3.1 Preparation and the chemical structure of PGP composites
Figure 1 presented a green and simple water evaporation-induced self-assembly method to prepare PGP composite. According to our study, the schematic diagram of the morphology of PGP composite and interactions inside it was proposed. Owing to abundant oxygen-containing groups (− OH, −COOH, and epoxy groups) on the surface of GO[41], the highly hydrophilic PVA molecules not only acted as a platform for the homogeneous dispersion of GO and PA[42], also crosslinked by PA molecules through hydrogen bonding (HB) interactions to form a polymer network at where localized the GO nanosheets[43–45], as shown in Fig. 1a. During the drying process, contraction of the polymer network and localization of GO would benefit the formation of a mechanically strong “mortar and brick” structure at where the GO stacked and aligned with the PVA molecules. Such a ternary system constructed a tough and flexible PGP composite film which could afford many times folding and bending without any cracks. Also, the organized stacking of GO nanosheets provided a rapid in-plane thermal transition capacity which accelerated the dehydration and graphitizationof PGP composites to trigger the deformation and conductivity change. Moreover, PA and GO collaboratively enhanced the flame retardancy of PGP composites through free radical trapping and catalytic dehydration and carbonization by PA, and the physical barrier effect of GO [33].
To identify the specific function of each component in the PGP composite, their chemical structure was characterized by FTIR, XRD, and XPS methods, as shown in Fig. 1b-1f. The FTIR spectrum of GO displayed the characteristic peaks at around 3392, 1737, 1620, and 1070 cm− 1, which were assigned to the stretching vibrations of the O − H, C = O, C = C, and C − O groups, respectively[29, 46]. In the case of PVA, the typical peaks observed at 3421, 1070, 920, and 847 cm− 1 were attributed to the stretching vibrations of O − H groups and C − O bonds, the bending vibrations of C − H bonds and the skeletal vibrations of polymer chains, respectively[37, 47]. For the pure PA, four characteristic peaks appeared at 1150, 1020, 1003, and 966 cm− 1, which were ascribed to the stretching vibrations of P = O, P − O−C, P − O and P − O−H bonds, respectively (Fig. 1b)[18]. The above-mentioned characteristic peaks were also observed in the spectra of PGP composites (Fig. 1b), while the peaks assigned to − OH stretching vibration gradually shifted from 3392 to 3433 cm− 1 (Fig. 1c), indicating the formation of interior HB interactions among GO, PVA, and PA after their combination. Interestingly, peaks of the C = C and C = O of GO were suppressed after combining with PVA, probably because of the immobilization effect of the PVA polymer chain to GO through entanglement and HB interactions that reduced the vibration of functional groups on GO (Fig. 1b)[48]. As shown in Fig. 1d, a new peak at 1064 cm− 1 assigned to the P − O−C bond appeared when adding PA[49], and shifted to lower wavenumbers with the increase of PA content, indicating potential covalent associations occurred between PA and PVA or GO as a result of the acidic catalyzed esterification[50]. The intensity of P = O and P-O-H peaks were gradually enhanced, PVA skeletal vibration peaks were gradually broadened, and the peak positions of C-H disappeared with PA content addition, also implying esterification.
The XPS spectra were carried out to confirm the detailed chemical structure and the interactions between PVA, GO, and PA in the PGP composites, as shown in Fig. 1e. The P1G1 sample possessed predominant peaks of C1s (~ 285.0 eV) and O1s (~ 532.0 eV) with the atomic percentages of 64.14 and 35.86%, respectively. While the P1G1P0.8 sample had four peaks of C1s, O1s, P2p (~ 131.0 eV), and P2s (~ 189.0 eV) with the atomic percentages of 36.15, 58.79, and 5.06%, respectively. High-resolution C1s, P2p, and O1s spectra and their summarized parameters were also given in Fig. S1 and Table S1. Compared to the C1s high-resolution spectrum of P1G1 in Fig. S1a, a new peak at the 286.0 eV which was assigned to the C − O−P bond presented in the spectrum of P1G1P0.8 (Fig. S1b), demonstrating again that PA associated chemically with PVA and GO. Additionally, in the C1s spectrum of P1G1P0.8, the fitted peaks assigned to C(= O)O, C = O, and C − O groups shifted to higher binding energy, while to C − C and C = C groups shifted to lower binding energy, indicating the HB interactions among PVA, GO, and PA. Three fitted peaks at 133.8, 134.8, and 134.5 eV which were assigned to P = O, P − O, and P − O−C bonds in the P2p spectrum of P1G1P0.8 (Fig. S1c), and two divided peaks attributed to − O− and = O in the O1s spectra of P1G1 and P1G1P0.8 (Fig. S1d-e), also confirmed again the HB interactions and esterification occurred between of PA, PVA and GO. Figure 2f showed the XRD patterns of PVA, GO, and PGP composites. GO showed a typical diffraction peak at 12.1° which was ascribed to the (0 0 2) crystal plane [46], while the PVA showed an amorphous wide peak around 19.2° which ascribed to the (1 0 \(\stackrel{-}{1}\)) crystal plane[51]. After combining PVA and GO, the characteristic peak of GO was significantly weakened, furtherly disappearing with the increase of PA, suggesting the crystallization of GO was suppressed which was affected by PVA and PA polymer chains as we explained the water evaporation-induced self-assembly process above. At the same time, the diffraction peak of PVA also was weakened, indicating the decreasing of crystallinity degree, which could lead to the enhancement of flexibility. This was consistent with the results of mechanical performance in the section 3.2.
3.2 Morphology and the mechanical strength of PGP composites
Surface morphology regarding the GO dispersion condition was observed through the fluorescence microscopy and SEM methods, as shown in Fig. 2. The PVA film containing bubbles showed relatively homogeneous images under the bright and dark field, and under the green light with a wavelength range of 460 ~ 550 nm (Fig. 2a1-c1). After combining with GO and PA, the composites presented turbid images under bright and dark fields and showed in-plane and well-dispersed GO sheets under green light irradiation (Fig. 2a2-c3 and Fig. 2a3-c3). Also, in-plane aligned GO sheets were mostly displayed in the observation field, confirming the parallel stacking of GO sheets in the PVA matrix as demonstrated above. However, different views on the surface PG and PGP composites were observed under SEM. As shown in Fig. 2d-2i, the flat and smooth surface of the PVA film became coarse and irregular and turned up some bulges with the increase of PA, indicating the formation of colloidal particles containing PVA, GO, and PA. The 3D laser microscope furtherly confirmed the increase of surface roughness (Sa, increase from 0.12 to 0.44) with the PA content, as shown in Fig. S2, which corresponded to the SEM result. The EDS mapping and spectra of the PG and PGP composites explained the element composition, content, and distributions, which were shown in Fig. 2j-2l, Fig. S3, and Fig. S4. The C, O, and P elements were evenly distributed on the sample surface (Fig. S3 and S4), suggesting that GO and PA had excellent dispersibility in PGP composites. The element ratio of C:O:P according to the EDS result was approximately agreed with the initial dosing of PVA, GO, and PA, showing no substrate loss during the sample preparation.
Good dispersion of GO and the chemical crosslinking of PA in the PGP composites significantly enhanced the mechanical properties, as illustrated by the tensile tests shown in Fig. 2m and 2n. Summarized results were also listed in Table S2. The introduction of GO dramatically increased the tensile strength σb from 60 to 121 MPa and the modulus E from 1.2 to 2.2 GPa, while decreasing the strain εb from 217 to 8% and work of fracture Wf from 115 to 10 MJ·m− 3, greatly raised the rigidity and brittleness owing to the HB interactions with PVA[42] but weakened the flexibility because of the original high modulus of GO[52], and that severely limited the application fields. After the following introduction of PA, the σb and E decreased with the PA content but the εb and Wf increased, which obtained better flexibility than pure PVA film in the case of P1G1P0.8 composite film (Fig. 2n), while the strength of this film could only catch the pure PVA one. The chemical crosslinking among PVA polymer chains and GO nanosheets by PA was responsible for such mechanical strength changes[50] that could easily manipulate the entire mechanical properties of PGP composites to satisfy practical application requirements. Additionally, as presented in Fig. 2o and Table S3, the comprehensive mechanical properties of the PGP composite films in the work were superior to those of most PVA composite films regarding their toughness and modulus, yet the final solid content of GO and PA was up to 64.3% which might require high cost in the practical applications that further optimizations should be made.
3.3 Thermal stability and flame-retardant performance of PGP composites
The thermal stability and flame-retardant performance of PGP composites are the foundation of being a fire-warning smart device demanding high integrity and good durability when exposed to high temperatures[35]. To assess the thermal stability, thermogravimetric (TG) tests were carried out and the results were shown in Fig. 3 and Table S4. PVA started to decompose at 230 oC (Tdep) and reached the maximum decomposition rate (Rdpm) of 1.13%·oC−1 at 270 oC (Tmax) which was mainly a dehydration process, followed by the weight loss of 50% at 281 oC (T50%) (Fig. 3a, Table S4). The second decomposition state was observed for PVA at 438 oC probably due to the cyclization of the PVA molecular skeleton (Fig. 3b) [53, 54]. The char residue at 800 oC for PVA was only 8 wt.%, showing its high flammability and thermal degradability. It is reported that the thermal degradation of PVA is promoted by the •OH radicals from the initial decomposition stage, which catalyze further chain breakage and generate flammable small molecules[55], and that would be responsible for the low char content of PVA. After combining with GO, the Tdep, Tmax, and T50% raised to 245, 333, and 330 oC, respectively, and the Rdpm decreased to 0.63 (Table S4), demonstrating reinforced thermal stability. However, the char residue remained at a low level of 9 wt.% at 800 oC, explaing that the GO proceeded with oxidation decomposition at a higher temperature and was unable to increase the char content. After introducing PA, all PGP composites showed two distinguish peaks, as shown in Fig. 3b. The Tdep and Tmax were lower than pure PVA and decreased with the PA content increasing (Table S4), which could be attributed to the lower dehydration temperature of PA[56], or the acidic PA accelerated the PVA dehydration[57]. While the T50% elevated to over 430 oC and even 525 oC in the case of P1G1P0.8, and the Rdpm furtherly decreased from 1.13 for pure PVA to 0.52 for P1G1P0.8, as shown in Fig. 3b and Table S4, showing better thermal stability. Also, the char residue of PGP composites ranged from 34 ~ 38% which was 4 more times than the PVA and PVA/GO one. The reinforcements on thermal stability and higher char level originated from the more effective dehydration carbonization and •OH radical trapping by phosphate compounds generated from PA[47, 58].
Subsequently, we evaluated the combustion behavior of PGP composites by microscale combustion calorimetry (MCC) to assess the flame-retardant performance, as shown in Fig. 3c and 3d. Key parameters such as the peak heat release rate (pHRR), the heat release capacity (HRC), and the total heat release (THR) were used to assess the flame-retardancy of PGP composites, which were summarized in Table S5. As shown in Fig. 3c, at around 275 oC, pure PVA showed a sharp peak which pHRR was calculated as 529.9 W·g− 1 and the HRC was 508.3 J·(g·K)−1, demonstrating that PVA burned intensely, which was agreed with the TG results. The introduction of GO significantly reduced the pHRR and HRC to 159.6 W·g− 1 and 141.8 J·(g·K)−1, respectively, which curve displayed the moderate-separated peaks at where the Tp raised to 364 oC, suggesting an enhanced flame retardancy due to the physical barrier action of GO nanosheets that preventing the heat transfer[59]. When the PA was incorporated, the pHRR and HRC of PGP composites decreased ulteriorly with much higher Tp (Table S5). A higher Tp and lower pHRR and HRC value indicated less risk of fire, confirming again the excellent reinforcement of flame retardancy by PA. Figure 3d furtherly showed the THR curves of each sample, what we found out was only the introduction of GO was unable to reduce the THR value instead generating more heat during the thermal degradation process of GO, even though its HRR prominently decreased probably owing to the physical barrier effect of GO. On the contrary, the PGP composites showed much lower THR values from 7.7 to 12.6 kJ·g− 1 compared to the PVA and PVA/GO one with a high THR value of 23.0 and 23.9 kJ·g− 1 (Table S5), verifying that PGP composites had excellent flame-retardant performance.
Vertical flame testing was also carried out to visualize the combustion behaviors of PGP composites, as shown in Fig. 3g-3h and Fig. S5. A 15 × 15 mm2 sample was directly exposed to fire and burned for 100s. After ignition, the PVA film burned quickly and completely burned down within 10 s. The introduction of GO did not enhance the flame retardancy which lost most of its shape and weight after 30s of burning (Fig. S5a). In contrast, the additional introduction of PA greatly promoted flame retardancy which was able to stand in fire for over 1 min with minor shape and weight loss (Fig. 3g and Fig. S5b-d). The bulk changes and shrinkages were observed in all PGP composites when the combustion time elongated. Also, with the increase of PA content, the shrinkage became more and more distinct as illustrated in Fig. 3h. Specifically, all PGP composites maintained integral char structure after exposing to flame attack for 60 min, while the planar size of their char decreased dramatically after 20 min. Moreover, the thickness of the char remarkably increases after combustion, as shown in Fig. 3e, which was responsible to the conductivity changes. Considering both the thickness and planar size of each PGP composite, the bulk volume was slightly larger than the unburned one, indicating gas released during combustion. The introduction of GO and PA increased the limited oxygen index (LOI) from 18% for pure PVA to almost 36% for P1G1P0.8, and the UL-94 level reached V-0 (Fig. 3f), explaining again an excellent flame retardancy of PGP composites.
3.4 Char layer characterizations and the mechanism for longtime fire resistance
To realize the longtime fire resistance, the physicochemical properties of the char layer produced at different temperature was important that profoundly affected the structural integrity and durable response during utilization. Here we carried out the measurements on the morphology, composition, and chemical structure features of the char residues from PGP composites to understand their fire resistance and even their combustion mechanism. The cross-section and surface morphology of the char layer of PG and PGP composites were acquired by SEM and presented in Fig. 4. Pure PVA was burned to ashes after combustion. Figure 4a1 showed a flocculent cross-sectional structure for P1G1 composites, while in the case of PGP composites, the cross-sectional morphology was mainly a porous structure, as shown in Fig. 4b1-e1. These internal pores were produced during combustion to trap the flammable substances generated from thermal degradation and isolate the oxygen to impede the combustion. Furthermore, as the content of PA increased, the pore walls became thicker because of a higher content of carbonization catalyzed by more phosphate compounds, which agreed with the result from Fig. 3g and 3h. A thick char layer with a better barrier effect against inflamers and oxygen and the radical scavenging effect of phosphate compounds was essential to construct a longtime fire resistance and was capable of continuous fire warning. The surface morphology of the char layers presented in Fig. 4a2−e2 showed many continuous and uncracked bulges owing to the gas release during combustion, also the bulges became denser with the PA content, confirming again the formation of the protective char layers. EDS was simultaneously performed to analyze the element composition in the char residues, as shown in Fig. S6. An increasing phosphorus content and atom ratio of oxygen and carbon caused by the introduction of PA indicated the increase of noninflammable compounds in the char layers which was beneficial to intercept the combustion process.
Spectral analysis was furtherly employed to identify the chemical structure of the char layer for a better understanding of the excellent fire resistance of PGP composites. The FTIR spectra of the PG composite showed characteristic peaks at 2925, 2850, 1441, and 1076 cm− 1 which were attributed to the stretching vibration of − CH2 − and CH3−, the deformation vibration of − CH2 − or CH3−, and the stretching vibration of C − O−C, respectively (Fig. 4f) [60]. While in the case of PGP composites, the above-mentioned peaks were weakened or even vanished, instead of generating new peaks at 1172, 1013, and 906 cm− 1 which were attributed to the stretching vibrations of P = O, P − O−C, and P − O−P, respectively (Fig. 4f) [61]. The phosphorus-containing compounds facilitated the dehydration and carbonization to form a high-quality char layer that improved the flame-retardant performance. Notably, the characteristic peak at 1561 cm− 1 attributed to the vibration of the graphene framework only displayed in the FTIR spectra of the char of PGP composites, meaning that the graphene structure was maintained after combustion.[30, 46] Raman spectra of PG and PGP composites also proved the unsaturated carbon structure through the comparison of the graphitization degree for their char residues, as shown in Fig. 4g. The D-band at ~ 1368 cm− 1 represents the vibration of disordered graphite and the G-band at ~ 1608 cm− 1 is related to the graphitic structure, and the ratio of the peak integral intensity between D-band and G-band (ID/IG) is used to characterize the graphitization degree.[62] A low ID/IG ratio value implied a high degree of graphitization, meaning that the compact char layer is formed during combustion. For PG and PGP composites, the values of ID/IG significantly decreased from 3.04 to 2.08 with the increase of PA content, indicating an increasing graphitization degree that probably interprets the promoted effect to convert amorphous char to graphitized char[63].
Such conversion was corroborated by the XRD patterns as shown in Fig. 4h. Without PA, the pattern only presented a board diffraction peak near 14.1° which was assigned to the char generated from thermal-oxidized and disintegrated GO compared to the original one in Fig. 1e [30, 46]. While with the addition of PA, this peak shifted to a lower 2θ value, indicating the increase of interlayer spacing in char residues, ascribed to the interlayer space being occupied by more thermal degradation products that induced a thicker char layer after combustion[64]. Another peak at around 25o was found with the addition of PA which was assigned to the (002) crystal plane of graphite [49, 52], implying the reduction of partial GO occurred with the catalyst of phosphate compounds during combustion instead of thermal degradation and maintained the structure integrity[48]. Also, the peak showed increased intensity and slightly shifted to higher 2θ with PA content, suggesting a higher content of graphite that stacked closer.
The XPS spectra in Fig. 4i and Fig. S7 furtherly revealed the chemical composition of the char layer. A higher atom ratio between oxygen and carbon (O/C) of the char of P1G1P0.8 (1.16) was observed compared to that of P1G1 (0.39), implying that more oxygen was immobilized in the char instead of acting as the free radicals to induce more chain breakages, which proved the free radical trapping effect by phosphate compounds that was responsible to the increase of the char layer[65–67]. Furthermore, the high-resolution XPS spectra of C1s, P2p, and O1s identified the bonding situation of oxygen atoms in the char, as shown in Fig. S7. The high-resolution XPS spectra revealed that the oxygen atoms were mainly immobilized by forming C = O, C − O, P − O, and P = O bonds according to the deconvolution results[18, 68–70]. The ratio between oxygen double bond (= O) and single bond (− O−) became lower by comparing the spectra of P1G1 and P1G1P0.8 (Fig. S7d and S7e), indicating a more inert chemical activity against thermal-oxidized degradation[71] that realize a longtime fire resistance. The new P − C bond found in the XPS spectra of the P1G1P0.8 char which was not detected in unburned (Fig. S1) also suggested the phosphorus atom participates in the formation of the char layers.
As we mentioned above, gas was released during combustion and formed a rough surface full of bulges (Fig. 4a2-e2). Hence, to explore the gaseous phase composition during combustion, the real-time TG-FTIR measurement was performed to detect the pyrolysis products under heat and air. Figure 4j and 4k displayed the continuous FTIR spectra with the dependence of wavenumber and temperature of P1G1P0.8 composites and the FTIR spectra at specific temperatures. Pre-drying treatment at 60 oC for the P1G1P0.8 sample was carried out before the measurement, and when the temperature was lower than 200 oC, almost no volatile substances were detected. Elevating the temperature would observe the signals for some degraded substances, as the spectra presented the peaks assigned to C = O, C = C, C − H, and phosphorus-containing pyrolysis fragments at 3500 ~ 3800 (water), 2350 (CO2), 1740 (C = O), 1640 (C = C),1440 (C − H), and 1210 cm− 1 (phosphorus-based compounds), respectively[72–74]. When the temperature exceeded 450 oC, new characteristic peaks attributed to the vibration of alkane groups and C − O appeared at 3100 ~ 2900 cm− 1 and 2100 ~ 2200 cm− 1, indicating that the PVA skeleton or GO hexatomic rings began to decompose[37, 75, 76]. Simultaneously, characteristic peaks of CO2 gradually increased, also suggesting a more vigorous thermal oxidation was proceeding in the entire PGP composite. In addition, the peaks of the phosphorus-containing compounds weakened and vanished after 550 oC, probably because the phosphorus was immobilized in the char layer by participating in the carbonization reactions. The inspection and analysis of the gaseous phase helped to understand the thermal degradation and oxidation process, combined with the above experimental results of the original sample and its char layer, we could figure out the reason for the longtime fire retardancy and deduce a proper mechanism for the combustion. During combustion, PA dehydrated PVA polymer chains at a relatively low temperature and then decomposed to phosphate compounds and captured •OH radicals that were generated from the thermal degradation of PVA and prevented further chain breakages. These phosphate compounds also participated in the formation of char layers which enhanced its structural integrity and thermal stability. The GO nanosheets increased the carbon content of the matrix which increased the thickness and density of the char layers and acted as a physical barrier to cut down the transition of heat, oxygen, and inflamers. The resulting thick, dense, and phosphorus char layers were the key to realizing a longtime fire resistance, also its high carbon composition containing in-plane oriented graphite nanosheets was able to continuously build up the circuit for electric warning.
3.5 Fire warning performance of PGP composite film
As the thermal-induced conductivity and longtime fire resistance of PGP composites were figured out, a fire warning system was constructed to verify the practical application. A simple circuit consisting of a PGP composites strip (2 × 0.5 cm2), LED light, power source, and wires, was built up as illustrated in Fig. 5a. The original PGP composites were insulators that could not trigger the LED light, while after heating and even burning, the conductivity of the sample increased due to the shape contraction and the graphitization of GO and lit up the LED light. As shown in Fig. 5b, the P1G1 sample could not trigger the LED light before it was burning down (< 10 s), while in the case of PGP composites, the LEG light was lit up after a short time of burning within 10 s (Fig. 5c). The duration time under fire of PGP composites increased with the PA content and excessed 2400 s in the case of P1G1P0.6 and P1G1P0.8, which also possessed a rapid response time of 3 s and 2 s, respectively. Owing to the shape contraction and the graphitization of GO sheets during heating as mentioned above, the resistivity of the PGP composite films significantly decreased, as shown in Fig. 5d and Table S6. Remarkably, the introduction of PA greatly lowered the resistivity of the composites from over 109 to less than 106 Ω·m, probably owing to the conjugation effect between PA molecules and GO sheets. During the burning process, the dehydration proceeded very fast and induced shape contraction, leading to a more compact structure that reduced the resistivity exponentially. After burning for 10 s, all PGP composite films possessed a very low resistivity lower than 103 Ω·m (Table S6), which was quite enough to light up the LED light with a 3 V power source. The introduction of PA not only promoted the fire resistance of char layers but also accelerated the dehydration process to enhance the responsiveness and enhanced conductivity by forming a dense and graphite-containing char structure, which agreed with the chemical structure analysis above.
Apart from the fast response under fire and longtime fire resistance, responsiveness under relatively low temperatures was also crucial for fire prevention. Generally, the ignition temperature of many polymeric materials ranges from 300 to 500 oC. Hence, if the fire warning works below 400 oC, that will effectively find out and prevent the occurrence of fire hazards. As shown in Fig. 6 and Fig. S8, a hot plate was employed to exert the different temperatures (100, 150, 200, 250, 300, and 350 oC) on the PGP composite films. The P1G1P0.8 film with the fastest response was able to light up the LED light after 20 s at 100 oC, as shown in Fig. 6a, however, the light intensity was so weak to observe. When the temperature was at 100 oC, all PGP composite films with low PA content (< 23 wt.%) failed to light up the LED (Fig. 6b & Table S7). While at a temperature larger than 150 oC, all PGP composite films could trigger the response and light up the LED light which light intensity increased with higher heating temperature (Fig. 6a and Fig. S8), and the response time decreased sharply as the PA content and temperature increased, as shown in Fig. 6b. The P1G1P0.8 sample showed the shortest response time within 10 s when the temperature between 150 ~ 300 oC and responded within 4 s at 350 oC which was before the burning point of most of the polymeric materials. This result agreed with the TG result that the P1G1P0.8 sample began to decompose at around 160 oC, which induced the shape contraction and the graphitization of GO that raised the conductivity. Moreover, with the PA catalyzed, the dehydration and graphitization of PGP at elevated temperature proceeded very fast, as shown in Fig. 6c, in which resistivity dropped sharply in a short time when the temperature was higher than 300 oC. But the equilibrium resistivity after 100 s at 350 oC was higher than that under burning (Fig. 5), indicating a lower extent of shape contraction or carbonization. Also, the constant resistivity below 350 oC reflected the limited shape contraction and thermal reduction which was affected by the dehydration extent of PGP and oxygen-containing groups breakages of GO[77], as illustrated in Fig. 6d. The weight of P1G1P0.8 film reached constant after heating, and the equilibrium weight was related to the heating temperature which was lower with a higher heating temperature. After heating at 300 oC, the P1G1P0.8 sample lost 43% weight and reached equilibrium, verifying again that dehydration of PGP and destruction of oxygen-containing groups in GO occurred before the burning point. While such a dehydration process was quite enough to induce obvious shape contraction and conductivity changes for turning on the circuit. Another aspect that decided the fast responsiveness was the thermal diffusivity of the composite film, which affected the heat accumulation and the rate of thermal-induced dehydration. Interestingly, the pure PVA film possessed a higher vertical thermal diffusivity (2.32 ~ 2.46 m2·s− 1) than the P1G1P0.8 sample (~ 1.43 m2·s− 1), as shown in Fig. 6e, demonstrating the heat barrier effect of GO nanosheets that benefited to the heat accumulation and faster dehydration process.
To find out the heat accumulation conditions regarding the lower vertical thermal diffusivity of the P1G1P0.8 sample, regional heating experiments were performed to verify the thermal sensitivity and accuracy. When utilized as an electronic device, the fire warning requires the sensor with a high sensitivity to deal with the local slight heating or flame induced by transient elevating room temperature, heat waves, and other hidden troubles. As a result, it was crucial to examine the sensitivity and accuracy under local heating conditions. As shown in Fig. S9, when the size of P1G1P0.8 film increased from 0.5 × 2 cm2 to 3 × 6 cm2 or 4 × 8 cm2, it still displayed a fast response within a few seconds under burning, and the light intensity strengthened with the burning time. But the light intensity weakened by reason of a decreased conductivity as the P1G1P0.8 enlarged, indicating the size dependence. Furtherly, we used an infrared thermometer to visualize the heat accumulation situation on a 0.8 × 4 cm2 P1G1P0.8 film under continuous air heating with a nozzle diameter of 0.5 cm (Fig. 6f and Fig. S10). When the heating temperature was 150 oC, the central temperature of the film reached 127.2 oC within 10 s, while the temperature at the heating edge was 82.1 oC. When the heating temperature was raised to 200 oC, the central and edge temperature was approximately 180.9 and 151.6 oC, respectively. The fast heating of P1G1P0.8 film and the temperature difference of around 30 ~ 40 oC between the heating center and edge showed a good regional thermal accumulation capacity and favorable temperature discrimination, implying a good and accurate sensitivity. The fast and accurate response capacity of P1G1P0.8 composites before the burning point was able to call attention earlier to deal with the potential fire hazard to better protect body health and properties. For example, the PGP composite film could be applied in the wallpaper-type fire alarm system by replacing the press-type and installed in the building (Fig. 6g). When abnormal heating or even a fire happened, the PGP wallpaper fire alarm system could quickly trigger the alarm automatically, which gave enough time for people and properties escape before the spreading of fire.
Table 1
Comparison of PGP composites fire alarm system with other previously reported fire alarm systems.
Component | Type of composites | Preparation method | Sustainability | tr a, s | td b, s | Tr c | [Ref] |
GO, MPMS d, L-ascorbic acid | Self-supporting paper | Evaporation induced self-assembly | No | 2 | > 180 | 175s at 150℃ | [30] |
GO, CNTs e, PVH f, boron nitride, phytic acid | Coating on polyurethane foam | Rod, spray, and brush-coating | No | 8 | > 120 | N/Ag | [78] |
GO, adenosine triphosphate | Self-supporting paper | Evaporation induced self-assembly | Yes | 1.4 | 1300 | NA | [79] |
GO, chitosan, CNTs, HPTCP h | Coating on polyurethane sponge | Layer-by-layer assembly | No | 1 | 2640 | 70s at 200℃ | [80] |
GO, aluminum oxide hydroxide | Coating on melamine foam | Layer-by-layer assembly | No | 2.5 | 256 | NA | [81] |
Graphene, nitrocellulose | Self-supporting film | Evaporation induced self-assembly | No | 5.8 | NA | 44s at 200℃ | [82] |
GO; boric acid | Self-supporting paper | Evaporation induced self-assembly | No | 0.8 | 205.4 | NA | [52] |
CNT, PVA, carbon black | Coating on cotton fabric | Dip-coating | No | 4.0 | > 30 | 8s at 350℃ | [83] |
GO, chitosan | Coating on melamine foam | Layer-by-layer assembly | Yes | 3 | 1280 | NA | [84] |
GO, phosphoric acid | Self-supporting paper | Evaporation induced self-assembly | No | 0.5 | 433.0 | NA | [49] |
GO, HCPA i | Self-supporting paper | Evaporation induced self-assembly | No | 1 | > 600 | 154s at 150℃ | [31] |
GO wide ribbons, silica | Coating on melamine sponge | Dip-coating | Yes | 3 | > 120 | 39s at 300℃ | [85] |
GO, tannic acid, P-CNFs j | Self-supporting paper | Evaporation induced self-assembly | No | 1 | > 600 | 154s at 150℃ | [61] |
GO, ammonium polyphosphate, TFTS k | Coating on polyurethane foam | Dip-coating | No | 2s | > 10 | 115s at 200℃ | [86] |
GO, PVA, PA | Self-supporting film | Evaporation induced self-assembly | Yes | < 2 | > 2400 | 10s at 150℃ | This work |
Notes:
a tr: Response time to flame; b td: Waring duration time; c Tr: Response temperature.
d MPMS: 3-Methacryloxypropyltrimethoxysilane.
e CNTs: Carbon nanotubes.
f PVH: Poly(sodium vinyl sulfonate-co-hydroxyethyl acrylate).
g N/A: Not mentioned.
h HPTCP: Phenoxycyclophosphazene.
i HCPA: modified hexachlorophosphazene.
j P-CNFs: phosphorylated cellulose nanofibrils.
k TFTS: Tetrahy-droperfluorodecyltrimethoxy silane.
3.6 PGP composite coating for fire protection and warning
The PGP composite films developed in this work showed a good fire warning capacity which was better than other reported works, regarding the compositions, preparation methods, sustainability, fire response, and waring duration time, as listed in Table 1. The preparation method of PGP composites was environmentally friendly and possessed ultrasensitive fire response, longtime fire resistance, and ultrafast fire-warning response, showing a great prospect in many applications required high fire safety. Profiting from the water evaporation-induced self-assembly method, the PGP composites could be presented in many application forms, not only the films we analyzed above, also the coating was feasible to build up a flame-retardant fire warning system. As presented in Fig. 7, the PGP composite (P1G1P0.8) was employed as the coating of wood and cotton fabric. In the Fig. 7b, the pure wood could be ignited within 36 s and burned down within 120 s. When the P1G1P0.8 coating was applied and connected to the electronic circuit, the coated wood showed good flame retardancy which was able to endure under fire for more than 120 s. At the same time, the carbonized PGP layer switched on the circuit and triggered the fire warning at 3 s. In addition, a flame-retardant cotton fabric coated with PGP composites achieved good flame retardancy, thermal sensitivity and longtime fire alarm capacity. The vertical combustion and MCC results presented in Fig. S11 and Table S8 showed only the cotton fabric coated with P1G1P0.8 could form an intact char layer, and its pHRR, HRC, and THR values decreased 74%, 75%, and 64%, respectively, suggesting the greatly enhanced flame retardancy, which made it display a fast response time within 3 s and persistent warning duration over 900 s (Fig. 7c). It also presented low-temperature warning performance which response time was 2400, 90, 20, 10, and 8 s at 150, 200, 250, 300, and 350 oC, respectively, as shown in Fig. S12. The P1G1P0.8 coated cotton fabric also possessed longtime fire resistance and fire warning ability, while in the case of P1G1 composite, it served no functions (Fig. S11a). In the work, the PGP composites were also capable as a green and efficient flame-retardant with fire warning capacity, showing the versatile applications for practical fire protection.