Janus particles stabilized asymmetric porous composites for thermal rectification

Abstract

Thermal rectification is a noteworthy phenomenon of asymmetric material, which enables the directional transfer of thermal energy. But the design and construction of such asymmetric thermal conductive materials with complex structures are full of challenges. Here, an additive manufacturing method is proposed to fabricate asymmetric porous composites from layer-by-layer cast emulsions, stabilized with Janus particles (JPs), for thermal rectification. The emulsions are remarkably stable, allowing each layer to be manipulated independently without interference, resulting in a porous structure with significant asymmetry. The thermal rectification of JPs-stabilized asymmetric porous composites (JAPCs) is investigated through both experiments and simulations. It is found that their thermal rectification ratios can be adjusted by altering the contrast between the two layers of the asymmetric porous composites, with a maximum value of 38%. This emulsion casting additive manufacturing method is suitable for large-scale production. A simple model demonstrates the potential of JAPCs to regulate thermal energy in ambient conditions with fluctuating temperatures. It is explored to achieve multilayer alternating porous composites, which cannot be achieved with gradient asymmetric approaches. This method provides a practical way to design and fabricate complicated porous structures with potential applications in additive manufacturing.

Asymmetric thermal conductive materials enable the directional transfer of thermal energy and their design and synthesis are challenging. Here, the authors show an additive manufacturing method to fabricate asymmetric porous composites from layer-by-layer cast emulsions, stabilized with Janus particles for thermal rectification.

Introduction

Global carbon reduction highlights the urgency of thermal management strategies to improve energy utilization efficiency^1–4. Thermal rectification, one such strategy, features heat flowing more easily in one direction while being hindered in the opposite direction, offering a potential solution for improving energy efficiency^5–7. To achieve thermal rectification, it is essential to fabricate asymmetric structures within composites. So far, the asymmetric structures for thermal rectification have been fabricated using different methods and at various scales^8–18. The advancement in preparing asymmetric structures at the nanoscale is impeded by the complexity and high cost of nano-fabrication techniques^19–21. The two-sectional composites containing phase change materials (PCMs) show promise for achieving asymmetric structures and improving rectification performance^22,23. However, this method is still limited by the phase change temperature range and risk of PCM leakage.

The porous structure in the nanoscale is effective for adjusting thermal conductivity and facilitates thermal rectification^19,24. Nevertheless, the synthesis methods remain inadequate for the preparation of porous structures in the macroscale with significant asymmetry. Asymmetry in porous structure is typically achieved by using gradient distribution of matrix precursors based on diffusion kinetics or by bonding separate layers with different porous structures^25–27. These methods, however, often result in insufficient asymmetry or require complicated and costly procedures. The remarkable stability of Pickering emulsions stabilized by Janus particles (JPs) opens up new possibilities for the advancement of fabricating asymmetric porous structures. JPs, with their amphiphilic nature and particulate Pickering effect, stabilize the oil-water interface by reducing interfacial tension and providing a mechanical barrier^28–34. The interface of emulsions is even stronger as the interface shrinks, further enhanced by the jamming behavior of JPs^35–40. Therefore, multilayer emulsions stabilized by JPs and the resulting asymmetric porous composites present a promising new approach for achieving improved thermal rectification performance.

Here, a simple strategy is proposed to construct asymmetric porous composites stabilized by JPs (JAPCs) for thermal rectification. The porous composites were fabricated from JPs-stabilized emulsions with varying microstructures, followed by layer-by-layer casting and polymerization. The emulsions stabilized with JPs are composed of dispersed water droplets and a continuous monomer phase. JPs endow the emulsions with remarkable stability and tunability, allowing each layer to be manipulated independently without interference. The as-prepared asymmetric porous composites consist of two distinct, yet seamlessly integrated, porous parts. After the internal water droplets were removed, JPs remained anchored in situ in the surface of pore walls stabilizing the structure throughout the whole preparation process and in the final composites. The distinct and integrated asymmetric porous structure generated the thermal rectification of JAPCs, whose performance can be tuned by optimizing the contrast between the two layers. Moreover, JAPCs are feasible and can be easily prepared on a large scale, making them suitable for practical applications like efficient energy harvesting from fluctuating environmental temperature in buildings and aerospace exploration. The layer-by-layer casting method using stable emulsions with JPs as solid surfactants can also be applied to fabricate alternating multilayer porous structures demonstrating the potential for delicate design and construction ability in additive manufacturing.

Results

Fabrication and characterization of the JPs stabilized asymmetric porous composites (JAPCs)

Emulsions stabilized with JPs were applied in layer-by-layer casting and fabrication of asymmetric porous composites. The fabrication procedure from emulsions is illustrated schematically in Fig. 1a. The stable emulsions were composed of monomer (n-butyl acrylate) and a small amount of crosslinker (ethylene glycol dimethacrylate) with a certain dosage of JPs, which served as the continuous phase and water as the dispersed phase. The emulsions were then sealed in the mold and heated for polymerization. The as-prepared typical JAPC sample is soft, lightweight (~0.26 g/cm³, Supplementary Fig. 1), and can be supported by the soft hairs of foxtail grass (Fig. 1b). The casting method is based on the outstanding stability of emulsions derived from amphiphilic JPs. JPs used in this procedure were synthesized by seed emulsion polymerization method⁴¹. These JPs consist of a hydrophobic polystyrene part and a hydrophilic silica part, which exhibit interfacial activity (Supplementary Fig. 2). The water-in-oil emulsions stabilized with JPs were characterized by the optical microscope with spherical water droplets in view. Emulsion structures can be adjusted across a wide range by varying the dosage of JPs and the internal phase volume fraction (Supplementary Fig. 3). When the fraction reached 65 and 75 vol%, the water droplets appeared to be densely stacked, resulting in high viscosity of the emulsions as shown by the lack of fluidity in tilted glass bottles. By controlling the JPs dosage between 3 wt% and 20 wt%, and internal phase volume fraction between 55 vol% and 75 vol%, the viscosities of emulsions were optimized to a moderate range ensuring adequate stability and fluidity for further operation. Two layers of emulsions with different structures and moderate viscosities were then cast layer-by-layer and exhibited no mutual interference between layers due to their satisfactory stability (Fig. 1c). The clear, sharp boundary between the two layers of emulsions strongly confirmed the stability and independence of each layer. (One layer of emulsions was colored for better observation.) The inner microstructure of each layer of emulsions is shown in Fig. 1d, e. The as-prepared emulsions can even be written directly underwater, indicating their outstanding stability and excellent formability (Fig. 1f, g). The emulsions stabilized with JPs also exhibited long-term stability with only minor changes in droplet size over 10 days (Supplementary Figs. 4 and 5).

Fig. 1. Fabrication and characterization of the Janus particles stabilized asymmetric porous composites.

a Schematic diagram of preparing asymmetric porous composites by casting two layers of JP-stabilized emulsions. b Typical JAPC sample from 7503/7520 double-layer emulsions (JAPC-7503/7520) placed on soft hairs of foxtail grass. c Digital photographs of the cast 7503/7520 double-layer emulsions. d, e Optical microscope images of individual layers of 7503/7520 double-layer emulsions: 7503 emulsions with 3 wt% of JPs (d) and 7520 emulsions with 20 wt% of JPs (e). The internal phase volume fractions of the two layers are both 75 vol%. f The digital graph of direct writing of emulsions under water as the simple model for 3D printing. g The written letters of emulsions under water. h–j SEM images of the cross-section of JAPCs. k Reconstructed 3D structure of JAPCs obtained from micro-CT characterization. l Cross-section images from micro-CT analysis. m Statistical analysis of the porous structure of JAPC-7503/7520 in terms of pore size and porosity. The left two blue bars refer to pore size. The right two orange bars refer to porosity. Data are mean (± s.d.) pore size (n > 50) and mean (± s.d.) porosity (n = 6). Scale bars, 10 mm (b, c); 200 μm (d); 50 μm (e); 10 mm (g); 100 μm (h); 4 μm (i); 500 nm (j); 100 μm (l).

After polymerization, the asymmetric composites were dried in a convection oven until the weight stayed constant to achieve the porous structure. The as-prepared JAPCs consist of two distinct but seamlessly integrated porous parts, each with a different pore size (Fig. 1h). The spherical pores resemble the water droplets in the original emulsions, suggesting that the continuous matrix maintained its structure through polymerization and indicating the stability of emulsions. The asymmetry of the porous composites is clear and significant with sharp contrast, making them superior to gradient structures and more suitable for thermal rectification. The sharp contrast is attributed to the high stability of emulsions derived from JPs. The supporting effect of JPs at the emulsion droplet interface is verified by their distribution inside JAPCs (Fig. 1i), where densely packed JPs anchor to the pore wall surfaces due to their interfacial activity. Most importantly, JPs exhibit clear orientation at the interface with the hydrophilic silica parts facing the voids left by water droplets and the hydrophobic polystyrene parts embedded into the mildly crosslinked poly(n-butyl acrylate) matrix (Fig. 1j). Micro-CT characterization further reveals the three-dimensional microstructure of porous composites without damage (Fig. 1k). It is worth noting that no actual interfacial boundary exists between the two distinct but seamlessly integrated porous parts (Fig. 1l). Instead, the interface consists of locally interpenetrating pores from two layers, which slightly mingle, and results in an integrated asymmetric porous structure. The pore size and porosity of the two distinct parts were analyzed based on cross-section images from micro-CT as shown in Fig. 1m. The part with larger pores has an average pore size of 160 μm and a porosity of 76%, while the part with smaller pores has an average pore size of 62 μm and a porosity is 60%. Although the initial internal phase volume fraction of two emulsions for the example JAPCs was set the same (75 vol%), the final porous structures of the two layers differ due to the additional volume contributed by JPs in the continuous phase.

Thermal rectification behavior of the JAPCs

The asymmetric porous structure has been shown effective for achieving thermal rectification, both theoretically in the system of silicon membranes and experimentally in porous graphene sheets^10,42. Here, the significant asymmetry in JAPCs with a sharp contrast between layers assured the thermal rectification behavior in heating experiments. JAPCs (20 × 20 × 7 mm³) were placed on the hot plate set to 100 °C with their upper surface exposed to naturally flowing air at room temperature (Fig. 2a). The temperature of the upper surface (T_s) was monitored (Fig. 2b). The JAPCs were placed in two different directions in one of which small-pore layer was positioned close to the heating plate (upper row in Fig. 2a, denoted as the forward direction, fwd) while in the other, the large-pore layer was close to the heating plate (bottom row in Fig. 2a, denoted as the reverse direction, rev). When JAPCs were heated in the forward direction, T_s increased more rapidly and reached a higher temperature plateau (~67 °C) compared to heating in the reverse direction (~65 °C), demonstrating a faster heat conduction. Therefore, the as-prepared JAPCs exhibited a significant thermal rectification behavior. Furthermore, the thermal conductivities of JAPCs in different directions were measured (Fig. 2c). The thermal conductivity in the forward direction (k_fwd) is 0.074 W/(m·K), which is higher than the value in the reverse direction (k_rev = 0.062 W/(m·K)). The thermal rectification ratio (TR) is calculated using Eq. (1):

$$\mathrm{TR}=\left(\frac{k_{fwd}}{k_{rev}}-1\right)\times 100\%$$ 1

Fig. 2. Thermal rectification behavior of the JAPCs.

a Surface temperature monitoring of JAPC-7503/7520 placed in different directions on the hot plate using an IR imaging instrument. b Monitored surface temperature of JAPCs at different times. c Thermal conductivities of JAPCs in different heat transport directions. Data are mean (± s.d.) thermal conductivity (n = 3). d Thermal conductivities of two porous composites with uniform pores from 7503 emulsions and 7520 emulsions, respectively at different temperatures. Data are mean (± s.d.) thermal conductivity (n = 3). e Temperature of the upper surface of JAPCs during heating simulation with the two equivalent layers model. f Simulation results of heat transport inside porous skeleton structures reconstructed from micro-CT. g Temperature distribution inside porous skeleton structures along the axis direction of heat conduction 1 s after the heating starts in simulation. h, i Illustrative graphs of thermal conductivity distribution of JAPCs along different heat transport directions. The solid lines represent the actual thermal conductivity of JAPCs and the dashed lines represent the trend of thermal conductivity. The blue lines: 7503 porous composites. The red lines: 7520 porous composites. j Schematic graph of thermal rectification in JAPCs analogous to electronic diodes. Scale bar, 10 mm (a).

The TR of JAPC-7503/7520 is approximately 20%, which is relatively high compared with materials in previous studies with the incorporation of PCMs⁴³.

The origin of thermal rectification behavior is first examined at the macroscale. JAPCs were regarded as a two-layer structure with different thermal properties in individual layers. The thermal conductivities of each separate layer (small-pore layer and large-pore layer) were measured at room temperature and elevated temperatures (Fig. 2d). The thermal conductivities of the two layers differed and the discrepancy remained at different temperatures. To verify that thermal rectification behavior resulted from the asymmetric two-layer structure, finite element simulations using an equivalent two-layer model were conducted based on the thermal properties of each layer. The simulation results closely matched the experimental results, displaying the same trend and a similar temperature discrepancy in two opposite directions (Fig. 2e).

Furthermore, a comprehensive investigation into asymmetric heat transport at the microscale via simulations was performed using the 3D microstructure with porous skeletons reconstructed from micro-CT analysis and the basic properties of the polymer matrix. The simulation results revealed a similar temperature increase, which confirmed that asymmetric heat transport arose from the asymmetric porous structure at the microscale (Fig. 2f). The high-temperature region was closely confined to the heat source when JAPCs were heated along the reverse direction, suggesting a strong retardation effect on heat transport. As a result, T_s increased more gradually in the reverse direction, which makes JAPCs potential for thermal shielding applications to protect objects from sudden thermal shock. The shielding effect became more pronounced with increasing thermal load, as shown in Supplementary Fig. 6. The temperature gradient along the heat conduction path provides a deeper insight into the asymmetric heat transport behavior (Fig. 2g). At a specific time point (e.g., 1 s), the temperature drops rapidly in the large-pore part and then decreases mildly in the small-pore part in the reverse direction. On the contrary, the temperature decreases more gradually through the entire JAPC sample in the forward direction. This variation originates from the effective thermal conduction speed along the axis direction of different parts with different porosity and pore sizes. The higher porosity and larger pore size result in slower thermal conduction along the axis, leading to a more severe temperature drop in the first part, followed by a smaller drop in the second part. This causes a large temperature gradient discrepancy between the two parts. When the part with lower porosity and smaller pore size is exposed to the heat source first, however, the temperature gradient difference between the two parts is smaller. The effective thermal conduction rate along the axis in the small-pore part is higher and results in a mild temperature decrease. The temperature drop in the second part is also small because the terminal end temperature is much higher than in the reverse direction due to more heat inflow from the former part. Meanwhile, the temperature gradient inside JAPCs is dynamic and changes over time (Supplementary Fig. 7), which will amplify the influence from the first part close to the heat source with an accumulation effect in the early heating stage. The influence is then weakened as heating time increases.

The thermal conductivity distribution inside JAPCs is illustrated in Fig. 2h, i. The total effective thermal conductivity is determined by the actual thermal conductivity of the two parts of JAPCs and varies along different directions. As a result, asymmetric porous composites have an overall thermal conductivity that is much lower in the reverse direction than in the forward direction, functioning like the electrical diodes as illustrated in Fig. 2j.

Thermal rectification ratios of different JAPCs

Thermal rectification of the as-prepared JAPCs originates from the difference in thermal conductivities between the two parts and their varied temperature dependence. Therefore, the thermal rectification ratios can be easily adjusted by altering the porosity and pore size contrast between the two parts. Benefiting from the stability of emulsions with JPs, the microstructure of each layer can be adjusted independently without interference across a wide range, ensuring significant contrast. Besides, the sharp contrast was also assured by the homogeneous (non-gradient) distribution of pore size within each layer, which is verified by the microstructure characterization with Micro-CT and SEM as displayed in the images in Supplementary Figs. 8–13. This adjustment is achieved by varying the dosage of JPs and internal phase volume fraction in emulsions, as displayed in the images from SEM and micro-CT characterization (Fig. 3a–h). The pore size contrast between the two parts of JAPCs was enhanced by increasing the dosage of JPs in the small-pore part from 5 wt% to 20 wt% improving thermal rectification ratios from 1% to 20% (Fig. 3i, j). This is because the pore size decreases as the dosage of JPs increases. The two approaches, varying JP dosages and internal volume fraction, were employed to successfully control the porosity and pore size of JAPCs as shown in the detailed analysis (Supplementary Fig. 14). These methods provide a more precise way to tailor the thermal rectification performance of JAPCs (Fig. 3k). Furthermore, the porosity difference between the two parts was widened by increasing the discrepancy of internal phase volume fraction in emulsions, leading to an improvement in rectification performance from 20% to 38%, the highest value observed to date in asymmetric porous structures (Fig. 3l, m).

Fig. 3. Thermal rectification performance of JAPCs with different pore size and porosity contrast.

a–d SEM images of JAPCs from different pairs of double-layer emulsions. 7503/7505 (a); 7503/7510 (b); 7503/6520 (c); 7503/5520 (d). e–h 3D reconstructed microstructure of different JAPCs corresponding to samples as shown in (a–d). i Pore size analysis of individual parts in different JAPCs. Large pores: blue bars, small pores: red bars. Data are mean (± s.d.) pore size (n > 50). j Thermal rectification performance of JAPCs with different pore size contrast. Data are mean (± s.d.) thermal conductivity (n = 3). Thermal conductivity in the reverse direction: blue bars, in the forward direction: red bars. k Variation of thermal rectification ratios with different pore size and porosity contrast. Porosity ratio: open squares and pore size ratio: filled circles. l Porosity analysis of individual parts in different JAPCs. Large pores: blue bars, small pores: red bars. Data are mean (± s.d.) porosity (n = 6). m Thermal rectification performance of JAPCs with different porosity contrast. Data are mean (± s.d.) thermal conductivity (n = 3). Thermal conductivity in the reverse direction: blue bars, in the forward direction: red bars. n Thermal rectification performance and thermal conductivity of JAPCs compared with macroscale materials in previous research. The blue region with half-filled squares represents the segmented composites containing PCMs⁴³ (The two segments’ thickness ratio is 1:1.). The green region with half-filled circles represents the composites with asymmetric filler alignment⁴⁴. The orange region with half-filled triangles represents the gradient porous structures⁴⁵. The red region with half-filled stars represent this work. Scale bars, 100 μm (a–d).

It is noteworthy that the thermal rectification ratio of JAPCs prepared in this work is relatively high among materials reported in previous studies (Fig. 3n). JAPCs, with their rectification performance derived solely from asymmetric porous structure, overcome the drawbacks of potential PCM leakage and the limitations of a specific phase change temperature range, making them more adaptable and suitable for practical applications⁴³. Additionally, JAPCs exhibit a more distinct and significant asymmetry derived from layer-by-layer cast stable emulsions compared to composites with asymmetric filler alignment or gradient pores and therefore have superior thermal rectification performance^44,45. Compared to other asymmetric materials with thermal rectification property, JAPCs have a lower thermal conductivity and therefore have the potential in deep space exploration, where directional and quantitative thermal management and basic insulation performance are both required. Even though the polymer porous composites could not be as strong as ceramic or metal, it will not influence their application in aerospace considering the combination of other support or protective components. Besides, the thermal stability of JAPC is also characterized with TGA test showing a good stability below 300 °C (Supplementary Fig. 15), which is potential for application in aerospace. And the thermal rectification performance is well-maintained up to 100 °C around 20% for JAPC-7503/7520 (Supplementary Fig. 16). Even after a severe heat treatment at 200 °C for 2 h in nitrogen atmosphere, the thermal rectification performance still exists with a decrease by only 35 % (Supplementary Fig. 17).

Thermal regulation performance of the JAPCs

JAPCs, with improved thermal rectification performance from significant asymmetry, provide an effective solution for thermal energy regulation. Based on the facile fabrication procedure, JAPCs can be prepared on a large scale to meet the requirements of real applications. Emulsions stabilized by JPs were first prepared in large quantities and then poured into a glass mold (Supplementary Fig. 18). The glass mold was subsequently sealed for polymerization. The resulting porous composites, with the dimensions of 30 cm × 30 cm × 1 cm, exhibited a uniform appearance, highlighting their great potential for mass production (Fig. 4a). The scale-up preparation of JAPCs was accomplished by pouring emulsions layer-by-layer followed by polymerization. The as-prepared JAPCs were then tailored with the desired size and a total thickness of 6 mm, 3 mm of each layer (Fig. 4b, c). Due to the soft nature of the matrix and the robust pore walls stabilized by JPs, the JAPCs were soft and tough enough to be bent without damage, making them ideal for practical use (Fig. 4d). The soft nature was also characterized by a compression test exhibiting a low modulus. The microstructure and mechanical performance of JAPC are well-maintained after heat treatment at 200 °C for 2 h in nitrogen atmosphere (Supplementary Figs. 19–21). A small-scale house model system was constructed to evaluate the thermal regulation performance of JAPCs (Fig. 4e). Four small model houses were used with different porous composites covering on the front roof planes, respectively (Fig. 4f). Due to the thermal rectification of JAPCs, the behavior of heat transfer between the model houses and the environment differed both during the day and at night depending on the orientation of JAPCs (Fig. 4g, h). During the daytime, the temperature recorded by sensor #2 (with JAPCs in the forward direction), increased more rapidly than that of sensor #3 (with JAPCs in the reverse direction), exhibiting the better heat conduction ability of composites (Fig. 4i). When shifted into night, the temperature of sensor #2, however, decreased more slowly than that of sensor #3, indicating the better thermal insulation properties (Fig. 4j). In contrast, the composites with uniform large pores functioned as a consistent thermal insulator throughout both day and night (Fig. 4i, j). The shifts between thermal conducting and thermal insulating of JAPCs is an ideal characteristic for thermal energy harvest from a single oscillating heat source by maintaining the temperature difference between the two sides of the heat engine or thermoelectric generator^46–49. Therefore, the small-scale house model demonstrates the significant potential of JAPCs in thermal energy regulation in an environment with fluctuating temperatures.

Fig. 4. Large-scale fabrication of JAPCs and their thermal regulation performance.

a Large-area production of the porous composites. b–d The digital photographs of tailored JAPCs from large-area production. e Small model house with JAPCs covering the front roof plane. The side and back surfaces of the model house are covered by insulation aerogel pads. f The illustrative graphs of the house model setup and temperature detection. The white bars represent the large-pore layer and the gray bars represent the small-pore layer. The brown bars denote the house roof planes. g The illustrative shift of heat transport behavior of JAPCs between day and night. h The monitored temperature inside different small model houses. i Temperature inside houses from 8:00 a.m. to 15:00 during the daytime. j Temperature monitored from 22:00 to the next day at 8:00 a.m. during night time. Scale bars, 10 cm (a); 10 mm (b); 2 mm (c); 10 mm (d, e).

General synthesis method for alternating multilayer porous composites

Due to the remarkable stability of emulsions with JPs as solid surfactants, the layer-by-layer casting method can be applied to fabricate alternating multilayer structures. The preparation process is illustrated in Fig. 5a. The alternating multilayers of emulsions stacked in a bottle are exhibited in Fig. 5b. The 7503 emulsion layers are blue and the 7510 emulsion layers are red. The clear boundaries between adjacent layers without diffusion demonstrate the stability of emulsions and the independence of each layer. Droplets inside emulsions were displayed in Fig. 5c, d. The distinct variation in droplet sizes results in a sharp contrast in porous structure between adjacent layers of composites. The digital photograph of alternating multilayer porous composites is displayed in Fig. 5e. The successful fabrication of seven layers, as demonstrated, also confirmed the ability to successively extend alternating layers. The boundaries between adjacent layers are marked by the red dashed lines. The panoramic SEM images of alternating porous composites clearly demonstrate the sharp contrast in porous structure regardless of the casting order of two different emulsions (Fig. 5f–k), which strongly confirms the remarkable stability of emulsions and the independence of each layer. The variation in average pore diameters inside the alternating multilayer composites was analyzed, revealing repeated transitions between small and large pores (Supplementary Fig. 22). It is evident that small pores from different layers have similar diameters, as do the large pores indicating no influence between adjacent layers.

Fig. 5. Alternating multilayer porous composites prepared by layer-by-layer casting method.

a Illustrative graph of layer-by-layer casting method to build alternating multilayer porous composites. b Photo of alternating multilayer emulsions stacked in a glass bottle. The water droplets of 7503 emulsions were dyed with methylene blue and those of 7510 emulsions were dyed with rhodamine as red. c, d Optical microscope images of emulsions used for layer-by-layer casting. 7510 emulsions (c); 7503 emulsions (d). e Digital photo of alternating multilayer porous composites. f–k, SEM images of alternating multilayer porous composites. l Schematic graphs of alternating multilayer porous composites stacked in different sequences. m The temperature at the terminal end of heat transport in different stacking sequences and the discrepancy between forward and reverse sequences. The data were extracted from finite element simulations of multilayer porous structures. Scale bars, 10 mm (b); 100 μm (c); 200 μm (d); 2 mm (e); 1 mm (f); 500 μm (g–k).

The layer-by-layer casting method provides a feasible tool to build alternating multilayer porous structures, which is impossible from gradient approaches and more efficient than the methods with additional bonding of different layers. The multilayer structure is also beneficial for enhancing the thermal rectification performance. Finite element simulations were conducted based on alternating multilayer structures with an odd number of layers to preserve the overall asymmetry, where one layer was defined as either a large-pore layer or a small-pore layer (Fig. 5l). As the stacking layers of porous composites increase, the discrepancy of temperature at the terminal end of heat transport in different stacking sequences widened implying the enhanced rectification performance (Fig. 5m).

Discussion

A method for fabricating asymmetric porous composites with enhanced thermal rectification performance has been developed based on JPs-stabilized emulsions, and their layer-by-layer casting and polymerization. The remarkable stability of JPs-stabilized emulsions enables the independent adjustment of emulsion structures without interference between adjacent layers, which results in the distinct and significant asymmetry of JAPCs. The discrepancy in thermal conductivity of the two layers and their varied temperature dependence both contribute to the thermal rectification behavior of JAPCs particularly during the dynamic heat transport. Two effective approaches, pore size and porosity contrast of two layers, were employed to adjust the asymmetry of JAPCs and their thermal rectification ratios with a maximum value of 38%. A small-scale house model based on large-area production of JAPCs demonstrates their great performance in thermal energy regulation and exploitation under fluctuating environmental temperature, which helps to advance the practical application of thermal rectification materials. Using this layer-by-layer emulsion casting method, alternating multilayer porous composites with sharp contrast between adjacent layers were achieved, which has potential for delicate design and fabrication in additive manufacturing.

Methods

Preparation of Janus particles

The Janus particles with snowman-like morphology were prepared through seed emulsion polymerization accompanied by sol-gel and phase separation process. Hollow polystyrene particles (HP-433, DuPont) were used as starting species. Divinylbenzene (DVB, J&K Scientific, purified with alkaline alumina) emulsions were prepared with sodium dodecyl sulfate (SDS, Sinopharm Chemical Reagent) and ultrapure water (homemade). The DVB emulsions were then added to the HP-433 dispersion in water. The mass ratio between HP-433 particles and DVB is 1:1. The mixture of HP-433 particles and DVB molecules was then maintained at room temperature for 8 h under continuous stirring followed by a reaction at 70 °C for 12 h to obtain the PS/PDVB particles. The PS/PDVB particles were then used as the seed particles. 3-Methacryloxypropyltrimethoxysilane (MPS, TCI) emulsions were prepared with aqueous SDS solution (0.24 wt%) and then added slowly into PS/PDVB dispersions with a peristaltic pump within 30 min. The mass ratio between PS/PDVB particles and MPS is 2:1. The mixture was then kept at 70 °C for 24 h to obtain JPs. The JPs were washed with water and ethanol through repeated centrifugation and redispersion and then freeze-dried for 48 h ready for use.

Preparation of JPs supported asymmetric porous composites (JAPCs)

JAPCs were prepared by casting two layers of emulsions stabilized by JPs followed by the polymerization at 70 °C for 24 h. The emulsions were prepared by adding water dropwise into the monomer phase containing JPs, crosslinker, and initiator molecules and treated with vortex mixing. The main component of the monomer phase is n-butyl acrylate (BA, J&K scientific, purified with alkaline alumina), accompanied by crosslinker ethylene glycol dimethacrylate (EGDMA, J&K Scientific, purified with alkaline alumina). The volume ratio between BA and EGDMA is 30:1. The concentration of initiator azobis(2-methylpropionitrile) (AIBN, J&K Scientific) is 1 wt% based on BA and EGDMA. Typically, for 7503 emulsions, the JPs dosage is 3 wt% based on BA and EGDMA, and the volume ratio between water and total volume of emulsion (water/(water + oil)) is 75 vol%. The as-prepared double emulsions, such as 7503/7520 emulsions were cast layer-by-layer. The cast double-layer emulsions were then sealed inside glass bottles for polymerization at 70 °C for 24 h. The products were then taken out from glass bottles and dried in a convection oven until the weight stayed constant.

Characterization of JPs stabilized emulsions

The microstructure of JP stabilized emulsions was investigated with an optical microscope (Olympus) just after emulsion preparation and ten days later. The droplet size of emulsions was analyzed using ImageJ software. The average diameters were evaluated as volume-weighted ones. More than 100 droplets were measured for each sample.

Microstructure characterization of JAPCs

The microstructure of JAPCs was characterized with the scanning electron microscope (SEM, JEOL JSM-7900F, Japan) after fracturing in liquid nitrogen to expose the cross-section. The pore size was analyzed using ImageJ software based on SEM images. Micro-CT was also applied to reveal the inner microstructure of JAPCs without damage (Diondo d2, Germany) and visualized with VGStudio MAX 3.4. The micro-CT characterization was performed with the help of ND Inspection, Shanghai. The porosity of JAPCs was evaluated based on the cross-sectional images obtained from micro-CT characterization.

Thermal conductivity characterization of JAPCs

The thermal conductivity of JAPCs in different directions was measured with Hot Disk TPS 2500S (Hot Disk AB, Sweden). The probing depth was carefully set via tuning heating and fitting parameters to match the total thickness of samples (6 ~ 7 mm, the thickness ratio of two layers is 1:1). Thermal conductivities at elevated temperatures were tested with an ancillary constant temperature test chamber (TMS9013-50, China).

Thermal stability of JAPCs

Thermogravimetric analysis was performed from 35 °C to 1000 °C with the heating rate of 10 °C/min in nitrogen atmosphere using NETZSCH TG 209 F3 (Germany). The heat treatment at 200 °C for 2 h was conducted in nitrogen atmosphere with tube furnace (GSL−1500X, China). The compression mechanical performance before and after heat treatment was characterized using DMA equipment (DMA-850, TA Instrument, USA).

Thermal conduction simulations of JAPCs

Thermal conduction simulations of JAPCs were performed on COMSOL Multiphysics software. The simulations of JAPCs with two equivalent layers were performed by building models with two layers in the dimension of 20 × 20 × 3.5 mm³ (length × width × thickness) each layer. The equivalent density was set as 0.16 g/cm³ for the large-pores layer and 0.36 g/cm³ for the small-pores layer. The equivalent thermal conductivity was set as the function of temperature T in the unit of Celsius degree. The function is k_l [W/(m·K)] = (1.31888 × 10⁻⁴)·T[°C] + 0.06107 and k_s [W/(m·K)] = (7.4467 × 10⁻⁵)·T[°C] + 0.07055 for large-pores and small-pores layer, respectively. The equivalent specific heat capacity was set as 1204 J/(kg·K) and 1334 J/(kg·K) for the large-pore layer and the small-pore layer. The heat source is set at a constant temperature of 100 °C. The initial and environment temperature was set as 25 °C. The surrounding and terminal end surfaces were set as natural convection boundaries with the convective heat transfer coefficient (h) as 10 W/(m²·K).

The simulations of asymmetric porous structures were performed with the reconstructed 3D microstructure from micro-CT characterization. The basic properties of JAPCs were set according to references. The density of the continuous matrix of porous composites was set as 1.087 g/cm³ and the specific thermal capacity was 1820 J/(kg·K). The heat source is set at a constant temperature of 100 °C. The boundary condition for the end surface of heat transport was set as natural heat convection with the convective heat transfer coefficient (h) as 5 W/(m²·K) to manifest the asymmetric phenomenon for clarity. Any other surfaces inside JAPCs were set as thermal insulation boundaries. The initial temperature of the whole composite is 20 °C, which is equivalent to the environmental temperature.

The simulations of multilayer alternating porous structures were performed by stacking a basic two-layer model reconstructed from micro-CT. The heat source was set at a constant temperature of 100 °C and the surrounding boundaries were set as natural heat convection with the convective heat transfer coefficient (h) as 10 W/(m²·K). The environment temperature is 25 °C.

Thermal regulation of JAPCs with small-scale house models

The porous composites were placed on the front roof plane of the model houses and the temperature inside the house was monitored with a sensor. The control group was set without a porous composite covering and was monitored with temperature sensor #1. Sensors #2 and #3 were used for model houses covered with JAPCs in different directions, where the small-pore part faced outward with sensor #2 and the large-pore part faced outward with sensor #3. The sensor #4 was used for a model house covered by porous composites with uniform large pores. The model houses were then put inside a polystyrene foam box and covered by the cling film to eliminate the influence of fierce air convection. The experimental setup was then settled on top of the building, which was exposed to the fluctuating environmental temperature. The experiment was conducted from November 16 to 18, 2024, in Beijing, China.

Author contributions

F.L. and C.J. conceived and designed the experiments. C.J. carried out the experiments. F.L. and C.J. analyzed the data. X.Z. contributed ideas and suggestions for thermal conductivity characterization. H.C. gave suggestions to thermal conductivity characterization. C.J., X.Y., and C.W. performed the finite element simulations. X.H. helped to carry out the experiments of direct writing of emulsions underwater and build the small house models and perform measurements. C.J. wrote the manuscript. F.L. revised and edited the manuscript. All authors discussed the results and commented on the manuscript.

Peer review

Peer review information

Nature Communications thanks Hortense Le Ferrand and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The source data generated in this study are provided in the Source Data files. Source data are also deposited as figshare items with the figshare 10.6084/m9.figshare.28418510.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-025-60792-4.