Skip to main content
Nature Communications logoLink to Nature Communications
. 2024 Jul 31;15:6470. doi: 10.1038/s41467-024-50876-y

Chemically and geometrically programmable photoreactive polymers for transformational humidity-sensitive full-color devices

Jongsun Yoon 1,#, Chunghwan Jung 1,#, Jaekyung Kim 2, Junsuk Rho 1,2,3,4,, Hyomin Lee 1,
PMCID: PMC11292010  PMID: 39085253

Abstract

Humidity-sensitive structural color has emerged as a promising technology due to its numerous advantages that include fast response, intuitiveness, stand-alone capability, non-toxicity, as well as resistance to thermal and chemical stresses. Despite immense technological advancements, these structural colors lack the ability to present independent multiple images through transformation. Herein, we present an approach to address this constraint by introducing a chemically and geometrically programmable photoreactive polymer which allows preparation of transformational humidity-sensitive full-color devices. Utilizing azido-grafted carboxymethyl cellulose (CMC-N3) allows adjustments in swelling properties based on the grafting ratio (Γ) of azido groups upon UV-induced crosslinking. Also, the distinctive photo-curability of the polymer enables precise geometric control to achieve vivid colors in combination with disordered plasmonic cavities. Our work culminates in the development of an advanced anti-counterfeiting multiplexer capable of displaying different full-color images with variation in humidity levels. The showcased color displays signify pivotal breakthroughs in tunable optical technologies, illustrating how chemical modifications in hydrogels provides additional degrees of freedom in the design of advanced optical devices.

Subject terms: Polymers, Metamaterials, Polymers, Nanophotonics and plasmonics


Humidity sensitive structural colour has potential in visualisation technologies, but display of multiple images with the same material is challenging. Here, the authors report a photoreactive polymer to prepare transformational devices that respond to humidity changes.

Introduction

Structural coloration is a natural colorimetric phenomenon that results from optical scattering, the interplay between incident light and structured materials18. Unlike dyes and pigments, structured colors have several advantages including high thermal and chemical stability, non-toxicity, ultra-high resolution, and resistance to fading unless the scatterer’s structure is severely disrupted911. Sophisticated light-matter interactions using dielectrics and plasmonics have enabled successful demonstrations of vivid colors based on a variety of structures: multi-layer interferometers5,12,13, high-refractive index disks1416, high aspect ratio nanopillars1720, and surface-relief metal slits21. However, there exists a significant drawback by its inability to change the color as well as the overall image once the structures had been formed, preventing its extension to applications beyond color filters.

Tunable structural colors have been recently proposed that can change their spectrum like a chameleon, modulating their optical properties in response to external triggers2228. Tunable structural colors present opportunities in anticounterfeiting applications, where the ability to dynamically switch colors in response to specific triggers provides an additional security feature that is difficult to replicate. This dynamic multi-color switching capability enhances encoding complexity beyond what is possible with static structural colors alone. In particular, these tunable structural colors have distinct operating mechanisms that allow their physical characteristics to be regulated in response to a variety of stimuli that include heat29, electricity3033, stretching34,35, and pH changes36. In this end, humidity-responsive colorimetric devices that respond to water vapor have attracted significant attention as they allow simple, fast, and intuitive visual discrimination without the aid of complex instruments while offering broad applicability to other uses, including self-powered gas sensors3742. In these sensors, optical resonance is controlled by the use of soft materials that change their refractive index and physical volume in response to specific gas concentrations. Utilization of various techniques such as nanoimprinting41 and electron-beam lithography42,43 in the fabrication of optical resonators such as Fabry-Pérot (F-P) resonators3739, disordered plasmonic scatterers40,42, and nano-waveguides41,44 has enabled the development of devices with special functionalities beyond simple color modulation that includes ultrafast response, reversibility control, and optical holograms to name a few40,41,45. Research reported so far have reconstructed different intensity and phase maps either by simply changing the colors of the image42, or intentionally mismatching the resonance modes of the unit cells45. Nevertheless, these systems are challenging to be truly reconfigurable, as the identical swelling tendency of the hydrogels has constrained the design freedom as well as the encoding capacity.

Here, we introduce an approach that allows transformational humidity-sensitive color printing, where the chemistry and geometry of the hydrogel structures can be simultaneously programmed in a single platform. By grafting azido (N3) groups to carboxymethyl cellulose (CMC) backbone, we can immobilize hydrogels with predetermined thicknesses and swelling properties through subsequent UV crosslinking; the hydrogel becomes less hydrophilic and its swelling tendency becomes significantly restricted with increase in the grafting ratio (Γ)37. The hydrogels can be also geometrically programmed in selective patterning sequences, in contrast to other conventional moisture-sensitive optical devices. By embedding these hydrogel structures in disordered plasmonic cavities, we demonstrate humidity-tunable full-color devices that exhibit diverse optical responses under different Γ in both simulations and experiments. We also showcase a large-area multi-layered structure with sophisticatedly designed CMC-N3-Γ to produce classic anti-counterfeiting devices and vivid color-printed pictures. Finally, we present a transformational anti-counterfeiting multiplexer that can encode/decode different full-color images depending on the triggering humidity levels, demonstrating that this device can be chemically and geometrically programmed with high precision and tunability.

Results

Working principle and fabrication process of CMC-N3-Γ

Recent studies have exploited the inherent ability of hydrogels to alter their volumetric structures in multi-layered cavities for preparation of tunable color devices; hydrogels with three-dimensional polymeric network become more flexible and absorb more water with decrease in crosslinking density46. We strategically adopt this concept to design a hydrogel interferometer which allows to precisely control the spectral shift tendency of the optical cavity through modulation in crosslinking density (Fig. 1a). We selected CMC as the hydrogel base material due to its high hygroscopicity. Introducing phenyl azido groups to CMC enables photochemical reactions in response to UV light, inducing crosslinking within these polymers and transforming the cellulose-azido derivatives (CMC-N3) into water-insoluble hydrogel form with different swelling behavior depending on the phenyl azido group grafting ratio (Γ). To determine the photo-patterning capability of CMC-N3 films as well as the swelling behavior with variation in Γ, we organized them into alternating square patterns using CMC-N3 with variation in Γ (Fig. 1b). We first coated the silicon wafer substrate (reflector) with an adhesive layer comprising of chitosan (CHI) to induce electrostatic attraction between the amine groups of CHI and carboxyl groups of CMC-N3 and thereby immobilize CMC-N3 onto the substrate. Here, CHI layer is deposited as thin as possible (12.5 ± 0.4 nm) to minimize its effect on the overall swelling behavior of the hydrogel (Supplementary Fig. 1). Next, CMC-N3 with predetermined grafting ratio (e.g. high Γ) is spin-coated and subsequently UV-cured using a photomask. As CMC-N3 is water-soluble, unexposed region can be easily removed through repeated rinsing with water. Moreover, sequential photolithography using CMC-N3 with different grafting ratio (e.g. low Γ) on the unexposed, uncoated region is possible via additional spin-coating followed by UV-curing through a separate complementary photomask. Also, the thickness of the resulting photopatterned hydrogel can be precisely controlled ranging from 10 to 300 nm by adjusting the initial concentration and spinning speed (Supplementary Fig. 2). Two sets of 480 µm square pixel arrays were prepared to visualize the swelling properties through color changes in response to humidity levels (Fig. 1c). An array of CMC-N3 having identically high Γ, the reference sample, exhibited a coherent color shift from sky blue to purple with increase in humidity (Fig. 1d). On the other hand, the array comprising of pixels with different Γ but prepared in similar thicknesses (142 ± 1 nm) showed initially same color (sky blue) but shifted to checker board pattern comprising both purple (high Γ) and dark blue (low Γ) (Fig. 1e).

Fig. 1. Functionality and manufacturing procedure of CMC-N3-Γ.

Fig. 1

a Schematics of the hydrogel interferometer under dry and humid environments. b Chemical structure of carboxymethyl cellulose-azido derivatives (CMC-N3) with different grafting ratios (Γ) and schematics depicting the process of photopatterning these derivatives on a reflector. c Simulated square pixels (480 µm × 480 µm) at dry state. Color change responses of crosslinked CMC-N3 in square pixels with identical Γ (d), and different Γ (e) under dry and humid environments. Scale bar is 100 µm for d and e.

Chemical characterization of CMC-N3

The photoreactive polymer, CMC-N3, was synthesized by coupling azidoaniline and carboxyl groups of CMC using carbodiimide chemistry and Γ was systematically varied by adjusting the proportion of reactants. CMC-N3 with four different Γ of, 4.7, 7.5, 17.1, and 23.7%, were synthesized and confirmed using 1H-NMR spectra (Fig. 2a) where the relative peak intensities of the phenyl protons at 7.0–7.5 ppm were compared to the sugar ring protons of the CMC at 2.8–4.5 ppm to acquire these values37. For simplicity, we will denote each Γ obtained from this estimation in the notion, CMC-N3-Γ, hereafter. Under UV exposure, the phenyl azido groups readily decompose into highly reactive phenyl nitrene intermediates which then engage with the nearby C-H bonds (Fig. 2b), resulting in formation of either intra- or intermolecular covalent bonds47. Fourier transform infrared (FT-IR) spectra confirms the presence of individual components within the CMC-N3 before UV illumination where we observe a rise in the peak intensity with increase in Γ at 2117 cm−1, corresponding to the stretching vibrations of azido groups (Fig. 2c). This is further supported by the peak rise at 1655 and 1550 cm−1, which represents the C=O and N-H bonds, respectively, that make up the amide bond after the photochemical reaction. UV-vis spectra were also acquired on various CMC-N3 solutions, each set at a concentration of 25 µg mL−1 in deionized (DI) water, to confirm the decomposition of phenyl azido groups caused by this photochemical reaction (Fig. 2d). Sharp decrease in absorbance at 270 nm with 10 min of UV exposure indicates the photochemical reaction of phenyl azido groups. With further UV irradiation, phenyl azido groups gradually decomposed, leading to increase in absorption at two isosbestic points, approximately at 238 and 354 nm, which is also confirmed by the kinetics of the photochemical reaction (Supplementary Fig. 3). As the extent of crosslinking depends on its chemically programmed Γ, it also affects the swelling capacity of the resulting crosslinked CMC-N3 (Fig. 2e). To validate this, we monitor the swelling behavior of sets of crosslinked CMC-N3 with similar thickness of 153 ± 3 nm but with different Γ using atomic force microscopy (AFM) in a controlled humidity environment (Fig. 2f). We find that there is a substantial swelling at relative humidity (RH) of 80% followed by a rapid increase in thickness above this threshold. Also, the maximum swelling capacity of water-saturated films confirm that CMC-N3 films with lower Γ correspond to higher swelling capacities (Fig. 2g) and that the volumetric expansion of the immobilized film, or the thickness ratio, can be precisely controlled in the range of 1.8–4.2 with variation in Γ value. We note that this is less than the maximum swelling thickness ratio of ~6 achievable when CMC-N3 film with the lowest Γ (CMC-N3-4.7) is completely submerged underwater where dynamic equilibrium with the surrounding environment is biased toward absorption by the excess amount of condensed water (Supplementary Fig. 4). To verify the origin of this pronounced difference in thickness ratio with respect to Γ, we additionally employed thermal analysis using differential scanning calorimetry (DSC). We observe enlargement of the endothermic peaks during the heating process (−60 to 30 °C) with increase in the amount of added water, indicating that the water molecules exist in weakly bonded48, freezable water state above a certain threshold for each CMC-N3-Γ (Supplementary Fig. 5). Detailed comparison among these sets of CMC-N3-Γ reveal that higher Γ corresponds to higher proportion of freezable water, as evidenced by the enlargement of the endothermic peaks for all conditions, presumably due to decrease in the polymer chain’s affinity toward water by the fewer available sites for hydrogen bonding. Overall, this confirms that the increase in Γ not only leads to more heavily crosslinked hydrogel film but also the CMC-N3-Γ comprising the hydrogel becomes less hydrophilic by fewer available sites for the water molecules to bind to, leading to substantial difference in thickness ratio with variation in Γ under humid environment. We also note that resulting CMC-N3 hydrogel films are stable in various humidity conditions, as evidenced by the marginal alteration in the saturated thickness ratio and reflectance under both standard conditions (25 °C and RH 50%) and high humidity conditions (25 °C and RH 95%) for up to 15 days (Supplementary Figs. 68).

Fig. 2. Chemical characteristics of CMC-N3-Γ.

Fig. 2

a 1H NMR spectra of CMC and azido derivatives with different grafting ratios (CMC-N3). The inset shows the chemical structure of photoreactive CMC-N3. b Scheme illustrating the photochemical reaction of phenyl azido groups in CMC. c FT-IR spectra of CMC and azido derivatives with different Γ. d UV-vis spectra obtained from CMC-N3 aqueous solution (25 µg mL−1) before and after UV exposure at various time points. e Scheme showing the inverse relationship between grafting ratio (CMC-N3) and the swelling capacity of the resulting crosslinked film. Plot showing the swelling tendency with respect to relative humidity (RH) (f) and saturated thickness ratio at RH 100% (g) of the crosslinked CMC-N3 film measured using atomic force microscopy (AFM). Error bars are displayed as mean ± σ (n  =  3 independent experiments).

Optical characterizations of the disordered plasmonic cavity

To amplify the optical phenomena during hydrogel swelling, a subwavelength plasmonic cavity was introduced into crosslinked CMC-N3-Γ film using a self-assembled Ag monolayer (Fig. 3a). This structure is fabricated through the deposition of an ultrathin Ag film on the patterned hydrogel film comprising CMC-N3-Γ, where disordered nanoislands can be formed due to the self-aggregation of the metal via optimization of the deposition rate and thickness. This collection of random nanoparticles not only offers cost-effectiveness and scalability due to its simple production process but also brings fast responsivity and high repeatability with no hindrance to water vapor diffusion (Supplementary Figs. 9 and 10). Measuring the response and recovery times, defined as the duration required to reach 90% intensity at equilibrium (T90) from 10% intensity at the initial state (T10) or vice versa, of CMC-N3-Γ films with disordered plasmonic cavities for different Γ confirms that they are fast (in the order of hundred milliseconds) and tend to increase with decreasing Γ; this is attributed to an increase in the available sites for the water molecules to bind to, which consequently increases the response time. In addition, the plasmonic top layer produces a wide range of bright colors without complex optical design due to broadband ohmic loss from free electron resonance. The reflectance spectrum relies on the island morphology and the optical path length within the hydrogel; the former depends on the Ag deposition conditions, while the latter is influenced by Γ and the humidity level.

Fig. 3. Optical behaviors of the disordered plasmonic cavity.

Fig. 3

a Schematics of disordered plasmonic island cavity with CMC-N3 hydrogel. Scanning electron microscopy (SEM) image (b) and AFM image (c) with grain size distributions of top layered Ag islands. d Top- and side-view of simulated electric field magnitude profiles of top layered Ag islands. e Simulated and experimental reflectance spectra obtained for crosslinked CMC-N3 (Γ = 4.7% and 23.7%) film at varying RH, along with corresponding optical images. The reflectance spectra of Γ = 23.7% saturate at RH 92% as opposed to Γ = 4.7%, which shifts constantly with increasing RH. f Refractive index of CMC-N3- Γ at RH 40%. g Reflection spectra map of the plasmonic cavity with varying thicknesses of CMC-N3-17.1 (white dashed lines: the corresponding modes of cavity resonances). h CIE 1931 color space derived from the reflectance spectra of the plasmonic cavity with varying thicknesses of CMC-N3-17.1. Scale bars represent 100 nm for bd.

Surface structural analysis using scanning electron microscope (SEM) and AFM was performed to clarify the structure of the Ag islands and to model rigorous electromagnetic simulations (Fig. 3b, c). Detailed fabrication conditions were empirically optimized to produce the widest color gamut, since the stochastic properties of both the morphology and spatial configuration of the deposited islands directly determine the optical interactions. Physical analysis of the island grains revealed that the hemi-ellipsoids were randomly scattered with partial overlap, following a Gaussian distribution with a top-view radius of 13.3 ± 5.1 nm. Employing the parameters of this plasmonic structure, we performed numerical analysis of optical properties using the finite-difference time-domain (FDTD) method. The X-Y and X-Z electric field profiles at the optical resonance not only clearly demonstrate the interactions among self-assembled plasmonic hemi-ellipsoids, but also explicitly indicate significant electromagnetic field enhancement between the scatterers, leading to robust absorption within the reflectance spectra (Fig. 3d).

The optical characteristics of a plasmonic self-assembled cavity using CMC-N3-Γ are clearly observed in simulation and experimentation. To elucidate the optical responses of hydrogel films prepared using different CMC-N3 derivatives, we simultaneously compared reflectance spectra and images for the devices where the Γ is 4.7 and 23.7%, both with h = 149 ± 4 nm at RH 40% (Fig. 3e). The simulated reflection spectra were derived by retrieving the effective index of the plasmonic nanoislands film with S parameters (Supplementary Fig. 11). The measured reflection spectra were obtained using a specialized optical setup (Supplementary Fig. 12), ensuring precise humidity control for reliable measurements. The strong agreement between simulated and experimented reflectance underlines the accuracy of optical modeling and the hydrogel’s swelling tendencies. At initial humidity (RH 40%), the small differences in refractive index (Fig. 3f) result in a slight spectral shift, but the overall spectrum and color are nearly identical. However, as external humidity gradually increases, the lower-Γ device expands rapidly and shows significant optical deviations due to unparalleled moisture absorption. It is worth noting that hygroscopic saturation occurs at RH 92% for CMC-N3-23.7, which is less hydrophilic and have high crosslinking density after UV-irradiation. Due to this early saturation behavior and the swelling becoming more pronounced at even higher humidity, CMC-N3-4.7 with lowest Γ leads to 2.5-fold maximum difference in physical height at RH 100%. Furthermore, considering the change in effective refractive index due to massive water absorption, the effective optical path differs by over 3-fold. This strong reliance on swelling tendencies and maximum swelling ratio on Γ offers exceptional design flexibility, allowing us to program desired colors based on humidity by controlling Γ and initial thickness. Moreover, rapid spectral changes in cavity resonance modes from the disordered plasmonic cavities (Fig. 3g), hint at their potential use for full-color displays; the presented device can cover approximately 80% of the sRGB gamut, depending on its thickness (Fig. 3h and Supplementary Fig. 13). The reflectance spectrum for oblique incident light is also calculated to verify the spectral variation of the device over the field of view (Supplementary Fig. 14). Moreover, the reflectance was monitored for extended periods of up to 15 days under both standard (25 °C and RH 50%) and high humidity (25 °C and RH 95%) conditions to confirm the long-term stability of the plasmonic cavity (Supplementary Figs. 15 and 16).

Structural color-printing for full-color images

High-resolution, large-area, and full-colored structural color-printing is achieved by integrating multi-stacked CMC-N3 patterns into the disordered plasmonic cavity. Figure 4a illustrates the process of fabricating stacked hydrogel films through repeated spin-coating and UV exposure cycles. This process culminates in forming a disordered plasmonic structure followed by Ag nanoparticle deposition. This series of steps allows to geometrically program the device by precise positioning of CMC-N3-Γ films at desired height and locations, enabling colorful visualization of arbitrary images.

Fig. 4. Full color images using structural color printing method.

Fig. 4

a Schematics illustrating the fabrication process of multi-layered disordered Ag islands-based CMC-N3-Γ device. b Microscopic snapshots of 7 layered CMC-N3-Γ of arabesque circles after each stacking process. Optical image (c) of the multi-layered hydrogel structure after Ag deposition and its magnified images of the center (d) and corner (e) positioned square, triangle, circle, and star marks. f The cross-sectional thickness profile of cornered square. g Red, green, and blue (RGB) pixels using plasmonic CMC-N3-Γ cavity. Structural color printing of The Starry Night and Sunflowers by Vincent van Gogh (h), and 5 cm × 5 cm landscape picture (i) on a 4” Si wafer. Scale bar is 1 cm for b, c and i, 5 mm for h, 500 µm for d, 250 µm for g, and 100 µm for e and inset of i.

We prepared arabesque circles sequentially stacked up to seven layers on a wafer scale to test the wide color control, large-area productivity, and high-resolution capabilities of this process (Fig. 4b). Using 1.0 wt% CMC-N3 solution and spin-coating at 3000 rpm for each iteration, geometric patterns with distinct colors were fabricated (Fig. 4b). After depositing Ag islands and thereby making disordered plasmonic cavity, Fig. 4c shows much clearer and brighter color compared to the bare hydrogel structure without Ag islands under identical illumination. Enlarged optical microscope pictures from the center and corners displayed precise geometric patterns (squares, triangles, circles, and stars) with high resolution (Figs. 4d, e). The high reliability of the CMC-N3 device was also confirmed using a profilometer, showing uniform stacking with a thickness profile of 57 ± 3 nm for each stack (Fig. 4f).

To verify the color reproduction capabilities of the presented color-printing, we made various pictures and an array of red, green, and blue (RGB) pixels. To achieve clear RGB coloration for 50 µm × 250 µm pixels, we deposited CMC-N3 films each with thicknesses of 285 ± 2, 370 ± 1, and 141 ± 1 nm, respectively (Fig. 4g). Our process enables facile production of large-scale and diverse artistic patterns through geometrical patterning of CMC-N3 film on a wafer, showcasing its wide adaptability to diverse graphics (Fig. 4h, i). We note that relatively large linewidth (~50 μm) was presented to showcase the versatility and capability of the synthesized photoreactive polymers in producing functional devices, even with commercially affordable lithographic methods. However, they are also compatible with state-of-the-art nanofabrication techniques such as UV-nanoimprinting that yields nanostructures, thus providing material candidates suitable for next-generation nanophotonic platforms (Supplementary Fig. 17).

Humidity-induced decryption and encryption

By utilizing the distinct differences in swelling ratios among CMC-N3-Γ, we can create a single-layered, tunable anti-counterfeiting device. Here, the design principle is to present only a single color in an encrypted state by ensuring that the effective optical path length of the plasmonic cavity remains uniform across every position. When decrypted, these optical path lengths adjust independently and transform into a different preprogrammed image. The key to successfully implementing this mechanism is to customize the Γ at each position to control how much each hydrogel film swells under humidity. We precisely calculated the optimal height and the corresponding Γ to make a blueprint using an optical simulation. For the basic configuration, a square pixel array with different Γ was prepared in Fig. 5a at an identical initial thickness (t = 149 ± 4 nm). While each pixel shows nearly identical color in dry state (RH 40%), the difference between each pixel in the array becomes distinct as humidity level increases (RH 90%) (Fig. 5b, c).

Fig. 5. Humidity-responsive image decryption and encryption.

Fig. 5

a Schematic illustration showing square pixels (480 µm × 480 µm), each containing CMC-N3 with different Γ. Simulated (b) and experimented square pixels (c) for the humidity-induced decryption. d Schematic illustration showing the design of humidity-induced decryption and encryption of the masterpiece, Sunflowers, by Vincent van Gogh. Simulated (e) and experimented results (f) of humidity-induced decryption in Sunflowers. Simulated (g) and experimented results (h) of humidity-induced encryption in Sunflowers. Scale bar is 250 µm for c, and 2.5 mm for f and h.

We can further extend this concept to realize humidity-induced decryption and encryption systems embedding complex information. To demonstrate this, Vincent van Gogh’s masterpiece, Sunflowers, was designed to be encrypted with four different Γ in both dry and humid states (Fig. 5d). Through repeated photolithography processing, we demonstrated that a single layer of chemically optimized CMC-N3-Γ can be patterned at desired positions; note that there exists some unavoidable experimental defects and slight vacancies at the boundaries that leads to an ambiguous shadow in the encrypted images. However, the image made in uniform thicknesses of 153 ± 3 nm can be hidden and encrypted, exhibiting single dark blue color at dry state but dramatically transforms into vivid multi-color image when the RH reaches 96% (Fig. 5e, f). Likewise, the image can be also encrypted in humid state through patterning films with different thicknesses (94, 102, 106 and 113 nm) (Fig. 5g, h). The image appears in brownish tone colors in dry state, but are suddenly concealed with bluish-white color by exhibiting a uniform thickness of 171 ± 2 nm throughout the device at RH 86%.

Transformative anti-counterfeiting display

Unlike previously reported devices that can only display static or singular encrypted images, the ability to simultaneously control both chemical and geometric aspects in our device allows complete transformation of entirely different images (Fig. 6a). For such transformative display to work, every local point of the device must reproduce desired colors at specific humidity levels. To achieve this, we decomposed the required swelling properties for each local point into separate layers each referred as, (grafting ratio (Γ), thickness, (t)). As these (Γ, t) pairs are orthogonal to each layer, the multi-stacking process allows to assign different swelling behaviors for any arbitrary point. Thus, this method enables optical multiplexing of various images, staying fully encrypted and revealing specific information only at designated humidity levels, making them a highly secured encryption system.

Fig. 6. Operating principles and image transformation in anti-counterfeiting display.

Fig. 6

a Schematics illustrating the chemically and geometrically programmable device for multi-image transformation. b Blueprint of the transformative 7-digit segment, which displays 2 and 5 in dry and humid environment, respectively. Simulated (c) and experimented results (d) of the transformative 7-segment display. e A plot showing the experimental thickness of the hydrogel film in four distinct areas before and after humidification. Error bars are displayed as mean ± σ (n  =  5 and 3 for indepenant dry and humid experiments). Simulated (f) and experimented results (g) of transformative QR code display. The scale bar is 500 µm for d and 5 mm for g.

To validate this concept, we simultaneously exploited the chemical and geometrical programmability of our device for transformative 7-segment display (Fig. 6b). Creating the reconfigurable device involves setting up four distinct areas with different thicknesses and suitable CMC-N3-Γ. First, we employed all regions except the iii zone with CMC-N3-17.1 of different thicknesses. For the iii zone, however, we used a film of CMC-N3-7.5, anticipating a relatively more pronounced swelling behavior; this enables the transformation from a yellow 2 against a blue background at dry state (RH 40%) to a purple 5 on a yellow background at humid state (RH 85%) (Fig. 6c). Indeed, we observed smooth transformation of the number 2 into 5 while maintaining color consistency with minimal deviations (Fig. 6d). When the RH reached 85%, the i, ii, and iv zones accurately displayed the simulated color, with their thicknesses increasing by 1.58, 1.60, and 1.56-fold, respectively, while the iii zone experienced 1.65-fold increase in thickness (Fig. 6e). To underscore the significance of chemical programmability, we also conducted simulations and fabrication for a case where the thickness increase factor was consistently set across all zones by employing only CMC-N3-17.1 (Supplementary Fig. 18). In such case where only geometrically programmability was utilized to realize 7-segment display, the regions that should have a yellow hue will instead have a green hue at RH 85%. This discrepancy could potentially cause cognitive misunderstandings in complex image transformation, highlighting the significance of tailoring the individual chemical properties.

These image transformations can be also extended to QR codes to improve information capacities and integrate with network systems (Fig. 6f). Here, the key design principle is the perception of each colored pixel as separate dots due to too high color contrast among these elements, making it impossible for the camera and the QR code software algorithm to recognize it as a QR code pattern. However, as the humidity reaches a specific RH value, these information-providing pixels shift as similar tones, while increasing the color contrast with the background and thus appearing as a QR code. Indeed, we observe that the resulting image remains unrecognized and hidden under low humidity levels. However, at RH 85%, a clear QR code linked to a specific website emerges (Fig. 6g and Supplementary Movie 1 and 2). More importantly, this QR code can be additionally shifted by further elevating the humidity level (RH 95%), transforming into a new QR code pattern. Overall, these proof-of-concept experiments clearly demonstrate that the photoreactive polymer-based multiplexing technology outlined in this work will have a broad impact on many advanced optical applications involving anti-counterfeiting and dynamic color printing to name a few.

In conclusion, we present an advanced toolbox in chemically and geometrically programming humidity-sensitive optical devices using azido-grafted CMC. The UV-curability of this hydrophilic polymer grants precise and orthogonal control over hydration behavior at any arbitrary point in the device through variation in grafting ratio (Γ) and multi-layering capability using lithography techniques. The resulting hydrogel structures are integrated with a disordered plasmonic system to express strong reflectance that covers RGB colors depending on the swollen thickness. Our device surpasses conventional humidity-responsive systems, demonstrating transformational anti-counterfeiting multiplexer that can encode/decode different full-color images based on humidity triggers. This research represents a leap in programmable optical devices, illustrating how the chemical modifications in polymers can unlock new controllable dimensions and functionality. More importantly, we anticipate that devices capable of transforming even more images can be potentially realized by leveraging the distinct refractive indices and swelling behaviors of each layer within the multi-layer structured hydrogel using additional chemical techniques, which further increases the design freedom in chemical programmability; some of these examples may include exploring other polymers40,42, or varying the molecular weight of polymers to realize different swelling tendency49, or adding non-swelling dielectric materials40. Therefore, we believe that our work paves the way for versatile applications in reconfigurable optical devices, such as colorimetric gas sensors, optical security devices, and structural color displays.

Online content

Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available.

Methods

Synthesis of CMC-N3-Γ

Photoreactive carboxymethyl cellulose-azido derivatives (CMC-N3) were synthesized by grafting phenyl azido groups onto CMC using carbodiimide reaction37,47,50. To begin with, 600 mg of CMC (Mw = 250,000 g mol−1) was dissolved in 60 mL of deionized (DI) water, and the pH was subsequently adjusted to 7.0 using 0.1 M NaOH or HCl. In the case of CMC-N3-4.7, 108 mg of N-(3dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDC), 108 mg of N-hydroxysuccinimide (NHS, 98%), and 23 mg of 4-azidoaniline hydrochloride were introduced into the CMC solution and allowed to react overnight at 4 °C. CMC-N3-7.5, CMC-N3-17.1, and CMC-N3-23.7 were prepared similarly but with feed ratios 1.67, 5.00, and 8.33 times that of CMC-N3-4.7, respectively. All solutions were dialyzed for 72 h with DI water, followed by lyophilization to retrieve CMC-N3.

Characterization

1H NMR spectra were acquired using a 500 MHz NMR spectrometer (Bruker). The Fourier transform infrared (FT-IR) spectra were collected using FT-IR spectroscopy (PerkinElmer). The absorbance spectra of the CMC-N3-Γ were measured using a spectrophotometer (UV−1800, Shimadzu). The refractive index of the CMC-N3-Γ were measured with an ellipsometer (M-2000, J.A. Woollam) using Cauchy model. The thickness and surface morphology of the polymeric patterns and Ag islands on these patterns were characterized using a stylus profilometer (Dektak XT, Bruker) and measured at a certain RH by AFM (Park System, XE7). Reflectance of optical devices under varying RH conditions was measured using a specialized humidity chamber designed to precisely control RH levels and the measurements were conducted at room temperature (20° ± 2 °C) utilizing a high-precision spectrometer (iHR320, Horiba). Response time, recovery time, and repeatability measurements were performed using a supercontinuum laser (SuperK FIANIUM, NKT Photonics), photodiode (SM05PD2A, Thorlabs), and data acquisition board (USB-6259, National Instruments).

Fabrication of disordered plasmonic devices

Prior to device fabrication, the substrate underwent a cleansing process using deionized (DI) water, ethanol, acetone, and isopropanol, and subsequently dried using nitrogen. For the adhesive layer, chitosan (CHI, low molecular weight) was firstly dissolved in water (1.0 wt%). Before dissolving CHI, 1.0% (v/v) acetic acid was added and the solution was then left to stir overnight at 1000 rpm and 65 °C. The CHI solution was applied to a plasma-treated silicon wafer substrate, followed by spin-coating at spinning speed of 6000 rpm. Subsequently, the CHI-coated wafer was placed on top of a heating plate set at 65 °C to facilitate thermal crosslinking51. By modulating the grafting ratio of CMC-N3-Γ, the solution concentration as well as the spinning speed, CMC-N3 layer with predetermined swelling behavior and thickness is assembled. Targeted UV exposure (15 mW cm−2 for 10 min) using a photomask ensures crosslinking to occur only in specified regions. Subsequent rinsing with water leaves behind the crosslinked areas while the unexposed region is easily removed. Further layers can be additionally assembled onto designated regions using a mask aligner (MDA-400M, MIDAS system) and photomasks. The process of UV curing and rinsing is repeated to achieve multi-layering. 5-nm Ag islands are deposited on the CMC-N3-Γ using an E-beam evaporator system (model KVE-C300160, Korea Vacuum Tech) with a high-purity Ag target under a base pressure of 3 × 10−6 Torr and a deposition rate of 1 Å/s.

Optical simulations

The simulated reflectance spectra were computed using the Finite-Difference Time-Domain (FDTD) method with a commercially available Lumerical FDTD solution. Ag islands were randomly positioned with Gaussian distributed radius across 500 μm × 500 μm area. To minimize the calculation error, FDTD simulations were iterated 10 times using random seeds. Additionally, S-parameter analysis was performed to derive the effective refractive index of the nanoparticle layer. This calculation was executed utilizing the embedded function within the FDTD solution and averaged over 100 repetitions to account for the layer’s randomness. The effective material properties were calculated through custom MATLAB code developed by others. The simulation used a refractive index that reflects the shift of the refractive index with humidity, based on experimentally measured data at RH 40%. The simulation of the glanced incident angle was calculated using the BFAST mode in FDTD. Chromaticity simulations were performed with OptProp, a MATLAB toolbox specialized in color-related analyses. For color transformations, the CIE 1931 observer function with D50 was employed.

Supplementary information

Peer Review File (3.2MB, pdf)
41467_2024_50876_MOESM3_ESM.pdf (77.4KB, pdf)

Description of Additional Supplementary Files

Supplementary Movie 1 (9.9MB, mp4)
Supplementary Movie 2 (15.2MB, mp4)

Source data

Source data (5.9MB, zip)

Acknowledgements

J.Y., and C.J. contributed equally to this work. This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. RS-2023-00208746, H.L.), (No. 2021R1A4A1021972, H.L.), (No. RS-2023-00260454, H.L.), (No. 2022M3C1A3081312, J.R.), and (No. RS-2024-00356928, J.R.). J.R. also acknowledges the Korea Evaluation Institute of Industrial Technology (KEIT) grant funded by the Korean government (MOTIE) (No. 1415185027 /20019169, Alchemist project), the POSTECH-Samsung Semiconductor Research Center program (IO201215-08187-01) funded by Samsung Electronics, the Samsung Research Funding & Incubation Center for Future Technology grant (SRFC-IT1901-52) funded by Samsung Electronics, and the POSCO-POSTECH-RIST Convergence Research Center program funded by POSCO.

Author contributions

J.Y. and H.L. conceived the idea and initiated the project. J.Y. and C.J. designed and fabricated the samples. J.Y. and C.J. characterized the materials and analyzed the data. C.J. and J.K. performed the experimental measurements and conducted optical simulations. J.R. and H.L. guided and supervised the research. All authors discussed the contents and wrote the manuscript.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Data availability

All relevant data supporting the findings of this study are available within the article and its Supplementary Information files as well as Source Data. Source data are provided with this paper. All data are available from the corresponding author. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Jongsun Yoon, Chunghwan Jung.

Change history

9/19/2024

A Correction to this paper has been published: 10.1038/s41467-024-52592-z

Contributor Information

Junsuk Rho, Email: jsrho@postech.ac.kr.

Hyomin Lee, Email: hyomin@postech.ac.kr.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-024-50876-y.

References

  • 1.Braun, P. V. Colour without colourants. Nature472, 423–424 (2011). [DOI] [PubMed] [Google Scholar]
  • 2.Zhao, Y., Xie, Z., Gu, H., Zhu, C. & Gu, Z. Bio-inspired variable structural color materials. Chem. Soc. Rev.41, 3297–3317 (2012). [DOI] [PubMed] [Google Scholar]
  • 3.Cuthill, I. C. et al. The biology of color. Science357, eaan0221 (2017). [DOI] [PubMed] [Google Scholar]
  • 4.Geng, J., Xu, L., Yan, W., Shi, L. & Qiu, M. High-speed laser writing of structural colors for full-color inkless printing. Nat. Commun.14, 565 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.ElKabbash, M. et al. Fano resonant optical coatings platform for full gamut and high purity structural colors. Nat. Commun.14, 3960 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Yang, Y. et al. Nanostructure-free crescent-shaped microparticles as full-color reflective pigments. Nat. Commun.14, 793 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.So, S., Mun, J., Park, J. & Rho, J. Revisiting the Design Strategies for Metasurfaces: Fundamental Physics, Optimization, and Beyond. Adv. Mater.35, 2206399 (2023). [DOI] [PubMed] [Google Scholar]
  • 8.Yang, Y. et al. Revisiting Optical Material Platforms for Efficient Linear and Nonlinear Dielectric Metasurfaces in the Ultraviolet, Visible, and Infrared. ACS Photonics10, 307–321 (2023). [Google Scholar]
  • 9.Kim, J. B., Chae, C., Han, S. H., Lee, S. Y. & Kim, S.-H. Direct writing of customized structural-color graphics with colloidal photonic inks. Sci. Adv.7, eabj8780 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Tittl, A. Tunable structural colors on display. Light Sci. Appl.11, 155 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Kim, S. et al. Self-assembled pagoda-like nanostructure-induced vertically stacked split-ring resonators for polarization-sensitive dichroic responses. Nano Converg.9, 40 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Choi, S. et al. Structural color printing via polymer-assisted photochemical deposition. Light Sci. Appl.11, 84 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Cencillo-Abad, P., Franklin, D., Mastranzo-Ortega, P., Sanchez-Mondragon, J. & Chanda, D. Ultralight plasmonic structural color paint. Sci. Adv.9, eadf7207 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Joo, W.-J. et al. Metasurface-driven OLED displays beyond 10,000 pixels per inch. Science370, 459–463 (2020). [DOI] [PubMed] [Google Scholar]
  • 15.Yang, W. et al. All-dielectric metasurface for high-performance structural color. Nat. Commun.11, 1864 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Kim, I. et al. Metasurfaces-Driven Hyperspectral Imaging via Multiplexed Plasmonic Resonance Energy Transfer. Adv. Mater.35, 2300229 (2023). [DOI] [PubMed] [Google Scholar]
  • 17.Dong, Z. et al. Schrödinger’s red pixel by quasi-bound-states-in-the-continuum. Sci. Adv.8, eabm4512 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Yang, B. et al. Ultrahighly Saturated Structural Colors Enhanced by Multipolar-Modulated Metasurfaces. Nano Lett.19, 4221–4228 (2019). [DOI] [PubMed] [Google Scholar]
  • 19.Jung, C. et al. Near-zero reflection of all-dielectric structural coloration enabling polarization-sensitive optical encryption with enhanced switchability. Nanophotonics10, 919–926 (2021). [Google Scholar]
  • 20.Lee, C., Lee, S., Seong, J., Park, D. Y. & Rho, J. Inverse-designed metasurfaces for highly saturated transmissive colors. J. Opt. Soc. Am. B41, 151–158 (2024). [Google Scholar]
  • 21.Song, M. et al. Versatile full-colour nanopainting enabled by a pixelated plasmonic metasurface. Nat. Nanotechnol.18, 71–78 (2023). [DOI] [PubMed] [Google Scholar]
  • 22.Xu, C., Stiubianu, G. T. & Gorodetsky, A. A. Adaptive infrared-reflecting systems inspired by cephalopods. Science359, 1495–1500 (2018). [DOI] [PubMed] [Google Scholar]
  • 23.Vatankhah-Varnosfaderani, M. et al. Chameleon-like elastomers with molecularly encoded strain-adaptive stiffening and coloration. Science359, 1509–1513 (2018). [DOI] [PubMed] [Google Scholar]
  • 24.Lopez-Garcia, M. et al. Light-induced dynamic structural color by intracellular 3D photonic crystals in brown algae. Sci. Adv.4, eaan8917 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Jung, C. et al. Metasurface-Driven Optically Variable Devices. Chem. Rev.121, 13013–13050 (2021). [DOI] [PubMed] [Google Scholar]
  • 26.Shao, L., Zhuo, X. & Wang, J. Advanced Plasmonic Materials for Dynamic Color Display. Adv. Mater.30, 1704338 (2018). [DOI] [PubMed] [Google Scholar]
  • 27.Chen, J., Song, G., Cong, S. & Zhao, Z. Resonant-Cavity-Enhanced Electrochromic Materials and Devices. Adv. Mater.35, 2300179 (2023). [DOI] [PubMed] [Google Scholar]
  • 28.Yang, Y. et al. Integrated metasurfaces for re-envisioning a near-future disruptive optical platform. Light Sci. Appl.12, 152 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Cao, X. et al. Challenges and Opportunities toward Real Application of VO2-Based Smart Glazing. Matter2, 862–881 (2020). [Google Scholar]
  • 30.Li, Y. et al. Colorful Electrochromic Displays with High Visual Quality Based on Porous Metamaterials. Adv. Mater.35, 2300116 (2023). [DOI] [PubMed] [Google Scholar]
  • 31.Xiong, K. et al. Video-Rate Switching of High-Reflectivity Hybrid Cavities Spanning All Primary Colors. Adv. Mater.35, 2302028 (2023). [DOI] [PubMed] [Google Scholar]
  • 32.Xiong, K. et al. Video Speed Switching of Plasmonic Structural Colors with High Contrast and Superior Lifetime. Adv. Mater.33, 2103217 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Badloe, T. et al. Liquid crystal-powered Mie resonators for electrically tunable photorealistic color gradients and dark blacks. Light Sci. Appl.11, 118 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Miller, B. H., Liu, H. & Kolle, M. Scalable optical manufacture of dynamic structural colour in stretchable materials. Nat. Mater.21, 1014–1018 (2022). [DOI] [PubMed] [Google Scholar]
  • 35.Nam, S., Wang, D., Kwon, C., Han, S. H. & Choi, S. S. Biomimetic Multicolor-Separating Photonic Skin using Electrically Stretchable Chiral Photonic Elastomers. Adv. Mater.35, 2302456 (2023). [DOI] [PubMed] [Google Scholar]
  • 36.Heuser, T., Merindol, R., Loescher, S., Klaus, A. & Walther, A. Photonic Devices Out of Equilibrium: Transient Memory, Signal Propagation, and Sensing. Adv. Mater.29, 1606842 (2017). [DOI] [PubMed] [Google Scholar]
  • 37.Dong, Z. et al. Photopatternable Nanolayered Polymeric Films with Fast Tunable Color Responses Triggered by Humidity. Adv. Funct. Mater.29, 1904453 (2019). [Google Scholar]
  • 38.Jang, J. et al. Self-Powered Humidity Sensor Using Chitosan-Based Plasmonic Metal–Hydrogel–Metal Filters. Adv. Opt. Mater.8, 1901932 (2020). [Google Scholar]
  • 39.Yoo, Y. J. et al. Large-Area Virus Coated Ultrathin Colorimetric Sensors with a Highly Lossy Resonant Promoter for Enhanced Chromaticity. Adv. Sci.7, 2000978 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Jung, C. et al. Disordered-nanoparticle–based etalon for ultrafast humidity-responsive colorimetric sensors and anti-counterfeiting displays. Sci. Adv.8, eabm8598 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Ko, B. et al. Tunable metasurfaces via the humidity responsive swelling of single-step imprinted polyvinyl alcohol nanostructures. Nat. Commun.13, 6256 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Ko, B. et al. Humidity-Responsive RGB-Pixels via Swelling of 3D Nanoimprinted Polyvinyl Alcohol. Adv. Sci.10, 2204469 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Dai, C. et al. Direct-Printing Hydrogel-Based Platform for Humidity-Driven Dynamic Full-Color Printing and Holography. Adv. Funct. Mater.33, 2212053 (2023). [Google Scholar]
  • 44.Dai, C. et al. Switchable unidirectional emissions from hydrogel gratings with integrated carbon quantum dots. Nat. Commun.15, 845 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Hu, W. et al. Hydrogel-Based Reconfigurable Visualization with Pixel-Scale Programmability by Vapor Exhalation. Laser Photonics Rev.17, 2200725 (2023). [Google Scholar]
  • 46.Inomata, H., Wada, N., Yagi, Y., Goto, S. & Saito, S. Swelling behaviours of N-alkylacrylamide gels in water: effects of copolymerization and crosslinking density. Polymer36, 875–877 (1995). [Google Scholar]
  • 47.Chen, X.-C., Huang, W.-P., Ren, K.-F. & Ji, J. Self-Healing Label Materials Based on Photo-Cross-Linkable Polymeric Films with Dynamic Surface Structures. ACS Nano12, 8686–8696 (2018). [DOI] [PubMed] [Google Scholar]
  • 48.Yoon, J. et al. Tailoring the Hydrophilicity for Delayed Condensation Frosting in Antifogging Coatings. ACS Appl. Mater. Interfaces14, 35064–35073 (2022). [DOI] [PubMed] [Google Scholar]
  • 49.Park, H. et al. Effect of Swelling Ratio of Injectable Hydrogel Composites on Chondrogenic Differentiation of Encapsulated Rabbit Marrow Mesenchymal Stem Cells In Vitro. Biomacromolecules10, 541–546 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Wu, G., Shi, F., Wang, Z., Liu, Z. & Zhang, X. Poly(acrylic acid)-Bearing Photoreactive Azido Groups for Stabilizing Multilayer Films. Langmuir25, 2949–2955 (2009). [DOI] [PubMed] [Google Scholar]
  • 51.Dong, J., Yu, D., Yu, Z., Zhang, L. & Xia, W. Thermally-induced crosslinking altering the properties of chitosan films: Structure, physicochemical characteristics and antioxidant activity. Food Packag. Shelf Life34, 100948 (2022). [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Peer Review File (3.2MB, pdf)
41467_2024_50876_MOESM3_ESM.pdf (77.4KB, pdf)

Description of Additional Supplementary Files

Supplementary Movie 1 (9.9MB, mp4)
Supplementary Movie 2 (15.2MB, mp4)
Source data (5.9MB, zip)

Data Availability Statement

All relevant data supporting the findings of this study are available within the article and its Supplementary Information files as well as Source Data. Source data are provided with this paper. All data are available from the corresponding author. Source data are provided with this paper.


Articles from Nature Communications are provided here courtesy of Nature Publishing Group

RESOURCES