Agarose/PNIPAAm semi-interpenetrating network hydrogels with enhanced mechanical and optical properties for thermoregulating smart windows

Abstract

We report a novel semi-interpenetrating network (s-IPN) hydrogel fabricated using a simple diffusion method that incorporates poly(N-isopropylacrylamide) (PNIPAAm) into agarose matrices. The agarose serves as a structural framework while PNIPAAm provides thermoresponsive capability, creating a straightforward, stable, and thermally responsive material for practical applications. This approach notably reduces volume shrinkage from 80–90% (typical of pure PNIPAAm) to approximately 12%, corresponding to only 4% linear thermal contraction, while preserving complete thermoresponsive functionality. The optimized composition (2% agarose/8% PNIPAAm) exhibits approximately 90% visible light transmittance at room temperature while becoming opaque above its lower critical solution temperature (LCST) of 32.1 °C. Thermogravimetric analysis and FTIR spectroscopy reveal enhanced thermal stability and molecular interactions between the agarose and PNIPAAm networks through hydrogen bonding. The properties of the PNIPAAm–agarose s-IPN hydrogel can be systematically controlled by simply adjusting the concentration of each polymer, enabling customization of the smart hydrogel properties. When incorporated into a glass–polymer–glass sandwich structure, these s-IPN hydrogels function effectively as smart window materials, providing autonomous temperature regulation by modulating solar transmittance in response to temperature changes.

---

1. Introduction

The interface between soft matter science and energy applications presents both fundamental scientific questions and practical engineering challenges.¹ Understanding how molecular interactions within multi-component hydrogel networks influence macroscopic properties such as mechanical integrity, optical transmittance, and thermal response kinetics is crucial for designing next-generation smart materials.² Stimuli-responsive hydrogels represent a fascinating class of soft materials that undergo reversible physical or chemical changes in response to external triggers such as temperature, pH, or light.^3,4 Among these, thermally responsive systems based on poly(N-isopropylacrylamide) (PNIPAAm) have been extensively studied owing to their well-defined phase transition behavior near physiological temperatures.^5,6 PNIPAAm exhibits a lower critical solution temperature (LCST) of approximately 32 °C in aqueous solutions, above which it undergoes a coil-to-globule transition accompanied by significant changes in hydrophilicity, volume, and optical properties.^7,8 Conventional PNIPAAm hydrogel fabrication requires complex chemical crosslinking reactions, typically involving initiators and crosslinkers for free radical polymerization.⁹ These methods suffer from harsh reaction conditions, difficult crosslinking density control, and high costs, limiting their large-scale applications. Conventional PNIPAAm hydrogels frequently also suffer from poor mechanical stability and slow response rates, which substantially restrict their practical applications.¹⁰

Recent advances in soft matter engineering have demonstrated that the strategic integration of different polymer networks can yield materials with programmable responses and enhanced functionality.^11,12 For instance, double-network hydrogels with exceptional mechanical strength have been achieved through careful design of network topology and interactions.¹³ Among the various multi-network architectures, known as interpenetrating polymer networks (IPNs), semi-interpenetrating polymer networks (s-IPNs) represent a particularly promising class. These systems consist of one crosslinked polymer network with another linear polymer chain interpenetrating at the molecular level, offering unique advantages in tailoring mechanical and responsive properties simultaneously.^14,15 Unlike full interpenetrating networks, s-IPNs maintain the mobility of the uncrosslinked polymer component, potentially preserving its rapid response to external stimuli while benefiting from the structural support of the host network.¹⁶

Conventional PNIPAAm hydrogels face critical limitations, including significant volume shrinkage (typically 80–90% upon heating above the LCST),¹⁷ poor mechanical stability,⁹ and fabrication difficulties¹⁸ that severely restrict their practical applications, particularly in smart windows where dimensional stability is essential. To overcome these challenges, we developed a novel s-IPN system strategically combining agarose as the structural framework with PNIPAAm as the responsive component. Agarose, a natural polysaccharide forming robust double-helix networks, was selected for its unique combination of optical transparency, thermal stability, and controllable pore architecture.^19–21 The pre-formed agarose matrix provides abundant hydrogen bonding sites and tunable mechanical properties while serving as an ideal host that constrains PNIPAAm volume changes during thermal transitions. This strategic design enables the s-IPN system to exhibit synergistic properties—maintaining thermoresponsive functionality while achieving enhanced mechanical integrity and dimensional stability that overcome the individual limitations of each component.²² This post-gelation diffusion method only requires immersing preformed agarose gels in a PNIPAAm solution, eliminating the need for any chemical crosslinkers and operating at room temperature.

In the context of energy-efficient building technologies, smart windows capable of dynamically regulating solar heat gain represent a promising application for responsive soft materials.^23,24 Traditional approaches to smart windows have predominantly relied on electrochromic²⁵ or thermochromic technologies,²⁶ often requiring complex manufacturing processes or external power sources. Hydrogel-based systems offer an alternative approach through their autonomous response to environmental temperature changes, potentially enabling passive thermal management without electrical input.^8,27

In this work, we report a novel approach to fabricating robust, thermally responsive hydrogels through the controlled diffusion of PNIPAAm into preformed agarose matrices. This strategy creates s-IPN structures with significantly enhanced mechanical properties compared to pure agarose gels, while maintaining excellent optical clarity at room temperature and efficient thermally induced opacity at elevated temperatures. We systematically investigate how the composition influences the structural, mechanical, and optical properties of these s-IPN hydrogels and demonstrate their application in smart window configurations for temperature regulation.

2. Materials and methods

2.1 Materials and hydrogel preparation

N-Isopropylacrylamide (NIPAAm, 97%) was purchased from Tokyo Chemical Industry, poly(N-isopropylacrylamide) (PNIPAAm) from Scientific Polymer Products, and agarose from USB Corporation. The heat-absorbing film (T73) was obtained from local suppliers.

The agarose/PNIPAAm semi-interpenetrating network (s-IPN) hydrogels were prepared using an immersion method (Fig. 1(a)). Agarose solutions (1%, 2%, and 4% w/v) were prepared with deionized water at 95 °C, cast into PMMA molds, and cooled for 2 hours. These preformed hydrogels were immersed in PNIPAAm solutions (1%, 4%, 6%, and 8% w/v) at room temperature and then rinsed with deionized water to remove the surface-adhered polymer before characterization. Sample designations follow “axny-s,” where “x” is the agarose concentration (%), “y” is the PNIPAAm concentration (%), and “s” indicates s-IPN.

Fig. 1 Fabrication and visual characteristics of agarose/PNIPAAm s-IPN hydrogels. (a) Schematic illustration of the preparation process using the immersion method. (b) Photograph of pure 2% agarose gel. (c) Photograph of a2n8-s-IPN hydrogel (2% agarose with 8% PNIPAAm) at room temperature showing high transparency. (d) The same a2n8-s-IPN hydrogel at 40 °C demonstrating thermally induced opacity. (e) Photograph of a2n1-IPN hydrogel (2% agarose with 1% PNIPAAm) at room temperature. (f) a2n1-IPN hydrogel at 40 °C showing the temperature-responsive opacity change. All samples are 1 cm³ in size.

2.2 Diffusion measurements

PNIPAAm diffusion into agarose matrices was measured using a one-dimensional setup. Cylindrical agarose gels (2% and 4% w/v, 10 × 50 mm) were prepared in glass tubes with one end exposed to PNIPAAm solutions. After heating to 45 °C, the visible boundary indicating the diffusion front was measured at predetermined intervals using a digital caliper. The diffusion coefficient (D) was calculated using D = x²/2t, where x is the diffusion length and t is the time. All measurements were performed in triplicate.

2.3 Optical and mechanical characterization

Optical properties were characterized using a UV-Visible spectrophotometer (USB-2000, Ocean Optics). Hydrogel samples in water-filled cuvettes were measured for transmittance over 500–900 nm at 25 °C (below the LCST).

Mechanical properties were evaluated using a custom-built apparatus with an electric displacement platform, a load cell, and a LabView control system. Samples were temperature-controlled using a water bath and removed for measurements.

2.4 FTIR spectroscopy

Fourier-transform infrared (FTIR) spectroscopy was performed using attenuated total reflectance (ATR) mode to investigate molecular interactions between PNIPAAm and agarose in the s-IPN hydrogels. Hydrogel samples were analyzed directly in their hydrated state without any drying or sample preparation. Fresh hydrogel specimens were carefully placed on a diamond ATR crystal and gently pressed to ensure complete contact with the crystal surface.

2.5 Thermal analysis

Differential scanning calorimetry (Netzsch 204 F1) was used to determine the LCST, while thermogravimetric analysis (TA Instruments TGA Q500) was used to evaluate thermal stability.

2.6 Smart window fabrication and evaluation

Smart window prototypes were fabricated in three configurations (Fig. 6(a)): (1) glass–PMMA–glass, (2) glass–PMMA–glass–T73 film, and (3) glass–PMMA (containing the a2n8-s-IPN hydrogel)–glass–T73. The PMMA spacer (1.4 mm) created a cavity for the hydrogel.

Thermal regulation performance was evaluated using a halogen lamp positioned 5.5 cm from the window (Fig. 6(b)). Temperatures at the exterior and interior surfaces were monitored using thermocouples connected to a data acquisition system.

3. Results and discussion

3.1 Fabrication and structural characteristics of s-IPN hydrogels

The fabrication of agarose/PNIPAAm s-IPN hydrogels involves a straightforward approach by simply immersing pre-formed agarose gels in a PNIPAAm solution. As illustrated in Fig. 1(a), this method allows controlled diffusion of PNIPAAm into agarose matrices, creating a semi-interpenetrating network structure. Pure agarose gels (Fig. 1(b)) showed moderate transparency but lacked thermoresponsive behavior. Notably, incorporating PNIPAAm into the agarose network significantly enhanced optical clarity, as evidenced in Fig. 1(c) and (e), demonstrating the synergistic optical effect of the s-IPN structure. During the immersion process, the overall gel dimensions remained largely unchanged, but subsequent placement in water resulted in slight volume increases that depended on both the PNIPAAm solution concentration and the agarose network density. Higher agarose concentrations led to smaller volume changes, while higher PNIPAAm concentrations resulted in greater swelling. Specifically, 2% agarose gels immersed in 8% PNIPAAm solution exhibited approximately 10% linear expansion when placed in water, attributed to the enhanced hydration capacity of the gel network due to the incorporated PNIPAAm chains increasing the overall hydrophilicity of the composite system.

When the a2n8-s-IPN hydrogel (2% agarose with 8% PNIPAAm) and the a2n1-IPN hydrogel (2% agarose with 1% PNIPAAm) are compared, both hydrogels maintained high transparency at room temperature, with the a2n8-s-IPN showing noticeably improved clarity (Fig. 1(c)versusFig. 1(e)). This observation indicates that a higher PNIPAAm content enhances the optical properties of the s-IPN system at temperatures below the LCST, likely due to improved refractive index matching between the two polymer networks. Upon heating to 40 °C, both hydrogels transition to an opaque state (Fig. 1(d) and (f)), with the a2n8-s-IPN exhibiting more pronounced opacity due to the higher volume fraction of the thermally responsive component. Importantly, dimensional analysis of the a2n8-s-IPN during thermal cycling revealed only a 4% linear contraction upon heating above the LCST, representing a dramatic improvement over pure PNIPAAm hydrogels while maintaining effective optical switching. This minimal volume change demonstrates the successful constraint of PNIPAAm's typical large-scale shrinkage by the rigid agarose framework, a critical advantage for practical applications requiring dimensional stability.

This fabrication approach offers several advantages over previously reported methods for preparing thermoresponsive hydrogels. Unlike conventional copolymerization strategies,²⁸ our post-gelation diffusion method preserves the inherent network structure of agarose while introducing a thermoresponsive functionality. The successful integration of PNIPAAm within the agarose network is further supported by thermal stability data, which will be discussed in a later section.

3.2 Optical properties and temperature-responsive behavior

The optical properties of the s-IPN hydrogels were systematically investigated, with results presented in Fig. 2. Fig. 2(a) shows hydrogel samples with varying compositions, while Fig. 2(b) displays transmittance spectra of 2% agarose hydrogels with different PNIPAAm concentrations. Pure agarose gel exhibits the lowest transmittance (∼40% at 500 nm), while adding PNIPAAm progressively increases transparency, with a2n10-s-IPN reaching nearly 95% transmittance at 900 nm.

Fig. 2 Optical properties of agarose/PNIPAAm s-IPN hydrogels. (a) Comparative visualization of hydrogel samples in cuvettes with different agarose and PNIPAAm concentrations. (b) Transmittance spectra of s-IPN hydrogels with 2% agarose and varying PNIPAAm concentrations (0%, 4%, 6%, 8%, and 10%), demonstrating increased transmittance with a higher PNIPAAm content across the visible and near-infrared spectrum (500–900 nm). (c) Transmittance as a function of PNIPAAm concentration for different agarose concentrations (1%, 2%, 3%, and 4%).

Fig. 2(c) quantifies transmittance against the PNIPAAm concentration for different agarose concentrations (1–4%). Across all series, transmittance increases with the PNIPAAm content, with the effect most pronounced in lower agarose concentrations. The 1% agarose series reaches ∼90% transmittance with 10% PNIPAAm, while 4% agarose maintains lower values. This transmittance enhancement is attributed to optical homogenization within the gel network. While agarose exhibits a refractive index of approximately 1.335,²⁹ similar to water, the opacity of pure agarose gels arises from refractive index fluctuations at polymer–water interfaces within the porous microstructure. The incorporation of hydrated PNIPAAm chains (n ≈ 1.35)³⁰ into the agarose matrix fills these pores and reduces refractive index discontinuities, creating a more optically homogeneous medium that minimizes light scattering. These findings demonstrate that PNIPAAm serves a dual function in the s-IPN system: providing thermoresponsive switching capability while simultaneously enhancing optical clarity below the LCST. This synergistic optical enhancement is particularly valuable for smart window applications that require maximum transparency in the passive state to optimize daylighting performance.

3.3 Diffusion behavior and mechanical properties

The diffusion of PNIPAAm into agarose matrices was studied using a one-dimensional diffusion model in a cylindrical tube, with the visible color change boundary serving as an indicator of penetration depth when samples were heated to 45 °C. As shown in Fig. 3(a), the diffusion behavior of PNIPAAm exhibits strong dependence on the agarose concentration. The penetration depth of the 6% PNIPAAm solution was significantly greater in 2% agarose gels compared to 4% agarose gels after equivalent immersion periods. From our analysis of the one-dimensional diffusion equation (D = x²/2t), we calculated the average diffusion coefficients over the first 12 hours to be 2.56 × 10⁻⁸ cm² s⁻¹ and 1.21 × 10⁻⁸ cm² s⁻¹ for 2% and 4% agarose networks, respectively.

Fig. 3 Physical and mechanical characterization of the hydrogels. (a) Comparison of diffusion lengths of 6% PNIPAAm in 2% and 4% agarose gels, showing the effect of agarose concentration on PNIPAAm penetration. (b) Young's modulus of hydrogel samples as a function of immersion time, demonstrating the time-dependent mechanical property development. (c) Comparative Young's modulus of pure agarose and polymer-incorporated hydrogels at 25 °C and 40 °C, highlighting temperature-dependent mechanical behavior changes.

This approximately two-fold difference in diffusion rates correlates directly with the network pore structure: 2% agarose gels exhibit larger pore sizes (∼370 nm) compared to 4% agarose gels (∼250 nm),³¹ providing less restricted pathways for PNIPAAm chain diffusion. The denser 4% agarose network creates a more tortuous diffusion path with smaller effective pore dimensions, substantially reducing the mobility of PNIPAAm chains and confirming that the agarose network structure is a critical factor controlling the interpenetration process.

The diffusion coefficient values we obtained are lower than those reported in the literature,³² suggesting additional interactions beyond simple Fickian diffusion. As observed in Fig. 3(a), the relationship between diffusion length and time shows an initial rapid penetration followed by a more gradual progression, indicating complex diffusion kinetics potentially involving interactions between PNIPAAm and agarose.

The mechanical property development during the immersion process (Fig. 3(b)) provides insights into the network formation dynamics. In this experiment, the agarose concentration was kept constant, while the immersion time in PNIPAAm solution was varied. The Young's modulus increases progressively with immersion time, indicating the gradual establishment of physical interactions between PNIPAAm chains and the agarose network. This time-dependent evolution suggests that PNIPAAm not only fills the pores of the agarose network but also establishes secondary interactions with the polysaccharide chains, contributing to the overall mechanical integrity of the system.

A critical finding of our study concerns the temperature-dependent mechanical behavior of the s-IPN hydrogels (Fig. 3(c)). Our s-IPN hydrogels exhibit decreased Young's modulus when heated above the LCST of PNIPAAm. This softening effect differs from the behavior of other PNIPAAm hydrogels, which typically undergo significant volumetric shrinkage and stiffening upon heating.⁹ The conventional stiffening behavior occurs because collapsed PNIPAAm chains create a denser, more concentrated polymer network with reduced water content and increased chain entanglements. However, our s-IPN system exhibits the opposite behavior due to its unique structural arrangement. We attribute this phenomenon to the unique structural arrangement in our s-IPN system, where the collapsed PNIPAAm chains form isolated hydrophobic domains within the agarose network. In this collapsed state, PNIPAAm chains lose their ability to interact effectively with the agarose matrix, eliminating the reinforcing mechanical interactions that existed between the two networks at lower temperatures.

The temperature-dependent mechanical behavior provides direct evidence for intermolecular interactions: Young's modulus decreases from 0.2 MPa at 25 °C to 0.1 MPa at 40 °C, representing a 50% reduction that correlates with the disruption of hydrogen bonding networks above the LCST. From a thermodynamic perspective, Young's modulus is related to the second derivative of free energy with respect to strain (E ∝ ∂²F/∂ε²), allowing estimation of interaction energies. This temperature-induced softening represents an interesting property of our s-IPN design, offering potential applications in controlled deformation systems. The magnitude of this effect correlates with the PNIPAAm concentration, with a higher PNIPAAm content resulting in more pronounced mechanical changes upon heating.

3.4 Molecular interactions and FTIR analysis

Fourier-transform infrared (FTIR) spectroscopy was employed to investigate the molecular interactions between PNIPAAm and agarose in the s-IPN hydrogels. Fig. 4(a) presents the FTIR spectra of PNIPAAm gel, agarose gel, and the s-IPN hydrogel, revealing distinct spectral changes that provide evidence of intermolecular interactions. Key spectral features are marked with arrows: (i) OH/NH stretching region (3200–3600 cm⁻¹), (ii) C=O stretching region around 1650 cm⁻¹, and (iii) a unique absorption feature at 2000 cm⁻¹ present only in the s-IPN system. The OH/NH stretching region (3200–3600 cm⁻¹) in the s-IPN exhibits a subtle red shift compared to pure PNIPAAm. This frequency decrease is characteristic of hydrogen bond formation,^33–36 where the stretching vibrations of donor groups are weakened due to intermolecular interactions.

Fig. 4 FTIR spectra: (a) pure PNIPAAm gel (8%) (blue line), pure agarose gel (2%) (red line), and s-IPN hydrogel (black line, 2% agarose/8% PNIPAAm). Key spectral features indicating molecular interactions are marked with arrows: (i) OH/NH stretching region (3200–3600 cm⁻¹) showing peak shift in the s-IPN, (ii) C=O stretching region around 1650 cm⁻¹ with increased intensity, and (iii) unique absorption feature at 2000 cm⁻¹ present only in the s-IPN system. (b) s-IPN hydrogel (black line, 2% agarose/8% PNIPAAm), s-IPN hydrogel above the LCST, and water.

Fig. 5 Thermal characterization and proposed mechanism of the thermo-responsive hydrogels. (a) Low-temperature differential scanning calorimetry thermogram showing the phase transition temperature. (b) Thermogravimetric analysis curves illustrating the thermal stability of the hydrogels. (c) First derivative of thermogravimetric analysis data revealing the decomposition temperature profiles. (d) Schematic model of the proposed thermo-response mechanism in agarose/PNIPAAm s-IPN hydrogels, illustrating the molecular interactions during thermal transitions.

Analysis of the relative absorption intensities between the C=O stretching (1650 cm⁻¹)^35,37 and OH/NH stretching (3200 cm⁻¹) regions reveals significant changes upon s-IPN formation. Converting transmittance to absorbance values, the ratio A₁₆₅₀/A₃₂₀₀ increases from 0.628 in pure PNIPAAm to 0.759 in the s-IPN. This 21% increase in the relative intensity ratio may indicate that the C=O stretching vibration shows proportionally stronger enhancement compared to the OH/NH region. Such changes in relative intensities are typically associated with preferential involvement of specific functional groups in intermolecular interactions.^38–41 The increased ratio may suggest enhanced participation of carbonyl groups in hydrogen bonding networks, consistent with the formation of C=O⋯H–O hydrogen bonds between PNIPAAm and agarose.

A distinctive absorption peak appears at approximately 2000 cm⁻¹ in the s-IPN spectrum that is absent in both pure PNIPAAm and pure agarose. This feature represents a unique spectral signature of the composite system. In infrared spectroscopy, absorption bands in the 1900–2100 cm⁻¹ region are typically attributed to combination bands involving multiple vibrational modes or overtones of fundamental vibrations.⁴² The appearance of this peak only in the s-IPN may suggest the development of new vibrational coupling between the two polymer networks. To further validate this interpretation, we measured temperature-dependent FTIR changes in the s-IPN system above the LCST, as shown in Fig. 4(b). We found that the characteristic 2000 cm⁻¹ peak significantly diminishes upon heating above the LCST, while the A₁₆₅₀/A₃₂₀₀ intensity ratio also changes, indicating that these spectral features are directly correlated with PNIPAAm's phase transition.

3.5 Thermal characterization and mechanism of thermoresponsiveness

The thermal properties and phase transition behavior of the hydrogels were characterized using differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) (Fig. 5). DSC analysis revealed a well-defined endothermic peak corresponding to the phase transition of PNIPAAm, with the transition temperature (LCST) determined to be 32.1 °C for the a2n8-s-IPN hydrogel, identical to that of pure PNIPAAm. This finding indicates that the phase transition behavior of PNIPAAm is preserved within the agarose network, with no significant shift in transition temperature due to confinement effects or specific interactions with the host network (Fig. 5(a)).

TGA results provide compelling evidence of molecular-level interactions between the two network components. As shown in our data (Fig. 5(b)), the s-IPN hydrogels exhibit improved thermal stability compared to pure agarose gels, with the decomposition temperature shifting from 240 °C for pure agarose to 263 °C for the a2n8-s-IPN sample. Analysis of the first derivative of the TGA curve (Fig. 5(c)) reveals two distinct decomposition events in the s-IPN hydrogels, corresponding to the separate degradation of agarose and PNIPAAm components. However, both decomposition temperatures are higher than those of the respective pure components (agarose: 240 °C → 263 °C; PNIPAAm: 310 °C → 350 °C), suggesting that the interpenetrating structure provides mutual thermal stabilization. Quantitative analysis of the TGA curves through peak deconvolution and area integration reveals that the a2n8-s-IPN hydrogel contains a PNIPAAm-to-agarose weight ratio of approximately 1 : 1.5, indicating that substantial amounts of PNIPAAm successfully diffuse into and are retained within the agarose matrix during the immersion process. This mutual thermal stabilization confirms intermolecular interaction that restricts molecular mobility, consistent with the mechanical reinforcement observed below the LCST.

The enhanced thermal stability observed in the TGA can be attributed to intermolecular interactions between the PNIPAAm and agarose networks. FTIR spectroscopy provides direct evidence for these interactions, showing a red shift in the OH/NH stretching region and an increase in the C=O/OH intensity ratio. These intermolecular hydrogen bonds restrict molecular mobility and increase the energy required for thermal degradation, explaining the enhanced thermal stability.⁴³ The presence of these intermolecular interactions further confirms the successful formation of a true semi-interpenetrating network, rather than a simple physical mixture of the two polymers. Based on our thermal and spectroscopic characterization, at temperatures below the LCST, FTIR-confirmed hydrogen bonding networks between PNIPAAm and agarose create molecular anchoring points that constrain volume changes while maintaining optical clarity. Upon heating above the LCST, PNIPAAm chains undergo coil-to-globule transition, but the agarose framework prevents catastrophic volume shrinkage typically observed in pure PNIPAAm hydrogels (Fig. 5(d)). The spectroscopic evidence provides the molecular foundation for the unique combination of thermoresponsive functionality and dimensional stability observed in the s-IPN system.

Upon heating above the LCST, PNIPAAm chains undergo a conformational transition from extended coils to collapsed globules, driven by the entropic gain associated with releasing structured water molecules. This phase transition leads to the formation of hydrophobic PNIPAAm domains within the agarose network, creating refractive index heterogeneities that scatter light and reduce transparency. Crucially, as these PNIPAAm chains collapse, they lose their ability to form effective molecular interactions with the agarose matrix that previously contributed to the mechanical reinforcement of the network. The collapsed, globular PNIPAAm structures no longer bridge across agarose chains, thereby reducing the structural integrity that was present in the hydrated state. This disruption of polymer–polymer interactions between the two networks leads to the observed decrease in Young's modulus at elevated temperatures, as the synergistic reinforcement effect between PNIPAAm and agarose is diminished when the thermosensitive component transitions to its hydrophobic, collapsed state.

The agarose network plays a crucial role in this mechanism by (1) providing a structural scaffold that limits the macroscopic shrinkage typically associated with PNIPAAm collapse, (2) creating spatial constraints that influence the size and distribution of collapsed PNIPAAm domains, and (3) maintaining the overall water content of the hydrogel through its hydrophilic character. This synergistic interaction between the two networks is key to achieving the unique combination of enhanced optical clarity below the LCST and efficient optical switching above the LCST.

3.6 Application in smart windows and performance evaluation

Our experimental evidence supports an interaction mechanism between PNIPAAm and agarose networks. Below the LCST, extended PNIPAAm chains form hydrogen bonds with agarose, creating molecular anchoring points that enhance mechanical properties while maintaining optical clarity. Above the LCST, PNIPAAm undergoes coil-to-globule transition, disrupting hydrogen bonding networks and causing mechanical softening while creating refractive index heterogeneities for optical switching. This mechanism is supported by convergent evidence from mechanical testing, thermal analysis, and FTIR spectroscopy, demonstrating that the unique properties arise from specific intermolecular interactions rather than simple physical mixing. The practical application of the s-IPN hydrogels in smart window configurations was investigated, with results presented in Fig. 6. Three window configurations were evaluated (Fig. 6(a)): (1) glass–PMMA medium–glass, (2) glass–PMMA medium–T73–glass, and (3) glass–PMMA medium (containing the 2% agarose/8% PNIPAAm hydrogel)–T73–glass. The experimental setup (Fig. 6(b)) allowed for the assessment of temperature regulation performance during exposure to a halogen lamp providing both light and heat.

The temperature profiles during a single heating cycle (Fig. 6(c)) demonstrate the thermoregulating capability of the hydrogel-containing window configuration. While the reference configurations (without the hydrogel) exhibited continuous temperature increases, the s-IPN hydrogel-containing window effectively limited the interior temperature rise. This thermoregulating effect can be attributed to the dynamic modulation of optical transmittance by the hydrogel layer, which transitions from highly transparent to opaque as its temperature exceeds the LCST.

The reproducibility of the thermoregulating effect was confirmed through ten heating–cooling cycles (Fig. 6(d)). The hydrogel-containing window maintained consistent performance throughout the cyclic testing, with no significant degradation in its temperature regulation capability. The agarose network plays a crucial role in thermoregulatory performance through three key mechanisms. First, it provides an optically transparent matrix that maintains excellent light transmission in the passive state. Second, it provides structural stability during optical switching, maintaining hydrogel integrity and preventing deformation.

Fig. 6 Application of agarose/PNIPAAm s-IPN hydrogels in smart windows. (a) Schematic diagrams and photographs of three window configurations: (top) glass–PMMA medium–glass, (middle) glass–PMMA medium–T73–glass, and (bottom) glass–PMMA medium (containing 2% agarose/8% PNIPAAm hydrogel)–T73–glass. Bar: 15 mm. (b) Experimental setup for smart window performance evaluation, with a halogen lamp providing light and heat, measuring temperatures at simulated exterior and interior positions. (c) Temperature profiles of interior and exterior surfaces of different window configurations during a single heating cycle. Inset image: s-IPN hydrogel window after heating above the LCST, demonstrating thermally induced opacity. (d) Temperature response of the s-IPN hydrogel-containing window over ten heating–cooling cycles, demonstrating stability and reproducibility of the thermoregulating effect.

The observed thermoregulating performance of our s-IPN hydrogel-based smart windows compares favorably with alternative technologies. Unlike electrochromic systems that require continuous electrical input, our hydrogel-based approach achieves autonomous temperature regulation through passive response to environmental conditions. Furthermore, the enhanced optical clarity of our s-IPN hydrogels at room temperature ensures excellent daylighting performance under moderate temperature conditions, an important consideration for energy-efficient building design.

4. Conclusion

We have developed thermally responsive s-IPN hydrogels through controlled diffusion of PNIPAAm into agarose matrices, yielding materials with enhanced optical clarity and tunable mechanical properties. These hydrogels exhibit approximately 90% visible light transmittance at room temperature while becoming opaque above the LCST of 32.1 °C. Notably, incorporating PNIPAAm into agarose networks synergistically enhances their mechanical properties below the LCST, particularly in the optimized 2% agarose/8% PNIPAAm composition. Our post-gelation diffusion approach offers distinct advantages over traditional in situ polymerization methods, including better preservation of the primary agarose network structure and enhanced control over the thermoresponsive characteristics through the diffusion process.

Our investigation has revealed how the interplay between agarose's rigid skeleton and PNIPAAm's responsive chains enables efficient optical and mechanical transitions while constraining macroscopic shrinkage. When incorporated into a glass–polymer–glass sandwich structure, these hydrogels function effectively as thermoregulating smart window materials with demonstrated stability over multiple heating–cooling cycles. This work both advances the fundamental understanding of multi-component responsive hydrogels and provides a practical pathway toward energy-efficient building technologies based on soft matter principles.

Author contributions

Chun-Yen Wu performed the experiments and contributed to discussions. Hong-Ren Jiang designed the experiments and supervised the entire process.

Conflicts of interest

There are no conflicts to declare.

Data availability

All data supporting the conclusions of this article are included in the figures of this manuscript. Raw data are available from https://doi.org/10.5281/zenodo.17096569.

Acknowledgements

This work was supported by the National Science and Technology Council, Taiwan, under grant no. MOST 114-2112-M-002-024. Special thanks to Dr Jung-Yen Yang from the Taiwan Semiconductor Research Institute, NIAR, for his assistance in FTIR measurements.

References

1. A. Kuzyk, M. J. Urban, A. Idili, F. Ricci and N. Liu, Sci. Adv., 2017, 3, e1602803.
2. Q. Wang, J. L. Mynar, M. Yoshida, E. Lee, M. Lee, K. Okuro, K. Kinbara and T. Aida, Nature, 2010, 463, 339–343.
3. M. A. C. Stuart, W. T. S. Huck, J. Genzer, M. Müller, C. Ober, M. Stamm, G. B. Sukhorukov, I. Szleifer, V. V. Tsukruk, M. Urban, F. Winnik, S. Zauscher, I. Luzinov and S. Minko, Nat. Mater., 2010, 9, 101–113.
4. M. C. Koetting, J. T. Peters, S. D. Steichen and N. A. Peppas, Mater. Sci. Eng., R, 2015, 93, 1–49.
5. H. G. Schild, Prog. Polym. Sci., 1992, 17, 163–249.
6. A. Halperin, M. Kröger and F. M. Winnik, Angew. Chem., Int. Ed., 2015, 54, 15342–15367.
7. M. Heskins and J. E. Guillet, J. Macromol. Sci., Part A: Pure Appl. Chem., 1968, 2, 1441–1455.
8. Y. Ding, Y. Duan, F. Yang, Y. Xiong and S. Guo, Chem. Eng. J., 2023, 460, 141572.
9. M. A. Haq, Y. Su and D. Wang, Mater. Sci. Eng., C, 2017, 70, 842–855.
10. Y. Guan and Y. Zhang, Soft Matter, 2011, 7, 6375.
11. J.-Y. Sun, X. Zhao, W. R. K. Illeperuma, O. Chaudhuri, K. H. Oh, D. J. Mooney, J. J. Vlassak and Z. Suo, Nature, 2012, 489, 133–136.
12. J. P. Gong, Y. Katsuyama, T. Kurokawa and Y. Osada, Adv. Mater., 2003, 15, 1155–1158.
13. Md. A. Haque, T. Kurokawa and J. P. Gong, Polymer, 2012, 53, 1805–1822.
14. E. S. Dragan, Chem. Eng. J., 2014, 243, 572–590.
15. D. Myung, D. Waters, M. Wiseman, P. Duhamel, J. Noolandi, C. N. Ta and C. W. Frank, Polym. Adv. Technol., 2008, 19, 647–657.
16. P. Matricardi, C. Di Meo, T. Coviello, W. E. Hennink and F. Alhaique, Adv. Drug Delivery Rev., 2013, 65, 1172–1187.
17. H. Kawasaki, S. Sasaki and H. Maeda, Langmuir, 1998, 14, 773–776.
18. Y. Zhou, Y. Cai, X. Hu and Y. Long, J. Mater. Chem. A, 2014, 2, 13550–13555.
19. J. Narayanan, J.-Y. Xiong and X.-Y. Liu, J. Phys.: Conf. Ser., 2006, 28, 83–86.
20. V. Normand, D. L. Lootens, E. Amici, K. P. Plucknett and P. Aymard, Biomacromolecules, 2000, 1, 730–738.
21. S. Arnott, A. Fulmer, W. E. Scott, I. C. M. Dea, R. Moorhouse and D. A. Rees, J. Mol. Biol., 1974, 90, 269–284.
22. A. S. Hoffman, Adv. Drug Delivery Rev., 2012, 64, 18–23.
23. R. Baetens, B. P. Jelle and A. Gustavsen, Sol. Energy Mater. Sol. Cells, 2010, 94, 87–105.
24. M. N. Mustafa, M. A. A. Mohd Abdah, A. Numan, A. Moreno-Rangel, A. Radwan and M. Khalid, Renewable Sustainable Energy Rev., 2023, 181, 113355.
25. C. G. Granqvist, Thin Solid Films, 2014, 564, 1–38.
26. G. Xu, Y. Lu, X. Zhou, N. Moloto, J. Liu, S.-Z. Kure-Chu, T. Hihara, W. Zhang and Z. Sun, Mater. Horiz., 2024, 11, 4867–4884.
27. Y. Zhou, Y. Cai, X. Hu and Y. Long, J. Mater. Chem. A, 2014, 2, 13550–13555.
28. C. Zhao, X. Zhuang, C. He, X. Chen and X. Jing, Macromol. Rapid Commun., 2008, 29, 1810–1816.
29. E. Fujiwara, T. D. Cabral, M. Sato, H. Oku and C. M. B. Cordeiro, Sci. Rep., 2020, 10, 7035.
30. H. J. Lee, C. H. Jeong, D. Y. Kim, C. K. Choi and S. H. Lee, Appl. Sci., 2021, 11, 3379.
31. N. Pernodet, M. Maaloum and B. Tinland, Electrophoresis, 1997, 18, 55–58.
32. K. Roger, O. Atasi and B. Lalanne, Phys. Chem. Chem. Phys., 2023, 25, 994–998.
33. K. Lewandowska, Eur. Polym. J., 2005, 41, 55–64.
34. J. Joseph and E. D. Jemmis, J. Am. Chem. Soc., 2007, 129, 4620–4632.
35. Y. Hirashima, H. Sato and A. Suzuki, Macromolecules, 2005, 38, 9280–9286.
36. B. Jović, A. Nikolić and S. Petrović, J. Mol. Struct., 2013, 1044, 140–143.
37. H. Cheng, L. Shen and C. Wu, Macromolecules, 2006, 39, 2325–2329.
38. G. Giubertoni, M. Hilbers, F. Caporaletti, P. Laity, H. Groen, A. Van Der Weide, D. Bonn and S. Woutersen, J. Phys. Chem. Lett., 2023, 14, 940–946.
39. T. Munk, S. Baldursdottir, S. Hietala, T. Rades, M. Nuopponen, K. Kalliomäki, H. Tenhu, J. Rantanen and C. J. Strachan, Polymer, 2013, 54, 6947–6953.
40. A. Kroning, A. Furchner, D. Aulich, E. Bittrich, S. Rauch, P. Uhlmann, K.-J. Eichhorn, M. Seeber, I. Luzinov, S. M. Kilbey, B. S. Lokitz, S. Minko and K. Hinrichs, ACS Appl. Mater. Interfaces, 2015, 7, 12430–12439.
41. A. Martinez-Felipe, A. G. Cook, J. P. Abberley, R. Walker, J. M. D. Storey and C. T. Imrie, RSC Adv., 2016, 6, 108164.
42. A. B. McCoy, T. L. Guasco, C. M. Leavitt, S. G. Olesen and M. A. Johnson, Phys. Chem. Chem. Phys., 2012, 14, 7205.
43. Y. Zhu, L. Lin, J. Zeng, X. Tang, Y. Liu, P. Wu and C. Xu, Soft Matter, 2020, 16, 2199–2207.

---