Porous Conductive Hybrid Composite with Superior Pressure Sensitivity and Dynamic Range

Porous conductive composites hold immense promise in flexible sensors and soft robotics due to their pressure‐responsive electrical conductivity. Unlike non‐porous composites whose pressure sensitivity is limited by relatively high elastic modulus, porous materials show improved pressure sensitivity owing to their lower stiffness. Despite this, existing porous composites still suffer from insufficient pressure sensitivity or narrow detection ranges, severely restricting their applications. This work presents a liquid metal hybrid filler porous composite to address these issues. Through experiment and simulation optimization, the composite exhibits a conductivity increase of five order‐of‐magnitude over 0–250 kPa, demonstrating a 900% higher pressure sensitivity than the best non‐porous counterpart in this work. The composite maintains a highly linear response (R2 of 0.999) over an exceptionally wide dynamic range up to 8.9 MPa, with a pressure sensitivity of 8.1 MPa–1, surpassing the state‐of‐the‐art in both pressure range and sensitivity. A proof‐of‐concept pressure sensor array further demonstrates the composite's excellent sensing performance, showing stable response under 100‐cycle loading with a measured pressure deviation of only 1.4%, outperforming existing commercial pressure sensors in terms of sensitivity, detection range and cyclic stability. The porous material design strategy opens doors for high sensitivity pressure sensors in wearable devices, flexible electronics, and soft robotics.


Introduction
In recent years, flexible materials with the ability to detect pressure stimuli have received significant attention as they find DOI: 10.1002/adfm.202309347[3] Many conductive elastic composites consisting of conductive fillers dispersed in polymer matrices are potential candidates due to their piezoconductive effect, i.e., their electrical conductivity changes with strain. [4]Numerous composites have been reported to show high sensitivity and stability, with some exhibiting up to six orders of magnitude change in conductivity upon deformation, rendering them highly desirable for pressure sensing endeavors. [5]8] Metal fillers are generally highly conductive, while carbon fillers can enhance the mechanical strength and oxidization resistance of composites. [9,10]Despite these advantages, solid fillers tend to impair the flexibility and stretchability of composites.To address this, intrinsically stretchable fillers, such as liquid metal and conductive polymers can be incorporated to improve flexibility. [11,12]19] Despite the high strain sensitivity of many reported composites, their suitability as pressure sensors remains limited due to their high elastic modulus.Under pressure, the change in electrical conductivity primarily arises from the spatial deformation of conductive particle networks during composite deformation.Therefore, a low elastic modulus increases the deformation and conductivity change, yielding a high pressure sensitivity.Moreover, reducing the modulus improves the material's flexibility, coveted in soft electronics such as wearable skin sensors. [20][23][24] Acting as particles with zero-stiffness, these pores lower the overall modulus of the composite, thereby enhancing flexibility and pressure sensitivity. [25]For instance, it has been shown that by creating micropores in a piezoconductive composite, a fivefold improvement in pressure sensitivity over the range of 0-100 kPa can be achieved. [26][29] Aiming for high initial sensitivity, prior studies generally reduce the modulus of porous composites by maximizing porosity, thereby achieving substantial deformation and conductivity change under low pressure.However, as pressure increases, the porous structure undergoes rapid compaction, leading to densification of composites and a surge in the elastic modulus. [28,30]This limits further compression of composites, causing their conductivity to stabilize. [31]As a result, these porous composites typically demonstrate limited working ranges below 500 kPa and severely decreased sensitivity at high pressures, with resistance changes below an order of magnitude over high pressure ranges.This significantly restricts the use of existing porous materials in pressure sensing applications demanding high sensitivity and wide dynamic range.
Here, we report a novel porous conductive composite consisting of spiked nickel (Ni) microparticles and eutectic galliumindium (EGaIn) liquid metal (LM) droplets as conductive fillers in a polydimethylsiloxane (PDMS) matrix, where pores are generated using 1,2-propanediol as a porogen.By optimizing the composition and processing method, the composite achieves a trade-off between high pressure sensitivity and wide measure-ment range, showing significantly increased electrical conductivity under wide pressure ranges.Compared with the non-porous Ni-LM-PDMS composite, we previously created, this optimized composite exhibits a very high pressure sensitivity as it demonstrates a conductivity growth of seven orders of magnitude under 1 MPa pressure, as well as a good linear response in a large pressure range of up to 8.9 MPa. [32]Through numerical simulations of the composite's behavior under compression, we elucidate the principles driving its exceptional sensitivity.Finally, we develop a proof-of-concept pressure sensor array based on this porous composite.The sensing array maintains stable and excellent sensing performance after prolonged cyclic loading, highlighting its promising potential in pressure sensing applications.

Preparation and Optimization of the Porous Composite
To fabricate the porous composite, we sequentially mix liquid PDMS, EGaIn liquid metal, and Ni microparticles, followed by the addition of the porogen (1,2-propanediol) (Figure 1a).This sequential mixing ensures dispersion of conductive fillers within the PDMS matrix only, so as not to fill the pores (see Figure S1, Supporting Information).The resulting mixture is cured to form a composite with interconnected porogen-filled pores.Subsequently, a brief post-cure heating step at high temperature removes the porogen, yielding the porous conductive composite illustrated in Figure 1b (see Experimental Section for details).The scanning electron microscope (SEM) image (Figure 1c) reveals LM microdroplets (4-30 μm diameter), irregular spiked Ni particles (2-5 μm diameter), and a network of interconnected pores in the porous composite.The corresponding energy-dispersive X-ray spectroscopy (EDS) element mapping (Figure 1d) identifies the distribution of EGaIn droplets, Ni particles and PDMS within the composite using their characteristic elements gallium (Ga), Ni, and silicon (Si), respectively.Note that the dark areas in the image are caused by the unevenness of the sample surface; the The X-ray photons emitted from the valleys are blocked by the composite and cannot be captured by the detector.From the element mapping, the continuous distribution of PDMS with Ni and EGaIn fillers dispersed therein is evident.In addition, as highlighted in Figure S2 (Supporting Information) and the red framed area in Figure 1c, the interconnected porous structures in composites facilitate the rapid removal of the porogen during post-cure heating.
In this work, a shorthand is used to indicate the composition of composites.The porogen (glycerol or 1,2-propanediol), nickel particle and EGaIn liquid metal are indicated by "Gly"/"Pro", "Ni" and "LM", respectively.Each of these components are followed by a subscript indicating the mass ratio of the component to the PDMS matrix.No subscript refers to a default mass ratio of 1:1.For example, a composition of Pro 0.8 Ni 6 LM refers to a porous composite with a 1,2-propanediol:nickel:EGaIn:PDMS mass ratio of 0.8:6:1:1.
Multiple parameters influence the mechanical and electrical properties of the porous composite, including pore size, porosity, conductive filler content, and processing method.To optimize the composite for maximum pressure sensitivity, we thoroughly investigate these effects through characterization of the above properties.We define the pressure sensitivity S to quantify the rate of increase in conductivity of the composite under pressure: where  and p are the electrical conductivity and pressure, respectively.In other words, S is the ratio of the slope of the conductivity-pressure curve to the conductivity at a given pressure.
The pore size is mainly governed by the size of porogen droplets, which form during mixing of the composite prior to curing, and typically range from 1 to 10 μm.The shear force applied during mixing, controlled by the mixing speed, influences this droplet size.SEM images in Figure 2a-c depict composites fabricated with increasing mixing speeds: mixing with an electric drill at 120, 350, and 1250 rpm, respectively.As indicated by the average pore size in the figure, the pore size decreases with increasing mixing speed.Figure 2d shows that reducing pore size leads to a decrease in the composite's elastic modulus, which is desirable for a high pressure sensitivity.Figure 2e,f illustrate the initial conductivity and initial pressure sensitivity of the composites with varying pore sizes.At constant porosity, reducing pore size considerably increases the number of pores, resulting in a decreased initial conductivity as more conductive pathways are blocked.On the other hand, this phenomenon enhances initial sensitivity as new conductive pathways form more readily, increasing the capacity for conductivity to rise under compression.For example, as the average pore size decreases from 8.3 to 3.7 μm, the number of pores increases by a factor of ten, resulting in a substantial decrease in conductivity by four orders of magnitude to 4 μSm -1 and an increase in sensitivity by 320% to 0.171 kPa -1 .However, there exists a trade-off between pressure sensitivity and reliability during cyclic loading.Although the medium-pore composite exhibits 10.3% lower pressure sensitivity than the smallpore composite, it has an order of magnitude higher initial conductivity.This significantly improves its reliability under cyclic loading while maintaining a sufficiently high pressure sensitivity (see detailed discussions in the next section).Furthermore, the medium-pore composite demonstrates more stable mechanical and electrical properties (indicated by its smaller error bars in Figure 2d-f).Therefore, we fabricate the porous composites with a moderate pore size of 3-4 μm for optimal pressure sensing performance.
In addition, the porosity of the composite significantly impacts its mechanical and electrical properties.Due to shrinkage of the composite during post-cure heating, the final porosity is less than the volume fraction of porogen used.Therefore in this study, we experimentally determine the porosity of composites by comparing their densities with non-porous variants.The porosity of the composite with a porogen/PDMS mass ratio of 0.4/0.6/0.8/1 is 13.7%/18.8%/22.6%/25.2%,respectively (see Figure S3, Supporting Information for details).
As porosity increases with porogen/PDMS mass ratio, the composite's elastic modulus experiences a sharp decline, as illustrated in Figure 2g. Figure 2h,i shows the initial conductivity and initial pressure sensitivity of composites with varying porosity.The initial conductivity decreases with increasing porosity from 2.14 Sm -1 (porosity of 13.7%) to 0.0176 μSm -1 (porosity of 25.2%), whilst the initial sensitivity peaks at 0.171 kPa -1 with a porogen/PDMS mass ratio of 0.8 (porosity of 22.6%).This decrease in conductivity arises from a reduced volume fraction of conductive particles when porosity is increased, resulting in a sparse particle network and thus poor conductivity.As discussed previously, composites with low elastic modulus tend to exhibit high pressure sensitivity.Furthermore, composites with a low porosity (Pro 0.4 /Pro 0.6 ) exhibit reduced interconnected porosity, leading to incomplete removal of porogen during post-cure heating (see Figure S4, Supporting Information).Conversely, composites with a high porosity (e.g., Pro 1 ) form excessive interconnected cavities.This reduces the amount of conductive pathways and hinder their generation under compression, resulting in a lower initial conductivity and pressure sensitivity.Composites with porogen mass ratios of 1.2 and 1.4 (porosity of 27.4% and 29.2%, respectively) exhibit insulating initial electrical behavior due to their high porosity, and only demonstrate piezoconductivity once strained above 6% and 13.6%, respectively, and hence are omitted from Figure 2h,i.Therefore, we fabri-cate the porous composites with a porogen/PDMS mass ratio of 0.8, which demonstrates an appropriate initial conductivity (30 μSm -1 ) and the highest pressure sensitivity (0.17 kPa -1 ).
In addition to the pore characteristics, we also optimize additional parameters to maximize the pressure sensitivity, including the porogen type and conductive filler content, as well as the post-cure heating temperature and time (see detailed results and discussions in Figure S5, Supporting Information).In brief, the composites based on a 1,2-propanediol porogen exhibit a more stable electrical response than those using glycerol as the porogen.During the optimization of conductive filler content, we find that a Ni:LM:PDMS mass ratio of 6:1:1 yields the optimal pressure sensing performance.Additionally, a post-cure heating process at 140 °C for 1.5 h resulted in the highest pressure sensitivity while ensuring complete evaporation of the porogen (see Figure S5, Supporting Information for details).

Mechanical and Electrical Properties of the Porous Composite
After optimization, the highest pressure sensitivity is achieved by the porous composite with a composition of 1,2-propanediol:Ni:EGaIn:PDMS of 0.8:6:1:1 (Pro 0.8 Ni 6 LM). Figure 3a shows its conductivity-pressure curve, along with comparison to a non-porous composite with the same conductive filler content (Ni 6 LM), and an optimized non-porous composite (Ni 3 LM) demonstrating the highest pressure sensitivity among non-porous composites, which we developed in previous work (see Figure S6, Supporting Information for details). [32]While the composites all exhibit an initial conductivity of approximately 10 -5 Sm -1 , the porous composite demonstrates a superior pressure sensitivity of 0.313 kPa -1 from 0 to 50 kPa, showcasing a significant improvement of over 900% and 1800% compared to the non-porous Ni 3 LM and Ni 6 LM composites, respectively.This enhanced sensitivity is attributed to the low elastic modulus which allows for greater deformation under pressure, as well as the high content of Ni particles that facilitate the generation and enhancement of conductive pathways.These Ni particles have surface spikes which further improve the pressure sensitivity as they give rise to conduction by quantum tunneling.When the composite deforms, the proximity of these spikes exponentially increases the probability of quantum tunneling between them.This greatly reduces the contact resistance of adjacent Ni particles, as well as the overall resistance of the composite, resulting in high sensitivity. [33]This effect is magnified by the large local electric field at the Ni spike tips. [34,35]Along with these phenomena, the closure of pores during compression generates additional conductive pathways, further enhancing the sensitivity.However, the effect of this is modest as the conductive fillers and pores are of comparable size.This is in contrast to other conductive composites, with small fillers (e.g., carbon nanotubes) and large pores, where the pores greatly obstruct the conductive pathways, resulting in a considerable and obvious increase in conductivity as pores are closed. [36,37]ver a wide dynamic range of 250 kPa (representing a strain of 3.3%), the porous composite demonstrates a substantial conductivity change of five orders of magnitude, showing a highly sensitive pressure response.By incorporating a microporous  [25,26,[28][29][30][38][39][40][41][42][43] d) Variation in conductivity of the porous composite over 1000 cycles loaded from 40-100 kPa, with insets showing ten cycles enlarged around cycles 0, 500, and 1000.
structure into the conductive composite, the elastic modulus of the porous composite is significantly reduced by 90% compared to the non-porous variant (see Figure S7, Supporting Information, for detailed comparison), thereby improving pressure sensitivity through greater mechanical deformation under pressure.
As pressure further increases, the conductivity of the composite increases by seven million times to 53 Sm -1 at 1 MPa, and shows a highly linear response (R 2 of 0.999) with a sensitivity of 8.1 MPa -1 over an extremely wide pressure range up to 8.9 MPa (see Figure 3b; Figure S8, Supporting Information).In comparison, non-porous composites exhibit lower sensitivity and poor linearity.38][39][40][41][42][43] Existing materials tend to have either a high pressure sensitivity or wide working range, whilst the material presented here demonstrates both, revealing new applications for pressure sensors with both high sensitivity and dynamic range (see Table S1, Supporting Information for detailed comparison of key characteristics and compositions of representative porous composites).Even compared with state-of-the-art reported porous materials, our composite still exhibits a two order of magnitude higher conductivity change over 0-1 MPa, as well as a dynamic range twice as wide (see Figure S8, Supporting Information for detailed comparison). [38]orous composites are susceptible to microstructural damage and significant elastic hysteresis under excessive mechanical deformations (see details in Figure S9, Supporting Infor-mation).These are caused by local stress concentrations due to internal friction between dense metal fillers during excessive deformation. [44]This phenomenon causes plastic and viscoelastic deformation of the material, as well as a residual strain determined by the maximum deformation (see Text S1, Supporting Information for details).While excessive deformation weakens the pressure sensing capabilities of the material at low strains, it does not impair its good linear response at high pressures (see Figure S9b, Supporting Information).During compression at lower strains, viscoelastic deformation has negligible impact on the pressure response of the conductivity.This deformation occurs predominately during the first loading cycle, with subsequent cycles exhibiting consistent mechanical and electrical responses (see Figure S10, Supporting Information).Therefore, we suggest that excessive deformation of porous composites during operation should be avoided to maintain a satisfactory low pressure response.
To assess the reliability of the porous composite, we perform long-term cyclic tests involving 1000 cycles of compression from 40 to 100 kPa at a rate of 12 kPa.s - .Figure 3d shows the change in conductivity over these 1000 cycles, with insets 1, 2, and 3 highlighting the behavior during the first, middle and last ten cycles, respectively.Initially, the conductivity decreases by two orders of magnitude over the first 100 cycles, which subsequently reaches a stabilized state.The enlarged insets indicate a 60% decrease in conductivity during the first ten cycles, followed by relative stability after 500 cycles.Despite the reduction in conductivity, the pressure sensitivity remains unaffected as the rate of conductivity change in each cycle remains constant.Notably, the composite exhibits high repeatability during the last 500 cycles, making it suitable for use as a pressure sensor.As discussed in last section, the low initial conductivity allows for significant conductivity increase under pressure, thereby maintaining a high pressure sensitivity across a broad dynamic range.However, Figure 3d suggests that an excessively low initial conductivity may lead to electrically insulating behavior during long-term cyclic loading, compromising the practical suitability and robustness of the composites (see Figure S11, Supporting Information).Therefore, it is crucial to strike a trade-off between high pressure sensitivity and reliability during cyclic loading, according to specific application requirements.

Pressure Response Simulation of the Porous Composite
The positive piezoconductive effect exhibited by the porous composite is due to the creation and enhancement of conductive path-ways during compression.As illustrated in Figure 4a, in the relaxed state, the LM droplets and Ni microparticles are separated by the PDMS layer, resulting in poor initial conductivity.Under compressive deformation, the conductive particles approach each other along the direction of compression to form conductive pathways.Large LM droplets serve as conductive network hubs to connect surrounding particles, while the spikes on the surface of Ni particles facilitate their contact.As a result, the electrical conductivity of the composite increases dramatically by orders of magnitude under compression.
To validate our hypothesis and gain further insights, we perform finite element simulations to compare the behavior of the porous conductive composite to non-porous composites under applied pressure.Figure 4b,c shows the simulated deformation and stress distribution of 2D models of porous and non-porous composites under the same pressure (stress magnitude indicated by color).The stress concentrates primarily in rigid Ni particles (red), whilst the liquid EGaIn droplets experience minimal stress (blue).Due to the deformation of the pores (as evidenced in Figure 4b), the porous composite deforms nine times that of the non-porous variant at the same pressure (≈300 kPa), illustrating a much lower elastic modulus.The simulated modulus of the porous composite is approximately 10% of the non-porous composite (Figure 4d), which agrees with experimental results with an error of 1.5%.
To investigate the electrical response, we simulate the potential drop through the composites by applying a fixed current density to a 2D model.This allows us to calculate the conductivity at different pressures, indicating the sensitivity (see Figure S12, Supporting Information for detailed simulation settings).Figure 4e,f shows this potential drop at an applied pressure of 50 kPa for the porous and non-porous composites, respectively (see Figure S13, Supporting Information for the simulation results in the relaxed state).The stress distribution simulations in Figure 4b,c demonstrates greater stress transfer to the Ni particles in the porous composite compared to the non-porous variant.This suggests a closer proximity of the Ni particles, enhancing the conductive pathways and increasing the conductivity at a given pressure.This hypothesis is supported by simulations of the current density distribution, where regions of higher stress exhibit elevated current densities, particularly in porous composites (see Figure S14, Supporting Information).Both experimental and simulated conductivity-pressure characteristics (Figure 3a respectively) validate this prediction, showing higher conductivity in the porous composite under pressure.On account of the lower elastic modulus and greater stress transfer to conductive fillers in the porous composite, the simulated pressure sensitivity can be increased several-fold by incorporating a porous microstructure (Figure 4g), aligning with experimental findings.

Pressure Sensing Applications
The porous conductive composite with superior pressure sensitivity and dynamic range show advantages in a wide variety of application scenarios, particularly in high sensitivity pressure sensors.To demonstrate its capabilities, we develop a proof-ofconcept pressure sensor array consisting of an array of individual porous composite sensor units.Figure 5a depicts one of these sensor units, showing its current-pressure characteristic under an applied voltage of 3.3 V.The photograph shows a single sensor unit with 20 nm gold electrodes on the top and bottom surfaces to improve the electrical contact (see Experimental Section for details).These sensor units have a thickness of 2 mm.The minimum thickness of the composite that can be used for sensors is 0.3 mm, below which, the composite loses its stable sensor performance (see Figure S15, Supporting Information for further discussion).The sensor unit demonstrates a dramatic five-orderof-magnitude current change across a wide dynamic range of 0-250 kPa, resulting in a high current-pressure sensitivity (S I ) of 0.306 kPa -1 at 50 kPa.][43] In contrast, the state-of-the-art commercial FSR (force sensitive resistor) sensor, FlexiForce™sensor (A201, Tekscan Inc., US), offers adjustable sensitivity and working range, but these parameters are interdependent.At the same sensitivity of 0.3 kPa -1 as the sensor developed in this work, the FlexiForce sensor shows a narrow detection range of less than 700 kPa.Conversely, when adjusted to its maximum pressure range of 6.2 MPa, its initial sensitivity drops to only 0.04 kPa -1 , significantly lower than that of our porous composite.In addition, our sensor unit exhibits no delay during cyclic compression at a rate of 46 kPa.s -1 (0.5%.s -1 ), outperforming existing commercial FSR sensors (see Figure S16, Supporting Information for details). [45]igure 5b illustrates the 3×3 array of individual pressure sensor units embedded in an Ecoflex 00-30 substrate.Ecoflex is chosen as it is 20 times softer than PDMS which improves stress transfer to the sensor units, thus enhancing the sensitivity. [46]In addition, the Ecoflex layer provides protection from direct skin contact due to its biocompatibility with the human body. [47][50] As discussed in Section 2.2, the composite experiences plastic deformation at large strains.To prevent accidental damage to the sensing unit caused by excessive loads, a 3D printed over-strain guard is implemented, limiting the applied strain to 10% (Figure 5c).Conductive fabric tape is utilized for electrical access to the sensor units through row-column indexing.We use a Raspberry Pi in combination with a potential divider circuit to measure the resistance of each sensor unit (see Figure S17, Supporting Information).The pressure sensor is calibrated based on the initial resistance for each sensor unit.In reference to the loading curve in Figure 5a, we employ a simple log-linear/linear approximation for calibration.This concise approach endows the proof-of-concept sensor array with high robustness (see cyclic loading test below).The main limitation of this pressure sensor comes from its simple potential divider circuit, as it can only accurately measure resistances within ±2orders of magnitude around the fixed resistor while maintaining stability (see Figure S18, Supporting Information).Notably, even this measurement accuracy has surpassed existing commercial sensors. [45]Alternative circuit designs, such as a resistance-to-frequency converter, can offer higher measurement accuracy at the cost of increased complexity. [51]In summary, our proof-of-concept application adequately demonstrates the sensing performance of the porous composite with a balance of accuracy and robustness.In contrast to piezoelectric materials which are commonly used as sensors in wearable electronics, this system exhibits a much wider detection range, whilst maintaining a high pressure sensitivity. [52]Additionally, compared to piezoelectric pressure sensors that typically involve complex and costly processing methods such as electrospinning, printing, and patterning techniques, our porous composites enable a simple and cost-effective manufacturing process. [53,54]Furthermore, our sensor array is able to measure static pressure, unlike most piezoelectric sensors whose output gradually falls to zero under constant pressure. [55]The above advantages further expand its utility to other fields such as industrial manipulators and flexible robots. [56]nder pressure, the Raspberry Pi outputs the real-time pressure distribution on the sensor array.As an example, the sensor array is loaded with an L-shaped rod (Figure 5d), and the corresponding real-time pressure distribution is shown in Figure 5e.Movie S1 (Supporting Information) demonstrates the calibration of the pressure sensing array and its real-time pressure distribution detection under static and dynamic loading.Despite the simplicity of our circuit design and calibration method, the sensor array accurately identifies the location and magnitude of applied pressure with a resolution of <0.1 kPa.This flexible sensor array is capable of operating on curved surfaces.As the curvature increases, the initial resistance and pressure sensitivity of the sensing units decrease due to bending strains.However, the sensor maintains a relatively stable initial resistance and reasonable sensitivity when the radius of curvature is greater than 1.6 cm (see Figure S19, Supporting Information for details).Therefore, the pressure sensor array is suitable for use in many flexible applications, such as wearable electronics where the sensor array can operate effectively on the wrist or other curved surfaces with lower curvature.
Figure 5f illustrates the loading method to evaluate the reliability of the sensor array under cyclic compression.A pressure of 50 kPa is applied to the region indicated in Figure 5g for 100 cycles at a rate of 10 kPa.s -1 .As illustrated in Figure 5g, the average pressures measured under the same load before and after cyclic compression are 18.75 kPa and 19.02 kPa, respectively, indicating a deviation of only 1.4%.As a control, half of the loaded sensor units do not undergo cyclic compression.In comparison, stateof-the-art commercial pressure sensors show significant drift or gradual loss of sensitivity during cyclic loading. [45]Despite the increase in sensor resistance during cyclic compression, as discussed in Section 2.2, the sensor array still accurately evaluates the magnitude of applied pressure after calibration.By adjusting the fixed resistor in the measurement circuit, the sensor array can maintain a high pressure resolution, even with orders of magnitude increase in sensor resistance.

Conclusion
In this work, we report a novel porous conductive composite and demonstrate its exceptional pressure sensitivity over a wide dynamic range through experiments and numerical simulations.The interconnected microporous network within the composite reduces its elastic modulus by 90%, enabling significant deformation under pressure.By incorporating a highly sensitive hybrid filler network of liquid metal and spiked Ni particles, the porous composite achieves a seven order of magnitude increase in conductivity across a wide dynamic range of 0-1 MPa, showing a high initial sensitivity of 0.31 kPa -1 .Such a considerable conductivity change not only exceeds published state-of-the-art porous materials by two orders of magnitude, but also outperforms the pressure sensitivity of an optimized non-porous composite by 900%.Furthermore, the composite exhibits highly linear pressure sensing capabilities over an extremely wide dynamic range up to 8.9 MPa with a sensitivity of 8.1 MPa -1 .Excellent repeatability and stability is maintained during loading tests over 1000 cycles, confirming its reliability in pressure sensing applications.To demonstrate its superior pressure sensing performance, we develop a proof-of-concept pressure sensor array, consisting of a 3×3 configuration of porous composite sensor units with a resolution of <0.1 kPa.In addition to high sensitivity and wide detection range, this sensing array maintains consistent sensor performance and exhibits no hysteresis or drift during long-term cyclic loading, surpassing state-of-the-art commercial pressure sensors across the board.Such excellent sensitivity, dynamic range, and cyclic stability of this porous composite provide unparalleled advantages for its application as high sensitivity pressure sensors in wearable electronics and soft robotics.
Preparation of the Porous Composite: The fabrication method used to prepare the porous composite is illustrated in Figure 1a.First, the PDMS base and curing agent, at a mass ratio of 9:1, was added to a disposable plastic cup.The Ni particles and EGaIn liquid metal was then added to the plastic cup with a Ni:EGaIn:PDMS(+base) mass ratio of of 6:1:1, before mixing the contents using a flat plastic stick (cross-section of 2×5 mm) equipped on an electric stirrer for approximately 5 min until well combined.Next, 1,2-propanediol with a mass ratio to PDMS of 0.8:1 was added, and the contents mixed for several minutes until well combined.Composites with small pore sizes, illustrated in Figure 2c, were prepared using an Einhell cordless TC-CD 18-2 Li Drill Driver to mix the composite at a speed of 1250 rpm.The mixture was degassed in a vacuum chamber for 5 min to remove any trapped gas before being poured into a 3D printed plastic mould and degassed for another 10 min, after which it was placed in an oven and cured at 80 °C for 16 h.An UltiMaker S5 3D printer was used to 3D print the mixing tools and curing moulds, and models were drawn using Autodesk Inventor.Once the composite was cured, further post-cure heating at 140 °C for 1.5 h was performed to allow the porogen to be removed from the composite by evaporation, resulting in the porous conductive composite.
Characterization of the Porous Composite: To allow characterization of the electrical properties, a 20 nm gold layer was first coated on the top and bottom of the composite samples (5×5×2 mm) to form surface electrodes using an AGB7341 Agar Auto Sputter Coater, improving the contact between the testing apparatus and the composite at low pressures (see Figure S20, Supporting Information for details).Note that the gold layer should be coated after the post-cure heating process to prevent cracking caused by composite shrinkage during heating (see Figure S21 and S22, Supporting Information).The electrical resistance was measured using a Keithley 2400 Standard Series Source-Measurement Unit.Mechanical characterization and tests were performed using an Instron 68TM-50 Universal Testing Machine, where 40% nominal strain tests were performed at a compression rate of 20%.min -1 , whilst all other tests were conducted at 10%.min -1 unless otherwise stated.To characterize the variation in mechanical and electrical properties between samples, standard deviations were calculated from several porous composites created separately.
An FEI Quanta 3D FEG Dual Beam Electron Microscope was used to obtain the SEM and corresponding EDS images of the composite in Figure 1, whilst other SEM images were taken using a secondary Magellan 400 SEM.The COMSOL Multiphysics 5.2 software package (Burlington, MA, USA) was used to simulate and calculate the deformation and resistance of the composite.
Demonstration Pressure Sensor Array: To fabricate the proof-of-concept sensor array as illustrated in Figure 5b, the array matrix and protection layer was created by combining a 1:1 ratio of Ecoflex Part A and Part B and transferring it to a 3D printed plastic mould.The Ecoflex was degassed in a vacuum chamber for 5 min to remove air bubbles before being cured at room temperature for 6 h.The plastic over-strain guards were printed using the aforementioned 3D printer.The porous conductive composite sensor units had dimension 5×5×2 mm and were row-column indexed using conductive fabric tape.A Raspberry Pi 3 Model B was used to measure the resistance of each sensor unit and perform data interpretation calculations, and code was written in Python.An MCP3008 Analogue to Digital Converter was used in the sensor array circuit (see Figure S17, Supporting Information).The sensor must be calibrated to the resistancepressure response of each individual sensor unit before use.Assuming a log-linear/linear characteristic, the software measured the resistance of each sensor unit with no applied pressure, and with an applied pressure of 200 kPa, allowing the characteristic to be approximately fitted.

Figure 1 .
Figure 1.Production of the porous conductive composite.a) Preparation process and b) 3D microstructure illustration of the composite.c) SEM image of the composite.d) Energy-dispersive X-ray spectroscopy (EDS) element mapping of the composite.

Figure 2 .
Figure 2. Optimization of pore size and porosity.a-c) SEM images of porous composites with a) large, b) medium, and c) small pore sizes.d) Elastic modulus (E), e) initial conductivity ( 0 ), and f) initial pressure sensitivity of porous composites with varying pore sizes.g) Elastic modulus, h) initial conductivity, and i) initial pressure sensitivity of porous composites with varying porosities.The value of the error bars illustrated is the standard deviation of the elastic modulus, initial conductivity, or sensitivity for five different samples.

Figure 4 .
Figure 4.The principle of high sensitivity in porous composites.a) Schematic of the piezoconductive effect.Simulated deformation and stress distribution of b) porous and c) non-porous composites under compression of 301 kPa.d) Simulated and experimental elastic modulus ratio of porous to non-porous composites, the value of the error bar illustrated is the standard deviation of the experimental data for five different samples.Conductivity simulation images of e) porous and f) non-porous composites under compression of 50 kPa.g) Simulated conductivity-pressure response for porous and non-porous composites under compression.

Figure 5 .
Figure 5. Pressure sensing applications.a) Current-pressure curve of an individual sensor unit, with an inset showing its photo.b) Schematic of the pressure sensor array.c) Photograph of the fabricated array excluding conductive tape.d) Photograph and e) Raspberry Pi output of the array loaded by an L-shaped rod.f) Photograph and g) Raspberry Pi output of the array loaded by a 200 g mass, before and after 100 cycles of compression at 50 kPa.