Boosted photochromic properties by carbon dots based on Förster resonance energy transfer

Abstract

Photochromic materials with rapid response and high stability are essential for progressive anti-counterfeiting and secure information encryption technologies. Herein, we report a Förster resonance energy transfer (FRET)-assisted strategy to boost the photochromic properties of highly crystalline C₃N₅ nanosheets (HC-C₃N₅) by integrating carbon dots (CDs). The incorporation of CDs significantly increased light absorption, fluorescence intensity, and energy transfer efficiency, leading to an ultrafast and reversible color transition from dark yellow to green under UV irradiation, with complete recovery within 180 s and excellent cycling stability. The transient photovoltage technique (TPV) test confirms a non-radiative energy transfer pathway between CDs and HC-C₃N₅, excluding the possibility of electron transfer. Building on the distinct photo response characteristics of bulk C₃N₅ (B-C₃N₅), HC-C₃N₅, and CDs/HC-C₃N₅, this study further explores their potential in multi-layered anti-counterfeiting labels and a time-resolved encryption system, enabling dynamic optical information encoding. This work not only reveals the key role of the FRET mechanism over CDs in modified photochromic materials, but also paves the way for next-generation anti-counterfeiting and secure data storage applications.

---

New concepts

While material optimization for enhanced photochromic performance typically focuses on compositional tuning or structural modifications, this work introduces a broader conceptual paradigm: mechanism-driven materials engineering for dynamic information encoding. The prevailing concept in photochromic material development has been to directly alter the electronic structure of the primary active component. In contrast, we propose and demonstrate a fundamental shift towards the rational integration of complementary nanoscale components to create a synergistic “energy transfer relay”. Our new concept lies not merely in combining carbon dots (CDs) with highly crystalline C₃N₅ nanosheets (HC-C₃N₅), but in deliberately exploiting the Förster resonance energy transfer (FRET) pathway as the primary design principle to bootstrap photochromic properties. This approach deliberately bypasses direct electron transfer, instead of creating a highly efficient, non-radiative energy channel that amplifies the photoresponse at a system level. This conceptual advance transforms photochromic materials from static color-changing entities into dynamic, time-resolved optical platforms. By moving beyond a single-material property to a multi-component system with distinct photo-response characteristics (B-C₃N₅, HC-C₃N₅, and CDs/HC-C₃N₅), we establish a novel framework for “information density” in optical security. This enables the creation of multi-layered anti-counterfeiting labels and encryption systems where information is not only spatially defined but also temporally revealed, paving the way for next-generation secure data storage with an unprecedented level of complexity and security.

1. Introduction

Photochromic materials, as a class of intelligent materials capable of undergoing reversible color changes upon exposure to specific wavelengths of light, have demonstrated irreplaceable utilization in secure anti-counterfeiting applications, dynamic information encryption, and optical storage in recent years.^1,2 Recently, significant strides have been made in developing advanced photo-responsive systems for high-level security and optical modulation, further highlighting the versatility of these intelligent materials.^3–5 The working mechanism of these materials relies on photoinduced structural or electronic modifications that regulate their absorption or emission spectra, thereby enabling controllable optical responses.⁶ An ideal photochromic system should exhibit rapid response kinetics, high color contrast, excellent stability, and reversible switching over multiple cycles.^7,8 However, conventional organic photochromic systems, such as spiropyrans and diarylethene, are generally limited by sluggish response kinetics (ranging from seconds to minutes), pronounced photo fatigue effects, and high environmental sensitivity, making them unsuitable for high-security applications that demand rapid response, long cycling stability, and multimodal regulation.^9–11 Although inorganic semiconductor materials (e.g., WO₃ and MoO₃) have improved photo-response rates to some extent through defect engineering and interfacial modulation, their intrinsically rigid lattice structures and limited excited-state lifetimes still hinder breakthroughs in reversible photochromic performance.^12–14 Therefore, the development of novel photochromic materials with enhanced light absorption, higher photoconversion efficiency, and superior stability is crucial for both fundamental research and technological applications.

In recent years, carbon nitride materials have emerged as a research focus in the field of photochromism due to their tunable band structures, abundant surface-active sites, and excellent photochemical stability. However, challenges such as high carrier recombination rates and low light absorption efficiency caused by poor crystallinity hinder their practical application in high-performance photochromic devices, necessitating breakthroughs through synergistic optimization of material design and energy transfer mechanisms.^15–18 Among emerging photochromic materials, nitrogen-rich carbon nitride (C₃N₅) has attracted increasing attention due to its high chemical stability, unique electronic structure, and tunable optoelectronic properties. Compared with traditional graphitic carbon nitride (g-C₃N₄), C₃N₅ possesses a higher nitrogen content and an extended conjugated system, enhancing the optical response and improving the charge separation efficiency.^19–21 Specifically, highly crystalline C₃N₅ nanosheets (HC-C₃N₅) have shown remarkable photophysical properties; nevertheless, the intrinsic photochromic efficiency of HC-C₃N₅ is still limited due to its average light-harvesting ability and relatively low exciton utilization efficiency,²² while enhancing the optical response of HC-C₃N₅ while maintaining its structural stability and reversibility is a challenge to be solved.

Carbon dots (CDs), as an emerging class of zero-dimensional luminescent nanomaterials with rich optical behaviors that can be efficiently regulated by conjugated domains, doped elements, and surface structural modifications,²³ have attracted much attention due to their excellent bio-compatibility, high stability, and low toxicity.²⁴ Additionally, CD materials are designed to exhibit various photoluminescence (PL) behaviors, including fluorescence,²⁵ phosphorescence,²⁶ organic afterglow,²⁷ and bimodal emission,²⁸ offering significant advantages in applications such as fluorescence bioimaging,²⁹ photodynamic therapy,³⁰ and, most importantly, anti-counterfeit documentation.³¹ Furthermore, CDs can be integrated with photochromic properties, exhibiting dynamically tunable color changes upon application of external stimuli, thereby demonstrating multifunctional optical response capabilities. For example, when CDs are anchored to a titanium dioxide (TiO₂) porous film, the film shows different spectral responses upon irradiation with light of varying wavelengths.³² Förster resonance energy transfer (FRET), as a non-radiative energy transfer mechanism, provides a novel paradigm for enhancing the performance of photochromic materials. The core of FRET lies in the directional transfer of excited state energy through dipole–dipole coupling between the donor and the acceptor, thus effectively avoiding charge recombination loss in the electron transfer process, and at the same time, by adjusting the energy level arrangement and spectral overlap between the donor and the acceptor, it can also realize the precise control of optical response dynamics.^33–36 Shan et al. designed a red afterglow developer consisting of rhodamine B (RhB) and CDs, in which CDs act as an energy donor and RhB as an energy acceptor, and an effective energy transfer from CDs to RhB can be achieved through the FRET process, demonstrating that the nanocomposite exhibits red persistent luminescence in aqueous environments.³⁷ Furthermore, other photoactive substances, such as diarylethene derivatives, have been introduced to modify CDs and establish FRET, which dynamically tunes the PL properties of photoactive materials with different light absorption capacities and FRET efficiencies.^38,39 However, despite the widespread application of the FRET mechanism in controlling the photophysical properties of photochromic materials, current research predominantly focuses on organic molecules and metal nanoparticle clusters, while its application in nitrogen-rich polymer semiconductors (such as C₃N₅) is still in its infancy, with significant challenges remaining in improving the photochromic response speed, enhancing stability, and expanding functional applications.

Herein, a CDs/HC-C₃N₅ composite based on the FRET mechanism is constructed, which significantly improves the light absorption ability, fluorescence intensity and energy transfer efficiency of HC-C₃N₅ through the introduction of CDs. More importantly, the tunable electronic states of CDs and their excellent compatibility with the semiconductor system enable synergistic interactions within the composite material, where their integration with HC-C₃N₅ not only facilitates effective energy transfer through dipole–dipole coupling but also enhances charge carrier dynamics, significantly improving the overall photochromic performance. Specifically, the composite material exhibits rapid and reversible color changes from dark yellow to green, and under UV light irradiation, it can return to its initial state within just 180 s, demonstrating excellent cycling stability. TPV analysis further confirms the non-radiative energy transfer pathway between the CDs and HC-C₃N₅, eliminating the possibility of electron transfer. This study explores the FRET interactions between CDs and HC-C₃N₅ and uses this mechanism to enhance the photochromic properties, which not only helps to deepen the understanding of the energy transfer process, but also provides a new idea for the development of highly efficient and tunable multifunctional dynamic photo-responsive materials.

2. Experimental section

2.1. Synthesis of CD powder

CDs were prepared using a typical electrochemical method as described in our previous studies.^40,41 Electrochemical synthesis of CDs was carried out by immersing two graphite electrodes into ultrapure water under a 30 V bias using a direct current (DC) power source and after 10 days of agitation, the black dispersion was purified through filtration and freeze-dried to acquire CDs in powder form.

2.2. Synthesis of HC-C₃N₅ nanosheets

HC-C₃N₅ nanosheets were prepared using an alkaline potassium salt-assisted thermal polymerization method.²² The process involved precisely mixing 3 g of 3-amino-1,2,4-triazole, 6 g of potassium chloride, and 0.3 g of potassium hydroxide in a mass ratio of 1 : 2 : 0.1, followed by gradual heating of the precursor in a muffle furnace at 2.5 °C min⁻¹ to 600 °C, where it was maintained for 2 h to ensure complete thermal polymerization. Following heat treatment, the product was thoroughly washed multiple times with boiling deionized water to remove residual metal salts, ensuring high purity, and subsequently dried to obtain the yellow-colored HC-C₃N₅ sample for further use. For comparison, a facile one-stage calcination approach was employed to synthesize the bulk C₃N₅ (B-C₃N₅).⁴² In this process, 3 g of 3-amino-1,2,4-triazole was used as the sole precursor, directly placed in a muffle furnace, rapidly heated to 550 °C at 5 °C min⁻¹, held for 2 h to complete calcination, and then washed with deionized water and ethanol before drying to obtain the B-C₃N₅ sample.

2.3. Synthesis of the CDs/HC-C₃N₅ composite

For the preparation of CDs/HC-C₃N₅, the synthesis steps were identical to those described previously for HC-C₃N₅, with the exception that a precise amount of CD powder (0.15 g) was added during the precursor addition stage, and the complete synthesis route is illustrated in Scheme 1.

Scheme 1 Schematic diagram of the synthesis process of the CDs/HC-C₃N₅ photochromic material.

3. Results and discussion

3.1. Morphology and chemical structure of CDs/HC-C₃N₅

The structure and morphology of B-C₃N₅, HC-C₃N₅ and CDs/HC-C₃N₅ were investigated by transmission electron microscopy (TEM) and high-resolution TEM (HRTEM), which revealed that both B-C₃N₅ and HC-C₃N₅ exhibit a distinct 2D nano-sheet structure, reflecting their intrinsic laminar nature (Fig. S1a and c). Furthermore, the HRTEM image in Fig. S1b indicates that B-C₃N₅ lacks clear lattice stripes, indicating its amorphous character, while HC-C₃N₅ nanosheets exhibited clear lattice stripes with the 1.02/(100) diffraction plane and the 0.32 nm/(002) diffraction plane (Fig. S1d), implying the improved crystallinity, which is consistent with the previously reported results.²² As illustrated in Fig. 1a and b, the CDs were firmly anchored on the surface of HC-C₃N₅ to form the CDs/HC-C₃N₅ composite, which notably retained the primary highly crystalline nanosheet morphology of HC-C₃N₅ and its lattice spacing and outline of CDs (Fig. 1c). Moreover, the energy dispersive X-ray (EDX) spectra (Fig. S2) unequivocally demonstrated the presence of K, C, and N elements within the CDs/HC-C₃N₅ composite. To further quantitatively characterize the thickness and verify the exfoliated nature of the synthesized materials, atomic force microscopy (AFM) measurements were performed. As displayed in Fig. S3, the bulk B-C₃N₅ exhibits a height of approximately 28.8 nm. In contrast, the HC-C₃N₅ nanosheets show a significantly reduced thickness of 10.2 nm, confirming effective exfoliation. The CDs/HC-C₃N₅ composite retains this ultrathin morphology (10.5 nm), indicating that the introduction of CDs preserves the layered structure of the matrix. The crystal structures of B-C₃N₅, HC-C₃N₅, and CDs/HC-C₃N₅ were elucidated by X-ray diffraction (XRD) analysis as illustrated in Fig. 1d. It can be seen that all of the samples exhibit two prominent peaks corresponding to the intralayer conformation (100) and the interlayer stacking feature (002) of g-C₃N₅, respectively.⁴³ Notably, the (100) peak of HC-C₃N₅ (7.8°), observed at a lower angle compared to that of B-C₃N₅ (13.1°), is attributed to the incorporation of K ions into the heptazine unit of HC-C₃N₅, while the (002) characteristic peak of HC-C₃N₅ (28°), shifted to a higher angle relative to B-C₃N₅ (27.6°), indicates that the strengthened π–π interactions between adjacent heptazine layers result in more compact interlayer stacking.⁴⁴ Furthermore, the slight shift of the (002) diffraction peak of CDs/HC-C₃N₅ to a higher angle with the addition of CDs suggests a decrease in the interlayer spacing of the C₃N₅ nanosheets.⁴⁵ For an in-depth comprehension of the molecular composition of the CDs/HC-C₃N₅ composites, Fourier transform infrared spectroscopy (FT-IR) was employed for the analysis. As indicated in Fig. 1e, all the samples revealed the typical peak characteristics of CN polymers, specifically the bending vibrational peak of the heptazine unit (C–N–C) at 810 cm⁻¹ and the stretching vibrational peak of the aromatic CN heterocyclic ring displayed in the range of 1100–1700 cm⁻¹. As for HC-C₃N₅ obtained after molten salt treatment, a more pronounced peak at 2173 cm⁻¹ is attributed to the asymmetric extension vibration of the cyano (C≡N) group,⁴⁶ and the FT-IR spectrum of the CDs/HC-C₃N₅ complex matches well with the characteristic vibrational modes of HC-C₃N₅, which confirms that addition of CDs has not changed its molecular structure. X-ray photoelectron spectroscopy (XPS) and detailed elemental analysis of the samples provide strong evidence for the presence of robust surface functional groups. In the XPS survey spectrum of Fig. S4, three prominent peaks located at 288 eV, 398 eV and 532 eV corresponding to C 1s, N 1s and O 1s orbitals, respectively, can be clearly observed, and it is noteworthy that the CDs/HC-C₃N₅ composite shows additional potassium (K) doping features (293 eV), which is due to the molten salt (KCl) introduced during the preparation of HC-C₃N₅. As can be seen in Fig. 1f, the high-resolution XPS spectrum of C 1s, deconvoluted into three well-defined peaks at 284.6 eV, 286.7 eV, and 288.1 eV, indicates that the peak at 284.6 eV can be ascribed to C–C/C=C bonds, predominantly arising from physiosorbed carbonaceous species,⁴⁷ whereas the pronounced signal at 286.7 eV is attributable to carbon atoms covalently bonded to three nitrogen neighbors (N₂–C=N),⁴⁸ and the peak at 288.1 eV corresponds to the triazole-functionalized CN framework.⁴⁹ Remarkably, compared to bulk C₃N₅ reported in previous literature,⁴² the significantly higher characteristic peak observed at 285.8 eV is mainly attributed to the similarity of the binding energies of C≡N and the amino group, where the presence of C≡N may be related to the dehydrogenation of the C–NH₂ moiety.⁵⁰ As revealed by the N 1s spectrum (Fig. 1g), the binding energies at 398.6 eV and 400.9 eV are indicative of nitrogen atoms covalently linked to two carbon atoms within the graphene sp² lattice and the N–(C)₃ configuration, respectively.⁵¹ Additionally, the O 1s spectrum (Fig. 1h) displays two peaks at 530.9 eV and 532.5 eV, confirming the presence of C=O and C–O groups, respectively.⁵² As evidenced by the K 2p peak at 295.4 eV and 292.7 eV in Fig. 1i, the successful incorporation of potassium into the HC-C₃N₅ framework is definitely confirmed, with the presence of CDs in CDs/HC-C₃N₅ exerting no discernible impact on the efficacy of this molten salt-mediated doping process.

Fig. 1 (a) TEM and (b) and (c) HRTEM images of the CDs/HC-C₃N₅ composite. (d) XRD patterns and (e) FT-IR spectra of B-C₃N₅, HC-C₃N₅ and the CDs/HC-C₃N₅ composite. High-resolution XPS spectra of (f) C 1s, (g) N 1s, (h) O 1s and (i) K 2p of the CDs/HC-C₃N₅ composite.

3.2. Investigation of photophysical properties

The optical photographs of the three samples, B-C₃N₅, HC-C₃N₅, and CDs/HC-C₃N₅, are shown in Fig. S5a, along with their corresponding Commission Internationale Eclairage (CIE) chromaticity coordinates. It is evident that the B-C₃N₅ sample exhibits a brown color, HC-C₃N₅ displays a relatively brighter yellow hue, and the combination of CDs with HC-C₃N₅ to form CDs/HC-C₃N₅ results in a significant color transformation to a dark yellow hue, as further confirmed by the CIE coordinate diagram in Fig. S5b, which highlights their distinct positions in the chromaticity space.

For a comprehensive understanding of the optical properties of these samples, further characterization was carried out using UV-vis absorption spectroscopy and fluorescence spectroscopy. As illustrated in Fig. 2a, the UV-vis absorption spectrum of CDs/HC-C₃N₅ reveals two prominent peaks at 291 and 400 nm, corresponding to π → π* and n → π* electronic transitions, respectively.⁵³ In comparison with B-C₃N₅ (Fig. S6a) and HC-C₃N₅ (Fig. S6b), the absorption peaks of CDs/HC-C₃N₅ display a redshift, primarily as a result of denser molecular packing that intensifies orbital overlap, leading to greater stabilization of the excited states and a consequent shift of the absorption peaks toward lower energy regions. To gain a more comprehensive understanding of the fluorescence properties of CDs/HC-C₃N₅, an in-depth analysis of its fluorescence emission spectra was conducted under various excitation wavelengths (Fig. S7), revealing its distinctive excitation-independent fluorescence behavior, with the maximum intense emission peak observed at 537 nm when excited at the optimal wavelength of 285 nm. The photochromic behavior of the CDs/HC-C₃N₅ material under ambient air conditions was comprehensively investigated through in situ absorption analysis, with significant alterations in its reflection spectrum being observed. As clearly illustrated in Fig. 2b, the CDs/HC-C₃N₅ material exhibits a distinct absorption peak in the wavelength range of 400 to 550 nm, indicating its high reflectance of blue light. At the initial stage of UV irradiation (0 s), the sample reflection spectrum exhibits an absence of relative characteristics, yet as the exposure time progresses to 50 s, a pronounced redshift in the absorption peak becomes evident, simultaneously accompanied by a gradual intensification of the color. As demonstrated in Fig. 2c, during the recovery process of the CDs/HC-C₃N₅ material after the cessation of UV light, the reflection spectrum exhibits a significant feature, with signs of recovery in the absorption peak position becoming evident within just 10 s, and the recovery continues until the reflection spectrum fully stabilizes and returns to its initial state after approximately 180 s. As depicted in Fig. S8, whereas B-C₃N₅ exhibited negligible alterations in its chromatic transitions within the triethanolamine (TEOA) solution, HC-C₃N₅ underwent reflection spectral modulations akin to those observed in CDs/HC-C₃N₅ with prolonged UV irradiation, signifying its remarkable photochromic responsiveness coupled with a progressive augmentation in light absorption capacity. Here, TEOA acts as a crucial hole scavenger. Upon UV irradiation, it captures the photogenerated holes from the valence band of HC-C₃N₅, thereby effectively suppressing electron–hole recombination. This process facilitates the accumulation of photogenerated electrons on the heptazine units, leading to the formation of colored radical species responsible for the macroscopic color transition from yellow to green. As clearly indicated in Fig. 2d, the fluorescence intensity of the CDs/HC-C₃N₅ composite gradually increases with prolonged UV irradiation time. During the initial irradiation stages, the relatively weak fluorescence intensity suggests that the photoactive centers are not fully activated, but as UV exposure extends, the emission spectrum stabilizes after 180 s with the peak at 588 nm in the green light region. Fig. 2e further reveals the fluorescence decay process of CDs/HC-C₃N₅ after the cessation of UV irradiation, which strongly demonstrates the excellent optical reversibility. The rapid decrease in fluorescence intensity can be attributed to the non-radiative transitions of electrons from the excited state to the ground state, as well as the gradual depletion of photoactive centers within the composite.^54,55 Correspondingly, Fig. 2f illustrates the fluorescence intensity variation curve, depicting the entire process of the material transitioning from the initial state to saturation. Similarly, the fluorescence intensity of B-C₃N₅ exhibited only a negligible enhancement (Fig. S9a and b), while that of HC-C₃N₅, despite a slight improvement due to its high crystallinity (Fig. S9c and d), remained considerably lower than the substantial increase observed in CDs/HC-C₃N₅, unequivocally highlighting the crucial role of CDs in significantly augmenting the photo response efficiency.

Fig. 2 (a) UV-vis absorption and fluorescence excitation and emission spectra of the CDs/HC-C₃N₅ powder. Normalized in situ reflection of CDs/HC-C₃N₅ in TEOA solution for (b) coloring and (c) bleaching under UV irradiation at different times. PL intensity curves with irradiation time during (d) coloring and (e) bleaching processes. (f) PL increased intensity and (g) CIE coordinates of CDs/HC-C₃N₅ after UV irradiation of different durations. (h) Remaining degree of color RC (%) and linear fitting equation and (i) reversibility of the photoactivated and deactivated CDs/HC-C₃N₅ material.

Additionally, the CIE chromaticity diagram in Fig. 2g, derived from the time-dependent reflection spectra, provides a vivid and detailed depiction of the color evolution trajectory of the CDs/HC-C₃N₅ sample under varying UV irradiation durations. In the initial state, the color mainly occupies the dark yellow area. However, consistent with the redshift observed in the reflection spectra, the gradual shift of the color coordinates with the accumulation of irradiation time vividly illustrates the macroscopic color transition of the deepening sample, with enhanced enhancement and green intensity. The color transition of CDs/HC-C₃N₅ in TEOA solution, as evidenced by the experimental photographs in Fig. S10, reveals a gradual shift from dark yellow to green under UV irradiation, aligning with the trend of the color coordinates in the CIE chromaticity diagram. To elucidate the critical factors governing this transformation, control experiments were conducted under varying conditions. As detailed in Fig. S11, the rapid coloration persisted even under an inert nitrogen atmosphere, whereas the bleaching rate was significantly reduced, indicating that oxygen participates primarily in the recovery process rather than the coloration. Conversely, replacing the TEOA solution with pure water resulted in a complete absence of photochromism (Fig. S12), and the pure CD solution exhibited no color change under identical irradiation (Fig. S13). These results unequivocally verify that the macroscopic color transition is specific to the CDs/HC-C₃N₅ composite and strictly dependent on the presence of TEOA. The photochromic behavior of B-C₃N₅ and HC-C₃N₅, along with their corresponding CIE chromaticity trends, is depicted in Fig. S14. Compared to CDs/HC-C₃N₅, B-C₃N₅ exhibited minimal color variation, highlighting its weaker photochromic performance, while HC-C₃N₅ gradually transitioned to a light green hue under UV irradiation, albeit with a less pronounced transformation than that observed in CDs/HC-C₃N₅. This disparity highlights the crucial role of CDs in significantly enhancing the photochromic properties of HC-C₃N₅, leading to a more pronounced color shift and superior reversibility in CDs/HC-C₃N₅. The dynamics of the color residual degree (RC%) decay over time following the cessation of UV excitation, along with the results of the linear fitting for the CDs/HC-C₃N₅ material, are presented in Fig. 2h. The experimental data reveal that after the light source is shut-off, the RC value of the material rapidly decreases and recovers to over 95% of its initial state within 180 s, with the decay rate showing a linear relationship with time (y = −0.0049x + 0.748) and a correlation coefficient of 0.847, demonstrating the material's highly efficient fading kinetics. The reversibility of CDs/HC-C₃N₅ was evaluated by multi-cycle photoactivation experiments, as given in Fig. 2i. The PL values were consistently maintained at more than 98% of the initial values during 10 consecutive cycles (each cycle consisted of 50 s of UV irradiation and 180 s of dark state restoration), and no significant change was undergone by the CIE, which demonstrated that the material possessed excellent cycling stability. In order to further assess the stability under saturated conditions, we conducted an additional cycling test with an extended irradiation time of 180 s per cycle (Fig. S15). The results show that the composite maintains excellent reversibility and structural integrity even when driven to its full colored state, ruling out photo-fatigue issues under prolonged exposure. In addition, no obvious structural degradation was observed from the unchanged XRD patterns (Fig. S16a) and the FT-IR spectra (Fig. S16b) after photochromism, which confirms the long-term stability of the material for photochromic applications. To further highlight the competitive advantages of the CDs/HC-C₃N₅ composite, we compared its key performance parameters with those of other representative photochromic systems reported in the literature (Table S1). Traditional organic photochromic materials (e.g., spiropyrans) typically suffer from severe fatigue and poor stability, while inorganic semiconductors (e.g., WO₃ and g-C₃N₄) often exhibit sluggish response and recovery kinetics due to limited charge transfer rates. In contrast, our CDs/HC-C₃N₅ system, driven by the FRET mechanism, achieves a superior balance between speed and stability. It delivers an ultrafast response (<50 s) and rapid self-recovery (180 s), significantly outperforming many existing inorganic systems, while maintaining excellent cycling stability that surpasses most of its organic counterparts.

3.3. Mechanistic analysis of Förster resonance energy transfer (FRET)

The energy transfer mechanism between CDs and HC-C₃N₅ was systematically investigated through the characterization of the photophysical properties of the donor (CDs) and the acceptor (HC-C₃N₅), as illustrated in Fig. 3a. The UV-vis absorption spectrum of the donor CDs is characterized by two distinct absorption peaks at 230 nm and 300 nm, attributable to π → π* electronic transitions of the aromatic ring structure and n → π* conversion of surface functional groups, respectively, and their fluorescence spectrum exhibits a strong emission peak at 480 nm, indicating that the CDs exhibit a long-lived triplet excitation state.⁵⁶ In addition, the UV-vis absorption spectrum of the receptor HC-C₃N₅ showed a broad absorption band at 330 nm, whereas its fluorescence emission peak was located near 473 nm, and the photoluminescence (PL) excitation spectrum (Fig. S17) shows that its optimal excitation wavelength was 263 nm. Crucially, there is a significant spectral overlap between the fluorescence emission spectrum of the CDs and the absorption spectrum of HC-C₃N₅. This significant overlap satisfies the primary condition for singlet-to-singlet Förster resonance energy transfer (FRET), which can be described as follows based on the equations for the energy transfer process (eqn (1)–(3)):

HC-C₃N_5(s1) → HC-C₃N_5(s1) + hv′ (3)

where CDs_(s0) and CDs_(s1) are the ground and excited states of the CDs; hv is the incident photon; HC-C₃N_5(s0) and HC-C₃N_5(s1) are the ground and excited states of HC-C₃N₅; and hv′ is the luminescent photon of HC-C₃N₅ from the excited state to the ground state.^57,58 The energy transfer pathway between CDs and HC-C₃N₅ is depicted in Fig. 3b, illustrating the FRET mechanism. Upon UV excitation, CDs undergo electronic excitation, and their excited-state energy is transferred non-radiatively to HC-C₃N₅ via dipole–dipole interactions. This process not only enhances the photoluminescence of HC-C₃N₅ but also significantly accelerates its photochromic response, with the FRET efficiency being dictated by the Förster distance (R₀) and the spectral overlap integral between the donor emission and acceptor absorption, thereby further confirming the occurrence of energy transfer. To quantify the energy transfer process, the spectral overlap integral (J) and R₀ were calculated and are summarized in Table S2. Based on the significant spectral overlap, the calculated J value reached approximately (2.1–2.5) × 10¹⁵ M⁻¹ cm⁻¹ nm. Using the typical quantum yield of electrochemical CDs (0.18–0.22), the R₀ was determined to be in the range of 3.6–4.0 nm, which falls well within the effective distance (1–10 nm) required for dipole–dipole coupling. Furthermore, based on the lifetime decay variations, the FRET efficiency (E) was calculated to be approximately 35–40%, providing robust quantitative evidence for the proposed non-radiative energy transfer mechanism.^59,60 The exciton evolution of CDs and HC-C₃N₅ is shown in Fig. 3c, where single-linear state excitons of CDs are generated upon excitation by UV light, and due to the effective spin–orbit coupling, some carriers are transferred from the single-linear to the triplet energy levels through the intersystem crossover (ISC) process, and radiative compositing of the resulting triplet excitons could cause long-lived emission. Due to the spectral overlap and dipole coupling, this energy is efficiently transferred non-radiatively to the singlet excited state of the adjacent HC-C₃N₅ nanosheets via the singlet-to-singlet FRET mechanism, resulting in a long-wavelength afterglow signal in the green region. This efficient energy transfer greatly enhances the photochromic response of HC-C₃N₅, resulting in a rapid and intense color change under UV irradiation. According to the above analysis, the photochromic switching mechanism of CDs/HC-C₃N₅ is controlled by the FRET process, as provided in Fig. 3d. Under UV irradiation, CDs absorb photon energy and their electrons jump from the ground state (S₀) to the excited singlet state (S₁), which satisfies the FRET condition due to the energy level match between CDs and HC-C₃N₅ and the large overlap of the two spectra. As a result, CDs transfer energy to HC-C₃N₅ in a non-radiative manner through FRET, triggering the rearrangement of the electron distribution of its molecular orbitals, which leads to the change of the absorption spectra, and then the photochromic phenomenon is generated. After the UV irradiation stops, the excited state electrons within HC-C₃N₅ gradually relax back to the ground state, and the main pathways include: (i) part of the carriers returning immediately to the ground state by non-radiative leaps and (ii) another part of the carriers releasing energy by radiative leaps, which may exhibit phosphorescence or afterglow phenomena. With this process, the absorption spectrum of HC-C₃N₅ is gradually restored to the initial state, and the color of the material fades away, demonstrating excellent reversibility of photochromism. While the FRET mechanism significantly boosts the utilization of light energy and the generation rate of excitons, the presence of TEOA is indispensable for stabilizing the colored state. Without TEOA, the rapid radiative or non-radiative recombination of electron–hole pairs would prevent the accumulation of reduced states required for photochromism. Therefore, the synergistic effect of FRET-enhanced absorption and TEOA-assisted hole scavenging ensures the ultrafast and high-contrast photochromic response of the CDs/HC-C₃N₅ composite.

Fig. 3 (a) Absorption and fluorescence emission spectra of the CDs (donor) and HC-C₃N₅ (acceptor). (b) Resonance energy transfer energy level diagram between the CDs and HC-C₃N₅. (c) Illustration of energy transfer from CDs to HC-C₃N₅. (d) Schematic of FRET-enhanced photophysical processes leading to photochromism in the CDs/HC-C₃N₅ system. (e) TPV relaxation curves, (f) maximum charge extraction time (t_max), (g) charge extraction efficiency (A) and (h) attenuation constants (τ) of HC-C₃N₅ and CDs/HC-C₃N₅.

To further elucidate the mechanism by which CDs reinforce the photochromic performance of HC-C₃N₅, the photo-response dynamics of HC-C₃N₅ and CDs/HC-C₃N₅ composites were comparatively investigated using transient photovoltage (TPV) measurements.⁶¹ As can be seen in Fig. 3e, the TPV decay curve of the CDs/HC-C₃N₅ sample exhibits a significantly slower voltage drop, indicating that the introduction of CDs effectively prolongs the lifetime of photo-generated carriers and sustains the excited-state energy within the composite system for a longer duration. The maximum charge extraction time (t_max) in Fig. 3f increases from 0.19 ms for HC-C₃N₅ to 0.50 ms for CDs/HC-C₃N₅, indicating that the FRET process transfers energy via dipole–dipole interactions rather than free-electron injection. No strong electric field is formed within a short timescale to drive charge separation. Meanwhile, as illustrated in Fig. 3g, the charge extraction efficiency (A) of CDs/HC-C₃N₅ improves only slightly from 19.82 to 58.42 compared to HC-C₃N₅, suggesting that the enhancement is limited and the significant charge separation typical of electron-transfer-governed systems is absent. The attenuation constant (τ) in Fig. 3h increases from 1.41 to 1.83 ms, further suggesting that the excited state relaxation process slows down in the system with a longer exciton lifetime. In addition, time-resolved photoluminescence (TRPL) measurements (Fig. S18) further validated the FRET-dominated energy transfer mechanism with an average fluorescence lifetime (τ_ave) of 0.71 ns for the CDs/HC-C₃N₅ composites. Meanwhile, the fluorescence lifetime of HC-C₃N₅ itself is τ_ave = 0.87 ns, which is slightly longer compared to that of the CDs/HC-C₃N₅ composite, suggesting that the energy receiving process in the excited state lifetime is lengthened, and this trend of lifetime change indicates the prolongation of the receptor lifetime, which is typical of the FRET process. Comprehensively, the above results indicate that the introduction of CDs has not significantly promoted the injection of electrons from CDs to HC-C₃N₅, but rather achieves an effective transfer of energy and prolongs the survival time of the excited state by non-radiative means. Combined with spectral overlap analysis, it was confirmed that the energy coupling between CDs and HC-C₃N₅ was mainly achieved through the FRET mechanism rather than the electron transfer route, which significantly enhanced the rate and stability of the photochromic response of HC-C₃N₅.

It should be emphasized that FRET itself does not directly induce macroscopic color change. Instead, FRET plays a crucial triggering role by facilitating non-radiative energy transfer from photoexcited CDs to the HC-C₃N₅ framework, thereby promoting charge separation and electron injection into the electron-accepting sites of HC-C₃N₅. Under light irradiation, the CDs are photoexcited and efficiently transfer excitation energy to HC-C₃N₅ via FRET, as evidenced by the shortened fluorescence lifetime and strong spectral overlap. The transferred energy accelerates the population of excited electrons in HC-C₃N₅, which are subsequently captured by intrinsic defect states and heptazine-based π-conjugated units, forming long-lived trapped electrons (electron storage states). These stored electrons induce the formation of mid-gap states and polaron-like species, leading to an enhanced visible-light absorption in the 450–550 nm region, which is directly responsible for the observed macroscopic color change. In the absence of photo-irradiation, the electrons gradually recombine with holes or are consumed by oxygen and other electron acceptors, resulting in the recovery of the original electronic structure and bleaching of the color. Notably, while the photoluminescence behaviour is strongly modulated by FRET-induced non-radiative pathways, the photochromic appearance change originates from light-induced charge accumulation and reversible electronic structure reconstruction in HC-C₃N₅ rather than FRET alone. Therefore, the photochromism of the CDs/HC-C₃N₅ system is governed by a synergistic mechanism involving FRET-assisted excitation, electron storage, and reversible charge recombination, whereas the photoluminescence modulation is primarily dictated by FRET-induced energy dissipation.

Besides, it is essential to distinguish the phenomenon observed in the CDs/HC-C₃N₅ system from classical organic photochromism. Classical systems typically undergo reversible structural isomerization upon irradiation.^62,63 Conversely, the color change in this composite is driven by photo-induced electron accumulation. The FRET-enhanced excitation, coupled with TEOA-mediated hole scavenging, leads to the injection and storage of electrons within the conduction band and defect states of the HC-C₃N₅ framework. This accumulation of electrons generates stable radical anions that absorb visible light. Therefore, the observed photochromism is intrinsically a manifestation of reversible electronic charging and discharging, rather than a molecular conformational change.

3.4. Potential applications of anti-counterfeiting

The photochromic responsiveness described above leads to the design of anti-counterfeiting and information encryption systems with high security. As illustrated in Fig. 4a, we constructed an anti-counterfeiting tag with layered response properties by material system design, where B-C₃N₅ (hull), HC-C₃N₅ (right sail) and CDs/HC-C₃N₅ (left sail) are used as the functional units, respectively. Under UV irradiation, the three materials exhibited different photochromic responses. In the initial state, the labels showed three different colors, and after 50 s of UV excitation, the hull remained orange, the right sail was yellow, and the left sail was light green due to the fast response of the FRET effect. After 180 s of continuous irradiation, the sail completed photochromism and formed a stable green color, which finally appeared as a three-color coexistence of orange-light green-dark green, which strictly matched with the thresholds of the material's photo responses (B-C₃N₅ < HC-C₃N₅ < CDs/HC-C₃N₅). As depicted in Fig. 4b, a dynamic information encryption system was constructed based on the differentiated photochromic properties of HC-C₃N₅ and CDs/HC-C₃N₅ by patterning the porous substrate. All units initially emit yellow, encoding the system as “88”, but after 1 min of UV irradiation, the CD/HC-C₃N₅ unit turns from yellow to green by the FRET enhancement mechanism, and the system is updated to “37”, while the HC-C₃N₅ unit still maintains the yellow color due to the slow light response, forming a static background. After 3 min of irradiation, the HC-C₃N₅ unit further turns green and the system is updated as green “88”, and finally, the green information can be spontaneously restored to the initial yellow information after 3 min of darkness, completing the conversion and reconstruction of information. In addition, a high-security message encryption system has been further designed and developed by integrating a ternary system, where bright yellow, dark yellow, and green represent “0”, “1”, and “2”, respectively. As illustrated in Fig. 4c, the system initially displays a yellow color, with the ternary message encoded as “111000100001” (corresponding to the photocode book “NAXB”, see Table S3). When the system is exposed to UV radiation, the initial message could gradually evolve into the word “JUST” upon first introducing TEOA. As more TEOA is added, the message could further transform into a “No information” state, at which point the system could cease recording any additional data. It is worth mentioning that the decoding process of the new message can be successfully decoded with the help of different codebooks implementing time-resolved message encryption techniques. These findings demonstrate the CDs/HC-C₃N₅-based system's capability to encrypt multilevel messages with high security, highlighting its potential for practical applications. As depicted in Fig. 4d, upon inscribing the word “JUST” on the substrate using C₃N₅, HC-C₃N₅, and CDs/HC-C₃N₅ inks, the information remains entirely concealed under both natural and UV illumination, and its decryption necessitates the selective introduction of TEOA into designated ink regions, thereby triggering a photochromic response through a stimulus-responsive mechanism. In Fig. 4e, the optical properties of CD/HC-C₃N₅ and HC-C₃N₅ inks are demonstrated, where the inscribed word “JUST” initially exhibits a dark yellow hue under sunlight, remains unchanged under UV irradiation, yet undergoes a distinct chromatic transition from dark yellow to green under UV light following the introduction of TEOA, with the HC-C₃N₅ ink displaying a less pronounced transformation, thereby underscoring the enhanced interaction between the HC-C₃N₅ matrix and TEOA upon the incorporation of CDs. If fluorophores such as B-C₃N₅, which intrinsically lack photochromic behavior, are applied to the substrate, data decryption becomes unfeasible. It is worth noting that these color-changing materials have intrinsic emission properties and take on different colors when interacting with TEOA, thus offering a wealth of possibilities for data decryption and anti-counterfeiting.

Fig. 4 (a) Application of disparate response-based CDs/HC-C₃N₅, HC-C₃N₅ and C₃N₅ in optical anticounterfeiting labels. (b) Information encryption application based on CDs/HC-C₃N₅ and HC-C₃N₅ functional units. (c) Time-resolved information encryption driven by UV light. (Bright yellow, dark yellow, and green colors were defined as “0”, “1”, and “2”, respectively.) (d) Schematic diagram of data writing and decryption processes. (e) Photographs of data under natural light and UV light, respectively.

4. Conclusions

In summary, this study successfully developed a FRET-based composite photochromic material, CDs/HC-C₃N₅, consisting of CDs and HC-C₃N₅, revealing the efficient energy transfer mechanism between CDs and HC-C₃N₅ and its significant enhancement of the photochromic properties. The introduction of CDs not only enhanced the light absorption and fluorescence emission intensity of HC-C₃N₅, but also accelerated the photochromic response through the FRET process, enabling the composite material to exhibit rapid, reversible color changes (from dark yellow to green) under UV light and recover to its initial state within 180 s, with excellent cycling stability. TPV test analysis further confirmed the non-radiative energy transfer pathway between CDs and HC-C₃N₅, eliminating interference from electron transfer and clarifying the dominant role of the FRET mechanism. Furthermore, based on the differentiated optical response characteristics of B-C₃N₅, HC-C₃N₅, and CDs/HC-C₃N₅, this study demonstrated their potential applications in multilayer anti-counterfeit tags and dynamic information encryption, successfully constructing a time-resolved encryption system that transitions from a “static background” to “dynamic information”, significantly enhancing anti-counterfeiting security. Comprehensively, the CDs/HC-C₃N₅ composite material not only provides new insights into the design of efficient photochromic materials but also lays the experimental foundation for the development of high-security anti-counterfeit technologies and information encryption systems. Despite these advances, the current reliance on TEOA restricts applications to solution-based environments. Future work will focus on developing solid-state electron donors or extending the excitation wavelength to the visible range, aiming to achieve intrinsic, self-sustaining photochromism for next-generation data storage and security devices.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Supplementary information (SI): specific experimental procedures for materials, characterization, and photochromic performance testing See DOI: https://doi.org/10.1039/d5nh00742a.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 22578190), the Science and Technology Planning Social Development Project of Zhenjiang City (No. JC2025004) and the Open Fund of the Key Laboratory of Solar Cell Electrode Materials in China Petroleum and Chemical Industry (2024A093).

Notes and references

1. L. Wang, W. Zhong, W. Gao, W. Liu and L. Shang, Chem. Eng. J., 2024, 479, 147490.
2. X. Li, M. Liu, J. Shao, H. Sun, Q. Zhang, D. Peng and F. Liu, Adv. Funct. Mater., 2024, 34, 2402603.
3. K. Zhang, Q. Fu, S. Sun, Z. Dong and M. Yue, ACS Appl. Mater. Interfaces, 2024, 16, 52833–52841.
4. J. Ni, Y. Zhuo, H. Zhang, Y. Zhang, H. Meng, K. Zhang, K. Zhu, L. Zeng, Y. Ai and Y. Li, Chem. Eng. J., 2025, 167799.
5. S. Sun, Q. Fu, Z. Dong and M. Yue, Chem. Eng. J., 2024, 499, 155972.
6. 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.
7. M. Zhu, B. Xu, T. Chen, J. Zhang and W. Sun, Adv. Funct. Mater., 2024, 34, 2406840.
8. J. Zhang, X. Liu, T. Chen, J. Han, T. Ahmed, X. Wang, Q. Wang, C. Fang and B. Xu, SusMat, 2025, e70023.
9. A. Parrot, A. Bernard, A. Jacquart, S. A. Serapian, C. Bo, E. Derat, O. Oms, A. Dolbecq, A. Proust, R. Métivier, P. Mialane and G. Izzet, Angew. Chem., Int. Ed., 2017, 56, 4872–4876.
10. C. B. Fan, L. Le Gong, L. Huang, F. Luo, R. Krishna, X. F. Yi, A. M. Zheng, L. Zhang, S. Z. Pu, X. F. Feng, M. B. Luo and G. C. Guo, Angew. Chem., Int. Ed., 2017, 56, 8015.
11. X. Hu, B. Fang, P. Li, J. Wang, J. Li, Y. Dai and M. Yin, Adv. Funct. Mater., 2025, 35, 2423793.
12. T. Chen, X. Mai, Y. Li, T. Wang, R. Gong, F. Chen, H. Huang, Z. Yan and F. Wang, Chem. Eng. J., 2024, 499, 155999.
13. Y. Zhou, Y. Lv, X. Guo, X. Wang, Z. Liu, J. Wang, Q. Sui, J. Wang, Z. Zeng and G. Cai, Adv. Mater., 2025, 37, 2502706.
14. Z. Xu, T. W. Lau, P. Xiong, J. Li, M. M.-J. Li, J. Yin and Y. Zhu, Nano Lett., 2024, 24, 9727–9733.
15. P. Huang, J. Huang, S. A. Pantovich, A. D. Carl, T. G. Fenton, C. A. Caputo, R. L. Grimm, A. I. Frenkel and G. Li, J. Am. Chem. Soc., 2018, 140, 16042–16047.
16. G. Liu, H. Yuan, Z. Wang, N. Qin, Y. Huang, Y. Cao, Y. Li, W. Lu, C. Zeng, Q. Liu and Z. Lu, Energy Storage Mater., 2024, 71, 103608.
17. S. Wang, Y. Wei, R. Wang, Q. Liu, Q. Wang, S. Luo, Y. Zhang, S. Yan, P. Hou, X. Liu, J. Guo and W. Mu, Composites, Part B, 2024, 285, 111759.
18. H. R. Kwon, J. W. Yang and H. W. Jang, Exploration, 2025, 5, 20240013.
19. P. Kumar, E. Vahidzadeh, U. K. Thakur, P. Kar, K. M. Alam, A. Goswami, N. Mahdi, K. Cui, G. M. Bernard, V. K. Michaelis and K. Shankar, J. Am. Chem. Soc., 2019, 141, 5415–5436.
20. A. M. Sadanandan, M. Fawaz, N. P. Dharmarajan, M. Huš, G. Singh, C. I. Sathish, B. Likozar, Z. Li, A. M. Ruban, C.-H. Jeon, J.-H. Yang, P. Kumar and A. Vinu, Appl. Catal., B, 2025, 362, 124701.
21. Z. Li, K. Tong, H. Dong, Y. Li, K. Wang, T. Ding, Y. Yun, Y. Zhou and M. Zhu, Angew. Chem., Int. Ed., 2025, 64, e202502390.
22. Y. Shen, X. Du, Y. Shi, L. J. Nguetsa Kuate, Z. Chen, C. Zhu, L. Tan, F. Guo, S. Li and W. Shi, Adv. Powder Mater., 2024, 3, 100202.
23. N. Tejwan, A. K. Saini, A. Sharma, T. A. Singh, N. Kumar and J. Das, J. Controlled Release, 2021, 330, 132–150.
24. Y. Pan, L. Qiao, Z. Zhu, X. Ran, Y. Kuang, Z. Chi, Y. He, R. Liu and L. Guo, Carbon, 2023, 213, 118212.
25. L. Ai, W. Xiang, J. Xiao, H. Liu, J. Yu, L. Zhang, X. Wu, X. Qu and S. Lu, Adv. Mater., 2024, 36, 2401220.
26. J. Li, H. Zhou, H. Qiu, Y. Yan, X. Wang, Z. Gao and Z. Wang, Coord. Chem. Rev., 2024, 503, 215642.
27. Y. Liu, D. Cheng, B. Wang, J. Yang, Y. Hao, J. Tan, Q. Li and S. Qu, Adv. Mater., 2024, 36, 2403775.
28. Q. Ran, X. Wang, P. Ling, P. Yan, J. Xu, L. Jiang, Y. Wang, S. Su, S. Hu and J. Xiang, Carbon, 2022, 193, 404–411.
29. J. Bian, Z. Wang, J. Ge, W. Gong, K. Nan, Z. Liu, J. Guo, Q. Zhao, X. Wang and L. Wan, Chem. Eng. J., 2025, 519, 165135.
30. Q. Jia, J. Ge, W. Liu, X. Zheng, S. Chen, Y. Wen, H. Zhang and P. Wang, Adv. Mater., 2018, 30, 1706090.
31. D. Chen, X. Guo, X. Sun, X. Feng, K. Chen, J. Zhang, Z. Zhu, X. Zhang, X. Liu, M. Liu, L. Li and W. Xu, Exploration, 2024, 4, 20230166.
32. J. Wu, S. Fu, X. Zhang, C. Wu, A. Wang, C. Li, G. Shan and Y. Liu, ACS Appl. Mater. Interfaces, 2020, 12, 6262–6267.
33. L. Yu, H. Liu, N. Feng, G. Yi, X. Xin, J. Hao and H. Li, Adv. Sci., 2024, 11, 2402565.
34. B. Zhou and D. Yan, Sci. China: Chem., 2021, 64, 509–510.
35. S. Yang, Y. Kuo and C.-C. Yang, Prog. Quantum Electron., 2025, 99, 100553.
36. L. Wu, C. Huang, B. P. Emery, A. C. Sedgwick, S. D. Bull, X.-P. He, H. Tian, J. Yoon, J. L. Sessler and T. D. James, Chem. Soc. Rev., 2020, 49, 5110–5139.
37. Y.-C. Liang, Q. Cao, K.-K. Liu, X.-Y. Peng, L.-Z. Sui, S.-P. Wang, S.-Y. Song, X.-Y. Wu, W.-B. Zhao, Y. Deng, Q. Lou, L. Dong and C.-X. Shan, ACS Nano, 2021, 15, 16242–16254.
38. A. Chen, R. Li, Y. Zhong, Q. Deng, X. Yin, H. Li, L. Kong and R. Yang, Sens. Actuators, B, 2022, 359, 131602.
39. Y. Jia, S. Wang, W.-J. Wang, B.-G. Li and S. Zhu, Macromolecules, 2019, 52, 7920–7928.
40. J. Lu, Y. Shi, Z. Chen, X. Sun, H. Yuan, F. Guo and W. Shi, Chem. Eng. J., 2023, 453, 139834.
41. J. Lu, Z. Chen, Y. Shen, H. Yuan, X. Sun, J. Hou, F. Guo, C. Li and W. Shi, J. Colloid Interface Sci., 2024, 670, 428–438.
42. C. Peng, L. Han, J. Huang, S. Wang, X. Zhang and H. Chen, Chin. J. Catal., 2022, 43, 410–420.
43. C. Fu, M. Zhao, X. Chen, G. Sun, C. Wang and Q. Song, Appl. Catal., B, 2023, 332, 122752.
44. H. Yu, R. Shi, Y. Zhao, T. Bian, Y. Zhao, C. Zhou, G. I. N. Waterhouse, L.-Z. Wu, C.-H. Tung and T. Zhang, Adv. Mater., 2017, 29, 1605148.
45. L. Ouyang, M. Ng, Z.-H. Zhou, H. Wu, M.-C. Tang and S. S. Chen, Adv. Sci., 2025, 12, 2417390.
46. E. Wirnhier, M. B. Mesch, J. Senker and W. Schnick, Chem. – Eur. J., 2013, 19, 2041–2049.
47. M. Shi, Q. Gao, J. Rao, Z. Lv, M. Chen, G. Chen, J. Bian, J. Ren, B. Lu and F. Peng, J. Am. Chem. Soc., 2023, 146, 1294–1304.
48. Y. Xu, M. Fan, W. Yang, Y. Xiao, L. Zeng, X. Wu, Q. Xu, C. Su and Q. He, Adv. Mater., 2021, 33, 2101455.
49. L. Lin, W. Ren, C. Wang, A. M. Asiri, J. Zhang and X. Wang, Appl. Catal., B, 2018, 231, 234–241.
50. Y. Shen, R. Xu, P. Shan, S. Zhang, L. Sun, H. Xie, F. Guo, C. Li and W. Shi, Small, 2024, 20, 2401566.
51. A. T. Smith, K. Shen, Z. Hou, S. Zeng, J. Jin, C. Ning, Y. Zhao and L. Sun, Adv. Funct. Mater., 2021, 32, 2110285.
52. P. Audebert, E. Kroke, C. Posern and S.-H. Lee, Chem. Rev., 2021, 121, 2515–2544.
53. X. Miao, D. Qu, D. Yang, B. Nie, Y. Zhao, H. Fan and Z. Sun, Adv. Mater., 2018, 30, 1870002.
54. L. Jiang, H. Ding, S. Lu, T. Geng, G. Xiao, B. Zou and H. Bi, Angew. Chem., Int. Ed., 2019, 59, 9986–9991.
55. S. Sun, Q. Fu, Z. Dong and M. Yue, Chem. Eng. J., 2024, 499, 155972.
56. T. Feng, Q. Zeng, S. Lu, X. Yan, J. Liu, S. Tao, M. Yang and B. Yang, ACS Photonics, 2018, 5, 502–510.
57. Y. Zheng, K. Arkin, Y. Bei, X. Ma and Q. Shang, Matter, 2023, 6, 4339–4356.
58. Y. Huang, Y. Zhu, C. Li, G. Zheng, C. Zhang, Y. Nan, Q. Lou, X. Chen, Y. Zhu and J. Zhu, J. Mater. Chem. C, 2024, 12, 14246–14252.
59. L. Jiang, H. Ding, S. Lu, T. Geng, G. Xiao, B. Zou and H. Bi, Angew. Chem., 2020, 132, 10072–10077.
60. Y.-C. Liang, Q. Cao, K.-K. Liu, X.-Y. Peng, L.-Z. Sui, S.-P. Wang, S.-Y. Song, X.-Y. Wu, W.-B. Zhao and Y. Deng, ACS Nano, 2021, 15, 16242–16254.
61. H. Yuan, H. Sun, Y. Shi, J. Wang, A. Bian, Y. Hu, F. Guo, W. Shi, X. Du and Z. Kang, Chem. Eng. J., 2023, 472, 144654.
62. T. Zhou, W. Zhao, Z. Wang, R. Li, X. Feng, Y. Du, L. Liu, J. Zhang, G. Yin and K. Wang, Adv. Energy Mater., 2025, 15, 2404850.
63. Z. Guo, Y. Bian, L. Zhang, J. Zhang, C. Sun, D. Cui, W. Lv, C. Zheng, W. Huang and R. Chen, Adv. Mater., 2024, 36, 2409361.

---