Sustaining Radicals in Aqueous Media Through Formation of a Trigger‐Inclusive Microenvironment in Hydrogen‐Bonded Frameworks for Antibacterial Activity

ABSTRACT

The stabilization of radical species in aqueous environments remains a central challenge for radical‐based antibacterial applications, as rapid quenching severely limits its efficacy. Herein, we report a strategy to sustain radical by coupling diffusional confinement with electronic delocalization in π‐conjugated hydrogen‐bonded organic frameworks (HOFs). Using a pyrene‐based HOF (PFC‐1) as a model system, we systematically modulate the longitudinal dimensions of 1D channels to generate frameworks with three distinct aspect ratios while maintaining topology. The higher‐aspect‐ratio channels impose a more pronounced kinetic barrier to the outward diffusion of radical‐generating trigger (ceric ammonium nitrate), thereby retaining a trigger‐inclusive oxidative microenvironment. Simultaneously, the extended π‐π stacking enhances electronic delocalization, stabilizing the resulting radicals through favorable host‐guest interactions. As a result, radicals confined within the longest‐channel framework remain persistent for over 30 days in water and up to 8 months under ambient conditions. This prolonged radical persistence enables potent antibacterial activity against Escherichia coli via radical‐induced oxidative stress. This work establishes a trigger‐inclusive microenvironment capable of simultaneously retaining both the radical species and its trigger as a generalizable design principle for engineering persistent radical states.

We dynamically stabilize radicals by engineering high‐aspect‐ratio 1D channels in hydrogen‐bonded organic frameworks (PFC‐1‐L‐R). The elongated channels create diffusion barriers that prevent trigger leaching, while long‐range π‐π stacking promotes electronic delocalization to strengthen host‐trigger interactions. This dual mechanism achieves remarkable stability (30+ days in water) with potent antibacterial activity.

1. Introduction

The rapid escalation of antimicrobial resistance (AMR) has exposed a fundamental limitation of conventional antibacterial strategies that rely on specific biochemical targets, which are inherently vulnerable to adaptive resistance. In contrast, radical‐based antibacterial systems offer a target‐agnostic mode of action, inducing concurrent damage to bacterial membranes, enzymes, and genomic deoxyribonucleic acid (DNA) through highly reactive intermediates [1]. Despite this conceptual advantage, the practical deployment of radical systems in physiological environments remains elusive, primarily due to the intrinsic difficulty in sustaining reactive radical species over relevant timescales in aqueous media [2]. Current efforts to address this challenge have largely focused on passive stabilization strategies, including steric protection and host‐guest encapsulation within rigid frameworks such as metal‐organic frameworks (MOFs), to physically shield the radical center from quenching [2, 3, 4, 5]. While these approaches can retard radical decay, They typically often overlooked a critical factor: the persistence of the radical‐generating trigger itself. In most confined systems, small‐molecule oxidants or triggers rapidly diffuse out of the host framework, leading to premature depletion of the oxidative microenvironment and, consequently, transient radical generation. We posit that a more robust design paradigm for radical‐based antibacterial platforms must therefore go beyond passive isolation and instead incorporate a trigger‐inclusive microenvironment that simultaneously retains both radicals and their triggers, enabling the framework‐stabilized triggers to sustain radical generation. Achieving such a configuration, however, presents a formidable challenge, particularly in open porous frameworks where the diffusion of small molecules is typically rapid. Although hindered molecular transport in high‐aspect‐ratio nanopores is well established within the framework of Fickian diffusion and steric confinement, its deliberate exploitation as a kinetic handle to regulate radical persistence has received surprisingly limited attention. Here, we hypothesize that π‐conjugated hydrogen‐bonded organic frameworks (HOFs) with elongated 1D channels provide a unique platform to couple transport kinetics with electronic stabilization of radical states. Specifically, extending the channel length introduces a substantial diffusional barrier that kinetically suppresses the leaching of small‐molecule oxidants, thereby maintaining a localized oxidative microenvironment. Concurrently, the expansion of the ordered π‐π stacking domain enhances electronic delocalization along the channel axis, which stabilizes the resulting radicals through a more favorable electronic landscape and strengthened host‐guest interactions. To test this concept, we employed our previously reported pyrene‐based HOF, PFC‐1, which exhibits a pronounced tendency for anisotropic growth along its π‐π stacking direction [6, 7, 8, 9]. By systematically modulating the longitudinal dimensions of the crystalline domains, we synthesized three morphological variants—PFC‐1‐S‐R, PFC‐1‐M‐R, and PFC‐1‐L‐R—while preserving the underlying framework topology. This model system enables a direct correlation between channel aspect ratio and radical persistence. Notably, the high‐aspect‐ratio channels in PFC‐1‐L‐R give rise to a synergistic stabilization mechanism: (i) kinetic confinement that suppresses the outward diffusion of ceric(IV) ammonium nitrate (CAN), sustaining a trigger‐inclusive oxidative microenvironment; and (ii) enhanced π‐electron delocalization that electronically stabilizes the generated radicals. As a result, radicals confined within PFC‐1‐L‐R exhibit exceptional persistence, remaining stable for over 30 days in aqueous media and up to 8 months under ambient conditions—substantially exceeding the performance of most reported materials (Table S1). Benefiting from this robust radical persistence, PFC‐1‐L‐R displays potent antibacterial activity against Escherichia coli (E. coli). Mechanistic investigations indicate that the persistent radical centers initiate an interfacial oxidative stress, triggering an endogenous reactive oxygen species (ROS) surge that leads to DNA fragmentation. This study demonstrates the innovative integration of diffusional confinement with electronic delocalization in high‐aspect‐ratio channels as a complementary strategy for radical stabilization, providing a foundation for designing resilient radical‐based therapeutics (Scheme 1).

SCHEME 1.

a) Schematic illustration of water‐stable radicals confined within PFC‐1‐L‐R. b) Schematic representation of the inhibitory mechanism of PFC‐1‐L‐R against E. coli.

2. Results and Discussion

Three distinct particle sizes of PFC‐1, designated as PFC‐1‐S (≈800 nm), PFC‐1‐M (≈20 µm), and PFC‐1‐L (≈150 µm), were synthesized by adapting a previously reported protocol [9, 10]. The precursor 1,3,6,8‐tetrakis(p‐benzoic acid)pyrene (H₄TBAPy), synthesized and confirmed by ¹H NMR spectroscopy (Figure S1), was dissolved in DMF, wherein the degree of anisotropic growth along the π‐π stacking axis was meticulously regulated by modulating the addition rate of methanol. The size‐dependent morphology, phase purity, and crystallinity of these PFC‐1 samples were systematically characterized by scanning electron microscopy (SEM) and powder X‐ray diffraction (PXRD), respectively (Figures S2 and S3). Chemical oxidation of the as‐synthesized PFC‐1 samples were performed using CAN in acetonitrile, yielding the corresponding radical‐active counterparts: PFC‐1‐S‐R, PFC‐1‐M‐R, and PFC‐1‐L‐R. This oxidative process was accompanied by a distinct color transition from yellow to green across all samples, providing a macroscopic manifestation of the underlying electronic transformation (Figure S4). Notably, the oxidative process did not compromise the structural integrity of the framework, with both the morphology and crystallinity of the radicalized PFC‐1 remaining robustly preserved (Figures S5 and S6). In addition, while the pristine H₄TBAPy building blocks within PFC‐1 initially adopt a closed‐shell layered configuration, the introduction of Ce (IV) triggers a concerted one‐electron transfer between adjacent pyrene units. This process facilitates the transformation of the framework into a system characterized by [TBAPy]₂ ⁺ radical‐cation pairs [6]. To balance the resulting positive charge, NO₃ ⁻ species are incorporated as counterions, ensuring neutrality within the radical‐cationic network. This mechanism was directly corroborated by X‐ray photoelectron spectroscopy (XPS) and Fourier transform infrared (FT‐IR) analyses, which clearly identified the presence of NO₃ ⁻ species (Figures S7 and S8).

Initially, the stability of the radicals within PFC‐1‐S/M/L‐R in aqueous media was investigated. As shown in Figure 1a, the radical form of PFC‐1‐S‐R underwent a discernible color transition from green to orange‐yellow within 6 h of water exposure, reverting completely to its original bright yellow after 12 h. Intriguingly, the stability of the radical state in aqueous media was significantly enhanced by the preferential elongation extension of the crystals along the π‐π stacking axis. For PFC‐1‐M‐R (≈20 µm), the green radical state persisted for 12 h before any noticeable fading (Figure 1b). Most remarkably, when the crystals were elongated to ≈150 µm (PFC‐1‐L‐R), the radical species exhibited exceptional persistence, remaining visually unchanged for at least 30 d in water (Figure 1c). To corroborate these colorimetric observations, the decay kinetics of the radical signals were monitored via electron paramagnetic resonance (EPR) spectroscopy, which revealed a signal attenuation trend perfectly congruent with the observed color transitions (Figure 1d–f). PXRD patterns of PFC‐1‐S‐R, PFC‐1‐M‐R, and PFC‐1‐L‐R recorded after immersion in water confirmed that their crystallinity was well preserved, with no apparent structural deterioration observed under aqueous conditions (Figure S9). Beyond its robust performance in aqueous media, the atmospheric stability of PFC‐1‐L‐R was evaluated under ambient conditions. Impressively, no appreciable decrease in EPR signal intensity was detected over a period of at least eight months (Figure 1g), signifying the exceptional ambient stability of the radicals within this long‐range π‐stacking framework. In sharp contrast, PFC‐1‐S‐R and PFC‐1‐M‐R exhibited markedly inferior radical persistence (Figure S10). Specifically, the EPR signal of PFC‐1‐S‐R was nearly exhausted within only two months, whereas PFC‐1‐M‐R underwent a 50% attenuation by the fourth month and a near‐complete loss of radical character by the eighth month. To quantitatively substantiate these stability profiles, we performed a kinetic analysis by fitting the time‐dependent EPR signal intensities (Figure S11) [11]. This analysis yielded distinct decay rate constants (k) that reflect the divergent stability of the three samples. PFC‐1‐S‐R displayed rapid signal decay with a rate constant of k = ‐0.43, whereas PFC‐1‐M‐R showed a significantly attenuated decay rate of k = ‐0.07. Most notably, PFC‐1‐L‐R exhibited the highest kinetic persistence with a minimal decay constant of k = ‐0.02, allowing it to retain the bulk of its radical character over extended durations. These results clearly delineate an aspect‐ratio‐dependent stability profile, in which enhanced crystal growth along the π‐stacking axis dramatically extends radical longevity by increasing the channel aspect ratio.

FIGURE 1.

Schematic representation of the time‐dependent color evolution of a) PFC‐1‐S‐R, b) PFC‐1‐M‐R, and c) PFC‐1‐L‐R upon immersion in water. Time‐dependent electron paramagnetic resonance (EPR) spectra of d) PFC‐1‐S‐R, e) PFC‐1‐M‐R, and f) PFC‐1‐L‐R after immersion in water. g) Time‐dependent EPR spectra of PFC‐1‐L‐R stored in air. h) N₂ sorption isotherms of PFC‐1‐S/M/L and PFC‐1‐S/M/L‐R. i) The cerium concentration was determined from the supernatants of PFC‐1‐S/M/L‐R after immersion in water.

To further elucidate the underlying mechanism governing the extraordinary radical stability afforded by the elongation of PFC‐1 along the π–π stacking axis, we performed an in‐depth characterization of its intrinsic structural features and electronic properties. As illustrated in Figure 1h and Figure S12, the pristine PFC‐1‐S/M/L samples exhibited nearly identical N₂ uptake capacities. However, upon radicalization via CAN treatment, a pronounced divergence in porosity was observed. Notably, PFC‐1‐L‐R displayed the most significant reduction in N₂ adsorption relative to its pristine counterpart, followed by PFC‐1‐M‐R and PFC‐1‐S‐R. This trend indicates that the anisotropic elongation along the π‐stacking axis facilitates a higher loading of CAN species within the 1D channels. This observation was further corroborated by quantifying the nitrate concentration within the digested samples. As shown in Figure S13 and Table S2, ion chromatography (IC) analysis revealed that PFC‐1‐L‐R possesses the highest NO₃ ⁻ concentration (347.5 ppm), whereas significantly lower values were recorded for PFC‐1‐M‐R (241.3 ppm) and PFC‐1‐S‐R (134.1 ppm). These quantitative data provide definitive evidence that longer crystalline domains enhance the incarceration of oxidant species, thereby establishing a stable, trigger‐inclusive microenvironment that sustains the radical state. Moreover, the leaching kinetics of cerium species into the aqueous phase were quantified via Inductively coupled plasma atomic emission spectrometry (ICP‐AES). As shown in Figure 1i and Table S3, after 24 h of water immersion, PFC‐1‐L‐R exhibited the most suppressed leaching profile with a cerium concentration of only 13.1 mg L⁻¹, significantly lower than those recorded for PFC‐1‐S‐R (56.5 mg L⁻¹) and PFC‐1‐M‐R (28.2 mg L⁻¹). This disparity became even more pronounced over extended periods. By the third day, the cerium concentration in the PFC‐1‐S‐R supernatant had surged 5.4‐fold to 306.8 mg L⁻¹, whereas PFC‐1‐L‐R maintained a far more controlled release, reaching only 49.5 mg L⁻¹ (a 3.7‐fold increase). These results collectively demonstrate that the elongated 1D channels, arising from long‐range π‐stacking, not only facilitate higher oxidant loading but also effectively inhibit leaching through a capillary‐like confinement effect. Such robust entrapment is instrumental in establishing a persistent, trigger‐inclusive microenvironment, ultimately endowing the framework with exceptional longevity in aqueous media.

Moreover, in sharp contrast to PFC‐1‐S‐R and PFC‐1‐M‐R, PFC‐1‐L‐R exhibits markedly enhanced optical absorption across the visible to near‐infrared region (Figure 2a). This pronounced red‐shift signifies the dominance of low‐energy optical transitions, characteristic of effective exciton delocalization across extended structural domains [12, 13]. Such delocalization, facilitated by the long‐range π‐stacking in PFC‐1‐L‐R, effectively lowers the localization energy of unpaired electrons, thereby thermodynamically favoring radical stabilization. Further insights into the electronic environment were gleaned from high‐resolution XPS analysis (Figure 2b). The C 1s spectra of PFC‐1 were deconvoluted into four distinct components: the aromatic C─C skeleton (284.8 eV), C─O (285.5 eV), C═O (289.4 eV), and the characteristic π–π* shake‐up satellite (291.4 eV) [14, 15]. Notably, PFC‐1‐L‐R exhibits a significantly diminished satellite peak area. Given that the intensity of the π–π* transition is intrinsically coupled to the spatial distribution of valence electrons and π‐conjugation, this attenuation suggests a profound reconfiguration of the electronic structure within the elongated framework [16]. Complementing this, the primary C 1s peak exhibits a progressive increase in asymmetry and a concomitant decrease in intensity with framework elongation (Figure S14). Peak fitting reveals a distinct component at ≈286.3 eV (highlighted in red), which attains its maximum relative area in PFC‐1‐L‐R, directly mirroring the trend of oxidant loading. Since this component correlates inversely with the contraction of the O 1s peak area attributed to C─O groups, we further examined the O 1s spectra to clarify its origin (Figure 2c). Notably, PFC‐1‐L‐R shows a significant contraction in the O 1s peak area attributed to C─O groups compared to PFC‐1‐S‐R and PFC‐1‐M‐R (from 1.45 to 1.19, then to 0.81). This downward trend, which inversely correlates with the growth of the newly emerged C 1s component, strongly indicates that this feature arises from a specific electronic interaction between the C─O moieties of the framework and the encapsulated CAN species. Collectively, these results indicate that the elongated 1D channels in PFC‐1‐L‐R not only facilitate superior oxidant confinement but also induce a significant redistribution of electron density at the host‐guest interface, creating an energetic environment conducive to radical persistence in aqueous media. Kelvin probe force microscopy (KPFM) was employed to further investigate the local potential profile (Figure 2d–f and Figure S15). The results reveal a progressive increase in surface potential contrast (ΔE) from PFC‐1‐S‐R (ΔE ≈ 22.6 mV) to PFC‐1‐M‐R (ΔE ≈ 34.2 mV), with PFC‐1‐L‐R reaching the maximum value of ΔE ≈ 44.1 mV. Such a pronounced enhancement in ΔE is indicative of a more potent built‐in electric field within the elongated framework [17, 18]. This intensified local field, likely stemming from the robust host‐guest interactions and extended electronic delocalization, may create a favorable electrostatic environment that facilitates the spatial separation and long‐term stabilization of radical species. Furthermore, Mott‐Schottky analysis reveals a progressive deepening of the HOMO level from ‐0.53 eV (PFC‐1‐S‐R) to ‐1.21 eV (PFC‐1‐L‐R) as the 1D channels extend (Figure S16). This downward shift in the energy level indicates a decreased propensity for electron loss, providing a thermodynamic rationale for the remarkable stability of the radical species within the PFC‐1‐L‐R framework.

FIGURE 2.

a) UV–vis–NIR absorption spectra of PFC‐1‐L, PFC‐1‐S‐R, PFC‐1‐M‐R, and PFC‐1‐L‐R. b) c) X‐ray photoelectron spectroscopy (XPS) of C 1s and O 1s spectra for PFC‐1‐S‐R, PFC‐1‐M‐R, and PFC‐1‐L‐R. d) e) f) Kelvin probe force microscopy (KPFM) surface potential maps of PFC‐1‐S‐R, PFC‐1‐M‐R, and PFC‐1‐L‐R.

The biological impact of radical species is dictated not merely by their intrinsic reactivity but, more critically, by their persistence and lifetime within competitive aqueous media [19]. Consequently, we evaluated the antibacterial performance of these stabilized radical HOFs against E. coli using a standard colony‐counting assay in water. As illustrated in Figure 3a, antimicrobial efficacy positively correlates with the aqueous stability of the series, with PFC‐1‐L‐R demonstrating the most potent activity (Figure S17 and Table S4). To definitively attribute this antimicrobial activity to the radical‐state framework, dose‐response assays were performed. Increasing the concentration of PFC‐1‐L‐R from 50 to 400 µg mL⁻¹ resulted in a systematic reduction of colony‐forming units (CFUs), confirming that the biocidal efficacy is a direct function of the radical‐HOF concentration (Figure S18 and Table S5) [20]. Systematic control experiments revealed that neither CAN nor pristine PFC‐1‐L exhibited discernible biocidal activity independently (Figure 3b,c), effectively ruling out individual precursors and highlighting the radical‐state framework as the active species (Figures S19 and S20, and Tables S6 and S7). Furthermore, the significant attenuation of the EPR signal following incubation with E. coli provides direct evidence of interfacial electron transfer, suggesting that PFC‐1‐L‐R functions as a potent oxidant that induces sustained oxidative stress on the bacteria (Figure S21) [21]. To evaluate intracellular oxidative stress, ROS levels were monitored using the 2′,7′‐dichlorofluorescent yellow diacetate (DCFH‐DA) fluorescent probe and visualized via confocal laser scanning microscopy (CLSM) [22]. While control group exhibited negligible fluorescence, PFC‐1‐L‐R treatment induced intense green fluorescence (Figure 3d), demonstrating that the radical‐state framework triggers a pronounced oxidative stress response in E. coli, which in turn induces the massive intracellular accumulation of ROS. In addition, SEM imaging was employed to visualize the morphological impact on the bacteria. While control groups and PFC‐1‐L‐treated samples retained their structural integrity, E. coli exposed to PFC‐1‐L‐R displayed significant membrane lysis [23]. This drastic transition substantiates that the radical‐state framework effectively breaches the bacterial protective barriers, leading to irreversible structural damage (Figure S22). To further assess the adaptive capacity of E. coli, several enzymatic and metabolic biomarkers, including superoxide dismutase (SOD), peroxidase (POD), malondialdehyde (MDA), and Na⁺/K⁺‐ATPase, were monitored across a concentration gradient of PFC‐1‐L‐R [24]. As illustrated in Figure 3e, compared to control, SOD, POD, and MDA levels were significantly elevated, signifying the activation of antioxidant defense system and the occurrence of radical‐mediated lipid peroxidation. Conversely, the suppressed ATP activity indicates that PFC‐1‐L‐R effectively inhibits bacterial energy metabolism and depletes intracellular ATP. Notably, SOD, POD, and MDA levels exhibited a distinctive biphasic trend: an initial ascent followed by a subsequent decline at higher concentrations. This profile suggests that at lower loadings, the bacterial antioxidant defense system is activated to scavenge accumulating ROS; however, elevated concentrations of PFC‐1‐L‐R impose a level of sustained oxidative stress that surpasses the cellular buffering capacity, leading to irreversible lipid peroxidation and system failure. Concurrently, the ATP levels exhibited a U‐shaped profile, characterized by an initial sharp decrease followed by a slight rebound at higher concentrations, indicating a severe disruption of energy metabolism that may trigger complex metabolic flux redirections as the cells approach death [25]. Furthermore, agarose gel electrophoresis analysis showed that, compared with the untreated control group, the band corresponding to intact genomic DNA became markedly weaker after treatment with PFC‐1‐L‐R, accompanied by obvious smearing in the lower molecular‐weight region, indicating fragmentation and degradation of genomic DNA induced by radical species (Figure S23) [26]. Collectively, these findings underscore that PFC‐1‐L‐R exerts bactericidal activity via oxidative stress that overwhelms endogenous defenses, leading to irreversible membrane rupture and metabolic collapse. Importantly, CCK‐8 assays using H9c2 rat cardiomyocytes revealed excellent cytocompatibility, with only marginal reductions in viability and unperturbed cell morphology (Figure S24), highlighting the substantial promise of PFC‐1‐L‐R for advanced biomedical applications [27].

FIGURE 3.

a) The number of colonies measured in colony‐forming units (CFUs) after being treated with control group (CK), PFC‐1‐S‐R, PFC‐1‐M‐R, and PFC‐1‐L‐R. b) CFU results of E. coli treated with CAN at different concentrations. c) CFU results after different being treated with control group, PFC‐1‐L, and PFC‐1‐L‐R. d) 2′,7′‐dichlorofluorescein (DCF) fluorescence images and corresponding bright‐field images after being treated with control group and PFC‐1‐L‐R. e) Superoxide dismutase (SOD), peroxidase (POD), malondialdehyde (MDA), and ATP levels in bacteria under different PFC‐1‐L‐R concentrations. Data represent mean ± SD (n = 3). Statistical significance was determined by one‐way analysis of variance (ANOVA) followed by Tukey's test (p < 0.05), with different letters indicating significant differences between groups.

3. Conclusion

In summary, we present a strategy for stabilizing radical species in aqueous environments by tailoring the aspect ratio of 1D channels within hydrogen‐bonded organic frameworks. Our findings demonstrate that the elongation of the crystalline domains exerts a synergistic dual effect: structurally, it imposes long‐range diffusional confinement that restricts the escape of redox triggers within the channels; electronically, it promotes exciton delocalization and establishes a favorable electrostatic potential, contributing to the thermodynamic stabilization of the radical state. The resulting trigger‐inclusive microenvironment affords exceptional radical persistence (>30 d in water) and sustained antibacterial activity against E. coli. Mechanistically, the biocidal efficacy of PFC‐1‐L‐R stems from oxidative stress, where persistent interfacial electron transfer triggers excessive intracellular ROS accumulation and subsequent irreversible membrane damage. This work establishes a trigger‐inclusive microenvironment capable of simultaneously retaining both the radical species and its trigger as a robust paradigm for stabilizing radicals, opening new avenues for designing durable functional materials for catalysis, sensing, and therapeutic applications.

Author Contributions

T.‐F. Liu conceived and designed the project. J.‐L. Li performed the experiments, and wrote the original draft. A.‐A. Zhang analyzed the data, revised and improved the manuscript. Y. Zou performed the biological experiments. X.‐P. Gou and Y.‐L. Li collected the SEM data. X.‐H. Pan supervised the biological experiments.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: smll73826‐sup‐0001‐SuppMat.docx.

Data Availability Statement

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