Chemical Engineering Journal 463 (2023) 142390 Contents lists available at ScienceDirect Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej Multistimuli-responsive smart windows based on paraffin-polymer composites Jaume Ramon Otaegui a, b, Daniel Ruiz-Molina b, Jordi Hernando a, *, Claudio Roscini b, * a Departament de Química, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Spain b Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and The Barcelona Institute of Science and Technology (BIST), Campus UAB, Bellaterra, Barcelona 08193, Spain A R T I C L E I N F O A B S T R A C T Keywords: Despite their potential to reduce energy consumption in buildings and increase user’s comfort, the general Smart windows application of smart windows is currently hampered by economic and technological factors. To overcome these Thermochromism limitations, herein we report a new strategy for the fabrication of thermoresponsive smart windows based on Photochromism polymer films loaded with paraffin nanoparticles. Proper selection of these components allows refractive index Electrochromism Energy saving matching when the paraffin nanoparticles are in their solid state - i.e., high visible and near-infrared light Photothermal heating transparency -, which is lost after melting, thereby resulting in pronounced light scattering - i.e., opacity. As a result, these films show ample modulation of sunlight transmittance upon heating, while additionally exhibiting other advantageous features: facile preparation from low-cost materials; high photostability, flexibility and scalability; and fine tunability of their thermal response. Moreover, by deposition onto transparent electrodes or incorporation of nanostructured photothermal agents, the solar transmittance modulation of our paraffin- polymer composites can also be triggered with low voltages - i.e., to warrant user’s control - or sunlight ab­ sorption - i.e., to improve self-adaptation to ambient conditions. Actually, films with combined thermal and photothermal activity allow high sunlight transmission in cold weather while efficiently reducing solar heat gain indoors in sunny hot days. 1. Introduction response to external stimuli is rising much attention [3-8]. One of the most appealing strategies toward SWs is based on chro­ Buildings account for ca. 40% of the primary energy consumption in mogenic materials that interconvert between noncolored and colored developed countries, about half of which is related to heating, ventila­ states – i.e., they reduce sunlight transmittance through the absorption tion and air conditioning systems (HVAC) [1,2]. One of the main ap­ of visible and near-infrared (NIR) photons when activated [3,8,12]. This proaches proposed to reduce building’s energy demand and its is the case of photochromic [9,13-15], thermochromic [16,17] and detrimental impact on CO2 emissions are smart windows (SWs) – i.e., electrochromic [18,19] windows responding to radiation, temperature glazing elements that adapt the transmittance of solar radiation to variation and electric currents, respectively. Although chromogenic SWs weather conditions [3-8]. As a result, SWs allow dynamic regulation of are already in the market, their widespread application is limited light and energy fluxes between outdoors and indoors: they are highly because of their elevated price and relatively low fatigue resistance, two transparent in winter and cloudy days to provide appropriate illumi­ drawbacks that are mainly ascribed to the use of stimuli-responsive dyes nation and subsequent solar heat build-up indoors; by contrast, SWs and pigments [3,12]. block sunlight entrance when warm, thus lowering the need of air Alternatively, dye-free SWs can be prepared from organic (e.g., conditioning. Aside from energy saving, smart glasses also increase hydro- [20-27] and ionogels [28-30], liquid crystals [31-37], co­ user’s comfort and privacy and can find application in other areas (e.g., polymers [38,39]) and inorganic (e.g., vanadium oxide [40,41]) sub­ in the ophthalmic [9,10] and automotive industry or in greenhouses to stances that undergo a phase transition between transparent and opaque increase horticulture productivity [11]). Accordingly, the development states upon application of external stimuli [42]. Modulation of light of materials capable of reversibly varying sunlight transmission in transmission in these systems is not accomplished by photoabsorption; * Corresponding authors. E-mail addresses: jordi.hernando@uab.cat (J. Hernando), claudio.roscini@icn2.cat (C. Roscini). https://doi.org/10.1016/j.cej.2023.142390 Received 16 December 2022; Received in revised form 21 February 2023; Accepted 8 March 2023 Available online 13 March 2023 1385-8947/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by- nc/4.0/). J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 instead, it takes place via changes in light reflectivity that occur when 2. Results and discussion the internal structure of the material varies due to phase transition – e.g., from a fully hydrated, transparent polymer hydrogel to a phase- 2.1. Thermoinduced transparency modulation segregated polymer-water mixture that strongly scatters light [20-27]. Though this feature largely increases the photorobustness of the To validate our strategy towards TR-SWs, we chose eicosane (EC), a resulting SW materials, they still suffer from a variety of limitations: common and low-cost C20 paraffin, for the preparation of light- high cost (e.g., for SWs based on liquid crystals), complex architectures scattering SLPs. As shown in Fig. 2a, EC exhibits an abrupt decrement to prevent solvent evaporation (e.g., sealed cells for polymer hydrogels), in RI upon solid-to-liquid transition: it goes from ca. 1.51 at room short lifetime, modest light transmission modulation, low scalability temperature to less than 1.44 when melting at TmEC = 36.5 ◦ C [49], a and/or poor adaptability (e.g., of transition temperature) to different behavior that can be ascribed to the change in molecular packing from weather conditions [20,42]. Moreover, most of these systems are only its crystalline form to the amorphous melt. By contrast, poly(vinyl responsive to a single type of stimulus, making them only capable of alcohol) (PVA), a regular hydrophilic film-forming polymer, presents a either user-controlled (by voltage application) or spontaneous (by rather constant RI over all this temperature range - from 1.510 at 25 ◦ C ambient light or temperature) modulation of light transmission. to 1.507 at 40 ◦ C –, where it does not suffer any phase change Consequently, novel multistimuli-responsive, dye-free technologies are (TgPVA =56–58 ◦ C) [50]. More importantly, the RI of PVA films lies very required to attain cost-effective, durable, scalable and tunable SWs. close to the value of solid EC (Fig. 2a), while they are immiscible and Herein we hypothesize that all these characteristics can be accom­ therefore should preserve their respective refractive indices and thermal plished at once by means of organic composites made of (a) a trans­ responses after integration in the final composite. As a result, refractive parent polymer matrix, where (b) paraffin particles (solid lipid particles, index matching should be selectively expected for EC-PVA mixtures SLPs) are dispersed that abruptly and reversibly change their refractive below TmEC and it must be lost after paraffin melting, thus enabling index (RI) upon melting. By properly selecting the nature of these two modulation of their light transparency properties. components, good RI matching could be ensured at room temperature To experimentally realize this behavior, we prepared EC-PVA com­ when SLPs are in their solid state. As a result, light scattering at the posite films, which required previous structuration of the paraffin into polymer-particle interfaces must be minimized and, therefore, the SLPs. For this we used an emulsion-cooling methodology previously composite materials must present high transmission of solar radiation. developed by us [45], which led to the formation of quasi-spherical In contrast, when heating above the melting temperature (Tm) of the particles. For comparison purposes, SLPs with variable diameters SLPs, a variation of their RI is expected, which should produce an in­ within the 25–2000 nm range were obtained (EC@SLP, Fig. 2b, Table S1 crease of light scattering and, consequently, turn the composite films and Fig. S1): 38 ± 13 nm, 74 ± 42 nm, 176 ± 63 nm and 1311 ± 514 opaque (Fig. 1). Together with their simplicity, low cost and facile nm. All these particles were stabilized by the addition of PVA (1 wt%) preparation, this behavior would make our composite films excellent during the emulsification process, which reduces the surface tension of candidates for the preparation of thermoresponsive smart windows (TR- the paraffin nanostructures formed and increases the viscosity of the SWs) for energy saving, in contrast to previously reported paraffin- suspension to prevent their creaming and coalescence [51]. Then, polymer mixtures that were found to be opaque at room temperature composite films were prepared by casting aqueous suspensions of the [43,44]. A further advantage of our approach is that the melting of SLPs SLPs (1 wt%) in the presence of dissolved PVA (10 wt%) onto poly­ can be induced by means of other external stimuli different from thermal styrene containers, which led to the formation of free-standing thin heating: (a) through the local heat released by nearby photothermal polymer layers after solvent evaporation that could be easily peeled off agents (e.g., noble metal nanostructures) when irradiated with light from the substrate (thickness ~0.2 mm). Electron microscopy analysis of [45]; and (b) by the Joule’s heating generated when electric current the composite films after treatment with chloroform to selectively passes through an adjacent electrode [44]. As a result, our composite dissolve the embedded SLPs demonstrated that they preserve their films could also enable the preparation of light- and electro-triggered structure and lie homogeneously dispersed within the surrounding PVA thermoresponsive SWs [27,42,44,46-48] that could be operated both matrix (Fig. 2c and Fig. S2). This good dispersibility must be ascribed to autonomously and under user’s demand. the small dimensions of the SLPs and the increasing viscosity of the PVA- rich medium as water evaporates during film formation, which disfavors Fig. 1. Structure and operating principle of our SW materials based on paraffin SLPs dispersed in polymer films. The refractive index matching between the particles (n2) and the surrounding polymer matrix (n1) reversibly varies upon paraffin solid–liquid transition, thus modulating the internal scattering effects and the transparency. 2 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 Fig. 2. a) Thermal variation of the refractive index of bulk eicosane and a PVA film. b) Transmission electron microscopy (TEM) image of EC@SLPs with an average diameter of 176 ± 63 nm. The particles were stained with uranyl acetate to obtain enhanced contrast. c) Scanning electron microscopy image of the cross-section of a cryo-fractured EC@SLP@PVA film treated with chloroform to dissolve the paraffin particles, which creates holes in the polymeric matrix showing the location of the SLPs into the film before their extraction. d) Transmittance spectra of an EC@SLP@PVA film at 25 ◦ C and 40 ◦ C, to which the solar irradiance spectrum ASTM G173- 03 Global Tilted 37◦ is superimposed. e) Photographs of an EC@SLP@PVA film in its transparent state at 25 ◦ C and in its opaque state at 40 ◦ C. f) Variation of the transmittance at λ = 550 nm of an EC@SLP@PVA film subjected to several consecutive heating–cooling cycles between 25 ◦ C and 40 ◦ C (theating = 40 s and tcooling = 180 s). All data in c-f) was obtained for composite films made from EC@SLPs with an average diameter of 176 ± 63 nm. particle creaming and aggregation. behavior that can be ascribed to the combination of two different factors Composite films with particles of different dimensions presented that favor light scattering: the large particle dimensions and the non- distinct degrees of transparency to naked eye as well as of thermally- negligible RI differences that exist between EC and PVA even in the induced transparency modulation. As expected, SLPs much smaller solid state at room temperature (RI25 C =1.514 and 1.510 for EC@SLPs ◦ than visible light wavelengths (38 ± 13 nm and 74 ± 42 nm) led to and PVA, respectively). Interestingly, an intermediate situation was transparent composite films with little transparency changes upon encountered for the film with SLPs of 176 ± 63 nm in diameter 25 C (EC@SLP@PVA). On the one hand, these particles are small enough to ◦ 25− 40 C temperature variation (% T 550nm >87.0%, % ΔT550nm <18%, films 1 ◦ and 2 in Table S1). In this case, the transparency properties are mainly minimize the effect of the tiny RI mismatch with the surrounding PVA determined by the reduced particle size irrespectively of the RI changes matrix at room temperature, thus yielding high transparency to visible EC that occur upon SLP melting when heating above Tm – i.e., minimal 25 C light (% T 550nm =72.2%). On the other hand, their dimensions are suf­ ◦ visible light scattering occurs at all temperatures. By contrast, the largest ficiently large to promote efficient visible light scattering when the RI SLPs (1311 ± 514 nm-in-diameter) resulted in opaque films below and difference between EC@SLPs and PVA increases upon paraffin melting EC 25 C 25− 40 C above Tm (% T 550nm =5.0%, %ΔT550nm = 4.5%, film 4 in Table S1), a ◦ ◦ (RI40 C =1.435 and 1.507 for EC@SLPs and PVA, respectively). As a ◦ 3 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 result, a clear transparent-to-opaque transition was observed to take 780–2500 nm) solar spectra. By combining high visible light trans­ place abruptly around TmEC (% ΔT550nm 25− 40 C =63.6%), whose actual mittance at room temperature with efficient solar irradiation modula­ ◦ amplitude depends on several parameters: the size and RI variation of tion upon heating (Fig. 2e), these results are clearly better than those EC@SLPs, the RI of the surrounding PVA matrix, the number of reported for some well-established TR-SW technologies (e.g., liquid dispersed SLPs and the thickness of the film (Fig. 2d-e and Fig. S3, film crystal- [31-35] and vanadium oxide-based [40,41] materials), and are EC@SLP@PVA in Table S1). not far from the best values described in the literature (%ΔTsolar = In view of the thermal modulation of transparency accomplished for 60–80% with % Tlum = 80–90% for hydrogels and copolymer films EC@SLP@PVA, we further investigated its optical performance for TR- [23,26,39,52], Fig. S4b). In addition, the thermally-induced optical SW applications. With this aim, we monitored its light transmittance response of EC@SLP@PVA was proven to be highly reversible and fa­ over the whole solar irradiance spectrum on Earth’s surface (λsolar = tigue resistant when applying sequential heating–cooling cycles 300–2500 nm), both at room temperature and after SLP melting at 40 ◦ C (Fig. 2f). (Fig. 2d and Fig. S4a). From these measurements, two main figures of merit were derived: (a) the transmittance in the visible or luminous 2.2. Gradual modulation of the thermal optical response range (λlum = 390–780 nm) for the transparent state of the material at 25 C room temperature, which we found to be % Tlum = 71.0%; and (b) the ◦ As many other TR-SW materials, EC@SLP@PVA undergoes a sudden EC transparent-to-opaque transition at a defined phase change temperature change in solar transmittance upon heating above Tm , for which an 25− 40 C – in our case, EC melting point. This limits the capacity to progressively overall value of %ΔTsolar = 46.5% was determined that can be ◦ adjust sunlight transmittance to the actual weather conditions, which broken down into the distinct variations achieved in the visible would require fine tuning the temperature at which solar transmission 25− 40 C 25− 40 C (% ΔTlum and NIR (% ΔTNIR ◦ ◦ =59.3%) =31.1%, λNIR = variation occurs. For this reason, strategies are needed to modulate the Fig. 3. a) DSC thermograms of EC@SLP@PVA, EC5.25HD1@SLP@PVA and EC1OC1@SLP@PVA (temperature ramp = 10 ◦ C min− 1). b) Variation of transmittance at λ = 550 nm of EC@SLP@PVA and EC5.25HD1@SLP@PVA for T = 21–43 ◦ C. c) Photographs of EC@SLP@PVA, EC5.25HD1@SLP@PVA and EC1OC1@SLP@PVA at T = 20, 30, 40 and 50 ◦ C. d) Variation of transmittance at λ = 550 nm of EC2OC1@SLP@PVA, EC1OC1@SLP@PVA and EC2OC3@SLP@PVA for T = 30–60 ◦ C. e) Time dependent variation of temperature and transmittance at λ = 550 nm of EC1OC1@SLP@PVA when heating from 30 to 60 ◦ C with isothermal steps of 120 s every 2 ◦ C of temperature increment. 4 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 thermal optical response of TR-SW systems – e.g., by adding dopants to respectively), which preserved the transparency modulation properties vanadium oxide films [41] or salts to hydrogels [20]. In our case, we 25− 40 C of the original composite materials (% ΔT550nm =62.1% and 61.6%, ◦ achieved this goal by simply mixing EC with paraffins of either lower or respectively). As shown in Fig. 4a, increased flexibility was observed for HD higher melting points: hexadecane (HD, Tm = 18.1 ◦ C [49]) and octa­ these composite layers, which we ascribe to the capacity of EG to act as a OC cosane (OC, Tm = 61.3 C [49]). ◦ plasticizer in PVA films without significantly affecting their RI proper­ In the case of EC-HD mixtures, SLPs of 276 ± 88 nm in diameter were ties (RI25 C =1.510 and 1.508 for SLP-free films of bare PVA and PVA ◦ prepared with a 5.25:1 paraffin ratio and then embedded in a PVA layer with 10 wt% of EG, respectively). through casting (EC5.25HD1@SLP@PVA, Fig. S5). Differential scanning Another desired property is composite film scalability, which we calorimetry (DSC) measurements of the resulting film showed a broad explored by preparing large TR-SW materials by mixing 25 mL of an EC endothermal peak at lower temperatures than Tm rather than two in­ EC@SLP aqueous suspension and 250 g of a PVA aqueous solution (10 EC HD dependent signals at Tm and Tm (Fig. 3a), in agreement with the re­ wt%) and casting them onto a large mold. After water evaporation and sults reported for bulk binary mixtures of other paraffins [53,54]. peeling off from the substrate, a round EC@SLP@PVA film of 42 cm in Actually, a similar behavior was encountered herein by DSC analysis of a diameter was obtained that exhibited thermally-induced transparency homogeneous bulk mixture of EC and HD with a 5.25:1 ratio (Fig. S6a), modulation (Fig. 4b and Fig. S9). This result clearly demonstrates that which further corroborates that: (a) the two paraffins remain mixed film formation can be easily scaled up using a mould of the desired di­ within the EC-HD SLPs; and (b) their solid–liquid phase change does not mensions. In addition, the capacity of this type of materials to adhere to take place isothermally, but through a range of temperatures where the window surfaces [55] was investigated by directly conducting the solid and liquid solutions of the mixed paraffins coexist [53,54]. casting process onto a 12x12 cm2 piece of glass. Because of PVA-glass Importantly, this must have a strong impact on the thermal variation of affinity, complete and homogenous coating with excellent adhesion their optical properties. In fact, a stepwise change of RI with tempera­ properties was observed. As expected, the obtained coated glass showed ture was measured for a homogenous bulk EC5.25-HD1 mixture - from RI 25 C high transparency at room temperature (% T550nm =90.3%) while it ◦ ~1.51 at 20 ◦ C to RI ~1.43 at 32 ◦ C -, which is in contrast with the became opaque when heated up above TmEC (% T550nm 40 C =41.8%, Fig. 4c). ◦ abrupt behavior of pure EC (Fig. S6b). As a result, a gradual and large The durability and stability of our paraffin-PVA films were also transparency change was observed for EC5.25HD1@SLP@PVA within the investigated under different relevant conditions: (a) after storage for 8 20− 40 C 20–40 ◦ C range (% ΔT550nm =48.5%) instead of the sudden variation ◦ months under ambient conditions – i.e., room temperature and ambient EC registered for EC@SLP@PVA at Tm (Fig. 3b-c). light; (b) upon accelerated aging at 70 ◦ C for several hours; and (c) when PVA films containing mixed EC-OC SLPs with 2:1, 1:1 and 2:3 ratios subjected to 10 consecutive freeze–thaw cycles to evaluate their frost and around 180 nm in diameter were also prepared resistance (Fig. S10). None of these treatments caused significant effects (EC2OC1@SLP@PVA, EC1OC1@SLP@PVA, EC2OC3@SLP@PVA, on the TR-SW performance of the EC@SLP@PVA and EC@SLP@PVA- Fig. S7). Homogenous mixing of the paraffins within the SLPs was again EG5% films tested, whose capacity to thermally modulate light trans­ revealed by DSC analysis, where broader thermal phase transitions were mission was preserved even after long storage times and exposure to observed extending from ca. 35 ◦ C to 55 ◦ C (Fig. 3a). This agrees with high and low temperatures. the transparent-to-opaque transitions measured for the EC-OC compos­ Finally, co-extrusion lamination of EC@SLP@PVA-EG5% with ther­ ite films, which took place gradually within the 35–52 ◦ C thermal range moplastic polyurethane and polycarbonate layers (5 × 5 cm2) was also (Fig. 3c-d). Moreover, their transmittance variation profile was found to carried out to provide even larger durability to the composite film. This be dependent on the actual EC-OC ratio of the SLPs, with required the application of high temperatures (up to 120 ◦ C) and pres­ EC1OC1@SLP@PVA exhibiting an almost linear transparency change sures to ensure proper adhesion of the multilayer structure. The result­ over a temperature range of ca. 15 ◦ C. The continuous thermomodula­ ing laminated material retained the thermo-optical properties of the tion of light transparency for this sample was additionally proven by original film: it switched between high transparency at room tempera­ temperature-dependent isothermal transmittance measurements carried 25 C 40 C ture (% T550nm =77.8%) and opacity at 40 ◦ C (% T550nm =20.1%), thus ◦ ◦ out upon fast temperature increments of 2 ◦ C followed by 120 s of demonstrating the capacity of our PVA-paraffin films to endure standard equilibration (Fig. 3e). Therefore, all these results combined unambig­ lamination conditions (Fig. 4d). In addition, lamination enhanced the uously demonstrate that the use of mixed paraffin SLPs allows the mechanical (e.g., anti-scratchability) and weathering (e.g., water resis­ preparation of TR-SW materials with (a) gradual transparency adapta­ tance) properties of our TR-SW materials, which are intrinsically limited tion to ambient temperature and (b) whose thermal response range can by the softness and aqueous solubility of PVA. be modulated by the particle composition. 2.3. Other properties of the paraffin-PVA films for smart window 2.4. Photo- and electrothermal transmittance modulation applications A highly desired feature for TR-SWs is multistimuli-responsive Together with their low cost, straightforward preparation and behavior [37,56], which cannot be currently accomplished with most optimal optical properties, other features from our paraffin-polymer technologies in the smart glass market. On the one hand, developing composite films might be of relevance for their practical implementa­ materials whose transparency not only varies thermally but also when tion as TR-SWs. First, we corroborated that the mechanical properties of photoactivated should allow for a better self-regulation of light and composite PVA films were not compromised by the presence of SLPs in energy fluxes through windows for energy-saving applications their interior. In fact, only a limited reduction in Young’s module, tensile [46,47,57]. On the other hand, if electroinduced transmittance modu­ strength, toughness and hardness was registered relative to bare SLP- lation is accomplished, user’s control of window transparency on de­ free PVA layers of the same thickness (Fig. S8 and Table S2). Another mand will also be enabled by applying an external voltage [30,44,58- interesting mechanical feature is film flexibility, which is required to 61]. In light of this, we explored the photo- and electrocontrolled favor manipulation, allow application on curved surfaces and increase operation of our TR-SW materials based on polymer-paraffin resistance to mechanical stress. Thus, control and tuneability of the composites. flexibility of SW materials are important to enhance their robustness, Based on our previous demonstration that SLPs’ melting can be durability and range of applications. To address this issue, two films of achieved photothermally under irradiation [45], we hypothesized that EC@SLP@PVA containing 5 wt% and 10 wt% of ethylene glycol (EG) the addition of small amounts of photothermal agents (PT) to our were prepared (EC@SLP@PVA-EG5% and EC@SLP@PVA-EG10% polymer-paraffin composite materials would facilitate their transparent- 5 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 Fig. 4. a) Photographs of EC@SLP@PVA (left), EC@SLP@PVA-EG5% (middle) and EC@SLP@PVA-EG10% (right) films when bent. b) Photograph of a 42 cm-in- diameter EC@SLP@PVA film attached on to a glass window. c) Photograph of a 12x12 cm2 glass substrate coated with EC@SLP@PVA. The upper right region of the material was selectively heated to show the contrast between its transparent and opaque states. d) Photographs of a laminated EC@SLP@PVA film at room tem­ perature (left) and after heating above 37 ◦ C (right). 6 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 to-opaque transition upon sunlight absorption and local generation of the visible light transmission of EC5.25HD1@SLP-AuNEs@PVA was heat. A mixture of visible-absorbing gold nanoparticles (AuNPs, average found to decrease progressively as NIR power increases between 0 and diameter = 18 ± 5 nm) and NIR-absorbing gold nanoshells (AuNSs, 100 mW mm− 2, which reproduces the gradual transparency response average diameter = 96 ± 34 nm) in a 1:2.2 wt ratio was chosen as PT previously demonstrated upon thermal heating. To further investigate agents because of (a) their superior properties in photostability and this behavior, we explored irradiation under relevant sunlight intensities photothermal heating [62], and (b) their complementary absorption - i.e., by illumination with a solar simulator operating at 100 mW cm− 2 that covers most of the solar irradiance spectrum (Fig. S11). Conse­ (1 sun, AM1.5), which reproduces the solar irradiation conditions in quently, very low concentrations of gold nanostructures (AuNEs) were moderate climates. Interestingly, after 3–4 min of illumination at 25 ◦ C, required to achieve photothermal SLPs’ melting (AuNEs ≤ 0.018 wt%), all the polymer-paraffin composite films bearing gold nanostructures which should allow minimizing their effect on the color, light scattering showed the expected modulation in optical transmittance (Fig. 5c-d and and RI properties of the final films. In fact, the two different EC-PVA Fig. S12), a result that could not be accomplished in the absence of the composites prepared containing 0.018 wt% (EC@SLP-AuNEs@PVA-1) PT agents (Fig. S13). After turning off the solar simulator, the AuNE- and 0.009 wt% of AuNEs (EC@SLP-AuNEs@PVA-2) showed a pale doped films recovered their initial transparent state spontaneously due neutral residual color and fine light transparency at room temperature in to cooling at room temperature and this behavior was found to be the dark (% Tlum 25 C =61.6% and 69.4% and haze = 3.8% and 3.6%, repeatable and robust through consecutive on–off cycles of sunlight ◦ respectively). In addition, they became opaque when heated above EC illumination (Fig. 5d). Overall, these results prove the capacity of pho­ melting temperature and allowed for high modulation of solar irradi­ totriggering our TR-SW materials without detrimentally affecting their 25− 40 C optical response. ance (% ΔTsolar =41.5% and 43.4%, respectively; Fig. 5a). Analo­ ◦ Because paraffin melting can also be accomplished thanks to the gously, when a film containing 0.009 wt% of AuNEs and Joule’s heating generated when current passes through a nearby EC5.25HD1@SLPs was prepared (EC5.25HD1@SLP-AuNEs@PVA), it also conductor [44,63], we also pursued the electroinduced operation of our exhibited good solar transmittance modulation through thermal heating TR-SW systems. For this we deposited a polymer composite of SLPs 20− 40 C 20 C (% ΔTsolar =33.7% with % Tlum = 56%, Fig. 5b). In all the cases, ◦ ◦ (EC@SLP@PVA) onto a transparent electroheater consisting of an ITO- however, a small variation in the TR-SW performance of these samples coated glass substrate (Fig. 6a). As expected, upon application of 25− 40 C was observed relative to analogous AuNE-free films (e.g., % ΔTsolar increasing voltages to this device at 25 ◦ C, the local temperature on the ◦ decreased in 5.0% and 3.1% for EC@SLP-AuNEs@PVA-1 and EC@SLP- surface of the polymer film highly increased, reaching values that are AuNEs@PVA-2, respectively). We ascribe this behavior to the non- well-above of paraffin melting point even for rather low potentials (ca. 6 negligible light absorption and scattering from AuNEs, which occurs V; Fig. 6b). At such voltages, efficient transmittance modulation was both in the transparent and opaque states of our films and, therefore, accomplished at room temperature (% T550nm 0 V 6 V =62.3% and % T550nm = slightly reduces their sunlight transmittance contrast. 15.4%), thus revealing the electrotriggered response of our TR-SW To analyze the photothermal response of these materials, we first materials (Fig. 6c-d). To further refine this behavior, we investigated considered irradiation with a NIR laser at λexc = 830 nm that selectively the electroinduced operation of EC1OC1@SLP@PVA, which we had excited the AuNSs in the films. As shown in Fig. 5b, photoinduced observed to present thermal gradual variation of transparency. Upon transparent-to-opaque transition could be accomplished at room tem­ adhesion onto ITO-coated glass, this film showed progressive modula­ perature for polymer-paraffin composites bearing both pristine EC and tion of light transmittance at 25 ◦ C depending on the applied voltage, mixed EC-HD SLPs. Thus, an abrupt and complete transmittance switch thereby allowing ultimate user’s control on the degree of transparency was measured for EC@SLP-AuNEs@PVA-2 at NIR powers around 125 achieved through windows (Fig. 6e-f). Importantly, this electroinduced mW mm− 2, which mimics its thermoresponsive behavior. By contrast, behavior was found to be robust and reproducible for several cycles of Fig. 5. A) Transmittance spectra of EC@SLP-AuNEs@PVA-1 (continuous line) and EC@SLP-AuNEs@PVA-2 (dashed line) at 25 ◦ C (black) and 40 ◦ C (red) in the dark. b) Variation of transmittance at λ = 550 nm and 25 ◦ C of EC@SLP-AuNEs@PVA-2 and EC5.25HD1@SLP-AuNEs@PVA under cw NIR irradiation at increasing power density (λexc = 830 nm). c) Photographs of EC@SLP-AuNEs@PVA-1 before (left) and 1 s after irradiation (right) with a solar simulator operating at 100 mW cm− 2 (1 sun, AM1.5). d) Variation of transmittance at λ = 550 nm of EC@SLP-AuNEs@PVA-1 during consecutive off–on cycles of solar simulator irradiation (100 mW cm− 2) at 25 ◦ C. The yellow areas represent the pe­ riods of time when the film was irradiated, while the white areas indicate when it was in the dark. The inset shows a magnified image of transmittance variation for the first irradiation cycle. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 7 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 Fig. 6. a) Scheme of our electroresponsive SW de­ vices. b) Variation of polymer surface temperature and current intensity with applied voltage for an electroresponsive device made from EC@SLP@PVA. c) Transmittance spectra of EC@SLP@PVA on ITO- coated glass at bias voltages of 0 V and 6 V. d) Pho­ tographs of EC@SLP@PVA on ITO-coated glass at bias voltages of 0 V and 6 V. e) Variation of trans­ mittance at λ = 550 nm and surface temperature with applied voltage for EC1OC1@SLP@PVA deposited onto ITO-coated glass. f) Photographs of EC1OC1@SLP@PVA on ITO-coated glass at increasing voltages. repetitive voltage application (Fig. S14). be accomplished photothermally in this case. This was indeed proven by monitoring the temperature on the surface of this film, which could only reach a maximum value around 31 ◦ C in the stationary state (Fig. 7d). By 2.5. Energy-saving performance contrast, significantly lower solar thermal heating was measured when using EC@SLP-AuNEs@PVA-1 and EC@SLP-AuNEs@PVA-2 as SWs To evaluate the performance of our SW materials for energy-saving (ΔT=7.9 and 8.9 ◦ C), which is consistent with their photothermal applications, we constructed an insulating box bearing a small 2.5 × operation triggered by the presence of AuNEs. For these two samples, 2.5 cm2 aperture as a house model and monitored the temperature in­ complex profiles of temperature variation with irradiation time were side the box and on the outer surface of these films, when exposed to determined that reveal the underlying energy-saving mechanism different relevant temperature and irradiation conditions (Fig. 7a and (Fig. 7b and d). First, a fast temperature increment inside the house Fig. S15a). In these experiments, both thermo- (EC@SLP@PVA) and model was observed until enough photothermal heating was produced photothermoresponsive polymer-paraffin composite films (EC@SLP- as to surpass the EC melting temperature on the surface of the films (t ~ AuNEs@PVA-1 and EC@SLP-AuNEs@PVA-2) were used as SWs to cover 1–2 min), which concomitantly induced their transparent-to-opaque the aperture of the house model, together with a blank PVA layer as a switch. Then, a subsequent decrease in temperature was registered in­ reference. side the insulating box due to reduced solar radiation transmission In a first step, we analyzed the energy-saving effect of our SW ma­ through the opaque films, while the surface temperature kept rising terials at a constant room temperature of 21 ◦ C and a sunlight irradiation under illumination (t ~ 2–4 min). Finally, a slight thermal increase was of 100 mW cm− 2 with a solar simulator (1 sun, AM1.5; Fig. 7b-c). As further detected inside the house model, until the corresponding pho­ expected, a high temperature increase was observed inside the house tothermostationary state was reached (t ~ 4–10 min). Overall, this model under these conditions for the reference PVA film (ΔT=12.8 ◦ C at resulted in a 30% (EC@SLP-AuNEs@PVA-2) and 38% (EC@SLP- the photothermostationay state), which remained highly transparent at AuNEs@PVA-1) reduction in solar heat gain inside the house model all times. A similar situation occurred for EC@SLP@PVA (ΔT=12.1 ◦ C) after 10 min, where the higher concentration of PT agents responsible because paraffin melting requires temperatures above 35 ◦ C and cannot 8 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 Fig. 7. a) Scheme of the set-up for energy-saving tests. b) Temperature variations inside the insulating box promoted by the heat gain during 10 min of irradiation with a solar simulator (1 sun, AM1.5, 100 mW cm− 2) through a blank PVA film, EC@SLP@PVA, EC@SLP-AuNEs@PVA-2 and EC@SLP-AuNEs@PVA-1. The room temperature during the experiment was 21 ◦ C. c) Effect of ambient temperature on the temperature variations achieved inside the insulating box after 10 min of irradiation with a solar simulator (1 sun, AM1.5, 100 mW cm− 2) through a blank PVA film, EC@SLP@PVA, EC@SLP-AuNEs@PVA-2 and EC@SLP-AuNEs@PVA-1. d) Temperature variations at the front surface of the SW film of the insulating box that are generated during 10 min of irradiation with a solar simulator (1 sun, AM1.5, 100 mW cm− 2) through a blank PVA film, EC@SLP@PVA, EC@SLP-AuNEs@PVA-2 and EC@SLP-AuNEs@PVA-1. The room temperature during the experiment was 21 ◦ C. for the photothermal melting of SLPs account for the largest (and fastest) 72.0 and 64.9% for EC@SLP@PVA, EC@SLP-AuNEs@PVA-1 and response measured for EC@SLP-AuNEs@PVA-1 (Fig. 7b-c). EC@SLP-AuNEs@PVA-2, respectively, which are slightly lower than the Next, we conducted the same energy-saving experiment under transmittance of the blank PVA film (% Tsolar25 C ◦ =91.1%). AM1.5 irradiation at two other different ambient temperatures: 13 ◦ C As for the measurements registered indoors with the solar simulator and 30 ◦ C to replicate the conditions of cold weather and a hot summer at 30 ◦ C, efficient energy saving due to transparency modulation was in day, respectively (Fig. 7c and Fig. S16a). For comparison purposes, an this case achieved not only for the phothermally-responding AuNE- additional measurement was performed outdoors at the solar noon doped films (31% and 40% in temperature reduction), but also for during a winter sunny day (Tambient = 12 ◦ C, Fig. S16b-c). Importantly, EC@SLP@PVA (25% in temperature reduction). This means that the no transparent-to-opaque transition was observed for any of the SW thermal mechanism for SLPs’ melting and, consequently, AuNE-free film materials tested at 12–13 ◦ C with the solar simulator or under direct opacity also become activated at high ambient temperatures and irra­ sunlight, as no thermal or photothermal melting of SLPs could be diation conditions, which we ascribe to the combination of two different induced at this low ambient temperature. Therefore, as desired for sunny factors: (a) the relatively low melting point of the EC@SLPs used cold days, all of them allowed for a large solar heat build-up inside the (TmEC =36.5 ◦ C [49]), which is only slightly above 30 ◦ C; and (b) the house model (ΔT>10.8 ◦ C) irrespective of the illumination conditions. heating of the EC@SLP@PVA layer, which results from both the diffu­ The small differences in temperature variation observed between the sion of the solar heat build-up caused inside the insulating box as well as reference PVA sample and our polymer-paraffin films must be attributed the intrinsic absorption of high-wavelength NIR light (λ > 1300 nm) by to the lower solar irradiation transmittance of the latter in their trans­ the PVA and EC molecules of the film (Fig. S15b). Actually, the 25 C parent state due to the presence of SLPs and AuNEs: % Tsolar = 75.1, ◦ transparent-to-opaque transition of EC@SLP@PVA was achieved after 9 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 2–3 min of solar simulator irradiation when a thermal increase of about First, 2 g of EC were melted by heating up to 60 ◦ C; then, they were 10 ◦ C was reached inside the house model (Fig. S16a), which must mixed with 20 mL of water containing 1 wt% of PVA as a stabilizer that induce SLPs’ melting at an external temperature of 30 ◦ C even in the was also heated at the same temperature. Afterwards, the mixture was absence of AuNEs. emulsified using an ultrasonic homogenizer operating at 70% amplitude Finally, energy-saving experiments were performed for films loaded (Branson 450 D Sonifier, 400 W–20 kHz, equipped with disruptor horn with SLPs made of paraffin mixtures: EC5.25HD1@SLP@PVA and and 13 mm flat tip) during 5 min in cycles of 15 s of sonication and 5 s of EC5.25HD1@SLP-AuNEs@PVA, for which very similar results were ob­ repose. Finally, the emulsion was poured onto 30 mL of precooled water tained (Fig. S17). Thus, whereas large solar heating gain under solar (4 ◦ C) to induce the rapid solidification of the EC droplets and generate a irradiation was enabled for these two samples at an ambient tempera­ suspension of the corresponding SLPs. For the preparation of the SLPs ture of 13 ◦ C (ΔT>12.5 ◦ C), they reduced heat build-up inside the house with an average diameter of 38 nm and 74 nm, the same procedure was model when warmer. As expected, this effect was specially efficient for applied but using 10 wt% and 5 wt% PVA concentrations during the the EC5.25HD1@SLP-AuNEs@PVA film containing photothermal nano­ emulsion process, respectively. As for the SLPs with an average diameter structures, which led to a temperature reduction of 32% and 31% under of 1311 nm, in this case we used a 10 wt% PVA concentration and high- solar irradiation at external temperatures of 21 ◦ C and 30 ◦ C, respec­ shear homogenization (IKA Ultra-Turrax® homogenizer) instead of an tively. Therefore, all these results combined demonstrate that our ultrasonic homogenizer. thermo- and photothermoresponsive materials based on paraffin- polymer composites can be used as adaptative and dynamic SWs for energy saving. 4.2. Preparation of SLPs of paraffin mixtures 3. Conclusions First, the desired mixtures of paraffins were prepared by mixing the corresponding amount of each paraffin in a vial and heating up the In this work we described a new strategy towards thermoresponsive sample at 10 ◦ C above the melting temperature of the paraffin with the smart windows, which consists in embedding solid lipid nanoparticles highest melting point. The melted mixture was stirred for 30 min and made from simple paraffins -e.g., eicosane - within polymer films - e.g., then subjected to the same procedure described above for the prepara­ poly(vinyl alcohol). These components are chosen to warrant refractive tion of SLPs of pure paraffins. In all the cases, a 1 wt% PVA concentra­ index matching when the paraffin nanoparticles are in their solid state, tion was used during the emulsion process, which was induced by means thus leading to high light transparency. This condition is lost upon of an ultrasonic homogenizer. paraffin melting, which results in pronounced light scattering at the nanoparticle-polymer interface and, consequently, opacity. As a result, 4.3. Preparation of gold nanoshells (AuNSs) and nanoparticles (AuNPs) our paraffin-polymer composites combine high transparency to the visible range of sunlight in the inactive state (~70–75%) with ample AuNSs were prepared following the method described by Guan et al. modulation of solar radiation transmittance upon heating (~45–50%). [64]. First, 20 µL of APTES and 9.2 mL of water were mixed under Although these figures are on the lower limit of the best values reported stirring for 30 s. Then, 640 µL of a 40 mM HAuCl4 aqueous solution were in the literature for other SW technologies, our films present many other added, obtaining a yellow emulsion due to the poor solubility of APTES simultaneous advantageous features as thermoresponsive smart win­ in water. The HAuCl4/APTES mixture was stirred for 30 s followed by dows. First, they are not only easily fabricated from low-cost and the addition of 800 µL of a 0.1 M aqueous suspension of NaBH4. readily-available materials, but are scalable and applicable to relevant Immediately, the color of the mixture changed to deep green and 800 µL flat and curved transparent surfaces - e.g., glass. Second, their thermal of BSA 0.1 M aqueous solution were then added to stabilize the formed response is highly reversible and photostable due to the lack of organic AuNSs. photoabsorbing components, while it can be finely tuned by changing AuNPs were prepared according to the procedure reported by Grabar the composition of the paraffin nanoparticles. In addition, the me­ et al. [65]. First, 250 mL of a 1 mM aqueous solution of HAuCl4 were chanical properties of the paraffin-polymer composites can be enhanced stirred under reflux. Then, 25 mL of a 40 mM sodium citrate aqueous to meet the requirements for actual applications by additive addition solution were added. The mixture was kept under reflux for another 15 and lamination - e.g., flexibility, anti-scratchability and weathering min. During this time, the color changed from pale yellow to deep red. resistance. Finally, these materials can be turned photo- and electro­ The final colloidal suspension was allowed to cool down to room tem­ sensitive by incorporation of photothermal nanoparticles and deposition perature while stirring. onto transparent electrodes, which (a) improves their self-adaptation to ambient conditions by simultaneously responding to sunlight and external temperatures, and (b) enables external user’s control of solar 4.4. Preparation of polymeric thermo- and photothermoresponsive PVA transmittance through on-command voltage application. Actually, films films (paraffin@SLP@PVA and paraffin@SLP-AuNEs@PVA) with combined thermal and phothermal activity were demonstrated to allow high sunlight transmission in cold weather conditions, while SLP-loaded PVA films with thermally- and photothermally-induced efficiently reduce solar heat gain indoors in sunny hot days (~30–40%). solar light transmittance modulation were prepared using the Because of these combinations of properties, our paraffin-polymer following procedure. First, the convenient volume of a water suspension composites emerge as excellent candidates for the development of of the SLPs of interest, typically 1.6 mL (0.04 g mL− 1), was added to 7 g simple, cost-effective and multi-stimuli-responsive smart windows, of an aqueous solution of PVA (10 wt%). The mixture was stirred until a overcoming the main issues related with other current TR-SWs, such as homogeneous suspension was obtained. For the preparation of photo­ high cost (e.g., liquid crystals) or the difficulty to scale-up and integrate thermoresponsive films, 500 or 250 µL of an AuNSs aqueous suspension them into real fenestration elements (e.g., hydrogels). (0.34 mg mL− 1) and 250 or 125 µL of an AuNPs aqueous suspension (0.30 mg mL− 1) were also added. In addition, when polymer films with 4. Materials and methods enhanced flexibility were desired, 0.09 or 0.045 g of ethylene glycol were added too. In all these cases, the homogeneous mixtures obtained 4.1. Preparation of solid lipid nanoparticles (SLPs) were poured into a 5.5 cm-in-diameter polystyrene Petry dish, and the water solvent was let to evaporate under ambient conditions for 72 h. EC SLPs with an average diameter of 176 nm were prepared through Afterwards, a free-standing film was obtained by easily peeling it off an emulsion-cooling methodology previously developed by us [45]. from the container. 10 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 4.5. Preparation of scaled-up thermoresponsive PVA films Acknowledgements The preparation of the 42 cm-in-diameter thermoresponsive PVA This work was supported by MINECO/FEDER (RTI2018-098027-B- film containing EC SLPs was analogous to the procedure described in C21, PID2019-106171RB-100 and PDC2022-133368-I00 projects) and previous section 4-4. In this case 25 mL of an aqueous suspension of Generalitat de Catalunya (2017 SGR00465 and 2021 PROD 00190 EC@SLPs (0.04 g mL− 1) were mixed with 250 g of an aqueous solution of projects). J. R. O. thanks the Generalitat de Catalunya (AGAUR) for his PVA (10 wt%). The mixture was left in repose during 12 h to allow the predoctoral FI fellowship. ICN2 acknowledges support from the Severo air bubbles to escape from the mixture. After that, the mixture was Ochoa program (MINECO, SEV-2017-0706) and the CERCA program slowly poured into the cavity formed by two independent pieces of 2 (Generalitat de Catalunya). mm-thick polyethylene terephthalate (PET) stacked onto each other: a flat base (50 cm × 70 cm) and an equally large layer in which a circular Data availability aperture with an inner diameter of 42 cm was cut. Then, the water solvent was let to evaporate under ambient conditions for 72 h. After­ The raw and processed data required to reproduce these findings is wards, the upper PET layer was separated from the base, where the film available in the CORA.RDR, Research Data Repository (https://data­ formed remained adhered (Fig. S9). Finally, it was peeled off from the verse.csuc.cat/): https://doi.org/10.34810/data683 (accessed on 20 base in order to obtain the 42 cm-in-diameter free-standing thermores­ March 2023). ponsive PVA film. Appendix A. Supplementary data 4.6. Preparation of glass with thermoresponsive EC@SLP@PVA coatings Supplementary data (the materials prepared and their optical prop­ The 12 × 12 cm2 glass coated with a thermoresponsive erties) to this article can be found online at https://doi.org/10.1016/j. EC@SLP@PVA film was prepared by pouring a mixture of 1.6 mL of an cej.2023.142390. EC@SLP suspension (0.04 g mL− 1) and 7 g of an aqueous solution of PVA (10 wt%) directly onto the glass substrate. In order to avoid the spilling References of the liquid and to confine it over the glass substrate, walls of 1 cm height were prepared using modelling putty along the glass borders. The [1] X. Cao, X. Dai, J. Liu, Energy Build. 128 (2016) 198–213. [2] M. González-Torres, L. Pérez-Lombard, J.F. Coronel, I.R. Maestre, D. Yan, Energy water solvent was let to evaporate over 24 h and the glass with a ther­ Rep. 8 (2022) 626–637. moresponsive coating was obtained. [3] Y. Ke, J. Chen, G. Lin, S. Wang, Y. Zhou, J. Yin, P.S. Lee, Y. Long, Adv. Energy Mater. 9 (2019), 1902066. [4] K. Allen, K. Connelly, P. Rutherford, Y. Wu, Energy Build. 139 (2017) 535–546. 4.7. Preparation of laminated thermoresponsive PVA film [5] H. Kim, S. Yang, Adv. Funct. Mater. 30 (2020), 1902597. [6] Y. Ke, C. Zhou, Y. Zhou, S. Wang, S.H. Chan, Y. Long, Adv. Funct. Mater. 28 (2018), 1800113. The preparation of the thermoresponsive composite film laminated [7] H. Khandelwal, A.P.H.J. Schenning, M.G. Debije, Adv. Energy Mater. 7 (2017), between polyurethane and polycarbonate layers was prepared as fol­ 1602209. lows. First, 3 mL of an EC@SLPs aqueous suspension (0.04 g mL− 1), 14 g [8] R. Tällberg, B.P. Jelle, R. Loonen, T. Gao, M. Hamdy, Sol. Energy Mater. Sol. Cells 200 (2019), 109828. of a PVA aqueous solution (10 wt%) and 0.9 g of ethylene glycol were [9] H. Torres-Pierna, D. Ruiz-Molina, C. Roscini, Mater. Horiz. 7 (2020) 2749–2759. mixed and poured into a polystyrene Petri dish of 9 cm in diameter. The [10] S. Nigel Corns, S.M. Partington, A.D. Towns, Color. Technol. 125 (2009) 249–261. suspension was left during 48 h at room temperature to evaporate the [11] G.H. Timmermans, S. Hemming, E. Baeza, E.A.J. van Thoor, A.P.H.J. Schenning, M.G. Debije, Adv. Opt. Mater. 8 (2020), 2000738. water solvent. Then, the film was peeled off from the substrate and a [12] Y. Wang, E.L. Runnerstrom, D.J. Milliron, Annu. Rev. Chem. Biomol. Eng. 7 (2016) central square section of 5 × 5 cm was cut. After that, the square piece 283–304. was sent to the company NovoGenio, where the lamination process was [13] Q. Meng, G. Wang, H. Jiang, Y. Wang, S. Xie, J. Mater. Sci. 48 (2013) 5862–5870. [14] S. Yamazaki, H. Ishida, D. Shimizu, K. Adachi, ACS Appl. Mater. Interfaces 7 conducted under the standard conditions established for the TPU (2015) 26326–26332. NovoGlass SF 1959 treatment. [15] L. Wang, Q. Li, Chem. Soc. Rev. 47 (2018) 1044–1097. [16] Y. Zhang, C.Y. Tso, J.S. Iñigo, S. Liu, H. Miyazaki, C.Y.H. Chao, K.M. Yu, Appl. Energy 254 (2019), 113690. 4.8. Preparation of electrothermoresponsive PVA films [17] S. Liu, Y. Li, Y. Wang, K.M. Yu, B. Huang, C.Y. Tso, Adv. Sci. 9 (2022), 2106090. [18] A. Cannavale, F. Martellotta, P. Cossari, G. Gigli, U. Ayr, Appl. Energy 225 (2018) 975–985. To prepare electrothermoresponsive smart windows, a thermores­ [19] R. Song, G. Li, Y. Zhang, B. Rao, S. Xiong, G. He, Chem. Eng. J. 422 (2021), ponsive PVA film (paraffin@SLP@PVA) of 3 × 2 cm was attached to the 130057. surface of an ITO coated glass (3 × 3 cm), previously covered by a [20] Y. Zhou, X. Dong, Y. Mi, F. Fan, Q. Xu, H. Zhao, S. Wang, Y. Long, J. Mater. Chem. A 8 (2020) 10007–10025. transparent adhesive layer obtaining a sandwich structure (see Fig. 6a in [21] T.-G. La, X. Li, A. Kumar, Y. Fu, S. Yang, H.-J. Chung, ACS Appl. Mater. Interfaces 9 the main text). The electrodes of the obtained paraffin@SLP@PVA@ITO (2017) 33100–33106. were connected to a power supplier to induce the transparency switch by [22] Y. Zhou, S. Wang, J. Peng, Y. Tan, C. Li, F.Y.C. Boey, Y. Long, Joule 4 (2020) 2458–2474. Joule’s heating when a specific voltage was applied. [23] X.-H. Li, C. Liu, S.-P. Feng, N.X. Fang, Joule 3 (2019) 290–302. [24] T. Jiang, X. Zhao, X. Yin, R. Yang, G. Tan, Appl. Energy 287 (2021), 116573. [25] L.-W. Xia, R. Xie, X.-J. Ju, W. Wang, Q. Chen, L.-Y. Chu, Nat. Commun. 4 (2013) Declaration of Competing Interest 2226. [26] Q. Zhang, Y. Jiang, L. Chen, W. Chen, J. Li, Y. Cai, C. Ma, W. Xu, Y. Lu, X. Jia, Z. Bao, Adv. Funct. Mater. 31 (2021), 2100686. The authors declare that they have no known competing financial [27] G. Xu, H. Xia, P. Chen, W. She, H. Zhang, J. Ma, Q. Ruan, W. Zhang, Z. Sun, Adv. interests or personal relationships that could have appeared to influence Funct. Mater. 32 (2022), 2109597. the work reported in this paper. [28] P. v Rathod, P.P. More, J.M.C. Puguan, H. Kim, Sol. Energy Mater. Sol. Cells 230 (2021) 111202. [29] X. Yang, S. Lv, T. Li, S. Hao, H. Zhu, Y. Cheng, S. Li, H. Song, ACS Appl. Mater. Data availability Interfaces 14 (2022) 20083–20092. [30] P. v Rathod, J.M.C. Puguan, H. Kim, Chem. Eng. J. 422 (2021) 130065. Data available in the CORA.RDR, Research Data Repository [31] A.J.J. Kragt, R.C.G.M. Loonen, D.J. Broer, M.G. Debije, A.P.H.J. Schenning, J. Polym. Sci. 59 (2021) 1278–1284. (https://dataverse.csuc.cat/): https://doi.org/10.34810/data683 [32] X. Liang, C. Guo, M. Chen, S. Guo, L. Zhang, F. Li, S. Guo, H. Yang, Nanoscale Horiz (accessed on 20 March 2023) 2 (2017) 319–325. 11 J.R. Otaegui et al. Chemical Engineering Journal 463 (2023) 142390 [33] Z.-Y. Kuang, Y. Deng, J. Hu, L. Tao, P. Wang, J. Chen, H.-L. Xie, ACS Appl. Mater. [48] J. Yang, T. Lim, S.-M. Jeong, S. Ju, ACS Appl. Mater. Interfaces 13 (2021) Interfaces 11 (2019) 37026–37034. 20689–20697. [34] W. Yoon, Y. Choi, S. Lim, J. Koo, S. Yang, D. Jung, S. Kang, K. Jeong, Adv. Funct. [49] D.R. Lide, CRC Handbook of Chemistry and Physics, CRC Press, Boca Raton, FL, Mater. 30 (2020) 1906780. 2016. [35] J. Wang, C. Meng, C.-T. Wang, C.-H. Liu, Y.-H. Chang, C.-C. Li, H.-Y. Tseng, H.- [50] J.E. Mark, Polymer Data Handbook, New York, 1999. S. Kwok, Y. Zi, Nano Energy 85 (2021), 105976. [51] G.F. Blehn, M.L. Ernsberger, Ind. Eng. Chem. 40 (1948) 1449–1453. [36] K.G. Gutierrez-Cuevas, L. Wang, Z. Zheng, H.K. Bisoyi, G. Li, L.-S. Tan, R.A. Vaia, [52] Y. Zhou, Y. Cai, X. Hu, Y. Long, J. Mater. Chem. A 2 (2014) 13550–13555. Q. Li, Angew. Chem. Int. Ed. 55 (2016) 13090–13094. [53] T. Kousksou, A. Jamil, T. el Rhafiki, Y. Zeraouli, Sol. Energy Mater. Sol. Cells 94 [37] L. Wang, H.K. Bisoyi, Z. Zheng, K.G. Gutierrez-Cuevas, G. Singh, S. Kumar, T. (2010) 2158–2165. J. Bunning, Q. Li, Mater. Today 20 (2017) 230–237. [54] B. He, V. Martin, F. Setterwall, Fluid Phase Equilib. 212 (2003) 97–109. [38] Y. Xie, F. Guan, Z. Li, Y. Meng, J. Cheng, L. Li, Q. Pei, Macromol. Rapid Commun. [55] B. Kim, K. Yoo, S. Kim, J.S. Park, D. Gi Seong, J. Yoon, Chem. Eng. J. 443 (2022), 41 (2020), 2000290. 136471. [39] Y. Liu, J. Fan, R. Plamthottam, M. Gao, Z. Peng, Y. Meng, M. He, H. Wu, Y. Wang, [56] Y.J. Hwang, S.B. Pyun, M.J. Choi, J.H. Kim, E.C. Cho, ChemNanoMat 8 (2022), T. Liu, C. Zhang, Q. Pei, Chem. Mater. 33 (2021) 7232–7241. e202200005. [40] X.P. Zhao, S.A. Mofid, T. Gao, G. Tan, B.P. Jelle, X.B. Yin, R.G. Yang, Mater. Today [57] H. Zhang, J. Liu, F. Shi, T. Li, H. Zhang, D. Yang, Y. Li, Z. Tian, N. Zhou, Chem. Eng. Phys. 13 (2020), 100205. J. 431 (2022), 133353. [41] Y. Cui, Y. Ke, C. Liu, Z. Chen, N. Wang, L. Zhang, Y. Zhou, S. Wang, Y. Gao, Y. Long, [58] P. v Rathod, J.M.C. Puguan, H. Kim, Chem. Eng. J. 455 (2023) 140874. Joule 2 (2018) 1707–1746. [59] J.M.C. Puguan, P. v Rathod, P.P. More, H. Kim, Chem. Eng. J. 437 (2022) 135157. [42] X. Zou, H. Ji, Y. Zhao, M. Lu, J. Tao, P. Tang, B. Liu, X. Yu, Y. Mao, Nanomaterials [60] S.J. Lee, S.H. Lee, H.W. Kang, S. Nahm, B.H. Kim, H. Kim, S.H. Han, Chem. Eng. J. 11 (2021) 3335. 416 (2021), 129028. [43] Z. Qiu, S. Wang, Y. Wang, J. Li, Z. Xiao, H. Wang, D. Liang, Y. Xie, Compos. Sci. [61] L. Zhang, Y. Du, F. Xia, Y. Gao, Chem. Eng. J. 455 (2023), 140849. Technol. 200 (2020), 108407. [62] B. Sepúlveda, P.C. Angelomé, L.M. Lechuga, L.M. Liz-Marzán, Nano Today 4 [44] R. Yoshikawa, M. Tenjimbayashi, S. Shiratori, A.C.S. Appl, Energy Mater. 1 (2018) (2009) 244–251. 1429–1434. [63] J.R. Otaegui, A. Carrascull-Marín, D. Ruiz-Molina, J. Hernando, C. Roscini, Adv. [45] J.R. Otaegui, D. Ruiz-Molina, L. Latterini, J. Hernando, C. Roscini, Mater. Horiz. 8 Opt. Mater. 10 (2022), 2200083. (2021) 3043–3054. [64] Y. Guan, Z. Xue, J. Liang, Z. Huang, W. Yang, Colloids Surf. A Physicochem. Eng. [46] Q. Chang, Z. Shen, Z. Guo, C. Xue, N. Li, J. Yang, S. Hu, ACS Appl. Mater. Interfaces Asp. 502 (2016) 6–12. 13 (2021) 12375–12382. [65] K.C. Grabar, R. Griffith, M.B. Freeman, M.J.N. Hommer, Anal. Chem. 67 (1995) [47] Z. Xu, S. Wang, X.-Y. Hu, J. Jiang, X. Sun, L. Wang, Solar RRL 2 (2018), 1800204. 735–743. 12