A biomimetic liquid metal cell

Abstract

Gallium-based liquid metals, as a broad category of emerging functional materials with unique physical, chemical, and biological properties, offer numerous possibilities for advancing intelligent systems. However, a basic query persistently remains for complex liquid metal systems: is there a minimal functional unit that can fully capture their diversity of morphology and function? Cells, as the most basic structural and functional units of life, are small in scale but have complex structures, functions, and life activities. Analogous to nature, this article proposes the concept of biomimetic liquid metal cells and systematically explores their construction routes, sensing capabilities, motion behaviors, and potential applications. We first construct a multi-phase composite structure with a liquid metal as the nucleus, an ionic solution as the cytoplasm, and a polymer as the cell membrane by developing a layered cryogenic molding method. Furthermore, we reveal that liquid metal cells exhibit inherently responsive characteristics and self-adaptive behaviors to thermal, pressure, chemical, electrical, and magnetic fields, indicating “small world, vast potential”. Based on these fundamental findings, we finally demonstrate the feasibility of utilizing liquid metal cells as sensors, fluidic valves, and material transport carriers in flow channels through dynamic control.

---

1. Introduction

Gallium-based room temperature liquid metals have emerged as revolutionary functional materials that attract considerable attention due to their unique physicochemical properties and widespread applications in heat transfer,¹ flexible devices,^2,3 reconfigurable systems,^4,5 catalytic synthesis,⁶ environmental governance,⁷ and biomedical technologies.⁸ Additionally, they exhibit excellent biocompatibility and environmental compatibility.^9,10 Recent discoveries have revealed their remarkable autonomous and external-field responsive behaviors, enabling innovations such as self-driven soft robots,¹¹ transition-state machines,¹² energy absorption and conversion,¹³ phagocytosis,¹⁴ memory function,¹⁵ and liquid-metal amoebae.¹⁶ These behaviors closely resemble biological processes, such as cell migration, cellular deformation, metabolism, cell differentiation, and active transport.¹⁷ These biomimetic characteristics indicate that liquid metal-based systems represent a highly promising platform for bridging the gap between artificial materials and living matter.^18,19 Moreover, they address the inherent limitations of conventional non-metallic artificial biomimetic systems, including those based on proteins,²⁰ polysaccharides,²¹ hydrogels,²² ionic conductors,²³ liquid crystal elastomers,²⁴ and silicone elastomers,²⁵ in replicating natural life processes. These limitations stem from the fundamental material properties, particularly in energy conversion, multi-field responsiveness, autonomous movement, environmental adaptability, and the scalability of functions and applications.^26–30 While these pioneering systems have provided valuable insights into artificial life, their material constraints often result in compromised performance in one or more of these critical aspects.

However, almost all the previously observed bio-like phenomena in liquid metals occur in centimeter-scale droplets due to the size limitations of surface tension-dominated fluid manipulation. The autonomous three-dimensional motion and deformation of large-scale liquid metal masses remain technically challenging. This scale limitation is a significant barrier to constructing a comprehensive artificial living matter system, which ideally should span from microscopic units to macroscopic assemblies. To overcome this constraint and expand the scale of liquid metal living matter systems, a bottom-up approach is essential.³¹ This approach requires developing fundamental structural units that both exploit the unique properties of small-scale liquid metals and enable the assembly of complex functional systems through ordered organization.

Learning from nature, cells are the most fundamental structural, functional, and biological units of life. Creating artificial cells with specific sensing capabilities through laboratory means would represent a significant breakthrough in the field of artificial life. The natural cellular structure, mainly consisting of a membrane, cytoplasm, and nucleus (Fig. 1A), provides valuable insights into liquid metal living matter systems. The realization of liquid metal living matter at various scales can be approached through the construction of minimal liquid metal cellular units. In this configuration, liquid metal droplets serve as the nucleus, providing basic perception and sensing functions. The surrounding aqueous solution acts as the cytoplasm, facilitating interactions between the metallic core and the external environment. The elastic cell membrane isolates individual liquid metal units while maintaining the dynamic dominance of the electric double layer over the small-scale liquid metal nucleus, simultaneously allowing material exchange with the environment. The term “cell” here represents a hierarchical and structural analogy to define an independent functional micro-unit within this artificial system. While fundamentally distinct from biological entities governed by metabolic processes, these engineered units successfully replicate the systemic role of natural cells. They establish defined boundaries, contain internal functional media, and exhibit dynamic environmental responsiveness. Conceiving this biomimetic architecture represents a crucial step toward realizing programmable and complex behaviors in liquid metal living matter.

Fig. 1 Fabrication procedure scheme and typical structure demonstration of LMCs. (A) Zebrafish fertilized egg; (B) cross-section of a typical LMC in the frozen state; (C) schematic illustration of the fabrication protocol of LMCs; (D) side view of a typical LMC at room temperature; (E) a variety of LMCs with different morphologies manufactured via the layered cryogenic molding method; (F) diffusion of a pigment solution encapsulated in the internal vesicle across the concentration gradient into the external solution environment; (G) microscopic structure of the cell membrane portion of the LMC cross-section.

Based on this hierarchical biomimicry, this study proposes for the first time the concept of biomimetic liquid metal cells (LMCs). As the fundamental structural and functional units of liquid metal living matter, LMCs are an extensible and multifunctional platform that fully exploits the comprehensive properties of liquid metals, possessing the potential for bottom-up assembly into more complex tissues and even systems. First, we elaborate on the structural design, material composition, and fabrication protocol of LMCs. Then, we explore the response of LMCs under various physical and chemical fields and disclose that LMCs have natural sensing functions and respond immediately to environmental changes. In addition, LMCs can also actively exhibit electrophilic and magnetotropic behaviors followed by directional movements. Finally, as an illustrative example, we demonstrate the potential application of LMCs as sensors, fluidic valves, and material transport carriers in a converging channel. The LMCs developed herein significantly expand the functional scope of liquid metals and offer a robust conceptual framework for constructing sophisticated artificial systems from the bottom up.

2. Results and discussion

2.1. Scheme of the fabrication protocol

Fundamental design and representative structure of LMCs. Inspired by the archetypal structure of natural cells, we built a biomimetic tri-layer architecture to show the essential design principle of LMCs (Fig. 1B). As a representative implementation, we selected EGaIn as the liquid metal core that provides dynamic responsiveness and adaptability characteristics (SI Note 1). The surrounding solution is 0.5 mol L⁻¹ NaOH solution, which acts as a cytoplasm, facilitating the exchange of ions and molecules with the external environment while maintaining the morphology and properties of the liquid metal. The outermost layer consists of a calcium lactate/sodium alginate-poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (CaL/SA-PEDOT:PSS) composite hydrogel membrane, chosen for its highly hydrophilic nature, tunable properties, facile crosslinking requirements, and excellent encapsulation capabilities.^32–34 This composite hydrogel combines the biocompatibility of the CaL/SA hydrogel with the high conductivity, optical transparency, flexibility, dispersibility, and environmental stability of PEDOT:PSS.^35–37 It offers enhanced mechanical strength and conductivity while maintaining essential membrane functions,³⁸ and its characteristic blue coloration facilitates structural observation. Benefiting from the formation of a semi-interpenetrating polymer network, the LMC membrane exhibits excellent stability in aqueous environments (SI Note 2).³⁹ During fabrication, Ca²⁺ ions instantly crosslink the alginate chains into a three-dimensional hydrogel network, which physically entraps and anchors the PEDOT:PSS chains to prevent their leaching.^37,40 The LMCs maintained their structural integrity without any PEDOT:PSS leakage during all dynamic experiments conducted in NaOH solutions. Furthermore, the composite hydrogel demonstrated long-term stability without macroscopic dissolution even after prolonged storage in deionized water (Fig. S1).⁴¹

Fabrication strategies and derivations. To achieve optimal integration of these layers with distinct viscosities into a unified structure while maintaining structural stability and functional flexibility, we developed a novel fabrication method, termed the layered cryogenic molding method (Fig. 1C). This scalable approach, versatile in both size and shape control, enables the fabrication of multi-layer structures with diverse morphologies at the millimeter scale and above. This modular strategy enables flexible customization through rational selection and combination of polymers, ionic solutions, and metallic cores, whose compositional complexity can be further expanded through materials genome engineering.^42,43 This versatile platform facilitates systematic investigation of the intrinsic properties and response behaviors of LMCs across diverse configurations.

Utilizing this method, we first fabricated spherical LMCs as fundamental units and investigated their properties and functions. Upon rewarming, the high-density liquid metal core naturally settles at the bottom of the structure while maintaining its spherical geometry within a specific volume range (Fig. 1D). Based on the capillary height formula, one has:

where γ_LM and ρ_LM denote the surface tension and density of the liquid metal, and g is the gravitational acceleration.⁹

We determine that a stable spherical morphology is maintained when the volume of the liquid metal (V_LM) is below 16.904 μL. Consequently, we standardize the liquid metal content from 0 to 20 μL for subsequent experiments. Furthermore, this fabrication approach enables us to achieve a diverse array of millimeter to centimeter scale shape-mimicking constructs. Demonstrating the exceptional physical plasticity of the liquid metal system, these engineered structures visually resemble various biological architectures, including the amoeba, paramecium, triceratium, nucleated erythrocyte, epithelial cell, muscle fiber, and neuron (Fig. 1E). The membrane thickness can be controlled by adjusting either the hydrogel solution concentration or the duration of the bilayer composite structure in the undercooled gel solution post-freezing. This tunability enhances the versatility of the method and enables the customization of structural properties for specific applications. While the upper size limit is theoretically unrestricted, the lower limit is governed by the coupled constraints of membrane thickness, assembly precision, and the increasing dominance of the oxide layer at smaller scales (Fig. S2). The hierarchical structure uses the hydrogel shell and internal solution to isolate the liquid metal from direct air exposure, preventing the spontaneous formation of a native oxide skin (∼1 nm in thickness).⁴⁴ However, a unique dynamic interfacial oxide layer still forms during the response processes. At the micrometer scale, the favorable role of surface tension and the adoption of advanced techniques such as microfluidics may enable further miniaturization beyond the current method.

2.2. Permeability of LMCs

The fundamental basis for a liquid metal to establish a living matter system lies in its ability to respond to external stimuli. It requires membrane structures and intermediate layers within LMCs to enable efficient metallic core–environment interactions, making membrane permeability a critical parameter. To validate this transport property, we conducted experimental investigations by encapsulating concentrated pigment solutions within LMCs using the layered cryogenic molding method. It was observed that, over a short period, the pigment molecules gradually permeated the surrounding area, encapsulating the LMC and accumulating at the top due to density differential (Fig. 1F and Movie S1). The underlying mechanism of this transport phenomenon lies in the porous structure and 3D interconnectivity of the pores of the hydrogel membrane, as evidenced by scanning electron microscopy (SEM) analysis (Fig. 1G and Fig. S3). Quantitative analysis revealed a high porosity of 81.568% and an average pore diameter of 7.876 ± 1.910 μm. Because typical pigment molecules and hydrated ions are several orders of magnitude smaller than these micro-scale pores, they can easily traverse the 3D network. Consequently, the permeability of LMCs is fundamentally governed by a passive diffusion process driven by concentration gradients.

These structural characteristics enable LMCs to function as independent systems capable of efficient material exchange and information communication with the external environment through the membrane structure. Furthermore, by exploiting the inherent pH-dependent and temperature-dependent swelling characteristics of the composite hydrogel, this permeation process can be dynamically regulated. External stimuli can tune the effective pore size, thereby endowing LMCs with immense potential for controlled drug delivery applications. The demonstrated permeability confirms the LMC's potential for dynamic interaction with external environments, setting the stage for exploring its responsive behaviors in subsequent experiments.

2.3. Response of LMCs under external fields

Timely response to unknown changes and risks is fundamental for life systems to achieve survival, adaptation, and evolution. As a system parallel to the natural life system, the liquid metal living matter system must also confront complex and dynamic environments, rather than being perpetually protected within ideal laboratory conditions. As the foundation for establishing this artificial system, LMCs facilitate the transition of liquid metals from dependency on solution environments to functioning as independent entities under open-air conditions, thereby directly addressing challenges. Given that the proportion of the liquid metal serving as the core component is a critical design parameter that fundamentally influences the properties and functions of LMCs, we have systematically investigated the impact of liquid metal core content on LMC behavior (Fig. 2A). To explore fundamental stimulus-response mechanisms of LMCs, we systematically evaluated their responses to three representative external stimuli: mechanical, chemical, and thermal fields (Fig. 2B). Because all constituent materials of the LMC are electrically conductive, we utilize electrical resistance as a key parameter to effectively capture the dynamic responses of LMCs. It serves as a sensitive and quantifiable indicator of the structural and compositional changes within the cells when exposed to external stimuli. We discover that LMCs possess an inherent sensing capability.

Fig. 2 Responses under pressure, chemical, and temperature fields of LMCs, respectively. (A) LMCs with various nuclear proportions (V_LM = 0, 5, 10, 15, and 20 μL) under ambient temperature and atmospheric conditions; (B) LMCs placed in a simulated complex environment; (C) variations in the electrical properties of LMCs during longitudinal compression from 0 to 3.27 mm at 26 °C; (D) instantaneous effect of external pH changes on the longitudinal resistance of a LMC (V_LM = 15 μL) at 26 °C; E. resistance changes of LMCs in the temperature range from 15 to 40 °C.

Mechanical response. Mechanical stress is ubiquitous in natural settings, whether from fluid dynamics, gravity, or physical contact with surrounding materials. To investigate their mechanical resilience, LMCs were subjected to uniform vertical compression at a rate of 2.4 mm min⁻¹ (Movie S2). During this time, they rapidly detected the increasing pressure and exhibited enhanced electrical conductivity properties as a characteristic response (Fig. 2C₁ and Fig. S4). Within the compression range considered in this study, the resistance decreased with increasing liquid metal content, showing a gradually diminishing rate of change. To elucidate the underlying mechanism of these behaviors, we developed a simplified geometric model to analyze the electrical resistance change during the compressive deformation of cells (Fig. 2C₂). Following the principle of charge transport along electric field lines, the overall resistance can be simplified into four series-connected components based on the current path. The first component is a spherical liquid metal core with radius r_LM. The second component is a truncated cone-shaped solution section, having a total height of H together with the first part. Its upper surface, with radius r_T, represents the circular contact area between the solution and the hydrogel membrane interfacing with the top electrode and its side extensions tangent to the core. The remaining two components represent the resistance of the hydrogel in contact with the top and bottom electrodes, with contact areas A_T and A_B and thickness t_h (SI Note 3). According to this model, the resistance of the LMC (R_LMC) under compression can be expressed as:where σ_s, σ_LM, and σ_h are the electrical conductivity of the solution, liquid metal, and hydrogel, respectively.

As the compression distance increases, this model predicts an overall decrease in electrical resistance. It is primarily attributed to the reduction in cell height H, and the expansion of r_T, A_T, and A_B. The theoretical predictions demonstrate excellent agreement with our experimental observations. Additionally, the resistance and its change rate gradually decrease with increasing nucleus proportion under equivalent compression distances, which can be explained by the monotonic variation of R_LMC and its derivative with increasing r_LM within the investigated range (eqn (2)). This further demonstrates that the incorporation of a liquid metal enhanced the environmental adaptability of the overall structure. No structural failure occurs during the compression tests, showing the structural integrity and deformation ability of these composite structures. These mechanical properties enable LMCs to extend their adaptability and physical reconfigurability across diverse pressure environments.

Chemical response. In addition to physical field variations, chemical gradients in the natural environment are inevitable factors to which living matters must adapt. Therefore, we further investigate whether the LMCs could adapt to rapid changes in external chemical environments by maintaining their dynamic equilibrium. When exposed to a hydrochloric acid solution (HCl, 0.5 mol L⁻¹, 350 μL) with equivalent concentration and volume to the internal cytosol solution, LMCs (V_LM = 15 μL) exhibit distinct morphological and physicochemical responses (Fig. 2D). Due to the neutralization reaction, ion concentration within the cytoplasmic region reduces,⁴⁵ and a continuously thickening oxide layer develops on the surface of the liquid metal core.⁴⁶ Thus, the overall electrical resistance of the LMC rapidly increases.

Benefiting from the excellent permeability of the membrane structure, the ionic concentration within the intramembrane area gradually equilibrates with the external HCl solution after 20 minutes. The excess HCl solution effectively dissolves the oxide layer from the liquid metal surface, similar to the effect of NaOH solution, restoring its spherical morphology and electrical properties.⁴⁷ Since the conductivity of 0.5 mol L⁻¹ HCl solution is slightly higher than that of NaOH solution (17.8 S m⁻¹ for HCl and 9.31 S m⁻¹ for NaOH),⁴⁸ the overall resistance of LMCs in excess hydrochloric acid solution decreases significantly compared to that in the neutralization phase, and is slightly lower than the initial resistance in NaOH solution. These response patterns directly reflect the combined influence of the conductivity of the solution and dynamic interfacial oxide skin on the overall resistance of LMCs, as predicted by the theoretical resistance model (eqn (2)). This chemical equilibration also highlights the fundamental material exchange mechanism of LMCs. While the current system utilizes passive diffusion for basic material exchange and has not yet achieved the complex active transport functions of natural cells, this biosimilar architecture provides a robust platform. In future iterations, by leveraging the intrinsic cation selectivity of PEDOT:PSS or the multi-field responsiveness of liquid metals, we can engineer advanced functionalities that extend beyond the capabilities of natural life.⁴⁹

Thermal response. Temperature fluctuations are inevitable in practical operational environments. To further investigate their behavior under varying thermal conditions, LMCs are maintained under controlled isothermal conditions across different temperature regimes from 15 °C to 40 °C. During these evaluations, each LMC is consistently compressed to a longitudinal height of 7.4 mm to ensure a uniform deformation state throughout the testing period. By real-time monitoring of the average temperature at three points (top, middle, and bottom) and longitudinal resistance, we explore the impact of both internal liquid metal content and external temperature variations on the overall electrical properties of LMCs. Due to the high conductivity of the liquid metal, the LMCs exhibit significantly lower resistance compared to the equal-volume hydrogel vesicles containing only solution. The resistance of LMCs exhibits an inverse relationship with temperature elevation (Fig. 2E). To analyze this phenomenon, we first examine the thermal effects on the conductivity of individual components. With increasing temperature, liquid metals show decreased electrical conductivity (σ_LM), resulting from enhanced electron scattering caused by intensified atomic thermal motion and wider interatomic spacing:

σ_LM = σ₀ − α(T − T_m) + β(T − T_m)² (3)

where σ₀ is the electrical conductivity at the melting temperature, T_m is the melting temperature, and α and β are the temperature coefficient of variation.⁵⁰

For the solution, a decrease in its viscosity, an increase in the mobility of the ions, and an increase in the number of ions in the cytoplasmic solution caused due to the increase in temperature will lead to an increase in the conductivity (σ_S):

σ_S = σ_S0(1 + δ(T − 25)) (4)

where σ_S0 is the conductivity of the solution at 25 °C, and δ is the temperature coefficient of variation.⁵¹

Given that the maximum nuclear proportion in the entire cell is 10.233%, the alkaline solution, which performs cytoplasmic functions, continues to dominate the overall electrical properties, as its conductivity differs from that of EGaIn by approximately six orders of magnitude within the testing range.⁴⁸ As the nuclear proportion increases, the electrical conductivity of the composite structure progressively increases under the same external conditions, accompanied by a gradual decrease in the rate of resistance change. This phenomenon indicates that the liquid metal gradually regains primary control over the LMC's overall properties, thereby enhancing the LMC's capacity to withstand thermal perturbations. LMCs’ real-time sensing of temperature variations in both external and internal environments ensures stable and efficient operation of liquid metal living matter systems.

Thermal evaluations were capped at 40 °C to prevent structural failure. Higher temperatures may disrupt hydrogen bonds and increase polymer mobility, softening the hydrogel shell and risking severe deformation. The selected temperature window also effectively precludes thermoelectric interference. Given that the Seebeck coefficients of gallium-based liquid metals and gallium oxide are intrinsically minimal (typically on the μV K⁻¹ scale),^52,53 any thermoelectric contribution in the present configuration is negligible under our experimental conditions. Therefore, the measured thermal response purely reflects resistance modulation and interfacial electrochemistry, not thermoelectric conversion. While this 15–40 °C window demonstrates standard ambient sensing, exploring the LMCs’ full adaptability requires investigating extreme thermal boundaries.

2.4. Capability of resisting low-temperature conditions

While the previous section examined LMC behavior under moderate temperature variations (15–40 °C), extreme thermal conditions present more significant challenges for functional materials. Understanding the stability and performance retention of LMCs under extreme low-temperature conditions is crucial for evaluating their operational reliability across extended environmental ranges. Here, we explore their response to rapid cooling from above 0 °C to −13 °C, focusing on the property changes of both integral and individual components. Due to the high water content (95.39%) and the permeability of the hydrogel membrane, it is considered a unified system with the solution, and both exhibit synchronized phase transitions.

To evaluate the effect of a liquid metal on cold resistance, this section compares LMCs with pure vesicles of the same size containing only solution as a control. The control group exhibits distinct thermal behavior characterized by a temperature plateau at −5 °C, with resistance changes showing a negative correlation with temperature variations before and after this point (Fig. 3A). In stage (i), the hydrogel maintains its gel state while the solution remains liquid. A significant 103.143% step increase in resistance marks the phase transition between two distinct stages. Then, both components solidify in stage (ii).

Fig. 3 Changes in the response characteristics of LMCs under extreme low-temperature conditions. (A and B) Resistance and temperature change curves, as well as phase transition schematic during the cooling process of (A) a pure vesicle without the addition of a liquid metal (V_LM = 0 μL); (B) a LMC (V_LM = 20 μL).

In contrast, LMCs demonstrate a more complex cooling process (Fig. 3B). During stage (iii), resistance increases with decreasing temperature while maintaining the cell membrane's gel state and liquid phases of both solution and liquid metal. The cooling rates remain similar through stages (i) and (iii) due to the low volume fraction of the liquid metal within the LMCs (10.234%) and extensive cooling source. Similarly, there is no significant influence on the phase transition time of the outer components. A 98.439% step increase in resistance marks the transition to stage (iv), where all components except the liquid metal are solidified.

Stage (v) is characterized by the liquid metal reaching its crystallization point at −7.3 °C after a period of supercooling. The sudden release of latent heat significantly elevates the surrounding temperature, temporarily reverting the adjacent frozen components back to their stage (iii) state. Finally, in stage (vi), complete solidification occurs with stabilized resistance. Although liquid metals typically possess a lower specific heat capacity than aqueous solutions, their pronounced supercooling phenomenon and the subsequent rapid release of latent heat during crystallization elevate the overall temperature descent curve, thereby delaying the occurrence of the minimum temperature.⁵⁴

Overall, the incorporation of the metallic core fundamentally modifies the system's thermal behavior through phase change-induced thermal buffering. During the initial cooling process, this latent heat release rapidly elevates the internal temperature to 90.76% of that observed in the blank control group, while the ultimate temperature drop is minimized to 73.07%. This thermal modulation capability enhances the robustness of LMCs, providing improved structural integrity and functional preservation under extreme environmental conditions, which is the critical feature for practical applications such as cryopreservation or deployment in harsh environments.

2.5. Motion of LMCs

Beyond static responses, directional movement capabilities are essential for artificial living systems to interact with and adapt to their environment and perform complex tasks. For liquid metal living matter, the development of movement strategies should capitalize on their unique physicochemical properties, particularly their exceptional responsiveness to electromagnetic fields.^55,56 Powered by external fields, the controlled responses of LMCs provide a foundation for developing their environmental adaptability and functional capabilities. This section focuses on the changes in the position and morphology of LMCs under the tuning of electric and magnetic fields in a solution environment.

Electrical tuning behavior – electrotaxis. Electric fields permeate natural environments, influencing various cellular behaviors and biological processes.¹⁷ The electrical response behavior of LMCs represents another sophisticated manifestation of their field-responsive properties. We explored the electrical tuning behavior of LMCs in a 17 mm wide and 50 mm long quartz channel containing the solution at a depth of 13 mm (Fig. 4A₁). It is observed that LMCs exhibit significant directional movement in the applied voltage of higher than 7.91 V (V_LM = 20 μL). After an initial acceleration period, LMCs maintained uniform velocities while moving along the electric field lines from the cathode to anode (Fig. S6 and Movie S3).

Fig. 4 Directional motion characteristics of LMCs. (A) Movement process of a LMC (V_LM = 20 μL) over a period of 12 s at a potential difference of 12 V between two electrodes spaced 5 cm apart; (B) effect of volume of EGaIn in LMCs on the activation voltage and average motion speed; (C) rolling phenomenon of LMCs near electrodes; (D) resonance phenomenon of LMCs under alternating electric fields; (E) spatiotemporal tracking trajectories of a magnetic LMC and an external magnet during an outward spiral motion; (F) multidimensional motion of magnetic LMCs guided by a magnet in a vertical plane: (F₁) two-dimensional and three-dimensional motion of a magnetic LMC; (F₂ and F₃) two different motion morphologies of a magnetic LMC: (F₂) dropping and (F₃) suspension.

To further investigate the relationship between LMCs’ composition and electrical tuning behavior, we varied the V_LM from 5 to 25 μL. Our results reveal that both the activation voltage and movement speed were significantly influenced by V_LM. Specifically, the minimum voltage required for initiating movement shows a negative linear correlation with V_LM within a certain range (Fig. 4B). Under a fixed potential difference of 12 V, the steady-state velocity of LMCs shows pronounced growth in the 5 to 15 μL range of V_LM. However, the velocity increase becomes slower beyond 15 μL, likely due to the combined effects of increased mass and droplet deformation.

To elucidate the underlying mechanism of this field-induced directional motion, we developed a theoretical framework focusing on the driving force analysis of LMCs. Cells containing only solution without LM cores showed no movement under the electric field, demonstrating that the motion of LMCs is primarily driven by the nucleus. Therefore, the force analysis focuses mainly on the LMC nucleus (Fig. 4A₂). It is primarily subjected to forces resulting from viscous drag caused by the surrounding solution (f_η), frictional forces from the hydrogel membrane at the bottom (f_β), and surface tension imbalance arising from the electric double layer present between the two hemispheric sides of an LMC parallel to the direction of the electric field (F_γ).⁵⁷

F _γ is the main driving force in elucidating the intrinsic mechanisms governing the LMCs’ dynamic behavior. According to our force analysis model of LMCs (SI Note 4), this driving force can be expressed as:

where q₀ is the amount of charge stored in the electrical double layer at the liquid metal surface when treated as a capacitor, U_F is the potential difference between the electrodes on either side, L_F is the distance between two electrodes, r_LM is the radius of the liquid metal droplet, and h₁ and h₂ are the width and height of the channel filled with liquid, respectively.

Mathematical analysis of the equation reveals that the driving force exhibits positive correlations with both U_F and r_LM, while showing an inverse relationship with L_E. These theoretical predictions align well with our experimental findings, where larger droplets demonstrated enhanced mobility due to increased driving forces while smaller droplets required higher activation voltages for effective movement. The theoretical model not only provides a quantitative basis for understanding the size-dependent electrical response behavior but also provides insights into the key parameters governing LMCs’ motion.

As the LMC approaches the anode region, it exhibits a continuous in situ rolling behavior (Fig. 4C). Bare liquid metal droplets usually experience severe oxidation and unrecoverable deformation upon direct contact with the anode. The encapsulating membrane of LMCs effectively isolates the metallic core from the electrode. This structural isolation prevents severe oxidation while allowing the core to undergo significant internal deformation to maintain dynamic equilibrium. Within the LMC, the core repeatedly climbs and falls within the inner cavity. This internal displacement drives the hydrogel shell to roll in the opposite direction of the falling trajectory. This phenomenon explicitly demonstrates the synergistic mechanism and distinct physical roles of the structural components. The liquid metal serves as the highly conductive active core that provides the driving mass. The electrical double layer at the core–solution interface generates the primary surface tension gradient under the electric field. The dynamic interfacial oxide skin modulates this surface tension and stabilizes the core during continuous deformations. The hydrogel shell provides a spatial constraint and acts as a mechanical transmission track.

Furthermore, under an alternating current field, the metallic core within the LMC enters a resonance state at lower frequencies compared to a bare droplet (Fig. 4D and Fig. S8). This arises from the physical constraint imposed by the hydrogel membrane, which fundamentally alters the internal vibration modes of the metallic core. Meanwhile, since the high voltages applied under direct current fields inevitably induce electrochemical reactions on the liquid metal surface, the alternating current field is expected to mitigate gas accumulation within the internal cavity through rapid electrode polarity reversal.⁵⁸

Magnetic tuning behavior – magnetotaxis. Magnetic fields, as one of the fundamental physical forces on Earth, play crucial roles in biological navigation and spatial orientation.^59,60 The diversity of the liquid metal composites further extends the functional expandability of LMCs. To expand the locomotion capabilities of LMCs beyond electric field response, we incorporate magnetic responsiveness by replacing EGaIn with a liquid metal ferrofluid through the layered cryogenic molding method. As demonstrated by the spatiotemporal motion trajectories (Fig. 4E), the internal magnetofluid closely tracks the external magnet, physically driving the entire LMC structure. Detailed kinematic analysis reveals a high position correlation, averaging 0.91 during outward spiral motion and exceeding 0.99 in linear segments. Although rapid directional changes can induce transient positional displacement between the core and the shell due to inertial lag (SI Note 5 and Fig. S9), the overall synchronization remains highly robust.

To further investigate this magnetic field-driven locomotion in complex terrains, we utilize a custom-designed experimental setup comprising two platforms at different heights, with the magnetic LMC immersed in a 0.5 mol L⁻¹ NaOH solution (Fig. 4F₁). The magnetic LMCs exhibit distinct locomotion characteristics with remarkable multi-dimensional mobility, including lateral movement, anti-gravity motion, and corner-crossing motion guided by a NdFeB magnet (Movie S4). The sequential motion trajectory (positions i–iii) illustrates one-dimensional lateral movement confined to a horizontal platform. The liquid metal core preserves high surface tension and remains at the bottom because of gravitational effects (Fig. 4F₂). When the magnetic force in the vertical direction counterbalances the gravitational force, the LMC achieves controlled suspension. Sequential motion trajectory (positions iv–viii) shows the flexible two-dimensional motion of the magnetic LMC under the guidance of the magnet. It enters a suspended state, with the liquid metal core stabilized centrally within the cell‘s inner membrane to the vertical wall (Fig. 4F₃).

Because the hydrogel shell serves as a physical barrier enclosing the active core, the magnetic response of LMCs is fundamentally an intrinsic property. This structural isolation theoretically enables LMC manipulation across diverse media, a capability supported by our preliminary observations of successful actuation under both air and acidic environments. The current field-driven motion still requires continuous macroscopic energy input, whose scale is orders of magnitude higher than that of the localized chemical energy utilized by natural biological cells or autonomous liquid metal motors. LMCs possess significant potential for functional evolution, but both practical operation and the future development of multi-field coupling or self-powered systems must rigorously account for the influence of the surrounding medium. Environmental parameters such as fluid viscosity, density, interfacial friction, and the pH-dependent swelling state of the hydrogel shell collectively determine the hydrodynamic drag. Therefore, these medium-related factors will profoundly influence the velocity, trajectory precision, and overall ease of controlled manipulation of LMCs in practical application scenarios.

2.6. Dynamic control of LMCs in deformed channels

The emergence of LMCs represents a pivotal step in transforming liquid metal-based living matter systems from conceptual frameworks into practical realities. They provide an extensible, multifunctional platform that potentially expands the structural and functional possibilities of liquid metals across various domains. To demonstrate this potential, we showcase their capabilities in deformed channels (Fig. 5A), where they function as sensors, valves, and delivery systems, integrating electrical conductivity, magnetic responsiveness, and morphological adaptability. This demonstration highlights their environmental adaptability and multifunctionality, illustrating how LMCs can operate effectively in dynamic environments.

Fig. 5 Movement and deformation of LMCs in a deformed channel. (A) LMCs demonstrate promising potential for flow control applications. (B) Self-deformation and blockage functions of the magnetic LMC within the channel and its reverse flow motion under the influence of an external magnetic field: (B₁) the LMC moves along the direction of the fluid flow and deforms to adapt to the surrounding environment; (B₂) upon reaching the maximum deformation, the LMC becomes stationary and fills the flow channel, resulting in blockage; (B₃) under the guidance of a magnet, the LMC moves in the direction opposite to the flow, accompanied by shape recovery; (B₄) states of the LMC and flow are restored.

Here we designed a converging-bifurcating channel system with a width gradient from 12 mm to 6 mm to validate the potential of magnetized LMCs in flow control applications (Movie S5). Fig. 5B shows the experimental phenomena and the corresponding deformation of LMCs and flow states of this system. In the initial phase, a magnetic LMC (V_LM = 20 μL) is positioned at the channel's widest cross-section, and an aqueous solution is introduced at a flow rate of 3.5 mL s⁻¹ (Fig. 5B₁). Guided by an external magnetic field, the LMC migrates steadily along the flow direction. During this process, it demonstrates remarkable deformability by adapting to the narrowing channel geometry while maintaining its forward momentum. Upon reaching a channel segment matching its volume or the bifurcation, the LMC establishes complete occlusion and effectively halts the fluid flow (Fig. 5B₂). Due to the fixed volume constraint of the system and the unidirectional flow direction, this blockage creates a significant hydraulic pressure differential across the LMC, generating a 5 mm liquid level difference within 7 seconds, evidenced by distinct liquid levels on either side.

The system's reversibility is demonstrated through magnetic guidance, where the LMC can be manipulated to restore flow patency. Under magnetic control, the magnetized LMC undergoes controlled movement (Fig. 5B₃), allowing fluid redistribution through the newly created channel space. The liquid level difference between the opposing sides of the LMC demonstrates a decrease over time. The magnetic force exerted on the ferrofluid core effectively overcomes the hydrodynamic forces, enabling precise positional control of the LMC. The system ultimately restores normal flow patterns (Fig. 5B₄).

Additionally, the photothermal effect and photoacoustic effect of liquid metals enable advanced multi-modal wireless sensing and remote control strategies in complex flow channels while providing imaging capabilities.^61,62 LMCs achieve dynamic control through these liquid metal cores and are further integrated with two additional material delivery modalities, specifically the hydrogel matrices and the internal solution.^63–65 Leveraging the extensible material properties of the layered cryogenic molding method, channel recanalization and targeted material delivery can be achieved through the controlled degradability of hydrogels.^66–68 LMCs create a versatile platform for complex flow channel management applications.

3. Conclusions

In this study, we systematically present the concept, preliminary theory, and technology of LMCs as the fundamental structural and functional units of liquid metal living matter and successfully establish their fundamental architecture using a systematic bottom-up fabrication strategy. Through the layered cryogenic molding method, we developed a multi-layer and multi-phase biomimetic construct wherein the liquid metal core serves as the nucleus for responsive functionality. The reducing electrolyte solution works as the cytoplasm, simultaneously maintaining the metallic core's reductive state advantages while facilitating information and material exchange between the internal and external compartments. The outer membrane is an alginate-based composite hydrogel with a porous architecture, providing structural support and permeability for molecular and ionic transport. This LMC unit with a tri-layer architecture effectively addresses previous limitations of liquid metal applications in scale and environment, maintaining the morphological and physicochemical stability of liquid metals.

As the basic unit of liquid metal living matter, the response properties of such LMCs to environmental stimuli are comprehensively explored. Through systematic measurement and analysis of characteristic resistance, we find that LMCs possess remarkable sensing capabilities and environmental adaptability. They exhibit instantaneous and well-regulated resistance changes in response to varying temperature, pressure, and chemical fields. To clarify the response mechanisms induced by different environmental stimuli, we established a geometric resistance model that links electrical resistance to the structural deformation of an LMC. Both experiment results and the resistance model demonstrate that the integration of the liquid metal core into the LMC structure significantly enhances its perceptual ability and intrinsic homeostasis, with the model successfully capturing the key scaling behaviors observed across different compression states.

The multi-field tuning capability of LMCs builds the foundation for developing various external field responsive behaviors, the most typical of which are electrical and magnetic tuning behavior. Therefore, the movement controllability of LMCs under electric and magnetic fields was studied by varying the liquid metal nucleus proportion. It was found that LMCs are capable of controlled directional movement under electric fields, and LMCs with higher liquid metal addition ratios show more sensitive tuning ability, manifested by higher moving speed and lower activation voltage. By modeling the force analysis of LMCs, we found that the key to the LMC motion regulation by an electric field lies in the pressure difference caused by the surface tension gradient generated across the liquid metal nucleus. The magnitude of this force is closely related to the structural ratio between the solution and the liquid metal nucleus within the hydrogel membrane. Mechanical modeling demonstrates that in the scale range of liquid metal cells that we have studied, the increase in the proportion of liquid metal nuclei has a significant gain effect on enhancing the driving force generated by the cell subjected to the electric field. In addition, comparative experiments with blank groups confirm that pure solution vesicles without added liquid metal nuclei cannot be driven by electric fields.

Furthermore, by virtue of the extensibility of liquid metal composite material systems, LMCs can be functionalized with magnetic-tuning ability by incorporating ferromagnetic particles into the liquid metal core. The magnetic LMCs demonstrate precise spatiotemporal control over their locomotion and positioning in three-dimensional space. Such a contactless motion control approach, coupled with the fully flexible and biocompatible composite structure of LMCs, as well as the viability and versatility of the structural design methodology, jointly endows the LMCs with navigational ability through adaptive shape modulation. These unique structural and responsive characteristics of LMCs make them particularly promising for applications in non-uniform channels where wired control is not feasible. Finally, as a practical illustration, we designed a converging-bifurcating channel system with a width gradient to validate the potential of magnetized LMCs as sensors, valves, and delivery systems.

In conclusion, the conceptual establishment, response characterization, and application demonstration of LMCs in this paper provide an initial insight into the unique features of liquid metal living matter systems. By establishing the foundational principles of multi-field responsiveness and functional integration in these engineered constructs, this work offers a conceptual framework for future explorations in adaptive materials. However, numerous aspects remain to be explored to further enhance the intelligence level and expand the survival scenarios. While current LMCs maintain robust functionality during short-term operations, extended environmental exposure eventually leads to hydrogel shrinkage, water evaporation, and core oxidation. Therefore, it is crucial to optimize LMCs’ long-term stability through targeted material cross-linking and structural reinforcement, explore diverse material combinations, and develop more sophisticated multi-field control strategies. Furthermore, investigating collective behaviors and interactions among identical LMCs would facilitate their assembly into higher-order liquid metal tissues. Through the controlled fabrication processes we established, developing diverse LMCs with distinct shapes, functions, and adaptabilities would advance this system toward organ-level complexity. As reliable basic building blocks, LMCs provide a solid foundation for the assembly of more sophisticated liquid metal living matter, potentially leading to unprecedented applications in various fields.

4. Experimental

4.1. Materials

Unless otherwise stated, all liquid metal mentioned in this study refers to pure EGaIn. EGaIn is prepared by mixing 75.5 wt% gallium and 24.5 wt% indium, and melting the mixture in a vacuum oven at 150 °C for 6 hours. This choice is based on the well-characterized properties and the suitable melting point of EGaIn, which enable us to fabricate LMCs within the experimental temperature window and systematically investigate the states of both liquid and solid cores.

Sodium alginate (90%) and calcium lactate hydrate (C₆H₁₀CaO₆·xH₂O, 98 wt%) were purchased from Shanghai Macklin Biochemical Technology CO., Ltd. Poly(3,4-ethylenedioxythiophene)–poly(styrene sulfonate) (PEDOT:PSS, 1.1% concentration in H₂O, without surfactants, and exhibiting high conductivity) was purchased from Shanghai Aladdin Biochemical Technology CO., Ltd. Iron nanoparticles (99.9% metal basis, 100 nm average particle size) were purchased from Shanghai Macklin Biochemical Technology Co., Ltd. Deionized water used for solution preparation was obtained from an ultrapure water system (Direct-Q 3 UV, Merck Millipore, Germany).

4.2. Preparation of SA-PEDOT:PSS composite gel solution

PEDOT:PSS solution was diluted, and then sodium alginate (SA) powder was added to the mixture. After manual stirring until achieving a homogeneous mixture, the solution was degassed under a vacuum. The final concentrations were 0.033 wt% PEDOT:PSS and 1.5 wt% SA.

4.3. Fabrication protocol of the layered cryogenic molding method

Step 1: preparation of the pre-frozen liquid metal core. Target volumes of EGaIn droplets were dropped onto pre-cooled smooth ice blocks prepared by freezing 0.5 mol L⁻¹ NaOH solution, causing phase transition and transformation into relatively smooth solid spheres.

Step 2: formation of the solution–liquid metal double-layer ice core. The target Ecoflex mold (Fig. S10) was filled halfway with a 0.5 mol L⁻¹ NaOH solution and lightly shaken to remove any trapped air bubbles. The mold was then frozen at −8 °C for 6 hours until the solution completely solidified. Subsequently, the pre-frozen EGaIn sphere from Step 1 was placed in the center of the solid upper surface. Using a pre-frozen core is essential to prevent the latent heat released during liquid metal solidification from melting the underlying NaOH ice block. After an additional 1 hour of freezing at −8 °C, a chilled NaOH solution (0 to 1 °C) was injected to completely fill the mold. This multi-step casting ensures that the liquid metal remains entirely protected by a solution layer, preventing direct metal–membrane adhesion during subsequent coating steps, which could otherwise compromise the structural integrity and properties of the final cell.

Step 3: hydrogel coating. The mold was frozen for another 6 hours to form the complete solution–liquid metal double-layer ice balls. These ice balls were then carefully removed from the mold and immediately immersed into the SA-PEDOT:PSS composite gel solution prepared in section 4.2 and maintained at 0 °C.

Step 4: crosslinking and membrane formation. The ice balls were subjected to multi-directional rolling to ensure a uniform coating of the hydrogel solution. They were then rapidly transferred into a 5 wt% calcium lactate (CaL) aqueous solution at 4 °C. The samples were kept stationary for approximately 5 minutes until the internal ice cores completely melted, allowing the calcium ions to thoroughly crosslink the alginate network. It is critical to strictly control this reaction time; prolonged exposure can lead to osmotic pressure imbalances and unwanted side reactions between the internal NaOH and external CaL, which negatively affects the permeability and optical transparency of the resulting cell membrane.

Step 5: washing and storage. The fully formed LMCs were carefully retrieved from the CaL crosslinking bath and rapidly rinsed three times with deionized water to remove any residual CaL from the membrane surface. Finally, the LMCs were transferred into a fresh 0.5 mol L⁻¹ NaOH solution for storage and subsequent experimental evaluations.

4.4. Characterization of the hydrogel layer cross-sectional morphology

The composite hydrogel membrane samples were prepared for SEM imaging using a combination of cryogenic fracturing and freeze-drying techniques to preserve their native microstructure. First, the samples were rapidly immersed in liquid nitrogen for 3 min to achieve cryogenic freezing, followed by mechanical fracturing to create a clean cross-sectional surface. Subsequently, the fractured samples were subjected to freeze-drying in a benchtop freeze dryer (Advantage 2.0, SP VirTis, America) under vacuum conditions at −50 °C for 12 hours to remove residual water while maintaining the structural integrity of the hydrogel network. The dried samples were then sputter-coated with a thin layer of gold (SPUTTER COATER 108auto, CRESSINGTON, UK) to enhance conductivity and reduce charging effects during SEM imaging. The hydrogel layer cross-sectional morphology images were captured using a field emission environmental scanning electron microscope (FESEM, QUANTA FEG 250, America) operated at an acceleration voltage of 15 kV, a working distance of 10 mm, and under high vacuum conditions.

4.5. Electrical testing of LMCs

All resistance values were obtained using the Kelvin four-terminal sensing method with a data acquisition system (DAQ970A, Keysight, America) unless otherwise noted. The potential difference between the two electrodes during the motion was supplied by a DC power supply (HLR-3660D, Henghui, China) and an arbitrary waveform function generator (DG1022, RIGOL, China).

4.6. Mechanical testing of LMCs

Mechanical compression tests of LMCs were performed on a universal stretching compressor (Model F105, MARK-10, America). The dimensions of two electrode plates constraining the longitudinal space of the LMCs were 25.4 mm × 76.2 mm × 10 mm, with an effective conductive area of 10 mm × 76.2 mm.

4.7. Temperature testing of LMCs

The mean temperature of the LMCs was determined by averaging the temperature values measured using three T-type thermocouples (KPS-T-T-30-SLE-1000-SMPW-G, Kaipusen, China) placed at the bottom, middle, and top of the LMCs. To ensure a standardized test configuration, the longitudinal compression of each LMC was precisely controlled at 7.4 mm using calibrated 3D-printed spacers. All temperature values were obtained on a data acquisition system (DAQ970A, Keysight, America).

4.8. Preparation of liquid metal ferrofluids

We preweighed 0.5 mL of EGaIn and 0.31 g of Fe particles (particle diameter = 100 nm) and mixed them in a beaker filled with 6 mL of 2 mol L⁻¹ HCl solution. Stirring the mixture gently with a glass rod for 15 minutes allowed the Fe particles to be completely internalized into the liquid metal. The desired magnetized liquid metal was then obtained by aspirating the liquid metal–Fe mixture with a syringe and rinsing it three times with deionized water.

4.9. Passive motion monitoring of magnetic LMCs

The motion of the magnetic LMCs and the magnet was monitored using a high-speed camera (Model: 28028 VEO-710L-18GB-M, Phantom, USA), which was used to capture high-frame-rate images of the LMCs’ movement.

Author contributions

Conceptualization: J. Y. L., and J. L.; methodology: J. Y. L., and J. L.; investigation: J. Y. L., R. X., J. W., and Y. S.; data curation: J. Y. L. and M. Q.; formal analysis: J. Y. L., Q. M., and X. Z.; validation: J. Y. L., Z. X., and Y. B.; visualization: J. Y. L., M. G., X. Z., and J. L.; writing—original draft: J. Y. L.; writing—review and editing: X. Z., and J. L.; supervision and funding acquisition: J. L. and X. Z.; all authors read and approved this manuscript.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: ten figures, six notes, and five movies to support the discussion and conclusion of the main text. See DOI: https://doi.org/10.1039/d6nr00291a.

Acknowledgements

This work was partially supported by the National Natural Science Foundation of China under Grant No. 91748206 and 12402324.

References

1. X. Wang, C. Lu and W. Rao, Appl. Therm. Eng., 2021, 192, 116937.
2. G. Li, M. Zhang, S. Liu, M. Yuan, J. Wu, M. Yu, L. Teng, Z. Xu, J. Guo, G. Li, Z. Liu and X. Ma, Nat. Electron., 2023, 6, 154–163.
3. Y. Yang, N. Sun, Z. Wen, P. Cheng, H. Zheng, H. Shao, Y. Xia, C. Chen, H. Lan, X. Xie, C. Zhou, J. Zhong, X. Sun and S.-T. Lee, ACS Nano, 2018, 12, 2027–2034.
4. B. Liu, P. Qin, M. Liu, W. Liu, P. Zhang, Z. Ye, Z. Deng, Z. Li and L. Gui, Micromachines, 2023, 14, 717.
5. L. Zhang, X. Huang, T. Cole, H. Lu, J. Hang, W. Li, S.-Y. Tang, C. Boyer, T. P. Davis and R. Qiao, Nat. Commun., 2023, 14, 7815.
6. T. Gan, W. Shang, S. Handschuh-Wang, Y. Zhang and X. Zhou, Chem. Mater., 2024, 36, 3042–3053.
7. X. Wu, X. Peng, L. Ren, J. Guan and M. Pumera, Adv. Funct. Mater., 2024, 34, 2410167.
8. B. Wang, N. Zhang, W. Feng, S. Chen, X. Zhu, X. Shan, R. Yuan, B. Yuan, H. Wang, G. Zhou, J. Liu and X. Sun, Adv. Funct. Mater., 2024, 34, 2407934.
9. Y. Ding and J. Liu, Appl. Phys. Lett., 2016, 109, 121904.
10. Q. Zhuang, K. Yao, M. Wu, Z. Lei, F. Chen, J. Li, Q. Mei, Y. Zhou, Q. Huang, X. Zhao, Y. Li, X. Yu and Z. Zheng, Sci. Adv., 2023, 9, eadg8602.
11. J. Zhang, Y. Yao, L. Sheng and J. Liu, Adv. Mater., 2015, 27, 2648–2655.
12. D. Yu, Z. Wang, G. Chi, Q. Zhang, J. Fu, M. Li, C. Liu, Q. Zhou, Z. Li, D. Chen, Z. Song and Z. He, Nat. Commun., 2024, 15, 1179.
13. Y. Lu, Y. Lin, Z. Chen, Q. Hu, Y. Liu, S. Yu, W. Gao, M. D. Dickey and Z. Gu, Nano Lett., 2017, 17, 2138–2145.
14. J. Tang, X. Zhao, J. Li, Y. Zhou and J. Liu, Adv. Sci., 2017, 4, 1700024.
15. Y. Yang and Y. Shen, Nat. Commun., 2024, 15, 3398.
16. L. Hu, B. Yuan and J. Liu, Sci. Rep., 2017, 7, 7256.
17. K. M. Yamada and M. Sixt, Nat. Rev. Mol. Cell Biol., 2019, 20, 738–752.
18. Y. Ding, F. Wu and C. Tan, Life, 2014, 4, 1092–1116.
19. J. Li, X. Zhao and J. Liu, Matter, 2024, 7, 2033–2065.
20. J. Sun, H. He, K. Zhao, W. Cheng, Y. Li, P. Zhang, S. Wan, Y. Liu, M. Wang, M. Li, Z. Wei, B. Li, Y. Zhang, C. Li, Y. Sun, J. Shen, J. Li, F. Wang, C. Ma, Y. Tian, J. Su, D. Chen, C. Fan, H. Zhang and K. Liu, Nat. Commun., 2023, 14, 5348.
21. Y. Li, J. Wang, H. Li, L. Huang, L. Chen, Y. Ni and Q. Zheng, Appl. Surf. Sci., 2022, 572, 151502.
22. H. Yuk, S. Lin, C. Ma, M. Takaffoli, N. X. Fang and X. Zhao, Nat. Commun., 2017, 8, 14230.
23. W. Chen, L. Zhai, S. Zhang, Z. Zhao, Y. Hu, Y. Xiang, H. Liu, Z. Xu, L. Jiang and L. Wen, Science, 2023, 382, 559–565.
24. T. S. Hebner, K. Korner, C. N. Bowman, K. Bhattacharya and T. J. White, Sci. Adv., 2023, 9, eade1320.
25. S. Terryn, J. Brancart, D. Lefeber, G. Van Assche and B. Vanderborght, Sci. Robot, 2017, 2, eaan4268.
26. J. Liu, Z. Guo and K. Liang, Adv. Funct. Mater., 2019, 29, 1905321.
27. Z. Xu, T. Hueckel, W. T. M. Irvine and S. Sacanna, Nature, 2021, 597, 220–224.
28. Y. Ma, H. Yang, X. Zong, J. Wu, X. Ji, W. Liu, P. Yuan, X. Chen, C. Yang, X. Li, Y. Chen, W. Xue and J. Dai, Biomaterials, 2021, 274, 120865.
29. M. Wei, Y. Lin and Y. Qiao, Giant, 2023, 13, 100143.
30. A. B. Cook, S. Novosedlik and J. C. M. van Hest, Acc. Mater. Res., 2023, 4, 287–298.
31. W. Jiang, Z. Wu, Z. Gao, M. Wan, M. Zhou, C. Mao and J. Shen, ACS Nano, 2022, 16, 15705–15733.
32. P. Treenate and P. Monvisade, Int. J. Biol. Macromol., 2017, 99, 71–78.
33. A. Karim, A. Rehman, J. Feng, A. Noreen, E. Assadpour, M. S. Kharazmi, Z. Lianfu and S. M. Jafari, Adv. Colloid Interface Sci., 2022, 307, 102744.
34. J. Ma, S. Li, X. Li, K. Guo, H. Wang and J. Yin, Chem. Eng. J., 2024, 492, 152433.
35. B. Lu, H. Yuk, S. Lin, N. Jian, K. Qu, J. Xu and X. Zhao, Nat. Commun., 2019, 10, 1043.
36. E. Dauzon, Y. Lin, H. Faber, E. Yengel, X. Sallenave, C. Plesse, F. Goubard, A. Amassian and T. D. Anthopoulos, Adv. Funct. Mater., 2020, 30, 2001251.
37. K. Zhao, Y. Zhao, J. Xu, R. Qian, Z. Yu and C. Ye, Chem. Eng. J., 2024, 494, 152971.
38. D. Ji, J. M. Park, M. S. Oh, T. L. Nguyen, H. Shin, J. S. Kim, D. Kim, H. S. Park and J. Kim, Nat. Commun., 2022, 13, 3019.
39. P. Matricardi, C. Di Meo, T. Coviello, W. E. Hennink and F. Alhaique, Adv. Drug Delivery Rev., 2013, 65, 1172–1187.
40. G. T. Grant, E. R. Morris, D. A. Rees, P. J. C. Smith and D. Thom, FEBS Lett., 1973, 32, 195–198.
41. L. Ferlauto, A. N. D'Angelo, P. Vagni, M. J. I. Airaghi Leccardi, F. M. Mor, E. A. Cuttaz, M. O. Heuschkel, L. Stoppini and D. Ghezzi, Front. Neurosci., 2018, 12, 648.
42. D. Wang, J. Ye, Y. Bai, F. Yang, J. Zhang, W. Rao and J. Liu, Adv. Mater., 2023, 35, 2303533.
43. L. Wang and J. Liu, Front. Energy, 2013, 7, 317–332.
44. F. C. Simeone, H. J. Yoon, M. M. Thuo, J. R. Barber, B. Smith and G. M. Whitesides, J. Am. Chem. Soc., 2013, 135, 18131–18144.
45. I. D. Zaytsev and G. G. Aseyev, Properties of aqueous solutions of electrolytes, CRC press, 1992.
46. S. Liu, S. N. Reed, M. J. Higgins, M. S. Titus and R. Kramer-Bottiglio, Nanoscale, 2019, 11, 17615–17629.
47. S. Handschuh-Wang, Y. Chen, L. Zhu, T. Gan and X. Zhou, Langmuir, 2019, 35, 372–381.
48. W. M. Haynes, CRC handbook of chemistry and physics, CRC press, 2016.
49. M. Guzinski, J. M. Jarvis, F. Perez, B. D. Pendley, E. Lindner, R. De Marco, G. A. Crespo, R. G. Acres, R. Walker and J. Bishop, Anal. Chem., 2017, 89, 3508–3516.
50. Y. Plevachuk, V. Sklyarchuk, S. Eckert, G. Gerbeth and R. Novakovic, J. Chem. Eng. Data, 2014, 59, 757–763.
51. J. A. Sorensen and G. E. Glass, Anal. Chem., 1987, 59, 1594–1597.
52. S. Park and H. J. Yoon, Nano Lett., 2018, 18, 7715–7718.
53. J. Jang, P. He and H. J. Yoon, Acc. Chem. Res., 2023, 56, 1613–1622.
54. T. Iida and R. I. L. Guthrie, The thermophysical properties of metallic liquids: Volume 1—fundamentals, Oxford University Press, 2015.
55. Y. Yu and E. Miyako, iScience, 2018, 3, 134–148.
56. L. Hu, H. Wang, X. Wang, X. Liu, J. Guo and J. Liu, ACS Appl. Mater. Interfaces, 2019, 11, 8685–8692.
57. M. F. Wang, M. J. Jin, X. J. Jin and S. G. Zuo, Phys. Chem. Chem. Phys., 2017, 19, 18505–18513.
58. X. Yang, S. Tan, B. Yuan and J. Liu, Sci. China: Technol. Sci., 2016, 59, 597–603.
59. K. J. Lohmann, K. M. Goforth, A. G. Mackiewicz, D. S. Lim and C. M. F. Lohmann, J. Comp. Physiol., A, 2022, 208, 41–67.
60. E. F. DeLong, R. B. Frankel and D. A. Bazylinski, Science, 1993, 259, 803–806.
61. S. A. Chechetka, Y. Yu, X. Zhen, M. Pramanik, K. Pu and E. Miyako, Nat. Commun., 2017, 8, 15432.
62. N. Yang, F. Gong, Y. Zhou, Y. Hao, Z. Dong, H. Lei, L. Zhong, X. Yang, X. Wang and Y. Zhao, Biomaterials, 2021, 277, 121125.
63. J. Jia, R.-K. Liu, Q. Sun and J.-X. Wang, Langmuir, 2023, 39, 17808–17817.
64. D. Xu, J. Hu, X. Pan, S. Sánchez, X. Yan and X. Ma, ACS Nano, 2021, 15, 11543–11554.
65. J. Li and D. J. Mooney, Nat. Rev. Mater., 2016, 1, 16071.
66. H. J. Kong, D. Kaigler, K. Kim and D. J. Mooney, Biomacromolecules, 2004, 5, 1720–1727.
67. K. T. Campbell, R. S. Stilhano and E. A. Silva, Biomaterials, 2018, 179, 109–121.
68. X. Zhang, Y. Li, Z. Ma, D. He and H. Li, Bioact. Mater., 2021, 6, 3692–3704.

---