A structural polymer for highly efficient all-day passive radiative cooling — TEI-tagged view

Tong WangpersName, Yi WupersName, Lan ShipersName, Xinhua HupersName, Min ChenpersName, Limin WupersName (2021date). “A structural polymer for highly efficient all-day passive radiative cooling.” Nature Communicationstitle·j.

Object of study — a hierarchically porous PMMAterm·dom film for day- and nighttime passive radiative coolingterm·dom

Research question — how to make efficient, low-cost radiative coolers that work all day and in all climates

Contribution — a porous polymer film whose microporesterm·key and nanoporesterm·key raise solar reflectanceterm·mth and thermal emittanceterm·key

Full text — tagged

A structural polymer for highly efficient all-day passive radiative coolingterm·dom Abstract All-day passive radiative coolingterm·dom has recently attracted tremendous interest by reflecting sunlight and radiating heat to the ultracold outer space. While some progress has been made, it still remains big challenge in fabricating highly efficient and low-cost radiative coolers for all-day and all-climates. Herein, we report a hierarchically structured polymethyl methacrylateterm·dom (PMMAterm·dom) film with a microporeterm·key array combined with random nanoporesterm·key for highly efficient day- and nighttime passive radiative coolingterm·dom. This hierarchically porous array PMMAterm·dom film exhibits sufficiently high solar reflectanceterm·mth (0.95) and superior longwave infraredterm·dom thermal emittanceterm·key (0.98) and realizes subambient cooling of ~8.2 °C during the night and ~6.0 °C to ~8.9 °C during midday with an average cooling powerterm·key of ~85 W/m 2 under solar intensity of ~900 W/m 2 , and promisingly ~5.5 °C even under solar intensity of ~930 W/m 2 and relative humidityterm·key of ~64% in hot and moist climate. The microporesterm·key and nanoporesterm·key in the polymer film play crucial roles in enhancing the solar reflectanceterm·mth and thermal emittanceterm·key.

Subject terms: Polymer synthesis, Polymers Introduction Although the climate-science community is attempting to find efficient solutions for the accelerating global warming and greenhouse gas emissions, few concrete actions have been taken to resolve climate change 1 – 3 . Conceptually, one of the most efficient strategies is to reduce the amount of solar irradiance absorbed by the Earth 4 , e.g., through solar radiation managementterm·mth (SRMterm·mth), to slow or reverse global warming 5 – 7 . The basic idea behind SRMterm·mth is to seed reflective particles into the Earth’s stratosphere to reduce solar absorption, which might cause potentially dangerous threats to the Earth’s basic climate operations 8 , 9 . One possibly alternative approach is passive radiative coolingterm·dom—a sky-facing surface on the Earth spontaneously cools by radiating heat to the ultracold outer space through the atmosphere’s longwave infraredterm·dom (LWIRterm·dom) transparency window (λ ~ 8–13 μm) 10 – 14 . However, passive daytime radiative coolingterm·dom (PDRCterm·dom) to a temperature below ambient under direct sunlight is a particular challenge because most of the naturally available thermal radiation materials also absorb incident solar irradiance and rapidly heat up under exposure to the Sun 15 , 16 . Accordingly, designing and fabricating efficient PDRCterm·dom with sufficiently high solar reflectanceterm·mth ρ ¯ solar (λ ~ 0.3–2.5 μm) to minimize solar heat gain and simultaneously strong LWIRterm·dom thermal emittanceterm·key ε ¯ LWIRterm·dom to maximize radiative heat loss is highly desirable 17 , 18 . When the incoming radiative heat from the Sun is balanced by the outgoing radiative heat emission, the temperature of the Earth can reach its steady state 4 , 19 . Thus, PDRCterm·dom technology is very promising to considerably decrease the use of compression-based cooling systems (e.g., air conditioners) and has a significant impact on global energy consumption 20 – 24 .

The first theoretical design of a metal-dielectric photonic structure for PDRCterm·dom was presented by Raman et al. 25 in 2013 by tailoring the material spectrum responses for continuous daytime radiative cooling. Then, they first experimentally achieved PDRCterm·dom via a precision-designed nanophotonic radiative cooler in 2014 26 . This cooler, consisting of seven alternating dielectric layers deposited on top of a silver mirror, cooled to 4.9 °C below ambient temperature by reflecting 97% of incident sunlight while strongly and selectively emitting in the atmospheric transparency windowterm·key. Nonetheless, many photonic structures suffer from a high manufacturing cost and large-scale production limits 23 , 26 – 28 . Another pioneering strategy was recently developed by Yin et al. 29 , who created a glass-polymer hybrid metamaterial thin film consisting of silica (SiO 2 ) microspheres randomly distributed in the matrix material of polymethylpentene via the scalable-manufactured roll-to-roll polymer extrusion process. With the assistance of a silver coating, the metamaterial was able to exhibit >93% infrared emissivity and reflect ~96% of solar irradiance, achieving a noontime radiative cooling powerterm·key of 93 W/m 2 under direct sunshine. The introduction of polymer-based radiative cooling materials can greatly improve the scalability and applicability of PDRCterm·dom systems in practical applications 18 , 30 – 33 . Very recently, instead of using a reflective metallic mirror, state-of-the-art PDRCterm·dom designs, such as porous polymer coatings 17 , 34 , 35 , polymeric aerogels 36 , white structural wood 37 and cooling paints 38 , 39 , have attracted considerable attention because of their high cooling performance, simplicity, applicability and economical efficiency. For example, Yu et al. 17 made remarkable progress in the design of PDRCterm·dom poly(vinylidene fluoride-co-hexafluoropropene) coatings with random micro-/nano-pores through a phase inversion-based method, demonstrating high solar reflectanceterm·mth (0.96 ± 0.03), as well as high longwave infraredterm·dom emittance (0.97 ± 0.02) that enabled cooling up to ~6 °C and ~3 °C below ambient temperature under direct sunlight in dry southwestern USA and south Asia, respectively. Nonetheless, almost all of the PDRCterm·dom prototypes reported so far performed well in an arid atmosphere (e.g., total precipitable water (TPW) <10 mm) and at relatively low environmental temperatures (e.g., below 25 °C) 26 , 29 , 36 , 37 , rather than in hot (e.g., 30 °C or above) and humid (e.g., TPW å 20 mm) regions 17 , 30 . Theoretically, atmosphere humidityterm·key or TPW heavily influences the absorbed power density of atmospheric radiation and the magnitude of the cooling performance. In a dry environment, the atmospheric window is open not only in the 1st atmospheric transparency windowterm·key (8–13 µm) but also in the 2nd atmospheric transparency windowterm·key (16–25 µm), making efficient thermal radiation possible 23 , 30 . However, in a hot and humid environment, the transmissivity of the atmosphere slightly decreases in the 1st atmospheric transparency windowterm·key and dramatically drops in the 2nd atmospheric transparency windowterm·key. Thus, the cooling performance naturally becomes lower and is even limited due to the increased downwelling atmospheric radiationterm·dom induced by higher humidityterm·dom and temperature 40 – 42 . Therefore, developing facile, scalable, and cost-effective PDRCterm·dom for practical thermal radiation applications, including hot and humid regions, still remains a great challenge.

Inspired by these intelligent reports and above considerations, herein, we design and report a hierarchically porous array PMMAterm·dom (PMMAterm·dom HPA ) film with a close-packed microporeterm·key array on the surface combined with abundant random nanoporesterm·key inside by a template method. The as-obtained PMMAterm·dom HPA film demonstrates excellent ρ ¯ solar (0.95) and ε ¯ LWIRterm·dom (0.98), with not only subambient cooling of as high as ~8.2 °C during the night and of ~6.0 °C to ~8.9 °C during midday with an average cooling powerterm·key of ~85 W/m 2 under solar intensity of ~900 W/m 2 , but also ~5.5 °C even under solar intensity of ~930 W/m 2 and relative humidityterm·key of ~64% in hot and moist subtropical marine monsoon climate, which was not reported previously. Both experimental evidence and theoretical calculations verify that both the dense microporeterm·key array on the surface and the random nanoporesterm·key in the PMMAterm·dom HPA film play crucial roles in enhancing the solar reflectanceterm·mth and thermal emittanceterm·key, which provides deep insight into devising superb PDRCterm·dom technologies.

Results and discussion ρ ¯ solar and ε ¯ LWIRterm·dom of the PMMAterm·dom HPA film Fig. 1a illustrates the fabrication process of PMMAterm·dom HPA . In brief, a monolayer of a hexagonally close-packed 5 μm SiO 2 array is fabricated using a facile unidirectional rubbing assembly method 43 , 44 , as shown in Fig. 1b, c . A dispersion of PMMAterm·dom and 200 nm SiO 2 nanospheres in acetone is then infiltrated into the SiO 2 monolayer template. After the rapid evaporation of acetone in air, the obtained PMMAterm·dom/SiO 2 composite displays randomly distributed SiO 2 nanospheres and regularly distributed SiO 2 microspheres, as indicated by the micrograph and EDS elemental mappings (Fig. 1d–g ). After removal of the SiO 2 nanospheres and the monolayer template by etching in hydrofluoric acid aqueous solution, a hierarchically porous array PMMAterm·dom (PMMAterm·dom HPA ) film with ordered symmetrical microporesterm·key (~4.6 μm diameter) and randomized nanoporesterm·key (~250 nm average diameter) can be obtained (Fig. 1h, i and Supplementary Figs. 1a–c ). Fourier transform infrared (FTIR) spectra and thermogravimetric analysis (TGA) confirm that PMMAterm·dom HPA is entirely composed of organic polymer without residual inorganic SiO 2 (Supplementary Figs. 1d and e ), while the 49.4 wt% of SiO 2 before etching indicates a high porosityterm·mth of PMMAterm·dom HPA (Supplementary Fig. 1f ).

Fabrication and characterization of the PMMAterm·dom HPA film. a Schematic illustration of the fabrication of PMMAterm·dom HPA with a hierarchically porous array. b , c SEM micrographs of hexagonally close-packed monolayer SiO 2 templates. d SEM micrograph of PMMAterm·dom/SiO 2 composite. e – g EDS elemental mappings of Si, O, and C in d , showing the randomly distributed SiO 2 nanospheres and regularly distributed SiO 2 microspheres of the composite. h , i SEM micrographs of PMMAterm·dom HPA showing an ordered symmetrical microporesterm·key array made of hierarchical randomized nanoporesterm·key.

Figure 2 demonstrates the spectral reflectance and emissivity of the PMMAterm·dom HPA film with ~160 μm effective thickness according to the normalized ASTM G173 Global solar spectrum and the LWIRterm·dom atmospheric transparency windowterm·key. The PMMAterm·dom HPA film with ~60% porosityterm·mth presents a high average solar reflectanceterm·mth ( ρ ¯ solar = 0.95, Fig. 2a and Supplementary Fig. 2a ), which ensures excellent reflection of sunlight from all incidences and minimizes the solar heat gain. Meanwhile, the film also shows a high thermal emittanceterm·key over a broad bandwidth in the mid-infrared and still emits a significant part of its thermal energy even at large emission angles ( ε ¯ LWIRterm·dom = 0.98, Fig. 2b and Supplementary Fig. 2b ), which can enable strong emission of heat to the cold sink of outer space through the atmospheric transparency windowterm·key at different angles in relation to the sky. One remarkable feature of our structural polymer is that the periodic ordered ~4.6 μm microporesterm·key can scatter sunlight of ultraviolet-visible-near-infrared (UV-Vis-NIR) wavelengths, and the abundant, random ~250 nm nanoporesterm·key greatly reduce the mean scattering path and transmission through the material, which further enhances the scattering of shorter visible wavelengths 45 , 46 . The combination of PMMAterm·dom with air voids, one of which has a high refractive index and the other has a low refractive index ( Δ n = n PMMAterm·dom − n air = 1.49 − 1 = 0.49 ) , could provide a sharp refractive index transition across polymer-air boundaries and yield the efficient solar scattering and the required strong sky window absorptance without surface obstruction 34 .

Spectroscopic response of the PMMAterm·dom HPA film. a , b Spectral reflectance and emissivity of the PMMAterm·dom HPA film with ~160 μm effective thickness along with the normalized ASTM G173 Global solar spectrum and the LWIRterm·dom atmospheric transparency windowterm·key. c , d Spectral refractive index (n) and extinction coefficient (κ) of PMMAterm·dom, showing negligible absorptivity in the solar range and multiple extinction peaks in the LWIRterm·dom wavelengths. e Absorbance spectrum of PMMAterm·dom HPA measured with ATR-FTIR spectroscopy. f Schematic diagram showing the periodic re-entrant structure with nano/microscale pores is conducive to enhance the total scattering efficiency by multiple reflections.

As one of the most widely used and low-cost polymers, pristine PMMAterm·dom film has ideal intrinsic properties to enable high-performance PDRCterm·dom applications 47 . Figure 2c, d show that PMMAterm·dom has negligible extinction coefficient in the solar wavelengths and multiple extinction peaks at the 8, 8.6, 10.3, 11.8, and 13.2 μm within the LWIRterm·dom window, which should result from the different vibrational modes of its molecular structure. These properties keep the heat gain from sunlight to a minimum and contribute to a large amount of infrared absorption/emission in the atmospheric transparency windowterm·key, which is responsible for the superior PDRCterm·dom. Furthermore, the absorbance spectrum measured with attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR) exhibits strong infrared absorption due to C–O–C stretching vibrations between 770 and 1250 cm −1 (8–13 μm, Fig. 2e ), which coincidently lie in the atmospheric transparency windowterm·key. Importantly, the periodic re-entrant structure with hierarchical nano/microscale pores is conducive to enhance the total scattering efficiency and increase the probability of infrared absorption/emission through multiple diffuse reflection at various incident angles (Fig. 2f ). All these features benefit the radiative heat exchange between the cooling structural polymer and the atmosphere, causing sufficiently high ρ ¯ solar and ε ¯ LWIRterm·dom of our PMMAterm·dom HPA film to achieve all-day passive radiative coolingterm·dom.

Different from transparent pristine PMMAterm·dom, the PMMAterm·dom/SiO 2 composite film is translucent due to the absorptance in the UV and visible regions, and the corresponding PMMAterm·dom HPA film has negligible transmittance because its plenty of pores can efficiently scatter sunlight of all wavelengths (Supplementary Figs. 3a–d ). The spectral transmittance of the PMMAterm·dom HPA film decreases with increasing thickness (Supplementary Fig. 4 ). Accordingly, the solar reflectanceterm·mth appears to have a more pronounced increasing trend with thickness than the thermal emittanceterm·key in the range of 8–13 μm (Fig. 3a–c ), which likely arises from the increased backscattering of light from the thicker, nonabsorptive, porous PMMAterm·dom layer. Evidently, even the emittance of the only effectively ~2 μm thick PMMAterm·dom HPA film can reach 0.85. This suggests that the porous structure is sufficient to augment the intrinsic emittance of PMMAterm·dom HPA . Further, we experimentally and theoretically demonstrated the influence of pore sizes and porosityterm·mth on the optical performance. As clearly seen in Supplementary Fig. 5a , the PMMAterm·dom HPA with ~500 nm nanoporesterm·key shows a high-level ρ ¯ solar , but its scattering of short wavelength region is inferior to that of ~200 nm nanoporesterm·key. The reflectance of PMMAterm·dom HPA with ~100 nm nanoporesterm·key drops gradually in the visible and NIR-to-SWIR ranges (0.4–2.5 μm), leading to a lower ρ ¯ solar . Such results coincide with the finite-difference time-domain (FDTD) simulation data (Supplementary Fig. 6a ), indicating nanoporesterm·key ranging in diameter from 200 to 300 nm are optimum to reinforce the scattering of UV and visible wavelength region. For the optimization of microporeterm·key sizes, we established a model of ordered monolayer microporeterm·key with different sizes in the range of 1 to 10 µm. Given the simulation results, one could derive that microporesterm·key ranging from 5 to 7 µm in diameter would highly contribute to the solar reflectanceterm·mth (Supplementary Fig. 6b ). Thus, combining the optimized nanoporeterm·key size of 200–300 nm and microporeterm·key size of 5–7 µm, the maximum ρ ¯ solar of our hierarchical porous PMMAterm·dom film can be obtained (Supplementary Fig. 6c ). Moreover, the solar reflectanceterm·mth of the PMMAterm·dom HPA film increases with increasing porosityterm·mth, while its thermal emittanceterm·key does not significantly change (Supplementary Fig. 5b ). FDTD simulation results also show that the nanoporeterm·key density or porosityterm·mth has a positive correlation with the reflectance (Supplementary Fig. 6d ). However, too high porosityterm·mth may reduce the mechanical strength. Based on the trade-off of the optical and mechanical properties, ~ 60% porosityterm·mth was adopted in the experimentterm·mth design and radiative cooling application.

Variation in ρ ¯ solar and ε ¯ LWIRterm·dom of the PMMAterm·dom HPA films with effective thickness and polarization angle. a Variation in ρ ¯ solar and ε ¯ LWIRterm·dom of the PMMAterm·dom HPA films with effective thickness. The error bars represent the standard deviation. b , c Measured reflectance and emissivity spectra of the PMMAterm·dom HPA films as a function of the film effective thickness. d Infrared emissivity spectra of a PMMAterm·dom HPA film at different polarization angles (θ) from 10° to 80°. e Polar distribution of the average emissivity across the atmospheric window of the PMMAterm·dom HPA film at different polarization angles (θ) from 10° to 80°. f Measured polarization-dependent infrared emissivity spectra of the PMMAterm·dom HPA films.

Figure 3d–f demonstrate the emissivity spectra of the PMMAterm·dom HPA film in the infrared range of 2.5–20 µm at different polarization angles. ε ¯ LWIRterm·dom shows a regularly decreasing trend from 10° to 80°, mainly owing to the ordered periodic microporeterm·key array on the top surface of the PMMAterm·dom HPA film. Excitingly, the average emissivity across the atmospheric window ε ¯ LWIRterm·dom is greater than 0.95 over a wide polarization angle range from 10° to 80°, indicating a stable emitted heat flux through the atmospheric transparency windowterm·key to the cold sink of outer space.

To investigate the influence of the hierarchical porous structureterm·dom on the optical properties of the PMMAterm·dom HPA film, we further compared it with three other types of PMMAterm·dom films, nanoporeterm·key PMMAterm·dom NP , monolayer microporeterm·key array PMMAterm·dom MPA and pristine PMMAterm·dom, as shown by the SEM images in Fig. 4a–d . Figure 4e shows that the PMMAterm·dom HPA film presents the highest average solar reflectanceterm·mth ( ρ ¯ HPA = 0.95) in UV-Vis-NIR wavelengths, while PMMAterm·dom NP drops gradually to a lower average solar reflectanceterm·mth ( ρ ¯ NP = 0.74) in the near-to-short wavelength infrared (NIR-to-SWIR) range (0.7–2.5 μm) because only these disordered nanoporesterm·key are too small to effectively scatter such wavelengths. In contrast, PMMAterm·dom MPA exhibits a low value at all wavelengths ( ρ ¯ MPA = 0.23) because the scattering of uniform monolayer microporesterm·key is weak. Similar results can be observed for pristine PMMAterm·dom ( ρ ¯ prestine = 0.10) due to its high transparency. The emission spectra show that all four films have strong thermal emittancesterm·key in the LWIRterm·dom transparency window ( ε ¯ HPA = 0.98, ε ¯ NP = 0.96, ε ¯ MPA = 0.95 and ε ¯ prestine = 0.92), as shown in Fig. 4f . However, the emissivity of PMMAterm·dom MPA and pristine PMMAterm·dom substantially drops in the range of mid-IR wavelengths compared to the PMMAterm·dom HPA and PMMAterm·dom NP films. This probably because that the existence of abundant hierarchically porous structure reduces the effective refractive index and leads to a more gradual refractive index transition across the polymer-air interface than PMMAterm·dom MPA and PMMAterm·dom 17 . Thus, the impedance matching between the porous polymer and surrounding air is improved 48 , which reduces the surface reflectance and results in a consistently higher emissivity for PMMAterm·dom HPA or PMMAterm·dom NP film in the mid-IR wavelengths. Besides, as we mentioned above, the porous structure with high specific surface area might increase the probability of infrared absorption through multiple diffuse reflection and enhances the emissivity.

SEM micrographs and optical properties of different types of PMMAterm·dom films. a – d SEM micrographs of the PMMAterm·dom HPA , PMMAterm·dom NP , PMMAterm·dom MPA , and pristine PMMAterm·dom films. e Reflectance spectra across the solar wavelengths of different types of PMMAterm·dom films. f Infrared emissivity spectra of different types of PMMAterm·dom films.

The experimental results of four types of PMMAterm·dom films were also theoretically verified by FDTD simulations (Supplementary Fig. 7 ) 49 . The numerical simulation results further reveal that hierarchical porous structureterm·dom containing dual-scale nano/micro cavities would significantly improve the broadband scattering performance contrast to uniform nanoscale porous structureterm·dom, especially in the range of NIR-to-SWIR, while the LWIRterm·dom thermal emissivity of the two kinds of structures has negligible changes. The theoretical model of PMMAterm·dom MPA verifies that the ordered microporesterm·key monolayer does not have enough high scattering coefficient, which matches well with the experimental result. We also simulated the reflectance spectra across the solar wavelengths of the PMMAterm·dom HPA film with thickness of ~5 µm (effectively ~2 µm, Supplementary Fig. 8 ). Surprisingly, such thin film is sufficient to yield an efficient scattering ( ρ ¯ HPA − 2 μm = 0.35), which further evidences the enhancement effect of hierarchical porous surface on the solar reflectanceterm·mth.

Besides, to verify the optical superiority of the periodic microporesterm·key array on the surface, we also investigated the optical properties of a hierarchically porous PMMAterm·dom (PMMAterm·dom HP ) film with loose-packed random microporesterm·key and nanoporesterm·key (Supplementary Figs. 9a–c ). The solar reflectanceterm·mth of the PMMAterm·dom HP film drops 8% in the range of NIR-to-SWIR compared to PMMAterm·dom HPA (Supplementary Fig. 9d ), which contains about 5% of sunlight ( ρ ¯ HP = 0.90). This suggests that the close-packed periodic arrangement of the hierarchical nano/microscale pores on the surface of our PMMAterm·dom HPA can maximize both the surface area and the amount of scatters per unit and increase the overall scattering efficiency, especially in NIR-to-SWIR range, although this distribution of microporesterm·key on the surface does not influence the infrared emissivity (Supplementary Fig. 9e ).

Furthermore, the PMMAterm·dom HPA film was modified by fluorosilane to become superhydrophobic with a water contact angle (WCA) of ~156° (Supplementary Fig. 10 ) for stable durability in various atmospheric humidityterm·key conditions. Both the solar reflectanceterm·mth and infrared emissivity are at high levels and vary negligibly after fluorosilane treatment (Supplementary Figs. 11a and b ). Even after accelerated weathering treatment for 480 h (each cycle including the UV irradiation at 310 nm wavelength with intensity of 0.71 W/m 2 at 60 °C for 4 h, followed by condensation at 50 °C for 4 h with UV lamps off), the PMMAterm·dom HPA films exhibit no blistering, peeling, cracking and color changing. The porous morphology of our PMMAterm·dom HPA films basically remains the same (Supplementary Figs. 12a and b ). The ATR-FTIR spectra further demonstrate that the surface-modified PMMAterm·dom HPA films before and after weathering treatment all have obvious absorption peaks of C=O at 1726 cm −1 , C–F at 1187 cm −1 , C–O–C at 1139 cm −1 and C–Cl at 779, 746, 700, and 651 cm −1 (Supplementary Fig. 12c ), owing to the protection of fluorosilane molecules. Moreover, the constant WCAs in Supplementary Fig 12d after weathering treatment also indicate the excellent durability and potential applicability. The reflectance and emissivity spectra of the PMMAterm·dom HPA films show slight fluctuations of solar reflectanceterm·mth ( ρ ¯ solar = 0.95 ± 0.02) and negligible variations in emissivity ( ε ¯ LWIRterm·dom = 0.98 ± 0.01) with the accelerated weathering time (Supplementary Figs. 12e and f ), which may be attributed to the inconspicuous yellowing effect in response to the UV irradiation. We also conducted the real outdoor exposure test for 40 days on a flat roof in Shanghai city and the results indicate the almost unchanged optical performance and WCAs (Supplementary Table S1 ).

All-day continuous passive radiative coolingterm·dom measurements of the PMMAterm·dom HPA film The radiative cooling performances of the structural polymers (~160 μm effective thickness and 100 mm × 100 mm in size) were measured from 20:00 on 09 Oct. to 20:00 on 10 Oct. 2019 using a 24-h uninterrupted thermal measurement on a flat roof of a five-story building under a clear sky in Shanghai, China. As shown in Fig. 5a , the strong optical scattering of sunlight gives our PMMAterm·dom HPA surface a matte and white appearance, which can be further confirmed by the CIE chromaticity coordinate analysis (Fig. 5b ). Under intense solar irradiance of ~930 W/m 2 and relative humidityterm·key of ~40% at noon (Fig. 5c ), the real-time temperature tracking of the air and four types of PMMAterm·dom films are shown in Fig. 5d . Evidently, although the pristine PMMAterm·dom and PMMAterm·dom MPA films can exhibit decreases in temperature by ~3.7 °C and ~4.9 °C at night, respectively, their temperatures dramatically rise to ~15.2 °C and ~14.1 °C above the ambient temperature at noon due to their transparency and translucency, respectively (Fig. 5e ). The PMMAterm·dom NP film can achieve a subambient cooling of ~6.5 °C during the night but maintains almost the same temperature as the ambient environment during midday. In contrast, the PMMAterm·dom HPA film exhibits fabulous passive radiative coolingterm·dom during both night and daytime. The average below-ambient temperature of the PMMAterm·dom HPA film is ~8.2 °C during the night (between 6 p.m. and 6 a.m.) and ~6.0 °C during midday (between 11 a.m. and 2 p.m.). This cooling performance is on par with those in previous reports (Supplementary Table S2 ).

Twenty-four-hour continuous passive radiative coolingterm·dom performance measurements. a Photograph of the cooling PMMAterm·dom HPA film showing its bright white appearance (inset, water contact angle image of PMMAterm·dom HPA ). b CIE chromaticity coordinates of the cooling PMMAterm·dom HPA film. c Relative humidityterm·key tracking of the air and solar irradiance at 12 PM on 10 Oct. 2019. d Temperature tracking of the air and PMMAterm·dom films. e Temperature difference between the ambient and structural PMMAterm·dom films.

Subambient PDRCterm·dom performance in various geographical regions and climates Radiative cooling performance in real-world applications is substantially affected by the geographical regions and climates. For instance, the net radiative cooling powerterm·key of the cooler is limited by the increased solar irradiation, humidityterm·key, cloud cover, local wind speed and ambient temperature 30 , 50 . Here, we further performed a series of experimentsterm·mth to evaluate and compare the effects of climates from different cities on the radiative cooling performance of our PMMAterm·dom HPA films. Three cities, Xiamen city (Southern China, Coastal), Shanghai city (Eastern China, Coastal) and Xuzhou city (Northern China, Inland), were chosen as typical test locations due to their different topographic and meteorological characteristics (Supplementary Fig. 13 ). The climate characteristics in Xuzhou, Shanghai and Xiamen are temperate monsoon climate, subtropical monsoon climate and subtropical marine monsoon climate, respectively, which provide a remarkable difference in humidityterm·key for our measurements. As shown in Fig. 6a–i , under clear skies with comparable solar intensity and wind speed but various humidityterm·key conditions (~38% in Xuzhou, ~47% in Shanghai, and ~54% in Xiamen), the subambient cooling of ~8.9 °C, ~7.8 °C and ~8.6 °C at noontime is observed, separately. Even in the same location (e.g., in Xiamen city), increasing solar intensity (from ~890 W/m 2 to ~930 W/m 2 ), wind speed (from ~0.6 m/s to ~1.1 m/s) and humidityterm·key (from ~54% to ~64%), the PMMAterm·dom HPA film still enables cooling up to ~5.5 °C below ambient temperature (Figs. 6g–l ), which has not been reported previously 40 , 41 . Comparing Xuzhou and Xiamen city (Fig. 6a–c and j–l ), we can conclude that increasing solar intensity (from ~860 W/m 2 to ~930 W/m 2 ), wind speed (from ~0.5 m/s to ~1.1 m/s) and humidityterm·key (from ~38% to ~64%), the subambient cooling temperatures of PMMAterm·dom HPA film are indeed influenced (decreasing from ~8.9 °C to ~5.5 °C). The primary driver of the good performance in various geographical regions and climates should be the synergistic result of visible white (high solar reflectanceterm·mth) and infrared black (high infrared emissivity in both the 1st and 2nd atmospheric transparency windowterm·key) that greatly minimizes the absorbing solar irradiance and the thermal radiation emitted by the atmosphere. In addition, we investigated the effect of surface modification on the daytime radiative cooling performance of the PMMAterm·dom HPA film. The results show that the PMMAterm·dom HPA films with and without surface modification can achieve subambient cooling temperatures of ~6.9 °C and ~6.5 °C, respectively (Supplementary Fig. 14 ). While the PMMAterm·dom HPA film without surface modification also demonstrates excellent passive radiative coolingterm·dom behavior, the fluorosilane treatment enables stable performance by restricting the effect of moisture and water under different levels of humidityterm·key.

Passive daytime radiative coolingterm·dom performance of the PMMAterm·dom HPA film in different locations and weathers. a–c Solar irradiance, wind speed and relative humidityterm·key, and temperatures of the air and the PMMAterm·dom HPA film in Xuzhou city. d–f Solar irradiance, wind speed and relative humidityterm·key, and temperatures of the air and the PMMAterm·dom HPA film in Shanghai city. g–i Solar irradiance, wind speed and relative humidityterm·key, and temperatures of the air and the PMMAterm·dom HPA film in Xiamen city. j–l Solar irradiance, wind speed and relative humidityterm·key, and temperatures of the air and the PMMAterm·dom HPA film in Xiamen city.

Passive radiative coolingterm·dom power measurements of the PMMAterm·dom HPA film When used in a building roof or external siding, our PMMAterm·dom HPA films can achieve all-day passive radiative coolingterm·dom through reflecting sunlight and radiating heat to the cold outer space under a clear sky (Fig. 7a ). Figure 7b schematically shows the direct thermal measurement system based on the net cooling equation. To further demonstrate the PDRCterm·dom capability of the cooling PMMAterm·dom HPA , we adopted a feedback-controlled heating system to measure its radiative cooling powerterm·key during the midday (Fig. 7c ). This feedback-controlled heating system maintains the surface temperature of the PMMAterm·dom HPA film at the measured ambient temperature to minimize the impact of conductive and convective heat losses (Fig. 7d ). Promisingly, the PMMAterm·dom HPA film attains an average cooling powerterm·key of ~85 W/m 2 under solar intensity of ~900 W/m 2 in April in Shanghai (Fig. 7e ).

Net cooling powerterm·key of the PMMAterm·dom HPA film during the midday. a Schematic of the basic principles of PDRCterm·dom. When used in a building roof or external siding, the PMMAterm·dom HPA film exhibits high solar reflectanceterm·mth and high infrared emissivity. b Two-dimensional schematic drawing of the thermal box apparatus with a feedback-controlled heater. The heater maintains the sample surface temperature at that of the ambient environment, minimizing convective and conductive heat losses. c Photo of the experimental apparatus on a rooftop in Shanghai, China. d Temperature tracking of the ambient and the PMMAterm·dom HPA film under solar intensity of ~ 900 W/m 2 and relative humidityterm·key of ~ 44% on April 23, 2020. e Continuous measurement of the radiative cooling powerterm·key of the PMMAterm·dom HPA film during the midday.

Figure 8a, b present the net cooling powerterm·key during the nighttime and daytime calculated using the radiative cooling theoretical model, respectively. More details of this model are given in the “Methods” section. The power of solar radiation is set to approximately 1000 W/m 2 for simplicity, and the ambient temperature T amb is assumed to be 298.15 K in both cases. A maximum cooling powerterm·key of 124.40 W/m 2 can be achieved for nighttime operation. For daytime operation, the calculated maximum net cooling powerterm·key is 74.40 W/m 2 at thermal equilibrium, which is lower than the measured daytime cooling powerterm·key due to the fluctuations in the ambient conditions, the uncertainty in the measurements and the theoretical model approximations.

Calculated net cooling powerterm·key with software MATLAB based on the theoretical simulation. a Calculated net cooling powerterm·key during the nighttime. b Calculated net cooling powerterm·key during the daytime. The variable h c is a combined nonradiative heat coefficient. Values of 0, 3, 6, 9, and 12 for h c are used in the calculations. For daytime calculations, 5% solar power absorption is considered.

In summary, we have demonstrated and fabricated a hierarchically structured PMMAterm·dom film with a dense microporeterm·key monolayer array and randomly distributed nanoporesterm·key for highly efficient all-day passive subambient radiative cooling in various geographical locations and climates. Our structural polymer film exhibits sufficiently high solar reflectanceterm·mth and thermal emittanceterm·key owing to its abundant periodic scattering microporesterm·key embedded with random nanoporesterm·key and ideal intrinsic properties. Without needing any silver or aluminum reflectors, our cooling structural PMMAterm·dom film realizes an average below-ambient temperature ~8.2 °C during the night and ~6.0 °C to ~8.9 °C during midday, and promisingly ~5.5 °C even under solar intensity of ~930 W/m 2 and relative humidityterm·key of ~64% in hot and moist subtropical marine monsoon climate, which is really an all-day and all-climate PDRCterm·dom system. And the superhydrophobized PMMAterm·dom HPA film can ensure cooling performance durability by eliminating the effect of moisture and water under different levels of humidityterm·key. This study has revealed the effects of microporesterm·key, nanoporesterm·key and their arrangement on optical performance, which may provide deep insight into the crucial roles of various pores and their array in solar reflectanceterm·mth and thermal emittanceterm·key and help us design and fabricate more efficient all-day passive subambient radiative cooling materials and systems.

Methods Fabrication of PMMAterm·dom HPA Monodisperse 5 μm SiO 2 microspheres were placed on top of a polydimethylsiloxane (PDMS)-coated glass sheet and rubbed with another PDMS substrate with slight palm pressure along a randomly chosen direction according to our reported method 43 , 44 . After being rubbed for 5 s, the SiO 2 microspheres had assembled into hexagonally close-packed monolayers on the PDMS surfaces. A 200 nm SiO 2 nanosphere dispersion in acetone was added to PMMAterm·dom to make a dispersion of SiO 2 -acetone-PMMA (1:10:1 mass ratio) under magnetic stirring at 50 °C for 2 h. This dispersion was then drop-cast onto the monolayer SiO 2 template. After the solvent was fully evaporated, a freestanding PMMAterm·dom/SiO 2 composite film was obtained by peeling the coating off the smooth surface. After removal of the SiO 2 template and nanospheres with 2-5 vol% hydrofluoric acid aqueous solution, PMMAterm·dom HPA films with various effective thicknesses of 2 ± 0.5 µm to 160 ± 5 µm were obtained by casting different amounts of solution. For comparison, PMMAterm·dom NP film was fabricated by the same procedure above without using the monolayer SiO 2 template, PMMAterm·dom MPA film was obtained without SiO 2 nanospheres, and the pristine non-porous PMMAterm·dom film was prepared at an acetone-PMMA (10:1 mass ratio) without using SiO 2 particles. In addition, PMMAterm·dom HP film was fabricated using random loose-packed monolayer 5 μm SiO 2 templates and 200 nm SiO 2 nanosphere. All the films above have the same effective thickness by controlling the PMMAterm·dom mass and film-forming substrate size.

PMMAterm·dom HPA surface modification To make superhydrophobic PMMAterm·dom HPA film, the sample was treated with 1H,1H,2H,2H-perfluorooctyltrichlorosilane (PFOTS) via a chemical vapor deposition (CVD) method. In detail, the PMMAterm·dom HPA film and a vial containing 1% PFOTS/ethanol solution were placed in a sealed vacuum desiccator. The desiccator was immediately pumped to vacuum for 15 min and placed at room temperature for 24 h for fluorination. The modified PMMAterm·dom HPA film was then baked in an oven at 80 °C for 1 h to remove the excessive PFOTS.

Characterization Optical characterization of the cooling structural polymers The spectral reflectance (ρ (λ)) and transmittance (τ (λ)) in the ultraviolet, visible and near-infrared (0.3–2.5 μm) wavelength ranges were separately determined in an UV-Vis-NIR spectrophotometer (Hitachi, U-4100, Japan) with a polytetrafluoroethylene integrating sphere. The ρ (λ) and τ (λ) in the mid-infrared wavelength ranges were characterized in an FTIR spectrometer (Nicolet 6700, Thermo Fisher Scientific, USA) equipped with a gold integrating sphere. We put one black substrate behind the samples during spectral reflectance measurement to eliminate the reflectance contribution of the substrate. The average value of more than five parallel measurements on different sites of the film was reported and the error bars represent the standard deviation. A polarizer was used in the FTIR spectrometer equipped with a smart diffuse reflectance accessory to measure the reflectance over a range of polarization angles (θ) from 10° to 80°. For any object at thermal equilibrium, the spectral absorptivity (α (λ)) and emissivity (ε (λ)) must be equal according to Kirchhoff’s law; thus, ε (λ) was calculated as ε (λ) = 1–ρ (λ)–τ (λ) 51 , 52 . The angular reflectance spectra in the wavelength range of 0.4–1.1 μm at different incidence angles were measured by an angle-resolved photonic spectral system (R1, Ideaoptics Technology Ltd., China). Refractive index and extinction coefficient measurements in the wavelength range of 0.2–14 μm at 60° were taken for pristine PMMAterm·dom films using a V-VASE and an IR-VASE ellipsometer (J. A. Woollam, USA).

Theoretical model of the radiative cooling performance When the structural polymer films are exposed to a clear sky, they are influenced by the solar irradiance and atmospheric downward thermal radiation. Meanwhile, heat can be transferred from the ambient surroundings to the polymers via conduction and convection due to the temperature difference between the cooling polymers and the ambient environment. To achieve PDRCterm·dom, the device must satisfy very stringent constraints as dictated by the power balance equation. The net cooling powerterm·key P cool of the structural polymers is expressed as 26 : where T is the surface temperature of the structural polymers and T amb is the ambient temperature. P rad ( T ) is the power radiated by the structural polymers, and P atm T amb is the absorbed atmospheric thermal radiation at T amb . P solar is the incident solar irradiation absorbed by the structural polymers, and P cond + conv is the power lost due to convection and conduction. The net cooling powerterm·key defined in Eq. ( 1 ) can reach a high value by increasing the radiative power of the structural polymers and reducing either the solar absorption or parasitic heat gain 34 . These parameters can be calculated by the following equations 26 , 53 – 55 : where A is the surface area of the radiative cooler. ∫ d Ω = 2 π ∫ 0 π / 2 d θ sin θ is the angular integral over a hemisphere. I BB T , λ = 2 h c 2 λ 5 1 e h c / λ κ B T − 1 is the spectral radiance of a blackbody at temperature T . h is Planck’s constant, κ B is the Boltzmann constant, and c is the speed of light. ε λ , θ is the directional emissivity of the surface at wavelength λ . ε atm λ , θ = 1 − τ λ 1 / con θ is the angle-dependent emissivity of the atmosphere; τ(λ) is the atmospheric transmittance in the zenith direction. P rad ( T ) and P atm T amb are determined by both the spectral data of the structural polymers and the emissivity spectrum of the atmosphere according MODTRAN of Mid-Latitude Summer Atmosphere Model 36 , 56 . In Eq. ( 4 ), the solar illumination is represented by the AM1.5 spectrum I AM 1.5 λ . For our PMMAterm·dom HPA film, approximately 95% of the input solar power can be reflected, and thus, the 5% absorption of the solar irradiance will reduce the net cooling powerterm·key. In Eq. ( 5 ), h c = h cond + h conv is a combined nonradiative heat coefficient that captures the collective effect of conductive and convective heating, which can be limited to a range between 0 and 12 W/m 2 /K.

The average solar reflectanceterm·mth ( ρ ¯ solar ) is defined as 17 : where λ is the wavelength of incident light in the range of 0.3–2.5 μm, I solar ( λ ) is the normalized ASTM G173 Global solar intensity spectrum, and ρ solar ( λ , θ ) is the surface’s angular spectral reflectance. The average emittance ( ε ¯ LWIRterm·dom ) in the LWIRterm·dom atmospheric transmittance window is defined as: where I BB ( λ ) is the spectral intensity emitted by a blackbody and ε LWIRterm·dom λ , θ is the surface’s angular spectral thermal emittanceterm·key in the range of 8–13 μm.

Thermal measurements of cooling temperature and cooling powerterm·key with a feedback-controlled heater We designed a feedback-controlled program to minimize both conductive and convective heat exchange to the structural polymers under strong solar irradiance based on Eq. ( 1 ) 29 . Our thermal box consisted of insulation foam covered by a layer of reflective foil. A 10-µm-thick transparent low-density polyethylene film was used to seal the thermal box and served as a wind shield. Moreover, a PMMAterm·dom HPA film with a size of 100 mm × 100 mm × 160 µm was placed on a 100 mm × 100 mm × 1 mm thick copper plate attached to a 100 mm × 100 mm × 0.2 mm thick Kapton heater. The Kapton heater was feedback-controlled by an ambient temperature-responsive thermostat to maintain the PMMAterm·dom HPA film at ambient temperature and accurately assess the cooling powerterm·key. The apparatus was elevated 1.1 meters above the ground to avoid heat conduction from the ground to the thermal box. An adhesive resistant temperature detector was directly mounted on the back surface of the polymer film to detect real-time temperature of the sample, which was continuously recorded by a datalogging thermometer with an uncertainty of ± 0.1 °C (CENTER309, CENTER corp, Taiwan, China). For comparison, a temperature detector was mounted outside the box to detect real-time temperature of the ambient. A relative humidityterm·key (RH) data logger with an accuracy of ± 0.1% RH (GSP-8, Elitech corp, China) was placed near the samples to measure the relative air humidityterm·key. The solar irradiation outside the box was simultaneously recorded using a datalogging solar radiometer with an accuracy of ± 5% (TES1333R, TES Electrical Electronic corp. Taiwan, China). It is worth mentioning that the sunlight our PMMAterm·dom films received is ~10% less than the pyranometer measured due to the polyethylene cover on the thermal box. The wind speed around our thermal boxes was measured using a digital anemometer with an accuracy of ±2.5% (AS856, Smart Sensor corp, China). All weather data were automatically tracked every 10 s. The heater was switched on to test the radiative cooling powerterm·key, but the heater and copper plate were removed to demonstrate the subambient cooling performance. As a control, we exposed the PMMAterm·dom NP , PMMAterm·dom MPA and pristine PMMAterm·dom films to the sky while mounted in the same apparatus to compare the subambient cooling performances. Demonstrations of the cooling performance of different types of PMMAterm·dom films during both day and night were carried out under a clear sky with a relative humidityterm·key of ~ 40% at noon in on a flat roof of a five-story building at Fudan University, Shanghai, China on October 09, 2019. The daytime cooling performance of the PMMAterm·dom HPA film in different locations and climates was conducted in April and May, 2020, such as Xiamen city (Southern China, Coastal, 24° 26’ 57” N, 118° 3’ 35” E), Shanghai city (Eastern China, Coastal, 31° 18’ 22” N, 121° 30’ 17” E) and Xuzhou city (Northern China, Inland, 34° 27’ 55” N, 117° 0’ 51” E).

Supplementary information Supplementary Information Peer Review File Acknowledgements We appreciate the financial support provided for this research by the National Key Research and Development Program of China (2017YFA0204600 and 2020YFE0100300) and the National Natural Science Foundation of China (52033003 and 51721002).

Author contributions L.W., T.W., and M.C. conceived the concept and designed the research. T.W. and Y.W. conducted the experimentsterm·mth. T.W. and L.S. conducted the FDTD simulations. T.W. created the schematics. L.W. and T.W. wrote the manuscript. All authors including X.H. discussed the results and commented on the manuscript.

Data availability All data needed to evaluate the conclusions in the paper are presented in the paper and/or the Supplementary information. Additional data related to this paper may be requested from the authors.

Competing interests The authors declare no competing interests.

Footnotes Supplementary information Supplementary information is available for this paper at 10.1038/s41467-020-20646-7.

References Associated Data Supplementary Materials Supplementary Information Peer Review File Data Availability Statement All data needed to evaluate the conclusions in the paper are presented in the paper and/or the Supplementary information. Additional data related to this paper may be requested from the authors.