Nuclear Spin Relaxation Modulation of Fluorinated Deep Eutectic Solvents for Radiofrequency Information Encoding and Display

ABSTRACT

Nuclear spin properties underpin nuclear magnetic resonance. The tunability of ¹⁹F spin systems renders fluorinated materials versatile platforms for diverse applications. Although frequency encoding of ¹⁹F nuclei enables high‐density molecular encoding, the use of spin relaxation remains limited. Here, a ¹⁹F magnetic resonance display platform that exploits spin relaxation modulation through synergistic material‐level design and instrument‐level pulse sequence control is presented. Incorporation of paramagnetic metal ions into fluorinated deep eutectic solvents enables precise tuning of ¹⁹F relaxation times for information encoding. Tailored relaxation‐selective pulse sequences allow selective visualization of the encoded information. Iron‐containing fluorinated deep eutectic solvents further enable reversible data editing via photoredox responses. In addition, UV‐printable paramagnetic fluorinated ionogels allow rapid and rewritable information display. This work expands nuclear magnetic information encoding by introducing controllable ¹⁹F spin relaxation as an additional information domain, thereby converting spin relaxation from a physical parameter into an information carrier and enabling a selective, responsive, and non‐luminescent display platform well suited for high‐confidential anticounterfeiting and information display.

Paramagnetic metal ions incorporated into fluorinated deep eutectic solvents precisely modulate ¹⁹F spin relaxations, enabling manual information encoding in the relaxation domain. Tailored relaxation‐selective pulse sequences provide selective readout of the encoded information. This bidirectional control, synergizing materials and instrumentation, converts spin relaxation from a passive physical parameter into a writable storage medium, enabling highly confidential radiofrequency information display.

1. Introduction

Spin is a fundamental quantum property of microscopic particles and provides a physical basis for various material functionalities at the atomic scale [1, 2]. In atomic nuclei, nuclear spin gives rise to a magnetic moment whose behavior is affected by the local molecular and electronic environment [3]. When placed in a static magnetic field, this magnetic moment can be coherently perturbed by radiofrequency (RF) irradiation, producing measurable signals that underpin nuclear magnetic resonance (NMR) [4]. The ability to manipulate and probe nuclear spins has led to broad scientific and technological advances [5]. NMR spectroscopy reveals the tiny differences of resonance frequencies from different nuclei, known as chemical shifts, for determining molecular structures [6, 7]. Clinical magnetic resonance imaging (MRI) exploits differences in proton spin relaxation, often enhanced by designed contrast agents [8, 9], to generate anatomical and pathological contrast that guide medical diagnosis [10, 11]. Beyond these applications, the manipulation of nuclear spins is being actively explored as a platform for quantum computing [12, 13]. In fact, many properties of spin have motivated the development of molecular and material design strategies for various functionalities [14].

Among various nuclear spin species, ¹⁹F is particularly attractive for NMR‐enabled functional materials owing to its high sensitivity (94% that of ¹H) and negligible environmental background [15]. Meanwhile, its wide chemical shift range (>350 ppm) [16] and relatively long relaxation times (usually from hundreds of milliseconds to seconds) provide ample opportunities for precise modulation of spin properties through molecular design [17]. For instance, transverse relaxation (T ₂) can be shortened by dipolar interactions or paramagnetic relaxation enhancement (PRE) [18], leading to signal quenching that could be exploited to develop stimulus‐responsive probes for accurate, background‐free detection and imaging. Longitudinal relaxation (T ₁) can be tuned by PRE [19, 20] or by restricting molecular motions [21], thereby accelerating signal acquisition for better NMR sensitivity. Apart from spin relaxation, resonance frequencies of ¹⁹F nuclei can be engineered through shielding effects [22], pseudocontact shifts [23], or paramagnetic chemical exchange [24], generating various distinguishable ¹⁹F MR signals. Such multiplexity has enabled a number of applications ranging from multiplexed sensing [24, 25, 26, 27, 28] to high‐density data storage [29, 30]. Together, these advances underscore the high programmability of ¹⁹F spin properties and establish fluorinated materials as tunable platforms for functional designs.

Among the various opportunities enabled by this tunability, ¹⁹F molecular encoding, which refers to encoding and storing information within fluorinated molecules [31, 32, 33], has emerged as a particularly promising direction, offering high information density by exploiting the large number of distinguishable ¹⁹F NMR signals across the wide chemical shift range [16]. Unlike optical or biochemical encoding approaches [34, 35, 36, 37, 38], the radiofrequency nature of NMR signals allows deep penetration (meters) through many nonconductive macroscopic obstacles [39], imparting RFID‐like capabilities [33] and enabling forensic‐level security with elevated confidentiality. However, current studies are mostly concentrating on the frequency domain of the ¹⁹F spin system [30, 33], whereas other intrinsic properties, especially spin relaxation, are considerably underexplored, which offers almost untapped opportunities for significant advancement.

In this work, we developed paramagnetic fluorinated materials with programmable ¹⁹F relaxation and established INSPIRE (Information Showcasing with Paramagnetic Fluorinated Deep Eutectic Solvent via MR Imaging in a Relaxation‐Selective Manner), a relaxation‐selective magnetic resonance display strategy that synergizes material‐level relaxation control with instrument‐level pulse sequence design for information encoding. Compared to clinical MRI, which unidirectionally extracts naturally embedded relaxation information, our approach enables bidirectional write‐in and readout of editable spin relaxation states (Figure 1). Paramagnetic metal ions were dissolved in fluorinated deep eutectic solvents (FDESs) to form paraFDESs, thereby allowing composition‐programmed control of ¹⁹F relaxation for encoding information in the “relaxation domain”. By adjusting ¹⁹F MRI pulse sequence parameters, distinct relaxation‐specific patterns could be selectively visualized from structurally identical fluorinated molecules. This strategy introduces an additional physical dimension for information encoding via ¹⁹F MRI, significantly enhancing both security and capacity. In addition, the reversible photoredox activity of iron‐based paraFDESs allows responsive relaxation control by UV irradiation or oxygen exposure and enables dual encryption together with time‐dependent pattern display. Moreover, FDESs were formulated into UV‐printable paramagnetic fluorinated ionogels (paraFGels), yielding a processable and rewritable format for practical labeling and rapid readout, providing a dynamic and secure non‐luminescent materials platform.

FIGURE 1.

Conceptual comparison between clinical MRI and this work. Clinical MRI unidirectionally extracts the relaxation differences of naturally occurring ¹H nuclei (mostly in H₂O) and reflects the corresponding anatomical information in the form of images for diagnosis. By exploiting the paramagnetic relaxation enhancement (PRE) effect of ferric ions, we enable a bidirectional process in which information can be both encoded in and retrieved from manually modulated ¹⁹F spin relaxation. Encoding is achieved by incorporating paramagnetic ions (e.g., Fe³⁺ ions) into fluorinated deep eutectic solvent (FDES) to generate paraFDES with programmed ¹⁹F relaxation times. Decoding is realized via tailored ¹⁹F MRI in a relaxation‐selective manner.

2. Results and Discussion

2.1. Information Showcasing with Paramagnetic Fluorinated DES and MR Imaging in a Relaxation‐Selective Manner (INSPIRE)

Fluorinated deep eutectic solvents (FDESs) are promising ¹⁹F MRI signal sources because they integrate high fluorine content, tunable ¹⁹F chemical shift [30], and the salt‐dissolving capability of DESs within a homogeneous liquid phase [40], thereby allowing fluorine‐rich signal generation and metal‐mediated relaxation regulation in one material system. Given these features, we envisioned that paramagnetic metal ions dissolving in FDESs (paramagnetic FDESs, paraFDESs for short) could offer tunable relaxation properties due to the paramagnetic relaxation enhancement (PRE) effect (Figure 2a). In a quick verification experiment, several metal chlorides (CrCl₃, MnCl₂, FeCl₃, CoCl₂, NiCl₂, and CuCl₂) were dissolved in DES‐A (made from choline chloride and 2,2,2‐trifluoroacetamide in a 1:2 molar ratio, with a ¹⁹F chemical shift at –76 ppm), yielding a series of transparent and colored liquids (Figure 2b). Compared to metal‐free DES‐A, paraFDES‐As showed peak broadening to varying degrees in ¹⁹F NMR spectra as expected, along with minor chemical shift changes (<2 ppm) due to the pseudocontact shift [23] (Figure 2c). The measurement results of T ₁, T ₂, r ₁, and r ₂ showed that the relaxation times of paraFDES‐As vary with metal types and are metal‐concentration dependent for paramagnetic metal ions (Figure 2d,e, see Tables S1 and S2 for details). Among them, FDES‐A containing diamagnetic ZnCl₂ showed no obvious concentration‐dependent relaxation change, paraFDES‐A containing FeCl₃ showed high r ₁ and moderate r ₂, which is favorable as a modulating agent for ¹⁹F relaxation times without significantly compromising ¹⁹F MRI signals [18]. Therefore, the relaxation times of paraFDESs can be finely modulated through the concentration of ferric ions. The slightly decreased viscosity of paraFDES‐A with increasing FeCl₃ concentration confirmed that relaxation times shortening mainly arose from the paramagnetic effect of Fe³⁺ ions rather than viscosity change (Table S3). The shift of amide C = O stretching vibration band in Fourier‐transform infrared spectroscopy (FTIR) spectra suggested the possible coordinating interactions between Fe³⁺ ions and trifluoroacetamide (Figure S1). Together, this tunable relaxation behavior provides a unique basis for encoding information into the relaxation dimension, which can subsequently be retrieved using tailored MR pulse sequences.

FIGURE 2.

Information Showcasing with Paramagnetic Fluorinated DES via MR Imaging in a Relaxation‐Selective Manner (INSPIRE). (a) Schematic representation of various interactions that are critical to the solvation of paramagnetic metal ions (mostly coordination and hydrogen bonding) in a paramagnetic fluorinated DES (paraFDES). (b) A photograph of various paraFDES‐As containing different paramagnetic metal ions. FDES‐A (made from choline chloride and 2,2,2‐trifluoroacetamide in a 1:2 molar ratio) was chosen as an example to investigate the solubility and relaxation properties of paraFDES. The concentration of metal ions was 20 mM. (c) Stacked ¹⁹F NMR spectra of paraFDES‐As with different dissolved paramagnetic metal ions. The NMR peaks showed significant line broadening compared to the blank (FDES‐A without any paramagnetic metal ions). Additionally, the ¹⁹F chemical shifts varied slightly (less than 2 ppm). (d) and (e), ¹⁹F longitudinal and transverse relaxation rates of various paraFDES‐A's. Paramagnetic metal ions could significantly accelerate ¹⁹F longitudinal (T ₁) and transverse (T ₂) relaxations in a concentration‐dependent manner. n = 3, graph bar = mean, error bars = standard deviation. (f) Schematic representation of INSPIRE. ¹⁹F MRI of these relaxation‐modulated paraFDESs with relaxation‐selective pulse sequences enables differential signal generation (T ₁ modulation) and acquisition (T ₂ modulation), allowing relaxation‐selective pattern display (vide infra). (g) The principles of T ₁‐selective imaging using an inversion recovery (IR) sequence. A 180° inversion pulse flips the longitudinal magnetization (M_z), which then relaxes from −M_z to +M_z. With a carefully chosen inversion time (TI), the excitation of ¹⁹F nuclei with a specific T ₁ is selectively inhibited due to M_z = 0 (indicated by the orange arrow), while the other ¹⁹F nuclei could still be excited by the 90° pulse for imaging (indicated by the blue arrows). (h) A proof of concept for T ₁‐selective INSPIRE. ¹⁹F MR images of paraFDES‐As containing FeCl₃ at varying concentrations, which were acquired with different TIs. See Table S1 for the T ₁ of ¹⁹F for different paraFDES species. (i) The principles of T ₂‐selective imaging using a spin‐echo sequence. A 90° pulse results in the generation of the transverse magnetization (M_xy), which undergoes T ₂ relaxation. With a carefully chosen echo time (TE), the signals from ¹⁹F nuclei with short T ₂ are selectively suppressed (indicated by the orange arrow), while those with longer T ₂ remain detectable (indicated by the blue arrows). (j) A proof of concept for T ₂‐selective INSPIRE. ¹⁹F MR images of paraFDES‐As containing FeCl₃ at varying concentrations, which were acquired with various TEs. See Table S1 for the T ₂ of ¹⁹F for different paraFDES species.

Based on these observations, we proposed INSPIRE (Information Showcasing with Paramagnetic Fluorinated DES and MR Imaging in a Relaxation‐Selective Manner), a strategy that integrates paraFDESs with tailored MRI pulse sequences for relaxation‐dependent information encoding and readout (Figure 2f). paraFDESs with different relaxation properties can be selectively visualized using relaxation‐selective settings of pulse sequences, allowing chemically identical species to yield distinct MR signals. For imaging paraFDESs with different T ₁ (T ₁‐selective imaging), an inversion recovery RARE (IR‐RARE) sequence was employed (Figure 2g). With a specific inversion time (TI), the ¹⁹F MRI signals from the paraFDES with a specific T ₁ were fully suppressed, allowing different readouts for the same ¹⁹F species. Simulated signal‐to‐noise ratios (Figure S2a) based on T ₁ measurements matched experimental ¹⁹F MRI results (Figure 2h, see Figure S2b,c for the full results). For imaging paraFDESs with different T ₂ (T ₂‐selective imaging), a conventional RARE sequence with relatively long echo time (TE) was used to suppress the ¹⁹F MRI signals from the paraFDES with short T ₂ (Figure 2i). A series of phantom samples with varying FeCl₃ concentrations showed distinct T ₂‐dependent imaging results (Figure 2j). Collectively, we established a relaxation‐centric framework in which paraFDESs could operate as programmable media for information processing in the spin‐relaxation domain, highlighting the bidirectional write‐read capability of relaxation‐based information that is fundamentally distinct from the one‐way acquisition in conventional MRI.

2.2. Information Display via Multiplexed INSPIRE

The multiplexed capacity of ¹⁹F MRI allows multi‐channel information display with appropriate molecular encoding. To integrate this feature into INSPIRE, we added different amounts of FeCl₃ (final concentrations: 0, 5, and 10 mM) to three selected fluorinated DESs, FDES‐A (vide supra), FDES‐B (from 2‐chloro‐2,2‐difluoroacetamide, ¹⁹F chemical shift at –63 ppm), and FDES‐C (from 2,2‐difluoroacetamide, ¹⁹F chemical shift at –125 ppm), to yield three series of samples with distinct relaxation properties within each series, i.e., paraFDES‐A, paraFDES‐B, and paraFDES‐C series. By adjusting inversion time (TI) and echo time (TE), five distinct relaxation‐selective patterns could be generated in each signal channel (Figure 3a), enabling up to 125 (5 × 5 × 5) possible pattern combinations from the three sets of samples for relaxation‐selective pattern display. To exemplify the feasibility of multiplexed INSPIRE, an English sentence “SHE DID NT LIKE PINEAPPLE PIE” was encoded in a 21 × 21 well plate, with indicated FDESs and paraFDESs loaded in separate regions for displaying different segments (Figure S3). Relaxation‐selective ¹⁹F MRI with different TI and TE combinations generated distinct patterns in each channel due to the difference in ¹⁹F relaxation (Figure 3b). When the patterns from different channels were merged, 18 (3 × 2 × 3) combinations could be generated, allowing differential displays of the original sentence to convey completely different messages, such as “SHE DID NT LIKE PINEAPPLE PIE”, “I DID LIKE APPLE”, or “HE DID LIKE PINEAPPLE” (Figure 3b).

FIGURE 3.

Information display via multiplexed INSPIRE. (a) A proof of concept of multiplexed INSPIRE with paraFDES‐A, paraFDES‐B, and paraFDES‐C. By tuning both TI and TE, five distinct patterns were generated using a paraFDES at a single signal channel, enabling 125 (5 × 5 × 5) possible pattern combinations across three signal channels (paraFDES‐A for −76 ppm, paraFDES‐B for −63 ppm, and paraFDES‐C for −125 ppm). (b) Relaxation‐selective display of different English sentences with multiplexed INSPIRE. With different pulse sequences, at least two distinct patterns were obtained within a single ¹⁹F MRI signal channel, yielding 18 (3 × 2 × 3) possible combinations across three channels. Merged patterns displayed different sentences (“SHE DID NT LIKE PINEAPPLE PIE”, “I DID LIKE APPLE”, and “HE DID LIKE PINEAPPLE”) with distinct meaning. See Figure S2 for the details of pattern design. (c) Relaxation‐selective display of a QR code with multiplexed INSPIRE. The correct QR code was revealed only when all three ¹⁹F MRI signals were acquired in a relaxation‐selective manner. See Figure S4 for the details of pattern design. (d) The display of a binary code with the parity bits encoded in a relaxation‐selective manner. The presence or absence of the signals from three ¹⁹F MRI channels were used to encode a piece of binary information, while the encoded parity bits could be read via independent relaxation‐selective ¹H MRI. (e) The correct information (“XMU”) was obtained only after parity check and the removal of error codes. Without parity check and the removal of error codes, the information was decoded as a meaningless letter combination “RHEMRE”. See Figure S6 for the details of the procedures for parity checking and decoding.

The diversity of signal patterns with relaxation‐selective imaging also offers opportunities for information encryption. For example, a QR code containing the link to our research group's website was encoded in a 21 × 21 well plate using three FDESs and corresponding paraFDESs (Figure S4). With a common imaging approach (relaxation‐nonselective), the merged ¹⁹F MR image contained no recognizable pattern (Figure 3c). In contrast, when a special imaging method with suitable TI and TE (relaxation‐selective) was used, the QR code was clearly displayed in the merged ¹⁹F MR image, demonstrating the successful decryption (Figure 3c). Furthermore, the use of FeCl₃, which can also change the relaxation properties of ¹H nuclei in the samples, provides an additional orthogonal “degree of freedom” to classical encoding based on ¹⁹F chemical shift [29, 30, 32, 33]. Differential displays of the samples containing FeCl₃ with relaxation‐selective ¹H MRI were achieved with optimized pulse parameters (Figure S5, see Methods for the detailed parameters). This feature allowed the incorporation of additional proof‐of‐concept check codes, such as parity bits, for information encoding and display with enhanced security (Figure 3d). The binary string decoded based on the presence (1) or absence (0) of ¹⁹F signals yielded no meaningful message. However, relaxation‐selective ¹H MRI revealed the parity bits for the removal of error codes, in this demonstration allowing for the recovery of the hidden message “XMU” (Figure 3e, see Figure S5 for the details of parity check). It should be noted that the parity‐checking rule used here is a simple error‐detection demonstration and can detect only an odd number of errors, whereas an even number of errors may remain undetected. These results demonstrate the promising potential of INSPIRE which extends the molecular encoding capabilities of ¹⁹F‐containing species beyond the conventional chemical shift domain by incorporating relaxation modulation for information encoding and readout, thereby enabling advanced functionalities such as selective information access, information obfuscation, and information encryption and decryption with inherent check codes.

2.3. Information Display via INSPIRE with Photoredox‐Enabled Dynamic Relaxation Modulation

Light is capable of inducing nuclear spin relaxation changes in certain paramagnetic molecule systems [41]. During the investigation of paraFDESs, we occasionally observed that FeCl₃‐containing paraFDESs turned from yellow to colorless upon UV irradiation (365 nm) and gradually restored the yellow color in the dark (Figure S7), accompanied by drastic changes in relaxation properties (Figure S8). A similar observation has been reported before [42], indicating that Fe³⁺ ions underwent photoreduction (UV 365 nm) to Fe²⁺ ions in the presence of citric acid, which are oxidized back to Fe³⁺ ions by atmospheric oxygen. Inspired by this phenomenon, we proposed that the paraFDES could undergo a photoredox cycle (Figure 4a), in which Fe³⁺ ions could be reduced to Fe²⁺ ions under UV light, with choline as a reductant being oxidized, followed by the oxidation of Fe²⁺ ions by oxygen. Since Fe³⁺ ions (5 unpaired electrons) possess a stronger PRE effect than Fe²⁺ ions (4 unpaired electrons), this cycle offers an approach for dynamic relaxation modulation. To assess the feasibility, we measured the T ₁ and T ₂ of paraFDES‐A with 20 mM FeCl₃ (designated as paraFDES‐A‐Fe(III)‐20) before and after UV irradiation. While Fe³⁺ initially caused pronounced shortening of both T ₁ and T ₂, UV exposure led to the obvious restoration of relaxation times (Figure 4b and Table S1). In contrast, the CF₃CONH₂ aqueous solution control showed negligible UV‐induced relaxation changes, indicating the importance of the FDES environment for the UV response (Figure S9). Relaxation rate analysis confirmed the considerable decreases in r ₁ and r ₂ after UV treatment (Figure 4c, Table S2). Fe²⁺/Fe³⁺ measurement assays established according to a classical 1,10‐phenanthroline‐based method [43] (Figure S10) revealed an almost quantitative photoreduction within 20 min (Figure 4d) and 53% complete oxidation in the dark after 5 d (Figure 4e). The replacement of choline with acetylcholine, which lacks a free hydroxyl group for oxidation, substantially suppressed Fe³⁺ reduction (Figure 4f), confirming the hydroxyl group's role as the reductant. High‐performance liquid chromatography‐mass spectrometry (HPLC‐MS) analysis of UV irradiated paraFDES‐A‐Fe(III)‐200 detected a peak for positive m/z 118.09, consistent with protonated betaine, supporting the oxidation of choline to betaine (Figure S11). Moreover, the oxidation of Fe²⁺ to Fe³⁺ was significantly suppressed under oxygen‐deficient conditions and accelerated in oxygen‐rich environments, confirming the role of oxygen as the oxidant (Figure 4g). Relaxation time measurements showed that the UV‐induced reduction in paraFDES‐A‐Fe(III)‐20 completed within 20 min, with T ₁ and T ₂ increasing from 60 and ≈1 ms to 570 and 57 ms, respectively (Figure 4h). In the dark, the oxidation over 4 d led to the decrease of T ₁ and T ₂ to 100 and ∼2 ms, respectively (Figure 4h). The process was of excellent reversibility for six cycles over 30 d (Figure 4i).

FIGURE 4.

Information display via INSPIRE with photoredox‐enabled relaxation modulation. (a) Schematic illustration of the Fe³⁺/Fe²⁺ photoredox cycle. UV (365 nm) reduces Fe³⁺ to Fe²⁺, weakening PRE and lengthening ¹⁹F relaxation, while dark oxidation restores Fe³⁺ and shortens relaxation. This reversible modulation arises from changes in unpaired electrons, with weak‐field ligands maintaining consistent spin states in paraFDESs. (b) ¹⁹F Relaxation times of paraFDES‐A containing FeCl₃ (20 mM) before and after UV irradiation. The presence of Fe³⁺ in the paraFDES‐A significantly reduced ¹⁹F relaxation times. After UV exposure, the ¹⁹F relaxation times were significantly restored. n = 3, graph bar = mean, error bars = standard deviation. (c) ¹⁹F relaxation rates of paraFDES‐A's containing FeCl₃ after UV irradiation. Compared to the ¹⁹F relaxation rates before UV exposure (Figure 2d,e), a substantial decrease was observed. See Table S1 for the details of ¹⁹F relaxation rates. n = 3, graph bar = mean, error bars = standard deviation. (d) Quantitative analysis on the relative Fe²⁺ and Fe³⁺ concentrations in paraFDES‐A during UV irradiation. UV irradiation converted nearly all Fe³⁺ ions to Fe²⁺. The initial Fe³⁺ concentration was normalized to 1. n = 3, graph bar = mean, error bars = standard deviation. (e) Quantitative analysis on the relative Fe²⁺ and Fe³⁺ concentrations in UV‐irradiated paraFDES‐A during dark storage. 53% of the Fe²⁺ ions were oxidized back to Fe³⁺ ions. n = 3, graph bar = mean, error bars = standard deviation. (f) The role of choline during the Fe³⁺/Fe²⁺ photoredox cycle. Acetylation of the hydroxyl group of choline significantly slowed down the UV‐induced reduction of Fe³⁺. n = 3, graph bar = mean, error bars = standard deviation. (g) The role of oxygen during the Fe³⁺/Fe²⁺ photoredox cycle. Oxygen depletion substantially suppressed the oxidation of Fe²⁺, whereas excess oxygen accelerated this process. n = 3, graph bars = mean, error bars = standard deviation, **p < 0.01, ****p < 0.0001. (h) The changes in ¹⁹F relaxation times during the Fe³⁺/Fe²⁺ photoredox cycle. T ₁ and T ₂ increased rapidly during the first 15 min of UV irradiation and gradually decreased over four days in the dark. n = 3, graph bars = mean, error bars = standard deviation. (i) The robustness of the Fe³⁺/Fe²⁺ photoredox cycle for relaxation modulation. The relaxation modulation remained highly robust over six cycles. (j) A demonstration on the encryption and decryption of a doubly relaxation‐encrypted colorcode. A color code encoded with paraFDES‐A, paraFDES‐B, and paraFDES‐C is concealed by paraFDES‐A‐Fe(III) “fake pixels”, which are selectively removed only under relaxation‐selective ¹⁹F MRI. UV irradiation converts Fe³⁺ to Fe²⁺ and temporarily erases this selectivity, while dark oxidation restores Fe³⁺ and permits a time‐gated successful decryption, see Figure S12 for the detailed colorcode design.

To demonstrate the dynamic relaxation modulation for INSPIRE, we designed a UV‐responsive color code with double encryption. In a 5 × 5 well plate, FDES‐A/B/C were used to encode a specific pattern, while paraFDES‐A‐Fe(III)‐5 was for “fake pixels” (Figure S12). With relaxation‐nonselective ¹⁹F MRI, the signals from the “fake pixels” led to decoding failure (Figure 4j). However, relaxation‐selective ¹⁹F MRI (TE = 5 ms, TI = 70 ms) completely suppressed the signals from the “fake pixels”, enabling successful decryption (Figure 4j). Notably, the preserved “real” pixels served as an internal control for verifying the selective removal of “fake pixels” without disrupting normal decoding. After UV irradiation, the relaxation properties of the paraFDES‐A changed, making the signals from the “real” and “fake” pixels indistinguishable even by relaxation‐selective ¹⁹F MRI and thus introducing a second encryption. The second encryption was removed after three days of dark storage, during which Fe²⁺ was re‐oxidized to Fe³⁺, allowing the signals from the “real” and “fake” pixels to be distinguished by relaxation‐selective ¹⁹F MRI again (Figure 4j). This temporal process unlocked access to the INSPIRE layer of encryption but did not directly reveal the concealed information (Figure 4j). The combination of time‐gated unlocking and INSPIRE “passwords” establishes a sequential dual‐encryption mechanism, thereby further enhancing the overall information security of the system. Taken together, we demonstrate that this photoredox cycle endows paraFDESs with environmentally driven control over spin‐relaxation states, a mechanism that enriches the INSPIRE framework by enabling dynamic and stimulus‐responsive modulation of spin relaxation in the time domain.

2.4. Programmed Dynamic Pattern Display via INSPIRE with Photoredox‐Driven Relaxation Modulation

As shown in Figure 4h, the photoredox process of FeCl₃‐containing paraFDESs, with UV‐induced reduction completed within minutes and the oxidation in the dark achieved over days, offers a feasible timeframe for dynamic information display. Given the pseudo‐first‐order kinetics of this redox process with respect to Fe species [44], we expect that tuning the initial FeCl₃ concentration could regulate the reaction rates to achieve programmed dynamic pattern changes (Figure 5a). With a high concentration of FeCl₃, paraFDESs initially are not “visible” for relaxation‐selective ¹⁹F MRI (“OFF”) due to the strong PRE effect of Fe³⁺ ions. Upon UV exposure, Fe³⁺ ions are gradually reduced to Fe²⁺ ions, which significantly attenuates the PRE effect. Fewer Fe³⁺ ions need to be consumed to weaken the PRE effect when the FeCl₃ concentration is low, resulting in earlier signal intensification (“ON”). Conversely, a high FeCl₃ concentration delays this transition. During the oxidation in the dark, however, high FeCl₃ concentration allows faster regeneration of Fe³⁺ ions, leading to earlier signal attenuation (“OFF”). We verified this hypothesis by tracking the changes in ¹⁹F relaxation times of paraFDES‐As with different FeCl₃ concentrations. Under UV light, both T ₁ and T ₂ of ¹⁹F increased with time, with delayed plateau appearance at higher concentrations (Figure 5b,c). In the dark, both ¹⁹F relaxation times decreased, with faster decay observed for higher FeCl₃ concentration (Figure 5d,e). We next illustrated dynamic patterning via relaxation‐selective ¹⁹F MRI. With T ₂‐selective imaging (TE at 10.9 ms) employed, the signals of UV‐irradiated paraFDES‐A samples containing FeCl₃ at 20, 40, 60, and 80 mM showed a stepwise turning “ON” in the order of increasing FeCl₃ concentrations (Figure 5f). In addition, a 5 × 5 well plate encoded with a designed gradient of FeCl₃ concentration revealed a programmed “Turn ON” dynamic pattern (Figure 5g and Movie S1, see Figure S13 for the detailed well‐plate design). During the oxidation, a faster transition of turning “OFF” was observed for the samples with higher FeCl₃ concentrations (Figure 5h). During dark storage after UV irradiation, the well plate of Figure 5g showed a dynamic pattern of “Turn OFF” (Figure 5i and Movie S2), which is the reverse process of that in Figure 5g. Notably, the process of “Turn OFF” in the dark lasted several days, making it well‐suited for time‐sensitive tagging. Given the high penetration capacity of MRI signals, a temporal encoding strategy exploiting this process is ideal for hidden anti‐counterfeiting tagging with time‐sensitive readouts. To implement this strategy with multiplexed ¹⁹F MRI, we first explored the iron concentration‐dependent “Turn OFF” process of paraFDES‐Bs and paraFDES‐Cs with ¹⁹F MRI signals at −63 ppm and −125 ppm, respectively (Figures S14 and S15). Then a proof‐of‐concept chip comprising three 5 × 5 layers with the patterns “M” (−76 ppm), “R” (−63 ppm), and “I” (−125 ppm), respectively, was constructed using paraFDES‐As, paraFDES‐Bs, and paraFDES‐Cs containing FeCl₃ at varying concentrations (Figure S16). The ¹⁹F MRI signals of all layers were initially “OFF” with T ₂‐selective imaging. After UV irradiation, the signals of all the wells were turned “ON” and the hidden patterns remained not visible. In the dark, the variation in oxidation rate resulted in the difference in relaxation properties for different wells, revealing the originally hidden patterns, which gradually faded as the majority of Fe³⁺ ions were regenerated (Figure 5j and Movie S3). Together, these findings reveal that the photoredox‐governed evolution of spin‐relaxation states in paraFDESs creates a mechanistic basis for encoding information into the temporal dimension, thereby extending INSPIRE from a relaxation‐selective framework into one that accommodates time‐dependent and dynamically addressable spin‐state encoding.

FIGURE 5.

Programmed dynamic pattern display via INSPIRE with photoredox‐driven relaxation modulation. (a) Schematic illustration of dynamic pattern display. The PRE effect in iron‐based paraFDESs is primarily governed by Fe³⁺, which determines the “ON/OFF” property. A T ₂‐selective INSPIRE approach (TE = 10.9 ms) was used to monitor these dynamics. Hence, the pattern dynamics are controlled by the initial FeCl₃ concentration of paraFDES, where low concentrations lead to faster “Turning ON” and slower “Turning OFF” transitions, and vice versa. (b,c), The changes in T ₁ and T ₂ relaxation times of ¹⁹F during UV irradiation. Both T ₁ (b) and T ₂ (c) increased slowly at first, followed by a rapid rise and eventual plateau, with higher FeCl₃ concentrations causing delayed plateaus and lower plateau values. n = 3, graph bars = mean, error bars = standard deviation. (d,e), The changes in ¹⁹F relaxation times during dark storage. Both T ₁ (d) and T ₂ (e) decreased rapidly after storage in the dark, and higher FeCl₃ concentrations resulted in faster declines. n = 3, graph bar = mean, error bars = standard deviation. (f) The “Turning ON” behaviors of paraFDES‐A. All the samples were UV‐irradiated and monitored by T ₂‐selective ¹⁹F MRI. The samples with lower FeCl₃ concentrations were turned ON earlier. (g) A programmed “Turn‐ON” dynamic pattern display. A 5 × 5 well plate containing paraFDES‐A with varying FeCl₃ concentrations was UV‐irradiated. The signals from the wells with lower FeCl₃ concentrations were turned “ON” earlier, generating a predefined dynamic pattern display. See Movie S1 and Figure S13 for details. (h) The “Turning OFF” behaviors of paraFDES‐A. All the samples were subjected to dark storage after UV irradiation and monitored by T ₂‐selective ¹⁹F MRI. The samples with higher FeCl₃ concentrations were turned “OFF” earlier. (i) A programmed “Turn‐OFF” dynamic pattern display. The well plate from (g) was subjected to dark storage after UV irradiation (30 min). The signals from the wells with higher FeCl₃ concentrations were turned “OFF” earlier, generating a dynamic pattern display that is the reversal of the one in (g). See Movie S2 for the details. (j) Dynamic pattern display via multiplexed INSPIRE with photoredox relaxation modulation. Three paraFDES based 5 × 5 arrays encoded the patterns M, R, and I which were invisible before or immediately after UV irradiation but emerged during dark storage and persisted for several days. See Movie S3 and Figure S19 for further details.

2.5. Editable Pattern Display with Paramagnetic Fluorinated Ionogels (paraFGels)‐incorporated INSPIRE

Fluorinated ionogels (FGels) have been reported as a well‐suited material for ¹⁹F MRI‐based molecular encoding²⁷. However, direct incorporation of Fe³⁺ ions into FGels is challenging, as the radical polymerization is significantly hindered by the radical scavenging capacity of Fe³⁺ ions. Two distinct post‐gelation modification strategies were introduced to overcome this obstacle. The first strategy is direct painting on FGels with FeCl₃ aqueous solution (Figure 6a). After painting, the diffusion of Fe³⁺ ions into FGels is restricted, resulting in the local relaxation modulation and the generation of painted patterns with suppressed ¹⁹F MRI signals via T ₂‐selective imaging. These patterns could be further edited with UV irradiation globally or regionally to achieve pattern erasing or editing, respectively. With this strategy, water was unavoidably introduced into paraFGels, which might compromise the quality and stability of ionogels. Therefore, a second strategy based on thermal diffusion was introduced, which allowed the formation of homogeneous paraFGels readily for UV editing (Figure 6b). Scanning electron microscopy (SEM) and energy‐dispersive X‐ray spectroscopy (EDS) elemental mapping were performed to characterize the morphology and elemental distribution of FGel‐A and paraFGel‐A (Figure S17). Compared with FGel‐A, paraFGel‐A showed a similar gel morphology, and the Fe signal was distributed uniformly throughout the paraFGel matrix together with the F signal. The ¹⁹F MRI signals of as‐prepared paraFGels were suppressed with T ₂‐selective imaging due to the strong PRE effect of Fe³⁺ ions, which could be selectively restored using UV‐induced reduction of Fe³⁺ to Fe²⁺ to implement relaxation modulation, therefore enabling spatially‐defined sequential pattern printing with specified intensity via temporal control over UV exposure (Figure 6b). Printed patterns in paraFGels gradually faded over 5 d (Figure 6c), likely due to Fe³⁺ regeneration and internal diffusion. Moreover, the UV‐printable patterns remained clear even after ten printing‐fading cycles across 50 days (Figure 6d). Scanning electron microscopy (SEM) images showed that paraFGel‐A displayed slight surface wrinkling after ten photo‐redox cycles (Figure S17). Signal analysis during the ten cycles showed that the ¹⁹F MRI intensity of the UV‐printed regions was largely retained (Figure S18b), indicating that the fluorinated component remained stable in the ionogel matrix. Although the signal‐to‐background ratio between the printed and nonprinted regions gradually decreased and then became relatively stable (Figure S18c), clear patterns could still be obtained after repeated printing and fading, confirming the sufficiently stable relaxation‐regulated imaging performance of paraFGels.

FIGURE 6.

Editable pattern display via ionogel‐incorporated INSPIRE. (a) Writable pattern display based on INSPIRE with FDES‐based ionogels (FGels). A FGel‐A was prepared with FDES‐A to generate a “bright” panel (blank) with relaxation‐selective ¹⁹F MRI. FeCl₃ aqueous solution (“ink”) could be used to generate “dark” patterns on the “bright” panel, which could be erased with UV irradiation (“eraser”). The erasure could be region‐selective for pattern editing. (b) UV‐printable pattern display based on paraFDES ionogels (paraFGels). The FGel was soaked with a paraFDES (paraFDES‐A in this case) to generate a UV‐printable “dark” panel with relaxation‐selective ¹⁹F MRI. With a UV‐proof mask, UV irradiation could selectively turn a patterned region “bright,” like printing a pattern. This UV printing process was repeatable, and the brightness was tunable by adjusting the UV‐irradiation time. (c) The erasure of the UV‐printed pattern. The UV‐printed pattern gradually faded away over five days, allowing new patterns to be printed. (d) The robustness of UV‐printable pattern display. The printed pattern remained clear even after 10 printing‐erasing cycles for over 50 days. (e) Sophisticated pattern display via multiplexed INSPIRE with paraFGels. paraFGels derived from paraFDES‐A, paraFDES‐B, and paraFDES‐C were subjected to individual UV‐printing and stacked to form a multiplexed paraFGels for sophisticated pattern display by merging the ¹⁹F MR images collected from different channels. (f) An example of sophisticated pattern display via multiplexed INSPIRE with paraFGels. Three emblematic elements were respectively UV‐printed and visualized with multiplexed ¹⁹F MRI to display the centenary logo of Xiamen University. (g) Conceptual schematic illustration of “multichannel flow line checking” with multiplexed ¹⁹F MRI. The high fluorine content and UV‐printable nature of paraFGels made them suitable as writable tags for rapid ¹⁹F MRI readout. These tags could be hidden inside cargo, enabling non‐invasive and “flow” tracking. (h) Schematic illustration showing the timing for cargo movement and image acquisitions. To minimize motion artifacts, cargo movement is restricted to the intervals between image acquisitions. (i) Representative images of “flow line” cargo checking with multiplexed INSPIRE. The information stored in the paraFGel inside each cargo could be successfully retrieved from the 3 signal channels in a sequential and fast manner. A single image was acquired within 2.25 s (6.75 s in total for 3 images from 3 channels), enabling rapid “flow line” cargo checking. See Movie S4 for the full animation.

We further increased the diversity and multiplicity of this pattern display by expanding the system to three MRI channels using paraFGel‐A, B, and C, which were prepared from FDES‐A, B, and C with ¹⁹F chemical shifts at −76, −63, and −125 ppm, respectively (Figure 6e). The capacity of these paraFGels for UV printing was quickly verified (Figure S19). Three distinct patterns, “XMU”, a centennial logo, and a gate symbol, which were UV‐printed in a paraFGel‐A, a paraFGel‐B, and a paraFGel‐C, respectively, could be clearly resolved with multiplexed ¹⁹F MRI (Figure 6f), demonstrating the potential of multiplexed INSPIRE with paraFGels for sophisticated pattern display. This capability shows a conceptual resemblance to a recently reported tricolor optical photoprinting system [45], while establishing a non‐luminescent, radiofrequency‐based approach for multiplexed pattern visualization. It is worth mentioning that the high fluorine content of FGels allows an acquisition time as short as 2.25 s for a single pattern with optimized MRI parameters (see Methods for details). We therefore investigated the implementation of “flowline check” for paraFGel‐based tags, where patterned paraFGels hidden inside the cargo on a plastic conveyor were sequentially read by an MRI scanner (Figure 6g). A program for this conceptual “flow line” was carefully designed, in which the motion was paused during signal acquisition to minimize potential artifacts (Figure 6h). A proof‐of‐concept of “flowline check” was successfully implemented with all the patterns were rapidly, continuously, and clearly visualized (Figure 6i and Movie S4). To further assess the applicability of MRI readout in practical cargo‐inspection scenarios, UV‐printed paraFGels were covered with different materials (Figure S20). The ¹⁹F MRI patterns remained clearly readable through plastic, carton, and lead glass, whereas aluminum foil partially attenuated or completely blocked the signal depending on its placement and enclosure geometry. Notably, ¹⁹F MRI‐based detection could successfully visualize paraFGel tags through certain heavy‐element nonmetallic materials, such as lead glass, which are typically opaque to X‐ray inspection. The integration of INSPIRE with paraFGels thus marks a transition from a conceptual proof‐of‐concept to a practical material platform that enables convenient, rewritable, and dynamic patterning, setting a promising foundation for future developments in real‐world applications such as radiofrequency‐readable tagging and anticounterfeiting.

3. Conclusions

In summary, we developed paramagnetic fluorinated materials with tunable ¹⁹F relaxation properties, including paraFDESs and paraFGels, by incorporating paramagnetic metal ions into FDESs. This design enables precise encoding of radiofrequency information in the domain of nuclear spin‐relaxation as well as in the traditional domain of ¹⁹F chemical shift. When combined with relaxation‐selective ¹⁹F MRI, the information encoded in the relaxation domain can be retrieved in a selective manner. This bidirectional process with tunable (and thus designable) features on both directions provides a solid basis for the INSPIRE strategy. The system further supports stimuli‐responsive editing and dynamic readout, which significantly enhances the flexibility and adaptability of INSPIRE. Leveraging the inherent advantages of MRI, such as deep penetration and multiplexed imaging, this approach offers an additional spin‐relaxation‐based domain for information encoding and decoding, thereby improving both security and density that are critical to many applications of information technology, such as anticounterfeiting. Its promising potential has been clearly demonstrated by a series of experiments showcasing its various uses for information encoding‐decoding, encryption‐decryption, and controlled display. Compared with fluorescence‐based molecular encoding and DNA‐based information storage, this strategy is particularly suitable for high‐security and high‐confidentiality information readout that requires noncontact and penetrative detection, such as concealed information tags embedded in objects or biological matrices, as well as dynamically rewritable information labels.

From a perspective in materials and applications. Further developments on media materials would significantly enhance the capability of MRI‐based systems. For example, expanding the structural diversity of FDESs, such as introducing fluorinated modifications to the choline component could readily provide access to up to 22 additional ¹⁹F signal channels that could be combined with relaxation modulation to generate a large space for information encoding. The tunability of DES properties, including viscosity and redox potential, also permits additional functional customization for various practical applications. From a more fundamental perspective, this work represents an initial step toward exploiting nuclear spin properties beyond conventional spin resonance frequency for information processing by introducing spin relaxation as an additional encoding physical dimension. Notably, relaxation‐based contrast is less dependent on spectral resolution, making this strategy potentially beneficial for low‐field NMR/MRI or open‐magnet systems. Together with the high ¹⁹F signal intensity of paraFDESs and paraFGels, which helps compensate for the sensitivity limitation of low‐field instruments, this feature may provide a new perspective for translating ¹⁹F NMR encoding toward more practical and accessible instrument systems. Future exploration of additional spin‐dependent physical properties, including scalar coupling and polarization, may unlock further opportunities for NMR information encoding. More importantly, the synergy between material‐level design and instrument‐level spin manipulation may serve as an important research paradigm for expanding the information‐processing capability of NMR‐based systems.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: adma73987‐sup‐0001‐SuppMat.docx.

Supporting File 2: adma73987‐sup‐0002‐MovieS1.mp4.

Supporting File 3: adma73987‐sup‐0003‐MovieS2.mp4.

Supporting File 4: adma73987‐sup‐0004‐MovieS3.mp4.

Supporting File 5: adma73987‐sup‐0005‐MovieS4.mp4.

Data Availability Statement

The data that supports the findings of this study are available in the supplementary material of this article.