Traditional polymer nanocomposites are characterized by strong friction of short-range segmental relaxation and free oscillation of several segments (Fig. 1). Segmental relaxation is accompanied by a precipitous decrease in the modulus of several orders of magnitude in a narrow frequency range, leading to dynamic mechanical instability. Therefore, to obtain stable dynamic mechanical polymer nanocomposites with high energy dissipation, we need to significantly broaden the relaxation time distribution of polymer chains in the system, achieved by tailoring full-scale polymer dynamics through the method of single-chain confinement (Fig. 1a(ii)). The relaxation modes can be mainly divided into segmental dynamics, Rouse dynamics and chain diffusion according to the length scale (Fig. 1b(ii)). Segmental dynamics is dominated by the dynamic heterogeneity of cooperative regions, where cooperative regions take 2–4 nm transient molecular clusters reorienting distinct configurations24,25. Segmental dynamics near the nanoparticle surface is heterogenous. Their relaxation time decreases smoothly from the solid surface into the matrix. The surface-to-surface distance (d) of nanoparticles is small enough that the interfacial zones overlap, broadening their relaxation time distribution. Rouse dynamics, also called intermediate dynamics, is the collective motion of several chain segments26. Polymer chains adsorbed atop a solid surface can form multiple conformations (tails, loops and trains)27 with different activated energy barriers of Rouse dynamics, leading to hierarchical relaxations of several chain segments. Due to the presence of nanoparticles, the chain diffusion is geometrically confined and subjected to viscous drag force, where the chains exhibit sticky reptation28. In this case, the particle obstacle hinders chain diffusion, different from the standard chain entanglements occurring in the flow state. Moreover, in the nanoconfined space of the chain size, all matrix chains dynamically couple with adsorbed chains on nanofiller interphase, providing an effective non-crosslinking stiffening mechanism for polymeric materials. This can further improve the dynamic mechanical stability of polymer nanocomposites, while also ensuring their high static mechanics.Fig. 1: Design concept for ultra-stable dynamic mechanics in polymer nanocomposites.a Mechanical properties and chain relaxation time distribution of traditional polymer nanocomposites and complex fluid nanocomposites. b Strategies for tailoring full-scale polymer relaxation induced by single-chain confinement in polymer nanocomposites. (i) Traditional nanocomposites have only strong frictional short-range segmental dynamics and free oscillating Rouse dynamics. (ii) Full-scale regulation of polymer dynamics across an exceptionally broad timescale can be achieved by the method of single-chain confinement in complex fluid nanocomposites. In this case, we regulate the chain dynamics at different length scales, divided into the segmental dynamics, Rouse dynamics and chain diffusion. The relaxation time of segmental dynamics decreases smoothly from the solid surface into the matrix, exhibiting heterogenous dynamics. The surface-to-surface distance (dss) of nanoparticles is so small that interfacial zones overlap. Rouse dynamics is the collective motion of several chain segments. Polymer chains adsorbed atop a solid surface can form multiple conformations (tails, loops and trains) with different activated energy barriers of Rouse dynamics, leading to hierarchical relaxations of several chain segments. Due to the existence of nanoparticle obstacles, the chain diffusion is geometrically confined, where chains are subjected to viscous drag force and their sticky reptation occurs.To validate the above proposed concept, these synthesized polymer nanocomposites should contain linear polymer fluids, nanofillers and multicomponent networks. We selected three polymers with extremely different glass transition temperatures (Tg) of poly(stearyl methacrylate) (PSMA), poly(lauryl methacrylate) (PLMA) and poly(tert-butyl acrylate) (PtBA) for random copolymerization to form a homogeneous copolymer network, regulating segmental dynamic heterogeneity in the system29 (Supplementary Fig. 1). Meanwhile, these polymers contain long side chains, which provide strong internal friction of chain segment relaxation. The polymer fluid is formed by flexible PLMA with a low Tg (−65 °C). By adjusting the chain length through the molecular weights of PLMA (Mn (PLMA fluid)), the polymer radius of gyration (Rg) of the polymer fluid can be precisely tailored, and their associated calculation is shown in Methods. The molecular characteristics of all synthesized PLMA fluids are provided in Supplementary Fig. 2 and Supplementary Table 1. The added nanofillers are hydrophobic silica (SiO2) nanoparticles, because their aggregate state is better controlled than that of other dimensions of nanofillers. The specific preparation strategy of our complex-fluid-gels (CFGs) is shown in the Supplementary Information. The mechanical properties of CFGs are mainly controlled by the following structural parameters: weight fraction of PLMA fluid (ΦPLMA fluid), molecular weight of PLMA fluids (Mn (PLMA fluid)), SiO2 nanoparticle size, weight fraction of nanoparticles (ΦNP), and weight fraction of network components (ΦNetwork). The molecular characteristics of all synthesized materials are listed in Supplementary Tables 2, 3, 7 and 8.We investigated the dynamic response properties of the CFGs with different molecular structures by employing an oscillatory strain rheometer, and constructed the dynamic master curves of CFGs by time-temperature superposition (TTS). We first analyzed the effect of chain length (Mn (PLMA fluid)) of the polymer fluid on the dynamic mechanical properties of CFGs. Without the polymer network, the rheological behavior of PLMA with various Mn was investigated at 25°C as shown in Supplementary Fig. 3. The low Mn PLMA fluids (Mn = 24k and 75k) display viscous flow behavior and lack the entanglement platform region, whose curves show the characteristics of unentangled polymer melts. In contrast, the higher Mn (PLMA fluid) fluids (Mn (PLMA fluid) larger than critical entanglement molecular weight (Me = 144k)) exhibit entanglement plateau region30. Subsequently, we further investigated the viscoelastic behavior of the polymer fluid with nanoparticles in this system. Supplementary Fig. 4 shows the frequency dependence of master curves of storage modulus (G′) and loss factor (tanδ) for PLMA fluids with ΦNP-14nm = 20% and varying Mn (PLMA fluid). Due to the existence of nanoparticles, their G′ is significantly improved, and their whole chain relaxation time is several orders of magnitude slower than that of pure PLMA fluid. As Mn (PLMA fluid) increases to 165k (larger than Me), the nanocomposite fluids do not exhibit viscous flow behavior in the low frequency region, but show gel-like characteristics. It arises from the dynamic percolating network formed between the entangled polymer chains and nanoparticles31. As Mn (PLMA fluid) is further increased, their G′ slightly increases and tanδ gradually decreases, because PLMA fluid is too entangled to relax.In the CFGs, G′ shows a similar trend to the nanocomposite fluids, increasing with the increase of Mn (PLMA fluid) (Fig. 2a(i)). The situation is quite different for tanδ in the mid/low frequencies. When Mn (PLMA fluid) is less than 165k (Me), tanδ increases with the increase of Mn (PLMA fluid) (Fig. 2a(ii)). The modulus (G′ and loss modulus (G′′)) of polymer fluids in the network are low, since chain length of polymer fluids is too short to be entangled and they are in the viscous flow state, as can be seen from Supplementary Fig. 4. In this case, their energy dissipation (tanδ) is low, like elastomers with crosslinked polymers. As Mn (PLMA fluid) is larger than Me, tanδ decreases with the increase of Mn (PLMA fluid), for the same reason for nanocomposite fluids. To ensure optimal mechanical and damping properties of CFGs, we should regulate Mn (PLMA fluid) of the added PLMA fluid slightly higher than its own Me. To further understand these interactions, we performed amplitude sweeps of CFGs with varying Mn (PLMA fluid). As shown in Supplementary Fig. 5, all CFGs do not show obvious modulus reduction (Payne effect) at small strain, indicating that there are no aggregations of nanoparticles in the PLMA matrix. With increasing strain, CFGs with high molecular weight exhibit a more pronounced Payne effect. It arises from the breakup of the bridging effects between the nanofiller surface and polymer chains (Supplementary Fig. 5b). In addition, CFGs containing high Mn PLMA show a more pronounced reduction in modulus after 100% cyclic strain (Supplementary Fig. 5a and c). This suggests that there are damaged physical interactions or topological structures that cannot be restored in the short term.Fig. 2: Rheological behaviour of complex-fluid-gels (CFGs).a Frequency dependence of master curves of storage modulus (G′) and loss factor (tanδ) for CFGs with Φnanoparticle = 20% and varying Mn (PLMA fluid). b (i and ii) Frequency dependence of master curves of G′ and tanδ for CFGs with Φnanoparticle = 5% and varying particle sizes. (iii and iv) Frequency dependence of master curves of G′ and tanδ for CFGs with varying Φnanoparticle of 14 nm nanoparticles. In the CFGs, changing the content and size of particles is essentially to regulate the surface-to-surface distance of particles (dss). In all CFGs, master curves were prepared by the time-temperature superposition treatment at 25°C with a constant shear strain of 0.5%, and the weight ratio of crosslinker / PLMA monomer / PLMA fluid is 3:100:150. Mn (PLMA fluid): molecular weight of PLMA fluids, Φnanoparticle: weight fraction of nanoparticle, ΦPLMA fluid: weight fraction of PLMA fluid.We then investigated how the surface-to-surface distance of nanoparticles (dss) affects the dynamic mechanical properties of CFGs. The dss is controlled by two structural parameters: size and content of nanoparticles. Master curves of CFGs with different particle sizes are shown in Fig. 2a(i) and (ii). The storage modulus (G′) and loss factor (tanδ) increase as the particle size decreases. The reason is that the dss of nanoparticles decreases as their size decreases, significantly improving the dynamic coupling between the interphase of nanoparticles and the polymer matrix. This can be confirmed by the fact that the whole chain relaxation of the PLMA fluid with ΦNP = 5% slows down as the particle size decreases (Supplementary Fig. 6). In addition, we further investigated the effect of dss on the mechanical properties of CFGs by controlling particle content (ΦNP) (Fig. 2b(iii) and (iv)). G′ increases with increasing the content of particles, while the peak value of tanδ becomes smaller and its changing trend displays more stable. Without the polymer network, the value of tanδ curve of the polymer fluid with high particle content (ΦNP = 20%) remains about 0.5 at frequencies from 10−2 to 102 rad/s (Supplementary Fig. 7), which shows frequency-insensitive energy dissipation property. As a consequence, we can achieve dynamic mechanical stability with high energy dissipation in CFGs by controlling the aggregation of complex fluids.To analyze such an enhancement mechanism of dynamic mechanical stability in CFGs, we investigated the law of chain relaxation with the particle aggregation. Figure 3a shows SAXS results of the CFGs with different contents of 14 nm nanoparticles. The curve of the CFG with low nanoparticle loading (Φ14nm-NP = 1%) shows good agreement with a single NP form factor, obtained from the dilute solution of well-dispersed nanoparticles (2 mg/mL in toluene). These data show that the nanoparticles are well dispersed in the PLMA matrix. All scattering curves of nanoparticle systems with different loading show a plateau in the low wave vector (q) region, which is a signature of essentially uniform distribution of nanoparticles. In this low-q region, we plotted ln(I) as a function of q2 (Supplementary Fig. 8). This curve is linear, an indication of the uniformity of size, and its slope is equal to −16.58 nm2. Since its slope is -Rg(NP)2/3 (Rg(NP), nanoparticle radius of gyration), 2Rg(NP) can be calculated to be 14.2 nm, in agreement with the values reported by the supplier32. Moreover, the Porod region slopes are all approximately equal to −4, further confirming the individual nanoparticle dispersion with sharp interfaces33. The inset for the photograph of CFG with Φ14nm-NP = 20% shows excellent transparency, also indicating the good dispersion of nanoparticles to some extent.Fig. 3: Stepwise chain relaxation of complex fluid in the CFGs.a SAXS results of the CFGs with varying content of 14 nm nanoparticles. The SAXS shows well-dispersed nanoparticle systems at the different loadings, as manifested by a low-q plateau. Moreover, the slopes in the Porod region are all −4, further confirming the individual nanoparticle dispersion with sharp interfaces. b The surface-to-surface distance (dss), dss / 2Rg (polymer radius of gyration) and dss / a (entanglement tube diameter) as a function of weight fraction of nanoparticles (ΦNP) in the CFGs. As ΦNP exceeds 10% (gray region), dss is smaller than the entanglement tube diameter and the chain size (2Rg), indicating that the polymer chains are geometrically constrained. c Frequency dependence of storage modulus (G′) of PLMA fluids with increasing ΦNP. In the low frequency region, G′ increases as an approximate power law function, G′(ω) ~ ωα. When ΦNP is larger than 10%, the power-law exponent α obviously less than 2. d Temperature dependence of the shift factor (αT) of PLMA with varying nanoparticle loading. The solid lines are fitted by WLF equation, where 20% data simultaneously is ideally matched with Arrhenius equation. Fitting parameters are listed in Supplementary Table 6. e The illustration of stepwise chain relaxation mechanism for CFGs.Based on these characterization results of nanoparticle dispersion in the polymer matrix, the surface-to-surface distance of nanoparticles (dss) can be calculated by the random distribution of spheres. The specific calculation is 2RNP[(2 / πΦNP)1/3 − 1], where RNP is the radius of the nanoparticle7. To analyze the effect of the nanoparticle confinement on chain relaxation, we calculated dss / 2Rg (polymer radius of gyration) and dss / a (entanglement tube diameter) of CFGs with varying Φ14nm-NP (Fig. 3b and Supplementary Table 4), where a was calculated in Methods. When ΦNP exceeds 10% (gray region in Fig. 3b), dss is smaller than the entanglement tube diameter and the chain size (2Rg), indicating that the polymer chains are geometrically constrained.We systematically compared the polymer chain dynamics of nanocomposite PLMA fluids with varying Φ14nm-NP. As the temperature dependent heat capacity (Cp) of nanocomposite PLMA fluids shown in the Supplementary Fig. 9, the glass transition exhibits a noticeable broadening when Φ14nm-NP increases, while the overall strength of this transition is not affected significantly. This can be inferred that the segmental relaxation of polymer chains adsorbed onto the nanofiller interface is not fully suppressed, while presents an obvious broadening of the relaxation time distribution34. Figure 3c shows the frequency dependence of their storage modulus (G′). In the low frequency region, G′ increases as an approximate power law function, G′(ω) ~ ωα, where the power-law exponent α equal to 2 is a characteristic of the whole chain relaxation of pure polymers (Φ14nm-NP = 0). This α is strongly dependent on the nanoparticle concentration. As Φ14nm-NP increases to 20%, the α decreases from 2 to 0.15, indicating that the frequency dependence of G′ becomes quite small and the neat liquid-like PLMA gradually transforms into a gel-like composite fluid7. In addition, the α (α = 0.15) of PLMA with Φ14nm-NP = 20% is almost unchanged over a broad frequency range (about 104 ~ 10-2 rad/s), reflecting that polymer chains exhibit a considerably wide distribution of short range chain relaxation.To understand this transition of chain relaxation, we summarized the essential information on the temperature dependence of chain dynamics of complex fluids in Fig. 3d, Supplementary Fig. 10 and Supplementary Table 5, where the frequency shift factor (αT) of PLMA with varying nanoparticle loading is plotted against the inverse temperature with Tref = 298 K. The solid lines are fitted by the WLF equation, and their fitting parameters are listed in Supplementary Table 6. More interestingly, the data of PLMA with Φ14nm-NP = 20% simultaneously is ideally matched with Arrhenius behavior with activation energy Ea = 112 kJ/mol, reflecting that interchain interactions become less important in the global dynamics35. These facts can be explained that the interfacial adsorbed chains are indeed highly mobile internally with no glassy nature while their center-of-mass diffusion and some Rouse modes are suppressed due to adsorption.Based on our results, we proposed the stepwise chain relaxation mechanism of CFGs to enhance the dynamic mechanical stability (shown in Fig. 3e). There are several relevant length scales to consider, namely, surface-to-surface distance of nanoparticles (dss), polymer chain size (2Rg), and entanglement tube diameter (a). At Φ14nm-NP = 20% where dss is significantly less than 2Rg, all polymer chains are adsorbed atop nanoparticles, forming multiple conformations (tails, loops, and trains) with different entropic barriers of Rouse dynamics for polymer chains. This can lead to hierarchical relaxations of several chain segments. Under these circumstances, the dss is also less than a. Due to the existence of particle obstacle, the chain diffusion is subjected to viscous drag force, where sticky reptation of chains will occur. Hence, we have realized the design concept for the regulation of stepwise chain relaxation with multiple length scales proposed in Fig. 1. Moreover, in the absence of the “free” chains at high ΦNP, all matrix chains dynamically couple with adsorbed chains on nanofiller interphase, providing an effective non-crosslinking stiffening mechanism. This endows polymer nanocomposites with high static mechanics, resulting in the improvement of their dynamic mechanical stability.To maximize the dynamic mechanical stability and energy dissipation performance of CFGs, we also need to regulate segmental dynamics of polymer chains, arrived at via adjusting polymer component heterogeneity (Fig. 4a). Specifically, the solution is to design a multicomponent network by random copolymerization of multiple polymers with extremely different glass transition temperatures (Tg), namely, poly(stearyl methacrylate) (PSMA, Tg = -100°C), poly(lauryl methacrylate) (PLMA, Tg = -65°C), and poly(tert-butyl acrylate) (PtBA, Tg = 44°C). Furthermore, these polymers contain long side chains that provide strong internal friction of chain segment relaxation. As a proof of principle, we prepared a series of CFGs by adjusting the proportion of polymer components, as shown in Supplementary Tables 7 and 8. First, we had to consider the phase separation behavior of polymer components, leading to mechanical instability of CFGs. Figure 4b and Supplementary Figs. 11 and 12 show AFM images of CFGs with different proportions of network components. As the PSMA content (ΦPSMA = 10%) of the CFG (CFG(2L, 4t, 2S, 0%NP)) is 10%, there is no phase separation behavior in this system (Fig. 4b). The area of white bright domain gradually increases with the increase of PSMA content, indicating that the degree of crystallization in CFGs increases. To confirm that the white bright domain is formed by PSMA crystallization, we used the PeakForce QNM model of AFM to characterize the property that the crystallization melting will be accompanied by a sharp change in the modulus. Supplementary Fig. 13 shows the modulus versus temperature of bright domain in the CFG’s AFM phase image (Fig. 4b). The modulus decreases sharply in the temperature range of 22 ~ 35°C, basically consistent with the reported crystallization melting temperature of PSMA.Fig. 4: Regulation of segmental dynamic heterogeneity.a Macromolecular structure design of segmental dynamic heterogeneity for the complex-fluid-gels (CFGs) by random copolymerization to form homogeneous copolymers. b AFM phase images of the CFGs with varying proportions of network components. c Dynamic master curves (i and ii) and DSC curves (iii) for the CFGs with varying proportions of network components, respectively. PSMA poly(stearyl methacrylate), PLMA poly(lauryl methacrylate), PtBA poly(tert-butyl acrylate).In addition, we employed the rheometer and differential scanning calorimetry (DSC) to investigate the phase separation behavior of CFGs with multicomponent network. The temperature dependence of master curves of G′, G′′ and tanδ for CFG(2L, 4t, 2S) shows a fairly stable trend (Fig. 4c(i) and (ii)), and there is no crystallization melting peak in its DSC curve (Fig. 4c(iii)). With the further increase of ΦPSMA, their dynamic master curves show a pronounced peak (Fig. 4c(i) and (ii)), due to the abrupt change of viscoelastic behavior caused by the melting of PSMA crystalline phase in CFGs. In this case, their DSC curves show crystallization melting peaks (Fig. 4c(iii)), where the peak value increases with the increase of ΦPSMA and the peak position shifts to higher temperatures, indicating the increase of crystallinity in the system. These results validate that we can adjust the proportion of network components to interfere with PSMA crystallization behavior, and obtain a multicomponent network with a homogeneous phase, ensuring dynamic mechanical stability of CFGs.We systematically studied the effect of network components on energy dissipation and mechanical properties of CFGs. When the mass ratio of PSMA, PLMA, and PtBA in the network of CFGs is 1:2:1, the CFG(2L, 4t, 2S) exhibits the widest frequency range of high energy dissipation (tanδ > 0.4) and the most stable mechanical properties, that is, the modulus change is the smallest in the range of tanδ > 0.4 (Supplementary Fig. 14). Its temperature-dependence master curves also show the same trend (Supplementary Fig. 15). Specifically, CFG(2L, 4t, 2S) exhibits considerably high tanδ (0.4 ~ 0.8) and G′ (100 ~ 102 MPa) in the frequency range of 10-1 ~ 107 Hz/temperature range of -35 ~ 85°C (Fig. 5a). The variation values of tanδ and G′ are only within 0.4 and two orders of magnitude, respectively. Notably, its G′ change rate (k) and tanδ change (Δ tanδ) in the temperature region of tanδ > 0.4 are far less than that of reported nanocomposites as shown in Fig. 5b and c and Supplementary Table 9. Their dynamic mechanical stability (rate of modulus change versus temperature (k)) is 10 times higher than that of conventional polymer materials, indicating that the CFG exhibits ultrahigh dynamic mechanical stability and energy dissipation over a broad frequency range. To visualize these properties of the CFG, acoustic absorption tests were conducted (Fig. 5d). The CFG can effectively attenuate sound waves over a broad frequency range. The sound absorption coefficient of the CFG (larger than 0.3) is significantly higher than that of commercial damping materials under any frequency band, and its curve displays a stable trend (Fig. 5e). In the electric motor vibration absorption demonstrative experiment, the CFG can effectively absorb vibration in all frequency bands, significantly better than commercial damping materials (Supplementary Fig. 16). In addition, the peak amplitude dissipation of vibration acceleration applied to our material is maintained at about 70% over a wide temperature range of −40 ~ 80°C (Fig. 5f and g), exhibiting stable dynamic mechanics with high energy dissipation. It can be seen that the application of our material can greatly improve the stability and reliability of electronic equipment in complex environments.Fig. 5: Stable dynamic mechanical CFGs with high energy dissipation.a Analysis of responsive mechanical properties of the CFG(2L, 4t, 2S). Dynamic master curves of storage modulus(G′) (square) and loss factor (tanδ) (circle) as a function of frequency (black) and temperature (red). b, c Comparison of rate of G′ change (k) and tanδ change (Δtanδ) in the temperature region of tanδ > 0.4 and temperature breadth of tanδ > 0.4 for CFGs and the reported polymer nanocomposites. k = log (G′0 / G′t) / ΔT, Δtanδ = 0.4 – tanδmax. All mechanical property parameters of the reported polymer nanocomposites are listed in Supplementary Table 9. d, e Acoustic absorption experiments and sound absorption coefficient as a function of frequency in the nanocomposites. The CFG can effectively attenuate sound waves over a broad frequency range. f, g Electric motor vibration absorption demonstrative experiments of polymer nanocomposites. NR natural rubber, PDMS polydimethylsiloxane.
