Graphene Nanoplatelet Exoskeleton on Polyurethane Foam to Produce Flame‐Retardant, Piezoresistive, and Electromagnetic Interference Shielding Surfaces

Polyurethane foam (PUF)’s porous structure, light weight, flexibility, and low‐cost properties make it useful in various cutting‐edge technologies. However, time‐consuming, costly, and complicated surface modification methods severely hinder its commercial applications. Herein, an ultrafast, simple, and cost‐effective surface modification method based on the evaporation of a low boiling point solvent to prepare a multifunctional graphene nanoplatelet (GNP)‐decorated PUF (GNP@PUF) is proposed. Due to the passive heat barrier of GNP sheets, the resulting sponge exhibits excellent flame retardancy by reducing the critical fire retardancy metrics, that is, peak heat release rate, total heat release, and total smoke release by 72%, 50%, and 81%, respectively. In addition, GNP@PUF can function as a piezoresistive sensor and electromagnetic interference (EMI)‐shielding material. As a piezoresistive sensor, it exhibits a wide‐compressive pressure (2.4–112 kPa)/strain (5–70%) range and ultra‐fast response/relaxation time (48/35 ms), wide‐stretching strain (5–100%) range, and it can detect minute human motions by being attached to different parts of the human body. Meanwhile, the composite foam displays good absorption‐dominant EMI shielding performance (≈38 dB), possibly due to conductive dissipation and multiple reflections/scattering of EM waves inside the 3D conductive graphene network. This study provides a simple coating technique for developing multifunctional lightweight foam materials.


Introduction
Polyurethane foam (PUF) is an inexpensive, porous, lightweight, and comfortable polymeric material.[3] Recent materials' demand and technological advancements have widened the application scope of PUF in cutting-edge technologies.An appropriate modification is necessary to utilize the PUF in these advanced applications. [4]This can be accomplished primarily through two methods: i) surface coating (posttreatment) or ii) incorporation of fillers during fabrication. [5]Surface coating has been regarded as one of the most straightforward and effective strategies for modifying polymeric foams.The two primary benefits of the surface coating technique are: i) it modifies PUF with minimal impact on the bulk properties [1,6] and ii) it deposits the modifying agent on the surface, exactly where it is required, as most advanced applications for polymeric foams are surface phenomena.Surface modifications impart multiple functionalities: flame retardancy, hydrophobicity, and electrical conductivity. [6]As a result, the modified PUF has uses in various cutting-edge technologies.Electromagnetic wave interference (EMI) shielding and pressure sensing are two promising applications of PUF within advanced technologies.In recent years, 5G wireless communications in electronic devices, military equipment, and smartphones have made significant strides.While providing conveniences, these devices have also introduced severe EMI problems. [7]In addition to interfering with the normal operation of electronic devices, EMI poses a grave threat to human health and the secure transmission of information.Thus, developing a high-performance EMI-shielding material is currently one of the hottest research topics.Among the many disclosed EMI-shielding materials, 3D porous polymeric foams modified with conductive nanomaterials (i.e., carbon nanotubes [CNTs], reduced graphene oxide [rGO], graphene nanoplatelet [GNP], polypyrrole [PPy], etc.) have demonstrated a high potential for replacing the conventional rigid metal-based shielding materials for the following reasons.First, they have an ultralow density, crucial for practical applications in the telecommunication, automobile, and aerospace industries.Second, the conductive nanomaterials/foam composite's ultrahigh porosity and 3D network structure convert the EM energy into thermal energy via conductive dissipation and multiple scattering/reflections. [8] Thus, foam-based EMI shielding materials can avoid the secondary EMI pollution caused by EMI reflection, a major drawback of metal-and carbon film-based EMI shielding materials. [9,10]Third, 3D polymeric foams are flexible and can withstand mechanical deformation, including compression, stretching, bending, folding, and twisting.Foams are also promising materials for fabricating piezoresistive devices that respond to deformation/stretching caused by mechanical forces and convert it into an electrical signal, owing to their unique porous structure.Piezoresistors have enormous applications, including fluid pressure measurement, wearable electronic devices for human motion detection, devices for health care monitoring, artificial electronic skins, tactile sensors, biochemical systems, monitoring robotic devices, and concrete structures of buildings. [11]The porous design of polymeric foams provides superior mechanical properties, that is, elasticity, flexibility, and compressibility which are fundamental for piezoresistance.Further, their porous structure imparts low modulus, resulting in a fascinating sensitivity to minimal changes in pressure or strain.Meanwhile, the surface modification of PUF boosts its flame retardancy, an important property for pressure sensing and EMI shielding applications. [2][14][15][16] The most common methods used for surface modification of 3D porous polymeric foams are i) dip coating, [8,17,18] ii) layer-bylayer (LbL) assembly, [16,[19][20][21] iii) hydrothermal, [22][23][24][25][26] iv) in situ polymerization, [27][28][29] v) polymer-assisted electroless plating, [13,30] or vi) a combination of two or more of these techniques. [7,31,32]ip coating and LbL assembly are simple, do not require complex equipment, and are straightforward but time-consuming. [13]eanwhile, the hydrothermal fabrication of graphene oxide (GO)-coated foams requires two steps.First, GO is applied to the surface of PUF by dipping it in a water-based GO dispersion.The GO sheets are then reduced to rGO hydrothermally in the presence of hydrazine. [23]Consequently, these modification processes will always involve complex and time-consuming procedures, severely impeding the widespread application and development of functional foams.In contrast, the in situ interfacial polymerization of self-polymerizing conductive polymers, such as PPy, polyaniline, and poly(3,4-ethylene dioxythiophene) polystyrene sulfonate, or the muss-inspired in situ polymerization of polydopamine (PDA) or tannic acid are regarded as relatively mild processes.Besides, PDA-assisted electroplating of metallic layer (i.e., Ag, Au, Cu) on the foam surface without electrical energy is a promising coating technique. [33]However, lengthy treatment times are unavoidable for both of these procedures.In addition, most of the coating strategies mentioned above generate a substantial amount of hazardous wastewater containing residual nanoparticles and toxic chemicals/dispersing surfactants, which are ultimately dangerous to the environment.Consequently, the development of a coating strategy that requires a short treatment time and minimizes the release of nanoparticle-polluted effluent remains a formidable obstacle.
[36][37][38][39][40] i) The solvent with a low boiling point can be evaporatively recovered with minimal energy input and without environmental contamination.ii) The rapid evaporation of the solvent prevents the fillers from aggregation. [34]iii) Highly conductive carbons such as graphene, GNP, and CNT can be utilized directly instead of the energyintensive additional reduction step required when using GO as a filler.][38][39] To date, no work has attempted to modify the surface of polymeric foams by dispersing fillers in solvents with a low boiling point.
Here, we report a simple coating strategy based on a low boiling point solvent evaporation technique for creating GNP-coated PUF.Initially, a stable dispersion of GNP was created in a lowboiling solvent using ultrasonication.The GNP-coated sponge was then prepared by dipping a lab-made PUF (Figure 1A) into the GNP suspension.Next, the wet PUF was heated until the solvent evaporated completely (Figure 1B).In this manner, the production of 3D conductive foam with hollow skeletons covered by GNP sheets can be accomplished in a matter of seconds.Moreover, recovery and recycling of the solvent through distillation have been demonstrated, making the coating process more cost-effective and environmentally friendly.Benefiting from the non-flammable nature of GNP sheets, the resulting sponge exhibited high flame retardancy.Meanwhile, the combination of the conductive GNP layer and the porous and flexible nature of PUF contributed to the excellent pressure/strain sensor and absorption-dominant EMI shielding.Chloroform (CMF), a widely used solvent in organic synthesis, [41] the extraction of valuable compounds, [42] 3D printing, [43] and the preparation of polyl-lactic acid, [44][45][46] poly(methyl methacrylate), [47,48] or epoxy [49][50][51][52] based composites, was used to demonstrate the proposed coating strategy.This is because CMF possesses two crucial properties that made the process simple and efficient: i) high dispersion capacity for GNP [35] and ii) a low boiling point (66 °C) that facilitated a straightforward process.However, other relatively green solvents, such as tetrahydrofuran (THF) and ethyl acetate, can also be used as low boiling point organic solvents for the proposed rapid coating process; but, the addition of cellulose acetate in a small portion is needed to ensure the dispersion of GNP in these solvents.
The proposed surface modification strategy can produce multifunctional foam through a simple and rapid coating procedure, which could pave the way for the production of pressure-sensing devices and EMI shielding materials.

Coating Characterization
In contrast to the neat white PUF, the obtained GNP@PUF had a uniform black appearance (Figure S1B, Supporting Information).The coated foam was cut into small pieces to verify that the GNP sheets could penetrate inside the foam.In addition, the fracture surface was uniformly black, indicating that GNP sheets had penetrated deeply into the PUF.At the 1st, 2nd, 3rd, and 4th dipping-evaporation cycles, the electrical conductivity of GNP@PUF was 0.1 ± 0.002, 0.27 ± 0.014, 0.30 ± 0.016, and 0.31 ± 0.024 S m −1 , respectively.This suggests that increasing the dipping-evaporation cycle from 3rd to 4th slightly boosts conductivity, which tends to stabilize after the 3rd cycle.Considering the coating time and conductivity, PUF dipped three times in the GNP suspension was chosen for the subsequent experiments.According to the mass change before and after coating, the GNP loading in the GNP@PUF foam after three times dipping was estimated to be ≈21 wt% (Figure S1B, Supporting Information).Coating durability is a key factor in PUF surface modification. [53]o determine the GNP coating's water resistance, the GNP@PUF was squeezed in deionized water several times and left for 15 days.The GNP coating remained intact on the PUF skeleton, proving the formation of a stable, interconnected GNP network on the PUF skeleton's surface (Figure S1C, Supporting Information).To learn more about the GNP-coated foam's durability, a slice of GNP@PUF (5 × 5 × 5 mm 3 ) was submerged for 24 h in acidic (0.1 m HCl, pH = 1) and alkaline (0.1 m, pH = 14) solutions.Neither solution showed any remarkable loss of GNP, as evidenced by the low coating mass detachment (<10 wt%) (Table S1, Supporting Information) and a slight decrease in conductivity (Table S2, Supporting Information).This proves that the GNP@PUF's performance, which relies on GNP loading and conductivity, will not be drastically altered.The following may account for the GNP coating's excellent stability: The amount of hydrophilic functional groups, which interacts with the water molecule via hydrogen bonding and coulombic repulsive forces, has been shown to have a significant effect on the dispersion of carbon-based materials in aqueous medium. [54,55]Unlike GO, hydrophobic interactions are more common in GNP because of the aromatic carbon rings that make up most GNP. [55]or this reason, GNP sheets prefer to adhere to the hydrophobic PUF frameworks rather than disassociation or peeling to the aqueous solutions.Other reported coatings, such as singlewalled carbon nanotube (SWNT)@cotton fabric [56] and graphene nanopowders@PUF, [57] also show a similar phenomenon.
The foam samples were subjected to scanning electron microscope (SEM) analysis to visualize the GNP coating.The structure of the PUF network in 3D was highly porous (Figure 2Ai-iii).After GNP deposition, the porous 3D structure of GNP@PUF did not change.However, the cell surface went from smooth to rough (Figure 2Bi-vi), indicating that the GNP sheets were successfully assembled on the PUF frameworks.Energy-dispersive X-ray (EDX) spectra of pristine PUF revealed its primary constituents to be C, O, and N (Figure S2, Supporting Information).GNP@PUF, in contrast, had only peaks for C and O (Figure S3, Supporting Information).In addition, the GNP coating significantly increased the C content while decreasing the O content.This is because GNP primarily consists of C (≥95%) and O-containing functional groups (≥2.5%).The absence of the N signal and the high C content for GNP@PUF suggest that a thin layer of GNP sheets completely concealed the surface of the foam.As is evident from the SEM images of GNP@PUF, the GNP sheets were assembled uniformly and without any discernible bare areas (Figure 2Bii-vi).We obtained a SEM image of GNP particles to compare their morphology before and following assembly on PUF.Due to the aggregation and stacking of numerous GNP sheets, the GNP initially resembled large particles (Figure S4Ai-ii, Supporting Information).After deposition on the PUF; however, the GNPs were thoroughly exfoliated and assembled on the surface wall of the foam (Figure S4Bi-ii, Supporting Information).The GNP sheets uniformly covered every available surface, creating a GNP layer that ran the length of the foam and provided fire protection.In addition, the uniform alignment of GNP sheets rendered the foam conductive, which was critical to its sensing and EMI shielding performances.
The mechanical properties of pristine PUF and GNP@PUF were evaluated by performing a compression test at 60% strain.The compressive stress of GNP@PUF was higher than that of unmodified PUF at the same strain (Figure 2C).For instance, at 60% strain, the compressive stresses of neat PUF and GNP@PUF were 12 and 35 kPa, respectively.This means that the mechanical compression strength had improved after GNP deposition.When GNP@PUF was under pressure, the GNP sheets wrapped on the PUF skeleton became denser (see Figure 5A in section 2.3) and more resistant to external force, increasing the stress.This is because GNPs are very strong materials, with Young's modulus of 1 Tpa and intrinsic strength of ≈130 Gpa for a single graphene layer. [58]Previous studies have also shown that PUF compression strength can be improved through the deposition of carbon-based nanomaterials. [59]The mechanical properties of the foams were investigated further by means of a cyclic compression test.The foams can be compressed to 60% of their original thickness and fully recover to their original shape, even after thousands of cycles, without significant plastic deformation (Figure 2D).The resilient behavior of the sponge was maintained after GNP deposition, as evidenced by the fact that the hysteresis loop of GNP@PUF decreases only slightly after 1000 cycles while the maximum compressive stress remains nearly unchanged.Moreover, SEM image confirms that the GNP coating remained stable after 1000 compression cycles (Figure S5, Supporting Information).Such excellent mechanical stability is vital for the durability of GNP@PUF.
To assess the effect of CMF on the structural integrity of PUF, SEM and X-ray diffraction (XRD) analyses were conducted after it had been soaked in CMF for three times.SEM images and X-ray diffractograms (Figure S6, Supporting Information) confirmed that CMF does not cause any structural damage to the foam.The mechanical property of PUF was also maintained after CMF soaking (Figure 2E).In addition, we determined the change in weight of the pristine foam after it was soaked in CMF for three times.About 0.73 ± 0.32 wt% of mass was lost after soaking in CMF (Table S3, Supporting Information), indicating no mass loss associated with CMF treatment of the foam.
Relatively green organic solvents such as acetone (b.p. 56 °C), THF (b.p. 66 °C), and ethyl acetate (b.p. 77 °C) were examined to determine whether our proposed rapid coating strategy can be implemented with other low boiling point solvents.In contrast to CMF, these solvents demonstrate poor GNP dispersion due to their inferior solvation properties (Figure S7A, Supporting Information).However, with the aid of cellulose acetate (CA), a stable suspension of GNP (up to 1 week) can be achieved in THF and ethyl acetate (Figure S7B, Supporting Information).CA, which dissolves readily in organic solvents, is compatible with graphene, and exfoliation of GNP in the presence of CA (GNP/CA ratio of 5:1) results in a uniform dispersion in these solvents. [60,61]he conductivity of GNP@PUF obtained with THF and ethyl acetate as low boiling point solvents is 0.32 and 0.29 S m −1 , respectively (Table S4, Supporting Information).These values are comparable to the conductivity values for GNP@PUF obtained using CMF; however, they exhibit a slightly higher mass addition (Table S4, Supporting Information), which is attributable to the contribution of CA.
According to the previously mentioned findings, a 3D conductive sponge with superior electrical and mechanical properties was fabricated using a simple and low-cost dipping-and-solventevaporation process.Several strategies for modifying the surface of PUF for use in various cutting-edge technologies have been reported.However, these procedures are time-consuming, multistep, intricate, or involve high-temperature treatment conditions (Table S5, Supporting Information).Contrarily, the low boiling point solvent evaporation-assisted coating proposed in this study successfully produces multifunctional GNP-coated 3D conductive PUF quickly.In addition, the solvent can be effectively recovered and recycled via a simple distillation system (Figure S8A, Supporting Information), making the proposed coating method economical and green for the production of functional PUF.Recovering and recycling the dispersion solvent after the coating is advantageous because it can significantly reduce the cost of the proposed coating process and minimize the production of hazardous liquid waste.More than 97% of the CMF was easily recovered over the course of five runs (Figure S8B, Supporting Information).

Thermal Stability and Flame Retardancy
Thermogravimetric analysis (TGA) was used to investigate the thermal stability of GNP, neat PUF, and GNP@PUF in a N 2 atmosphere (Figure S9A,B and Table S6, Supporting Information).At 700 °C, the GNP exhibited minimal thermal degradation, losing only 1.8% of its initial weight, demonstrating its superior thermal stability and inflammability.The thermal degradation of the foams occurred in two stages: the first stage at T = 295 °C for PUF and T = 317 °C for GNP@PUF was due to the decomposition of urea and urethane hard segments, and the second stage at T = ≈379 °C for PUF and T = 384 for GNP@PUF due to the pyrolysis of the polyether soft segment. [1,2]As demonstrated by the derivative thermogravimetry (DTG) curve (Figure S5B, Supporting Information), GNP modification significantly decreased the thermal decomposition rate of the foam.In addition, the onset temperature (T −5% ) and the amount of char residue that remained at 700 °C were significantly higher for the GNP@PUF than for the pristine PUF, indicating that the deposition of the non-flammable GNP sheets on the PUF surface had improved the foam's thermal stability.
The peak heat release rate (pHRR) was determined using micro-combustion calorimetry (MCC) to evaluate the flame retardancy of the foams.The pHRR is a point in a fire incident where the produced heat is sufficient to spread the flame to nearby flammable materials.Thus, pHRR is regarded as an important fire safety metric.Due to its inherent flammability, PUF burns quickly and has a high pHRR value (598 W g −1 ) (Figure 3A).In contrast, the pHRR of GNP@PUF was 248 W g −1 , 57% less than that of pristine PUF.Thus, the risk of fire hazard was significantly reduced with GNP@PUF compared to neat foam.The foams' flame retardant (FR) properties were further investigated using a cone calorimeter (CC).65][66] Compared to MCC, the CC stimulates large-scale, real-world fire accidents and provides additional vital FR parameters for smoke emission and gas production.As shown in Figure 3B, PUF attained a pHRR of 356 kW m −2 , whereas GNP@PUF demonstrates a pHRR of 99 kW m −2 , representing a 72% decrease.Moreover, the average heat release rate (avHRR), the maximum average rate of heat emission (MARHE), and the total heat release (THR) are reduced by 49%, 54%, and 50%, respectively (Figure 3C,D; Table S7, Supporting Information).These values indicate that the GNP sheets significantly improve the fire safety performance of GNP@PUF.
When PUF is ignited, massive quantities of deadly smoke and gases are released. [67,68]In evaluating the overall FR performance of the foams, smoke and gas production-related data are; therefore, equally vital metrics.The peak rate of smoke release (pRSR), total smoke release (TSR), total smoke production (TSP), and smoke optical density values for GNP@PUF were 0.016 m 2 s −1 , 22 m 2 m −2 , 0.2 m 2 , and 43 m 2 kg −1 , with reductions of 54%, 81%, 80%, and 83%, respectively, when compared to PUF (Figure 3E-G; Table S7, Supporting Information).Such remarkable reduction suggests that the GNP coating effectively reduces the foam's burning rate, and fire risk offers a special smoke suppression effect.Releasing less hazardous and less-dense smoke during a fire incident facilitates evacuation and firefighting.Figure 3H,I depicts the foam's CO 2 P and COP curves, respectively.The peak CO 2 P (pCO 2 P) value for GNP@PUF was 0.07 g s −1 , a 63% reduction compared to PUF, further confirming the improved FR behavior of the foam after GNP coating.However, there was no discernible difference in COP between the two foams; although, the coated foam required slightly more time to reach peak COP release.Overall, it was determined that the fire safety of GNP@PUF is superior to that of the neat PUF.

Application as Piezoresistor
Foams are promising materials for fabricating piezoresistive devices due to their porous structure and excellent mechanical properties, that is, flexibility, stretchability, and compressibility.Figure 4A shows the experimental setup used to assemble and characterize the GNP@PUF piezoresistive sensor.Several key parameters are used to evaluate the performance of piezoresistors, including resistance change (ΔR/R 0 ), applied pressure or strain, and sensitivity (s) or gauge factor (GF).As shown in Figure 4B,C, when a compressive or tensile force is applied to the sensor, the length of the foam decreases by ΔL.In both cases, the strain is denoted as  = ΔL/L 0 = |L 0 -L|/L 0 × 100%, where L 0 is the original length of the foam and L is the compressed length of the foam.Meanwhile, the resistance change is expressed as ΔR/R 0 = |R 0 -R|/R 0 × 100%, where R 0 is the initial resistance of the conductive foam, and R is the resistance of the conductive foam when compressed.The piezoresistor strain sensitivity (GF) is calculated as GF = (ΔR/R 0 )/, and the piezoresistor pressure (P) sensitivity is calculated as s = (ΔR/R 0 )/ΔP.Piezoresistors based on 3D conductive foams under compression mode typically have negative piezoresistive properties (i.e., decreasing resistance with increasing pressure).This is because when compressed, the porous structure squeezes and eventually contacts each other, densifying the conductive GNP sheets on the foam's surface (Figure 5A), resulting in more conductive pathways and a consequent reduction in electric resistance.Figure 5B,C illustrates the compression mode working principle of foam-based piezoresistors. [22]In this model, the foam skeleton encased by a conductive material, such as GNP, is considered a resistor network.At low pressure, the pores in the outermost layers are flattened, resulting in the formation of new conductive networks and a reduction in resistance.When the pressure is increased further, the pores of the middle layers are bent and become an oval shape.At this stage, the amount of deformation increases, but no new conductive connections are formed.As a result, the resistance decreases slightly owing to the shortened distance of conduction pathways.At high pressure, all pores come in contact, resulting in numerous conductive networks and a substantial decrease in resistance.In addition to the increased cell wall contact, the contact of conductive GNP sheets also increases at high pressure.Increasing the pressure beyond this stage results in a saturation of conductive networks and resistance change stabilizes.Due to this phenomenon, the sensitivity of 3D conductive foams varies with the applied pressure or strain.
Figure 6B illustrates the dynamic resistance values of the GNP@PUF sensor (shown in Figure 6A), which was tested by cyclic compressing-releasing at varying strain levels (5%, 10%, 20%, 30%, 40%, 50%, and 60%).The sensor's response was consistent and stable under the same strain because the conductive network of the GNP suffered the same level of damage at the same strain.With increasing strain, the maximum value of the relative resistance of the GNP@PUF decreased due to an increase in contact points and a decreased distance between the GNP layers.This electromechanical property endowed the GNP@PUF to monitor various actions, a crucial feature in practical applications, that is, human−machine interaction, human movement monitoring, and electronic skin.The relative resistance changes versus the strain were plotted to calculate the GF of the piezoresistive sensor.Due to the variation in the porosity of the conductive foam, the sensor's sensing range was divided into two regions (Figure 6C).In region I (5% <  < 50%), the resistance decreased precipitously, and a GF of 2.4 was determined.In region II (50% <  < 70%), the resistance displayed minimal variation, with a GF as low as 0.24.At higher strains, nearly all the porous skeleton of the foam contracted, and the tunnel distance between the conductive layers decreased with compression, resulting in a decrease in tunneling resistance.Thus, the variation in resistance was small compared to region I.Meanwhile, the highest GF was obtained at  = 40% (Figure 6D). Figure 6E illustrates the relationship between relative resistance change and pressure values, highlighting the high sensitivity of the GNP@PUF sensor over a broad pressure range.Specifically, s = 1.52 kPa −1 at pressure < 55 kPa and 0.29 kPa −1 between 55 and112 kPa.This indicates that the sensor's sensitivity was high at low pressures.The GF and s values demonstrate that the GNP@PUF-based sensor is susceptible to strain and pressure.
The pressure sensor's sensing performance at low and high compression frequencies hardly changed (Figure S10, Supporting Information), which is significant to its practical application.The response time and recovery time of the GNP@PUF sensor were 48 and 35 ms when it was compressed and released with 50% strain, respectively (Figure S11A, Supporting Information), which is comparable with reported foam-based pressure sensors (Table S5, Supporting Information).Such ultrafast response and recovery time ensure real-time response to instantaneous pressure.Figure S11B, Supporting Information depicts the pressure sensor's stability.The sensor maintained a stable response after 500 compression cycles at  = 50%, showing good repeatability and stability, demonstrating excellent structural integrity of the 3D conductive GNP network.The pressure sensor could be connected to a 9 V circuit, and the light-emitting dieode (LED) glowed brighter when compressive pressure was applied to the sensor (Figure S12 and Video S1, Supporting Information), indicating the sensor's great potential in real-world applications.According to the preceding discussion, the GNP@PUF sensor has excellent sensitivity, fast response/recovery, and stability.A series of demonstration experiments was conducted to detect various human motions (Figure 7A-E) to demonstrate its viability as a wearable electronic device.The sensor adhered to the joint of an index finger using adhesive tape to detect finger bending.The finger joint caused a compressive deformation of the conductive sponge, which resulted in a variation of the relative resistance signal (Figure 7A; Video S2, Supporting Information).The curve exhibited constant and repeatable response peaks responding to repeated bending, indicating the GNP@PUF-based sensor's stability.The sensor also showed excellent repeatability and stability upon repeated gentle hand presses (Figure 6B; Video S3, Supporting Information), providing real-time finger activity data.The sensor was attached to the insole of a shoe to acquire the sensor response from two distinct pressure points (toe and heel) to detect foot motion.The sensor exhibited a strong and repeatable signal of resistance variation upon successive toe and heel stamping (Figure 7C,D; Videos S4 and S5, Supporting Information).The sensor was then attached to the calf to detect squatting motion.The sensor demonstrated remarkable stability and reproducibility in response to repeated squats (Figure 7E; Video S6, Supporting Information).All these results indicate that the GNP@PUF piezoresistive sensor possesses excellent pressure sensing and human motion detection performance, demonstrating the sensor's vast application potential in human health, sports monitoring, and intelligent robots.
The stretch mode sensing property of the GNP@PUF sensor was also investigated.Figure 7F illustrates the typical relationship between the change in relative resistance and the applied tensile strain (up to 100%).The curve displayed two linear regions: region I (5-30% strain) with a GF of 1.66 and region II (30-100% strain) with a GF of 5.82.An increase in relative resistance with increasing strain was observed, typical of foam-based stretch sensors. [69,70]The conductive GNP@PUF foam was connected to a 9 V battery and an LED to form a simple circuit and exploit the sensor's electronic applications.The LED light's brightness decreased with increasing tensile strain due to an increase in resistance, and this phenomenon was reversible (Figure 7G-J; Video S7, Supporting Information).An SEM image of the stretched foam was obtained to elucidate the sensing mechanism of the sensor under stretching mode.In the sensor's original state ( = 0%), the GNP sheets were self-overlapping and densely packed on the surface of the PUF framework (Figure 5A), forming a conductive network that gave the composite foam a low resistance.With the application of a 50% tensile strain, the pores of the foam elongated without cell wall contacts (see Figure S13, Supporting Information), resulting in long conduction paths; a closer look into the cell wall surface revealed that the GNP sheets became separated and loosely stacked (Figure 5A).Therefore, it can be concluded that the main A circuit constructed with GNP@PUF foam reveals that the LED is bright G,H) before stretching the foam, but its brightness decreases I,J) after stretching.
reason for increasing resistance with increasing tensile strain is a combination of increased length and broken conduction paths (Figure 5D).The more the foam is stretched, the longer the conduction paths and the more paths are damaged, resulting in a remarkable increase in resistance and GF value, as depicted in Figure 7F.

Application as EMI Shielding
As depicted in Figure 8A,B, the EMI properties of the foam were determined using the S parameters (including S 11 , S 22 , S 12 , and S 21 ) characterized by a vector network analyzer.The shielding effectiveness (SE), including total SE (SE total ), Figure 8. A,B) Experimental setup for measuring EMI shielding using a vector network analyzer with a waveguide.C) EMI shielding mechanism of GNP@PUF.D) EMI shielding performance (SE total) of the foams in the frequency range of 8-12.5 GHz and E) thickness effect, F) SE reflection, and G) SE absorption for GNP@PUF.
Coefficients of shielding mechanism: (1) EMI shielding effectiveness (SE): SE total = 10 log Typically, the EMI shielding effect is realized by three mechanisms, that is, the reflection of the incident radiation (SE ref ) from the front of the shield, the absorption of the incident radiation (SE abs ) within the material, and the multiple internal reflections (which is insignificant when SE total is higher than 15 dB). [7]he EMI SE of neat PUF and the GNP@PUF was measured over the frequency range from 8.2 to 12.4 GHz (X-band).Due to the insulating properties of neat PUF, it was nearly transparent to the electromagnetic waves, with a SE total value of ≈0 dB (Figure 8A).In contrast, the SE total of GNP@PUF was significantly enhanced due to the increased conductivity (0.30 S m −1 ) with values of 18 dB for 5 mm thickness and even higher values of 25 and 38 dB for 10 and 15 mm sample thicknesses, respectively (Figure 8B).This value far exceeds the minimum required for commercial applications (20 dB) by a significant margin. [13]To elucidate the EMI shielding mechanism of the GNP@PUF foam, the SE abs and SE ref were calculated from the measured S parameters.As depicted in Figure 8C,D, the SE abs values were considerably greater than the SE ref values, indicating that absorption is the predominant EMI shielding mechanism for the GNP@PUF.For instance, the SE total , SE abs , and SE ref of the 15 mm thick GNP@PUF were ≈38, 37, and 1 dB, respectively.Further, the coefficients of reflection (R), absorption (A), and transmission for the GNP@PUF_15 mm sample were calculated using the Equations (1)-(3) and were determined to be 0.28, 0.71, and 0.00015, respectively.The obtained values manifest that the manufactured GNP@PUF foam samples exhibited strong EMWs absorption within the conductive porous structure.With its strong absorption of EMWs, the GNP@PUF could reduce the secondary EMI pollution caused by microwave reflection, a major drawback of carbon film-based EMI shielding materials in which most EMI was reflected and caused secondary EMI pollution. [9,10]Based on these results, the possible EMI shielding mechanism of the GNP@PUF foams was predicted, as shown in Figure 8C.The highly porous 3D network structure of the GNP@PUF could reduce the impedance mismatch at the airmatrix interfaces, thereby diminishing the back reflection and facilitating the deep penetration of most of the incident EM waves into the foam.The 3D conductive GNP network within the foam could then convert the entered EM energy into thermal energy via conductive dissipation and multiple scattering and reflections. [26]he specific EMI SE (SSE t ) was calculated from the ratio of SE to the thickness (t) and density ( = 0.075 g cm −3 ) of the sample, according to Equation ( 7) [71] to analyze the practical viability of GNP@PUF foam samples for real EMI shielding applications.
At 10-and 15-mm thickness, the SSE t values of GNP@PUF (with 0.30 S m −1 conductivity) were 333 and 338 dB cm 2 g −1 , respectively.The closest comparable system in the literature rGO@PUF, [26] with a 0.25 S m −1 conductivity, afforded an EMI shielding performance of 210-320 dB cm 2 g −1 at a much higher thickness (20-60 mm).Aside from the low EMI shielding performance, the previous work used hydrothermal treatment in the presence of hydrazine and required a longer processing time (>2 h).Other studies demonstrated higher EMI shielding performance (467-5250 dB cm 2 g −1 ) due to their higher conductivity (105-1724 S m −1 ) (Table S5, Supporting Information).However, these studies utilized time-consuming (up to 96 h), complex (i.e., several steps and several axillary chemicals), and costly surface modification processes.Moreover, they discharged a massive amount of nanoparticle-polluted effluent into the environment.In contrast, GNP@PUF can be produced rapidly and with few steps, and the solvent can be recovered and reused through a straightforward distillation process.Effective recovery and recycling of the solvent reduce the loss of GNP particles to the environment and the cost of producing the foam composite.

Conclusion
To deposit GNP onto PUF, an ultrafast, simple, and inexpensive surface coating method was developed based on low boiling point organic solvent evaporation.The resultant sponge (GNP@PUF) exhibited excellent flame retardancy with multiple functionalities; for instance, it could act as a piezoresistive sensor and an EMI shielding material with potential applications in the civil and military sectors.Due to the passive heat barrier of GNP sheets, GNP@PUF demonstrated excellent flame retardancy by lowering the critical fire retardancy metrics, that is, peak heat release rate, total heat release, and total smoke release by 72%, 50%, and 81%, respectively.When used as a piezoresistive sensor, it had a wide pressure (2.4-112 kPa)/strain (5-70%) range and ultrafast response/relaxation time (48 ms/35 ms), wide-stretching strain (5-100%) range.It could be attached to various parts of the human body to detect human motion.Further, GNP@PUF showed a good absorption-dominant EMI shielding performance (≈38 dB), possibly because of both conductive dissipation and multiple reflections and scattering of EM waves within the 3D conductive graphene network.This study provides a simple and efficient method for the surface functionalization of PUF to impart flame retardancy and multiple functions.
Synthesis of PUF: The PUF was prepared using a one-pot free-rise method (Figure 1A). [72]Briefly, the polyol (100 g) and DI water (1 g) were mixed well in a glass beaker by vigorous mechanical stirring for 30 s. Next, MDI (36 g) was added to the mixture and stirred for 5 s.The mixture was immediately transferred into a mold, allowed to rise freely and cured for 3 h at room temperature.
Fabrication of GNP@PUF: The GNP@PUF was fabricated by a simple solvent evaporation method, as illustrated in Figure 1B.First, stable GNP suspension (0.5 mg mL −1 ) was prepared by exfoliation of GNP in CMF with the help of bath ultrasonication for 3 h (Figure S1A, Supporting Information).Then, PUF was immersed into the GNP dispersion and repetitively squeezed for ≈10 s.After this, the GNP@PUF was removed from the dispersion and directly dried (for ≈60 s) in a Rotavap at 80 °C or by blowing hot air.These steps were repeated three times.During the drying process, the evaporation of CMF would lead to the self-assembly of GNP sheets onto the surface of PUF skeletons, forming a thin GNP layer.
Characterization: SEManalysis was conducted using a scanning electron microscope equipped with Energy Dispersive X-ray spectroscopy (Hitachi SU-70/ HORIBA, Energy X-MaxN EDS) operating at 2.0 kV.The pristine PUF was sputter-coated with Au before imaging, whereas the GNP@PUF was analyzed as it is.XRD was obtained using a Bruker AXS D8 focus X-ray diffractometer (Cu-K radiation, step size 0.03°at 4°min −1 scan speed) between 5°and 90°.The densities of the foams were measured by the ratio of mass to volume.The dimensions of cube-shaped foams were measured using a digital caliper.The density was calculated using the relation  (g cm −3 ) = m/V, where m and V are the mass of the foam samples in g and the volume of the samples in cm 3 , respectively.The compression (ISO 844: 2004 standard) test was performed using a universal testing machine (LLOYD LF Plus).A thermogravimetric analyzer (Mettler Toledo, TGA2, Korea) was used to perform the TGA of the foams under an N 2 atmosphere (50 mL min −1 ) and a heating rate of 10 °C min −1 .An FAA Micro Calorimeter (model number 11 311) was used to perform the MCC test according to ASTM D7309.The test was done by heating the samples (≈15 mg) from 100 °C to 700 °C at a heating rate of 1 °C s −1 under a N 2 atmosphere.ISO 5660-1 standard was used for the samples' CC test.Foams (10.2 × 10.2 × 2.5 cm 3 ) were exposed to a heat flux of 35 kW m −2 (Fire Testing Technology, West Sussex, UK).The cuboid GNP@PUF was nipped between two aluminium sheets, and the resistance of GNP@PUF was recorded by a KEYSIGHT 34450A 5 1/2 digital multimeter.The volume electrical conductivity () of the foams was calculated using the following Equation ( 8): [73]  = L R × A (8) L and A are the length and cross-sectional area, measured by a digital caliper, respectively, and R is the resistance.A schematic drawing of the experimental setup used for piezoresistive characterizations is given in Figure 4A.The bottom and top sides of the GNP@PUF were connected to aluminum plates with copper wires using conductive silver paste.The piezoresistive sensing performance (variation of resistance as a function of compressive pressure) of the GNP@PUF was measured using a KEYSIGHT 34450A 5 1/2 digital multimeter, and loadings of compressive pressure/strain were performed using a universal testing machine (LLOYD LF Plus).Data acquisition from the digital multimeter was carried out using BenchVue Software v3.0 (BV0000A).The average resistance values of three samples are reported herein.The sample size for piezoresistive characterization was (15 × 21 × 21 mm 3 ), while a (10 × 10 × 10 mm 3 ) sample size was utilized for human motion monitoring.
The S parameters (S ij ), where i and j are the EM waves receiving and transmitting ports, respectively (see Figure 8A,B), were measured using a vector network analyzer (VNA, E5071C) using the waveguide method in the X-band (8.2-12.4GHz) region.Foams (22.5 × 10.0 mm 2 ) with different thicknesses were cut to fit the WR-90 rectangular waveguide holders well.
Statistical Analysis: The data in figures and tables were presented as mean value (μ) ± standard deviation () computed using the following equations by Excel (Microsoft 365, Version 2303).
Mean Value: Standard Deviation: where x is experimental data, and n is the number of experimental replications performed.

Figure 1 .
Figure 1.A) Schematic representation of the synthesis of PUF.B) Manufacturing of GNP@PUF by coating GNP sheets onto PUF structures and evaporating the solvent.

Figure 2 .
Figure 2. SEM images of Ai-iii) PUF and Bi-vi) GNP@PUF.C) Compression and D) cyclic compression of PUF and GNP@PUF.E) Compressive strength of PUF after CMF soaking.The average value was used for creating the stress versus strain plots (n = 2).

Figure 4 .
Figure 4. A) Experimental setup for the pressure sensor characterization.Schematic showing the circuit diagram of GNP@PUF under B) compressive and C) stretching mode.D) Key performance evaluation parameters for the pressure sensor.

Figure 5 .
Figure 5. A) SEM images of GNP@PUF at three different strains (−60%, 0%, and +50%) and the red-circling indicates the contact points under compression mode and elongated pores under stretching mode.B) Schematic illustration of the change of network structure of the 3D conductive sponge and C) its equivalent circuit diagram to explain the pressure sensor operating principle under compression mode.D) The sensing mechanism of the GNP@PUF sensor under stretching strain-sensing mode.

Figure 6 .
Figure 6.A) The GNP@PUF pressure sensor, B) the sensor response to varying compressive loading, C) the correlation between resistance change and compressive strain change, D) the calculated gauge factor for varying compressive strains, and E) the correlation between resistance change and pressure change.The mean value ± standard deviation was used for plotting the strain-resistance, GF for each strain, and pressure-resistance curves (n = 2).

Figure 7 .
Figure 7. Monitoring of human motion using the GNP@PUF.Resistance change to the activities of A) finger movement, B) gentle hand press, C) toe press, D) heel press, and E) squat.The inset photographs show the corresponding body movement.F) The correlation between resistance change and stretching strain change (values are mean ± standard deviation, n = 2).A circuit constructed with GNP@PUF foam reveals that the LED is bright G,H) before stretching the foam, but its brightness decreases I,J) after stretching.