中文   |   English   |   Email Alert
Fig/Tab
Advanced Search
image

New Strategies for Flame Retardancy in Polyurethane Composites：A Review of Biomimetic and Smart Flame Retardants

Yuanyuan Kang , Zun Gao , Yangyang Cui , Shuting Wu , Fei Xin

Progress in Chemistry ›› 2026, Vol. 38 ›› Issue (6) : 1085 -1092.

PDF
Progress in Chemistry ›› 2026, Vol. 38 ›› Issue (6) : 1085-1092 DOI: 10.7536/PC20251111
research-article
Scan to read & Share
New Strategies for Flame Retardancy in Polyurethane Composites：A Review of Biomimetic and Smart Flame Retardants
Author information +
History +
PDF

Abstract

Polyurethane（PU）is an indispensable polymer material widely used in daily life. However，its flammability（limiting oxygen index LOI < 20%）and release of toxic smoke during combustion restrict its application in emerging fields such as smart wearables and robotic skins. Therefore，overcoming the limitations of traditional flame-retardant strategies has become an urgent issue in materials science. This review systematically summarizes key advances from 2021 to 2025，focusing on new flame-retardant paradigms based on biomimetic structural design and intelligent response mechanisms. Multiscale biomimicry includes macro-shape mimicry，such as sunflowers and cacti，and micro-structural mimicry，such as nacre-like layering and lotus-leaf superhydrophobic surfaces. This approach effectively balances flame retardancy with mechanical properties and introduces additional functions like oil-water separation，solar ice melting，and underwater antibacterial performance. Furthermore，this paper presents the concept of intelligent active protection. Integrating self-healing，fire early warning，and shape memory creates a synergistic system that forms a closed-loop protection mechanism covering sensing，alerting，and action. This advances flame-retardant technology from static，passive defense toward dynamic，adaptive response. Nevertheless，this field still faces three major challenges. First，environmental adaptability is limited，as most self-healing and shape memory behaviors require specific activation conditions. Second，complex biomimetic nanostructures are difficult to fabricate on a large scale at low cost. Third，the long-term cyclic stability and durability of coatings and composites need improvement. To address these challenges，future work should integrate artificial intelligence for material pre-design，combine animal and plant features to build multi-level defense systems，and develop biomass-based green closed-loop solutions across the material life cycle. These efforts will promote the development of biomimetic intelligent flame-retardant polyurethanes that can coexist sustainably with the environment.

Contents

1 Introduction

2 Bionic flame retardant

2.1 Shape bionics

2.2 Structural bionics

2.3 Surface coating

3 Intelligent flame retardant

3.1 Self-healing

3.2 Fire warning

3.3 Shape memory

4 Conclusion and outlook

Graphical abstract

Keywords

polyurethane / bionic flame retardancy / structural biomimicry / intelligent flame retardancy

Cite this article

Download citation ▾
Yuanyuan Kang, Zun Gao, Yangyang Cui, Shuting Wu, Fei Xin. New Strategies for Flame Retardancy in Polyurethane Composites：A Review of Biomimetic and Smart Flame Retardants. Progress in Chemistry, 2026, 38(6): 1085-1092 DOI:10.7536/PC20251111

登录浏览全文

4963

注册一个新账户 忘记密码

1 Introduction

Polyurethane (PU) is a multifunctional polymer material obtained through the step-growth polymerization of isocyanates and polyols. By adjusting the ratio of soft segments (polyols) to hard segments (isocyanates) and the synthesis process, the chain segment structure can be controlled to form various forms such as flexible foams, rigid foams, thermoset elastomers (non-meltable and extremely durable), thermoplastic elastomers (meltable, reshaped, and more elastic), coatings, and adhesives, covering almost all industrial fields including transportation, construction, home appliances, healthcare, aviation, and railways.[1]. However, its flammability (Limiting Oxygen Index LOI < 20%, classified as a flammable material) creates a sharp contradiction with its wide range of application scenarios. Furthermore, combustion generates large amounts of toxic smoke (such as hydrogen cyanide (HCN), carbon monoxide (CO), etc.), posing severe hazards to individual life, health, and the natural environment.[2].
Traditional flame-retardant modification mainly relies on the physical blending of flame retardants, such as halogen-free flame retardants, reactive flame retardants, and group synergistic enhancement systems. Although these can achieve flame-retardant effects to a certain extent, they also bring new challenges: high addition amounts lead to a decline in the mechanical properties of materials[3]; poor interfacial compatibility between the flame retardant and the matrix causes the flame retardant to migrate and exude over time[4]; some halogen-containing flame retardants release toxic gases during combustion, making them unfriendly to the environment[5]. These limitations have prompted researchers to turn their attention to new strategies that combine sustainability with multifunctional integration.
In recent years, the development of bionics and smart materials has provided new structural inspirations and design concepts for flame-retardant PU: bio-strategies that mimic the shapes and structures of plants and animals in nature to improve interfacial compatibility achieve a multiplication of flame retardancy efficiency and smoke suppression effects; combining dynamic covalent bonds (such as disulfide bonds/borate ester bonds) and stimulus-responsive units (such as photothermal/pH/thermal triggers) constructs adaptive protection systems. Reviewing the breakthrough progress made between 2021 and 2025, biomimetic intelligent flame-retardant systems can simultaneously optimize flame retardancy performance (significant increase in LOI) and mechanical strength (doubling of tensile strength), while endowing materials with novel properties, further expanding the application scope of polyurethane to wearable electronic devices[6], piezoelectric sensors[7-8], artificial skin[9], and other emerging industries. This review systematically summarizes the excellent performance of biomimetic structural design and intelligent response mechanisms in flame-retardant PU, and looks ahead to trends in green intelligent manufacturing, providing a theoretical basis for designing the next generation of high-performance, multifunctional, and intelligent flame-retardant polyurethane materials.

2 Biomimetic Flame Retardant

Biomimetic strategies provide a new paradigm for resolving the contradiction between the flame retardancy and mechanical properties of PU by simulating structural optimization mechanisms in nature. Based on differences in biomimetic scale, they can be classified into three categories: macroscopic shape biomimicry, microscopic structure biomimicry, and interfacial coating biomimicry.

2.1 Shape Bionics

The unique topological structures of many plants in nature (such as the radial petals of sunflowers, the layered expansion of hydrangeas, and the directional arrangement of cactus spines) inspire interface design: enhancing interfacial anchoring by increasing specific surface area to improve the dispersion and flame retardancy of nanofillers in PU. Ce-4APD@BMC synthesized inspired by sunflowers (Ce-4APD: rare earth metal coordination compound; BMC: basic magnesium carbonate)[10], its radial channels extend the diffusion path of pyrolysis gases, reducing the total heat release (THR) of thermoplastic polyurethane (TPU) composites by 42.6% and total smoke production (TSP) by 40.8%, with carbon dioxide (CO2) and CO production rates reduced by 98.2% and 52.0%, respectively. Hydrangea-like hierarchical structures can optimize filler dispersion, such as Ti3C2Tx MXene@FeNi, etc.[11-12] reduce TSP by 16.1%~67.0% through catalytic char formation; APTES silane coupling agent-modified BN@MOF-LDH (BN: boron nitride; LDH: layered double hydroxide)[13] reduces THR and TSP of flexible polyurethane foam (FPUF) by 15.8% and 15.7%, respectively, via metal oxides and silicates protecting the char layer. Mimicking the spine-like leaf structure of cacti can enhance mechanical interlocking effects. In BP@MoS2 hybrids (BP: black phosphorus; MoS2: molybdenum disulfide)[14], vertically grown MoS2 (cactus-spine-like) increases contact force with the waterborne polyurethane (WPU) matrix. After adding 2 wt%, TSP of WPU/BP@MoS2-2.0 decreased by 52.0%; tensile strength and breaking strength increased by 57.7% and 36.5%, respectively. The synergistic mechanisms of these shape-biomimetic strategies focus on two aspects: first, capturing free radicals in the gas phase; second, catalyzing char formation to create a barrier effect in the condensed phase.Table 1 provides an intuitive comparison of the flame retardant effects and synergistic mechanisms of the aforementioned shape-biomimetic strategies.

2.2 Bionic Structure

From macroscopic shapes to microscopic structures, nature similarly provides abundant design references. The superhydrophobic phenomenon of lotus leaves, the strength and toughness of seashells, and the refraction of natural light by insect wings all originate from their own exquisite micro-nano structures. Imitating macroscopic forms only grasps the surface; designing microscopic structures better embodies the essence of bionics.

2.2.1 Superhydrophobic surface design

The representative work of superhydrophobicity in nature is the lotus leaf, from which the term "lotus effect" is derived (Figure 1). Inspired by the "lotus effect," flame retardancy and hydrophobicity can be simultaneously optimized: dopamine-modified SiO2/Fe3O4@PU sponge[15] has a pHRR of 139.7 kW/m2, a reduction of 71.2%; THR is 4.6 kJ/g, a reduction of 74.2%. Water contact angle (WCA) = 151.3° (as shown in Figure 2a), with an oil adsorption capacity up to 33.63 times its own weight (stable over 10 cycles). Waste PU sponges can also be designed with superhydrophobic surfaces to achieve waste recycling and reuse[16]. One-step dip-coating with silicone/Fe3O4/EG (EG: expanded graphite) yields a WCA = 156.2° ± 1.4°, water sliding angle (WSA) = 8.0° ± 0.5°, while light oil adsorption capacity, separation efficiency, and flux reach 33.2 g/g, 99.5%, and 31,611 L/m/h, respectively. Additionally, pHRR is 56.05 kW/m2, reduced by 41.57%; THR is 10.67 kJ/g, reduced by 39.56%. The "lotus effect" also exists in certain animals (Figure 1). Inspired by the compound eye structure of mosquitoes[17], ADP/SiO2 modified (ADP: aluminum diethyl phosphinate) PU foam (Figure 2b) has an LOI = 21%, with pHRR reduced by 58.0% and THR reduced by 44.6%. Meanwhile, the char residue increases from 0.4% to 23.5%. Additionally, the dichloromethane adsorption rate is 69 g/g (applicable under turbulent conditions). Its dynamic adaptability stems from the response of the biomimetic compound-eye hierarchical porous structure to environmental changes. Whether it is the static "lotus effect" or the dynamically adaptive insect compound eye structure, the essence lies in constructing micro-nano secondary rough structures to trap air, thereby achieving superhydrophobicity and ingeniously combining it with flame retardant performance.

2.2.2 Layered Enhanced Structure

Beyond surface phenomena, the microscopic internal structures of organisms are equally worthy of emulation. Observing a shell cross-section under a microscope reveals a layered stacking; this "brick-and-mortar" multilayer model achieves material strengthening and toughening through ordered stacking: MA@BP nanosheets (MA: melamine) enhance interfacial bonding via MA hydrogen bond reorganization[18], reducing the pHRR of TPU composites to 660.6 kW/m2, and decreasing THR by 22.8%. Meanwhile, the BNNS-CP/MP "brick-and-mortar" system (BNNS: hexagonal boron nitride nanosheets; CP: copper salt of hydroxyethyl diphosphonic acid; MP: meclizine hydroxyethyl diphosphonate)[19]enables TPU composite films to achieve an LOI = 30.2% (UL-94 V-0), with pHRR and THR reduced by 77.2% and 80.6% respectively, and a thermal conductivity of 9.12 W/m/K. Introducing more functional MXene nanosheets would generate new synergistic effects[20]. Electrostatic self-assembled MXene@SnO2 nanosheets (SnO2: tin dioxide) reduce the pHRR of TPU composites by 50.5% and the total smoke production rate (SPR) by 33.1%, while achieving solar de-icing in 40 s at -20 ℃ (compared to 600 s for pure PU). Similar layered structures also exist in plants. TPU-MSSO nanocomposites (M: MXene; SSO: strontium hydroxytin nanorods)[21]exhibit a leaf-vein-inspired hierarchical structure during stretching, increasing elongation at break and toughness by 55% and 126%, respectively. Moreover, harmful pyrolysis products during combustion are significantly reduced, with CO and CO2 generation decreased by 96.04% and 89.85%, respectively. Additionally, inspired by the "thorn-seed-bark" structure of crabapples[22], PU composites were designed using copper ethylenediamine methylphosphonate as "thorns" and functionalized aluminum oxide microspheres as "seeds," achieving a 354% increase in tensile strength, a thermal conductivity of 2.51 W/m/K, and an LOI of 29.1% (UL-94 V-0). Applications in high-power CPU cooling materials have been developed. In summary, the core of biomimetic design for layered reinforcement structures lies in achieving multifunctional synergy in mechanical, thermal, and flame-retardant properties through ordered stacking or multi-level interface strengthening. Whether it is the shell's "brick-and-mortar" model, the hierarchical support of leaf veins, or the crabapple's "thorn-seed-bark" structure, all reflect strategies adopted by organisms to maximize performance in natural environments, providing important insights for the multifunctional integration of PU composites.

2.2.3 Biosorption effect

To adapt to the complex and variable marine environment, marine organisms have evolved epidermal structures integrating multiple functions over hundreds of millions of years; mimicking these structures promises to enable multifunctional integration: shark skin-inspired β-FeOOH coating[23]reduced the pHRR of FPUF by 45.0% and smoke density by 69.1%, while also providing antibacterial properties (Figure 3a); drawing inspiration from octopus suckers (Figure 3b), an M@CN hierarchical network was designed (M: MXene; CN: carbon nanotubes)[24], where the 1D-2D hierarchical structure increased the tensile strength and elongation at break of the TPU composite by 79.5% and 113.5%, respectively. Furthermore, pHRR decreased by 54.1%, and the release rates of CO and CO2 were reduced by 62.7% and 33.0%, respectively.

2.3 Surface coating

The aforementioned biomimetic strategies are mostly achieved through filler blending; however, while physical blending can improve flame retardancy and mechanical properties, on one hand, it disrupts the structure of the PU matrix (especially foams), and on the other hand, the mixing process damages the biomimetic design. The rapid advancement of coating technologies modified via interface engineering provides a complementary and efficient solution.

2.3.1 Layered barrier coating

The "brick-and-mortar" inorganic-organic alternating layered structure of shells also provides a model for designing high-performance flame-retardant coatings: gelatin/THPS aerogel[25](THPS: tetrakis(hydroxymethyl)phosphonium salt); at thicknesses greater than 100 μm, LOI > 80% (pure FPUF ≈ 18%), and compressive strength increases by 375% (Figure 4a); BP/SiO2 ceramic scales[26] reduce the pHRR and TSP of PU foam by 50.3% and 54.9%, respectively, with LOI = 45% (Figure 4b). If hybridized with BN/glass powder[27], a thin layer of 200 μm enables PU foam to achieve UL-94 V-0 (LOI 35.8%). The core mechanism of such barrier biomimetic strategies is: the presence of a layered topological structure enhances char compactness during combustion through geometric constraints that entangle molecular chains, ultimately forming a barrier effect.

2.3.2 Superhydrophobic interface coating

Although layered barriers improve fire resistance limits, they struggle to prevent moisture penetration and erosion. In contrast, superhydrophobic coatings guided by the "lotus effect" and desert beetle microstructures achieve multi-functional protection through gas-liquid interface stabilization (Cassie-Baxter state). FPUF@MOF-LDH@HDTMS sponge[28]Superhydrophobic modification of FPUF is achieved via in-situ growth of micro-nano structures; after multiple compressions, the sponge still exhibits a WCA > 150°. Metal-organic framework materials (such as ZIF-8 and UiO-66) are also used to enhance the oil-water selectivity of PU sponges due to their unique structures, but the flammability of organic ligands limits their further application. The PU-Cu(BDC)-NGPs sponge prepared by co-dipping with nanographite sheets (NGPs)[29]not only possesses excellent thermal stability, recyclability, and mechanical stability but also effectively performs oil-water separation. The separation efficiency reaches up to 96% and adapts to different pH environmental conditions. The mechanism of such coatings lies in utilizing low-surface-energy substances and rough structures to jointly construct a stable air layer, achieving synergy in water repellency, oil repellency, and flame retardancy.

2.3.3 Biomimetic Adhesive Coating

Superhydrophobic coatings are prone to failure under mechanical wear, whereas "bio-inspired adhesive interfaces" enhance durability through chemical bonding and topological interlocking. HEA-VS copolymers (HEA: hydroxyethyl acrylate; VS: sodium vinyl sulfonate)[30]possess phase-separated micro-nano structures. After coating treatment, PU achieves UL-94 V-0 rating (LOI 35.5%), with pHRR reduced by 87%, SPR reduced by 71%, and char yield reaching 80.3 wt%. Alternating copolymers of acrylic acid and hydroxyethyl acrylate[31]achieve UL-94 V-0 while simultaneously exhibiting water-mediated self-healing. TA/PVA@BNNS coatings prepared via a simple brush-coating method (TA: tannic acid; PVA: polyvinyl alcohol)[32]reduce the pHRR and peak smoke production rate (PSPR) of rigid polyurethane foam (RPUF) by 52.4% and 53.4%, respectively. The core of such adhesive bio-inspired strategies is to achieve strong adhesion of flame-retardant coatings to the substrate surface through mechanisms such as interfacial hydrogen bonding and mechanical interlocking, thereby keeping the flame as far away as possible from the flammable substrate.

2.3.4 Function-oriented coating

After chemical adhesion resolved interface durability issues, bionic concepts further guided the development of functional coatings for specific application scenarios, expanding the scope of material applications to photothermal and acoustic fields. Among them, the insect compound eye anti-reflective coating PDMS@SiO2modified graphene (PDMS: polydimethylsiloxane)[33]reduced light reflectance by 20%; for the coated PU composite, the de-icing time at -20 °C was 65 s (compared to 600 s for pure PU), and the pHRR was also reduced by 58%. The ZIF@DHPPC nanomaterial, mimicking the dropstone nuclear pore structure,[34]achieved a noise reduction of 35 dB through the acoustic vortex dissipation mechanism, while simultaneously reducing the CO and CO2generation by 54.5% and 21.4%, respectively.

3 Intelligent Flame Retardant

Biomimetic coatings endow PU with static protection capabilities, whereas real fires represent an extremely dynamic and evolving environment, imposing new demands on the "proactivity" and "intelligence" of materials. Therefore, it is necessary to integrate intelligent response mechanisms to achieve real-time damage remediation or early fire warning, forming the core of intelligent flame retardancy.

3.1 Self-healing

The emergence of self-healing not only enables a more dynamic and flexible response to fire situations but also extends the service life of materials. Based on triggering mechanisms, they can be classified into: (1) Photothermal activation type: PDA@EG/PU coating (PDA: polydopamine)[35]Near-infrared (NIR) irradiation for 3 min raises the temperature to 82.9 °C, triggering disulfide bond reorganization to achieve 93.79% toughness recovery (within 10 min), and it is repeatable. Although photothermal healing systems respond rapidly, they rely on external light sources. In contrast, bio-based dynamic bond networks can autonomously heal through molecular thermal motion: NVP-xnetwork (xrepresents different molar ratios of naringenin to vanillin-based diol)[36]achieves 98.1% healing efficiency at 100 °C for 40 min via dynamic phenol-urethane bonds, while simultaneously exhibiting an LOI of 28.9% (UL-94 V-1). (2) Spontaneous dynamic bond type: DBPA-modified DPU (DBPA: hyperbranched phenolic compound)[37]achieves room-temperature self-healing (efficiency: 71.2% at 25 °C/50% RH) through multi-level hydrogen bonding interactions. Trace addition (0.6 wt%) reduces the heat release rate (HRR) by 13%, decreases total heat release (THR) by 18%, and extends time to ignition (TTI) by 7 s (a 21% increase). Besides this room-temperature hydrogen bond reorganization self-healing, boronate ester bonds can also drive self-healing. PEI/APP/GO coating (PEI: polyethyleneimine; APP: ammonium polyphosphate; GO: graphene oxide)[38]enables humidity-driven self-healing in FPUF samples (3 days at 50% RH), while increasing compressive strength by 175%. This method is simple and feasible, but healing strength and time are typically limited. (3) Thermal-driven type: Reaction between vanillin-based polyols and hexamethylene diisocyanate (HDI) trimer yields a bio-based polyurethane network possessing dynamic disulfide bonds and imine bonds[39]. This polyurethane network exhibits a char residue of 23% at 800 °C and an LOI value of 30% (UL-94 V-0). Treatment at 120 °C for 90 min achieves nearly 100% self-healing efficiency.Table 2presents a comparison of the efficiency results for the three self-healing mechanisms.

3.2 Fire Warning

Although the self-healing function can dynamically repair damage, endowing the material with "post-event remediation" capabilities, it cannot proactively avoid fire risks. An ideal intelligent protection system requires more "pre-event perception"; therefore, it is necessary to combine a real-time fire warning system with subsequent shape memory responses to form a closed loop of "perception-warning-protection." The sensor CHAP-PU was prepared by depositing coatings of A-CNTs (amino carbon nanotubes), HGM (magnesium hydroxide), and ADP (ammonium phosphate) onto the surface of a PU sponge skeleton via dip-coating, followed by modification with polydimethylsiloxane (PDMS).[40]. Upon exposure to fire, the coating carbonizes, causing a sharp drop in electrical resistance and triggering an audible and visual alarm within 2 s. However, a single-layer structure cannot sustain this for long; thus, a three-layer "sandwich" flame-retardant coating was constructed using phytic acid (PA), a flame-retardant copolymer (PVH), GO, CNTs, and BN.[41]. After coating treatment, the PU foam achieved an LOI of 58.0%, with pHRR, TSR, and peak carbon monoxide production rate (PCOP) reduced by 49%, 41%, and 67%, respectively. It exhibits a rapid response within 8 s at high temperatures and can provide continuous warning. Although the "sandwich" coating offers rapid warning, its fabrication process is complex. In contrast, FGO/CNT@PUS foams prepared via a one-step impregnation method (FGO: functionalized graphene oxide; CNT: chitosan-functionalized carbon nanotubes)[42]compressed the fire alarm time to 1 s, triggered an alarm within 2 s at 500 ℃, and sustained operation for 2640 s. Simultaneously, HRR was reduced by 63.2%. The aforementioned warning methods all require external alarms, which are not intuitive enough. When the GEL-CR-BA-M aerogel coating[43]was applied to FPUF, the LOI exceeded 70% (UL-94 V-0), THR was reduced by 90%, and TSR was reduced by 70%. The phenol red-boric acid complex turns red above 100 ℃ and possesses reversible characteristics. If combined with a color recognition system, it can achieve remote intelligent early fire warning. Different warning systems can control the response time within 10 s, yet deficiencies remain regarding alarm timing and stability. The BFPU system[44]utilizes a bio-graphene sensor to achieve a 5.3 s response plus an ultra-long continuous alarm duration of 3600 s, while maintaining an LOI of 46.5% (UL-94 V-0). These studies reveal that the key to fire warning response lies in constructing a sensitive conductive network capable of undergoing drastic and stable changes in electrical or optical properties under high-temperature impact, thereby achieving intelligent perception.Figure 5provides a simple schematic of the fire warning circuit mechanism.

3.3 Shape memory

The sensitive response of the warning system is only one part of intelligent protection; the shape memory function will assist in completing active protection driven by the warning. Specifically, shape memory can satisfy: (1) Deformation precision control. Covalently connecting flame-retardant groups into the network of the PU elastomer[45], the PU-DP elastomer possessing dynamic borate ester bonds can recover 83.6% of its elongation at break (at 130 °C), while simultaneously achieving a UL-94 V-0 rating. (2) Directional migration of flame retardants. Embedding polycaprolactone diol (PCL-diol) and tris(hydroxymethyl)phosphine oxide (THPO) into the PU main chain[46], the presence of THPO enhances the cross-linking network density within the elastomer, increasing tensile strength and toughness by 450% and 117%, respectively. The glass transition temperature (Tg) between 30 °C and 40 °C enables the material to recover its preset shape after heating. If additional flame retardants are incorporated, the motion during soft segment recovery can drive the directional enrichment of flame retardants toward the surface. (3) Matching warning time. A simple electrospinning and vacuum filtration process can be used to fabricate an SMPU/MXene paper composed of shape memory thermoplastic polyurethane (SMPU) and MXene[47]. This paper undergoes shape recovery within seconds when the ambient temperature exceeds the glass transition temperature, realizing a circuit connection alarm function. Furthermore, the pHRR and THR are reduced by 66.0% and 49.8%, respectively. Shape memory materials meeting these requirements convert temperature signals into mechanical deformation via glass transition, achieving functional complementarity with fire warning systems, representing a key step in constructing future intelligent fire protection systems.

4 Conclusion and Outlook

Polyurethane flame retardant technology, through the integration of bionic structural design and intelligent response, has achieved a leapfrog development from "passive flame retardancy" to "active flame retardancy": progressing from basic imitation to structural design, and further to interface coverage. Multi-scale bionics transforms polyurethane composites from flammable materials into difficult-to-ignite materials; beyond improving mechanical properties, it also opens up multi-functional integrated capabilities including oil-water separation, underwater antibacterial action, building sound absorption, and solar ice melting. The development of smart materials brings new opportunities for polyurethane flame retardancy: the introduction of self-healing enhances "sensing" capabilities, the design of fire warning structures provides "sensing" signals, and shape memory handles signal conversion and perfects protection. If these three elements interact to form a complete chain, polyurethane flame retardant technology will reach even greater heights.
Despite significant achievements, there is still a long way to go from the laboratory to industrialization. The main challenges include: (1) Imperfect environmental adaptability. Many self-healing and shape memory effects require specific temperature, humidity, or external stimulus conditions to be realized, and their applicability to natural extreme environments remains unknown. (2) Challenges in scalable preparation. The sophisticated biomimetic nanostructures (such as hierarchical rough surfaces and ordered layered stacking) require complex processes, making it difficult to achieve low-cost, large-scale, and stable production. (3) Insufficient long-term durability. Although coating research aims to extend material lifespan, few coatings can maintain excellent performance after multiple thermal cycles, mechanical wear, and environmental aging; how to suppress degradation is also one of the issues to consider.
To break the aforementioned dilemmas in the future, the following aspects can be considered: (1) Multi-source bionic fusion intelligent design: Since single-structure bionics offers limited performance enhancement, consider combining animal characteristics (such as scale-like protection and suction cup adhesion) with plant characteristics (such as leaf transpiration and phototropism) to construct a dual-protection system. Simultaneously, leverage artificial intelligence to customize structures, predict and optimize designs in advance, and align with emerging fields such as robotic skin and 5G electromagnetic shielding. (2) Develop green sustainable technologies: Material design based on bionics and intelligence can connect with frontiers in biology to develop flame retardants and matrices derived from biomass raw materials, while designing pathways for biodegradability after use, forming a fully bio-based green closed loop encompassing synthesis, flame retardancy, application, and recycling. (3) Enhance performance in practical environmental applications: Future research needs to focus more on evaluating material performance under actual service conditions, constructing composite material systems with greater environmental tolerance, longer service life, and easier maintenance. Ultimately, through interdisciplinary collaborative innovation, enable bionic intelligent flame-retardant polyurethane composite materials to self-perceive, dynamically adjust, and coexist harmoniously with the environment.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]

Amini Jahromi S, Farahavar G, Nasiri Zarandi M, Tamaddon A M. J. Ind. Eng. Chem., 2024, 130: 73.
[2]

Dong H H, Li S J, Jia Z X, Luo Y F, Chen Y J, Jiang J, Ji S. Polymers, 2024, 16(22): 3182.
[3]

Pan G F, Wang Z, Kong D Q, Sun T W, Zhai H, Tian T, Wang Y F, Xing R G, Zhang B W. J. Appl. Polym. Sci., 2022, 139(6): 51598.
[4]

Qi P, Li Y C, Sun J, Wang X G, Wang K H, Meng D, Gu X Y, Li H F, Zhang S. Compos. Part B Eng., 2022, 247: 110262.
[5]

Jeong S H, Heo J H, Lee J W, Kim M J, Park C H, Lee J H. ACS Appl. Mater. Interfaces, 2021, 13(19): 22935.
[6]

Yan K, Wang J, Zong Y, Xu Q N. Chem. Eng. J., 2024, 489: 151315.
[7]

Tang S, Han J H, Xia Y H, Yu Y Y, Li Z M, Wu F, Xie Y H, Xie D L, Huang H, Feng D. Ind. Eng. Chem. Res., 2024, 63(19): 8673.
[8]

Sun Y, Liu K, Bu F, Meng R J, Xie G Q, Guo K, Cao A D, Tu L X. Sens. Actuat. A Phys., 2024, 375: 115524.
[9]

Chen Y A, Lai R H, Lin W C, Huang H Y, Chen S J, Yeh C M, Huang H L, Elsenety M M, Hu C C, Yu C H, Chou H H. ACS Appl. Polym. Mater., 2024, 6(12): 6976.
[10]

Xu J Y, Yu F H, Cheng Z F, Song Z M, Fang J, Li H, Wang B B, Song L, Hu Y. Polym. Degrad. Stab., 2025, 233: 111173.
[11]

Jia P F, He R F, Song L, Wang B B, Hu Y. Appl. Mater. Today, 2023, 32: 101803.
[12]

Zhou Y F, Chu F K, Ding L L, Yang W H, Zhang S H, Xu Z M, Qiu S L, Hu W Z. Chemosphere, 2022, 297: 134134.
[13]

Zhou Y F, Qiu S L, Chu F K, Yang W H, Qiu Y, Qian L J, Hu W Z, Song L. J. Colloid Interface Sci., 2022, 609: 794.
[14]

He L X, Chu F K, Zhou X, Song L, Hu Y. Polym. Degrad. Stab., 2022, 202: 110027.
[15]

Hu C, Xue T D, Ma R L, Chai B, Zhang S Y, Wei Q H, Wang M Y, Jiang Q R, Mei Y J. J. Environ. Chem. Eng., 2025, 13(3): 116230.
[16]

Han Z S, Lv X M, Li Y G, Gao M H, Tang Z L, Su X Y, Zhang Z Y, Li H D, He J, Zheng Z H, Liu Y. J. Mol. Liq., 2024, 409: 125430.
[17]

Xu Z, Chu F, Jiang S, Hu Y, Song L, Hu W. Mater. Today Chem., 2022, 26: 101127.
[18]

Cai W, Xing W Y, Cui T Y, Wang J L, Lin B C, Li Z X, Qi L Y, Hu X, Ming Y, Xiao P, Bian F L, Fei B, Hu Y. Compos. Part A Appl. Sci. Manuf., 2024, 178: 108006.
[19]

Wang Y B, Pan K C, Fan Y, Fu Y F, Tu J Y, Zhao W Q, Guo W J, Liu Z F, Qiu J. Chem. Eng. J., 2024, 490: 151619.
[20]

Cai W, Li Z X, Cui T Y, Feng X M, Song L, Hu Y, Wang X. Compos. Part B Eng., 2022, 244: 110204.
[21]

Lu J Y, Jia P F, Liao C, Xu Z M, Chu F K, Zhou M T, Yu B, Wang B B, Song L. Compos. Part B Eng., 2022, 228: 109425.
[22]

Wang Y B, Fan Y, Pan K C, Liu Z F, Zhao W Q, Zhou X, Qiu J. Small, 2025, 21(5): 2405971.
[23]

Meng D, Wang K H, Wang W J, Sun J, Wang H Q, Gu X Y, Zhang S. Chemosphere, 2023, 312: 137060.
[24]

Jia P F, Lu J Y, He R F, Jiang G Y, Jiang X, Wang B B, Song L, Hu Y. Chem. Eng. J., 2022, 450: 138184.
[25]

Yang Y Q, Wang S Z, Wang S Q, Wang C, Jiao Y H, Liu H M, Ma H Y. Prog. Org. Coat., 2024, 196: 108702.
[26]

Qiu S L, Su F H, Zhou Y F, Li Y T, Wang X, Zhang L B. Chem. Eng. J., 2024, 499: 156137.
[27]

Ma Z W, Zhang J Z, Maluk C, Yu Y M, Seraji S M, Yu B, Wang H, Song P G. Matter, 2022, 5(3): 911.
[28]

Piao J X, Lu M J, Ren J Y, Wang Y F, Feng T T, Wang Y X, Jiao C M, Chen X L, Kuang S P. J. Hazard. Mater., 2023, 444: 130398.
[29]

Habibi N, Faraji S, Pourjavadi A. Colloids Surf. A Physicochem. Eng. Aspects, 2023, 676: 132186.
[30]

Ma Z W, Liu X C, Xu X D, Liu L, Yu B, Maluk C, Huang G B, Wang H, Song P G. ACS Nano, 2021, 15(7): 11667.
[31]

He H F, Liu L, Liu X H, Ding H L, Wang C S, Zhang W, Lei Y, Wang L C, Yu B. Chem. Eng. J., 2025, 519: 165179.
[32]

Huang J, Yang L T, Wang W, Yu B, Cao C F, Bi Z Y, Wang J, Shen J, Song L. Prog. Org. Coat., 2024, 186: 107995.
[33]

Cui T Y, Zheng Y P, Hu M D, Lin B C, Wang J W, Cai W, Fei B, Zhu J X, Hu Y. Small, 2024, 20(35): 2312083.
[34]

Wang H, Wang Y J, Li T Y, Yu C, Lin P, Liu J P, Lan Y X, Pan Y T. Adv. Funct. Mater., 2025, 35(25): 2500800.
[35]

Cui Y Y, Zhang M Y, Li X, Guo L J, Cai Y K, Li H J, Wu Y C. Polym. Degrad. Stab., 2025, 240: 111499.
[36]

Hu C Y, Liu F Q, Luo Z Y, Yi Q Q, Zeng Y N, Chen Y H, Wang C Y. Prog. Org. Coat., 2024, 187: 108191.
[37]

Li J L, Song Y H, Song G J, Li Z W, Li Z J, Yang X R, Li X R. Chem. Eng. J., 2024, 498: 155628.
[38]

Meng D, Liu X D, Wang S H, Sun J, Li H F, Wang Z W, Gu X Y, Zhang S. Compos. Part B Eng., 2021, 219: 108886.
[39]

Hu C Y, Li J W, Pan X J, Zeng Y N. Ind. Crops Prod., 2023, 200: 116828.
[40]

Qu M N, Dong Y R, Liu Q H, Wang Y Q, Feng P, Zhang Y, Deng Y, Zhang R Z, Sun C L, He J M. ACS Appl. Mater. Interfaces, 2025, 17(1): 2105.
[41]

Ma Z W, Zhang J Z, Liu L, Zheng H, Dai J F, Tang L C, Song P G. Compos. Commun., 2022, 29: 101046.
[42]

Chen Z, Chen W H, Liu P J, Liu Y, Liu Z X. Compos. Part A Appl. Sci. Manuf., 2021, 150: 106598.
[43]

Sun S J, Li T T, Yan G L, Jiao Y H, Xu J Z, Ma H Y. Chem. Eng. J., 2025, 518: 164618.
[44]

Yang X M, Gao S L, Xiao X Q, Lv J X, Wang Q T, Gohs U, Harre K, Xiao D. Constr. Build. Mater., 2025, 482: 141530.
[45]

Zhang T T, Huo S Q, Ye G F, Wang C, Zhang Q, Xue Y J, Song P G, Wang H, Liu Z T. Polym. Degrad. Stab., 2024, 230: 111047.
[46]

Xing W Y, Xi J C, Qi L Y, Hai Z B, Cai W, Zhang W J, Wang B Y, Chen L, Hu Y. Compos. Part A Appl. Sci. Manuf., 2023, 169: 107512.
[47]

Zhang L, Huang Y B, Dong H R, Xu R Z, Jiang S H. Compos. Part B Eng., 2021, 223: 109149.

Funding

National Natural Science Foundation of China(52173213)

PDF (3660KB)

120

Accesses

0

Citation

Detail
Sections
Recommended

/

〈 〉

Address of Editorial Office: No. 33 Beisihuan Xilu, Zhongguancun , Beijing 100190, China
Tel: +86 10 8262 7757
Fax: +86 10 8262 7757
E-mail: progchem@mail.las.ac.cn
http://www.progchem.ac.cn
    • 　
    Copyright © Editorial Office of Progress in Chemistry