n/π-electron hybrid systems is illustrated in Fig. 30g, highlighting the types and strength of TSI for CL. Compared with pure π–π or π–n TSI, n–n TSI represents a more efficient approach to redshift CL wavelength. Importantly, o-2Md with a small relative molecular mass of 172 and NIR luminescence exceeding 800 nm also represents the milestone of CL and organic emitters, which is even smaller than the recently reported single-benzene-based NIR fluorophores.²⁵⁸

Fig. 30 Clusteroluminescence of diarylmethanes with heteroatoms, which greatly redshift their emission wavelength. (a)–(f) Excitation-dependent photoluminescence spectra of pure (a) o-2Ox oil, (b) o-2Th crystals, (c) o-2Pr crystals, (d) o-1Py–1Ph oil, (e) o-2Py oil, and (f) o-2Md oil. The corresponding photographs were taken under 365 nm UV irradiation. (g) The illustration diagram of the mechanism for clusteroluminescence, which is regulated by different types and strengths of through-space interaction (TSI), as well as molecular conformation. Reproduced with permission from ref. 47. Copyright 2025, Elsevier.

The involvement of heteroatoms and lone pairs also introduces the non-uniform electron distribution, resulting in D–A structures and TSCT.^259–261 Following the design principle of conjugated D–A-based luminogens, Zhang et al. synthesized a type of non-conjugated compound consisting of triphenylmethylamine as the electron-donating unit, Schiff base as the electron acceptor, and saturated ethyl linker (Fig. 31).²⁶² Although these three compounds show the D–A skeleton, the absorption maximum at ≤300 nm suggests their non-conjugated nature. Interestingly, they display dual fluorescent peaks with different intensity ratios under different excitation wavelengths, suggesting multiple emissive channels after photoexcitation. For instance, TPMI-Br shows dual peaks at around 420 nm and 500 nm, respectively, and the long-wavelength peak becomes weaker with the increased excitation wavelength (Fig. 31c). The short-wavelength peak comes from the electron-donating triphenylmethylamine unit, which is almost consistent with the previous reported TPM.¹⁹⁵ Meanwhile, the long-wavelength emission arises from TSCT between the spatially separated donor and acceptor, mediated by nitrogen lone pairs. It is noteworthy that the triphenylmethylamine moiety is not only the fragment possessing TSC but also acts as an electron donor to achieve TSCT, forming the protein-like secondary TSI.²⁶³ Besides, a comparatively weak RTP is also observed at 600 nm in the delayed PL spectra, which belongs to the n–π* transition of the Schiff base. These three emission channels finally result in single-molecule white-light emission of CLgens. This example indicates that the involvement of heteroatom-based units endows CLgens with multiple TSI and various photophysical processes and properties.²⁶⁴

Fig. 31 Clusteroluminescence from triphenylmethane derivatives with through-space charge transfer. (a) The mechanism of conjugated molecules with multiple emissions. (b) The mechanism illustration of non-conjugation molecules with multiple emissions. (c) The excitation-dependent photoluminescence spectra of TPMI-Br and the photograph of its crystals under 365 nm UV irradiation. (d) The hole–electron analysis of three pathways for excited-state relaxation, which produces the while-light emission of TPMI-Br. Reproduced with permission from ref. 262. Copyright 2022, Springer Nature.

Analogous TSCT effects and multi-channel emission have been successfully engineered in polymeric phenolic resin systems through the strategic incorporation of heteroatom-functionalized units. Phenolic resins are widely utilized synthetic resins and the first plastic to be sold commercially, but their photophysical properties are negligible due to their non-emissive feature.^266–268 By learning from the strategies of small-molecule CLgens discussed above, a series of phenolic resin derivatives is constructed by increasing the number of isolated fragments, enhancing electron density, and introducing D–A structures (Fig. 32a).²⁶⁵ Typically, An-MO-PR with methoxy-substituted phenyl rings display an excitation-independent emission at 585 nm with a greatly increased PLQY of 47%, indicating its rigid polymeric chain and strong electronic TSC (Fig. 32b). Besides, with the cyanophenyl moiety as the electron donor, An-CN-PR exhibits redshifted emission maximum of CL at 680 nm due to the formation of TSCT, which can further extend to more than 800 nm (Fig. 32c). Accordingly, these strategies endow the non-emissive phenolic resins with an unexpected luminescence performance with the help of n/π-electron hybrid TSI (Fig. 32d). Apart from the TBC-based emission from isolated phenolic units, they also display long-wavelength emission from through-space locally excited (LE) and TSCT states.

Fig. 32 Clusteroluminescence of non-conjugated phenolic resins. (a) The molecular engineering strategies based on traditional phenolic resins for near-infrared (NIR) emission. (b) The photoluminescence (PL) spectra of An-MO-PR in THF solution with a concentration of 0.1 M under different excitation wavelengths. (c) The PL spectra of An-MO-PR in the solid state under different excitation wavelengths. (d) The proposed potential energy surface for their clusteroluminescence with multiple emission peaks. Reproduced with permission from ref. 265. Copyright 2023, John Wiley & Sons, Inc.

In 2019, Wan et al. reported the polymerization-induced emission (PIE) of hyperbranched polytriphenylmethanols (HPTPM), exploring the importance of polymerization on luminescence.^269,270 By using the Barbier reaction and AB₂-type monomers with one carbonyl group and two chloride groups in one skeleton, they synthesized the p,p′,p″-HPTPM, which showed a non-conjugated nature with an absorption maximum at 248 nm. Apart from its typical AIE effect in THF/water mixtures, this hyperbranched polymer shows reaction time-dependent luminescence, although the monomer is non-emissive (Fig. 33a–c). That is, with the gradually increased polymerization time from 15 minutes to 24 hours, the emission color changes from blue at 466 nm to yellow at 596 nm (Fig. 33d). To investigate the mechanism of the PIE effect, they found that the hyperbranching effect played a crucial role in restricting intramolecular motions, which facilitated the intramolecular TSC and TSCT. Meanwhile, the monomer types, connecting positions, synthetic methods, and polymeric structures (linear or hyperbranched) are verified to influence the luminescence from the PIE approach and this non-conjugated polymeric system.^271–273 Further introducing electron-donating dimethylamino units into the monomer, the non-emissive monomers become polytriphenylmethanols with different structures and luminescent colors. For instance, the linear PDMATPM shows bright green emission after polymerization (Fig. 33e), and the polymeric N,N-dimethyl-triphenylmethanol via reversible addition–fragmentation chain transfer displays blue-light luminescence (Fig. 33g, left).^274,275 Interestingly, the heteroatom-containing alcoholic hydroxyl and dimethylamino groups synergistically make these polymers responsive to acid–base stimuli, demonstrating dynamic luminescence of n/π-electron hybrid systems. After acid stimuli, their appearance color changes from colorless to orange, but the original emission disappears, which is attributed to the structural change from triphenylmethanol to carbocation-quinoid form after acid stimuli (Fig. 33g, right). These examples clearly demonstrate the role of PIE as a versatile chemical strategy for activating luminescence in non-emissive monomers or dynamically modulating the emission of non-conjugated polymeric systems, significantly expanding the toolbox for designing stimulus-responsive luminescent materials.^271,276,277

Fig. 33 Clusteroluminescence of hyperbranched polytriphenylmethanols (HPTPM). (a) The synthetic route of p,p′,p″-HPTPM with polymerization-induced emission effect and photographs of the products with increased polymerization times taken under UV irradiation. (b) The photoluminescence (PL) spectra of p,p′,p″-HPTPM in water/THF mixtures with different water fractions (f_w), concentration = 0.1 mg mL⁻¹. (c) The plot of photoluminescence quantum yields (PLQY) versus f_w spectra of p,p′,p″-HPTPM in water/THF mixtures and photographs of mixtures taken under sunlight (left) and UV irradiation (right). (d) The PL spectra of synthesized p,p′,p″-HPTPM in the solid state, which shows redshifted emission wavelength with the increased polymerization time. Reproduced with permission from ref. 269. Copyright 2019, American Chemical Society. (e) The synthetic route of hyperbranched polytriphenylmethanols with dimethylamino groups (PDMATPM) and photographs of the reactants and products in both solution and solid states taken under sunlight and UV irradiation. (f) The photographs of synthesized PDMATPM upon acid and base stimuli. Reproduced with permission from ref. 274. Copyright 2024, Elsevier. (g) The polymerization of N,N-dimethyl-triphenylmethanol via RAFT and its luminescence and structural response to acid stimuli, verified by NMR spectra. Reproduced with permission from ref. 275. Copyright 2023, American Chemical Society.

Apart from aromatic systems, nonaromatic polyester (PE) is another famous class of non-conjugated polymer that has been widely utilized as plastics, fibers, coatings, etc. Although the intrinsic luminescence properties of poly(ethylene terephthalate) (PET) were reported early in 1969, the basic origin of its luminescence and PL spectra was controversial.^278–280 Some early reported biobased and hydroxyl-terminated hyperbranched PE also show low efficiency and short wavelength.^281,282 In recent years, growing mechanistic insight into TSI and CL has spurred renewed interest in the photophysics of polyester systems. The ester group (a hybrid n/π-electron unit), combined with the structural diversity and synthetic accessibility of PE, makes them an ideal platform for establishing clear structure–property relationships in CL and developing high-performance luminescent polymers.²⁸³ Due to the spatial electronic coupling of n/π electrons and confinement of polymeric chains in the aggregate state after polymerization, the originally forbidden (n,π*) transition of isolated carbonyl and ester groups can be reactivated, promoting the long-wavelength CL of PE. Besides, the comparatively flexible main chains endow PE with the feasibility to form clusters and electronic coupling in the excited state, increasing the efficiency of CL.

Focusing on the structural framework of the PE chain, Chu et al. first reported the influence of main-chain rigidity on the luminescence properties of aliphatic PE. Through copolymerization of propylene epoxide and cyclic anhydrides, they synthesized four types of PE with different emission colors and efficiency dependent on the rigidity of the polymer chains (Fig. 34a).²⁸⁴ From cyclohexane, vinyl, cyclohexene to phenyl moieties, these polymers show gradually increased glass transition temperature (T_g) and decreased segmental mobility from PE1 to PE4 (Fig. 34b). Notably, their emission characteristics do not follow a simple linear trend. The maximum emission wavelength is 448 nm for PE1, 464 nm for PE2, 504 nm for PE3, and 445 nm for PE4, and the corresponding PLQYs are 2.8% for PE1, 12.7% for PE2, 37.9% for PE3, and 0.9% for PE4, respectively. The longest emission wavelength and highest efficiency of PE3 are the results of the aggregate of ester units, which form the TSC of carbonyl groups and promote the (n,π*) transition. These results suggest that the balance of structural flexibility and rigidity is a key point in promoting efficient CL as flexibility allows structural mobility for cluster formation, while the rigid skeleton provides a stable environment for the formed clusters.

Fig. 34 Clusteroluminescence (CL) and the structure–property relationship of aliphatic polyesters. (a) Synthetic routes of polyesters from the ring-opening copolymerization of epoxide and anhydrides. (b) Chemical structures, CL properties of P1, P2, P3, and P4 with gradually increased segmental rigidity. Reproduced with permission from ref. 284. Copyright 2022, John Wiley & Sons, Inc. (c) Chemical structures, CL properties, and luminescence photographs (taken under 365 nm UV irradiation) of a series of polyesters with different main chains and side chains. (d) An illustration diagram of the effect of primary structure and secondary structure on the photophysical properties of aliphatic polyesters. Reproduced with permission from ref. 286. Copyright 2022, American Chemical Society.

On the other hand, the balance of structural flexibility and rigidity could be modulated by side chains.²⁸⁵ By altering the monomer from succinic anhydride to trans-maleic anhydride, cis-MA, and citraconic anhydride, four series of aliphatic PEs with different lengths of side chain originating from epoxides were synthesized (Fig. 34c).²⁸⁶ In the case of the same anhydride units, T_g suggests that increasing the length of the side chain can increase segmental mobility, which affects the primary structure of PE. From the summary of their photophysical properties and luminescence photographs, it is obvious that the PLQY of the PEs shows gradually increased and further decreased values within the same series from top to bottom. For instance, in the PE-5 series with the gradually increased length of the side chain, the PLQY increases from 3.6% to 9.2%, 20.3%, and then declines to 14.0%, 15.3%, and 14.9%. Typically, the highest PLQY of each series belongs to these PEs with a middle length of the side chain (e.g., ethyl or propyl unit). This result is consistent with the previous study and highlights that the equilibrium of structural flexibility and rigidity is a crucial factor in improving CL efficiency.²⁸⁴ Meanwhile, although their wavelengths within the same series are nearly the same, these PEs exhibit redshifted emission wavelengths along with the change from succinic anhydride to trans-maleic anhydride, cis-MA, and citraconic anhydride (from PE-5 to PE-8 series in Fig. 34c). Theoretical simulation proves that the secondary structures of these PEs dominated by the conformation of the main chain modulate the emission wavelength of these PEs. These different anhydrides determine the secondary conformation of PE from a helix to straight and sheet confirmations (Fig. 34d). That is, the spatially separated carbonyl groups form unstable ester clusters in the helical structure, generating short-wavelength emission at 470 nm from the (n,π*) transition. Upon transitioning to a straight conformation, better packing of PE in the aggregate state produces partially stabilized carbonyl clusters and enhances intermolecular TSC, leading to the emergence of long-wavelength emission at 540 nm that coexists with the original peak at 470 nm. When the system changes to sheet-like multifold structures, the carbonyl groups aggregate into denser clusters with shortened O⋯C distances and stronger through-space n–π* interactions, resulting in a complete emission redshift to 570 nm. These studies provide a general picture of how polymeric skeletons regulate their electronic TSC and corresponding photophysical properties.

With the high sensitivity of CL to secondary structures of PE, Li et al. developed a novel strategy to determine the β-transition in PE via CL signals.²⁸⁷ Unlike conventional techniques or earlier fluorescence methods that often depend on external probes and lack sub-glass transition sensitivity, the presented approach monitors the temperature-dependent CL intensity of polyesters. By analyzing the first derivative of emission intensity, both the T_g and the sub-T_g transition (T_β) are unambiguously identified. This study not only demonstrates the utility of CL in real-time and non-destructive polymer dynamics characterization but also reports the first label-free method capable of simultaneously probing T_g and T_β, offering a new approach in material analysis via CL.

Analogous to their role in aromatic systems, heteroatoms significantly influence cluster formation and luminescence behavior in aliphatic PE. With the help of ring-opening copolymerization of epoxide and anhydrides, Liu et al. designed a series of electron-bridged aliphatic PE, containing heteroatoms of oxygen, nitrogen, and sulfur (Fig. 35a).²⁸⁸ As traditional PE without heteroatoms, PE-9 shows dark blue emission at 468 nm, which is attributed to the (n,π*) transition of individual carbonyl groups.²⁸⁹ With the gradual change of the bridged atoms from oxygen to nitrogen, their maximum emission wavelength in the solid state also redshifted from sky blue of PE-10 at 474–558 nm and yellow of PE-11 at 518–596 nm, respectively, indicating the gradually increased strength of TSI and (n,π*) transition. In particular, solid-state PE-12 containing sulfur atoms exhibits NIR CL with the wavelength at 680–740 nm (Fig. 35b). PE-12 is also the first single-component luminogen that can emit full-color light from blue to NIR by simply changing its concentration (from 1 × 10⁻⁵ M to 1 × 10⁻¹ M). These results demonstrate the gradually increased strength of TSI, which follows the order of sulfur > nitrogen > oxygen > carbon in this system, which is controlled by both electronic structure and its van der Waals radius. Nevertheless, the generality of this trend to influence CL efficiency and color still needs to be proved in other systems. In addition, these heteroatoms are proven to facilitate the TSI between neighboring ester groups via the electronic bridging effect, although they are connected by saturated σ bonds. A higher degree of electron-rich property and a larger radius of heteroatoms are beneficial for the formation of closely packed carbonyl clusters and the enhancement of TSI of aliphatic PE, finally promoting NIR CL (Fig. 35c).

Fig. 35 The effect of heteroatoms on clusteroluminescence (CL) of aliphatic polyesters. (a) Synthetic routes of heteroatom-involved polyesters from the ring-opening copolymerization of epoxide and anhydrides. (b) The CL properties and luminescence photographs (taken under 365 nm UV irradiation) of the synthesized polyesters (PE9, PE10, PE11, and PE12). (c) The effect of heteroatoms on cluster formation, CL wavelength of aliphatic polyesters. Reproduced with permission from ref. 288. Copyright 2024, Royal Society of Chemistry.

An intriguing phenomenon has emerged wherein identical polymers synthesized via different routes exhibit markedly distinct luminescence behavior, drawing increasing research interest.^292,293 Beyond the known influence of synthetic methods on secondary structure, the choice of initiator also plays a critical role in CL, which is always neglected due to its very low proportion in the whole polymer. In 2024, Chu et al. suggested that the end-group effect caused by various initiators was an efficient approach to regulating the CL properties of PE.²⁹⁰ First, by utilizing different molar ratios of triethylamine (TEA) (a%) from 0.5% to 5.0% as initiators in the copolymerization of epoxides and cyclic anhydrides, they obtained polyester PE-13 with TEA attached to polymer chain ends (Fig. 36a). For pure PE-13 without the TEA ends, it shows blue CL in the solid state, which is attributed to the (n,π*) transition of easter groups. However, apart from the short-wavelength emission at 474 nm, PE-13-1TEA exhibits an additional long-wavelength emission peak at around 600 nm, resulting in white-light emission. With the gradually increased molar ratios of TEA, the emission intensity ratio of the long-wavelength peak versus the short one increases, which shows redshifted CL and enhanced TSI. Besides, using the other three organic amines (i.e., DBU, mTBD, and TBD), the same CL spectra and luminescence behaviors can be obtained, verifying the importance of amines in realizing red-to-NIR CL of PE. In addition, extending this concept to phosphorus-based initiators (such as tri-n-butylphosphine, tricyclohexylphosphine, triphenylphosphine, tri-tert-butylphosphine, etc.), a series of PE, namely PE-4-P, was synthesized (Fig. 36c).²⁹¹ Compared to amine-capped PE with a colorless appearance and weak emission, phosphine-capped polymers show a dark appearance with a longer-wavelength absorption band, a longer-wavelength CL peak, and a higher luminescence efficiency under the same ratio of initiators. For instance, the phosphine-capped PE with a 0.5% molar ratio of phosphines as the initiator can achieve red CL at 665 nm and a high PLQY of 32%, respectively. The above examples verify that the end-group effect induced by amine/phosphine initiators is a general engineering strategy to tune the photophysical properties of PE.²⁹⁴

Fig. 36 The amine-capped and phosphine-capped effects on clusteroluminescence (CL) of polyesters. (a) The amine-capped effect on CL color of PE-13, which shows redshifted CL with the increased molar ratios (a%) of amine. Photographs of solid-state PE-13 were taken under 365 nm UV irradiation. (b) The illustration diagram of the amine-capped effect on cluster formation, through-space interaction, and CL properties of polyesters. Reproduced with permission from ref. 290. Copyright 2024, Springer Nature. (c) Chemical structures and luminescence photographs (taken under 365 nm UV irradiation) of a series of phosphine-capped PE-4 with a 0.5% molar ratio of phosphines. Reproduced with permission from ref. 291. Copyright 2024, American Chemical Society.

The working mechanism behind the end-group effect is also illustrated through extensive photophysical characterization and dynamical experiments. As shown in Fig. 36b, monomers and amines would form short-range complexes before polymerization, and these host–guest interactions are too weak to produce luminescence. Following polymerization, however, the diverse microenvironments within the polymeric architecture facilitate the formation of stable amine–ester complexes with varying strengths of TSI. The clusters of isolated esters are stabilized and emit blue-color CL, and the amine–ester clusters can form short-range and long-range complexes with enhanced TSI and intra/inter-chain charge transfer, resulting in red and NIR CL with the help of the electron-donating end of amines. Notably, this phenomenon is not exclusive to polymeric systems. Bright red emission has also been observed in small-molecule complexes between imide derivatives and organic bases, providing a mechanistic explanation for the color development and luminescence observed in maleimide-based color reactions.²⁹⁵

Due to the excellent bioactivity, amide-based CL systems have also attracted great interest and have been explored. For example, poly(amidoamine) (PAMAM) dendrimers represent a typical class of amide-based CL system, which has been reported to emit blue emission at around 440 nm.^112,296,297 However, the luminescence origin of PAMAM is controversial. In the early stage, some researchers believed that their intrinsic fluorescence was due to the formation of oxidized tertiary amine species, while others argued that tautomerization of amide bonds to their imidic acid forms and protonation of tertiary amine groups led to their luminescence.^298–300 Alternatively, hydrogen-bonded amide assemblies have also been proposed as a potential source of intrinsic fluorescence of PAMAM dendrimers.¹⁴⁷ Later, Turker et al. suggested the potential involvement of (n,π*) transitions in the CL mechanism without providing clear evidence.¹¹² In 2015, Zhu et al. reported the CL of linear (l) and hyperbranched (hb) PAMAMs synthesized by Michael-type polycondensation-addition (Fig. 37).²⁹⁶ Although these two PAMAMs show the typical CTE phenomenon, the emission intensity of linear PAMAM increases more rapidly than that of hyperbranched PAMAM with the increased concentration from 0.1 mg mL⁻¹ to 20 mg mL⁻¹, likely due to the more compact aggregated structure of linear PAMAM (Fig. 37a). Meanwhile, both linear and hyperbranched PAMAMs form solid films on quartz substrates that exhibit strong excitation-dependent CL. The PL spectra of these films are redshifted with the increased excitation wavelength, but with varying magnitude, indicating the influence of structural topology on electronic TSI and corresponding CL. Beyond spatial delocalization of n and π electrons, accumulating evidence highlights structural rigidification as a key factor that enhances the CL of clusteroluminogens by suppressing nonradiative channels. Despite these insights, the underlying CL mechanism in amide-based systems is still not well resolved, especially the origin of their characteristic blue emission.

Fig. 37 Clusteroluminescence of poly(amido amine)s (PAMAM). (a) The synthetic route of linear l-PAMAM and hyperbranched hb-PAMAM. (b) and (c) The excitation-dependent photoluminescence spectra of l-PAMAM and hb-PAMAM solid films under excitation wavelengths of 380 nm, 460 nm, and 530 nm, respectively. Photographs were taken under irradiation of UV (top), blue (middle), and green (bottom) lights. Reproduced with permission from ref. 296. Copyright 2015, Springer Nature.

To provide a deeper understanding of the emitting sources, Zhang et al. carefully studied three non-conjugated polypeptides, namely poly(γ-benzyl-L-glutamate) (PBLG), poly(L-glutamic acid) (PLGA), and its sodium salt (PLGA-Na). All three polypeptides exhibit excitation-independent blue clusteroluminescence at 440 nm, albeit with differing emission efficiencies (Fig. 38a).²⁸⁹ Variable temperature spectral tests of these polypeptides and model small molecules of N-acetyl-L-alanine and N,N-diethylacetamide with the identical 440 nm emission strongly verify that their emission at 440 nm comes from (n,π*) transition of an individual amide instead of amide clusters. Theoretical calculations based on the lowest singlet state also suggest that the electronic transition of the (n,π*) transition of individual amides with comparatively low oscillator strength (Fig. 38b). Accordingly, they also highlighted that intra- and inter-chain interactions induced by both polymerization and clusterization are critical factors to activate the conventionally forbidden (n,π*) transition and enhance the CL efficiency of those non-conjugated polypeptides. This mechanistic hypothesis is accepted by other researchers and verified by other examples, establishing this transition of amide units as a general source of blue emission in non-conjugated peptide and polymeric systems.^301–304

Fig. 38 Clusteroluminescence of polypeptides. (a) Chemical structure, photoluminescence spectra of poly(L-glutamic acid). Photographs were taken in the solution and crystalline state under 365 nm UV irradiation, respectively. (b) The typical electron transition of the carbonyl group, which is responsible for the blue emission observed in polypeptides. Reproduced with permission from ref. 289. Copyright 2022, Elsevier.

In 2017, Ye et al. systematically studied the photophysical properties of oligopeptides and polypeptides derived from the natural amino acids of L-alanine (ALA), L-valine (VAL), and L-isoleucine (ILE).³⁰⁵ As shown in Fig. 39a, while the individual amino acids are non-emissive in the solid state, their corresponding oligopeptides (namely, OALA, OVAL, and OILE) exhibit blue emission at about 400 nm with a moderate PLQY of 20% under UV irradiation. Besides, using a high-temperature dehydration method and solid-phase peptide synthesis, two ALA-based polypeptides, PALA-HT and PALA-SS, are synthesized. These two polypeptides show cyanic color with a maximum emission peak at 440 nm, which should come from the (n,π*) transition of amide groups as discussed above. Theoretically, the luminescent species of these polypeptides form through hydrogen-bond-mediated clustering of amide groups, where any disruption of this ordered assembly would directly modulate their photophysical properties. Interestingly, the piezo-enhanced fluorescence is observed in the poly-L-proline (PPRO) system, which is a fluffy fibrous polymer with negligible emission (Fig. 39b). Initial compression transforms the fluffy PPRO into a thin film with markedly increased fluorescence intensity, yielding an increased brightness index of 4.5 compared to the pristine sample. Subsequent compression under a higher pressure further boosts the brightness index of PPRO to 6.3. With the help of a traditional Chinese stamping, a fluorescent micropattern can be further created, where the compressed region shows substantially brighter emission than uncompressed areas. This piezochromic behavior aligns with the CTE phenomenon that the dense packing of amide clusters and reduced inter-amide distances under high pressures promote the (n,π*) transition and restrict molecular motions, leading to turn-on CL.

Fig. 39 Clusteroluminescence of oligopeptides and polypeptides. (a) Chemical structure of ALA, VAL, and ILE, and photographs of amino acids, oligopeptides, and polypeptides taken under the 365 nm UV lamp (left). Photoluminescence spectra of oligopeptides and polypeptides. (b) Illustration of poly-L-proline (PPRO) in response to external pressure under daylight and UV light. Reproduced with permission from ref. 305. Copyright 2017, Royal Society of Chemistry.

Nylon polymers, characterized by their regular alternation of polyethylene segments and amide groups, typically adopt a well-ordered and rigid solid-state structure stabilized by extensive hydrogen bonding.³⁰⁶ This structural regularity makes them ideal platforms for investigating how molecular arrangement governs CL.^301,307,308 In 2020, Yuan et al. reported the influence of the antiparallel and parallel molecular arrangement of polyamide-6 (PA-6) on their CL.¹¹⁹ Through the precipitation of formic acid-dissolved PA-6 using THF and DMF as nonsolvents, they successfully obtain two solid samples, namely PA-6a and PA-6b, which show different fractions of α and γ form aggregates (Fig. 40a). Their dilute solutions (0.001 wt%) are non-emissive upon irradiation, but gradually enhanced visible emission could be observed with the concentration increases from 1 wt% % to 15 wt%. Differently, under 365 nm UV light, concentrated PA-6a/FA and PA-6b/FA solutions (15 wt%) display green and blue emissions, respectively. The maximum emission wavelength of these two solutions is highly dependent on the excitation. Specifically, PA-6a/FA shows two obvious emission peaks at 404 nm and 490 nm under 312 nm irradiation, while only the peak at 490 nm could be detected under 400 nm irradiation (Fig. 40b). On the other hand, the maximum emission wavelength of PA-6a/FA redshifts from 402 nm to 524 nm along with the gradually increased excitation wavelength (Fig. 40c). These spectral features unambiguously indicate the coexistence of multiple emissive species, arising from distinct molecular arrangements induced by different non-solvents during precipitation. X-ray diffraction measurements also verify that the fractions of the α form of PA-6a and PA-6b are 78.4 and 73.3%, respectively. In addition, through different preparation processes, the cast film and electrospun film of PA-6 exhibit totally different molecular arrangements.³⁰⁹ The former shows the α form fraction of 97.9% while the latter is dominated by the γ form of 94.2%. Correspondingly, the antiparallel molecular packing in the α-form creates a denser structure compared to the parallel arrangement in the γ-form, resulting in a higher PLQY of the cast film (8.7%) than that of the electrospun film (3.8%), directly linking tighter molecular arrangement to enhanced luminescence efficiency.

Fig. 40 Clusteroluminescence of polyamide. (a) The synthetic route of PA-6a and PA-6b with different molecular arrangements and their luminescence photographs in formic acid (FA) solution. (b) and (c) The excitation-dependent photoluminescence spectra of 15 wt% (b) PA-6a/FA and (c) PA-6b/FA solutions. Reproduced with permission from ref. 119. Copyright 2020, American Chemical Society.

Apart from these typical and widely reported polyamides and polyesters,³¹¹ some carbonyl-based aliphatic polymers also show similar CL behavior. In 2022, Zhang et al. reported similar CL from poly(monothiocarbonate) (PMTC) and polycarbonate (PC).³¹⁰ These two polymers containing carbonyl groups on the main chains are synthesized via ring-opening reactions of epoxide with carbonyl sulfide and carbon dioxide, respectively (Fig. 41a).³¹² Both polymers display concentration-dependent emission in DMF solution, with intensity progressively increasing at higher concentrations (Fig. 41b and c). Interestingly, despite the substitution of sulfur in PMTC with oxygen in PC, the two polymers exhibit nearly identical emission maxima at 440 nm and lack excitation dependence, which is consistent with the characteristic (n,π*) transition of carbonyl groups as established in earlier systems. Although the quantum yield of PC (Φ = 4.5%) is moderately higher than that of PMTC (Φ = 1.7%), the overall brightness of both systems remains considerably lower than that of previously reported PE analogs. This diminished performance may be attributed to the high conformational flexibility of the PMTC and PC backbones.

Fig. 41 Clusteroluminescence of poly(monothiocarbonate) (PMTC) and polycarbonate (PC). (a) The synthetic route and CL properties of PMTC and PC. (b) and (c) The photoluminescence spectra of (b) PMTC and (c) PC in DMF solutions with different concentrations. Reproduced with permission from ref. 310. Copyright 2022, John Wiley & Sons, Inc.

The 440 nm emission of CL produced by the (n,π*) transition of carbonyl groups is influenced by both electronic structures and conformational flexibility. Accordingly, relocating the carbonyl unit from the polymer backbone to the side chain offers a viable route to modulate CL efficiency.³¹³ In 2023, Liu et al. reported the CL of poly(vinylene carbonate) (PVC) and poly(vinylethylene carbonate) (PVEC), where the five-membered cyclic carbonates are the active groups for electronic TSI.³¹⁴ The large steric hindrance and strong polarity endow these two polymers with higher T_g of 130.4 °C and 162.5 °C, respectively, indicating their rigid polymeric skeletons. Similar to the previous examples, carbonyl-based PVC and PVEC also show cyan-color CL in both solid state and concentrated solutions (Fig. 42a). It is worth noting that the molecular weight of PVC can be well controlled by polymerization approaches and the ratio of initiator and monomer, resulting in tunable molecular weight from 2.5 to 1436 kDa for comparison of their CL properties. As shown in Fig. 42b, the plot of emission efficiency versus degree of polymerization (DP) exhibits four stages, including cluster formation, rapid growth, decelerated growth, and the saturation stage. After the critical point, the emission efficiency first rapidly increases, then slowly increases, and eventually saturates along with the gradually increased DP, suggesting that the CL properties can only be modulated within a certain range of molecular weight. In addition, these two polymers show excitation-dependent emission and the main peak centers at around 440 nm (Fig. 42c and d). Although the authors claimed that the formed intramolecular and intermolecular cluster of oxygens were responsible for such emission (Fig. 42e), it is now clear that the 440 nm CL of PVC and PVEC should come from carbonyl groups, which could be finely regulated by chain conformation and electronic structures in the clustered state. Benefiting from its high molecular weight, PVC could be further fabricated into fibers, microspheres, or sponges via different methods, which may widen new application prospects for such CL materials in the future.

Fig. 42 Clusteroluminescence of poly(cyclic carbonate)s. (a) The synthetic route of poly(vinylene carbonate) (PVC) and poly(vinylethylene carbonate) (PVEC). Photographs in both solution and solid state were taken under UV irradiation. (b) The plot of emission efficiency versus average degree of polymerization (DP) of PVC in DMSO solutions (concentration = 1 × 10⁻¹ M). (c) and (d) The excitation-dependent photoluminescence spectra of solid-state (c) PVC and (d) PVEC. (e) The schematic illustration of the through-space interactions of electrons in PVC. Reproduced with permission from ref. 314. Copyright 2023, Royal Society of Chemistry.

Following the CL of PVC, poly(hydroxyurethane) (PHU) was reported as another type of n/π-electron hybrid system with excellent luminescence properties, which could be simply synthesized via the straightforward ring-opening of polyfunctional cyclic carbonate with amine without the use of toxic isocyanates.^317–321 In 2024, Xu et al. studied the influence of the strength of hydrogen bonding and steric hindrance on the CL properties of poly(hydroxyurethane) with hexylamine, ethanolamine, and cyclohexylamine as the functional amines, respectively (Fig. 43a).³¹⁵ All of them show a special PL spectra in both solution and solid states, with initially excitation-independent emission and subsequently excitation-dependent emission. For example, solid-state PHNCs-C exhibit excitation-independent emission at 500 nm with the increased excitation wavelength from 320 nm to 400 nm, while its emission peak gradually redshifts from 500 nm to more than 650 nm with further increase in excitation wavelength (Fig. 43b). The shorter-wavelength emission is assigned to the intrinsic (n,π*) transition of carbamate group, while the longer-wavelength band arises from TSI of hydroxyurethane groups in the aggregate state, where hydrogen-bonding interactions and polymerization play key roles.²⁸⁹ In particular, due to the bulky steric hindrance of the hexyl group and the resulting largest FWHM, PHNCs-C emits white-light emission under a 365 nm UV lamp with Commission International de L’Eclairage (CIE) coordinates of (0.32, 0.36). In contrast, the solid PHNCs-H with flexible alkyl chain and PHNCs-E with hydrogen bonds produce blue fluorescence under 365 nm irradiation (Fig. 43c and d). It is noted that the solid-state PLQYs of these three polymers (1.89% for PHNCs-C, 6.3% for PHNCs-H, and 3.9% for PHNCs-E) are lower than in the solution state, owing to the intense reabsorption that is often observed in CL materials.

Fig. 43 Clusteroluminescence of poly(hydroxyurethane)s. (a) The synthetic route of PHNCs-H, PHNCs-C, and PHNCs-E. (b)–(d) The excitation-dependent photoluminescence spectra of solid-state (b) PHNCs-H, (c) PHNCs-C, and (d) PHNCs-E. Photographs were taken under daylight (left) and UV irradiation (right). Reproduced with permission from ref. 315. Copyright 2024, Elsevier. (e) The synthetic route of TPTE-based poly(hydroxyurethane) and its application for white light-emitting diodes. Reproduced with permission from ref. 316. Copyright 2020, John Wiley & Sons, Inc.

Using a similar synthetic approach, Liu et al. developed a white-light-emitting microsphere based on PHU from the crosslinking of trimethylolpropane tri(cyclic carbonate) ether and 1,6-hexanediamine (Fig. 43e).³¹⁶ By regulating the synthetic temperature, the crosslinked PHUs show gradually redshifted CL due to the different crosslinking degree and hydrogen-bonding strength, resulting in white-light-emitting microspheres with PLQYs of 6.40% at 180 °C and 6.11% at 200 °C, respectively. In addition, with the help of the phosphor-converted architecture of light-emitting diodes (LEDs), these two PHU-based microspheres are further fabricated as white-light LEDs by combining them with a 365 nm chip. Apart from the good CIE coordinates of (0.346, 0.374), these fabricated LEDs also achieve excellent color rendering of 95 and corresponding CCT values of 5025 K, displaying comparable properties as traditional white LEDs prepared from RGB-based phosphors.³²²

Analogous to cyclic carbonates in chemical functionality, simple heteroatom-containing succinimide and maleic anhydride are other sources for CL. In 2022, He et al. reported the bright blue-green CL from aromatic-free maleimide and succinimide (Fig. 44a).³²³ Different from other carbonyl-based systems that rely on a polymeric skeleton, these examples also verify the (n,π*) transition of carbonyl groups with the help of stabilized cluster excitons of small molecules in the condensed state. Besides, with the introduction of alkyl thiols to their molecular skeleton, the emission wavelength further redshifts to around 510 nm with a higher emission efficiency (Fig. 44b), which is closely associated with the dominant (n,π*) transition. On the other hand, maleic anhydride was utilized as a monomer with vinyl acetate to synthesize a well-known non-conjugated alternating polymer with pure oxygen atoms, poly[(maleic anhydride)-alt-(vinyl acetate)] (PMV), which shows special solvent-dependent CL.^118,324 In dilute THF solution (concentration = 5 mM), PMV is colorless and shows no obvious absorption at more than 300 nm. Correspondingly, it shows blue emission at 440 nm, which should be related to the emission from the cluster of carbonyl groups. Interestingly, it shows a new and intense absorption band at 560 nm in electron-rich N-methyl-2-pyrroldone (NMP), dimethyl sulfoxide (DMSO), and dimethylformamide (DMF), and its emission color also changes to green in NMP and DMSO and even red in DMF with a new emission peak at around 610 nm. It is found that the formed polymer/solvent complex is responsible for the longer-wavelength emission observed in electron-rich solvents, which is similar to the end-group effect discussed before.^290,291 Following the strong CL from PMV, Qiao et al. further fabricated PMV-derived polymers and reported their photophysical properties.³²⁵ By dropping PMV/alcohol suspension into strong alkaline (i.e., LiOH, NaOH, and KOH), they obtain these charged and water-soluble PMV-derived polymers that can emit bright blue CL in both concentrated solutions and the solid state. Among them, LiOH-treated and KOH-treated polymers even show the highest PLQY of 87% and 76% in the solid state compared to other non-conjugated CL polymers, which may be related to the rigidity-aggregation environment. In addition, these polymers even show emission-color transformation from the blue to red-light region around 600–670 nm upon heating treatment at 170 °C. These two examples clearly prove the ability and variability of maleic anhydride-based polymers for CL.

Fig. 44 Clusteroluminescence of succinimide and poly[(maleic anhydride)-alt-(vinyl acetate)] (PMV). (a) The chemical structure and fluorescent photographs of succinimide and its derivatives. (b) The photoluminescence (PL) spectra of succinimide and its derivatives. Reproduced with permission from ref. 323. Copyright 2021, John Wiley & Sons, Inc. (c) The synthetic route of PMV. (d) Absorption spectra and photographs of PMV in different solvents taken under daylight. (e) PL spectra and photographs of PMV in different solvents taken under 365 nm UV irradiation. Reproduced with permission from ref. 118. Copyright 2014, American Chemical Society.

Through a rough comparison of pure n-electron systems, pure π-electron systems, and n/π-electron hybrid systems, it is obvious that n/π-electron hybrid architectures show the best CL performance and a wider regulating range. Specifically, some non-conjugated small molecules (e.g., pyrimidine-based o-2Md)⁴⁷ and polymers (e.g., modified phenolic resins of An-MO-PR)²⁶⁵ can achieve red and NIR emission, which cannot be observed in the other two systems. As illustrated by the above examples, the CL properties of these non-conjugated materials are mainly dependent on the strength of the electronic TSI and stability of molecular skeletons upon photoexcitation. Accordingly, the excellent performance of some n/π-electron hybrid systems is beneficial from the synergistic combination of electron-rich heteroatoms that largely increase the strength of TSI and the π-electron subunits (such as isolated double bonds or phenyl rings) that provide suitable rigidity for the whole skeleton.

With a greatly increased number of non-conjugated materials with CL properties and a deeper understanding of their working mechanisms, some regulation strategies of CL are summarized as general guidance for materials design and synthesis. Although the detailed structure–property relationship may have been discussed before, it is too scattered. Therefore, a schematic summary is also provided here to give researchers a more systematic outline of these manipulation approaches (Fig. 45). (1) Conformational regulation. The spatial orientation of n- and π-electron orbitals fundamentally governs the TSI of electrons in CLgens, as illustrated by the odd–even effect of multi-aryl-substituted alkanes with different face-to-face and staggered conformations.^205,212 Hence, conformational regulation is a basic method to regulate the properties of CLgens. (2) Electronic regulation. Consistent with fundamental photophysical principles, enhanced electron density strengthens conjugation effects. Correspondingly, electronic regulation, especially introducing electron-donating substituents, would be a useful approach to promote TSI efficiency.¹⁹⁵ (3) TSCT effect. In contrast to locally excited TSC in CLgens, TSCT featuring complete D–A separation drives larger redshifts and heightened environmental sensitivity of the emission.^104,262,326 (4) Skeleton engineering. Molecular skeletons, especially for non-conjugated polymers with flexible alkyl chains, play an essential role in promoting and stabilizing the TSI for CL. Hence, balancing the rigidity and flexibility via main-chain and side-chain subunits is crucial to achieve long-wavelength and high PLQYs of CL.^284,286,327 (5) Heteroatom incorporation. Incorporation of heteroatoms introduces electron-rich lone pairs that promote cluster formation with stronger TSI, yielding a pronounced redshift in CL relative to phenyl-based systems. Heteroatom effects should be viewed as pivotal for overcoming the current limitations of CLgens, as exemplified by a pyrimidine-based system that delivers near-infrared cluster luminescence and a photoluminescence quantum yield (PLQY) of 25%.³²⁸ (6) End-group/complex effect. End groups of polymers induced by initiators and complex interactions between CLgens and solvents have been proved to form TSCT upon photoexcitation, promoting a redshift of emission wavelength to the red and NIR regions.^118,290,291 (7) Hydrogen-bond engineering. Hydrogen bonding is a double-edged sword. It can stabilize the formed TSI through the hydrogen bonding network, but it may also hinder the formation of effective TSI between lone pairs of electrons.^257,329 Therefore, balancing the number and position of hydrogen bonds in non-conjugated molecules and hydrogen bonds in the external environment is also important for regulating CL. (8) Thermal post-modification. Modifying the chemical structure of non-conjugated materials by thermal treatment has also been reported as a means of tuning their CL properties, which can effectively enhance the luminescence efficiency and red-shift the wavelength.^325,330 Due to the small number of examples, the working mechanism of the thermal treatment is still unclear and needs to be studied.

Fig. 45 Schematic summary of regulation strategies of clusteroluminescence from non-conjugated small molecules and polymers. (a) Conformational regulation of TSC. Reproduced with permission from ref. 205. Copyright 2023, American Chemical Society. (b) Electronic regulation of TSC. Reproduced with permission from ref. 195. Copyright 2021, American Chemical Society. (c) Through-space charge transfer with primary and secondary levels. (d) Balanced flexibility and rigidity of polymer skeletons. (e) Heteroatom effect for clusteroluminescence. Reproduced with permission from ref. 288. Copyright 2024, Royal Society of Chemistry. (f) End-group and complex effect for clusteroluminescence. (g) Hydrogen-bonding effect for clusteroluminescence. Reproduced with permission from ref. 329. Copyright 2024, Springer Nature. (h) Thermal treatment for intensified and redshifted clusteroluminescence.

3.4. Potential applications of clusteroluminescence materials

The growing mechanistic understanding of CL and the expanding library of CL-active materials have accelerated the exploration of their practical applications. The unique advantages of non-conjugated structures, unique working mechanisms, and diverse photophysical properties, which differ fundamentally from those of conventional conjugated luminogens, enable them to fulfill specialized application requirements that is comparable to conventional conjugated luminogens, or cannot be achieved by these conjugated systems (Fig. 46). Herein, a general introduction to the potential applications of current CL materials is necessary to draw the whole picture of TSI and CL, as well as promote the development of these newly emerging luminescent materials.

Fig. 46 Schematic summary of potential applications of clusteroluminescence materials. (a) Multichannel and two-photon bioimaging. Reproduced with permission from ref. 167. Copyright 2021, Springer Nature and ref. 331. Copyright 2025, John Wiley & Sons, Inc. (b) Biological process monitoring from normal protein to amyloid fibrils. Reproduced with permission from ref. 332. Copyright 2023, Springer Nature. (c) Non-doped luminescent fiber based on non-conjugated polyester. Reproduced with permission from ref. 291. Copyright 2024, American Chemical Society. (d) Encryption and anticounterfeiting based on CLgens. Reproduced with permission from ref. 120. Copyright 2018, American Chemical Society and ref. 205. Copyright 2023, American Chemical Society. (e) Non-conjugated fluorescent hydrogen for 3D printing. Reproduced with permission from ref. 329. Copyright 2024, Springer Nature. (f) Chemosensor for ion detection and humidity visualization. Reproduced with permission from ref. 284. Copyright 2022, John Wiley & Sons, Inc. and ref. 340. Copyright 2024, John Wiley & Sons, Inc.

A key advantage of CLgens lies in their exceptional biocompatibility owing to non-conjugated molecular structures, making them particularly suitable as a novel class of biological probes (Fig. 46a). For bioimaging, the oxygen/sulfur-containing dendrimer demonstrates excitation wavelength-dependent emission behavior with an exceptionally broad excitation window spanning 450–600 nm, which has been successfully employed for multi-channel bioimaging of 4T1 cancer cells.¹⁶⁷ However, it is still challenging for conventional conjugated probes with excitation-independent emission to achieve multi-channel bioimaging. In addition, some CLgens can also exhibit nonlinear optical properties (i.e., two-photon excited CL) as observed in traditional conjugated luminogens. For instance, small-molecule CLgens based on diarylmethane derivatives have been engineered into nanoparticles exhibiting two-photon excited CL properties.³³¹ These nanoparticles have been effectively utilized for in vivo angiography in mouse models, providing high-resolution three-dimensional reconstruction images of blood vasculature with both acceptable tissue penetration and excellent biocompatibility. More significantly, certain native biomacromolecules have been discovered to function as intrinsic CLgens with the help of their non-covalent interaction networks that form during biological molecular aggregation processes, enabling the visualization of biological processes through their inherent CL signals without requiring external probe labeling. A notable example is using the intrinsic CL properties of hen egg white lysozyme to monitor its complete aggregation pathway from native protein state through protofibril intermediates to mature amyloid fibrils (Fig. 46b).³³² Remarkably, this label-free clusteroluminescence (CL) assay exhibits higher sensitivity than conventional staining with the commercial amyloid dye Thioflavin T (ThT).

In addition, CL materials with non-conjugated architectures offer major advantages in synthetic accessibility and processing flexibility. They can be built from simple, inexpensive precursors in fewer steps under mild conditions, tolerate diverse functional groups, and dissolve or melt readily for scalable solution or melt processing (e.g., spin-coating, extrusion, and printing). These modular backbones also enable precise performance control via end-group engineering, heteroatom content, and supramolecular interactions, yielding excellent manufacturability and tunable optical outputs. Hence, they can be fabricated as fibers and films with intrinsic luminescence properties. A representative example is the straightforward fabrication of large-scale, bright-red CL-emitting polymer fibers through manual extrusion-spinning of a phosphine-capped polyester system (Fig. 46c).²⁹¹ Recently, researchers also synthesized cross-linked polymeric films via interfacial polymerization of triyne and water, and achieved precise control over their luminescent colors with a gradual bathochromic shift from blue to yellow and quantum efficiency enhancement from 31.1% to 45.7% via modulating the interfacial polymerization duration.³³³ These free-standing polymeric films with CL properties show a Janus character of two sides and can work as a vapor-responsive soft actuator. In addition, Wang and Jiang et al. developed a red-fluorescent hydrogel system based on poly(N-acryloylsemicarbazide), which has been successfully implemented for high-resolution additive manufacturing using 3D printing technology.³²⁹ As demonstrated in Fig. 46e, this red-fluorescent hydrogel could be fabricated as sophisticated architectures, such as the Eiffel Tower. Furthermore, they also manufacture a bioinspired fluorescent jellyfish-mimetic swimming robot that operates through a centrally positioned servo motor that drives the rhythmic expansion and contraction of support rods, which in turn induce coordinated undulatory movements of the hydrogel tentacles in the aqueous environment. This movement of the fabricated jellyfish-mimetic robot could be visualized by its intrinsic red CL under water.³³⁴

Due to the involvement of n/π electrons in their molecular skeleton that can promote the efficiency of ISC of luminogens, many CL materials show both fluorescence and RTP at the same time.³³⁵ Accordingly, they can be utilized for time-dependent information encryption and anticounterfeiting. For instance, Yuan et al. reported the CL and RTP properties of the plant-derived sodium alginate and utilized them as a link for the painting of a flower (Fig. 46d).^120,336 Blue CL is observed under continuous UV irradiation. However, after the UV source is turned off, the green phosphorescence gradually diminishes and eventually disappears. On the other hand, the CL properties of these materials exhibit a strong dependence on non-covalent TSI of electrons, rendering them exceptionally sensitive to external stimuli as manifested by detectable changes in their CL signals. Notably, TSI-based CLgens exhibit far higher environmental sensitivity than traditional TBC-type conjugated luminophores, because their emission stems from non-covalent electronic interactions and cluster packing that respond acutely to changes in polarity, hydrogen-bonding strength, viscosity, temperature, and pressure. This enhanced responsiveness enables novel applications in dynamic optical materials. For instance, Xiong et al. have successfully integrated two structurally similar luminogens (TPB and s-TPB) that share identical emission colors but display distinct solvent responsiveness into a single platform.²⁰⁵ This system has been effectively employed for advanced information encryption and dynamic patterning applications through controlled solvent fuming. Besides, CL materials could be utilized as sensitive sensors (Fig. 46f).^337–339 As a chemosensor, we once reported the CL properties of PE as discussed above and demonstrated PE as an exceptionally selective and sensitive detector with a rapid response time (about 1.2 min), low detection limit (0.78 µM), as well as high specificity for iron ions.²⁸⁴ In addition, due to the involvement of hydrogen bonding interactions, some CL materials display remarkable sensitivity to water and can visualize environmental humidity via luminescence. A representative example includes an amphiphilic polymer functionalized with water-responsive carboxylate groups, which serves as an effective humidity sensor by exhibiting a bathochromic shift in its CL emission from 508 nm to 534 nm as the relative humidity decreases from 90% to 10%.³⁴⁰

4. Through-space interactions for thermally activated delayed fluorescence

Thermally activated delayed fluorescence (TADF) refers to a luminescence process where molecules harness thermal energy to convert the triplet excited state to the singlet excited state via reverse intersystem crossing (RISC), resulting in delayed fluorescence.^341–345 Due to the RISC ability to realize the utilization of both singlet and triplet excitons, TADF materials can realize the utilization of both singlet and triplet excitons, enabling theoretically 100% internal quantum efficiency. Renowned for their high brightness, long operational lifespans, and ability to operate at low voltages, TADF materials are ideally suited for applications such as organic light-emitting diodes, flexible displays, and energy-efficient lighting systems. TADF is governed by the energy gap (ΔE_ST) between the S₁ and T₁ states as described by the energy gap law, where a minimal ΔE_ST is crucial to facilitate rapid RISC.^346–348 Therefore, one conventional strategy to minimize ΔE_ST relies on TBC to achieve electronic separation of HOMO and LUMO, employing molecular architectures with spatially separated donor and acceptor units connected by covalent linkages. Unlike traditional through-bond charge transfer mechanisms, TSCT with non-covalently linked donor and acceptor units facilitates complete charge separation and enables an exceptionally small ΔE_ST, while simultaneously maintaining a relatively high radiative decay rate and overall PLQY.^349–356

The advantages of TSCT are particularly pronounced at low drive voltages, where energy efficiency is paramount, because the spatially separated D–A states enable facile charge injection and radiative recombination under weak electric fields, reducing power consumption and thermal load.^357–360 For instance, in OLED displays and lighting technologies, TSCT-based TADF materials enable devices to achieve high brightness and longevity while consuming less power. Furthermore, TSCT-based TADF materials exhibit superior energy efficiency, stability, and durability compared to their traditional counterparts. These properties make them compelling candidates for next-generation optoelectronics, which demand not only high efficiency and stability but also low-energy, solution-based fabrication and benign chemistries for improved sustainability. By significantly reducing energy costs while delivering enhanced performance, these materials align perfectly with the global push toward sustainable and advanced technologies. Beyond OLEDs and lighting, TSCT-based TADF materials show promise for a wide range of applications, including sensors,^361–363 solar cells,³⁶⁴ and even biomedical devices.^365–367 Their versatility and superior performance characteristics position them as a transformative technology across multiple industries. As research continues to advance, TSCT-based TADF materials are poised to play a pivotal role in shaping the future of optoelectronics, offering a compelling combination of efficiency, durability, and sustainability.

Compared with the developed TBC-based theories, TSCT is a newly proposed strategy to realize the modulation of TADF properties. Although TSCT has been utilized in systems with conjugated donor and acceptor subunits and CL materials as discussed above, it has not been applied in totally non-conjugated systems. Besides, the structure–property relationship and manipulating strategies of TSCT-regulated TADF materials are ambiguous. Thus, a summary of the intrinsic mechanisms and modulation principles of TSCT will undoubtedly offer both fundamental theoretical insights and practical value for the development of TADF materials, especially for non-conjugated skeletons. In this section, we introduce several small-molecule and polymeric TADF systems, focusing on how TSCT governs their photophysics and on practical strategies to manipulate performance.

4.1. Small molecules

Small-molecule TADF luminophores have risen to prominence owing to their concurrent advantages in efficiency, solution processability, film-forming quality, and molecular design freedom, alongside remarkable thermal stability that supports high-temperature fabrication and operation. This kind of property suite directly aligns with the needs of next-generation OLEDs, solid-state lighting, and sensing platforms seeking high brightness at low bias and durable performance. Furthermore, their well-specified molecular scaffolds constitute a clean testbed for probing TSCT, facilitating quantitative links between microstructure and macroscopic TADF metrics such as color purity, efficiency roll-off, and lifetime.^369–374

A representative study by Liao et al. demonstrated the design and performance of space-confined small-molecule TADF emitters with TSTC, showcasing their improved efficiency and potential for next-generation optoelectronics.³⁶⁸ Compared with randomly mixing donors and acceptors or utilizing a flexible skeleton to link them, a fixed linker with close distance and confined rotation of subunits are regarded as the general platform to construct TADF emitters with high PLQY (Fig. 47a). Following this consideration, they construct a series of small molecules (namely, DM-B, DM-Bm, DM-G, DM-X, and DM-Z) by systematically varying the rigidity of the linker and the distance between 10-phenyl-9,10-dihydroacridine as the donor and 2,4,6-triphenyl-1,3,5-triazine as the acceptor motifs (Fig. 47b). The absorption spectra, measured in DCM solution (1 × 10⁻⁵ M), reveal strong intramolecular CT transitions with peaks at 278 nm of DM-B, 274 nm of DM-Bm, and 286 nm of DM-G (Fig. 47c). Interestingly, fluorescence and phosphorescence spectra of these three compounds at 77 K are almost overlapped with each other, suggesting their same energy levels of singlet and triplet states. Accordingly, the ΔE_ST are calculated to be 0.17 eV, 0.08 eV, and 0.11 eV for DM-B, DM-Bm, and DM-G, respectively. In addition, single-crystal X-ray structures confirm co-facial arrangements of the donor and acceptor units with short distances of 3.16 Å (DM-B), 3.06 Å (DM-Bm), and 2.83 Å (DM-G), promoting efficient coupling for charge transfer and exciplexes with excellent emission properties. As a result, DM-B, DM-Bm, and DM-G achieve a high PLQY of 78%, 69%, and 51% in toluene solution, respectively. As a comparison, although they show similar absorption and emission spectra, the flexible DM-X and electronically less coupled DM-Z display a much lower PLQE of only 12% and 6% in toluene. By further solidifying these compounds in the bis[2-(diphenylphosphino)phenyl]ether oxide matrix to restrict intramolecular motions, they all exhibit different degrees of PLQY improvement. Notably, the PLQE of DM-B is near unity of 96%, which is rarely observed even on covalently bonded D–A systems.

Fig. 47 Thermally activated delayed fluorescence from space-confined small-molecule emitters. (a) Illustration of models with different donor/acceptor (D/A). (b) Chemical structures of DM-B, DM-Bm, DM-G, DM-X, and DM-Z, along with their molecular counterparts (DMTPA donor and TPZ acceptor). (c) Normalized absorption (room temperature, 1 × 10⁻⁵ M in DCM), fluorescence, and phosphorescence emission spectra (77 K, 1 × 10⁻⁵ M in toluene) of the rigid exciplex emitters. Insets show single-crystal X-ray structures of DM-B, DM-Bm, and DM-G. (d) EL spectra of different emitters (DM-B, DM-Bm, DM-G, DM-X, DM-Z) with varying doping ratios of DM-B; EQE versus luminance characteristics; power efficiency versus luminance characteristics. The color legend is consistent with panel a. Reproduced with permission from ref. 368. Copyright 2020, Springer Nature.

Inspired by their excellent photophysical behaviors, these emitters are further fabricated into OLED devices to check electroluminescence (EL) performance. The EL spectra show emission peaks ranging from 488 nm (DM-B) to 500 nm (DM-Bm, DM-G), with further red-shifting for DM-X and DM-Z, reflecting variations in molecular structure and D/A distances. The DM-B-based device achieves an external quantum efficiency (EQE) of 27.4% at 67 cd m⁻², with minimal roll-off to 24.4% at 1000 cd m⁻², demonstrating exceptional stability and performance. In contrast, devices based on DM-Bm, DM-G, DM-X, and DM-Z exhibit lower EQEs (21.7%, 18.5%, and significantly lower for DM-X/DM-Z), underscoring the critical role of TSCT strength in device performance. Power efficiency data further highlight the superiority of DM-B, achieving 68.1 lm W⁻¹ at 10 wt% doping, with performance declining at higher concentrations due to concentration-related effects. These results demonstrate that rigid D–A configurations, such as those in DM-B, are essential for achieving high EQE, low-voltage operation, and strong power efficiency in OLEDs. The space-confined TSCT strategy effectively addresses previous limitations in exciplex-based TADF materials, such as low emission efficiency and slow RISC processes, paving the way for advanced optoelectronic applications.

Owing to the characteristic twisted U-shaped conformation, some small-molecule TADF emitters simultaneously exhibit AIE and mechanochromic behavior. As illustrated in Fig. 48a, Swager and co-workers reported a molecular design strategy that integrates TADF with AIE.³⁷⁵ The compound, namely XPT, is based on donor–acceptor systems bridged by a 9,9-dimethylxanthene spacer, which enforces a co-facial alignment of the donor and acceptor moieties with intermolecular distances of 3.3–3.5 Å. This spatial configuration enables efficient TSCT, serving as the fundamental mechanism for high-performance TADF. Furthermore, the resulting TSCT-active XPT display pronounced responsiveness to mechanical and vapor stimuli, highlighting their potential for stimuli-responsive optoelectronics. Single crystals grown via vapor diffusion of pentane into a 1,2-dichlorobenzene solution exhibit green color with an emission peak at 524 nm, while crystals obtained by solvent evaporation show a slightly redshifted emission maximum at 536 nm. Upon mechanical grinding, the emission shifts to yellow with emission peak changes to 569 nm, which is ascribed to the synergistic effect between the AIE attribute of XPT and its crystal packing adaptability. This mechano-induced yellow emission can be fully reverted to the original green state upon exposure to dichloromethane vapor or by heating above its glass-transition temperature. This reversible transition remains robust over multiple grinding-recovery cycles, indicating excellent fatigue resistance.

Fig. 48 Thermally activated delayed fluorescence from a TSCT-based small molecule. (a) U-shaped through-space architecture of XPT comprising a donor and an acceptor bridged by 9,9-dimethylxanthene. The inset is the single-crystal structure of XPT. (b) Luminescence photographs of XPT samples under a 365 nm LED lamp upon grinding, sublimation, heating, and DCM vapor. (c) Photoluminescence (PL) transient spectra of XPT in toluene under saturated oxygen and without oxygen at room temperature. The concentration of XPT is 1 × 10⁻⁴ M and λ_ex = 336 nm. (d) PL spectra of XPT in THF/water mixtures and the variation of normalized PL peak intensity with different water fractions. The concentration of XPT is 1 × 10⁻⁵ M and λ_ex = 320 nm. Inset: Luminescence images of XPT in THF/water mixtures with water fractions of 0% and 99% under 365 nm UV light. Reproduced with permission from ref. 375. Copyright 2017, American Chemical Society.

The photophysical behavior of XPT is further studied. As shown in Fig. 48c, the PL transient decay of XPT in toluene under a nitrogen atmosphere reveals a distinct delayed fluorescence component, characteristic of TADF. This delayed emission is effectively quenched in the presence of oxygen, confirming the triplet-involved TADF mechanism. In addition, with the increased water fractions in THF/water mixtures, XPT shows initially decreased but further increased PL intensity, which can be attributed to both TICT and the AIE effect (Fig. 48d). Once the water fraction exceeds 80%, a remarkable emission enhancement is observed, with the PL intensity rising to about 20 times that in pure THF. This turnaround is a hallmark of AIE, where aggregate formation in poor solvent conditions restricts intramolecular motion, thereby activating radiative decay pathways. These results collectively establish XPT as a prototypical TADF-AIEgen, combining high emission efficiency in the solid state with dynamic stimuli-responsiveness. Such a dual functional profile makes this material class highly appealing for smart OLEDs, where on-demand color tuning and real-time environmental adaptability enable dynamic displays, self-calibrating lighting, and responsive sensor-emitter platforms.

Multiple resonance (MR) TADF emitters, typically characterized by their rigid boron–nitrogen (B–N) embedded skeletons, have emerged as a privileged class of narrowband luminescent materials for high-definition OLEDs.^377–381 Their distinct electronic structure, arising from the precise separation of HOMO and LUMO on adjacent atoms, enables both narrowband emission and efficient RISC process. Despite these merits, conventional MR-TADF systems often suffer from an intrinsically slow RISC process and pronounced ACQ, which collectively limit their device efficiency and operational stability. To address these challenges, Yang's group recently devised an innovative “sandwich” configuration, in which a conventional B–N core is strategically flanked by two rigid electron-donating units.³⁷⁶ This architecture not only preserves the narrow emission profile inherent to the MR fragment but also introduces TSCT excited states. As illustrated in Fig. 49a, the TSCT state resides energetically between the local singlet (S₁) and triplet (T₁) states of the MR unit, effectively reducing the ΔE_st gap and providing an additional spin-flipping channel to accelerate RISC. Moreover, the peripheral donors exert a steric shielding effect that suppresses detrimental π–π interactions in the condensed phase, thereby alleviating ACQ and maintaining high PLQY.

Fig. 49 Thermally activated delayed fluorescence from multiple-resonance-based small molecules. (a) Illustration of the design strategy of introducing TSCT states to perturb the emission properties of MR B–N molecules. (b) Chemical structures of NBNN1 and NBNN2. (c) Photoluminescence spectra of NBNN1 and NBNN2 in mCP at 298 K. (d) Transient photoluminescence characteristics of NBNN1 and NBNN2 in mCP at 298 K in Ar. (e) The fluorescence (298 K) and phosphorescence (77 K) spectra of NBNN1 and NBNN2 in mCP. (f) Proposed excited state diagrams for NBNN1 and NBNN2 in mCP. (f) Calculated hole (green) and electron (blue) distributions (isosurface = 0.002) and energies of the ^1,3MRCT and ^1,3TSCT excited states for NBNN1 and NBNN2. Reproduced with permission from ref. 376. Copyright 2020, John Wiley & Sons, Inc.

Two representative emitters, NBNN1 and NBNN2, are synthesized to validate this design concept (Fig. 49b). When doped into a 1,3-bis(9-carbazolyl)benzene (mCP) host at 10 wt%, both compounds exhibit green emission with peaks at 520 nm and 531 nm, respectively (Fig. 49c). Notably, NBNN2 shows a narrower FWHM of 38 nm compared to 73 nm for NBNN1, highlighting its superior color purity. Transient photoluminescence decay profiles recorded under an argon atmosphere at 298 K confirmed TADF characteristics, with prompt and delayed fluorescence lifetimes of 65 ns and 6.8 µs for NBNN1 and 25 ns and 10.6 µs for NBNN2 (Fig. 49d). Low-temperature spectroscopic analyses further reveal differences in their triplet manifolds. NBNN1 displays a single phosphorescence component with an FWHM of 41 nm at 77 K, which indicates a dominant ³MRCT characteristic. However, NBNN2 exhibits dual phosphorescence bands, attributed to the coexistence of ³TSCT and ³MRCT states (Fig. 49e). The slightly higher-lying ³TSCT state in NBNN2 facilitates efficient RISC by bridging the ¹MRCT and ³MRCT states, as depicted in the proposed energy-level diagram (Fig. 49f). This synergistic interplay between MR and TSCT states enhances spin–orbit coupling and promotes faster RISC. Through this molecular design, their study addresses key challenges of MR-TADF emitters, such as long-delayed fluorescence lifetimes and aggregation issues, while enhancing OLED performance in terms of efficiency and operational stability. Overall, the design supplies a versatile, mechanism-anchored strategy to enhance MR emitters in OLEDs, marrying ultrasharp emission with efficient RISC and manufacturability suited to next-generation display and lighting technologies.^382–385

4.2. Polymers

The investigation of TSCT phenomena in polymeric systems represents an area of significance comparable to analogous studies in small-molecule systems, while offering distinct advantages rooted in macromolecular architecture.^387–390 The extended π-conjugated backbone inherent to polymers enables superior charge transport capabilities compared to discrete molecular systems, effectively addressing critical limitations in charge transfer efficiency and stability that often plague small molecules constrained by finite dimensions. This structural advantage facilitates not only enhanced PLQY through minimized non-radiative decay pathways but also provides unprecedented flexibility in emission property engineering. Through systematic modulation of donor and acceptor moieties within the polymeric framework, precise spectral tuning across the visible spectrum (400–700 nm) can be achieved, which is a crucial requirement for advanced applications in full-color displays and white-light-emitting technologies.^391,392

The above molecular design paradigm through TSCT polymers exhibiting programmable emission characteristics was systematically reported by Wang et al. (Fig. 50).³⁸⁶ As illustrated in Fig. 50a, chromatic control could be achieved via strategic D/A pairing, enabling emission spanning from deep blue (450 nm) to red (650 nm) regions. Meanwhile, efficient white-light generation is demonstrated through simultaneous incorporation of Ac/TRZ-H (as the blue emitter) and Ac/TRZ-CN (as the yellow emitter) pairs, with emission intensity ratios finely controlled through acceptor composition tuning (CIE coordinates: 0.33–0.36, 0.33–0.38). Optical characterization reveals well-defined UV absorption signatures (300–400 nm) coupled with broad emission profiles (Fig. 50b), while transient photoluminescence analysis of solid-state films confirms pronounced TADF behavior with lifetime components ranging from 0.36 µs to 1.98 µs (Fig. 50c). Notably, green-emitting variants (P3-05 and P3-50) exhibit superior delayed fluorescence contributions (91–93%) compared to their red-emitting counterpart (P5-05: 6%), highlighting a clear emission wavelength dependence of TADF efficiency. EL performance evaluation further validates the viability of this design strategy for optoelectronic applications, demonstrating dual emission bands at 472 nm (blue) and 550 nm (yellow) with intensity-modulation capability through TRZ-CN content variation (Fig. 50d). This established structure–property relationship positions TSCT polymers as versatile platforms for developing advanced OLEDs, combining broad spectral tunability with efficient triplet harvesting mechanisms. The demonstrated white-light generation via intramolecular exciton allocation to distinct D/A pairs marks a substantive advance over conventional blend-based approaches, delivering single-component emission with enhanced color stability and a simplified device stack.

Fig. 50 Thermally activated delayed fluorescence from polymeric systems based on π-conjugated donor and acceptor units. (a) Molecular design and chemical structures of TSCT polymers with full-color and white emission. The polymers are composed of donor (D) and acceptor (A) units, with emission in B (blue), G (green), and R (red) regions. (b) The absorption and photoluminescence (PL) spectra of the TSCT polymers in the solution state with a concentration of 1 × 10⁻⁵ M, with excitation wavelengths of 350, 375, 385, 400, and 440 nm for P1-05, P2-05, P3-05, P4-05, and P5-05, respectively. (c) Transient PL decay curves in film under a N₂ atmosphere. (d) Electroluminescence spectra of the solution-processed OLEDs based on TSCT polymers with white emission. Reproduced with permission from ref. 386. Copyright 2019, John Wiley & Sons, Inc.

Building on the design principles of TSCT polymers for emission tuning, Xu et al. further report a paradigm-shifting approach for optimizing TADF through precise regulation of intra- and interchain TSCT.³⁹³ By engineering non-conjugated copolymers incorporating spatially optimized donor (DMAC) and acceptor moieties (DPPO, DPT, DPOT, and TPOT), the researchers achieved unprecedented control over triplet exciton harvesting (Fig. 51). The strategic use of diphenylphosphine oxide–triazine hybrid acceptors (e.g., DPOT) introduces synergistic electronic and steric effects, enabling simultaneous enhancement of TSCT efficiency (PLQY = 95%) and suppression of non-radiative decay. This molecular design leads to record electroluminescence performances, with the optimized copolymers surpassing previous benchmarks in external quantum efficiency and power efficiency among solution-processable TADF materials.

Fig. 51 Thermally activated delayed fluorescence from through-space charge transfer (TSCT)-based copolymers. (a) Molecular design, structures, and simulated electronic characteristics of non-conjugated donor–acceptor (D–A) copolymers featuring TSCT. Two chains of non-conjugated copolymers with pendant donor and acceptor groups, in which three parts with different D–A interactions are highlighted with the features of intrachain and interchain TSCT, and invalid charge-transfer-free interactions, in contrast to only interchain TSCT in blends of homopolymers, respectively containing D or A groups. (b) Chemical structures of copolymers and selected acceptor groups. (c) Photoluminescence (PL) spectra of all synthesized polymers in films. (d) Time decay curves of promoted PL in nanosecond scale (inset) and delayed fluorescence (DF) in microsecond scale for the films of copolymers and corresponding blends. (e) The schematic diagram illustrates the modulation mechanism of through-space charge transfer (TSCT). In this diagram, k represents the rate constant. Reproduced with permission from ref. 393. Copyright 2023, John Wiley & Sons, Inc.

In this molecular design, donor–acceptor coupling spans intrachain TSCT along an individual backbone, interchain TSCT across adjacent chains, and a residual manifold of non-contributing interactions, collectively defining the operative TSCT regime (Fig. 51a). This multi-mode framework contrasts with conventional blend systems, which rely predominantly on interchain pathways. The chemical evolution of the acceptors from the basic DPPO unit to sterically optimized TPOT derivatives enables systematic tuning of electronic conjugation and spatial confinement (Fig. 51b). Optical characterization shows pronounced bathochromic shifts (Δλ > 50 nm) in PL maxima across the copolymer series, correlating directly with the electron-withdrawing strength of the acceptors (Fig. 51c). Time-resolved spectroscopy also highlights the balance between prompt and delayed fluorescence, with DPOT-based systems exhibiting a threefold increase in delayed contribution relative to the DPPO reference (Fig. 51d). The proposed TSCT-modulation mechanism defines a dual-parameter optimization strategy in which inductive substituent effects enhance frontier-orbital overlap to compress ΔE_ST and facilitate RISC, while steric confinement channels the charge-transfer pathway and curbs triplet–triplet annihilation by limiting exciton diffusion and deleterious aggregation (Fig. 51e). This synergistic approach achieves a remarkable exciton utilization efficiency of 92% in the best-performing system, which is a 40% improvement over prior TADF polymer designs. This work substantially advances the management of triplet excitons in disordered polymer matrices. By decoupling electronic and spatial contributions within TSCT processes, it provides a universal design blueprint for next-generation OLED materials that combine solution processability with high device efficiency. The ability to rationally tune emission color via acceptor engineering elevates these copolymers to versatile platforms for full-color displays and energy-efficient solid-state lighting, enabling precise control of chromaticity (CIE coordinates), minimized spectral cross-talk, and stable output across drive currents.

The integration of TSCT with TADF represents a significant leap forward in the molecular design of organic optoelectronic materials. This paradigm enables precise modulation of excited-state dynamics, leading to markedly improved RISC rates, suppressed non-radiative decay, and exceptionally high PL quantum yields. Furthermore, the ability to program intra- and interchain TSCT pathways offers unprecedented control over emission color and efficiency across the visible spectrum, including stable white-light emission. These advances not only provide a robust framework for developing high-performance, solution-processable OLEDs with reduced efficiency roll-off but also establish TSCT-TADF as a versatile and powerful design platform for next-generation displays and solid-state lighting technologies. Nevertheless, almost all present examples are constructed based on TSCT of conjugated donors and acceptors with non-conjugated linkers (e.g., small molecules with “sandwich” configurations and tailored D–A copolymers), and non-conjugated skeletons that show a TADF effect via the TSC mechanism are expected in the near future.

5. Through-space interactions for room-temperature phosphorescence

Due to their long-lived and time-resolved luminescence characteristics, room-temperature phosphorescent (RTP) materials have become an important component of organic optoelectronic materials.^394–399 Compared with well-studied traditional inorganic RTP materials, organic RTP materials have attracted considerable attention in recent years owing to their advantages, including wide tunability of luminescent color, excellent processability, biocompability, and potential applications spanning OLEDs, flexible devices, information encryption and anti-counterfeiting, and bioimaging.^400–404 However, the weak spin–orbit coupling (SOC), spin-forbidden T₁ to S₀ transitions, and strong non-radiative decay of excitons inherent in organic materials often result in low phosphorescence efficiency and short lifetime, which significantly limit their practical applications.⁴⁰⁵ To address these challenges, various design strategies have been developed and reviewed to enhance the RTP performance of organic molecules, including heavy atom incorporation, molecular rigidification, molecular orbital engineering, host–guest doping, and aggregation.^406–410

Conventional organic RTP materials typically feature extended π-conjugation and distinct D–A structures, which facilitate efficient ISC by minimizing the singlet–triplet energy gap.⁴¹¹ These proposed strategies have been proven to greatly increase the efficiency and lifetimes of π-conjugation-based systems.⁴¹² Intriguingly, recent studies have unexpectedly observed RTP phenomena in numerous non-conjugated and even non-aromatic heteroatom-containing aliphatic systems.^413–418 Beyond the conventional orbital hybridization involving n and π electrons, non-covalent TSI of electrons, including TSC and TSCT, has been identified as the intrinsic mechanism responsible for their RTP properties. Although a few regulating strategies are occasionally reported, it is still challenging to draw a complete picture to realize feasible modulation of RTP efficiency and lifetimes of these non-conjugated systems. Thus, a comprehensive understanding of the intrinsic mechanisms and performance modulation principles of TSI will undoubtedly provide both fundamental theoretical insights and practical value for the development of novel organic materials with unique non-conjugated structural features and excellent RTP properties. Accordingly, this section will systematically examine various types of RTP materials, with particular emphasis on elucidating how TSI can be effectively generated and manipulated to achieve efficient RTP.

5.1. Single-component systems

As natural products without conjugated structures and π electrons, starch, and cellulose were reported as typical examples to show RTP by Tang and Yuan et al. in 2013 (Fig. 52).¹¹⁴ These materials are non-luminescent in the solution state but highly luminescent upon cooling at 77 K to form a solid glass state. Interestingly, they show bright blue emission after crystallization, which is also proven to play a critical role in achieving RTP with the help of strong and abundant intra-/inter-molecular interactions that rigidify their skeleton and block nonradiative decay. Thus, the concept of crystallization-induced phosphorescence was proposed. Although they cannot provide direct evidence, the electron-rich oxygen and glucose units and their appropriate interactions in the crystalline environment are proposed to be responsible for the observed RTP, which has been proven by numerous examples at present. Later, Peng et al. reported the stepwise rigidification and structure-modification approaches for achieving color-tunable and long-lived RTP from cellulose.^419–421

Fig. 52 Room-temperature phosphorescence from natural products. (a) The chemical structures of starch and cellulose, and photographs of their solid powders, taken under UV irradiation. (b) The photoluminescence spectra of starch and cellulose in solution and powder states. Reproduced with permission from ref. 114. Copyright 2013, Springer Nature.

In addition, following the intrinsic visible emission of amino acids, RTP of poly(amino acid) was also observed and reported by Yuan et al. in 2017.⁴²² Similar to the phenomenon of CL, several amino acids (e.g., L-glycine, L-alanine, L-serine, L-arginine, and L-lysine) are non-emissive in dilute solution, but their concentrated solutions and crystalline forms can emit blue or green light under UV irradiation with varying PLQY of 0.1–7.4%. Persistent phosphorescence from their solid samples at 77 K after ceasing the irradiation is also observed, which is previously believed to be exclusively stemming from aromatic systems. Inspired by this behavior, they further studied the photophysical properties of their corresponding polymers and detected the RTP from ε-poly-L-lysine (ε-PLL). As a solid power, ε-PLL exhibits excitation-dependent fluorescence around 410–490 nm with the highest PLQY of 7.9% (Fig. 53a and b). With a delay time of 0.1 ms, green RTP at 490 nm is measured, indicating its rigid conformation assisted by the polymeric chain effect and multiple intra- and inter-molecular interactions. Besides, similar behavior was observed in the biomacromolecular luminogen of sodium alginate, a plant-derived anionic polysaccharide with a similar composition to starch.^120,336 Since the α-L-guluronic acid (G) and 1,4-linked β-D-mannuronic acid (M) residues show different rigidity, two typical chain sequences of GM_r (∼1.53 M/G blocks) and G_rM (∼0.69 M/G blocks) exhibit different RTP properties (Fig. 53c). Although they display blue-white fluorescence and green RTP, GM_r processes a longer phosphorescent lifetime of 27.4 ms than G_rM of 9.1 ms because the G block is stiffer and more sterically hindered. Due to the high ion-binding ability, the RTP lifetimes could be further adjusted by Ca²⁺ ions. Accordingly, the lifetimes of film-Ca²⁺ GM_r and film-Ca²⁺ G_rM extended to 49.0 ms and 30.0 ms, respectively (Fig. 53d). Based on these results, the photophysical properties and working mechanism of SA are summarized, which are closely related to the clustering of oxygen atoms and carboxylates (Fig. 53e).⁴²³ Isolated molecules in dilute solution are non-emissive due to a lack of rigidity, but clustering upon concentration or aggregation facilitates O⋯O, C=O⋯C=O, and H–O⋯C=O interactions and promotes electronic TSI among adjoining n and π electrons, achieving visible-light emission.⁴²⁴ Hydrogen bonding and Ca²⁺ ion binding further rigidify conformations and enhance spatial conjugation, which promotes ISC and finally realizes RTP in the solid state.⁴²⁵

Fig. 53 Room-temperature phosphorescence from nonaromatic poly(amino acids) and sodium alginate. (a) Chemical structure of ε-poly-L-lysine (ε-PLL) and its photographs taken before and after UV irradiation. (b) Normalized photoluminescence spectra of solid-state ε-PLL solids with a delay time of 0 ms (solid line) and 0.1 ms (dashed line) under different excitation wavelengths. Reproduced with permission from ref. 422. Copyright 2017, Springer Nature. (c) Chemical structure of G and M units and typical chain sequence of sodium alginate. (d) Photographs of GM_r and G_rM in the form of powder, film, and Ca²⁺ incorporated film taken under 312 nm UV light or after ceasing the UV irradiation. (e) The illustration diagram of sodium alginate from isolated to aggregated states and possible intra- and intermolecular interactions within the clusters for observed RTP. Reproduced with permission from ref. 120. Copyright 2018, American Chemical Society.

Apart from the above natural systems, some artificial nonaromatic systems are also designed and synthesized, in which these electron-rich heteroatoms play a critical role in achieving RTP. Inspired by the noncovalent interactions of DNA, hydantoin (HA), which resembles thymine in DNA, was selected and proved to show efficient RTP in 2020.⁴²⁶ In the presence of carbonyl groups and nitrogen heteroatoms, the promoted SOC and subsequent ISC process endow crystalline HA to show a yellow-green RTP with a high phosphorescent QY of 21.8% (87.5% PLQY in total) and lifetime of 1.74 s under the 365 nm irradiation (Fig. 54a–d). Through single-crystal analysis, the planar conformation with effective intermolecular interactions is regarded as an important factor to restrict molecular motions and generate TSI for its efficient CL and RTP. In addition, two dimeric derivatives of HA, namely 1,1′-methylenedihydantoin (MDHA) and 1,1′-(ethane-1,1-diyl)dihydantoin (EDHA), were also synthesized and compared (Fig. 54c and d). Both MDHA and EDHA exhibit excitation-tunable CL and RTP similar to that of HA crystals, indicating the presence of multiform emission species induced by TSI of subgroups bearing π and n electrons. However, the phosphorescent lifetimes and efficiencies of these two derivatives are lower than that of HA, which may be ascribed to the destruction of electronic TSI and the more active molecular motions caused by the steric hindrance of methyl and ethyl groups.

Fig. 54 Room-temperature phosphorescence from nonaromatic small molecules. (a) Chemical structure and typical features of hydantoin (HA). (b) The prompt and delayed photoluminescence spectra of crystalline HA under different excitation wavelengths. (c) The synthetic route and emissive photographs of MDHA and EDHA with intramolecular through-space conjugation. (d) The comparison of quantum yields and lifetimes of HA, MDHA, and EDHA. Reproduced with permission from ref. 426. Copyright 2020, The Royal Society of Chemistry. (e) Chemical structures and photographs of CBSI and OBSI taken under 312 nm and 365 nm UV light or after ceasing the UV irradiation. (f) The delayed photoluminescence spectra of CBSI and OBSI under different excitation wavelengths. (g) The frontier molecular orbitals of CBSI trimer and tetramer, which show typical through-space conjugation of electrons in LUMO. Reproduced with permission from ref. 427. Copyright 2020, John Wiley & Sons, Inc.

Similar systems with multiple carbonyl groups, namely N,N′-carbonylbissuccinimide (CBSI) and N,N′-oxalylbissuccinimide (OBSI), were also obtained and reported to show AIE, CL, and tunable RTP (Fig. 54e).^427,428 These two compounds are non-emissive when molecularly dissolved in solution but show visible-light emission in concentrated solutions or crystalline state, suggesting their typical AIE feature. CBSI displays tunable RTP, which red-shifts from 505 to 575 nm when the excitation wavelength increases from 290 to 400 nm. In contrast, OBSI shows less obviously varied RTP wavelength in the range of 550 to 575 nm. Interestingly, these two compounds exhibit better tunability of RTP wavelength under cryogenic conditions (Fig. 54f).^429,430 For instance, under an excitation wavelength of 250–450 nm, the phosphorescence of CBSI crystals changes from 415 to 636 nm and accompanies the CIE coordinates changing from (0.18, 0.15) to (0.48, 0.45), and the crystalline OBSI also gradually redshifts from 410 to 596 nm. Theoretical calculations based on their crystal packing also reveal the intensive TSC among carbonyl and imide subgroups on the LUMO of their trimers and tetramers, resulting in diverse clusters with a narrowing energy gap and tunable RTP.

Apart from the above-discussed non-conjugated and even nonaromatic systems, electronic TSI within conventional conjugated compounds also plays an important role in regulating their optoelectronic properties. Although noncovalent intramolecular interactions (e.g., S⋯O, Se⋯O, S⋯F, etc.) have been employed to modulate the photoluminescence properties of organic/polymeric semiconductors, the underlying relationship between them is still elusive.^{51,133,432–434} In 2023, Huang et al. successfully achieved manipulation of triplet states for efficient RTP and a large wavelength difference between fluorescence/phosphorescence (Fig. 55).⁴³¹ With the help of the carbonyl group in chromone, they first designed a series of chromone derivatives with intramolecular TSI and planar conformation. Through single-crystal structure, NMR spectrum, and natural bond orbital analysis (Fig. 55b), the strength order of noncovalently intramolecular interactions is proved to gradually increase from BC to 3TC and 2TC. Surprisingly, although they show almost the same wavelength of fluorescence, the wavelength of phosphorescence gradually redshifts, resulting in the largest wavelength difference between emission peaks of 279 nm for 2TC. As a comparison, the wavelength differences are 165 nm for BC and 188 nm for 3TC, respectively. Meanwhile, the lifetime of RTP gradually declined from 81.7 ms for BC to 6.0 ms for 3TC and 0.3 ms for 2TC, respectively. Theoretical calculation also suggests that with the gradually stronger noncovalently intramolecular interactions, the energy levels of S₁ states are similar for all three compounds, while T₁ decreases and T_2/3 is closer to S₁, finally resulting in efficient ISC as well as achieving a large wavelength difference between fluorescence/phosphorescence (Fig. 55c and d). Accordingly, utilizing the temperature-dependent lifetime of RTP and intensity ratio of fluorescence/phosphorescence, 3TC is further fabricated as nanoparticles as a new dual-ratiometric thermometer for in vitro and in vivo bio-thermometry.

Fig. 55 Room-temperature phosphorescence assisted by noncovalent intramolecular interactions (NIIs). (a) Chemical structures of BC, 3TC, and 2TC. (b) Natural bond orbital analysis of BC, 3TC, and 2TC, which shows the overlap of orbitals of NIIs. (c) The diagram of energy levels and regulation of triplet states through NIIs to achieve a large Δλ. Ex: excitation; Fl: fluorescence; IC: internal conversion; P: phosphorescence. (d) The normalized prompt and delayed photoluminescence spectra of BC, 3TC, and 2TC. The wavelength shift (Δλ) between fluorescence and phosphorescence is indicated. Reproduced with permission from ref. 431. Copyright 2023, John Wiley & Sons, Inc.

In addition to modulating triplet energy levels, increasing the strength of ISC is also critical to achieve efficient RTP. With this consideration, heavy atoms such as bromine and iodine are often utilized to increase the TSC of pure organic systems for excellent RTP.^436–439 The simplest way is to introduce these heavy atoms as anions to interact with luminogens. In 2018, Wang et al. demonstrated a representative example using bromide anions to achieve RTP from single-component organic aggregates with strong noncovalent anion–π⁺ interactions. In their study, a series of 1,2,3,4-tetraphenyloxazoliums with different counterions as organic salt compounds was synthesized (Fig. 56a–c).⁴³⁵ Among them, TPO-Br and TPO-I show outstanding fluorescence and RTP at the same time, while the other compounds only produce fluorescence under UV irradiation. Taking TPO-Br as an example, the single-crystal analysis reveals that two bromide anions are positioned on either side of the positively charged oxazolium rings at distances of 3.494 Å and 3.601 Å, forming effective anion–π⁺ interactions (Fig. 56a). These interactions, along with the twisted molecular conformation, suppress nonradiative decay and enhance emission in aggregates. Importantly, the small energy gap between S₁ and T₃/T₄ states (0.18 eV) together with strong SOC provided by the bromine atoms facilitates efficient ISC, leading to intense RTP (Fig. 56d). Notably, since the ratio of fluorescence and phosphorescence can be simply tuned by the degree of crystallization or environmental rigidity, it is easy to balance dual emission and achieve efficient white-light emission. Accordingly, the doped film of TPO-Br in PEG produces high-quality white light with CIE coordinates of (0.31, 0.33), closely matching the standard white point (0.33, 0.33) (Fig. 56e).^440,441 Since PEG is a good 3D printing material, TPO-Br was also utilized as an additive of 3D printing materials for lampshades with white light emission (Fig. 56f), showing one potential application of these RTP materials.

Fig. 56 RTP assisted by through-space anion–π⁺ interactions. (a) Chemical structure of TPO-I, TPO-Br, TPO-Cl, TPO-F, and TPO-P with different counter ions, and single-crystal structure of TPO-Br with anion–π⁺ interactions. (b) Normalized photoluminescence (PL) spectra of TPO-I, TPO-Br, TPO-Cl, TPO-F, and TPO-P in the solid state. (c) The time-resolved PL decay of TPO-I and TPO-Br detected at their maximum emission wavelength. The RTP photographs were taken under 365 nm UV irradiation. (d) The calculated energy levels diagram and SOC constants of TPO-Br in the crystalline state using the ONIOM method. (e) The PL spectra and CIE 1931 coordinates of the PEG film containing TPO-Br (2%, m/m). (f) The 3D printed lampshades without (left) and with (right) TPO-Br taken under daylight (top), UV light irradiation (middle), and UV-LED lamps (bottom). Reproduced with permission from ref. 435. Copyright 2018, Springer Nature.

Despite the great success of the heavy-atom effect in improving the RTP efficiency, controlling the intermolecular packing between anions and cationic luminogens remains challenging for consistently strong and fast ISC.^48,443–445 Thus, the idea of an intramolecular heavy-atom effect was proposed and verified, which is more controllable for constructing anion-π interactions (Fig. 57).⁴⁴⁶ For instance, based on carbazole units, a couple of molecules of PDCz without Br atoms and PDBCz with two Br atoms were synthesized (Fig. 57a).⁴⁴² Under UV irradiation, PDCz shows blue emission at about 410 nm, which is assigned to the fluorescence with a nanosecond lifetime. With the delay of 5 ms, another emission peak belonging to phosphorescence at 600 nm is detected under ambient conditions (Fig. 57b). In contrast, apart from a weak emission peak within the blue region, PDBCz with two Br atoms exhibits bright yellow light at about 600 nm, which is overlapped with its delayed spectra, indicating its strong RTP. The shorter RTP lifetime of PDBCz (217 ms) than that of PDCz (560 ms) also suggests its faster radiative decay compared to the nonradiative decay, which finally results in its very high phosphorescent PLQY of 38.1% (Fig. 57c). To better understand their differences, single-crystal analysis and theoretical calculations were carried out. The results indicate that each carbazole subunit is surrounded by one intramolecular C–Br⋯π interaction (3.591 Å) and multiple intermolecular interactions (such as C–Br⋯π, C–Br⋯N, and C–H⋯N), while only intramolecular and intermolecular C–H⋯N interactions exist in PDBCz. All these interactions provide a stable external environment to stabilize triplet excitons for ultralong RTP. Besides, the strong intramolecular C–Br⋯π interaction of PDBCz endows it with large SOC constants and small energy gaps between singlet and triplet states (Fig. 57d), greatly increasing its ISC efficiency and finally achieving excellent RTP. In addition, based on PDBCz powder, several small patterns of flexible films and 3D models with bright-yellow RTP were also fabricated, exhibiting their potential in 3D-printing luminescent materials and flexible electronics (Fig. 57e).

Fig. 57 Room-temperature phosphorescence (RTP) assisted by noncovalent intramolecular halogen interaction. (a) Chemical structures of PDCz intramolecular halogen interaction and PDBCz with intramolecular halogen interaction. (b) The normalized prompt and delayed photoluminescence spectra of PDCz and PDBCz. (c) The decay curve of RTP peaks and quantum yields of PDCz and PDBCz. (d) The calculated energy level and SOC constants of PDCz and PDBCz in the crystalline state. (e) The fabricated patterns of flexible films and 3D models with bright RTP based on PDBCz powder. Reproduced with permission from ref. 442. Copyright 2019, American Chemical Society.

Even though many strategies have been developed to increase ISC, achieving efficient ISC and subsequent RTP without heavy atoms remains challenging.⁴⁴⁸ As discussed above, TSCT has emerged as a promising strategy for achieving efficient TADF via a tunably small energy gap between singlet and triplet states and strong electronic coupling between donor and acceptor units with an enhanced reverse ISC process. However, the influence of TSCT on the forward ISC channel and RTP has remained relatively unexplored. With this consideration, three space-confined bridged phosphors were designed by He et al., in which phenothiazine is connected to dibenzothiophene, dibenzofuran, and carbazole units via a 9,9-dimethylxanthene bridge.⁴⁴⁷ Taking dibenzothiophene-based XPT as an example (Fig. 58a), its single crystal exhibits a face-to-face arrangement between the donor and acceptor phenyl groups with a short π–π distance of 3.34 Å. The absorption reflects the sum of xanthene, phenothiazine, and dibenzothiophene contributions at short wavelength, suggesting their non-conjugated nature in the ground state. Its PL spectra of crystals show perfect overlapping between the prompt and delayed spectra with bright-green color at both room temperature and 77 K, indicating the efficient ISC and dominant component of phosphorescence (Fig. 58b and c). The lifetimes of the phosphorescence in the crystalline state were also measured as 5 ms at room temperature and 77 ms at 77 K (Fig. 58d).

Fig. 58 Room-temperature phosphorescence (RTP) assisted by through-space spin-orbital coupling. (a) Chemical structure of XPT. (b) The prompt and delayed photoluminescence (PL) spectra of XPT at room temperature. The photographs were taken under 365 nm UV light or after ceasing the irradiation. (c) The prompt and delayed PL spectra of XPT at 77 K. The photographs were taken under 365 nm UV light or after ceasing the irradiation for different times. (d) The decay curve and lifetime of XPT crystals at room temperature detected at 532 nm. (e) The calculated energy level diagram and SOC constants of XPT. (f) The frontier molecular orbitals of XPT. Reproduced with permission from ref. 447. Copyright 2021, American Chemical Society.

Besides, an interesting phenomenon is observed that nearly pure RTP is solely from the phenothiazine segment in these three compounds without the obvious influence of the bridged segments. The feasible vibration of nonplanar phenothiazine within the bridged molecular conformation may account for this observation. Besides, the through-space SOC also plays a critical role in promoting the ISC process.⁴⁴⁹ It is worth noting that, different from covalently bonded D–A systems with a strong electronic push–pull effect to realize HOMO–LUMO separation, through-space bridged systems are insensitive to the electronic features of the donor and acceptor, which enables non-conjugated system with a small difference of the counterparts in electronic properties to achieve complete HOMO–LUMO separation and very small energy gap between singlet and triplet states (Fig. 58e and f). In addition, accompanied by the hybrid orbital configurations of S₁ and T_n (which benefits the spin-vibronic coupling for ISC), the rate constant of ISC is measured as 7.1 × 10⁸ s⁻¹, which is much larger than most reported heavy-atom-free RTP systems (0.12–2.2 × 10⁸ s⁻¹). Therefore, this example provides a general strategy and mechanistic understanding for effective RTP materials based on a through-space bridged weak D–A scaffold.^450–452

Following the concept that non-conjugated D–A systems with complete HOMO–LUMO separation and a small energy gap for efficient ISC, numerous TSCT-based RTP systems have been designed.^453–456 Most of them display a locally excited state (³LE) character with comparatively long lifetimes.^457–459 By contrast, the development of charge transfer excited state (³CT)-featured RTP materials remains a significant challenge. For those twisted or coplanar D–A systems based on TBCT, they show strong SOC or effective D–A coupling through the conjugated π-bridge, but the forbiddenness degree of the radiative ³CT → S₀ transition hampers the achievement of a long lifetime, according to El-Sayed's rules. Inspired by the strategy of the TSCT mechanism, Lu et al. developed a ³CT-based system based on non-conjugated donor and acceptor units with a CH₂(sp³) linker, namely NIC-DMAC, to achieve a prolonged RTP lifetime (Fig. 59a).⁴⁶⁰ At room temperature, steady-state PL spectra show two broad emission bands at about 500 nm and 580 nm with a structureless nature, which implies that the CT transition feature between donor and acceptor in NIC-DMAC (Fig. 59b). With the delay time of 1 ms, only the emission band at 580 nm persists, indicating its RTP feature. At 77 K, the delayed PL of NIC-DMAC shows a fine-structured phosphorescence peak centered at 583 nm, which is similar to the shape and position of locally excited phosphorescence of the acceptor part (³LE_A), suggesting the T₁ state with ³LE_A feature of NIC-DMAC (Fig. 59c). Given that the delayed emission at room temperature exhibits a blue shift compared to its T₁-state phosphorescence, the observed RTP likely originates from a higher-lying triplet excited state exhibiting through-space charge transfer (³TSCT) characteristics. To explore this possibility, the natural transition orbitals (NTO) were calculated, which reveal that the S₁ state exhibits a predominantly charge-transfer character (¹CT), whereas the T₁ and T₂ states of NIC-DMAC are characterized by locally excited triplet (³LE_A) and through-space charge-transfer triplet (³TSCT) nature, respectively. Interestingly, with the increased temperature from 77 K to 293 K, it displays a notable enhancement in PL intensity at around 580 nm, accompanied by a reduction in spectral fine structure. This observation supports the occurrence of a kinetic process involving reverse internal conversion (RIC) from ³LE_A to ³TSCT at room temperature. Consequently, the RTP observed in NIC-DMAC results from a sequential photophysical process involving ¹CT (S₁) → ³LE_A (T₁) → ³TSCT (T₂), achieving a long lifetime of 210 ms and a high phosphorescent efficiency of 6.1%.

Fig. 59 Room-temperature phosphorescence (RTP) assisted by through-space charge transfer (TSCT). (a) Illustration diagram of through-bond charge transfer (TBCT) and TSCT for RTP, and the chemical structure and photophysical properties of NIC-DMAC. (b) The temperature-dependent photoluminescence spectra of NIC-DMAC mixed in PMMA film (concentration = 1.5 wt%). (c) The phosphorescence spectra of two isolated segments, NI-Br and DMAC, and the synthesized NIC-DMAC at 77 K. (d) Natural transition orbitals of NIC-DMAC and its excited-state electron processes of NIC-DMAC, which involve sequential transitions from S₁ to T₁, and then to T₂. Reproduced with permission from ref. 460. Copyright 2024, John Wiley & Sons, Inc.

To obtain a long lifetime and high efficiency simultaneously, it is essential to augment the ISC efficiency and reduce the fluorescence rate and nonradiative decay rate. Although NIC-DMAC displays a comparatively long RTP lifetime, its phosphorescence quantum yield of 6% is low. This is attributed to the comparatively low ISC rate of 1.4 × 10⁷ s⁻¹, even though the spin–orbit charge-transfer (SOCT) effect minimizes its energetic splitting. With the consideration that a larger dihedral angle between donor and acceptor moieties can promote larger SOC matrix element (SOCME) values and subsequent faster ISC rate,⁴⁶¹ they further proposed a conformational strategy to increase phosphorescent efficiency by modulating the U-shaped molecules into a V-shaped analog.⁴⁶² Accordingly, V-1, featuring the same donor and acceptor subunits as NIC-DMAC, and V-2, featuring the identical acceptor but a bulkier donor, were designed and synthesized (Fig. 60a).⁴⁶³ Compared to the parent compound of NIC-DMAC (also named U-1), V-1 and V-2 also show similar excited-state transitions where the S₁, T₁, and T₂ states are dominated by ¹CT, ³LE_A, and ³TSCT features, respectively. Although V-1 and V-2 display similar PL spectra, delayed spectra, and lifetimes to U-1, their phosphorescent efficiency is determined to be 25% and 13%, respectively, which are much higher than that of U-1 (Fig. 60b and c). With the help of calculated geometries of these three compounds based on the S₁ state, the V-shaped conformation endows V-1 and V-2 with a larger dihedral angle between donor and acceptor subunits of 59° and 64°, respectively. As a comparison, the dihedral angle of U-1 is only 32°. Therefore, the SOCME value is calculated to be 0.31 cm⁻¹ and 0.37 cm⁻¹, which is two-fold higher than that of U-1 (Fig. 60d and e). Accordingly, V-1 and V-2 demonstrate large ISC rates of 2.4 × 10⁷ s⁻¹ and 3.2 × 10⁷ s⁻¹, respectively, which is attributed to their increased SOCME between their S₁ and T₁ states and dihedral angle. This work clearly demonstrates the structural refinement in non-conjugated systems with TSCT features to extend lifetimes and elevate the efficiency of RTP.

Fig. 60 Molecular shape-dependent RTP with through-space charge transfer features. (a) A diagram illustrating the conformational change from U-shape to V-shape for long-lifetime and efficient RTP. The optimized geometries and dihedral angles of three compounds in the S₁ state are shown with their RTP lifetimes and efficiencies. (b) Normalized prompt and delayed photoluminescence spectra of the three compounds in different states. (c) The decay profiles of the three compounds mixed in the PMMA film. (d) The temperature dependence of the sum of the rate constants of radiation and nonradiative transition processes from the triplet excited state. (e) The calculated natural transition orbitals of S₁ and T₁ states and corresponding SOC constants of the three compounds. Reproduced with permission from ref. 463. Copyright 2025, John Wiley & Sons, Inc.

5.2. Multiple-component systems

Compared with the single-component RTP systems based on purely non-conjugated molecules, multi-component systems are in their infant stage with few reported examples. The idea of multi-component systems may come from the nature that some inorganic materials, such as luminous pearls and glow-in-the-dark stones with ultralong afterglow, where impurity is considered to produce a charge-separated state and result in slow charge recombination to realize afterglow (i.e., doublet luminescence).^{394,465–471} Although some host–guest systems with the charge-separated state are constructed following this idea, the multi-component organic systems and detailed photophysics of molecular phosphorescence (i.e., RTP from triplet excited state) have not been completely investigated. Several purely organic host–guest systems have been reported, but they still focus on the conventional strategy of external and internal heavy-atom effects.^472,473 As a result, both S_n–T_n and T_n–S₀ electronic transitions are facilitated, resulting in enhanced ISC and RTP quantum yield but decreased lifetime.^411,474 It is noteworthy that the guests within these systems are generally considered to undergo the excitation and emission processes, while the host only dilutes or rigidifies guest molecules to restrict nonradiative decay. However, different from those inorganic systems where both host and guest play active roles in electronic transitions, whether the electronic transition processes involve the host species within multi-component organic RTP systems has not been answered so far.

With this consideration, Zhang et al. proposed a study to verify the role and photophysical processes of the host species using commercially available, heavy-atom-free, and electron-deficient host/guest systems.⁴⁶⁴ As shown in Fig. 61a, 1,8-naphthalic anhydride (NA) is selected as the guest while pentachloropyridine (PCP), phthalic anhydride (PA), and 1,2-dicyanobenzene (DCB) are regarded as the host species, respectively. For all three systems, they all show two main emission bands at 400 nm and 540 nm under steady-state UV irradiation; the former is verified as fluorescence, while the latter is phosphorescence as evidenced by their millisecond lifetimes of more than 360 ms (Fig. 61b–e). After ceasing the irradiation, yellow emission lasting for several seconds is observed in all systems, with the main peak centered at 540 nm and two shoulder peaks at 585 nm and 636 nm. This phosphorescent peak is completely overlapped with the 540 nm band in the steady-state PL spectrum and even displays an overwhelming ratio in the NA/PCP system with a total PLQY of 17.3%, indicating their high RTP efficiency. Interestingly, by comparing the PL spectra of single components, it is rational that both the fluorescence and ultralong yellow RTP of NA/PCP, NA/PA, and NA/DCB systems all come from the guest species of NA with fine vibration energy levels. To determine the role of the host species during the luminescence processes, they utilized a femtosecond transient absorption (fs-TA) technique (Fig. 61f). Upon excitation by the 267 nm pulsed laser, the excited state transitions of the pure PCP film from a singlet state (the narrow band centered 355 nm) to a triplet state (the newly formed broad peak at 460 nm) over hundreds of picoseconds. However, adding a trace amount of NA (forming NA/PCP film) drastically accelerates this process, reducing the transition time from 636 ps to just 126 ps. These results clearly indicate that NA, as the guest species, substantially speeds up the excited-state dynamics of PCP as the host species.

Fig. 61 Room-temperature phosphorescence (RTP) assisted by the formation of cluster excitons. (a) Chemical structures of the guest compound, NA, and host compounds of PCP, PA, and DCB. (b)–(d) Steady-state and delayed photoluminescence (PL) spectra of (b) NA/PCP, (c) NA/PA, and (d) NA/DCB systems under ambient conditions. Insert: corresponding photographs were taken under 254 nm UV irradiation or after ceasing the irradiation for different times. (e) The time-resolved PL decay of NA/PCP, NA/PA, and NA/DCB systems detected at 540 nm and around 400 nm, respectively. (f) Femtosecond transient absorption (fs-TA) spectra of the PCP film and NA/PCP film at different delay times after 267 nm excitation, and the normalized fs-TA spectra of the NA/PCP film with a delay time of 126 ps and PCP film with delay times of 144 ps and 636 ps. (g) Illustration diagram of emission from pure PCP and NA/PCP host/guest systems with cluster excitons. (h) The Jablonski diagram and excited-state processes of the excitation and emission from pure PCP and NA/PCP host/guest systems with cluster excitons. Reproduced with permission from ref. 464. Copyright 2019, Springer Nature.

Based on the above analysis, a model of cluster exciton in organic guest/host systems is proposed, where a molecular cluster of PCP and NA is formed to participate in the excited states (Fig. 61g). Different from the single-component system, the formed cluster in resonance (energetically close) is excited under irradiation to form transient species, which undergoes accelerated ISC to the triplet state and further rapidly decay into the more stable and localized triplet state of NA, prompting the ultralong yellow RTP(Fig. 61h). Crucially, several prerequisites should be considered to select host and guest molecules in the cluster exciton model. First, the energy level of the excited cluster should be close to either the host/guest singlet or guest triplet state to enable efficient radiationless transition. Second, the host and guest molecules should be in electronic matching (with similar electron-donating or withdrawing ability) to avoid nonradiative charge transfer. Third, the melting point of the host species should be sufficiently higher than room temperature to suppress nonradiative decay.

Another multi-component RTP system was built using the simplest polycyclic aromatic hydrocarbon, naphthalene (NL), as the guest and a series of benzene derivatives as the host to verify this cluster exciton mechanism (Fig. 62).⁴⁷⁵ As expected, naphthalene and all benzene derivatives (including 1,4-dimethoxybenzene, 1,4-dichlorobenzene, 1,2,4,5-tetrachlorobenzene, and 1,4-dibromobenzene) themselves only show weak or very faint fluorescence without phosphorescence. By the melt-casting process of NL and different hosts at a mass ratio of 1/100, several guest/host samples were created (Fig. 62a–e). With the halogen-free 1,4-dimethoxybenzene, the guest/host mixture exhibits weak blue fluorescence and a faint green afterglow lasting for ∼8 s with a low RTP quantum yield of ∼1%. This indicates that 1,4-dimethoxybenzene only provides an inelastic matrix but cannot significantly enhance ISC in NL. In contrast, halogen-containing guest/host systems can produce much more intense green RTP, with the RTP band of NL in these host matrices dominating the steady-state spectra. Among them, NL/1,4-dichlorobenzene shows the brightest emission, the longest afterglow (∼11 s), the highest RTP quantum yield (15.6%), and the longest RTP lifetime (760 ms) (Fig. 62f). Although the above result indicates that the halogen-containing hosts play a critical role in promoting RTP efficiency of NL in the cluster exciton mechanism, there is no quantitative correlation with the external heavy atom effect (e.g., atomic number or count of halogen atoms). Instead, the RTP efficiency is believed to depend on how closely the NL π-electron cloud can approach the halogen nucleus on the host molecule. In addition, since NL is a ubiquitous air pollutant that widely exists in our daily life, this highly efficient system was successfully applied as a specific and ultrasensitive “turn-on” RTP sensor for detecting NL vapor at room temperature (Fig. 62g). Solid 1,4-dichlorobenzene is fabricated as the sensor, which is initially non-luminescent. It rapidly develops intense, long-lasting green RTP upon exposure to sublimed NL molecules, enabling the creation of a simple visual detection device.

Fig. 62 Room-temperature phosphorescence (RTP) assisted by the naphthalene-based cluster excitons. (a)–(e) The steady-state and delayed photoluminescence of naphthalene-based systems with different host matrices. Inset: Chemical structure of the host molecule and the photograph of solid-state samples before and after ceasing UV irradiation. (f) Time-resolved decay curves of naphthalene/host systems measured at the maximum emission peak. (g) The diagram of the detection device and process for naphthalene utilizing the “turn-on” RTP phenomenon, and the RTP photographs of the 1,4-dichlorobenzene-based sensor after non-contact exposure to naphthalene for 2 hours in a sealed container. Reproduced with permission from ref. 475. Copyright 2022, Elsevier.

6. Through-space interactions for charge transport

Charge transport in organic materials, characterized by the hopping of charge carriers through localized states, serves as the operational foundation for numerous electronic and optoelectronic devices.^476–480 The efficiency of this process, quantified by charge carrier mobility, has long been pursued through molecular designs centered on TBC.^481–483 In TBC-based systems, high-density charge delocalization is achieved via extended π-conjugated backbones formed by covalent linkages, where orbital overlap along the molecular skeleton creates a continuous pathway for charge migration.⁴⁸⁴ While this approach has yielded high-performance semiconductors, its inherent limitations are evident. Charge transport is highly sensitive to molecular conformation and conjugation length, and tuning electronic properties often requires laborious synthetic modification of the covalent framework.

By contrast, charge transport governed by TSI operates on a fundamentally different principle, leveraging non-covalent orbital overlap between spatially proximate molecular units.^485–489 Key interactions such as π–π stacking, σ–σ interactions, and other van der Waals contacts create efficient charge-hopping channels that are largely independent of the covalent backbone.^490–493 This paradigm shift enables distinct molecular design strategies. Rather than optimizing conjugation length and planar area, TSI-based design focuses on controlling intra- and inter-molecular arrangements and packing models to maximize orbital overlap across space. The comparative advantage of TSI lies in its decoupling of electronic conjugation from the covalent bond. Whereas TBC systems require rigid, planar structures to maintain conjugation, TSI-active materials can incorporate flexible, non-conjugated spacers while maintaining efficient charge transport through optimized spatial organization.^494–496 This architecture enables controllable multi-channel conduction pathways and enhanced responsiveness to external stimuli.^497,498 Moreover, the modular nature of the TSI design allows for precise property tuning by systematic variation of molecular shape and packing motifs, rather than through complex covalent synthesis.

By transcending the traditional reliance on covalent conjugation, TSI-based charge transport opens new avenues for developing adaptable, processable, and multifunctional organic electronic materials with tailored charge transport characteristics. In this section, the working mechanism of TSI and typical molecular design for charge transport will be introduced from aromatic and non-aromatic systems, as well as their potential applications in mechanical potentiometer, multi-valued logic gate, molecular circuit, and more.

6.1. Aromatic systems

In molecular electronics, charge transport has traditionally been dominated by TBC, where electrons delocalize along covalently linked π-conjugated backbones. This mechanism typically supports single-channel conductance along a linear pathway, with its efficiency fundamentally constrained by the molecular structure and conjugation quality.^501–503 For instance, linear polyphenylene-based molecular wires (PP-SM) exhibit conductance values around 2.45 nS, representing a benchmark for conventional TBC systems (Fig. 63a, right).⁴⁹⁹ By contrast, TSI offers a distinct transport mechanism, where non-covalent interactions, particularly π-orbital conjugation between spatially proximate aromatic rings, enable efficient electron transfer without direct covalent bonding. This interaction, driven by long-range Coulombic forces and optimized molecular arrangement, facilitates charge transport over extended distances and enables multi-channel conduction. A compelling example is hexaphenylbenzene-derived HPB(OM)₃-SM, which achieves a remarkable conductance of 12.28 nS, approximately five times higher than PP-SM (Fig. 63a, left). This enhancement stems from the synergistic cooperation of parallel transport pathways. While the molecular backbone maintains limited through-bond conduction, strong TSI between stacked aromatic rings provides additional, highly efficient channels for charge migration. The multi-channel nature of TSC not only significantly boosts conductance but also improves system robustness (Fig. 63b). When conformational distortions or structural disruptions impair through-bond pathways, through-space channels can compensate, maintaining efficient charge transport. This redundancy makes TSC-based molecular wires more resilient to structural fluctuations and external perturbations than their TBC counterparts.

Fig. 63 Multichannel conductance assisted by through-space interactions. (a) Left: A single-molecule wire model, namely HPB(OM)₃-SM, with both through-bond conjugation (TBC) and through-space interaction (TSI). Right: A traditional single-molecule wire model, namely PP-SM, with only TBC. (b) The illustration of the multichannel and single-channel conductance of two molecular wires with different characteristics of TBC and TSI. Reproduced with permission from ref. 499. Copyright 2018, American Chemical Society. (c) Chemical structure and crystal structure of o-PP-2SMe. (d) Schematic of a mechanical single-molecule potentiometer based on o-PP-2SMe. The tensile force from the Au tip movement alters the contact conformations of o-PP-2SMe, thereby modulating its conductance to control circuit current. This conformational change also affects the charge transport mechanism. The inset shows the circuit diagram incorporating the potentiometer. Reproduced with permission from ref. 500. Copyright 2021, Springer Nature.

Zhao et al. provided compelling experimental validation of these principles through ortho-linked phenylene systems.⁵⁰⁰ As shown in Fig. 63c, the molecular structure of o-PP-2SMe features an ortho-substituted phenylene backbone that enables conformational flexibility. This flexibility allows the molecule to adopt multiple stable states, with crystallographic data revealing close interplanar distances and sulfur–sulfur contacts conducive to through-space coupling. The interplay between these structural features creates a tunable multi-channel transport system, where conductance can be modulated across two orders of magnitude through conformational changes. This exceptional tunability forms the basis for a mechanical single-molecule potentiometer, where applied tensile force induces conformational transitions that systematically modulate conductance (Fig. 63d). The ability to precisely control current through molecular conformation, via both through-bond and through-space pathways, demonstrates the unique advantages of TSC-based designs for adaptive molecular electronics.^238,504,505

Building on the concept of conformation-dependent conductance, they further demonstrated how o-PP-2SMe foldamers exploit multiple conductive pathways to achieve multi-valued logic circuits by integrating both through-space and through-bond interactions (Fig. 64).⁵⁰⁶ The distinct quantum interference effects generated in these pathways enable precise control over charge transport. The chemical structure of o-PP-2SMe, with its inherent conformational flexibility, facilitates transitions between different electronic states under mechanical stretching. In the through-space pathway, orbital overlap between LUMO and LUMO+1 plays a decisive role in modulating quantum interference, switching between constructive quantum interference (CQI) and destructive quantum interference (DQI) to directly regulate conductance (Fig. 64a). Experimental evidence from conductance-displacement traces reveals characteristic conductance decay upon junction elongation, with f-34Th maintaining significantly higher conductance than f-34Fu throughout the mechanical deformation (Fig. 64b). This differentiation is quantitatively validated in one-dimensional conductance histograms, showing well-defined peaks at 10^{−4.68±0.03} G₀ for f-34Th and 10^{−4.95±0.07} G₀ for f-34Fu, consistent with theoretical predictions of their molecular designs (Fig. 64c).

Fig. 64 Multi-valued output of molecular junction via through-space and through-bond pathways. (a) The molecular design strategy with different heteroatom-participated aromatic rings to adjust through-space and through-bond pathways. The conversion between constructive quantum interference (CQI) and destructive quantum interference (DQI) is shown based on the overlap of molecular orbitals (LUMO and LUMO+1). (b) Conductance-displacement traces of f-34Th and f-34Fu under mechanical stretching, showing the variation in conductance with displacement. (c) 1D conductance histograms of f-34Th and f-34Fu, with white dashed lines labeling the conductance peaks, demonstrating the different conductance levels. (d) 2D histograms of f-34Th, showing detailed displacement-conductance data and a Gaussian fit for the relative displacement distribution. (e) 2D histograms of f-34Fu, indicating the molecular behavior during mechanical stretching; insets are the relative displacement distribution histogram with Gaussian fitting. (f) Schematic illustration of the molecular junction with multi-valued output logic based on o-PP-2SMe. Reproduced with permission from ref. 506. Copyright 2022, Springer Nature.

Further insights emerge from two-dimensional conductance-displacement histograms, which illustrate distinctive evolution patterns for each configuration during stretching (Fig. 64d and e). The Gaussian-fitted displacement distributions quantified the correlation between conformational changes and conductance modulation, highlighting the precise tunability of these systems. Building on these quantum interference characteristics, a multivalued logic system emerges where the synergistic through-space and through-bond interactions create four distinct quantum-interfered states (Fig. 64f). The through-space CQI-dominated configuration (f-34Th) enables efficient conductance corresponding to logic “2”, while the DQI-dominated state (f-34Fu) suppresses conductance to logic “1”. When through-space CQI combines with through-bond suppression (f-23Th), reduced conductance yields another logic “1” state, and simultaneous through-space DQI with through-bond suppression (f-23Fu) produces minimal conductance for logic “0”. Clearly, this system with both TSC and TBC enables the creation of molecular logic gates with multiple outputs, surpassing the binary logic limitations and paving the way for more complex molecular electronics.^507,508

Building upon the foundation of multi-pathway conduction in foldamers, Zhao and colleagues have recently advanced the theoretical understanding of electro-responsive quantum interference switching, demonstrating its practical implementation in volatile memory devices.⁵⁰⁹ As schematically illustrated in Fig. 65a, the applied electric field induces molecular polarization that alters the stacking geometry between heterocycle and benzene units in f-Fu and f-Th foldamers, triggering reversible transitions between distinct conductance states. Specifically, f-Fu exhibits a voltage-gated “turn-on” characteristic, switching from a low-conductance (LC) to a high-conductance (HC) state, while f-Th displays a complementary “turn-off” behavior from the HC to LC state. This bidirectional switching capability enables their potential application in true random number generators (TRNG) and biologically inspired axon-like voltage-gated channels.^510,511

Fig. 65 Voltage-triggered quantum interference switching assisted by through-space. (a) The illustration of polarization-induced electro-responsive quantum interference switching in foldamers and potential applications in true random number generators (TRNG) and axon-like voltage-gated channels. The low-conductance state is labeled as '0', and the high-conductance state as '1'; The green memristor represents f-Fu, while the blue one represents f-Th. (b) The chemical structures of benzene, thiophene, and furan, along with the electrostatic potential distribution (ESP) of benzene–benzene stacking and heterocycle-benzene stacking models, highlighting their dipole moments. (c) The chemical structures of f-Fu and f-Th, accompanied by their corresponding single-crystal structures. The torsion angles of the stacked arms and the distances between sulfur atoms of the thiomethyl groups are labeled. (d) and (e) The high-conductance (HC) and low-conductance (LC) states of f-Fu, observed through two-electrode electrochemical gating experiments under varying bias voltages. (f) Statistical conductance data for f-Fu, f-Th, and f-Ph under different applied biases in electrochemical gating measurements, with the switching ratios (f) indicated at −0.5 V. Reproduced with permission from ref. 509. Copyright 2023, Springer Nature.

In this system, the molecular origin of electro-responsiveness is elucidated through systematic structural analysis. The electrostatic potential distributions of benzene, thiophene, and furan reveal significantly larger dipole moments in the heterocycles that enhance their sensitivity to electric fields (Fig. 65b). This asymmetric dipole configuration in heterocycle-benzene stacking models is crucial for the observed quantum interference modulation. Further structural insights based on single-crystal structures of f-Fu and f-Th highlight key parameters of torsion angles (25.8° for f-Fu and 40.1° for f-Th) and inter-sulfur distances that collectively determine the TSI efficiency and junction dimensions (Fig. 65c). Experimental validation of the voltage-gated switching mechanism is provided through electrochemical gating studies. The HC state of f-Fu exhibits enhanced conductance with increasing negative bias, suggesting field-induced structural relaxation that optimizes the high-conductance pathway (Fig. 65d). Conversely, the LC state also becomes more conductive under negative bias, though to a lesser extent, confirming the bidirectional nature of the switching behavior (Fig. 65e). This contrasting response to electrochemical gating underscores the sophisticated tunability of quantum interference in these foldamer systems, establishing a solid foundation for their implementation in advanced molecular electronic devices.

In double-channel molecular circuits, particularly those involving charged systems, TSI and electrostatic gating effects operate synergistically to control electron transport.^512–515 The electrostatic gating effect enables one molecular channel to modulate the conductance of an adjacent channel through non-contact electrostatic interactions. This occurs as electrostatic coupling between parallel channels alters molecular orbital energies, effectively tuning their quantum states. Specifically, the gating effect can shift LUMO into resonance with the Fermi level, thereby reducing the tunneling barrier and significantly enhancing conductance. Unlike neutral systems, where channels operate independently, charged cyclophane systems exhibit strong inter-channel electrostatic coupling that intertwines their quantum states through TSI. This mutual dependence enables one channel to function as a “gate” that adjusts the orbital alignment of its counterpart, facilitating near-resonant transport and collectively enhancing conductance beyond what isolated channels could achieve.

In 2020, Chen et al. provided compelling experimental evidence using tetracationic cyclophane-based molecular junctions (Fig. 66a).⁵¹⁵ The intramolecular circuit of a double-channel cyclophane, where parallel conduction pathways are coupled via electrostatic interactions. Theoretical transmission spectra reveal distinct behaviors that, compared with the single-channel system (gray), the severed double-channel configuration (red) shows a gating-induced leftward shift, while the intact double-channel system (blue) exhibits CQI that enhances transmission (Fig. 66b). The strategic design of molecular structure highlights the double-channel system (1-DS) maximizes inter-channel electrostatic interactions compared to the single-channel counterpart (1-S) (Fig. 66c). With the experimental characterizations, the one-dimensional conductance histograms demonstrate an 8-fold enhancement in the most probable conductance for 1-DS (10^{−4.43±0.01} G₀, 2.88 nS) compared to 1-S (10^{−5.36±0.02} G₀, 0.34 nS) (Fig. 66d). In addition, two-dimensional conductance-displacement histograms show consistently higher conductance for 1-DS throughout junction elongation (Fig. 65e and f), with flicker noise analysis revealing distinct scaling exponents (G^1.4 for 1-S and G^2.0 for 1-DS). This differential noise signature indicates TSI-dominated transport in the double-channel system, contrasting with the through-bond transport mechanism predominant in the single-channel configuration. The higher noise power in 1-DS further supports the presence of chemical gating effects between the two parallel channels, which play a key role in the enhanced conductance observed in the double-channel system. Together, these results establish that the synergistic combination of TSI electrostatic gating and constructive quantum interference enables unprecedented conductance enhancement in charged molecular circuits, providing a foundation for advanced quantum devices and molecular electronics.

Fig. 66 Intramolecular circuits assisted by the inter-channel gating effect. (a) A schematic representation of an intramolecular circuit traversing a tetracationic double-channel cyclophane. (b) Model transmission spectra for single-channel (silver), double-channel (blue), and severed double-channel (red) circuits, illustrating a leftward shift induced by the gating effect and a rightward shift resulting from constructive quantum interference. HOMO refers to the highest occupied molecular orbital, while LUMO refers to the lowest unoccupied molecular orbital. (c) Molecular structures of two cationic molecules, along with their graphical representations, showing a single-channel configuration (1-S, left) and a double-channel configuration (1-DS, right), respectively. (d) 1D Conductance histograms for solvent (black), 1-S (gray), and 1-DS (blue). (e and f) 2D Conductance-displacement histograms for (e) 1-S and (f) 1-DS, respectively. Insets: 2D histograms of normalized flicker noise power plotted against average junction conductance for (e) 1-S and (f) 1-DS, respectively. Reproduced with permission from ref. 515. Copyright 2020, Elsevier.

6.2. Nonaromatic systems

Recent studies have revealed the unexpected capability of σ–σ TSI to mediate efficient charge transport in non-aromatic molecular systems, a role traditionally ascribed to π–π interactions in conjugated aromatic frameworks.^516–518 The relevant studies focus on non-conjugated molecules such as cyclohexanethiol and adamantane, showing that these molecules can exhibit electrical conductivity through σ–σ stacking interactions, despite lacking the typical aromatic conjugation. These interactions allow for through-space charge transport, facilitated by the proximity and overlap of Sigma bonds between adjacent molecules. Meanwhile, by comparing the behavior of non-aromatic systems with traditional π–π stacking systems, these studies highlight that TSI in non-conjugated molecules can achieve similar conduction efficiency, challenging the reliance on conjugated systems for charge transport.

The representative study by Hong et al. systematically compared conductance characteristics in π-conjugated and non-conjugated systems using break junction measurements, elucidating distinct electron transport mechanisms governed by π–π versus σ–σ stacking (Fig. 67).⁵⁰ The experimental setup based on scanning tunneling microscope break junction (STM-BJ) is constructed, depicting both π–π stacked benzenethiol junctions and σ–σ stacked cyclohexanethiol dimer configurations (Fig. 67a). As shown in Fig. 67b, the one-dimensional logarithmic conductance histogram of benzenethiol reveals two primary conductance features. The high-conductance state (GH) peaks at approximately 10^−3.10G₀, which corresponds to a conductance of 61.6 nS, while the low-conductance state (GL) is observed at 10^−4.40 G₀ (approximately 3.1 nS). The inset shows typical conductance traces, with black traces representing tunneling decay in a pure solvent, providing a baseline measurement without the target molecules. The corresponding two-dimensional conductance–distance histogram further resolves how these states with respect to junction elongation, and the inset displaying stretching distance distributions that reflect junction stability under mechanical deformation (Fig. 67c). Parallel measurements for cyclohexanethiol similarly identify two conduction states where the high-conductance state (GH) is seen at 10^−3.50G₀ (around 24.5 nS), while the low-conductance state (GL) peaks at 10^−5.30G₀ (about 0.4 nS) (Fig. 67d). The two-dimensional histogram delineates the evolution of these conductance features with junction displacement, while the stretching distribution inset provides mechanical insights into junction formation (Fig. 67e). Despite the absence of π-conjugation, cyclohexanethiol exhibits a well-defined high-conductance state attributable to σ–σ TSI, demonstrating efficiency comparable to certain π-systems. This comparative analysis underscores σ–σ TSI as a viable mechanism for charge transport in non-conjugated molecular junctions, offering an alternative design paradigm for molecular electronic devices that leverages TSI pathways.

Fig. 67 Supramolecular junctions via through-space σ–σ stacking. (a) The schematic of STM-BJ measurements for supramolecular junctions and the configuration of π–π stacked molecular junctions for benzenethiol and σ–σ stacked molecular junctions for cyclohexanethiol dimers. The atoms are color-coded as follows: red for sulfur, pink and blue for carbon, grey for hydrogen, and gold for gold atoms. (b) and (d) The one-dimensional logarithmic conductance histograms are presented for (b) benzenethiol and (d) cyclohexanethiol, with insets showing typical individual conductance traces. The histograms show two key features: GH for high-conductance and GL for low-conductance. The insets depict tunneling decay traces, where black traces represent the pure solvent conductance without the target molecules. (c) and (e) The two-dimensional (2D) conductance–distance histograms for (c) benzenethiol and (e) cyclohexanethiol are presented, with the color bar indicating the number of counts. The insets show the corresponding relative stretching distance distributions of the molecular junctions, offering insight into the mechanical behavior of the junctions during the measurement. Reproduced with permission from ref. 50. Copyright 2022, Springer Nature.

Recent research has expanded the scope of through-space charge transport to σ-delocalized systems based on non-carbon elements such as selenium (Se), where lone-pair electrons actively participate in σ-orbital formation. Unlike π-orbitals that align perpendicular to the charge transport direction, σ-orbitals are oriented parallel to the current flow, enabling highly efficient charge conduction along the transport axis (Fig. 68a). A representative system, hexakis(methylselanyl)benzene, features six Se atoms anchored to a benzene ring, each contributing lone-pair electrons to create a peripherally σ-delocalized electronic structure. This configuration fundamentally modifies electronic properties compared to conventional π-conjugated systems, with several occupied orbitals (i.e., from HOMO to HOMO-5) demonstrating significant electron delocalization across Se atoms that directly facilitates charge transport (Fig. 68b). Experimental evidence from single-molecule junction measurements confirms the efficacy of this design. The 2D conductance versus stretching distance histogram for hexaselenylbenzene reveals stable conductance plateaus characteristic of efficient charge transport through σ-delocalized pathways (Fig. 68c). Further insight comes from current–voltage characterization, where 10 000 individual measurements identify two distinct conductance states: a high-conductance state attributed to direct Au–Se bonding, and a low-conductance state associated with van der Waals contacts between methyl groups and gold electrodes (Fig. 68d). Thermopower analysis corroborates this interpretation, showing near-zero thermopower for the low state (indicative of weak van der Waals coupling) and significantly higher thermopower for the high state, confirming efficient charge transport through Au–Se covalent bonding (Fig. 68e). The coexistence of these states reflects different electrode connection geometries that modulate internal electron transport pathways, with Au–Se bonds providing superior conductance through enhanced electronic coupling.

Fig. 68 Charge transport of single-molecule junctions with a σ-delocalized system. (a) Schematic illustration of charge transport through π- and σ-delocalized systems. The p-orbitals are aligned perpendicular and parallel to the charge transport direction for π-delocalized and σ-delocalized systems, respectively. Chemical structure of hexakis(methylselanyl)benzene. (b) Optimized geometries and electronic distribution of different orbitals. (c) 2D histograms of conductance versus stretching distance traces for single-molecule junctions. (d) 2D map of the I–V curves of single-molecule junctions. (e) Distribution of thermopower (S) at ΔT = 10.6 K. The distribution consists of 5000 measurements. The S value was calculated as S = −V_th/ΔT. The distribution of S (red line) is fitted with two Gaussian functions. The dotted line is the total fitting result, and each Gaussian function is represented by blue and orange lines. The peak values are +16 (low state) and +50 µV K⁻¹ (high state). Reproduced with permission from ref. 519. Copyright 2024, American Chemical Society.

Due to the diversity of TSI, including van der Waals forces, hydrogen bonds, π–π stacking, and even σ–σ interactions, these noncovalent interactions create multiple pathways in the microstructures of materials, allowing electrons or charge carriers to migrate through different routes. The multi-channel conductivity characteristics enable charge transport not only through a single channel but also through multiple parallel channels, enhancing the efficiency of carrier migration and significantly improving conductivity. In the future, as the understanding of TSI deepens, material design will increasingly rely on the precise control of various types of interactions. Molecular engineering and nanotechnology will allow for the fine-tuning of interaction strength and arrangement, further enhancing the conductivity of materials. With the advancement of nanotechnology, regulating TSI at the nanoscale will become crucial. For example, designing these interactions at the nanoscale can significantly boost conductivity, driving advancements in fields such as nanoelectronics, nanosensors, and nanomaterials. Meanwhile, by designing eco-friendly materials with high conductivity, innovations in green energy and smart grid technologies can be accelerated. At the same time, it is important to strictly compare the advantages and limitations of through-space and through-bond systems, especially focusing on different application scenarios. In summary, TSI provides a versatile handle to regulate conductivity via structure-encoded coupling, and continued mechanistic and processing advances will broaden the palette of high-performance organic electronic materials across information, energy, and environmental technologies.

7. Conclusions and perspectives

Inspired by the widely existing luminescence and charge transport from these unconventional compounds with a non-conjugated skeleton, TSI and related optoelectronic materials have witnessed rapid development during the past twenty years.⁵²⁰ Different from traditional TBC-based theories, TSI is dependent on spatial orientation and interaction instead of covalent bonds. With partially or totally non-conjugated structures, these unconventional materials display high controllability of properties, stimuli responsiveness, and even biocompatibility. From through-space nuclear spin–spin coupling in NMR techniques to through-space orbital interactions in chemical reactions and noncovalent interactions in biological systems and supramolecules, the development of these non-conjugated optoelectronic materials endows TSI with a new mission and connotation that can modulate photophysical properties via changing nonbonded electronic structures. At present, no systematic understanding on the mechanism and structure–property relationship of TSI in optoelectronic materials exists. Based on this consideration, this review provides a comprehensive and detailed summary of how TSI promotes or regulates optoelectronic properties, including CL, TADF, RTP, and charge transport.

For CL, three systems from n-electron to π-electron and n/π-electron hybrid systems are systematically introduced, which also cover non-conjugated small molecules and polymers. TSI, especially TSC, endows non-conjugated compounds with a strong conjugation effect of electrons via spatial overlap in the excited state. In common, these CLgens exhibit typical features of clustered emitting species, unmatched absorption and excitation spectra, excitation-dependent emission, etc. For pure n-electron and π-electron materials, the emission wavelength of CL is almost limited within the blue-color region, which, to some extent, restricts their performance. With the help of both electron-rich heteroatoms and rigidity of π-electron skeletons, TSI could be strengthened and stabilized, resulting in n/π-electron hybrid systems that show tunable emission color from blue to red and even NIR and controllable PLQY. From typical CLgens examples, the working mechanism and photophysical picture of TSI for CL have been constructed, providing a complement to TBC-based theories. Some manipulating strategies of TSI and corresponding CL are also introduced, including conformation and electronic regulation, flexibility and rigidity balance, heteroatom effects, end-group effects, etc. With different features, these CL materials show potential applications in bioimaging, monitoring of biological processes, sensors, information encryption, and the emission-fiber industry. On the other hand, TSCT, as a type of TSI, plays an unreplaceable role in achieving TADF. With the complete separation of the HOMO and LUMO, TSCT endows these co-facially arranged small molecules and D–A copolymers with extremely small ΔE_st, efficient RISC processes, and relatively high radiative decay rates and overall PLQY. By carefully refining the distance, orientation, strength of donor and acceptor units, and rigidity of molecular skeletons, the strength and stability of TSCT could be modified, resulting in some TADF materials with tunable emission colors, white-light emission, narrow-band emission with high color purity, and even 100% PLQY. Some compounds with TSCT features have been verified to be excellent candidates for advanced OLEDs with high EQE, low-voltage operation, and efficiency. Since the hybrid n/π-electrons that are beneficial for ISC, TSI in these non-conjugated materials can promote the transition from the singlet state to triplet state at room temperature, resulting in RTP. Apart from natural amino acids and sodium alginate, some artificial systems without extended conjugation can also show RTP properties under the mechanism of TSI. Different strategies based on TSI, such as noncovalent conformation lock, inter-/intra-molecular anion–π⁺ modulation, through-space SOC, and TSCT regulation, have been developed to pursue highly efficient and long-lifetime RTP from these non-conjugated luminogens. It is noteworthy that, apart from single-component systems, the concept of cluster exciton is introduced to illustrate the role of host species and the observed RTP in multi-component systems.^521–523

In addition to different luminescence properties, TSI also plays a critical role in promoting charge transport of molecules. Different from conventional methods based on TBC, TSI endows these non-conjugated compounds with high controllability, multi-channel conduction pathways, enhanced responsiveness to external stimuli, as well as comparable conduction efficiency even without any π electrons. Several twisted foldamers are introduced to show multi-channel pathways, which could be adjusted by the types and connecting portions of central aromatic rings, showing potential applications in multi-valued logic circuits and volatile memory devices. Meanwhile, the electrostatic gating effect is also studied in charged systems to synergistically control the electron transport of supramolecular junctions. On the other hand, σ–σ TSI has recently been observed to show unexpected capabilities in mediating efficient charge transport in non-aromatic systems, which exhibits comparable conductance to traditional π–π interactions in conjugated aromatic frameworks, opening an avenue for molecule-based electronic materials.^524,525

With rapidly increasing examples of non-conjugated optoelectronic materials and the illustration of TSI, the working mechanisms and photophysical processes have become clearer, and some optoelectronic materials with excellent performance and application values have been constructed, which offer a great complement to the TBC-based theories and OFMs. Nevertheless, compared to the widely studied TBC systems, TSI-based mechanisms and corresponding OFMs are still in their infancy. Therefore, continuous and deeper exploration is expected to further uncover all aspects of TSI in optoelectronic materials. Accordingly, several perspectives are provided for future development:

(1) Elucidation of TSI working mechanisms and intrinsic features via advanced theoretical analysis. Compared to strong TBC, the relatively weak and often transient nature of TSI poses significant challenges for direct experimental characterization. While current theoretical approaches, such as frontier molecular orbitals (FMO), anisotropy of the induced current density (AICD), independent gradient model based on Hirshfeld partition (IGMH), atoms-in-molecules (AIM) analysis, natural transition orbitals (NTO), etc., have provided valuable insights into the forms and functions of TSI, more advanced methods are needed to fully unravel their fundamental features and working mechanisms.^526–528 New theoretical theories for noncovalent electron structures and novel models for aggregates and aggregate-like electronic clusters should be constructed and developed.^529–532 These advances will not only deepen the understanding of TSI but also establish a robust chemical and photophysical foundation for designing new optoelectronic materials.

(2) Real-time and high-resolution analysis of aggregate-state structure and dynamics.^533–535 Unlike conventional conjugated systems that often perform well in monodisperse states, non-conjugated materials typically require stable aggregate environments to stabilize TSI and achieve high luminescence performance. Thus, characterizing the precise structure of these aggregates, monitoring their dynamic evolution, and even visualizing electronic structures of TSI under varying external conditions will be essential. Thus, some advanced instrumental techniques, including real-space electron density imaging, scanning tunneling microscopy-based force spectroscopy, high-resolution transmission electron microscopy, stimulated emission depletion microscopy, small-angle X-ray scattering, single-crystal X-ray crystallography, etc., may play a critical role. Such studies will help clarify structure–property relationships and refine strategies for tuning TSI and optoelectronic properties of these materials.

(3) Constructing dynamically responsive TSI-based functional systems. Dynamic changes endow materials with versatile functions. The structural flexibility inherent in non-conjugated molecular frameworks offers unique opportunities for stimuli-responsive behavior.^536–538 However, most current TSI-based optoelectronic materials rely on static molecular conformations and predefined packing motifs, limiting their dynamic adaptability. Introducing molecular and supramolecular designs that respond to optical, electrical, magnetic, or mechanical stimuli with the help of molecular switches and artificial molecular motors will open pathways to highly adaptive and even biomimetic optoelectronic systems with on-demand performance modulation. Specifically, the switchable control of TSI and related optoelectronic properties will also become a breakthrough for non-conjugated materials to surpass traditional conjugated systems.

(4) Property optimization and functional demonstrations of TSI-based materials. Although some reported non-conjugated materials already rival conventional conjugated systems in CL, RTP, and TADF, most systems still operate in the visible or short-wavelength NIR region with limited quantum efficiency. Achieving near-unity quantum yields, emission in the second near-infrared window, long-lived RTP, or CPL remains challenging but intriguing. Modulating TSI between non-conjugated units, controlling aggregate structures,⁵³⁹ and integrating TBC- and TSI-based fragments may be helpful strategies to further optimize and achieve the above properties. In the context of charge transport, the multi-channel pathways enabled by the synergy of TSI and TBC offer new advantages, yet the realization of highly conductive or even quantum-effect-dominated molecular junctions based primarily on TSI requires deeper exploration.

(5) Artificial intelligence (AI)-accelerated investigation of TSI-based optoelectronic materials. Although extensive experimental and theoretical studies have established design strategies for TSI-based optoelectronic materials, their development and synthesis still largely rely on traditional trial-and-error approaches. In addition, the structure–property of non-conjugated structures is more complex than that of traditional conjugated counterparts. Leveraging the rapidly advancing technologies of AI, machine learning, and large language models offers the potential to uncover hidden modulation strategies from entirely new perspectives,^540–545 thereby accelerating the targeted synthesis and property-oriented design of TSI-based optoelectronic materials. Similar to many other fields in materials science, research on TSI-based optoelectronic materials currently faces challenges such as a limited number of reported materials and a lack of standardized evaluation protocols. Therefore, advancing AI and machine learning techniques tailored for small datasets could bring about transformative changes.

(6) Development of unique applications of TSI-based materials. While some reported materials already demonstrate promise in conventional applications such as bioprobes, OLEDs, information encryption, and molecular wires, their distinctive features, including inherent biocompatibility, sustainability, mechanical flexibility, multi-channel charge transport, etc., remain underexploited. Future opportunities may include biocompatible imaging probes, functional polymeric films for energy conversion and the semiconductor industry, TSI-based luminescent and optical devices, large-scale and low-cost fabrication using naturally derived precursors,^546,547 as well as stimulus-responsive molecular junctions or single-molecule wires for integrated chip architectures. Fully leveraging these intrinsic advantages will help establish a unique application landscape for TSI-based optoelectronic systems.

This review is believed to construct a clear structure–property relationship for TSI and further offer general strategies and guidance for the design and manipulation of non-conjugated organic luminescent and electronic materials. Meanwhile, the systematic construction of TSI-based photophysics will be promoted as a supplement to the well-established TBC-based theories for future organic functional materials.

Author contributions

J. Z. and H. Z. conceived and designed the content of the review. J. Z., Z. X., J. S., B. Z. T., and H. Z. participated in the discussion of figures and details. J. Z. and Z. X. wrote the original manuscript, and X. L., Q. X., H. Z., J. W. Y. L., and B. Z. T. revised the manuscript. All authors agree with the final version of the manuscript.

Conflicts of interest

The authors declare no competing interests.

Data availability

No primary research data, software, or code were generated or analyzed in this study. All materials discussed are derived from previously published sources.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (52503253; 22575211, and 524B2036); the Fundamental Research Funds for the Central Universities (226-2025-00091; 226-2025-00031; 2024BSSXM04); The Open Fund of the State Key Laboratory of Luminescent Materials and Devices (South China University of Technology); and the Seed Program of China Petroleum & Chemical Corporation (Sinopec).

References

1. L. A. Dobrzański, J. Mater. Process. Technol., 2006, 175, 133–148.
2. D. Raabe, Chem. Rev., 2023, 123, 2436–2608.
3. B. Tawiah, E. A. Ofori and S. C. George, in Nanotechnology in Societal Development, ed. S. C. George and B. Tawiah, Springer Nature Singapore, Singapore, 2024, pp. 1–64 DOI:10.1007/978-981-97-6184-5_1.
4. L. Himanen, A. Geurts, A. S. Foster and P. Rinke, Adv. Sci., 2019, 6, 1900808.
5. M. A. Little and A. I. Cooper, Adv. Funct. Mater., 2020, 30, 1909842.
6. F. Yang, S. Cheng, X. Zhang, X. Ren, R. Li, H. Dong and W. Hu, Adv. Mater., 2018, 30, 1702415.
7. W.-J. Liu, H. Jiang and H.-Q. Yu, Chem. Rev., 2015, 115, 12251–12285.
8. X. Zhu, J. Duan, J. Chen, R. Liu, Z. Qin, H. Chen and W. Yue, Angew. Chem., Int. Ed., 2024, 63, e202311879.
9. Y. Fu, J. Zhou, X. Zou, L. Shi, X. Zhang, A. M. Doze, M. F. Hilbers, W. J. Buma, J. Zhang and B. L. Feringa, J. Am. Chem. Soc., 2025, 147, 35903–35912.
10. K.-W. Lee, Y. Wan, Z. Huang, Q. Zhao, S. Li and C.-S. Lee, Adv. Mater., 2024, 36, 2306492.
11. H. Chen, Y. Cai, Y. Han and H. Huang, Angew. Chem., Int. Ed., 2024, 63, e202313634.
12. Z. Xie, D. Liu, C. Gao, X. Zhang, H. Dong and W. Hu, J. Am. Chem. Soc., 2025, 147, 2239–2256.
13. Z. Xie, D. Liu, C. Gao, H. Dong and W. Hu, Nat. Rev. Mater., 2024, 9, 837–839.
14. W. Li, Y. Li, Z. Song, Y.-X. Wang and W. Hu, Chem. Soc. Rev., 2024, 53, 10575–10603.
15. J. Zhu, J. Xia, Y. Li and Y. Li, ACS Appl. Mater. Interfaces, 2025, 17, 5595–5608.
16. M. Sharma, C. M. S. Negi, P. A. Alvi and S. K. Gupta, Smart Materials for Science and Engineering, 2024, pp. 17–37 DOI:10.1002/9781394186488.ch2.
17. Y. Huang, L. Ning, X. Zhang, Q. Zhou, Q. Gong and Q. Zhang, Chem. Soc. Rev., 2024, 53, 1090–1166.
18. K. Chen, I. Song, L. You and J. Mei, Acc. Chem. Res., 2025, 58, 24–35.
19. L. Lan and H. Zhang, Angew. Chem., Int. Ed., 2025, 64, e202509140.
20. S. Klassen, Sci. Educ., 2011, 20, 719–731.
21. T. Baumgartner and R. Réau, Chem. Rev., 2006, 106, 4681–4727.
22. A. Facchetti, Chem. Mater., 2011, 23, 733–758.
23. A. Mateo-Alonso, Chem. Mater., 2023, 35, 1467–1469.
24. F. Saal, F. Zhang, M. Holzapfel, M. Stolte, E. Michail, M. Moos, A. Schmiedel, A. M. Krause, C. Lambert, F. Wurthner and P. Ravat, J. Am. Chem. Soc., 2020, 142, 21298–21303.
25. J. Li, Q. Xia, J. Li, X. Yu, Z. Feng, Y. Zhang, T. Wu, Z. Xu, H. Lin and J. Qian, Aggregate, 2025, 6, e70148.
26. H. V. Babu, M. G. M. Bai and M. Rajeswara Rao, ACS Appl. Mater. Interfaces, 2019, 11, 11029–11060.
27. Z. Chen, Q. Gu, M. Li, W. Qiu, Y. Jiao, X. Peng, W. Xie, D. Liu, K. Liu, Z. Yang and S.-J. Su, Adv. Opt. Mater., 2024, 12, 2302503.
28. M. Xu, C. Wei, Y. Zhang, J. Chen, H. Li, J. Zhang, L. Sun, B. Liu, J. Lin, M. Yu, L. Xie and W. Huang, Adv. Mater., 2024, 36, 2301671.
29. G.-e Fu, H. Yang, W. Zhao, P. Samorì and T. Zhang, Adv. Mater., 2024, 36, 2311541.
30. K. Yang, Z. Li, Y. Huang and Z. Zeng, Acc. Chem. Res., 2024, 57, 763–775.
31. Y. Zhang, S. Fukuma, R. Shang and E. Nakamura, Nat. Synth., 2024, 3, 1349–1359.
32. X. Zhang, Z. Tong and S. Dong, Chin. J. Chem., 2025, 43, 536–542.
33. L. Zhao, H. Liu, W. Li, Y. Yang, X. He, Z. Zhang, Y. Zhao, Y. Yao, L. Sun, K. Yang and Z. Zeng, Angew. Chem., Int. Ed., 2025, 64, e202507603.
34. I. Pozo and L. Bogani, Trends Chem., 2024, 6, 581–595.
35. C. Li, A. Yu, W. Zhao, G. Long, Q. Zhang, S. Mei and C.-J. Yao, Angew. Chem., Int. Ed., 2024, 63, e202409421.
36. H. Qin, W. Li, S. Z. Li, Z. Gao, Q. J. Xie, W. J. Zhang, R. R. Tang, S. Gu, J. T. Tang and G. P. Yu, Polym. Int., 2024, 75, 19–33.
37. P. A. Kollman, Acc. Chem. Res., 1977, 10, 365–371.
38. A. Haque, K. M. Alenezi, M. S. Khan, W. Y. Wong and P. R. Raithby, Chem. Soc. Rev., 2023, 52, 454–472.
39. R. C. Ollier and M. J. Webber, Adv. Mater., 2025, 37, 2507397.
40. X. Tang, J. Pang, J. Dong, Y. Liu, X.-H. Bu and Y. Cui, Angew. Chem., Int. Ed., 2024, 63, e202406956.
41. T. Kim, G. Shin, T. Park and M. Kim, Adv. Funct. Mater., 2025, 35, 2412267.
42. Z. Wu, H. Qian, X. Li, T. Xiao and L. Wang, Chin. Chem. Lett., 2024, 35, 108829.
43. C.-Z. Liu, S.-S. Hou, C. Zhang, T. Wang and B. Zhai, Eur. J. Inorg. Chem., 2025, e202400652.
44. X. M. Zhang, Y. H. Bai, J. W. Deng, X. S. Zhong, J. S. Xiao, W. D. Xie and H. L. Wang, Aggregate, 2025, 6, e70063.
45. X. Zhang, Y. Bai, J. Deng, P. Zhuang and H. Wang, Aggregate, 2024, 5, e517.
46. J. Liu, H. Zhang, L. Hu, J. Wang, J. W. Y. Lam, L. Blancafort and B. Z. Tang, J. Am. Chem. Soc., 2022, 144, 7901–7910.
47. Z. Xiong, J. Zhang, L. Wang, X. Liu, J. Z. Sun, H. Zhang and B. Z. Tang, Chem, 2025, 11, 102299.
48. T. Zhu, T. Yang, Q. Zhang and W. Z. Yuan, Nat. Commun., 2022, 13, 2658.
49. Z. Zhang, Z. Xiong, J. Zhang, B. Chu, X. Liu, W. Tu, L. Wang, J. Z. Sun, C. Zhang, H. Zhang, X. Zhang and B. Z. Tang, Angew. Chem., Int. Ed., 2024, 63, e202403827.
50. A. Feng, Y. Zhou, M. A. Y. Al-Shebami, L. Chen, Z. Pan, W. Xu, S. Zhao, B. Zeng, Z. Xiao, Y. Yang and W. Hong, Nat. Chem., 2022, 14, 1158–1164.
51. H. Huang, L. Yang, A. Facchetti and T. J. Marks, Chem Rev, 2017, 117, 10291–10318.
52. Z. Wang, M. Huang, J. Dong, X. Wang, Y. Li, M. Sun and S. Chang, J. Phys. Chem. C, 2023, 127, 2518–2523.
53. M. Lian, Y. Mu, Z. Ye, Z. Lu, J. Xiao, J. Zhang, S. Ji, H. Zhang, Y. Huo and B. Z. Tang, Aggregate, 2024, 5, e560.
54. S. Furukawa, M. Fujita, Y. Kanatomi, M. Minoura, M. Hatanaka, K. Morokuma, K. Ishimura and M. Saito, Commun. Chem., 2018, 1, 60.
55. F. Cozzi, R. Annunziata, M. Benaglia, M. Cinquini, L. Raimondi, K. K. Baldridge and J. S. Siegel, Org. Biomol. Chem., 2003, 1, 157–162.
56. L. Viglianti, N. Xie, H. H. Y. Sung, A. A. Voityuk, N. L. C. Leung, Y. Tu, C. Baldoli, I. D. Williams, R. T. K. Kwok, J. W. Y. Lam, E. Licandro, L. Blancafort and B. Z. Tang, Angew. Chem., Int. Ed., 2020, 59, 8552–8559.
57. H. Zhang, X. Zheng, N. Xie, Z. He, J. Liu, N. L. C. Leung, Y. Niu, X. Huang, K. S. Wong, R. T. K. Kwok, H. H. Y. Sung, I. D. Williams, A. Qin, J. W. Y. Lam and B. Z. Tang, J. Am. Chem. Soc., 2017, 139, 16264–16272.
58. C. Zheng, L. Zhu, J. Zhang, X. Deng, Z. Zhong, X. He, Q. Zhang, J. Liu, M. Meng, J. W. Y. Lam, Y. Li, M. Chen and B. Z. Tang, Chem. Sci., 2025, 16, 20060.
59. G. Saielli, ChemPhysChem, 2024, 25, e202300963.
60. J.-C. Hierso, Chem. Rev., 2014, 114, 4838–4867.
61. H. Wu, H. Hanayama, M. Coehlo, Y. Gu, Z. H. Wu, S. Takebayashi, G. Jakob, S. Vasylevskyi, D. Schollmeyer, M. Klaui, G. Pieters, M. Baumgarten, K. Mullen, A. Narita and Z. Qiu, J. Am. Chem. Soc., 2024, 146, 7480–7486.
62. J. Hilton and L. H. Sutcliffe, Prog. Nucl. Magn. Reson. Spectrosc., 1975, 10, 27–39.
63. A. D. Buckingham and J. E. Cordle, J. Chem. Soc., Faraday Trans. 2, 1974, 70, 994–1004.
64. A. Saika and H. S. Gutowsky, J. Am. Chem. Soc., 2002, 78, 4818–4819.
65. B. J. Kimber, J. Feeney, G. C. Roberts, B. Birdsall, D. V. Griffiths, A. S. Burgen and B. D. Sykes, Nature, 1978, 271, 184–185.
66. F. B. Mallory, J. Am. Chem. Soc., 2002, 95, 7747–7752.
67. F. B. Mallory, E. D. Luzik, C. W. Mallory and P. J. Carroll, J. Org. Chem., 2002, 57, 366–370.
68. E. Fermi, Nature, 1930, 125, 16.
69. N. F. Ramsey, Phys. Rev., 1953, 91, 303–307.
70. A. C. Diz, R. H. Contreras, M. A. Natiello and H. O. Gavarini, J. Comput. Chem., 2004, 6, 647–651.
71. O. L. Malkina and V. G. Malkin, Angew. Chem., Int. Ed., 2003, 42, 4335–4338.
72. A. Soncini and P. Lazzeretti, J. Chem. Phys., 2003, 119, 1343–1349.
73. A. Soncini and P. Lazzeretti, J. Chem. Phys., 2003, 118, 7165–7173.
74. B. V. Cheesman and R. F. M. White, Can. J. Chem., 1984, 62, 521–525.
75. Y. Xiao, J. T. Mague and R. A. Pascal Jr., Angew. Chem., Int. Ed., 2018, 57, 2244–2247.
76. M. Dračínský, M. Buchta, M. Buděšínský, J. Vacek-Chocholoušová, I. G. Stará, I. Starý and O. L. Malkina, Chem. Sci., 2018, 9, 7437–7446.
77. M. Dračínský, P. Jansa and P. Bouř, Chem. Eur. J., 2012, 18, 981–986.
78. E. I. Brodskaya, G. V. Ratovskii and G. V. Mikhail, Russ. Chem. Rev., 1993, 62, 919.
79. R. Hoffmann, Acc. Chem. Res., 1971, 4, 1–9.
80. S. A. Cowling, R. A. W. Johnstone, A. A. Gorman and P. G. Smith, J. Chem. Soc., Chem. Commun., 1973, 627–628.
81. V. A. Bren, D. D. Alexander, I. M. Vladimir and V. A. Chernoivanov, Russ. Chem. Rev., 1991, 60, 451.
82. F. Hemauer, H.-P. Steinrück and C. Papp, ChemPhysChem, 2024, 25, e202300806.
83. A. S. Aslam, M. Shamsabadi, R. J. Salthouse, J. Andréasson and K. Moth-Poulsen, ChemSusChem, 2025, 18, e202501005.
84. R. Hoffmann, Pure Appl. Chem, 1971, 28, 181–194.
85. J. Cerny and P. Hobza, Phys. Chem. Chem. Phys., 2007, 9, 5291–5303.
86. M. Juanes, R. T. Saragi, W. Caminati and A. Lesarri, Chem. Eur. J., 2019, 25, 11402–11411.
87. K. T. Mahmudov, M. N. Kopylovich, M. F. C. Guedes da Silva and A. J. L. Pombeiro, Coord. Chem. Rev., 2017, 345, 54–72.
88. C. R. Martinez and B. L. Iverson, Chem. Sci., 2012, 3, 2191–2201.
89. M. H. Sak and E. N. Jacobsen, Chem. Rev., 2025, 125, 8319–8369.
90. J. Gutiérrez-Flores, E. Ramos, C. I. Mendoza and E. Hernández-Lemus, Biophys. Chem., 2018, 233, 26–35.
91. S. Jena, J. Dutta, K. D. Tulsiyan, A. K. Sahu, S. S. Choudhury and H. S. Biswal, Chem. Soc. Rev., 2022, 51, 4261–4286.
92. F. Blanco, B. Kelly, G. Sánchez-Sanz, C. Trujillo, I. Alkorta, J. Elguero and I. Rozas, J. Phys. Chem. B, 2013, 117, 11608–11616.
93. A. J. Néel, M. J. Hilton, M. S. Sigman and F. D. Toste, Nature, 2017, 543, 637–646.
94. A. K. Wortman and C. R. J. Stephenson, Chem, 2023, 9, 2390–2415.
95. Y. Chen, Q. Zhen, F.-J. Meng, P. Yu and C. Xu, Chem. Rev., 2024, 124, 13370–13396.
96. C. A. Hunter, C. M. Low, J. G. Vinter and C. Zonta, J. Am. Chem. Soc., 2003, 125, 9936–9937.
97. H. Kawashima, S. Ukai, R. Nozawa, N. Fukui, G. Fitzsimmons, T. Kowalczyk, H. Fliegl and H. Shinokubo, J. Am. Chem. Soc., 2021, 143, 10676–10685.
98. Y. Morisaki and Y. Chujo, Prog. Polym. Sci., 2008, 33, 346–364.
99. S. Y. Yang, Y. K. Qu, L. S. Liao, Z. Q. Jiang and S. T. Lee, Adv. Mater., 2022, 34, e2104125.
100. Y. Morisaki and Y. Chujo, Angew. Chem., Int. Ed., 2006, 45, 6430–6437.
101. I. Ciofini, T. Le Bahers, C. Adamo, F. Odobel and D. Jacquemin, J. Phys. Chem. C, 2012, 116, 11946–11955.
102. D. Jacquemin, T. L. Bahers, C. Adamo and I. Ciofini, Phys. Chem. Chem. Phys., 2012, 14, 5383–5388.
103. A. Batra, G. Kladnik, H. Vázquez, J. S. Meisner, L. Floreano, C. Nuckolls, D. Cvetko, A. Morgante and L. Venkataraman, Nat. Commun., 2012, 3, 1086.
104. W. Tu, Z. Xiong, L. Wang, J. Zhang, J. Z. Sun, H. Zhang and B. Z. Tang, Sci. China Chem., 2024, 67, 3121–3130.
105. W. Zeng, S. Gong, C. Zhong and C. Yang, J. Phys. Chem. C, 2019, 123, 10081–10086.
106. L. Zheng, N. F. Polizzi, A. R. Dave, A. Migliore and D. N. Beratan, J. Phys. Chem. A, 2016, 120, 1933–1943.
107. J. R. Lakowicz, in Principles of Fluorescence Spectroscopy, ed. J. R. Lakowicz, Springer, US, Boston, MA, 1999, pp. 1–23 DOI:10.1007/978-1-4757-3061-6_1.
108. K. Chen, X. Liu, X. Li, J. Zhang, J. Z. Sun, H. Zhang, X.-H. Zhang and B. Z. Tang, Sci. China Chem., 2025, 68, 6693–6698.
109. F. Bacon, The Advancement of Learning, Macmillan and Co., London, 1605.
110. S. S. Yanari, F. A. Bovey and R. Lumry, Nature, 1963, 200, 242–244.
111. D. L. Richter-Egger, J. C. Landry, A. Tesfai and S. A. Tucker, J. Phys. Chem. A, 2001, 105, 6826–6833.
112. C. L. Larson and S. A. Tucker, Appl. Spectrosc., 2001, 55, 679–683.
113. S. Chen, H. Bai, H. Ma and L. Han, Eur. Polym. J., 2024, 206, 112760.
114. Y. Gong, Y. Tan, J. Mei, Y. Zhang, W. Yuan, Y. Zhang, J. Sun and B. Z. Tang, Sci. China Chem., 2013, 56, 1178–1182.
115. T. W. Huang, Z. Qin, A. Sun, J. Tang and B. Tang, Acta Chim. Sin, 2013, 71, 979–990.
116. 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, 2019, 32, 275–292.
117. Q. Zhu, Y. Zhang, H. Nie, Z. Zhao, S. Liu, K. S. Wong and B. Z. Tang, Chem. Sci., 2015, 6, 4690–4697.
118. E. Zhao, J. W. Y. Lam, L. Meng, Y. Hong, H. Deng, G. Bai, X. Huang, J. Hao and B. Z. Tang, Macromolecules, 2014, 48, 64–71.
119. X. Chen, T. Yang, J. Lei, X. Liu, Z. Zhao, Z. Xue, W. Li, Y. Zhang and W. Z. Yuan, J. Phys. Chem. B, 2020, 124, 8928–8936.
120. X. Dou, Q. Zhou, X. Chen, Y. Tan, X. He, P. Lu, K. Sui, B. Z. Tang, Y. Zhang and W. Z. Yuan, Biomacromolecules, 2018, 19, 2014–2022.
121. L.-L. Du, B.-L. Jiang, X.-H. Chen, Y.-Z. Wang, L.-M. Zou, Y.-L. Liu, Y.-Y. Gong, C. Wei and W.-Z. Yuan, Chin. J. Polym. Sci., 2019, 37, 409–415.
122. M. Li, X. Li, X. An, Z. Chen and H. Xiao, Front Chem, 2019, 7, 447.
123. Q. Zhou, T. Yang, Z. Zhong, F. Kausar, Z. Wang, Y. Zhang and W. Z. Yuan, Chem. Sci., 2020, 11, 2926–2933.
124. Q. Zhou, T. Yang, Z. Zhong, F. Kausar, Z. Wang, Y. Zhang and W. Z. Yuan, Chem. Sci., 2020, 11, 2926–2933.
125. X. Chen and W. Z. Yuan, in Encyclopedia of Aggregation-Induced Emission, ed. B. Z. Tang, Z. Zhao and Z. Qiu, Springer Nature Singapore, Singapore, 2025, pp. 1–12 DOI:10.1007/978-981-97-1574-9_8-1.
126. H. Zhang and B. Z. Tang, Handbook of Aggregation-Induced Emission, 2022, pp. 153–175 DOI:10.1002/9781119643098.ch5.
127. J. Zhou, L. Gan and J. Huang, Curr. Eng. Lett. Rev., 2025, 2, E2666948X375153.
128. J. Zhang, Z. Xiong, H. Zhang and B. Z. Tang, Nat. Commun., 2025, 16, 3910.
129. J. Mei, Y. Hong, J. W. Lam, A. Qin, Y. Tang and B. Z. Tang, Adv. Mater., 2014, 26, 5429–5479.
130. H. Zhang and B. Z. Tang, JACS Au, 2021, 1, 1805–1814.
131. Y.-Q. Han, Y. Xie, J. Zhang, S. Tan, H. Zhang, B. Z. Tang, J. L. Sessler and F. Huang, CCS Chem., 2024, 6, 1739–1747.
132. Y. Huang, J. Xing, Q. Gong, L.-C. Chen, G. Liu, C. Yao, Z. Wang, H.-L. Zhang, Z. Chen and Q. Zhang, Nat. Commun., 2019, 10, 169.
133. M. Lian, Y. Mu, Z. Ye, Z. Lu, J. Xiao, J. Zhang, S. Ji, H. Zhang, Y. Huo and B. Z. Tang, Aggregate, 2024, 5, e560.
134. J. Zhang, H. Zhang, J. W. Y. Lam and B. Z. Tang, Chem. Res. Chin. Univ., 2021, 37, 1–15.
135. V. Torregrosa-Rivero, M. Rosell, J. I. Castro Castro, F. D. L. Cruz-Martínez, J. A. Castro-Osma, A. Lara-Sánchez, S. Nonell, J. García-Martínez, R. Bresolí-Obach, E. Serrano and C. Martín, Carbon, 2025, 243, 120540.
136. K. Huang, T. Jiang, J. Liang, X. Chen, W. Liu, W. Cai, C. Zeng, J. Chen, J. Zhang, C. Cao and X. Feng, Chem. Eng. J., 2025, 523, 168741.
137. W. Xu, D. Hu, Z. Wang, G. Wang, K. Liu, J. Liang, R. Miao and Y. Fang, J. Phys. Chem. Lett., 2022, 13, 5358–5364.
138. J. Deng, H. Wu, W. Xie, H. Jia, Z. Xia and H. Wang, ACS Appl. Mater. Interfaces, 2021, 13, 39967–39975.
139. L. Yuan, H. Yan, L. Bai, T. Bai, Y. Zhao, L. Wang and Y. Feng, Macromol. Rapid Commun., 2019, 40, 1800658.
140. Z. Wang, X. Cheng, Y. Xie, S. Liu, M. Dong, J. Zhao, F. Liang, Z. An and W. Huang, CCS Chem., 2023, 5, 292–309.
141. S. Mandal, P. Durairaj, B. Kommula, S. Sarkar and S. Bhattacharyya, J. Phys. Chem. C, 2023, 127, 2430–2439.
142. X.-G. Yang, Z.-M. Zhai, X.-Y. Liu, J.-Y. Li, F.-F. Li and L.-F. Ma, Dalton Trans., 2020, 49, 598–602.
143. D. Ding, X. Xu, F. Li, A. Saparbaev, E. Zakhidov and M. Sun, Small, 2025, 21, e03142.
144. H. Zhang, Y. Guo, Z. Wu, Y. Wang, Y. Sun, X. Feng, H. Wang and G. Zhao, J. Lumin., 2021, 232, 117864.
145. Y. Liu, X. Wen, X. Chen, J. Chan, Z. Shuai, Z. Wang and W. Jiang, Angew. Chem., Int. Ed., 2026, 65, e202515124.
146. T. Chivers, Acc. Chem. Res., 1984, 17, 166–171.
147. C. Liu, Q. Cui, J. Wang, Y. Liu and J. Chen, Soft Matter, 2016, 12, 4295–4299.
148. Y. Wang, X. Bin, X. Chen, S. Zheng, Y. Zhang and W. Z. Yuan, Macromol. Rapid Commun., 2018, 39, 1800528.
149. Y. Takahashi and H. Tadokoro, Macromolecules, 2002, 6, 672–675.
150. C. Sun, X. Jiang, B. Li, S. Li and X. Z. Kong, ACS Sustainable Chem. Eng., 2021, 9, 5166–5178.
151. Y. Delgado Arcaño, O. D. Valmaña García, D. Mandelli, W. A. Carvalho and L. A. Magalhães Pontes, Catal. Today, 2020, 344, 2–14.
152. L. Pastor-Pérez, Y. Chen, Z. Shen, A. Lahoz and S. E. Stiriba, Macromol. Rapid Commun., 2007, 28, 1404–1409.
153. N. Jiang, C. Y. Zhu, K. X. Li, Y. H. Xu and M. R. Bryce, Macromolecules, 2024, 57, 5561–5577.
154. Q. Li, Y. Qiao, Z. Li, R. Wu, Y. Zhang, Y. Zhao, W. Tian, R. Wang, W. Feng and H. Yan, J. Mater. Chem. B, 2025, 13, 11687–11695.
155. D. A. Tomalia, B. Klajnert-Maculewicz, K. A. M. Johnson, H. F. Brinkman, A. Janaszewska and D. M. Hedstrand, Prog. Polym. Sci., 2019, 90, 35–117.
156. D. A. Tomalia, H. Baker, J. Dewald, M. Hall, G. Kallos, S. Martin, J. Roeck, J. Ryder and P. Smith, Polym. J., 1985, 17, 117–132.
157. S. Niu, H. Yan, Z. Chen, L. Yuan, T. Liu and C. Liu, Macromol. Rapid Commun., 2016, 37, 136–142.
158. Y. Zhao, B. Liu, R. Lou, Y. Qi, M. He, S. Long, W. Feng and H. Yan, Biomater. Adv., 2022, 137, 212848.
159. L. Bai, Y. Zhang, H. Yan and X. Liu, Biomacromolecules, 2022, 23, 4617–4628.
160. L. Bai, X. Liu, H. Yan, S. Zhao and X. Han, Langmuir, 2023, 39, 12053–12062.
161. Y. Feng, T. Bai, H. Yan, F. Ding, L. Bai and W. Feng, Macromolecules, 2019, 52, 3075–3082.
162. T. Bai, Y. Zhang, L. Wang, H. Yan and W. Zhou, Aggregate, 2023, 4, e267.
163. L. Bai, P. Yang, L. Guo, S. Liu and H. Yan, Biomacromolecules, 2022, 23, 1041–1051.
164. L. Bai, H. Yan, T. Bai, Y. Feng, Y. Zhao, Y. Ji, W. Feng, T. Lu and Y. Nie, Biomacromolecules, 2019, 20, 4230–4240.
165. R. Wu, S. Lian, Y. He, Z. Li, W. Feng, Y. Zhao and H. Yan, J. Mater. Chem. B, 2024, 12, 10584–10592.
166. R. Wu, Y. He, Z. Li, Z. Yu, X. Guo and H. Yan, Chem. Eur. J., 2025, 31, e202501133.
167. Z. Zhang, H. Zhang, M. Kang, N. Li, D. Wang and B. Z. Tang, Sci. China Chem., 2021, 64, 1990–1998.
168. H. Fang, Y. Chen, Z. Jiang, W. He and Z. Guo, Acc. Chem. Res., 2023, 56, 258–269.
169. Y. Li, Q. Chen, X. Pan, W. Lu and J. Zhang, Top. Curr. Chem., 2022, 380, 22.
170. A. Mizuno, R. Matsuoka, T. Mibu and T. Kusamoto, Chem. Rev., 2024, 124, 1034–1121.
171. Z. Zhou, K. Yang, L. He, W. Wang, W. Lai, Y. Yang, Y. Dong, S. Xie, L. Yuan and Z. Zeng, J. Am. Chem. Soc., 2024, 146, 6763–6772.
172. T. Huang, K. Yang, W. Hu, L. Feng, Z. Wu, H. Chen, J. Wang and Z. Zeng, Chem, 2025, 11, 102621.
173. X. Ai, E. W. Evans, S. Dong, A. J. Gillett, H. Guo, Y. Chen, T. J. H. Hele, R. H. Friend and F. Li, Nature, 2018, 563, 536–540.
174. H. Guo, Q. Peng, X.-K. Chen, Q. Gu, S. Dong, E. W. Evans, A. J. Gillett, X. Ai, M. Zhang, D. Credgington, V. Coropceanu, R. H. Friend, J.-L. Brédas and F. Li, Nat. Mater., 2019, 18, 977–984.
175. A. Abdurahman, T. J. H. Hele, Q. Gu, J. Zhang, Q. Peng, M. Zhang, R. H. Friend, F. Li and E. W. Evans, Nat. Mater., 2020, 19, 1224–1229.
176. Z. Wang, X. Zou, Y. Xie, H. Zhang, L. Hu, C. C. S. Chan, R. Zhang, J. Guo, R. T. K. Kwok, J. W. Y. Lam, I. D. Williams, Z. Zeng, K. S. Wong, C. D. Sherrill, R. Ye and B. Z. Tang, Mater. Horiz., 2022, 9, 2564–2571.
177. Z. Zhang, J. Zhang, J. Zhi Sun, H. Zhang, X. Zhang and B. Zhong Tang, Chem. Eur. J., 2025, 31, e202403493.
178. F. P. Schwarz and S. P. Wasik, Anal. Chem., 1976, 48, 524–528.
179. M. Suto, X. Wang, J. Shan and L. C. Lee, J. Quant. Spectrosc. Radiat. Transfer, 1992, 48, 79–89.
180. A. Braendle, A. Perevedentsev, N. J. Cheetham, P. N. Stavrinou, J. A. Schachner, N. C. Mösch-Zanetti, M. Niederberger and W. R. Caseri, J. Polym. Sci., Part B: Polym. Phys., 2017, 55, 707–720.
181. A. Perevedentsev, A. Francisco-López, X. Shi, A. Braendle, W. R. Caseri, A. R. Goñi and M. Campoy-Quiles, Macromolecules, 2020, 53, 7519–7527.
182. B. Ellis, P. G. White and R. N. Young, Eur. Polym. J., 1969, 5, 307–314.
183. P. Muller, Pure Appl. Chem., 1994, 66, 1077–1184.
184. J.-R. Mistry, S. Montanaro and I. A. Wright, Mater. Adv., 2023, 4, 787–803.
185. A. Braendle, A. Perevedentsev, N. J. Cheetham, P. N. Stavrinou, J. A. Schachner, N. C. Mösch-Zanetti, M. Niederberger and W. Caseri, Chimia, 2017, 71, 733.
186. W. Tu, Z. Xiong, Z. Zhang, J. Zhang, L. Wang, Y. Xie, Y. Wang, H. Zhang, J. Z. Sun and B. Z. Tang, Smart Mol., 2023, 1, e20220006.
187. A. Mellace, J. E. Hanson and J. Griepenburg, Chem. Mater., 2005, 17, 1812–1817.
188. Z. Pan, J. P. Wicksted and H. Liu, Phys. Rev. B, 1993, 48, 844–852.
189. R. Xu, H. Liu and W. Shi, J. Polym. Sci. B Polym. Phys., 2006, 44, 826–831.
190. L. Wang, Z. Xiong, J. Zhi Sun, F. Huang, H. Zhang and B. Zhong Tang, Angew. Chem., Int. Ed., 2024, 63, e202318245.
191. Y. Wang, M. Deng, X. Shen, J. Liu and Y. Xiao, J. Appl. Polym. Sci., 2025, 143, e58129.
192. E. Hontz, W. Chang, D. N. Congreve, V. Bulović, M. A. Baldo and T. Van Voorhis, J. Phys. Chem. C, 2015, 119, 25591–25597.
193. X.-K. Chen, D. Kim and J.-L. Brédas, Acc. Chem. Res., 2018, 51, 2215–2224.
194. H. Liu, Y. Fu, J. Zhang, X. Dong, N. Zheng, D. Yang, X. Qiao, D. Ma, J. Sun, J. W. Y. Lam, B. Z. Tang and Z. Zhao, Angew. Chem., Int. Ed., 2025, 64, e202511525.
195. J. Zhang, L. Hu, K. Zhang, J. Liu, X. Li, H. Wang, Z. Wang, H. H. Y. Sung, I. D. Williams, Z. Zeng, J. W. Y. Lam, H. Zhang and B. Z. Tang, J. Am. Chem. Soc., 2021, 143, 9565–9574.
196. Q. Xu, J. Zhang, J. Z. Sun, H. Zhang and B. Z. Tang, Nat. Photon., 2024, 18, 1185–1194.
197. Y. Wang, J. Zhang, Q. Xu, W. Tu, L. Wang, Y. Xie, J. Z. Sun, F. Huang, H. Zhang and B. Z. Tang, Nat. Commun., 2024, 15, 6426.
198. N. Su, J. Wang, Y. Yang, Z. Yan, L. Zhou, Y.-X. Zheng and J. Ding, Angew. Chem., Int. Ed., 2025, 64, e202512717.
199. A. P. Olbrykh, A. V. Tsorieva, V. M. Korshunov, A. F. Smol'yakov, I. A. Godovikov, A. A. Korlykov, I. V. Taydakov, A. A. Titov, O. A. Filippov and E. S. Shubina, Inorg. Chem. Front., 2025, 12, 812–820.
200. Y. Iida, Y. Shimomura, M. Tokita and G.-I. Konishi, Liq. Cryst., 2024, 51, 2032–2045.
201. A. Rajca, J. Am. Chem. Soc., 2002, 112, 5890–5892.
202. A. Rajca, S. Utamapanya and J. Xu, J. Am. Chem. Soc., 1991, 113, 9235–9241.
203. X. Cheng, H. Hu, Y. Wu, Z. Ma and Z. Ma, ACS Appl. Mater. Interfaces, 2022, 14, 56185–56192.
204. D. Schilter, Nat. Rev. Chem., 2017, 1, 0097.
205. Z. Xiong, J. Zhang, J. Z. Sun, H. Zhang and B. Z. Tang, J. Am. Chem. Soc., 2023, 145, 21104–21113.
206. A. Jakli, Liq. Cryst., 2021, 49, 1010–1019.
207. C. Du, Z. Wang, J. Chen, A. Martin, D. Raturi and M. Thuo, Angew. Chem., Int. Ed., 2022, 61, e202205251.
208. J. A. Pradeilles, S. Zhong, M. Baglyas, G. Tarczay, C. P. Butts, E. L. Myers and V. K. Aggarwal, Nat. Chem., 2020, 12, 475–480.
209. J. Gierschner, L. Lüer, B. Milián-Medina, D. Oelkrug and H.-J. Egelhaaf, J. Phys. Chem. Lett., 2013, 4, 2686–2697.
210. Y. Kong, Y.-C. Wang, X. Huang, W. Liang and Y. Zhao, Aggregate, 2024, 5, e395.
211. O. P. Dimitriev, J. Zirzlmeier, A. Menon, Y. Slominskii and D. M. Guldi, J. Phys. Chem. C, 2021, 125, 9855–9865.
212. Z. Xiong, J. Zhang, L. Wang, Y. Xie, Y. Wang, Z. Zhao, H. Zhang, J. Z. Sun, F. Huang and B. Z. Tang, CCS Chem., 2023, 5, 2832–2844.
213. B. Jiang, C.-W. Zhang, X.-L. Shi and H.-B. Yang, Chin. J. Polym. Sci., 2019, 37, 372–382.
214. K. S. Sharath Kumar, Y. R. Girish, M. Ashrafizadeh, S. Mirzaei, K. P. Rakesh, M. Hossein Gholami, A. Zabolian, K. Hushmandi, G. Orive, F. B. Kadumudi, A. Dolatshahi-Pirouz, V. K. Thakur, A. Zarrabi, P. Makvandi and K. S. Rangappa, Coord. Chem. Rev., 2021, 447, 214135.
215. Z. Hu, H. Zhang, Y. Chen, Q. Wang, M. R. J. Elsegood, S. J. Teat, X. Feng, M. M. Islam, F. Wu and B. Z. Tang, Dyes Pigm., 2020, 175, 108175.
216. D. D. La, S. V. Bhosale, L. A. Jones and S. V. Bhosale, ACS Appl. Mater. Interfaces, 2018, 10, 12189–12216.
217. H.-T. Feng, Y.-X. Yuan, J.-B. Xiong, Y.-S. Zheng and B. Z. Tang, Chem. Soc. Rev., 2018, 47, 7452–7476.
218. J. Sturala, M. K. Etherington, A. N. Bismillah, H. F. Higginbotham, W. Trewby, J. A. Aguilar, E. H. C. Bromley, A. J. Avestro, A. P. Monkman and P. R. McGonigal, J. Am. Chem. Soc., 2017, 139, 17882–17889.
219. A. N. Sussardi, G. F. Turner, J. G. Richardson, M. A. Spackman, A. T. Turley, P. R. McGonigal, A. C. Jones and S. A. Moggach, J. Am. Chem. Soc., 2023, 145, 19780–19789.
220. D. Xie, S. Sun, Q. Zhou, C. Wang, L. Cao, W. Zhang and G. Li, Chem. Commun., 2025, 61, 10408–10417.
221. D. Yan, Y. Huang, J. Zhang, Q. Wu, G. Song, J. Ji, Q. Jin, D. Wang and B. Z. Tang, J. Am. Chem. Soc., 2023, 145, 25705–25715.
222. S. Xu, Y. Duan and B. Liu, Adv. Mater., 2020, 32, e1903530.
223. A. T. Turley, P. K. Saha, A. Danos, A. N. Bismillah, A. P. Monkman, D. S. Yufit, B. F. E. Curchod, M. K. Etherington and P. R. McGonigal, Angew. Chem., Int. Ed., 2022, 61, e202202193.
224. J. Zhang, H. Zhang and B. Z. Tang, Matter, 2023, 6, 3702–3704.
225. G. F. Turner, N. Stapleton, J. Brookes, D. Spagnoli, A. N. Sussardi, A. C. Jones, P. R. McGonigal and S. A. Moggach, CrystEngComm, 2024, 26, 459–467.
226. Y. Wang, Z. Xiong, Y. Wang, A. Li, Y. Fang, L. Li, K. Wang, Q. Li and H. Zhang, Angew. Chem., Int. Ed., 2025, 64, e202420502.
227. Z. Zhuang, P. Shen, S. Ding, W. Luo, B. He, H. Nie, B. Wang, T. Huang, R. Hu, A. Qin, Z. Zhao and B. Z. Tang, Chem. Commun., 2016, 52, 10842–10845.
228. W. Luo, H. Nie, B. He, Z. Zhao, Q. Peng and B. Z. Tang, Chem. Eur. J., 2017, 23, 18041–18048.
229. J. M. Teng, D. W. Zhang and C. F. Chen, ChemPhotoChem, 2021, 6, e202100228.
230. M. Gong, L. Yuan, Y. X. Zheng and W. H. Zheng, Adv. Funct. Mater., 2024, 34, 2314205.
231. C. Liao, Y. Zhang, S. H. Ye and W. H. Zheng, ACS Appl. Mater. Interfaces, 2021, 13, 25186–25192.
232. H. Y. Woo, J. W. Hong, B. Liu, A. Mikhailovsky, D. Korystov and G. C. Bazan, J. Am. Chem. Soc., 2005, 127, 820–821.
233. A. Yanagawa, M. Tsuchiya, R. Inoue and Y. Morisaki, J. Mater. Chem. C, 2023, 11, 986–993.
234. K. A. Lyssenko, M. Y. Antipin and D. Y. Antonov, Chemphyschem, 2003, 4, 817–823.
235. I. Majerz and T. Dziembowska, ACS Omega, 2020, 5, 22314–22324.
236. I. Majerz and T. Dziembowska, J. Phys. Chem. A, 2016, 120, 8138–8147.
237. Z. Zhao, B. He, H. Nie, B. Chen, P. Lu, A. Qin and B. Z. Tang, Chem. Commun., 2014, 50, 1131–1133.
238. J. Li, P. Shen, Z. Zhao and B. Z. Tang, CCS Chem., 2019, 1, 181–196.
239. P. Shen, Z. Zhuang, X. F. Jiang, J. Li, S. Yao, Z. Zhao and B. Z. Tang, J. Phys. Chem. Lett., 2019, 10, 2648–2656.
240. B. Das, Chem. Commun., 2025, 61, 6391–6416.
241. J. Zhang, S. Ma, H. Fang, B. Xu, H. Sun, I. Chan and W. Tian, Mater. Chem. Front., 2017, 1, 1422–1429.
242. L. Guo, L. Wu, T. Jia, H. Zhang, J. Song, X. Xie, M. H. Jee, H. Ma, S. Liu, G. Lu, H. Y. Woo, Z. Wang, F. Gao and Y. Sun, Angew. Chem., Int. Ed., 2026, 65, e202516421.
243. R. Posey, N. Gill, D. Fernandez, L. G. Fonseca Dos Santos, H. Garza, J. Tran, B. Alfaro, N. Payne, T. A. Oyshi, C. Cashman, B. Salinas, I. Vasquez, A. Mdzinarishvili, I. Srivastava, U. Bickel, H. Lischka and J. Tropp, J. Mater. Chem. B, 2025, 13, 12918–12925.
244. M. Maafi, in Photokinetics: A New Perspective, ed. M. Maafi, Springer Nature Switzerland, Cham, 2025, pp. 73–117 DOI:10.1007/978-3-031-98985-8_4.
245. B. He, H. Nie, L. Chen, X. Lou, R. Hu, A. Qin, Z. Zhao and B. Z. Tang, Org. Lett., 2015, 17, 6174–6177.
246. S. Mukhopadhyay, S. P. Jagtap, V. Coropceanu, J. L. Bredas and D. M. Collard, Angew. Chem., Int. Ed., 2012, 51, 11629–11632.
247. B. S. Nehls, F. Galbrecht, A. Bilge, D. J. Brauer, C. W. Lehmann, U. Scherf and T. Farrell, Org. Biomol. Chem., 2005, 3, 3213–3219.
248. P.-C. Shen, Z.-Y. Zhuang, Z.-J. Zhao and B. Z. Tang, Chin. Chem. Lett., 2016, 27, 1115–1123.
249. B. He, S. Zhang, Y. Zhang, G. Li, B. Zhang, W. Ma, B. Rao, R. Song, L. Zhang, Y. Zhang and G. He, J. Am. Chem. Soc., 2022, 144, 4422–4430.
250. Z. Yu and S. Hecht, Chem. Commun., 2016, 52, 6639–6653.
251. T. Kinoshita, Y. Haketa, H. Maeda and G. Fukuhara, Chem. Sci., 2021, 12, 6691–6698.
252. P. Shen, S. Jiao, Z. Zhuang, X. Dong, S. Song, J. Li, B. Z. Tang and Z. Zhao, Angew. Chem., Int. Ed., 2024, 63, e202407605.
253. Z. Wang, Y. Li, H. Zhang, N. Jiang, J. Xu, D. Zhu and M. R. Bryce, ACS Appl. Polym. Mater., 2025, 7, 12337–12344.
254. N. D. Epiotis, W. R. Cherry, F. Bernardi and W. J. Hehre, J. Am. Chem. Soc., 1976, 98, 4361–4364.
255. M. D. Watson, A. Fechtenkotter and K. Mullen, Chem. Rev., 2001, 101, 1267–1300.
256. C. C. Price, Chem. Rev., 1941, 29, 37–67.
257. Z. Xiong, J. Zhang, Z. Zhang, L. Wang, S. Wei, X. Liu, J. Z. Sun, H. Zhang and B. Z. Tang, ACS Mater. Lett., 2024, 6, 3941–3950.
258. R. Huang, Q. Qiao, D. Seah, T. Shen, X. Wu, F. de Moliner, C. Wang, N. Ding, W. Chi, H. Sun, M. Vendrell, Z. Xu, Y. Fang and X. Liu, J. Am. Chem. Soc., 2025, 147, 5258–5268.
259. A. A. Aly and A. A. Hassan, in Advances in Heterocyclic Chemistry, ed. A. R. Katritzky, Academic Press, 2014, vol. 112, pp. 145–181.
260. K. Mahesh, S. Karpagam and K. Pandian, Top. Curr. Chem., 2019, 377, 12.
261. N. Obernikhina, M. Kachaeva, V. Shchodryi, Y. Prostota, O. Kachkovsky, V. Brovarets and Z. Tkachuk, Polycyclic Aromat. Compd., 2020, 40, 1196–1209.
262. J. Zhang, P. Alam, S. Zhang, H. Shen, L. Hu, H. H. Y. Sung, I. D. Williams, J. Sun, J. W. Y. Lam, H. Zhang and B. Z. Tang, Nat. Commun., 2022, 13, 3492.
263. K. D. Stigers, M. J. Soth and J. S. Nowick, Curr. Opin. Chem. Biol., 1999, 3, 714–723.
264. S. Kawano, T. Ida, M. Muraoka and M. Shizuma, ACS Mater. Lett., 2025, 7, 3868–3875.
265. Z. Zhang, J. Zhang, Z. Xiong, B. Chu, C. Zhang, J. Z. Sun, H. Zhang, X.-H. Zhang and B. Z. Tang, Angew. Chem., Int. Ed., 2023, 62, e202306762.
266. M. Asim, N. Saba, M. Jawaid, M. Nasir, M. Pervaiz and O. Y. Alothman, Curr. Anal. Chem., 2018, 14, 185–197.
267. Z.-L. Yu, Y.-C. Gao, B. Qin, Z.-Y. Ma and S.-H. Yu, Acc. Mater. Res., 2024, 5, 146–159.
268. H. Zheng and Y. Zu, J. Phys. Chem. B, 2005, 109, 16047–16051.
269. Y.-N. Jing, S.-S. Li, M. Su, H. Bao and W.-M. Wan, J. Am. Chem. Soc., 2019, 141, 16839–16848.
270. H. Xiao, Q.-X. Shi, Q. Li, H.-W. Cai, X.-L. Sun, W.-M. Wan and Q.-R. Qian, Chem. Eur. J., 2024, 30, e202303292.
271. B. Liu, H. Zhang, S. Liu, J. Sun, X. Zhang and B. Z. Tang, Mater. Horiz., 2020, 7, 987–998.
272. B. Li, B. Feng, J. Wang and Y. Qin, Luminescence, 2023, 39, e4630.
273. Y.-J. Sheng, M. Su, H. Xiao, Q.-X. Shi, X.-L. Sun, R. Zhang, H. Bao and W.-M. Wan, Chem. Eur. J., 2022, 28, e202201194.
274. L.-T. Wu, D.-S. Li, Q.-X. Shi, H. Xiao, Y.-J. Chen, X.-L. Sun, H. Nan and W.-M. Wan, Dyes Pigm., 2023, 219, 111554.
275. H. Xiao, Q. X. Shi, M. Su, X. L. Sun, H. Bao and W. M. Wan, ACS Macro Lett., 2023, 12, 40–47.
276. J.-Y. Mao, S.-X. Ye, Y. Yang, C.-J. Lu, X.-L. Sun, H. Xue and W.-M. Wan, Polym. Chem., 2025, 16, 3432–3442.
277. C.-X. Chen, H.-W. Cai, X.-L. Sun, W.-M. Wan and B.-Q. Huang, ChemPlusChem, 2025, 90, e202400633.
278. X. Chen, Z. He, F. Kausar, G. Chen, Y. Zhang and W. Z. Yuan, Macromolecules, 2018, 51, 9035–9042.
279. Y. Takai, K. Mori, T. Mizutani and M. Ieda, J. Polym. Sci., Polym. Phys. Ed., 2003, 16, 1861–1868.
280. D. H. Phillips and J. C. Schug, J. Chem. Phys, 1969, 50, 3297–3306.
281. Y. Du, H. Yan, W. Huang, F. Chai and S. Niu, ACS Sustainable Chem. Eng., 2017, 5, 6139–6147.
282. X. Liu, J. Mater. Sci. Appl., 2022, 6, 1–10.
283. B. Chu, H. Zhang, X. Zhang and B. Z. Tang, Acc. Chem. Res., 2025, 58, 1924–1935.
284. B. Chu, H. Zhang, L. Hu, B. Liu, C. Zhang, X. Zhang and B. Z. Tang, Angew. Chem., Int. Ed., 2022, 61, e202114117.
285. H. Bai, L. Han and H. Ma, Macromolecules, 2024, 57, 1819–1828.
286. B. Chu, H. Zhang, K. Chen, B. Liu, Q. L. Yu, C. J. Zhang, J. Sun, Q. Yang, X. H. Zhang and B. Z. Tang, J. Am. Chem. Soc., 2022, 144, 15286–15294.
287. X. Li, X. Liu, Y. Lv, Z. Zhang, K. Chen, B. Chu, J. Z. Sun, H. Zhang, X. Zhang and B. Z. Tang, Mater. Horiz., 2025, 12, 6195–6201.
288. X. Liu, B. Chu, Z. Xiong, B. Liu, W. Tu, Z. Zhang, H. Zhang, J. Z. Sun, X. Zhang and B. Z. Tang, Mater. Horiz., 2024, 11, 1579–1587.
289. Z. Zhang, W. Yan, D. Dang, H. Zhang, J. Z. Sun and B. Z. Tang, Cell Rep. Phys. Sci., 2022, 3, 100716.
290. B. Chu, X. Liu, Z. Xiong, Z. Zhang, B. Liu, C. Zhang, J. Z. Sun, Q. Yang, H. Zhang, B. Z. Tang and X.-H. Zhang, Nat. Commun., 2024, 15, 366.
291. B. Chu, X. Liu, X. Li, Z. Zhang, J. Z. Sun, Q. Yang, B. Liu, H. Zhang, C. Zhang and X. H. Zhang, J. Am. Chem. Soc., 2024, 146, 10889–10898.
292. C. Zhang, X. Geng, X. Zhang, Y. Gnanou and X. Feng, Prog. Polym. Sci., 2023, 136, 101644.
293. S. Wang, D. Wu, S. Yang, Z. Lin and Q. Ling, Mater. Chem. Front., 2020, 4, 1198–1205.
294. S. Ye and Y. Bao, Chem. Mater., 2024, 36, 5878–5896.
295. Y. Xie, W. Tu, Z. Xiong, D. Liu, J. Zhang, J. Z. Sun, Z. Zhao, D. Wang, F. Huang, H. Zhang and B. Z. Tang, Macromolecules, 2023, 56, 10082–10091.
296. R.-b Wang, W.-z Yuan and X.-y Zhu, Chin. J. Polym. Sci., 2015, 33, 680–687.
297. Y.-J. Tsai, C.-C. Hu, C.-C. Chu and T. Imae, Biomacromolecules, 2011, 12, 4283–4290.
298. W. I. Lee, Y. Bae and A. J. Bard, J. Am. Chem. Soc., 2004, 126, 8358–8359.
299. C.-C. Chu and T. Imae, Macromol. Rapid Commun., 2009, 30, 89–93.
300. Y. Ji, X. Yang and Y. Qian, RSC Adv., 2014, 4, 49535–49540.
301. X. Yan, M. Wei, X. Miao, X. Qiao, X. Zhang, P. Fu, X. Pang, M. Liu, Q. Zhao, S. Chen and Z. Cui, Polym. Chem., 2023, 14, 573–586.
302. H.-J. Li, Y. Chen, H. Wang, H. Wang, Q. Liao, S. Han, Y. Li, D. Wang, G. Li and Y. Deng, Adv. Funct. Mater., 2023, 33, 2302862.
303. W. Zhao, M. Gao, L. Kong, S. Yu, C. Zhao and C. Chen, Biomacromolecules, 2024, 25, 1897–1905.
304. J. Bao, C. Tong, M. He and H. Zhang, Luminescence, 2024, 39, e4594.
305. R. Ye, Y. Liu, H. Zhang, H. Su, Y. Zhang, L. Xu, R. Hu, R. T. K. Kwok, K. S. Wong, J. W. Y. Lam, W. A. Goddard and B. Z. Tang, Polym. Chem., 2017, 8, 1722–1727.
306. R. Liu, X. Chen, Z. Hayouka, S. Chakraborty, S. P. Falk, B. Weisblum, K. S. Masters and S. H. Gellman, J. Am. Chem. Soc., 2013, 135, 5270–5273.
307. Y. Li and W. A. Goddard, Macromolecules, 2002, 35, 8440–8455.
308. S. Dasgupta, W. B. Hammond and W. A. Goddard, J. Am. Chem. Soc., 1996, 118, 12291–12301.
309. H. Arimoto, J. Polym. Sci. A Gen. Pap., 1964, 2, 2283–2295.
310. Z. Zhang, Z. Xiong, B. Chu, Z. Zhang, Y. Xie, L. Wang, J. Z. Sun, H. Zhang, X. H. Zhang and B. Z. Tang, Aggregate, 2022, 3, e278.
311. M. Yang, Z. Wang, L. Zhang, J. Wang, X. Li, W. Zhang, W. Zhao, G. Shi, Y. He, Y. Zhang, X. Zhang, P. Fu, Z. Cui, X. Pang and M. Liu, J. Mater. Chem. C, 2025, 13, 13029–13041.
312. M. Luo, X.-H. Zhang and D. J. Darensbourg, Acc. Chem. Res., 2016, 49, 2209–2219.
313. B. Liu, in Encyclopedia of Aggregation-Induced Emission, ed. B. Z. Tang, Z. Zhao and Z. Qiu, Springer Nature Singapore, Singapore, 2025, pp. 1–10 DOI:10.1007/978-981-97-1574-9_99-1.
314. B. Liu, G. Huang, H.-L. Lu, K. Chen, Z. Yan, Y.-L. Wang, B. Chu, F.-D. Ren, Y. Yang and X.-H. Zhang, J. Mater. Chem. C, 2023, 11, 13142–13150.
315. H.-l Lu, G. Huang, Y.-L. Wang, Z. Yan, F.-d Ren, Y. Yang, B. Liu and B. Xu, Eur. Polym. J., 2024, 210, 112991.
316. B. Liu, B. Chu, Y.-L. Wang, Z. Chen and X.-H. Zhang, Adv. Opt. Mater., 2020, 8, 1902176.
317. Z. Feng, W. Zhao, Z. Liang, Y. Lv, F. Xiang, D. Sun, C. Xiong, C. Duan, L. Dai and Y. Ni, ACS Appl. Mater. Interfaces, 2020, 12, 11005–11015.
318. J. Shen, D. Chen, H. Deng, H. Zeng, D. Fang, X. Huang, C. Hu, Y. Zhou, H. Xu and C. Zhang, Prog. Org. Coat., 2024, 195, 108608.
319. B. Liu, Y.-L. Wang, W. Bai, J.-T. Xu, Z.-K. Xu, K. Yang, Y.-Z. Yang, X.-H. Zhang and B.-Y. Du, J. Mater. Chem. C, 2017, 5, 4892–4898.
320. N. Jiang, Y.-J. Meng, X. Pu, C.-Y. Zhu, S.-H. Tan, Y.-H. Xu, Y.-L. Zhu, J.-W. Xu and M. R. Bryce, ACS Mater. Lett., 2025, 7, 24–31.
321. C.-Y. Zhu, Y.-J. Meng, Y.-H. Xu, C.-Q. Xia, N. Jiang, J.-W. Xu and M. R. Bryce, J. Mater. Chem. C, 2025, 13, 8523–8530.
322. Z. Wang, F. Yuan, X. Li, Y. Li, H. Zhong, L. Fan and S. Yang, Adv. Mater., 2017, 29, 1702910.
323. B. He, J. Zhang, J. Zhang, H. Zhang, X. Wu, X. Chen, K. H. S. Kei, A. Qin, H. H. Y. Sung, J. W. Y. Lam and B. Z. Tang, Adv. Sci., 2021, 8, 2004299.
324. W. Zhu, T. Li, X. Zhang, B. Xia, J. Huang, S. Wang, Y. Wang and W. Dong, ACS Macro Lett., 2025, 14, 1060–1067.
325. Z. Guo, Y. Ru, W. Song, Z. Liu, X. Zhang and J. Qiao, Macromol. Rapid Commun., 2017, 38, 1700099.
326. T. Sarmah, S. Pal, D. Barman, D. Barman, B. G. Jaganathan and P. K. Iyer, Adv. Opt. Mater., 2025, 13, 2403060.
327. L. He, X. Nan, T. Wang, Z. Liu, C. Wang, T. Gu and P. Bai, Aggregate, 2026, 7, e70176.
328. Z. Xiong, J. Zhang, L. Wang, X. Liu, J. Z. Sun, H. Zhang and B. Z. Tang, Chem, 2025, 11, 102299.
329. J. Wu, Y. Wang, P. Jiang, X. Wang, X. Jia and F. Zhou, Nat. Commun., 2024, 15, 3482.
330. X. Chen, C. Hu, Y. Wang, T. Li, J. Jiang, J. Huang, S. Wang, T. Liu, W. Dong and J. Qiao, ACS Appl. Mater. Interfaces, 2023, 15, 23824–23833.
331. J. Zhang, H. Shen, Z. Xiong, L. Du, M. Li, X. Ou, X. Zhu, J. W. Y. Lam, T. M. Liu, C. Xu, H. Zhang and B. Zhong Tang, Angew. Chem., Int. Ed., 2025, 64, e202413751.
332. Z. Zhang, L. Zhu, J. Feng, H. Zhang, X. Zhang, J. Z. Sun and B. Z. Tang, Mater. Chem. Front., 2023, 7, 713–719.
333. B. Song, J. Zhang, J. Zhou, A. Qin, J. W. Y. Lam and B. Z. Tang, Nat. Commun., 2023, 14, 3115.
334. N. Jiang, H. Zhang and M. R. Bryce, Trends Chem., 2025, 7, 603–616.
335. B. Zhao, S. Yang, X. Yong and J. Deng, ACS Appl. Mater. Interfaces, 2021, 13, 59320–59328.
336. X. Dou, T. Zhu, Z. Wang, W. Sun, Y. Lai, K. Sui, Y. Tan, Y. Zhang and W. Z. Yuan, Adv. Mater., 2020, 32, 2004768.
337. M. Mahapatra, M. Bourguignon, P. Stiernet, S. Melo, B. Grignard, M. Vandevenne, M. Galleni and C. Detrembleur, Adv. Funct. Mater., 2025, e23095.
338. T. Xiong, L. Xu, Y. Qiu, L. Yang, J. Wang and X. Wang, Polym. Chem., 2025, 16, 4709–4717.
339. L. Li, G. Yang, T. Xiong, T. Duan, J. Wang and X. Wang, Chin. Chem. Lett., 2025, 36, 111374.
340. X. Chen, C. Hu, Y. Wang, T. Li, J. Jiang, J. Huang, S. Wang, W. Dong and J. Qiao, Adv. Sci., 2024, 11, e2304946.
341. 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.
342. 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.
343. J. M. Dos Santos, D. Hall, B. Basumatary, M. Bryden, D. Chen, P. Choudhary, T. Comerford, E. Crovini, A. Danos, J. De, S. Diesing, M. Fatahi, M. Griffin, A. K. Gupta, H. Hafeez, L. Hämmerling, E. Hanover, J. Haug, T. Heil, D. Karthik, S. Kumar, O. Lee, H. Li, F. Lucas, C. F. R. Mackenzie, A. Mariko, T. Matulaitis, F. Millward, Y. Olivier, Q. Qi, I. D. W. Samuel, N. Sharma, C. Si, L. Spierling, P. Sudhakar, D. Sun, E. Tankelevičiūtė, M. Duarte Tonet, J. Wang, T. Wang, S. Wu, Y. Xu, L. Zhang and E. Zysman-Colman, Chem. Rev., 2024, 124, 13736–14110.
344. N. Xue, H.-Y. Zhou, Y. Han, M. Li, H.-Y. Lu and C.-F. Chen, Nat. Commun., 2024, 15, 1425.
345. L. Ai, W. Xiang, J. Xiao, H. Liu, J. Yu, L. Zhang, X. Wu, X. Qu and S. Lu, Adv. Mater., 2024, 36, 2401220.
346. M. Banerjee and A. Anoop, Chem. Eur. J., 2024, 30, e202304206.
347. M. Hagai, N. Inai, T. Yasuda, K. J. Fujimoto and T. Yanai, Sci. Adv., 2024, 10, eadk3219.
348. F. Di Maiolo, D. K. A. Phan Huu, D. Giavazzi, A. Landi, O. Racchi and A. Painelli, Chem. Sci., 2024, 15, 5434–5450.
349. X. Li, Z. Kuang, Z. Situ, T. Huang, R. Jing, J. Kong, Y. Li, D. Song, M. Zhou, L. Duan and A. Xia, J. Phys. Chem. C, 2024, 128, 6612–6620.
350. Z. Pang, S. Shen, X. Xie, X. Lv, Y. Liu, J. Liu and Y. Wang, J. Mater. Chem. C, 2025, 13, 9232–9238.
351. W. Yang, C. Xie, T. Chen, X. Yin, Q. Lin, S. Gong, Z. Quan and C. Yang, Angew. Chem., Int. Ed., 2024, 63, e202402704.
352. Y. Wang, R. Feng, Y. Yu and Q. Song, CCS Chem., 2024, 6, 1165–1173.
353. J. Liu, S. Wang, W. Lv, J. Wu, Q. Ling and Z. Lin, Adv. Funct. Mater., 2024, 34, 2402578.
354. X. Yan, S.-Y. Song, S. Liang, K.-K. Liu, X.-T. Liu and C. Lu, Chem. Sci., 2025, 16, 16770–16779.
355. K. Pang, H. Liu, J. Fan and Y. Xu, Mol. Phys., 2025, 123, e2443533.
356. Y. Song, K. Zhang, Y. Li and L. He, Chem. Eng. J., 2025, 521, 166755.
357. Q. Feng, S. Zhu, B. Wang, F. Yu, H. Li, M. Yu, M. Xu and L. Xie, Adv. Funct. Mater., 2024, 34, 2312622.
358. Y. Sun, X.-F. Fu, C.-L. Hou, T.-T. Lin, D.-H. Zhang, J. Liu, J.-X. Hu, F.-L. Lin, L. Zhou, L. Meng, X.-L. Chen and C.-Z. Lu, Angew. Chem., Int. Ed., 2025, 64, e202511921.
359. P. Chen, C. Jiang, N. Li, X.-F. Song, X. Wan, H. Liu, J.-G. Yang and K. Li, Org. Electron., 2024, 133, 107114.
360. Z. Ma, Y. Wang, J. Pu, G. Li, Z. Mao, J. Zhao, Z. Yang, S.-J. Su and Z. Chi, Adv. Opt. Mater., 2024, 12, 2302386.
361. H. Li, C. Sun, M. Zhang, W. Yan and Z. Kang, Adv. Funct. Mater., 2024, 34, 2405669.
362. L. Zhang, K. Zhang, M. Wu, J. Ding, M. Feng, X. Li, S. Yang and H. Ma, Anal. Chem., 2025, 97, 3344–3351.
363. N. Acharya, M. Upadhyay, R. Deka and D. Ray, J. Mater. Chem. C, 2024, 128, 8750–8758.
364. X. Du, J. Chen, Z. Guo, L. Cao, X. Li, H. Lin, G. Yang, C. Zheng and S. Tao, Chem. Eng. J., 2025, 524, 169130.
365. J. F. Berbigier, L. C. da Luz and F. S. Rodembusch, Chem. Rec., 2024, 24, e202400146.
366. Z. Li, J. Lu and X. Li, Chem. Eur. J., 2024, 30, e202401001.
367. Y. Chu, X. Jin, G. Ji, P. Li, S. Xiao, W. Wang and Z. Song, ACS Nano, 2025, 19, 680–690.
368. X. Tang, L. S. Cui, H. C. Li, A. J. Gillett, F. Auras, Y. K. Qu, C. Zhong, S. T. E. Jones, Z. Q. Jiang, R. H. Friend and L. S. Liao, Nat. Mater., 2020, 19, 1332–1338.
369. S.-Y. Yang, Z.-Q. Feng, C. Liu, J. Zhang, Z. Deng, K. Zhang, F.-C. Kong, P. C. Y. Chow, R. T. K. Kwok, Z.-Q. Jiang, J. W. Y. Lam, L.-S. Liao and B. Z. Tang, Sci. China Chem., 2025 DOI:10.1007/s11426-025-2899-1.
370. S. Shi, X. Wang, L. Zhao, J. Lv, S. Shao and L. Wang, Adv. Opt. Mater., 2025, 13, 2402490.
371. Y. Song, K. Zhang, P. Wang, Y. Cao and L. He, Adv. Funct. Mater., 2025, 35, 2411957.
372. J. Wang, Y. Niu, Y. Jiang, C. Yao, Z. Dai, Y. Deng and J. Zhang, Chem. Eng. J., 2025, 524, 169199.
373. Y. Song, K. Zhang, Y. Li, Y. Shu, Y. Xiu and L. He, Adv. Funct. Mater., 2025, 35, 2422927.
374. Q. Li, H. Zhao, M. Li, Y. Liu, S. Yan and Z. Ren, Angew. Chem., Int. Ed., 2025, 64, e202506654.
375. H. Tsujimoto, D. G. Ha, G. Markopoulos, H. S. Chae, M. A. Baldo and T. M. Swager, J. Am. Chem. Soc., 2017, 139, 4894–4900.
376. S. Luo, J. Wang, N. Li, X. F. Song, X. Wan, K. Li and C. Yang, Angew. Chem., Int. Ed., 2023, 62, e202310943.
377. X. Wu, S. Ni, C.-H. Wang, W. Zhu and P.-T. Chou, Chem. Rev., 2025, 125, 6685–6752.
378. Y. Zou, M. Yu, Y. Xu, Z. Xiao, X. Song, Y. Hu, Z. Xu, C. Zhong, J. He, X. Cao, K. Li, J. Miao and C. Yang, Chem, 2024, 10, 1485–1501.
379. X.-F. Song, S. Luo, N. Li, X. Wan, J. Miao, Y. Zou, K. Li and C. Yang, Angew. Chem., Int. Ed., 2025, 64, e202413536.
380. T. Wang, X. Yin, X. Cao and C. Yang, Angew. Chem., Int. Ed., 2023, 62, e202301988.
381. G. Chen, J. Miao, X. Huang, Z. Zhang, Z. Xue, M. Huang, N. Li, X. Cao, Y. Zou and C. Yang, Light: Sci. Appl., 2025, 14, 81.
382. S. Li, J.-M. Jin, J.-X. Chen, R. Wang, B. Liu, Y.-Z. Su, X.-L. Liu, L. Xing, Y. Gong, Y.-L. Wu, W.-C. Chen, Y. Huo and H.-L. Zhang, ACS Appl. Mater. Interfaces, 2025, 17, 21560–21567.
383. L. Guo, W. Cui, L. Li, Y. Pu, K. Wang, P. Zheng, Y. Wang and C. Li, Adv. Mater., 2025, 37, 2500269.
384. R.-H. Liu, Z.-Q. Feng, S.-J. Ge, Y. Wang, Z.-H. Yu, J.-R. Wu, H.-Y. Yan, D.-Y. Zhou, L.-S. Liao and Z.-Q. Jiang, Angew. Chem., Int. Ed., 2025, 64, e202424950.
385. R.-J. Wang, F. Zheng, J. Hu, Y.-L. Wu, X.-L. Liu, J.-X. Chen, L. Xing, Q.-E. Wu, Y. Huo and W.-C. Chen, Chem. Eng. J., 2025, 514, 163327.
386. J. Hu, Q. Li, X. Wang, S. Shao, L. Wang, X. Jing and F. Wang, Angew. Chem., Int. Ed., 2019, 58, 8405–8409.
387. Y. Bao, Langmuir, 2024, 40, 3275–3282.
388. P. R. Bommireddy, N. Mameda, C. S. Musalikunta, Y.-W. Lee, Y. Suh, M. Godumala and S.-H. Park, Mater. Chem. Front., 2025, 9, 367–402.
389. J. Zhang, D. Li, W. Li, Y. Wu, X. Mu, C. Liu, K. Fang and Z. Ge, Sci. China Chem., 2024, 67, 1270–1276.
390. Y. Sang, R. Feng, Y. Wang and Q. Song, Polym. Chem., 2025, 16, 62–68.
391. S. Ye, N. Meftahi, I. Lyskov, T. Tian, R. Whitfield, S. Kumar, A. J. Christofferson, D. A. Winkler, C.-J. Shih, S. Russo, J.-C. Leroux and Y. Bao, Chem, 2023, 9, 924–947.
392. Q. Li, F. Huang, X. Xu, X. Wang, Z. Ma, Z. Quan, J. Liu, H. Tong, S. Shao and L. Wang, Chem. Commun., 2025, 61, 4864–4867.
393. Y. Xin, Y. Zhu, R. Chi, C. Duan, P. Yan, C. Han and H. Xu, Adv. Mater., 2023, 35, e2304103.
394. Z. Yin, Z. Wu and B. Liu, Adv. Mater., 2025, e2506549.
395. X. Zou, N. Gan, Y. Gao, L. Gu and W. Huang, Angew. Chem., Int. Ed., 2025, 64, e202417906.
396. Y. Yang, Q. Li and Z. Li, Mater. Chem. Front., 2025, 9, 744–753.
397. H. Chen, Y. Zhang, J. Shan, M. Dong, Z. Qian, A. Lv, H. J. Qian, H. Ma, Z. An, L. Gu and W. Huang, Angew. Chem., Int. Ed., 2025, 64, e202500610.
398. J. Shan, H. Chen, Y. Gao, M. Dong, A. Lv, H. Ma, Y. Zhao, W. Huang and L. Gu, Angew. Chem., Int. Ed., 2025, 64, e202512424.
399. F. Lu, X. Xu, W. Wan, X. Liang, J. Xia, W. Cao, Z. Chen and M. Hu, J. Mater. Chem. C, 2025, 13, 2091–2120.
400. Z.-A. Yan, C. Yin, H. Tian and X. Ma, Angew. Chem., Int. Ed., 2025, 64, e202417397.
401. J. Zeng, S. Song, Y. Fu, X. Peng, B. Z. Tang and Z. Zhao, Sci. Adv., 2025, 11, eadt7899.
402. J. Gu, W. Yuan, K. Chang, C. Zhong, Y. Yuan, J. Li, Y. Zhang, T. Deng, Y. Fan, L. Yuan, S. Liu, Y. Xu, S. Ling, C. Li, Z. Zhao, Q. Li, Z. Li and B. Z. Tang, Angew. Chem., Int. Ed., 2025, 64, e202415637.
403. X. Wei, J. Ye, L. Shi and R. Jia, Adv. Opt. Mater., 2025, 13, 2500106.
404. Y. Zhao, X. Ma, S. Cui, J. Qu, Y. Wen, K. Zhang, G. Li, Y. Zhang, X. Huang, B. Mei, T. Wang, P. Sun, J. Shi, B. Tong, H.-Y. Xie, Z. Liu, Z. Cai and Y. Dong, Angew. Chem., Int. Ed., 2026, 65, e202511784.
405. H. Guo, H. Cheng, R. Liu, X. Chen, L. Wang, C. Yang, S. Li, S. Liu, J. Li, Q. Pan, T. D. James and Z. Chen, Angew. Chem., Int. Ed., 2025, 64, e202421112.
406. M. He, C. Ding, H. Guo and Q. Li, Responsive Mater., 2024, 2, e20240014.
407. W. Zhao, Z. He and B. Z. Tang, Nat. Rev. Mater., 2020, 5, 869–885.
408. Q. Gao, B. Lü and F. Peng, Prog. Mater. Sci., 2025, 148, 101372.
409. C. Li, M. Wu, Z. Zhao, J. W. Y. Lam, B. Xu and B. Z. Tang, Chem, 2025, 11, 102654.
410. X. Huang, Y. Zhang, G. Li, T. Wang, P. Sun, J. Shi, B. Tong, J. Zhi, Z. Li, Z. Cai and Y. Dong, ChemPhotoChem, 2024, 9, e202400315.
411. Q. Peng, H. Ma and Z. Shuai, Acc. Chem. Res., 2021, 54, 940–949.
412. L. Hou, T. Wang, S. Yu, X. Xu and X. Yu, Aggregate, 2025, 6, e70071.
413. W. Yan, P. Dong, Q. Zhang, Y. Zhang, R. Deng, L. Zhu, T. Wang, J. Shi, J. Zhi, B. Tong, P. Sun, Z. Cai and Y. Dong, Coord. Chem. Rev., 2025, 535, 216629.
414. J. Zhang, Y. Jin, X. Lu, C. Sun, W. Ma, Y. Li, L. Zhang and R. Chen, Chem. Sci., 2025, 16, 7829–7837.
415. C. Sun, X. Zhang, Y. Yang, J. Fu and L. Xu, Int. J. Biol. Macromol., 2025, 307, 142328.
416. Z. Zhang, Y.-e Shi, Y. Liu, Y. Xing, D. Yi, Z. Wang and D. Yan, Chem. Eng. J., 2022, 442, 136179.
417. R. Yang, H. Su, J. Song, C. Xie, B. Chen, X. Li, J. Jiang, G. Zhang, Y. Luo, X. Zhang and B. Z. Tang, J. Am. Chem. Soc., 2025, 147, 37385–37397.
418. Y. Zuo, H. Yang, K. Wang, Y. Song, L. Jiang, W. Huang, X. Xue, Q. Jiang, B. Jiang and G. Zhang, Macromolecules, 2023, 56, 5854–5864.
419. Q. Gao, M. Shi, J. Rao, Z. Su, G. Chen, B. Lü, J. Bian and F. Peng, Adv. Funct. Mater., 2024, 34, 2403977.
420. F. Peng, Y. Chen, H. Liu, P. Chen, F. Peng and H. Qi, Adv. Mater., 2023, 35, 2304032.
421. Q. Gao, M. Shi, Z. Lü, Q. Zhao, G. Chen, J. Bian, H. Qi, J. Ren, B. Lü and F. Peng, Adv. Mater., 2023, 35, 2305126.
422. X. Chen, W. Luo, H. Ma, Q. Peng, W. Z. Yuan and Y. Zhang, Sci. China Chem., 2017, 61, 351–359.
423. Q. Zhou, Z. Wang, X. Dou, Y. Wang, S. Liu, Y. Zhang and W. Z. Yuan, Mater. Chem. Front., 2019, 3, 257–264.
424. N. Jiang, K.-X. Li, J.-J. Wang, C.-S. Li, X.-Y. Xu, Y.-H. Xu and M. R. Bryce, J. Mater. Chem. C, 2024, 12, 1040–1046.
425. Y. Wang, Y. Zhao, X. Chen, T. Li, X. Zhang, J. Huang, J. Qiao and W. Dong, Macromolecules, 2025, 58, 354–362.
426. Y. Wang, S. Tang, Y. Wen, S. Zheng, B. Yang and W. Z. Yuan, Mater. Horiz., 2020, 7, 2105–2112.
427. S. Y. Zheng, T. W. Zhu, Y. Z. Wang, T. J. Yang and W. Z. Yuan, Angew. Chem., Int. Ed., 2020, 59, 10018–10022.
428. Q. Zhang, Z. Yin, T. Zhu, W. Dong and W. Z. Yuan, Dyes Pigm., 2025, 241, 112904.
429. H. Wang, H. Shi, W. Ye, X. Yao, Q. Wang, C. Dong, W. Jia, H. Ma, S. Cai, K. Huang, L. Fu, Y. Zhang, J. Zhi, L. Gu, Y. Zhao, Z. An and W. Huang, Angew. Chem., Int. Ed., 2019, 58, 18776–18782.
430. Q. Wang, X. Dou, X. Chen, Z. Zhao, S. Wang, Y. Wang, K. Sui, Y. Tan, Y. Gong, Y. Zhang and W. Z. Yuan, Angew. Chem., Int. Ed., 2019, 58, 12667–12673.
431. H. Chen, Y. Sun, M. Liu, F. Li, Q. Peng and H. Huang, Angew. Chem., Int. Ed., 2023, 62, e202302629.
432. T. Dong, L. Lv, L. Feng, Y. Xia, W. Deng, P. Ye, B. Yang, S. Ding, A. Facchetti, H. Dong and H. Huang, Adv. Mater., 2017, 29, 1606025.
433. H. Li, Y. Wang, L. Yu, C. Liu, C. Zhou, S. Sun, M. Li, Y. Tao, G. Xie, H. Xu, W. Huang and R. Chen, Chem. Eng. J., 2021, 425, 131487.
434. S. Yu, A. Peng, S. Zhang and H. Huang, Sci. China Chem., 2018, 61, 1359–1367.
435. 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.
436. J. Li, Q. Yan, X. Zhai, G. Wang, X. Li and K. Zhang, Adv. Funct. Mater., 2025, 35, 2422121.
437. W. Dai, X. Yang, K. Lv, L. Li, Y. Peng, H. Ma and Z. An, Chem. Eur. J., 2024, 30, e202401882.
438. H. Li, H. Ma, P. Zhang, Z. An and X. He, Angew. Chem., Int. Ed., 2025, 64, e202419366.
439. T. Xu, C. Zhai, Z. Liu, X. Yang, S. Hu, Y. Shang, L. Yue, J. Dong, R. Liu, Q. Li, M. Yao and B. Liu, Nat. Commun., 2025, 16, 3550.
440. X. Li, Y. Wu, S. Zhang, B. Cai, Y. Gu, J. Song and H. Zeng, Adv. Funct. Mater., 2016, 26, 2435–2445.
441. W. Huang, X. Zhao, J. Zhang, L. Zhang, T. S. Cheung, Y. Xiao, F. Wang, Z. Zhao, S. Chen, L. Xu, Q. Shen and B. Z. Tang, Adv. Funct. Mater., 2025, e12647.
442. H. Shi, L. Song, H. Ma, C. Sun, K. Huang, A. Lv, W. Ye, H. Wang, S. Cai, W. Yao, Y. Zhang, R. Zheng, Z. An and W. Huang, J. Phys. Chem. Lett., 2019, 10, 595–600.
443. X. Ma, J. Wang and H. Tian, Acc. Chem. Res., 2019, 52, 738–748.
444. W. Dai, X. Niu, X. Wu, Y. Ren, Y. Zhang, G. Li, H. Su, Y. Lei, J. Xiao, J. Shi, B. Tong, Z. Cai and Y. Dong, Angew. Chem., Int. Ed., 2022, 61, e202200236.
445. Z. Hua, H. Yu, H. Gong, P. Li, T. Zhang, Z. Deng, W. Sun, Y. Tian and H. Fu, J. Phys. Chem. Lett., 2024, 15, 4729–4736.
446. P. Li, Q. Lv, C. Sun, P. Zhang, X. Wang, C. Yin, Y. Pan and R. Chen, J. Phys. Chem. Lett., 2024, 15, 9787–9794.
447. J. Yu, H. Ma, W. Huang, Z. Liang, K. Zhou, A. Lv, X. G. Li and Z. He, JACS Au, 2021, 1, 1694–1699.
448. Q. Zheng, Y.-K. Qu, P. Zuo, H.-T. Yuan, Y.-J. Yang, Y.-C. Qiu, L.-S. Liao, D.-Y. Zhou and Z.-Q. Jiang, Chem, 2025, 11, 102353.
449. F. Ma, B. Wu, S. Zhang, J. Jiang, J. Shi, Z. Ding, Y. Zhang, H. Tan, P. Alam, J. W. Y. Lam, Y. Xiong, Z. Li, B. Z. Tang and Z. Zhao, J. Am. Chem. Soc., 2025, 147, 10803–10814.
450. H. Zhu, J. Liu, Y. Wu, L. Wang, H. Zhang, Q. Li, H. Wang, H. Xing, J. L. Sessler and F. Huang, J. Am. Chem. Soc., 2023, 145, 11130–11139.
451. J. Li, G. Xing, J. Wu, Y. Zhang, J. Wei, S. Liu, Y. Ma and Q. Zhao, Laser Photonics Rev., 2023, 18, 2300704.
452. J. Liu, W. Xu, L. Gao, B. Li, J. Zhang, M. Zong, Y. Li, L. Chen, X. Yan and B. Wang, Chem. Eng. J., 2025, 521, 166239.
453. B. Li, Z. Yang, W. Gong, X. Chen, D. W. Bruce, S. Wang, H. Ma, Y. Liu, W. Zhu, Z. Chi and Y. Wang, Adv. Opt. Mater., 2021, 9, 2100180.
454. L. Han, Y. Lan, X. Zhang, L. Bu, J. Bai, J. Ma, Z. Ma, M. Chen, Y. Guan and Z. Ma, ACS Appl. Mater. Interfaces, 2025, 17, 25602–25612.
455. Q. Shi, N. Ding, Z. Wang, X. Gou, L. Peng, J. Ma and Y. Fang, J. Phys. Chem. Lett., 2024, 15, 2995–3001.
456. Y. Xie, S. Wu, Z. Zhu, J. Wang, Z. Kuang, L. Zhang, A. Abdurahman, Q. Peng and X. Ai, Angew. Chem., Int. Ed., 2025, 64, e202515346.
457. J. Fang, P. Li, X. Xue, H. Li, D. Cui, J. Zhang, Y. Wang, T. Debnath, Y. Jie, W. Huang and R. Chen, Chem. Mater., 2023, 35, 6405–6414.
458. D. Dou, X. Zhou, T. Wang, Q. Yang, X. Tan, Z. Ling, M. Manz, X. Liu, G.-J. A. H. Wetzelaer, X. Li, M. Baumgarten, P. W. M. Blom and Y. Li, Adv. Opt. Mater., 2024, 12, 2400976.
459. I. Partanen, C. H. Hsu, E. H. Shi, I. Maisuls, T. Eskelinen, A. J. Karttunen, J. J. Saarinen, C. A. Strassert, A. Belyaev, P. T. Chou and I. O. Koshevoy, Angew. Chem., Int. Ed., 2025, 64, e202503327.
460. K. Chen, Y. Luo, M. Sun, C. Liu, M. Jia, C. Fu, X. Shen, C. Li, X. Zheng, X. Pu, Y. Huang and Z. Lu, Angew. Chem., Int. Ed., 2024, 63, e202314447.
461. Z. E. X. Dance, S. M. Mickley, T. M. Wilson, A. B. Ricks, A. M. Scott, M. A. Ratner and M. R. Wasielewski, J. Phys. Chem. A, 2008, 112, 4194–4201.
462. R. Huang, K. Yu, S. Chen, K. Chen, Y. Luo, Z. Lu, F. B. Dias and X. Zheng, Adv. Opt. Mater., 2024, 12, 2400210.
463. K. Chen, B. Li, C. Liu, Y. Luo, K. Yu, M. Jia, Y. Huang and Z. Lu, Angew. Chem., Int. Ed., 2025, 64, e202508776.
464. X. Zhang, L. Du, W. Zhao, Z. Zhao, Y. Xiong, X. He, P. F. Gao, P. Alam, C. Wang, Z. Li, J. Leng, J. Liu, C. Zhou, J. W. Y. Lam, D. L. Phillips, G. Zhang and B. Z. Tang, Nat. Commun., 2019, 10, 5161.
465. C. Chen, Z. Chi, K. C. Chong, A. S. Batsanov, Z. Yang, Z. Mao, Z. Yang and B. Liu, Nat. Mater., 2021, 20, 175–180.
466. C. Chen, K. C. Chong, Y. Pan, G. Qi, S. Xu and B. Liu, ACS Mater. Lett., 2021, 3, 1081–1087.
467. T. Wang, M. Liu, W. Feng, R. Cao, Y. Sun, L. Wang, D. Liu, Y. Wang, T. Wang and W. Hu, Adv. Opt. Mater., 2023, 11, 2202613.
468. R. Kabe and C. Adachi, Nature, 2017, 550, 384–387.
469. 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.
470. Z. Yin, Q. Sun, Z. Wu, Y. Xu, Z. Xie and B. Liu, Aggregate, 2025, 6, e70141.
471. Y. Bi, J. Yu, J. Xiao, K. Zhu, B. Wang, X. Xu and S. Lu, Angew. Chem., Int. Ed., 2025, 64, e202506162.
472. B. Chen, W. Huang, X. Nie, F. Liao, H. Miao, X. Zhang and G. Zhang, Angew. Chem., Int. Ed., 2021, 60, 16970–16973.
473. D. Wang, Y. Xie, X. Wu, Y. Lei, Y. Zhou, Z. Cai, M. Liu, H. Wu, X. Huang and Y. Dong, J. Phys. Chem. Lett., 2021, 12, 1814–1821.
474. X. Wang, J. Jin, C.-L. Chen, Z. Wang, Y. Lei, X. Huang, Y. Guo and W. Liu, J. Mater. Chem. C, 2025 10.1039/D5TC03334A.
475. X. Zhang, J. Liu, B. Chen, X. He, X. Li, P. Wei, P. F. Gao, G. Zhang, J. W. Y. Lam and B. Z. Tang, Matter, 2022, 5, 3499–3512.
476. S. Fu, J. Zhang, X. Li, E. Jin, L. Gao, R. Dong, Z. Wang, X. Feng, H. I. Wang and M. Bonn, Nat. Rev. Mater., 2025 DOI:10.1038/s41578-025-00840-z.
477. V. Coropceanu, J. Cornil, D. A. da Silva Filho, Y. Olivier, R. Silbey and J.-L. Brédas, Chem. Rev., 2007, 107, 926–952.
478. J. Nelson, J. J. Kwiatkowski, J. Kirkpatrick and J. M. Frost, Acc. Chem. Res., 2009, 42, 1768–1778.
479. Y. Shirota and H. Kageyama, Chem. Rev., 2007, 107, 953–1010.
480. D. M. Guldi, H. Nishihara and L. Venkataraman, Chem. Soc. Rev., 2015, 44, 842–844.
481. B. Van Averbeke, D. Beljonne and E. Hennebicq, Adv. Funct. Mater., 2008, 18, 492–498.
482. G. Pace, I. Bargigia, Y.-Y. Noh, C. Silva and M. Caironi, Nat. Commun., 2019, 10, 5226.
483. S. M. Quintero, L. Van Nyvel, N. Roig, J. Casado and M. Alonso, J. Phys. Chem. A, 2024, 128, 6140–6157.
484. C. Schubert, J. T. Margraf, T. Clark and D. M. Guldi, Chem. Soc. Rev., 2015, 44, 988–998.
485. L. Chen, Y.-H. Wang, B. He, H. Nie, R. Hu, F. Huang, A. Qin, X.-S. Zhou, Z. Zhao and B. Z. Tang, Angew. Chem., Int. Ed., 2015, 54, 4231–4235.
486. S. Jiao, P. Shen, J. Li, X. Dong, B. Z. Tang and Z. Zhao, Angew. Chem., Int. Ed., 2025, 64, e202414801.
487. H. Chen and J. Fraser Stoddart, Nat. Rev. Mater., 2021, 6, 804–828.
488. F. Tao, Y. Wei, J. Chen, Y. Li, Y. Wang, Z. Zhang, Y. Pu, K. Wang, H. Li, X. Li, Q. Feng, S. Chen, T. Yang, L. Liu, L. Xie and W. Huang, Small, 2025, 21, 2406628.
489. J. Wang, X. Wang, C. Yao, J. Xu, D. Wang, X. Zhao, X. Li, J. Liu and W. Hong, Langmuir, 2025, 41, 5705–5735.
490. Y. Zhou, S. Ji, Y. Zhu, H. Liu, J. Wang, Y. Zhang, J. Bai, X. Li, J. Shi, W. Su, R. Huang, J. Liu and W. Hong, J. Am. Chem. Soc., 2024, 146, 33378–33385.
491. M. O. Hight, J. Y. Wong, A. E. Pimentel and T. A. Su, J. Am. Chem. Soc., 2024, 146, 4716–4726.
492. H. Ren, P. Li, J. Hao, W. Xiong, L. Pei, B. Wang, C. Zhao, S. He, S. Jin, J. Lü, J. Wang, C. Jia and X. Guo, ACS Mater. Lett., 2025, 7, 754–760.
493. Q. Ren, W. Hao, L. Wang, K. Zhou, L. Liu and H.-C. Wu, J. Am. Chem. Soc., 2025, 147, 34118–34125.
494. Y. Zhu, Y. Zhou, L. Ren, J. Ye, H. Wang, X. Liu, R. Huang, H. Liu, J. Liu, J. Shi, P. Gao and W. Hong, Angew. Chem., Int. Ed., 2023, 62, e202302693.
495. A. Nath, V. Kumar, H. N. Ghosh and S. Mandal, J. Phys. Chem. Lett., 2025, 16, 1238–1246.
496. Y. Zhang, Y. Zhang, L. Wang, C. Zhang and H. Chen, Trends Chem., 2025, 7, 299–316.
497. T. Zhang, T. Li, J. Li, Z. Yu, H. Chen, M. Li, J. Yu, S. Zhen and Z. Zhao, Chem. Eng. J., 2025, 520, 165691.
498. Y. Guo, M. Li, C. Zhao, Y. Zhang, C. Jia and X. Guo, JACS Au, 2024, 4, 1278–1294.
499. S. Zhen, J. C. Mao, L. Chen, S. Ding, W. Luo, X. S. Zhou, A. Qin, Z. Zhao and B. Z. Tang, Nano Lett., 2018, 18, 4200–4205.
500. J. Li, P. Shen, S. Zhen, C. Tang, Y. Ye, D. Zhou, W. Hong, Z. Zhao and B. Z. Tang, Nat. Commun., 2021, 12, 167.
501. T. A. Su, M. Neupane, M. L. Steigerwald, L. Venkataraman and C. Nuckolls, Nat. Rev. Mater., 2016, 1, 16002.
502. W. Chen, J. R. Widawsky, H. Vázquez, S. T. Schneebeli, M. S. Hybertsen, R. Breslow and L. Venkataraman, J. Am. Chem. Soc., 2011, 133, 17160–17163.
503. K. Brunner, A. Tortschanoff, C. Warmuth, H. Bässler and H. F. Kauffmann, J. Phys. Chem. B, 2000, 104, 3781–3790.
504. M. Wielopolski, A. Molina-Ontoria, C. Schubert, J. T. Margraf, E. Krokos, J. Kirschner, A. Gouloumis, T. Clark, D. M. Guldi and N. Martín, J. Am. Chem. Soc., 2013, 135, 10372–10381.
505. Z. Tan, W. Jiang, C. Tang, L.-C. Chen, L. Chen, J. Liu, Z. Liu, H.-L. Zhang, D. Zhang and W. Hong, CCS Chem., 2022, 4, 713–721.
506. J. Li, Z. Zhuang, P. Shen, S. Song, B. Z. Tang and Z. Zhao, J. Am. Chem. Soc., 2022, 144, 8073–8083.
507. Y. Tsuji, K. Okazawa, T. Tatsumi and K. Yoshizawa, J. Am. Chem. Soc., 2024, 146, 32506–32518.
508. S. Song, J. Li, Y. Xiong, B. Z. Tang and Z. Zhao, SmartMat, 2025, 6, e70021.
509. J. Li, P. Shen, Z. Zhuang, J. Wu, B. Z. Tang and Z. Zhao, Nat. Commun., 2023, 14, 6250.
510. K. S. Woo, A. Zhang, A. Arabelo, T. D. Brown, M. Park, A. A. Talin, E. J. Fuller, R. S. Bisht, X. Qian, R. Arroyave, S. Ramanathan, L. Thomas, R. S. Williams and S. Kumar, Nat. Commun., 2024, 15, 4656.
511. F. Yang, M. Anderson, S. He, K. Stephens, Y. Zheng, Z. Chen, S. N. Raja, F. Aplin, Y. Guan and G. Fridman, Sci. Adv., 2018, 4, eaaq1438.
512. P. Zhou, J. Zheng, T. Han, L. Chen, W. Cao, Y. Zhu, D. Zhou, R. Li, Y. Tian, Z. Liu, J. Liu and W. Hong, Nanoscale, 2021, 13, 7600–7605.
513. A. W. Ghosh, T. Rakshit and S. Datta, Nano Lett., 2004, 4, 565–568.
514. R. Karnik, R. Fan, M. Yue, D. Li, P. Yang and A. Majumdar, Nano Lett, 2005, 5, 943–948.
515. H. Chen, H. Zheng, C. Hu, K. Cai, Y. Jiao, L. Zhang, F. Jiang, I. Roy, Y. Qiu, D. Shen, Y. Feng, F. M. Alsubaie, H. Guo, W. Hong and J. F. Stoddart, Matter, 2020, 2, 378–389.
516. A. A. Fokin, D. Gerbig and P. R. Schreiner, J. Am. Chem. Soc., 2011, 133, 20036–20039.
517. H. Ju, B. Wang, M. Li, J. Hao, W. Si, S. Song, K. Mei, A. C. Sue, J. Wang, C. Jia and X. Guo, J. Am. Chem. Soc., 2024, 146, 25290–25298.
518. T. A. Su, H. Li, R. S. Klausen, J. R. Widawsky, A. Batra, M. L. Steigerwald, L. Venkataraman and C. Nuckolls, J. Am. Chem. Soc., 2016, 138, 7791–7795.
519. S. Fujii, S. Seko, T. Tanaka, Y. Yoshihara, S. Furukawa, T. Nishino and M. Saito, J. Am. Chem. Soc., 2024, 146, 19566–19571.
520. Z. Zhao, A. Li and W. Z. Yuan, Acc. Chem. Res., 2025, 58, 612–624.
521. Q. Zhou, X. Zhang, L. Ning, Y. Song, Y. Wang, J. Feng, Q. Gong and Y. Huang, J. Mater. Chem. C, 2024, 12, 18991–19016.
522. Z. Li, Q. Yue, H. Zhang and Y. Zhao, Mater. Today, 2024, 78, 209–230.
523. Z. Wu, J. Nitsch and T. B. Marder, Adv. Opt. Mater., 2021, 9, 2100411.
524. C. Yan, C. Fang, J. Gan, J. Wang, X. Zhao, X. Wang, J. Li, Y. Zhang, H. Liu, X. Li, J. Bai, J. Liu and W. Hong, ACS Nano, 2024, 18, 28531–28556.
525. J. Guo, P.-K. Chen and S. Chang, Anal. Chem., 2024, 96, 9303–9316.
526. N. J. Hestand and F. C. Spano, Acc. Chem. Res., 2017, 50, 341–350.
527. S. Li, W. Li and J. Ma, Acc. Chem. Res., 2014, 47, 2712–2720.
528. M. Liu, W. Wang, P. Su, J. Ren and Q. Peng, J. Phys. Chem. Lett., 2025, 16, 8731–8738.
529. W. Gao, C. Addiego, H. Wang, X. Yan, Y. Hou, D. Ji, C. Heikes, Y. Zhang, L. Li, H. Huyan, T. Blum, T. Aoki, Y. Nie, D. G. Schlom, R. Wu and X. Pan, Nature, 2019, 575, 480–484.
530. C. Addiego, W. Gao, H. Huyan and X. Pan, Nat. Rev. Phys., 2023, 5, 117–132.
531. D. J. Müller, A. C. Dumitru, C. Lo Giudice, H. E. Gaub, P. Hinterdorfer, G. Hummer, J. J. De Yoreo, Y. F. Dufrêne and D. Alsteens, Chem. Rev., 2021, 121, 11701–11725.
532. G. A. DiLabio and A. Otero-de-la-Roza, Reviews in Computational Chemistry, 2016, pp. 1–97 DOI:10.1002/9781119148739.ch1.
533. H. Wu, L. Zhang, J. Yang, R. Bo, H. Du, K. Lin, D. Zhang, M. Ramachandran, Y. Shen, Y. Xu, X. Xue, Z. Ma, A. R. Lindstrom, R. Carney, T.-Y. Lin and Y. Li, Adv. Funct. Mater., 2020, 30, 1910348.
534. P.-P. Sun, B.-L. Han, H.-G. Li, C.-K. Zhang, X. Xin, J.-M. Dou, Z.-Y. Gao and D. Sun, Angew. Chem., Int. Ed., 2022, 61, e202200180.
535. Y.-L. Ying, Y.-J. Li, J. Mei, R. Gao, Y.-X. Hu, Y.-T. Long and H. Tian, Nat. Commun., 2018, 9, 3657.
536. N. Li, J. Pang, F. Lang and X.-H. Bu, Acc. Chem. Res., 2024, 57, 2279–2292.
537. D. Yan, Z. Wang and Z. Zhang, Acc. Chem. Res., 2022, 55, 1047–1058.
538. F. Gu and X. Ma, Chem. Eur. J., 2022, 28, e202104131.
539. R. Wu, H. Dai, W. Guan and C. Lu, Small, 2025, 21, 2412133.
540. K. M. Jablonka, P. Schwaller, A. Ortega-Guerrero and B. Smit, Nat. Mach. Intell., 2024, 6, 161–169.
541. S. Yu, N. Ran and J. Liu, Artif. Intell. Chem., 2024, 2, 100076.
542. Q. Zhang, Y. Hu, J. Yan, H. Zhang, X. Xie, J. Zhu, H. Li, X. Niu, L. Li, Y. Sun and W. Hu, Adv. Mater., 2024, 36, 2405163.
543. J. Zhong, W. Zhu, S. Shen, N. Zhou, M. Xi, K. Du, D. Wang and B. Z. Tang, Aggregate, 2025, 6, e70089.
544. G. Lei, R. Docherty and S. J. Cooper, Digit. Discovery, 2024, 3, 1257–1272.
545. Z. Zheng, N. Rampal, T. J. Inizan, C. Borgs, J. T. Chayes and O. M. Yaghi, Nat. Rev. Mater., 2025, 10, 369–381.
546. C. Sun, J. Fu and L. Xu, Bioorg. Chem., 2025, 166, 109029.
547. Z. Ren, J. Liu, W. Wang, Y. Wang, Y. Li and X. Leng, ACS Appl. Polym. Mater., 2025, 7, 9378–9388.

---

Footnote

† These authors contributed equally to this work.

---