Bidirectional optical response hydrogel with adjustable human comfort temperature for smart windows

Abstract

Smart windows are effective in reducing the energy consumption of air conditioning and lighting systems, while contributing to maintaining the comfort zone of temperature in the indoor environment. Currently used smart windows mainly rely on traditional single-phase thermochromic material in which only one abrupt optical change occurs during temperature changes, and their inherent characteristics may not be suited for a practical balance of energy saving and privacy protection. Here, we developed a novel bidirectional optically responsive smart window (BSW) with unique bidirectional optical response features by introducing sodium dodecyl sulfate (SDS)/potassium tartrate (PTH) micelles into PNIPAM hydrogel to form a composite hydrogel, which was encapsulated in two glass panels. The upper critical solution temperature (UCST) and lowest critical solution temperature (LCST) of the material can be individually adjusted and are capable of matching the human comfort zone of temperature. In addition, the smart window exhibits remarkable transparency (92.5%), visible light transmission ratio (T_lum = 91.31%), and excellent solar modulation (ΔT_sol,UCST = 76.34%, ΔT_sol,LCST = 76.75%). Moreover, it possesses selectivity in transmitting light in the infrared band of solar radiation and can complete the “transparent-opaque” transition in a very narrow temperature range (<1 °C). When at comfortable temperatures, the highly transparent smart windows facilitate interior light and appreciation of the view. At low temperatures, SDS/PTH micelles aggregate to form large micelles, blocking the transmission of light and protecting customer privacy. At high temperatures, PNIPAM can undergo a “sol–gel” transition, thus blocking incident solar radiation. Taken together, these proposed materials with bidirectional optical response characteristics would be harnessed as a promising platform for building energy conservation, anti-counterfeiting, information encryption, and temperature monitoring.

---

New concepts

This study presents a new concept for designing bidirectional optically responsive smart window (BSW) by introducing sodium dodecyl sulfate (SDS)/potassium tartrate (PTH) micelles into PNIPAM hydrogels to form a composite hydrogel, which is then encapsulated into two glass panels. Smart windows are effective in reducing energy consumption of air conditioning and lighting systems, while contributing to maintaining comfortable temperatures in the indoor environment. Herein, this work fabricated a novel bidirectional optically responsive smart window with unique bidirectional optical response features. The UCST and LCST of the material can be individually adjusted and are capable of matching the human comfort zone for temperature. In addition, the smart window exhibited remarkable transparency and visible light transmission ratio with excellent solar modulation. Additionally, it possesses selectivity in transmitting light in the infrared band of solar radiation and can complete the “transparent-opaque” transition in a very narrow temperature range. BSW was developed to block light transmission at night (low ambient temperature) or mid-day (high ambient temperatures), thus combining the two demands of privacy protection and energy saving. These proposed materials with bidirectional optical response characteristics can be harnessed as a promising platform to develop energy conservation, anti-counterfeiting, information encryption, and temperature monitoring.

Introduction

In the past few decades, energy consumption in buildings has become a major focus. With residential and commercial buildings accounting for 40% of global energy consumption, and as industrialization continues, energy consumption continues to exacerbate global warming trends.^1,2 Windows are regarded as the most energy-wasting part of the buildings. Interestingly, smart windows can reduce energy consumption by heating ventilation, and air conditioning (HVAC) systems and lighting through sunlight blocking or light-harvesting according to the external environment.³ Therefore, among numerous solutions proposed to reduce building energy consumption, smart windows are considered a reasonable and green strategy for building energy efficiency. Smart windows are generally known for their thermochromic, mechanochromic, and electrochromic properties which can dynamically realize optically modulation functions under external stimuli, such as heat, strain, and electricity.^4,5 Among these, thermochromic smart windows have been most widely studied for their rational stimulus, low-cost and zero-energy-consumption properties.^6,7

As warm-blooded animals, humans have a constant body temperature (37 °C) and a limited comfortable temperature range. The american society of heating, refrigerating, and air-conditioning engineers (ASHRAE) found that the standard temperature for human comfort was 22–27 °C. Numerous studies have proven that in warm or cold uncomfortable environments, human physiological and psychological performances can be adversely affected.^8–10 Therefore, smart windows must possess a reasonable thermochromic transition temperature (T_k), considering that they are primarily intended for use in human residential environments.^11,12 However, most of the existing smart windows are constructed from traditional thermochromic materials, such as vanadium dioxide (VO₂, T_k ≈ 68 °C), hydroxypropyl cellulose (HPC, T_k ≈ 54 °C), poly(N-vinyl caprolactam) (PNVCL, T_k ≈ 34 °C), poly (N-isopropyl acrylamide) (PNIPAM, T_k ≈ 32 °C), and so on.^13–16 The T_k of these materials is much higher compared to the comfortable temperature range (22–27 °C). More specifically, smart windows based on traditional thermochromic materials may not change their optical characteristics to modulate the incident solar radiation even if the outdoor temperature is already higher than the comfortable temperature.¹² This incident solar radiation continues to contribute heat to the indoors, which may cause the users to feel uncomfortable or consume more air conditioner cooling energy. Thus, a smart window with a reasonable thermochromic transition temperature has great advantages in practical applications. At the same time, considering the climate temperature differences in different regions and individual user preferences, the adjustable thermochromic transition temperature is also crucial to the thermochromic smart windows.^17,18 Furthermore, the majority of thermochromic smart windows can become opaque during hot midday (high ambient temperature) to block the incoming sunlight.^2,19 However, these smart windows will become transparent at night (when the ambient temperature is low), which may lead to a potential risk of privacy leakage.³ With the increasing emphasis on personal privacy protection, it is also considered a part of the critical function of next-generation smart windows. Accordingly, it is a significant challenge to break through the inherent limitations of traditional thermochromic materials and prepare energy-saving smart windows with a reasonable thermochromic transition temperature, while considering energy conservation and privacy protection.

Bidirectional optical response is defined as a material that exhibits high transparency within a specific temperature range and exhibits a decrease in bidirectional transparency as the temperature changes.²⁰ Different from traditional smart windows, which only show one abrupt optical change at the thermochromic transition temperature, smart windows with bidirectional optical response features are more suitable for meeting the two major demands of energy conservation and privacy protection; they can also achieve on-demand modulation of incident light. Here, a novel bidirectional optically responsive smart window (BSW) was designed by introducing sodium dodecyl sulfate (SDS)/potassium tartrate (PTH) micelles into the poly(N-isopropyl acrylamide) (PNIPAM) to constitute a composite hydrogel, which was sandwiched in-between two glass plates. PNIPAM is one of the most well-known thermo-sensitive polymers, which undergoes LCST phase transition in aqueous solution at ≈32 °C. Under the stimulation of temperatures, PNIPAM produces a phase transition between coil-globule structural, thus changing the transparency of PNIPAM-based hydrogel.²¹ Anionic surfactant sodium dodecyl sulfate (SDS) was chosen as a thermal response unit due to the formation of micelles adjusted by the addition of potassium tartrate (PTH), which aids in tuning the transmittance of light.²² This bidirectional optical response was based on the dynamic refractive index change induced by the change in particle size within the BSW. The two response temperatures of the BSW can be individually manipulated across a wide temperature range (UCST: 9.5–25.5 °C, LCST: 25.5–50 °C) by adjusting the micelle size and gel formation conditions, respectively, BSW allows flexible regulation of the temperature range to suit the user's demands and permit on-demand light modulation. BSW exhibits high transmittance (91.33% at 550 nm) and visible light transmission (T_lum = 91.31%) in its transparent state. It is worth noting that the “transparent-opaque” switch can be completed within a narrow temperature range. The transparency of the BSW shows a dramatic difference in transparency before and after the phase transition, with visible light modulation rates of 90.95% and 91.21% around UCST and LCST, respectively. The obtained smart windows can selectively allow visible light to pass through and block most infrared light and ultraviolet at room temperature, while blocking the light of almost all wavelengths at higher or lower temperatures. This BSW hydrogel material with precise response to environmental temperatures can have promising applications, such as information encryption,²³ temperature monitoring, and smart thermochromic windows.

Experimental section

Materials

Sodium dodecyl sulfate (SDS, 98.5%), N-isopropyl acrylamide (NIPAM, 98%), and potassium tartrate (PTH, 99%) were obtained from Macklin (China). Aluminum chloride hexahydrate (AlCl₃·6H₂O, 97%), N,N,N,N-tetramethylethylenediamine (TEMEDA, 98%), and ammonium persulfate (APS, 98%) were purchased from Sinopharm (China). All reagents obtained were of analytical grade and were used without any additional refining. Deionized water (DI, 18.3 MΩ) was utilized throughout the experimental process.

Preparation of PNIPAM hydrogel

The PNIPAM gel was prepared by free radical polymerization. In a typical process, magnetic stirring was used to dissolve 0.75 g of NIPAM in 30 mL of DI at room temperature to obtain a homogeneous aqueous solution. Then, an additional 0.08 g of APS and 20 μL of TEMEDA were injected into the above solution and the reaction was carried out at 40 °C for 4 h.

Preparation of BSW

The preparation procedure for BSW composite hydrogel was as follows: firstly, SDS (0.6–4.2 g) was dissolved in deionized water (30 mL) and stirred until the solution was transparent. Then, AlCl₃·6H₂O (0.75 g) and PTH (1.125–4.5 g) were added to the above SDS solution, and stirring was continued for 1 h to obtain the SDS/PTH composite micelle solution. At last, PNIPAM hydrogel and SDS/PTH micelle solution with a volume ratio of 1 : 1 were mixed at 20 °C to obtain BSW. A series of composite hydrogel BSW were prepared using different conditions that are summarized in Table S1 (ESI†).

Characterization

Using a UV-Vis-NIR spectrophotometer (UV-3600, Shimadzu, Japan), the transmittance measurements of BSW were conducted in the range of 200–2500 cm⁻¹, and the thickness of the samples was 10 mm. For the cycle test of transparency, the BSW was heated from 10 to 25 °C; it was further heated to 30 °C and then cooled back to 10 °C for 50 cycles. The samples were maintained for 3 min to attain equilibrium at the preset temperature before measurements. The required functional groups of PTH, SDS, SDS/PTH micelle, PNIPAM, and BSW hydrogel were identified using a Fourier transform infrared spectrometer (FTIR) (IS10, Nicolet) in the range of 4000 to 500 cm⁻¹. A high-intensity discharge lamp (CEL-S500L, China) was utilized as the illumination light for the demonstration of solar modulation on the top of the window. Field emission scanning electron microscopy (FESEM, JEOL JSM-7100F, Japan) was performed to study the microstructures of the BSW and pure PNIPAM hydrogels. Before being tested, both hydrogels were freeze-dried for 24 h and gold-sputtered to enhance their conductivity.

Dynamic light scattering (DLS) measurements were performed on Dynamic Nanostar (Zetasizer Nano ZS90, UK) to characterize the diameter of SDS/PTH micelles and BSW. The transmittance at 550 nm for every cycle was recorded and used for durability analysis. To quantify the overall optical transmittance performance of materials, the luminous transmittance (T_lum, 380–780 nm), (T_sol, 300–2500 nm), and ΔT was calculated as follows:

ΔT_{lum|sol|IR,UCST} = T_lum|sol|IR (25 °C) − T_lum|sol|IR (10 °C) (2)

ΔT_{lum|sol|IR,LCST} = T_lum|sol|IR (25 °C) − T_lum|sol|IR (30 °C) (3)

where T(λ) denotes spectral transmittance, and T is the testing temperature. φ_lum(λ) is the standard luminous efficiency function for photopic vision (the wavelength range of 380–780 nm), and φ_sol/IR(λ) is the solar irradiance spectrum at air mass 1.5 (AM, 1.5) when the sun stands at 37° above the horizon. Due to the bidirectional optical features of BSW, the optical modulation capabilities at UCST and LCST were quantified with ΔT_{lum/sol/IR,UCST}, and ΔT_{lum/sol/IR,LCST}, respectively. For all tests, temperature measurements and recordings were investigated by a hand-held thermometer (OMEGA HH306A, USA) with an accuracy of ±0.1 °C. The optical images of samples were collected by a digital camera (Panasonic Lumix DC-GF10, Japan).

Results and discussion

Inspired by the responses of warm-blooded animals to environmental temperature, in this study, we proposed a novel bidirectional smart optical material with both LCST- and UCST-type transitions and high transparency. The main design principles are illustrated in Fig. 1a. Briefly, PNIPAM hydrogel was obtained by free radical polymerization. The hydrogel appeared as a transparent liquid at room temperature. When heated above the T_k, the PNIPAM hydrogel turned into a white opaque solid.²⁴ The smart bidirectional, optical material BSW was obtained by introducing the micelles formed by PTH, SDS, and Al³⁺ into the PNIPAM hydrogel network. The as-prepared smart windows showed a unique bidirectional optical response effect with potential applications in smart windows, information camouflage, and other applications.²⁵Fig. 1b illustrates the light transmission effect of BSW and the corresponding working scheme. When the ambient temperature was between UCST and LCST, the incident light could be permitted to pass through the thermally responsive BSW material, and the smart window material presented itself as optically transparent. At low temperatures, hydrophobic large-size micelles were formed by H-bonds aggregation interactions between PTH and SDS, which led to the smart window adopting an opaque state. When the temperature was increased to LCST or higher, the PNIPAM inside the BSW underwent a reversible “sol–gel” transition, which caused the refractive index mismatch between large-size PNIPAM microspheres and water, leading to intensive light scattering.²⁶ The smart window then changed from the transparent to the opaque state.^14,27 It is worth noting that the T_k of most traditional smart window materials (PNIPAM, PNVCL, HPC, and VO₂) was above the comfort temperature range (22–27 °C) and was difficult to regulate. Furthermore, the UCST of the obtained BSW could be precisely adjusted by varying the amount of PTH during the preparation of hydrogel. Moreover, the additional amount of SDS in the hydrogel can affect its LCST. According to this characteristic of the BSW hydrogel, the transparent temperature range of the smart window can be controlled by adjusting the content of SDS and PTH to meet the needs of different users (Fig. 1c).

Fig. 1 (a) Schematic diagram of the fabrication process of the BSW. (b) Schematic illustrations and corresponding photos of BSW at different temperatures. (c) Adjustable range of thermochromic transition temperature for BSW. (d) FT-IR spectra of SDS, PTH, SDS/PTH, PNIPAM, and BSW.

The chemical structures of SDS, PTH, SDS/PTH micelle, PNIPAM, and BSW are illustrated in Fig. 1d. It can be seen that the peak at 1218 cm⁻¹ for SDS was associated with the S=O group; the carboxylate asymmetric absorption and the hydroxyl bending vibration absorption peaks for PTH appeared at 1606 cm⁻¹ and 1394 cm⁻¹. For the SDS/PTH curve, the absorption peak of the S=O group underwent a left shift from 1218 cm⁻¹ to 1228 cm⁻¹, indicating the local environment of SDS was changed. Simultaneously, the hydroxyl vibration peak appearing at 3421 cm⁻¹ and 1394 cm⁻¹ of PTH were shifted to low wave numbers at 3415 cm⁻¹ and 1386 cm⁻¹, respectively, proving the formation of the H-bonds between SDS and PTH within the SDS/PTH composite micelles.²⁸ PNIPAM hydrogel displayed three prominent vibrations at 1651 cm⁻¹, 1542 cm⁻¹, and 3301 cm⁻¹, corresponding to the stretching C=O, bending N–H, and stretching N–H vibrations, respectively. In the infrared curve of BSW, the peaks at 1130 cm⁻¹ (from SDS), 1387 cm⁻¹(from PTH), 1646 cm⁻¹, and 3268 cm⁻¹ (from PNIPAM) were observed. In addition, the hydroxyl vibration peak moved to 3435 cm⁻¹ and widened, indicating that the presence of SDS/PTH micelles and PNIPAM in BSW is different from that in the bulk material; there was extensive hydrogen bonding between the as-prepared BSW hydrogels.²⁹

Fig. 2 shows the microstructures of BSW and PNIPAM hydrogel after freeze-drying. It can be seen that the PNIPAM hydrogels were mainly irregular sheets with rough surfaces, and the network structures were not observed on the surface of the hydrogels (Fig. 2a and b). In contrast, the surface of BSW was relatively smooth and distributed with many pores (Fig. 2c). In addition, the presence of tiny bumps on the surface and within the pores can be observed in the high-magnification images of BSW (Fig. 2d). This may be because the micelles formed by SDS aggregation inside BSW were embedded in the network structure of PNIPAM hydrogel and occupy a certain space. After freeze-drying, the size of the micelles became smaller due to dehydration, and vacancies were formed on the surface of the dried BSW hydrogel.

Fig. 2 The microstructure of the PNIPAM (a) and (b) and BSW (c) and (d).

To explore the effect of SDS and PTH contents on the thermo-responsive behaviors, a series of modified PNIPAM hydrogels were prepared with PTH and SDS contents in the range of 1.125 to 4.5 g and 0.6 to 4.2 g, respectively. To dynamically satisfy energy-saving and personal privacy purposes in different areas and also consider the needs of different customers for comfortable temperatures, it is crucial to modulate the LCST and UCST of the BSW by conditioning the phase separation in the smart window. The amount of SDS was fixed to 0.6 g, while the variations in UCST and LCST of BSW with respect to the PTH content added are presented in Fig. 3a. Notably, the UCST of the smart window increased by 16 °C as the PTH content increased from 1.125 to 4.5 g, whereas the LCST only varied slightly by less than 1.5 °C. At low temperatures, the large-sized micelles formed between SDS and PTH through hydrogen bonding could scatter incident light, while when heated above UCST, the breakage of hydrogen bonding led to the disintegration of large-sized micelles. This led to the disappearance of the scattering effect and BSW transformed into transparent material. The change in micelle size driven by the hydrogen bonding was responsible for the optical abruptness of BSW at UCST. When the PTH addition amount increased from 1.125 to 4.5 g, the hydrogen bonding density within the smart windows increased. Therefore, to disintegrate the larger micelles, more thermal energy was required to disrupt the hydrogen bonding interactions in the micelles,^30,31 which led to the increase in the UCST from 9.5 to 25.5 °C. Meanwhile, the LCST of BSW decreased from 27.5to 26 °C, mainly due to the decrease in the range of electrostatic repulsion forces caused by the increase in the amount of electrolyte PTH, making it easier to form particle aggregates based on polymer–polymer interactions.^32,33

Fig. 3 UCST and LCST of BSW with different amounts of (a) PTH and (b) SDS.

In addition, we further evaluated how to control the LCST of the smart windows. The effect of SDS amount on the response temperature of BSW with a PTH amount of 1.5 g is shown in Fig. 3b. The LCST of the BSW gradually decreased from 27.5 to 25.5 °C as the SDS content increased from 0.6 to 2.4 g. However, as the SDS content continued to rise, the LCST increased gradually and reached 50 °C, when the additional amount of SDS was 4.2 g. This phenomenon was attributed to the fact that at lower SDS content, the interaction between PNIPAM and SDS was achieved through the formation of polymer-surfactant aggregates.³⁴ Na⁺ released from the hydrolysis of SDS affected the aggregation of PNIPAM by reducing the range of electrostatic repulsion, resulting in a decrease in the LCST of BSW.^19,32 However, with higher SDS content added, free micelles could form that antagonized the formation of polymer-bound aggregates due to a further increase in the additional amount of surfactant SDS, which led to the elevation of the LCST of BSW.³⁴ In addition, the amount of SDS also influenced the UCST of BSW. It was found that the different SDS content had little effect on the UCST of BSW and that the UCST of BSW increased only from 14 to 16.9 °C. PTH and SDS demonstrated desirable thermochromic transition temperature which affected the independence in smart windows. This valuable characteristic could be used to accurately tune both UCST and LCST of smart windows, which was important for the bidirectional optically responsive smart windows. The BSW samples prepared with PTH and SDS amounts of 1.5 and 0.6 g, respectively, have the advantages of suitable temperature range (UCST = 14 °C, LCST = 26.9 °C) for most regions of China and convenient differentiation and temperature controllability, making it ideal for further discussion in energy-efficient smart windows to regulate thermochromic performances.

Furthermore, to explore the optical transmittance characteristics of BSW, the transmittance of BSW in both transparency (25 °C) and opacity (10 and 30 °C) at a wavelength from 200 to 2500 nm are shown in Fig. 4a. When BSW was present at a high temperature (30 °C), it was highly shielded from the whole region of incident light, preventing the outdoor heat from entering, which was beneficial for heat insulation and energy-saving. While at low temperatures (10 °C), the BSW was opaque to protect indoor privacy. Simultaneously, indoor light and heat were not allowed to be transmitted to the outside through opaque smart windows, reducing the loss of indoor heat and realizing the energy-saving function. In the transparent state, the water-rich BSW exhibited high transmittance to visible light and due to the strong absorption of infrared radiation by water,³⁵ it could selectively block the infrared part of the incident light, while allowing ultraviolet and visible light through, which is beneficial for indoor illumination and insulation of the heat introduced by infrared light. Therefore, the room temperature could be stabilized at a comfortable range. Moreover, the transmittance of BSW in the visible region (380–780 nm) at different temperatures was explored (Fig. 4b). It is worth noting that the smart window exhibited high transparency (91.33% at 550 nm) at 25 °C in the visible range, which was quite close to the transparency of deionized water (92.53% at 550 nm). High transparency was essential for the windows to fulfil the function of natural lighting and viewing, which is a prerequisite for the applications of smart windows. Meanwhile, the light transmission of BSW was 0.09% at 30 °C and 0.35% at 10 °C. The dramatic contrast between the transparency of BSW in the transparent and opaque state enables the modulation of the incident light effectively.

Fig. 4 (a) The UV-Vis-NIR spectroscopy (200–2500 nm) and (b) the UV-Vis spectroscopy (200–800 nm) of BSW in transparent (25 °C) and opaque (10 and 30 °C). (c) BSW at different phase transitions corresponding to the light transmittance modulations (ΔT_lum, (380–780 nm), ΔT_IR (780–2500 nm), and ΔT_sol, (300–2500 nm)). (d)–(f) Optical photos and (g) transmittance spectra of smart windows of various thicknesses (2, 5, and 10 mm) at different temperatures (10, 25, and 30 °C). (h) Comparison of different thicknesses (2, 5, and 10 mm) of BSW in terms of light transmittance (T_lum) and optical modulation capability (ΔT_lum,UCST and ΔT_lum,LCST). (i) At 550 nm, reproducibility in transmittance of the BSW (550 nm) at 10, 25, and 30 °C.

To evaluate the modulation capability of BSW for different wavelengths of the incident light, the modulation performance of the BSW for visible light (380–780 nm), sunlight (300–2500 nm), and infrared light (780–2500 nm) were analyzed and calculated, respectively. Due to the special bidirectional optical response features of the BSW, its modulation rates at both UCST and LCST were characterized. The visible light modulation rates of the BSW were 90.95 and 91.21% at UCST and LCST, respectively, while the sunlight modulation rates were 76.34 and 76.75% at UCST and LCST, respectively. The as-prepared smart windows exhibited excellent visible and solar light modulation performances under both high and low-temperature conditions. In addition, under UCST and LCST conditions, the IR light modulation rates of the BSWs were 50.85 and 51.48%, respectively. Moderate ΔT_IR could be attributed to the selective blocking of IR light by the BSW, while other smart windows allow IR light from solar radiation to pass through (Fig. 4c). The as-prepared BSW has an excellent blocking effect on IR, which accounts for 53% of the total solar energy (UV 3%; visible light 44%).³ This is crucial for reducing the energy consumption in HVAC. In summary, the prepared smart windows can achieve effective optical modulation, which is vital for the energy-saving and privacy-protection performance of next-generation smart windows.

To display the influence of different thicknesses on the transparency and optical modulation ability, BSWs with different thicknesses (2, 5, and 10 mm) were prepared. The optical photos for different thickness samples at 10, 25, and 30 °C are shown in Fig. 4d–f. At 25 °C, all samples were transparent and the plant underneath the samples could be visible, the luminous transmittance was not affected by thickness, showing remarkably high transparency. When the temperature (30 °C) was above LSCT, all the samples turned opaque, and the under-plant became invisible. When the temperature (10 °C) was lower than UCST, the 2 mm sample became translucent, and the plants behind could be seen faintly. In contrast, the 5 mm and 10 mm samples were opaque and the plants behind them could not be observed. The optical photographs were consistent with the spectra-temperature dependence.

The detailed transmittance of BSW with different thicknesses in both transparent (25 °C) and opaque (10 and 30 °C) states at 200–800 nm are shown in Fig. 4g. All the samples showed high transmittance in the visible range at 25 °C (the transmittances of 2, 5, and 10 mm samples were 92.50, 92.10, and 91.33%, respectively). The transmittance of the samples gradually lowered with the increase in the sample thickness. Meanwhile, the samples of three different thicknesses all showed excellent light-shielding performance at high or low ambient temperatures. At 10 °C, the transmittance of the 2 mm sample at 550 nm was 3.4%, which gradually decreased to 1.9 and 0.4% with the gradual increase in the thickness to 5 and 10 mm, respectively. At 30 °C, the transmittance at 550 nm decreased from 1.3 to 0.1% and 0.03% with the increase in thickness. It is noteworthy that in the transparent state, the samples with different thicknesses exhibited higher light transmission in the visible range. To evaluate the ability of smart windows with different thicknesses to modulate illumination, the visible light modulation performance of BSW was tested and analyzed (Fig. 4h). The increase in thickness decreased the integrated transmittance of the sample, and as the BSW thickness increased from 2 to 5 mm and then to 10 mm, the T_lum decreased from 92.54 to 92.51% and, finally, to 91.31%. Although the 2 mm sample exhibited the highest T_lum, the other thickness samples had lower ΔT_lum (ΔT_lum,UCST = 89.59%, ΔT_lum,LCST = 91.33%). Moreover, the 10 mm sample had the highest ΔT_lum,UCST (90.95%). Five millimetres of the sample exhibited the highest ΔT_lum,LCST (92.22%), while its ΔT_lum,UCST (90.32%) was quite close to that of the 10 mm sample (ΔT_lum,LCST = 91.21%).

The cycle stability of BSW was critical for its application potential. The transmittance was recorded at 550 nm per cycle in transparent (25 °C) and opaque (10 and 30 °C) states (Fig. 4i). The transmittances were about 89.74 and 91.45% in the transparent states, whereas they fluctuated at 0.63% in the opaque states. After 50 transition cycles, the transmittance of BSW did not show significant attenuation in either the cold or hot states. The transparency change of BSW can be cycled by heating and cooling. After repeated measurements of the transmittance during the cycling, an obvious and stable change in transmissivity between the two states could be observed. The cycle measurement indicated that BSW is reliably reversible and cyclically stable, which has a positive impact on future practical applications.

An optical transmittance test was conducted on BSW from 4 to 41 °C at 550 nm to study the temperature dependence of the optical properties. Fig. 5a illustrates the relationship between temperature and BSW transmittance. In the range from 4 to 41 °C, the transmittance of BSW exhibited a huge difference in transparency. It can be noticed that the initial transmittance was relatively low at 1% as the measuring temperature increased. Remarkably, the transmittance of BSW was >90% in the range between 14 and 26.9 °C. With increasing temperature, the transmittance decreased sharply, and BSW exhibited an optical opacity when the temperature was higher than 26.9 °C. It was worth noting that BSW could complete the “transparent-opaque” transition in a very narrow temperature range of less than <1 °C. In contrast to BSW, the micelles formed by SDS, PTH, and Al³⁺ exhibited opacity throughout the temperature change (4–41 °C), while pure PNIPAM showed only one abrupt optical change at 32 °C (Fig. S1, ESI†).

Fig. 5 (a) Temperature-dependent transmittance of BSW at the wavelength of 550 nm; the insets show the photos of BSW at 10, 25, and 30 °C, respectively. The particle size distribution of BSW at different temperatures, (b) 4–20 °C, (c) 24–30 °C. (d) Digital photographs of BSW at different temperatures.

To further assess the mechanism of the bidirectional smart windows transparency, the average particle size and size distribution of samples at different temperatures were determined by DLS measurements. The micelles formed by SDS, PTH, and Al³⁺ were relatively stable and difficult to destroy even upon heating. When heated from 5 to 45 °C, the particle size of the micelles remained at ∼1000 nm with moderate variability (Fig. S2a, ESI†). Since Al³⁺ bonded to the anionic head group of SDS, the resulting metal complexes were very stable and difficult to destroy even by heating.¹² Below the phase transition temperature, the particle size of PNIPAM was approximately 10 nm. However, when the temperature was higher than the phase transition temperature, the size of PNIPAM sharply increased to about 1000 nm (Fig. S2b, ESI†). The particle sizes of BSW varied with the measuring temperature. When the temperature increased from 4.0 to 12.0 °C, the semicircular diameter of the BSW dramatically decreased from more than 2200 nm to around 800 nm. However, the sizes of the particles were still above the wavelength range of visible light (380–780 nm) at this temperature range, which could scatter the majority of the incident light, and the smart window also performed in an opaque state. In addition, a rapid narrowing of the particle size distribution range of the smart window could be observed during this process. When heated to 16.0 °C, the temperature of the BSW sample became higher than that of the LCST (14.0 °C); the particle size of 350 nm was lower than the visible wavelength, and the BSW sample transformed to transparent (Fig. 5b). As the temperature continued to 26 °C, the sample retained a particle size of less than 100 nm and the smart window exhibited high transparency. Notably, as the temperature was higher than the LCST of BSW (26.9 °C), the particle size increased rapidly due to the transition of PNIPAM molecular chains within the smart window from the coil phase to the globular phase.^36,37 At 28 °C, the particle diameters of the samples were more than 1000 nm, and the smart windows turned from transparent to opaque (Fig. 5c). The variation in the light transmission of the BSW samples at different temperatures was directly reflected in the corresponding digital photographs (Fig. 5d). At 16–26 °C, BSW samples were almost transparent and the flower-like pattern underneath could be observed, whereas at other temperatures, the samples were opaque and the flower-like pattern underneath was obscured. This phenomenon was attributed to the fact that the sample became transparent at specific temperatures, when the size of the particles inside the BSW, was smaller than the wavelength of the incident visible light. On the contrary, when the temperature was above or below this temperature range, the large-sized particles inside the smart windows reflected significantly and had scattering effects on the incident light, thus transitioning the BSW to an opaque state. In summary, temperature could influence the change in the size of the particles inside the BSW, which was reflected in the macroscopic changes in the transparency of the BSW.

The energy-saving performance test was designed to study the solar modulation and energy-saving performance of BSW. As illustrated in Fig. 6a, BSW of 5 mm thickness was sandwiched between two pieces of glass sheets and sealed, which were then installed in a glasshouse (15 cm × 15 cm × 25 cm) to investigate the temperature change. A plain glass window with the same thickness was installed to serve as a control. In addition, an HID light source was used to simulate sunlight to simultaneously illuminate two model houses of BSW and bare glass windows, respectively, and a handheld thermometer was used to monitor the air temperature inside the model house. Fig. 6b shows the temperature changes on the window surface and indoor air during 200 s of sunlight exposure. It was obvious that the interior temperature of the BSW-equipped model house was significantly lower than that of the model house with bare glass windows. At an exposure time of 200 s, the interior temperature of both model houses increased rapidly and then stabilized (61.1 and 37.9 °C for the glass model house and the BSW model house, respectively). The surface temperature of the house with smart windows was 5.5 °C lower than that of the ordinary glass house. Excitingly, the interior air temperature of the smart window house was 23.2 °C lower than that of the ordinary glass house, indicating that BSW hydrogel has excellent insulation ability. The designed BSW hydrogel was suitable for smart windows to keep the indoor temperature at a comfortable state, thus reducing the consumption of heating, ventilation, and air conditioning systems. It effectively achieved the thermal regulation function of energy-saving buildings (Table S2, ESI†).

Fig. 6 (a) Schematic diagram of the energy-saving performance test. (b) Time-dependent temperature evolution of ordinary windows and BSW surface and indoor air under xenon lamp exposure. (c) Schematic diagram of the multilayer structured encrypted label based on BSW. (d) Digital photographs of the encrypted label of the multilayered structure based on BSW at different temperatures. (e) Photographs of the bidirectional optical response smart material for monitoring temperature.

Furthermore, based on this bidirectional optical response feature of BSW, the prospect of its use in the field of information security is exciting. We designed a multilayer structure that encapsulated the BSW and QR code into a sandwich-type encrypted label to realize the potential of BSW's functionality as an information encryption device. This device was characterized by a simple decryption method as well as high cost-effectiveness. The schematic diagram of the multilayer structure with the encrypted label is shown in Fig. 6c. When the temperature of the label was in a specific temperature range, the QR code was visible through the BSW, and it could be easily scanned and identified by smartphones or other devices. When the temperature was out of this temperature range, the QR code beneath was hidden and unintelligible, so no information could be scanned and recognized. Such encrypted label can only successfully identify the information within the correct temperature range (Fig. 6d). From the above analysis (Fig. 3), a series of BSW hydrogels can be obtained by adjusting the components of the hydrogels, and thus the temperature range of the encrypted labels can be regulated, which endows the encryption labels with better security and flexibility. Moreover, a smart temperature monitoring device was designed (Fig. 6e), whose schematic diagram is shown in Fig. S3 (ESI†). Only when the temperature is within the preset temperature range, the device is transparent and shows the “safety” label located below. Outside the temperature range, the label is not visible, which alerts that the changes in room temperatures should be noticed by the relevant operator immediately. In combination, such a design can be installed easily and massively in a variety of locations and shows good potential applications in special workplaces, such as greenhouses, farms, warehouses, etc.

Conclusions

In summary, a novel thermally responsive smart window was designed based on a combination of common and easily accessible PNIPAM and SDS/PTH micelles. The BSW exhibits a unique bidirectional optical response that makes it transparent only in a specific temperature range and opaque when the temperature is outside the transparent temperature range. The BSW with a thickness of 10 mm had high transparency (92.5%) and visible light transmission ratio (T_lum = 91.31%) at room temperature. As the temperature deviated from room temperature (heating or cooling), the bidirectional transparency decreased, and BSW exhibited an excellent ability to modulate visible light (ΔT_lum,UCST of 90.95% and ΔT_lum,LCST of 91.21%) and solar energy (ΔT_sol,UCST of 76.34% and ΔT_sol,LCST of 76.75%). Meanwhile, BSW showed a special wavelength-selective transmission effect, blocking most of the infrared radiation in the transparent state without affecting the visible transmittance. Moreover, the transparent temperature range of BSW can be readily regulated by adjusting the hydrogel composition to suit various climatic conditions or user preferences. Most critically, its UCST and LCST can be independently adjusted and matched to the human body's comfort zone of temperature. In the indoor thermal test, BSW smart windows achieved a temperature reduction of 23.2 °C compared to the normal glass panel, indicating its great potential for use in energy-efficient buildings. In addition, BSW's unique optical response modes, individually adjustable response temperature, high transparency, and optical modulation capabilities allow it to create new opportunities for smart optical materials, such as smart windows, information encryption, temperature monitoring, optical switches, etc.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was funded by the open project of Hubei Provincial Engineering Center of Performance Chemicals (GCZX-2022-03), Hubei University, China.

Notes and references

1. Y. Niu, Y. Zhou, D. Du, X. Ouyang, Z. Yang, W. Lan, F. Fan, S. Zhao, Y. Liu and S. Chen, Adv. Sci., 2022, 9, 2105184.
2. Y. Ding, Y. Duan, F. Yang, Y. Xiong and S. Guo, Chem. Eng. J., 2023, 460, 141572.
3. G. Xu, H. Xia, P. Chen, W. She, H. Zhang, J. Ma, Q. Ruan, W. Zhang and Z. Sun, Adv. Funct. Mater., 2022, 32, 2109597.
4. D. Cao, C. Xu, W. Lu, C. Qin and S. Cheng, Solar Rrl, 2018, 2, 1700219.
5. X. Zou, H. Ji, Y. Zhao, M. Lu, J. Tao, P. Tang, B. Liu, X. Yu and Y. Mao, Nanomaterials, 2021, 11, 3335.
6. S. Wang, Y. Zhou, T. Jiang, R. Yang, G. Tan and Y. Long, Nano Energy, 2021, 89, 106440.
7. R. Baetens, B. P. Jelle and A. Gustavsen, Sol. Energ. Mater. Sol. C., 2010, 94, 87–105.
8. K. Janprom, W. Permpoonsinsup and S. Wangnipparnto, J. Control Sci. Eng., 2020, 2020, 1–13.
9. O. Deschenes, Energ. Econ., 2014, 46, 606–619.
10. Z. J. Schlader, S. E. Simmons, S. R. Stannard and T. Mündel, Physiol. Behav., 2011, 103, 217–224.
11. Y. Zhou, F. Fan, Y. Liu, S. Zhao, Q. Xu, S. Wang, D. Luo and Y. Long, Nano Energ., 2021, 90, 106613.
12. Z. Yu, Y. Yang, C. Shen, L. Mao, C. Cui, Z. Chen and Y. Zhang, J. Mater. Chem. C, 2023, 11, 583–592.
13. Y. Q. Feng, M. L. Lv, M. Yang, W. X. Ma, G. Zhang, Y. Z. Yu, Y. Q. Wu, H. B. Li, D. Z. Liu and Y. S. Yang, Molecules, 2022, 27, 1638.
14. R. L. Sala, R. H. Gonçalves, E. R. Camargo and E. R. Leite, Sol. Energ. Mat. Sol. C., 2018, 186, 266–272.
15. J. Qian, B. Li, S. Tian, B. Liu and X. Zhao, Appl. Surf. Sci., 2022, 605, 154680.
16. X. Liu and Y. Wu, Sol. Energy, 2021, 220, 722–734.
17. Z. Sun, X. Xie, W. Xu, K. Chen, Y. Liu, X. Chu, Y. Niu, S. Zhang and C. Ren, ACS Sustainable Chem. Eng., 2021, 9, 12949–12959.
18. J. Tian, H. Peng, X. Du, H. Wang, X. Cheng and Z. Du, Prog. Org. Coat., 2021, 157, 106287.
19. Y. Zhang, S. Furyk, D. E. Bergbreiter and P. S. Cremer, J. Am. Chem. Soc., 2005, 127, 14505–14510.
20. J. Li, X. Lu, Y. Zhang, X. Wen, K. Yao, F. Cheng, D. Wang, X. Ke, H. Zeng and S. Yang, Small, 2022, 18, 2201322.
21. M. Podewitz, Y. Wang, P. K. Quoika, J. R. Loeffler, M. Schauperl and K. R. Liedl, J. Phys. Chem. B, 2019, 123, 8838–8847.
22. H. Khan, J. M. Seddon, R. V. Law, N. J. Brooks, E. Robles, J. T. Cabral and O. Ces, J. Colloid Interf. Sci., 2019, 538, 75–82.
23. J. Guo, S. Wu, Y. Wang, J. Huang, H. Xie and S. Zhou, Mater. Horiz., 2022, 9, 3039–3047.
24. H. Zhang, J. Liu, F. Shi, T. Li, H. Zhang, D. Yang, Y. Li, Z. Tian and N. Zhou, Chem. Eng. J., 2022, 431, 133353.
25. L. Feng, G. Mi, Q. Wang, M. Guo, J. Hao, Z. Li, J. Yang, G. Qin, G. Sun and Q. Chen, J. Power Sources, 2023, 580, 233453.
26. G. Li, J. Chen, Z. Yan, S. Wang, Y. Ke, W. Luo, H. Ma, J. Guan and Y. Long, Mater. Horiz., 2023, 10, 2004.
27. Y. Zhou, S. Wang, J. Peng, Y. Tan, C. Li, F. Y. C. Boey and Y. Long, Joule, 2020, 4, 2458–2474.
28. Y. Gao, K. Li, X. Ren and G. Gao, Chem. – Eur. J., 2023, e202302147.
29. H. M. Said, S. G. Abd Alla and A. W. M. El-Naggar, React. Funct. Polym., 2004, 61, 397–404.
30. K. Li, S. Meng, S. Xia, X. Ren and G. Gao, ACS Appl. Mater. Interfaces, 2020, 12, 42193–42201.
31. K. Li, H. Gao, X. Ren and G. Gao, ACS Sustainable Chem. Eng., 2019, 7, 15036–15043.
32. B. A. Humphreys, E. J. Wanless and G. B. Webber, J. Colloids Interf. Sci., 2018, 516, 153–161.
33. M. Irie, Y. Misumi and T. Tanaka, Polymer, 1993, 34, 4531–4535.
34. Y. Mylonas, G. Staikos and P. Lianos, Langmuir, 1999, 15, 7172–7175.
35. A. Barth, Nature, 2020, 577, 34–35.
36. J. Tian, J. Gu, H. Peng, H. Wang, Z. Du, X. Cheng and X. Du, Compos. Part A: Appl. S., 2021, 149, 106538.
37. B. Sun, Y. Lin, P. Wu and H. W. Siesler, Macromolecules, 2008, 41, 1512–1520.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3mh01376f

---